INFORMATION TO USERS
This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted.
The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction.
1.The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity.
2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame.
3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete.
4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced.
5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received.
University Microfilms International 300 North Zeeb Road Ann Arbor, Michigan 48106 USA St. John's Road. Tyler's Green High Wycombe, Bucks, England HP10 8HR 77-17,088 FINNEY, Stanley Charles, 1947- GRAPTOLITES OF THE MIDDLE ORDOVICIAN ATHENS SHALE, ALABAMA. (VOLUMES I AND II) The Ohio State University, Ph.D., 1977 Paleozoology
Xerox University Microfilms, Ann Arbor. Michigan48 ioe GRAPTOLITES OF THE MIDDLE ORDOVICIAN
ATHENS SHALE, ALABAMA
VOLUME I
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
By
Stanley C. Finney, B.S., M.S. *****
The Ohio State University
1977
Reading Committee: Approved by
Professor S. M. Bergstr8m, Chairman Professor W. C. Sweet Professor J. W. Collinson fif^ I r 7 Adviser J Department of geology and Mineralogy ACKNOWLEDGMENTS
Professor Stig M. BergstrBm and Dr. Valdar Jaanusson suggested and supervised this project. Professor BergstrBm has been an enthusiastic adviser, and without his interest, guidance, time, and support this project could not have been completed. The project was initiated during Dr. Valdar Jaanusson's tenure as a
Distinguished Visiting Professor at The Ohio State University, and the writer was able to spend the Winter of 1974-75 studying with Dr. Jaanusson at the Swedish Museum of Natural History,
Stockholm. Many of the paleobiological ideas presented in this study resulted from stimulating discussions with Dr. Jaanusson.
Thanks are also due to Professors Walter C. Sweet and James C.
Collinson, who read the manuscript and offered helpful advice.
The first season of field work was supported by the Geological
Survey of Alabama (Alabama State Contract No. GSA-73-13). Mr. Charles
Copeland and Dr. James Drahovzal of the Alabama Geological Survey aided the writer in the field and directed him to the Pratt's
Syncline section. I thank the Martin Marietta Cement Company for permission to work in the shale quarry at Calera.
The writer's tenure in Stockholm was supported by a Thord
Gray Memorial Fund Fellowship awarded by the American-Scandinavian
Foundation. I especially thank Dr. Tor 0rvig for the hospitality shown me by the Swedish Museum of Natural History, Maria Tatarczuch
ii and Uno Samuelson for their help in illustrating many of my specimens, and Dr. Hans Tralau for permission to use his hydrofluoric acid laboratory.
I thank the following persons at geologic-paleontologic museums for allowing me access to type collections and working space and for the loan of type specimens: Dr. Norman Newell,
The American Museum of Natural History; Dr. Donald W. Fisher, The
New York State Museum; Mr. Frederick Collier, U. S. National
Museum; Mrs. J. S. Lawless, Yale-Peabody Museum; Dr. Stephen
J. Gould, Harvard University; Professor Gerhard Regnell, Museum of the Palaeontological Institute, Lund University; Professor
Richard Reyment, Museum of the Palaeontological Institute,
University of Uppsala; Dr. David Bruton, Palaeontological Museum of Oslo; Dr. Barrie Rickards, Sedgifick Museum; and Dr. Tom Bolton,
Geological Survey of Canada.
I also thank Mr. Ragnar Nilsson, Lund University, for granting me access to his collections and unpublished data on the Koïngen boring core. Linda Massie andDuncan Foley assisted in the final assembly of the manuscript. Karen Tyler drafted several of the text-figures, and Robert Wilkinson photographed all of the text-figures.
Grants from the friends of Orton Hall Fund paid for much of the expenses for field work and drafting supplies.
Ill VITA
October 14, 1947 ...... Bom, Chula Vista, California
1969 ...... B. S., University of California, Riverside, California
1 9 7 1 ...... M. S., University of California, Riverside, California
1969-1971 ...... Teaching Assistant, Department of Geology, University of California, Riverside, California
1971-1976 ...... Teaching Associate, Research Associate, Department of Geology and Mineralogy, The Ohio State University, Columbus, Ohio
1974-1975 ...... Fellow of the American- Scandinavian Foundation, Swedish Museum of Natural History, Stockholm, Sweden
PUBLICATIONS
Matti, J. C., M. A. Murphy, and S. C. Finney. 1974. Summary of Silurian and Lower Devonian Basin and Basin-Slope Limestones, Copenhagen Canyon, Nevada: Geology, v. 2, no. 12, p. 575-577.
Matti, J. C., M. A. Murphy, and S. C. Finney. 1975. Silurian and Lower Devonian Basin and Basin-Slope Limestones, Copenhagen Canyon, Nevada: Geol. Soc. Amer., Special Paper159, 48 pp.
Finney, S. C. 1975. Morphology of the Graptolite Families Nemagraptidae and Dicranograptidae (abs.): Geol. Soc. Amer., Abs. with Progr., v. 7, no. 6 , p. 755.
IV Finney, S. C. 1976. Graptolite biostratigraphy of the Middle Ordovician Athens Shale, Alabama: The boundary between the Zones of Glyptograptus teretiusculus and Nemagraptus gracilis (abs.): Geol. Soc. Amer., Abs. with Progr., v. 8 , no. 2, p. 171.
Finney, S. C. 1976. Unusual Structures in Graptoloids from the Athens Shale, Alabama, and their bearing on the mode of life of Graptoloids (abs.): Geol. Soc. Amer., Abs. with Progr., V. 8, no. 4, p. 476-477. t a b l e o f c o n t e n t s
VOLUME I
Page
ACKNOWLEDGEMENTS ...... ii
VITA ...... iv
PUBLICATIONS ...... iv
LIST OF TABLES ...... viii
LIST OF TEXT-FIGURES ...... ix
INTRODUCTION...... 1
Methods of Study ...... 10 Previous Biostratigraphic Studies ...... 13
PHYSICAL STRATIGRAPHY ...... 24
Introduction ...... 24 Lithology of Measured Sections ...... 26 Facies Relations and Environment of Deposition...... 42
BIOSTRATIGRAPHY...... 44
Graptolite Biostratigraphy in Alabama ...... 44 Concepts of the Zones of IJ. gracilis and G. sp. cf. teretiuscuius ...... 56 Correlation...... 62 Summary and Conclusions ...... 6 8
SYSTEMATIC PALEONTOLOGY ...... 70
Classification ...... 70
Class Graptolithina Bonn, 1846 ...... 78
Order Graptoloidea Lapworth, 1875 ...... 78
Suborder Didymograptina Lapworth, 1880 ...... 78
VI Page Family Dichograptidae Lapworth, 1873 ...... 78
Genus Pterograptus Holm, 1881 ...... 78 Genus Didymograp tus M'Coy, 1 8 5 1 ...... 82 Genus Azygograptus Nicholson, 1875 ...... 84
Family Abrograptidae Mu, 1958 ...... 87
Genus Reteograptus Hall, 1859 ...... 89
Family Nemagraptidae Lapworth, 1873 ...... 102
Genus Nemagrap tus Emmons, 1855 ...... 105 Genus Amphigraptus Lapworth, 1873 ...... 156 Genus indet. Nemagraptid sp. A ...... 165
Family Dicranograptidae Lapworth, 1873 ...... 176
Genus Dicellograptus Hopkinson, 187 1 ...... 176 Genus Leptograptus Lapworth, 1873 ...... 298 Genus Pierancgraptus Hall, 1865 ...... 324
Vll LIST OF TABLES
Table Page
1. Sicular measurements on specimens from locality PS-109.2...... 125
2. Measurements on figured thecae of N. gracilis...... 130
3. Distance in mm between successive thecal apertures in specimens of N. gracilis ...... 142
4. Diagnostic features used by previous workers for taxa of Nemagrap tu s ...... 149
5. Classification of Dicellograptus species in terms of the vagus and elegans groups ...... 180
6 - Correlation of growth stages of Dicellograptus ......
7. Variation in stipe width in JD. gurleyi gurleyi in relation to state of preservation...... 235
vxxi LIST OF TEXT-FIGURES
Text-figure Page
1. Map of portion of United States ...... 3
2. Sketch map of central Alabama ...... 5
3. Sketch map of area of Pratt’s Ferry and Pratt’s Syncline sections ...... 7
4. Sketch map of area of Calera section ...... 9
5. Stratigraphie section at Pratt’s Ferry ...... 29
6 . Stratigraphie section at Pratt’s Syncline ...... 34
7. Stratigraphie section at Calera ...... 38
8 . Stratigraphie distribution of graptolites in the Pratt’s-Ferry-Pratt’s Syncline section ...... 40
9. Stratigraphie distribution of graptolites in the Calera section ...... 46
10. Correlation of the Pratt’s Ferry-Pratt's Syncline sections ...... 51
11. Correlation of the graptolite biostratigraphy in Alabama with that of other geographic areas ..... 55
12. Method of measuring dicellograptid specimens ...... 72
13. Method of measuring diplograptid specimens ...... 73
14. Non-isolated specimens of Pterograptus sp., Didymograptus sp., and Azvgograptus incurvus EkstrBm ...... 80
15. Isolated specimens of Reteograptus geinitzianus Hall ... 91
16. Isolated specimens representing early growth stages of Reteograptus geinitzianus Hall ...... 93
IX Text-figure Page
17. Isolated specimens of Nemagraptus gracilis (Hall) .... 108
18. Isolated specimens representing early growth stages of Nemagraptus gracilis (Hall) ...... Ill
19. Isolated specimens representing thecae of Nemagraptus gracilis (Hall) ...... 113
20. Isolated specimens and reconstructions of Nemagraptus gracilis (Hall) ...... 115
21. Non-isolated specimens of Nemagraptus gracilis (Hall) .. 117
22. Non-isolated specimens of Nemagraptus gracilis (Hall) .. 119
23. Non-isolated specimens of Amphigraptus ...... 159
24. Isolated and non-isolated specimens of Nemagraptid sp. A ...... 158
25. Proposed phylogenetic relationships within Dicranograptidae ...... 182
26. Stratigraphie variation in thecal spines in specimens of Dicellograptus alabamensis Ruedemann ...... 212
27. Stratigraphie variation in thecal spines in specimens of Dicellograptus gurleyi gurleyi Ruedemann from the Calera section ...... 237
28. Stratigraphie variation in thecal spines in specimens of Dicellograptus gurleyi gurleyi Ruedemann from the Pratt's Ferry-Pratt's Syncline section ... 239
29. Stratigraphie variation in sicular orientation in specimens of Dicellograptus sextans (Hall) from the Calera section ...... 259
30. Stratigraphie variation in sicular orientation in specimens of Dicellograptus sextans (Hall) from the Pratt's Ferry-Pratt's Syncline section ...... 261
31. Isolated specimens of Dicellograptus alabamensis R u e d e m a n n...... 266 Text-figure Page
32. Isolated specimens of Dicellograptus alabamensis Ruedemann ...... 268
33. Non-isolated specimens of Dicellograptus alabamensis Ruedemann ...... 270
34. Isolated specimens representing early growth stages of Dicellograptus alabamensis Ruedemann .... 272
35. Isolated specimens representing early growth stages of Dicellograptus alabamensis Ruedemann .... 274
36. Non-isolated and isolated specimens of Dicellograptus bispiralis bispiralis (Ruedemann) and Dicellograptus bispiralis n. ssp. A ...... 276
37. Isolated specimens representing early growth stages of Dicellograptus bispiralis bispiralis (Ruedemann) ...... 278
38. Non-isolated specimens of Dicellograptus geniculatus Bulman and Dicellograptus gurleyi n. ssp. A .. 280
39. Isolated specimens of Dicellograptus gurleyi gurleyi Ruedemann ...... 282
40. Non-isolated specimens of Dicellograptus gurleyi gurleyi Ruedemann ...... 285
41. Isolated specimens representing early growth stages of Dicellograptus gurleyi gurleyi Ruedemann...... 287
42. Isolated specimens representing early growth stages of Dicellograptus gurleyi gurleyi Ruedemann ...... 289
43. Isolated and non-isolated specimens of Dicellograptus sextans (Hall) ...... 291
44. Non-isolated specimens of Dicellograptus sextans (Hall). 293
45. Isolated specimens representing early growth stages of Dicellograp tus sextans (Hall) ...... 295
46. Isolated specimens representing early growth stages of Dicellograptus sextans (Hall) ...... 297
xi Text-figure Page
47. Isolated and non-isolated specimens of Leptograptus trentonensis Ruedemann ...... 307
48. Isolated specimens showing rhabdosomal development in Leptograptus trentonensis Ruedemann ...... 3 0 9
49. Isolated specimens showing rhabdosomal development in Leptograptus trentonensis Ruedemann ...... 3 1 1
50. Non-isolated specimens of Dicranograptus irregularis Hadding ...... 327
Xll INTRODUCTION
Graptolite-bearing rocks of Middle Ordovician age are exposed
throughout the southern Appalachians in northwest trending fold belts
(Text-fig. 1). These rocks are here collectively referred to as the
Athens Shale. Their graptolite fauna first received scientific notice with Ruedemann's (1908) description and illustration of five species
and varieties from a locality in Bibb County, Alabama, that is known
as Pratt's Ferry (Text-figs. 1-3).
Since this initial study, graptolites have been collected from the
Athens Shale in Alabama and at many other localities throughout the
Southern Appalachians by field parties of the U.S. Geological Survey,
by university geologists, and by Decker (1952). Yet the only additional
descriptive study of these graptolites is that of Ruedemann (1947), in
which five new species and varieties are described. Two of these
were defined on the basis of material collected at Pratt's Ferry; one
was based on material from a locality in Shelby County, Alabama known
as Calera (Text-fig. 1,3-4); and the remaining two were based on
material collected near Bristol, Tennessee (Text-fig. 1). However,
Ruedemann's (1908, 1947) taxonomy and faunal lists, as well as the
faunal lists compiled by previous workers (e.g. Decker, 1952), are EXPLANATION OF TEXT-FIGURE 1
Map of portion of the United States showing geographic localities mentioned in the text and outcrops of Middle Ordovician rocks in the Southern Appalachians. (A) Albany, Hudson River Valley, New York; (B) Bristol, Tennessee; (G) Gadsden, Alabama; (V) Vincent, Alabama; (C) Calera, Alabama; (PF) Pratt's Ferry, Alabama; (M) Marathon Region, Texas. -Ni
• M 100 300 500 miles
500 k m
Text-figure 1 w EXPLANATION OF TEXT-FIGURE 2 Sketch map of Central Alabama showing distribution of Ordovician rocks and location of sections mentioned in the text. Stippled pattern denotes Athens Shale and Lenoir Limestone, dashed-line pattern other Ordovician and Cambrian rocks. White areas represent post-Ordovician rocks. (PF) Pratt's Ferry; (PB) Pratt's Bluff; (PS) Pratt's Syncline; (C) Calera; (SS) Simpson Spring; (S) Saginaw. BIRMINGHAM ALABAMA
AREA SHOWN BELOW PSJ
JEFFERSON CO
[s»J
TUSCALOOSA CO. SHELBY CO
Tuscaloosa
[271
CHILTON CO
PS; 25] Montgomery Tuscaloosa
Centerville 62]
BIBB CO
Montgomery
Text-figure 2 EXPLANATION OF TEXT-FIGURE 3
Sketch-map showing location of the Pratt's Ferry and Pratt's Syncline sections. 01 denotes Lenoir Limestone; Oa Athens Shale including the Pratt Ferry beds; Df and Mfp the Frog Mountain Sandstone and the Fort Payne Chert, respectively. (P.P.) Pratt's Ferry; (P.B.) Pratt's Bluff; (P.S.) Pratt's Syncline. 30 29 28 2 7
PB' 36
Df & Mfp
[25J
4. ï milts ^ TO CENTEKVIUE 1km
Text-figure 3 EXPLANATION OF TEXT-FIGURE 4
Sketch-map showing location of the Calera section. Lines with bars at both ends represent the section. The barred line within the hachured line represents the quarry section, and the other barred line located near highway 25 represents the highway 25 roadcut section. The light stippled pattern denotes the Lenoir Limestone, the diagonal line pattern the Athens Shale, and the densely stippled pattern the Frog Mountain Sandstone and the Fort Payne Chert. % m ile
I— V a k m — I
2 4 25
1 Va m ile s TO CALER/I
TO MONTEVAllO
Text-figure 4 10 today badly out of data and in need of revision. In addition, very few of the many graptolite collections from the Athens Shale are precisely located stratigraphically, and thus they are of limited biostratigraphic value.
In 1971, it was brought to the writer's attention that the Athens
Shale is calcareous locally and that it might be possible to isolate
graptolite specimens by means of acid treatment from the calcareous
facies of the Athens Shale at Pratt’s Ferry and Pratt's Syncline (Text-
fig. 2,3). In addition, the excavation of a shale quarry near Calera,
Alabama proved to expose a continuously graptolitiferous section of the black shale facies of the Athens Shale, which could be precisely located
stratigraphically. Because biostratigraphic studies of concdonts from beds immediately underlying the Athens Shale at Pratt's Ferry and Calera had been carried out (Sweet and Bergstrom, 1962; Bergstr&n, 1971a; Berg
strom and Drahovzal, 1972), it seemed possible that the graptolite biostratigraphy, once established, could be correlated with the condonont biostratigraphy. Thus, the project on which this report is based was
undertaken from the following purposes:
1) To study isolated graptolite specimens with regard to their
detailed morphology and its bearing on the taxonomy and iden
tification of non-isolated specimens.
2) To work out the graptolite biostratigraphy at Calera, Pratt's
Ferry, and Pratt's Syncline.
3) To decipher in Alabama the east-west facies relations within
the Athens Shale and immediately adjacent strata on the basis 11
of graptolite biostratigraphy and with the aid of conodont
information.
4) To correlate the ; graptolite sequence of the Athens Shale in
Alabama with those of other areas in North America and north
western Europe.
Methods of Study
Most of the field work was carried out during the Spring seasons of 1973 and 1974. At those times, the Pratt's Ferry, Pratt's Syncline and Calera sections were measured, described, and sampled. Two sets of graptolite collections were obtained from the Pratt's Ferry and
Pratt's Syncline sections, one for acid preparation, the other for thé study of non-isolated specimens. At Calera, nearly every bedding plane is covered with graptolites; therefore, as many horizons as possible were collected in the available time. In order to sample the section systematically, an attempt was made to collect horizons that were evenly distributed stratigraphically throughout the section, but special emphasis was placed on the biostratigraphically most important, lower part of the section.
In order to isolate the graptolites, rock specimens were etched first with acetic acid, then with hydrofluoric acid. Etching by strong acetic acid required two to six weeks depending on the carbonate to clay ratio of the rock specimens. Many samples, primarily the lime mudstone that predominates in the upper two-rthirds of the Pratt's Syn cline section, broke down rapidly to a mud and thus required no further acid treatment. Décalcification of samples with a high clay content, which originated primarily from the Pratt's Ferry section and the lower 12 third of the Pratt's Syncline section, produced an extremely porous rock.
After being washed for five to seven days, the decalcified rock was im mersed for one to seven days in concentrated hydrofluoric acid until it broke down to a mud. The final product of the acid treatment was a muddy sludge, \diich was washed and screened gently to remove a large fraction of the undissolved clay. A fine sable-hair brush was then used to pick the graptolite specimens from the concentrated residues and to transfer them to storage containers s i m i lar to those described by Skoglund (1961). Several attempts were made to clear the specimens with various bleaching agents, but none was successful. The failure of the bleaching techniques was more than accounted for by the large number of specimens that was obtained.
Specimens intended for photographic illustration were either stored in water or dried and mounted on plastic slides with deep wells.
Many transparent early growth stages and stipe fragments were mounted on glass slides for study at high magnifications. These specimens were first dehydrated by the normal alcohol series and then mounted with Histoclad, a mounting medium manufactured by Clay-Adams, Inc.,
New York City. Specimens intended for illustration by camera-lucida line drawings were stored in either water or glycerine. Specimens stored in glycerine could be oriented in many different positions in which they would remain for some time; however, while preparing camera- lucida drawings, heat from the illuminator caused the specimen to migrate throu^ the medium. This phenomena did not affect specimens stored in water, but because the water did not support the specimens,
the number of aspects that could be illustrated for each specimen 13 were usually limited to two.
Both isolated and non-isolated specimens were studied with a model
M5 Wild binocular microscope. Surface morphology of opaque specimens was most visible in direct light. Transmitted light reflected off a white background proved to be best for the study of transparent specimens.
A 20X calibrated eyepiece was used for measurements, and line drawings of both isolated and non-isolated specimens were made with the aid of a Wild camera lucida.
Previous Biostratigraphic Studies
The first descriptive report on graptolites from the Athens Shale is that of Ruedemann (1908), in which is listed a fauna that was collec ted from the locality designated herein as the Pratt’s Ferry section.
Ruedemann (1908, p. 12) considered this fauna of eight species, two new, to be essentially, but not exactly, identical to the Normanskill fauna of New York State. Ruedemann’s (1908) Pratt’s Ferry fauna is as follows;
Nemagraptus gracilis var. surcularis (Hall) Dicellograptus smithi nov. moffatensis Carruthers var. alabamensls nov. 2 " cf. mensurans nov. Diplograptus foliaceus var. alabamensis nov. Climacograptus cf. putillus (Hall) Cryptograptus tricomis (Carruthers) Glossograptus ciliatus Emmons
In his report on the geology of Alabama, Butts (1926; PI. 22, fig.
10; PI. 23, figs. 1-8, 11-12) illustrated, but did not describe, several graptolite species from the Athens Shale in Alabama. From Pratt’s Ferry,
Butts reported, in his Plate explanations, Dicellograptus moffatensis var. alabamensis, D. smithi, Glossograptus ciliatus, and Diplograptus foliaceus.
From Pratt’s Ferry and Simpson Spring (Text-fig. 2), he reported 14
Nemagraptus gracilis, and from Pratt's Bluff (Text-fig. 2), Climacograp- tus scharenbergl. Butt's illustrations of the first four species listed above are copied from Ruedemann (1908). His illustration of N^. gracilis is copied from Hall (1859, fig. 6 ). His figured specimens of C^. scharenbergi are stored at the U.S. National Museum (U.S.N.M. 71489).
In 1947, Ruedemann again reported on the graptolites of the Athens
Shale in Alabama. In a U.S. Geological Survey collection (U.S.G.S. loc.
440 0., Ulrick, Butts, Mesler coll.) from Pratt's Ferry, Ruedemann
(1947, p. 80) identified the following species:
Thamnograptus capillaris Hall Didymograptus sagitticaulis Gurley D^. subtenuis (Hall) Nemagraptus exilis (Lapworth) JN. gracilis var. sub tenuis n. var. Dicellograptus moffatensis var. alabamensis Ruedemann jD. gurleyi Lapworth var. exilis n. var. smithi Ruedemann Dicranograptus furcatus (Hall) var. bispiralis n. var. Corynoides gracilis Hopkinson Diplograptus (Orthograptus) calcaratus Lapworth var. incisus D. (O'.) calcaratus var. acutus Lapworth (£•) angustifolius Hall Glyptograptus tamariscus var. euglyphus Lapworth Climacograptus modestus Ruedemann C^. modestus var. meridionalis n. var. Lasiograptus sp. Glossograptus ciliatus Emmons G^. ciliatus var. debilis Ruedemann G. whitfieldi (Hall) Cryptograptus tricomis n. sp. Retiograptus geinitzianus Hall
In another U.S. Geological Survey collection (U.S.G.S. loc. 323V, 323Vj^) this one from Simpson Spring, Ruedemann (1947, p. 81) identified the following:
Thamnograptus capillaris (Emmons) Didymograptus sagitticaulis Gurley D. subtenuis Gurley 15
Leptograptus flaccidus mut. trentonensis Ruedemann Nemagraptus gracilis (Hall) IÎ. exilis Lapworth Dicellograptus smithi Ruedemann jD. divaricatus var. bicurvatus Ruedemann Corynoides curtus Lapworth C^. gracilis Hopkinson Diplograptus (Orthograptus) calcaratus var. incisus Lapworth angustifolius (Hall) Glyptograptus tamariscus var. euglyphus Lapworth Climacograptus bicomis (Hall) C^. eximius Ruedemann C^. scharenbergi Lapworth C^. modestus var. meridionalis Ruedemann
Finally, Ruedemann (1947), p. 81) described a collection from a road cut near Calera. Exposures in this road cut are in the upper part of the Calera section described below. The collection available to Ruede mann is stored at the New York State Museum (Mus. loc. 5391), and in it
Ruedemann reported the following species:
Thamnograptus poori Ruedemann Corynoides calicularis Nicholson Didmograptus subtenuis (Hall) Didymograptus (Isograptus) lyra Ruedemann Dicellograptus sextans (Hall) Diplograptus (Orthograptus) calcaratus var. alabamensis Ruedemann Diplograptus (Glyptograptus) angustifolius (Hall) 2» (G.) euglyphus Lapworth 2» (G.) euglyphus var. pygmaeus Ruedemann Climacograptus modestus Ruedemann C^. caudatus Lapworth Cryptograptus tricomis Carruthers
Even though he noted important faunal differences between the Athens
Shale of Alabama, and the Normanskill of New York State, Ruedemann
(1947, p. 79) correlated the Athens Shale with the lower Normanskill shale (Zone of Nemagraptus gracilis). He accounted for the faunal differences by proposing the existence of two separate faunal provinces.
Decker (1952) collected, identified, and illustrated Athens Shale graptolites from a very large number of localities throughout the 16 southern Appalachians. At many of the larger exposures. Decker measured sections, and even though most of his collections spanned stratigraphie intervals several feet thick, he was able to recognize "...three some what overlapping major zones" (Decker, 1952, p. 2). Decker did not de fine these zones, nor did he relate them to any particular section. He assigned his Alabama graptolite faunas to the lower two of his three zones and considered them to be of Trenton age.
At Pratt's Bluff Decker measured a section in the Athens Shale that extended downward from the overlying Frog Mountain Sandstone. On the basis of his faunal lists. Decker (1952, p. 16) thought that the "basal part of the Athens Shale was not deposited here, as the section lacks the dominant lower zone of the climacograptids and diplograptids." The species in Decker's faunal lists from Pratt's Bluff are as follows:
1) From a 10 ft. interval, the top of which is 230 ft. below the base of the Frog Mountain Sandstone:
Azygograptus simplex Ruedemann Climacograptus parvus Hall Dicellograptus intortus Lapworth moffatensis var. alabamensis Ruedemann D^. sextans (Hall) Nemagraptus exilis Lapworth N. gracilis (Hall)
2) From an 8 ft. interval, the top of which is 290 ft. below the base of the Frog Mountain Sandstone:
Azygograptus simplex Ruedemann Climacograptus modestus Ruedemann Cryptograptus tricomis (Carruthers) Dicellograptus gurleyi var. exilis Ruedemann D^. intortus Lapworth D. moffatensis var. alabamensis Ruedemann D^. sextans (Hall) Dicranograptus nicholsoni Hopkinson Didymograptus sagitticaulis Gurley D. serratulus (Hall) 17
Glossograptus quadrimucronatus Lasiograptus (Hallograptus) mucronatus (Hall)
3) From a 10 ft. interval, the top of which is 298 ft. below the base of the Frog Mountain Sandstone:
Climacograptus antiquus Lapworth C^. modestus Ruedemann C^. parvus Hall Dicellograptus moffatensis var. alabamensis Ruedemann D. sextans (Hall) Dicranograptus sp.
4) From a 4 ft. interval, the top of which is 308 ft. below the base of the Frog Mountain Sandstone:
Didymograptus sagitticaulis Gurley D^. serratulus (Hall) Diplograptus vespertinus Ruedemann
From Pratt's Ferry, Decker (1952) reported the following species, which he assigned to the middle of his three Athens graptolite zones:
Climacograptus caudatus Lapworth C^. modestus Ruedemann C^. parvus Hall Cryptograptus tricomis (Carruthers) Dicellograptus divaricatus var. Salopiansis Elies & Wood 2. gurleyi var. exilis Ruedemann 2» intortus Lapworth D^. moffatensis var. alabamensis Ruedemann D^. sextans (Hall) Dicellograptus aextans var. exilis Elies & Wood p^. smithi Ruedemann Dicranograptus nicholsoni var. parvangulus Gurley p^. spinifer Lapworth Didymograptus sagitticaulis Gurley Diplograptus (Glyptograp tus) euglyphus Lapworth (Orthograptus) calcaratus var. alabamensis Ruedemann Glossograptus ciliatus Emmons Lasiograptus (Thysanograptus) eucharis (Hall) Nemagraptus exilis Lapworth
Near Calera, Decker (1952) collected from a measured section loca ted in a road-cut. This exposure is included in the upper part of the
Calera section of the present study, and it is the same exposure near
Calera from which Ruedemann (1947) reported a fauna (see above). From 18
this locality Decker identified 39 species and varieties, which he as
signed to his middle Athens graptolite zone and the uppermost part of his
lower zone. From a 5 ft. interval immediately below the overlying Frog
Mountain Sandstone, Decker reported the following graptolite taxa.
Climacograptus antiquus var. lineatus Elies & Wood C^. modestus Ruedemann C^. modestus var. meridionalis Ruedemann C^. parvus Hall Corynoides gracilis Hopkinson Cryptograptus tricomis (Carruthers) Dicellograptus divaricatus var. bicurvatus Ruedemann 2» divaricatus var. salopiensis Elies & Wood forchammeri (Geinitz) 2.* Rurleyi var. exilis Ruedemann D^. intortus Lapworth mensurans Ruedemann D. minimus Ruedemann D. moffatensis var. alabamensis Ruedemann 2" sextans (Hall) sextans var. exilis Ruedemann jD. smithi Ruedemann Didymograptus sagitticaulis Gurley 2" serratulus (Hall) D^. sub tenuis (Hall) Diplograptus (Amplexograptus) macer Ruedemann D. (Glyptograptus) euglyphus Lapworth ]D. (Glyptograptus) euglyphus var. pygmaeus Ruedemann Di. (Glyptograptus) teretiusculus mut. occidentalis Ruedemann 2" (Glyptograptus) vespertinus Ruedemann (Orthograptus) calcaratus var. acutus Lapworth 2» (Orthograptus) calcaratus var. incisus Lapworth Glossograptus ciliatus var. debilis Ruedemann Lasiograptus (Hallograptus) bimucronatus (Nicholson) L. Thysanograptus eucharis (Hall) Ij. (Hallograptus) mucronatus (Hall) Nemagraptus exilis Lapworth N. gracilis (Hall) Thamnograptus capillaris (Emmons poori Ruedemann
From a 5.5 ft. interval, the top of which is 5 ft. below the Frog
Mountain Sandstone, Decker identified the following species:
Climacograptus eximius Ruedemann C^. modestus Ruedemann C^. scharenbergi Lapworth 19
Cryptograptus tricomis (Carruthers) Dicellograptus intortus Lapworth D. mensurans Ruedemann moffatensis var. alabamensis Ruedemann D. sextans (Hall) D. smithi Ruedemann Didymograptus sagitticaulis Gurley D^. subtenuis (Hall) Diplograptus (Glyptograptus) euglyphus Lapworth D^. (Orthograptus) calcaratus var. alabamensis Ruedemann Blossograptus ciliatus Emmons Lasiograptus (Hallograptus) mucronatus (Hall) Nemagraptus gracilis (Hall) Thamnograptus poori Ruedemann
From a 2 ft. interval, the top of which is 10.5 ft. below the Frog
Mountain Sandstone, Decker identified the following species:
Climacograptus antiquus Lapworth C^. modestus Ruedemann jC. modestus var. meridionalis Ruedemann Pazvus Ruedemann C^. scharenbergi Lapworth Cryptograptus tricomis (Carruthers) Dicellograptus divaricatus var. salopiensis Elies & Wood D^. gurleyi var. exilis Ruedemann D^. mensurans Ruedemann D. moffatensis var. alabamensis Ruedemann D. patulosus Lapworth D. sextans (Hall) Didymograp tus sagitticaulis Gurley subtenuis (Hall) Diplograptus (Orthograptus) calcaratus var. alabamensis Ruedemann
From a 13.5 ft. interval, the top of which is 12.5 ft. below the Frog
Mountain Sandstone, Decker identified the following species:
Dicellograptus moffatensis var. alabamensis Ruedemann Diplograptus (Glyptograptus) euglyphus var. pygmaeus Ruedemann ]D. (Orthograptus) calcaratus var. acutus Lapworth
From a 4 ft. interval, the top of which is 26 ft. below the Frog Mountain
Sandstone, Decker identified the following species:
Climacograptus eximius Ruedemann C^. modestus Ruedemann C^. scharenbergi Lapworth 20
Dicellograptus intortus Lapworth D. sextans (Hall) Didymograptus sagitticaulis Gurley D^. subtenuis (Hall) Diplograptus (Glyptograptus) euglyphus var. pygmaeus Ruedemann 2" teretiusculus (Hisinger) D. (Orthograptus) calcaratus var. acutus Lapworth D. (Orthograptus) calcaratus var. alabamensis Ruedemann 2« vespertinus Ruedemann
Near Simpson Spring, 2.5 miles northwest of the town of Calera, Dec ker (1952) obtained collections from three separate localities, which he did not locate stratigraphically. In these collections. Decker identified
38 species and varieties, which are as follows:
1. West side of road, 5 ft. stratigraphie interval:
Climacograptus modestus Ruedemann modestus var. meridionalis Ruedemann 2* parvus Hall 2« scharenbergi Lapworth Cryptograptus tricomis (Carruthers) Dicellograptus divaricatus var. salopiensis Elies & Wood 2* gurley var. exilis Ruedemann 2« intortus Lapworth 2« moffatensis var. alabamensis Ruedemann 2» sextans (Hall) Diplograptus (Glyptograptus) euglyphus var. pygmaeus Ruedemann 2- (Glyptograptus) vespertinus Ruedemann 2* (Orthograptus) calcaratus var. acutus Lapworth 2* (Orthograptus) calcaratus var. alabamensis Ruedemann 2« (Orthograptus) calcaratus var. incisus Lapworth Glossograptus ciliatus Emmons Nemagraptus gracilis (Hall) Thamnograptus poori Ruedemann
2. East side of road, 4 ft. stratigraphie interval:
Climacograptus eximius Ruedemann 2» modestus var. meridionalis Ruedemann 2* parvus Hall 2» scharenbergi Lapworth D icellograp tus divaricatus var. bicurvatus Ruedemann 2» sextans (Hall) 2* sextans var. exilis Elies & Wood Didymograptus sagitticaulis Gurley 2* serratulus (Hall) Diplograptus (Glyptograptus) euglyphus Lapworth 21
D^. (Glyptograptus) teretiusculus mut. occidentalis Ruedemann 2" (Glyptograptus) vespertinus Ruedemann (Orthograptus) calcaratus var. acutus Lapworth D^. (Orthograptus) calcaratus var. alabamensis Ruedemann 2" (Orthograptus) calcaratus var. incisus Lapworth Thamnograp tus capillaris (Emmons)
3. 200 ft. east of road, 2 ft. stratigraphie interval:
Azygograptus canadensis Ruedemann Climacograp tus antiquus Lapworth 2- modestus Ruedemann 2- modestus var. meridionalis Ruedemann £. parvus Hall 2» scharenbergi Lapworth C. typicalis (Hall) Corynoides gracilis mut. perungulatus Ruedemann Dicellograptus forchammeri (Geinitz) D. gurleyi Lapworth 2- gurleyi var. exilis Ruedemann 2« mensurans Ruedemann 2» sextans (Hall) 2- sextans var. perexilis Ruedemann 2- smithi Ruedemann D idymograp tus sagitticaulis Gurley Diplograptus (Amplexograptus) amplexicaulis (Hall) 2- (Glyptograptus) euglyphus Lapworth 2» (Glyptograptus) teretiusculus (Hisinger) 2* (Glyptograptus) teretiusculus mut. occidentalis Ruedemann 2« (Glyptograptus) vespertinus Ruedemann 2» (Orthograptus) calcaratus var. acutus Lapworth 2» (Orthograptus) calcaratus var. alabamensis Ruedemann 2* (Orthograptus) calcaratus var. incisus Lapworth Leptograptus flaccidus mut. trentonensis Ruedemann Nemagraptus gracilis var. distans Ruedemann 2- gracilis var. surcularis (Hall) Thamnograptus capillaris (Emmons)
Decker assigned this last collection to the top of the lowest of his three major Athens graptolite zones.
Near Saginaw, Alabama (text-fig. 2), Decker (1952) reported, from the upper part of the Athens Shale, eleven graptolite species and varie ties, which he assigned to the middle of his three major graptolite zones.
His faunal list is as follows:
Callograptus hepaticus Ruedemann 22
Climacograptus modestus Ruedemann C^. parvus Hall Dicellograptus moffatensis var. alabamensis Ruedemann sextans (Hall) Diplograptus (Glyptograptus) euglyphus Lapworth (Orthograptus) calcaratus Lapworth (Orthograptus) calcaratus var. acutus Lapworth (Orthograptus) calcaratus var. alabamensis Ruedemann (Orthograptus) calcaratus var. incisus Lapworth Thamnograp tus capillaris (Emmons)
Decker reported three species from a locality near Vincent, Alabama
(text-fig. 1 ), which he believed represented the middle of his three major Athens graptolite zones. These species are: Climacograptus nodes- tus Ruedemann, C^. scharenbergi Lapworth, and Nemagraptus gracilis (Hall).
Decker identified graptolites in three U. S. Geological Survey col lections that were collected by P. E. Cloud near Gadsden, Alabama
(text-fig. 1). The species, identified by Decker in these collections, are:
Locality AF-62 Didymograptus sagitticaulis Gurley Diplograptus (Orthograptus) calcaratus var. incisus Lapworth Glossograptus whitfieldi (Hall)
Locality AF-75 Climacograptus caudatus Lapworth Diplograptus (Glyptograptus) euglyphus var. sepositus Keble and Harris D. (Amplexograptus) perexcavatus Lapworth
Locality AF-78 Cryp tograp tus tricomis (Carruthers) Glossograptus ciliatus Emmons G. whitfieldi (Hall)
The Pratt's Ferry section has served as an important locality, not only for studies of graptolites, but also for studies of brachiopods
and condonts. The exposed section, which is described in detail below, begins in the upper part of the Lenoir Limestone and extends through the
Pratt Ferry beds and into the Athens Shale (stratigraphie nomenclature 23 after Drahovzal and Neathery, 1971).
As the result of the classical brachiopod work of Cooper (1956), the Pratt’s Ferry section has become known worldwide. Cooper referred the upper part of what is mapped as Lenoir Limestone to the Little Oak
Formation because it contains a Christiania fauna. From the overlying
Pratt Ferry beds. Cooper obtained an extremely diverse fauna comprising more than 62 genera. He considered the fauna from the Pratt Ferry beds, together with the Christiania fauna from the upper part of the Lenoir
Limestone, to be closely related to those of the Little Oak Formation of Cahaba Valley, Alabama, to the Arline Formation of northeastern Ten nessee and southwestern Virginia, and to the Stinchar-Balclatchie brachiopod faunas of the Girvan District of Scotland. Cooper assigned the Christiania-bearing beds of theLenoir, the Pratt Ferry beds, and the immediately overlying Athens Shale (Columbiana Shale of Cooper, 1956) to his Porterfield Stage.
Sweet and BergstrSm (1962) obtained conodonts referable to 37 spe cies from the Pratt Ferry beds at the Pratt’s Ferry section. This fauna is remarkable in that it shows an extremely close affinity to fau nas from a relatively limited interval in the Middle Ordovician sequence of Sweden. Bergstrom (1971) was able to place the boundary between the
Pygodus serrus and Pygodus anserinus conodont zones 0.3 m below the top of the Pratt Ferry beds at the Pratt’s Ferry section. At Calera, Berg strom (1971) and BergstrSm and Drahovzal (1972) reported conodonts from the basal beds of the Athens Shale (Bergstrom's, 1971, Columbiana Shale) that are indicative of the lowest subzone of the P. serrus zone. PHYSICAL STRATIGRAPHY
Introduction
Two of the rock-stratigraphic units used herein, namely the Lenoir
Limestone and Athens Shale were first established in Alabama by Butts
(1926) and they are presently recognized by the Geological Survey of
Alabama (Drahovzal and Neathery, 1971). However, these units have been re-defined at varous times in the past.
Hayes (1894) originally applied the name Athens Shale to what was described as black graptolitic shale that he mapped in the Kingston and Cleveland quadrangles of Tennessee. He designated a type section at an exposure along the Nashville, Tellico and Charleston Railroad about
1.5 miles (2.4 km) northeast of Athens, Tennessee. During the following
50 years, the term Athens Shale was applied by many workers to black graptolitic shales exposed throughout the entire length of the Southern
Appalachians.
After re-examining Hayes' type section. Cooper (1956, p. 43) repor ted that the type Athens contains only a small amount of shale and consists mostly of "silty or sandy, cobbly limestone that weathers to a yellow sandy, shaly rock." Because the term Athens was mostly used for black graptolitic shales, so clearly unlike anything in the type section, he proposed that the term Athens be restricted to the type area where it is a facies of the Arline Formation. For the black graptolitic shales that had previously been termed Athens, Cooper and
24 25
Cooper (1946) assigned the name Paperville Formation in the Bristol area and Cata\vba Valley of Virginia, and Cooper (1956) introduced the term
Columbiana Formation for the unit previously referred to as the Athens in Alabama. Additionally, Neuman (1965) proposed the name Blockhouse
Formation for the black graptolitic shales exposed along the base of the
Great Smoky Mountains in eastern Tennessee.
Butts (1926) introduced the name Athens to Alabama stratigraphy and applied it to black graptolitic shale. He included a fossiliferous limestone that occurs at the base of the shale at Pratt's Ferry within his definition of the formation. When Cooper and Cooper (1946) proposed the name Columbiana Formation for the shale, they raised the basal lime stone to formation status and proposed the name "Pratt Ferry Formation" for it. They considered the Pratt Ferry Formation to be equivalent to a part of the Little Oak Limestone, which they regarded as a lateral facies of the Columbiana Formation.
The Lenoir Limestone was named by Safford and Killebrew (1876) for blue-gray limestone of Middle Ordovician Age, which unconformably overlies the Knox Dolomite at Lenoir City, Tennessee. On the basis of similarities in lithology and stratigraphie position. Butts (1926) extended the usage of this term to rocks of early Middle Ordovician
Age in Alabama. Cooper and Cooper (1946) restricted the use of the
term "Lenoir" to beds equivalent to the sequence exhibited at the type
locality at Lenoir City. So restricted, the Christiania-bearing beds
at the top of the Lenoir at Pratt's Ferry were excluded from the Lenoir,
and because of faunal similarities, they were considered by Cooper
(1956) to represent the Little Oak Limestone, even though these two 26 units differ lithologically. Cooper (1956) considered his Pratt Ferry
Formation to be a manifestation of, and to be situated within, his Little
Oak Limestone because of faunal similarities.
Obviously, Cooper's (1956) modification of Butts' (1926) strati graphie terminology is based primarily on faunal similarities, and his terminology has never gained wide acceptance in Alabama. The terminology employed in the present study is that used by the Geological Survey of ■
Alabama. The Lenoir Limestone includes the Christiania-bearing beds.
Cooper's (1956) Pratt Ferry Formation has a very limited thickness and areal extent. Although it is considered to be a basal unit of the Athens
Shale, it is herein referred to separately as the Pratt Ferry beds. The term Athens Shale is used in the present study because of its wide usage and recognition; however, it is doubtful that this unit is continuous with black graptolitic shale throughout the Southern Appalachians, and as it is used herein, it does not refer to the Pratt Ferry beds.
Lithology At Measured Sections
Graptolites were collected from measured sections through the Athens
Shale at Pratt's Ferry, Pratt's Syncline, and Calera. The state of preservation of the graptolite specimens differs greatly between the sections, and reflects differences in lithology between the sections.
At Calera, the Athens Shale is black shale. The lower part of the Athens
Shale at Pratt's Ferry and Pratt's Syncline consists of gray calcareous shale and lime mudstone, and the upper part of the Athens Shale at .
Pratt's Syncline is medium-to thick-bedded lime mudstone. The litho logy of these two facies, the eastern black shale facies and the western 27 carbonate faciès, is described below for each of the measured sections.
Pratt's Ferry Section
The Pratt's Ferry section is in a road-cut on the west side of highway 27 approximately 0.25 miles (0.4 km) east of Pratt's Ferry bridge over the Cahaba River in Bibb County, Alabama (Text-figs. 2-3).
The measured section is 25.8 m thick. From a horizon 3.75 m below the top of the Lenoir Limestone, the section extends upward through the
Pratt Ferry beds to the highest exposed bed of the Athens Shale (Text- • fig. 5).
The Lenoir Limestone is a massively bedded, medium gray, bioclastic wackestone. The bedding is defined by irregular clay horizons 10 to 40 cm apart. Large fossils such as Maclurites are common, and Cooper (1956) reported a Christiania brachiopod fauna from this locality. However, most of the fossil material is fossil debris made up of fragments up to
3 mm in size. The bioclastic debris is primarily crinoidal; however, sponge' spicules and fragments of bryozoans, brachiopods, and trilobites are common. A single algal-coated grain was observed in one thin sec
tion.
The contract between the Lenoir Limestone and the overlying Pratt
Ferry beds is placed at 3.75 m in the measured Pratt's Ferry section
(Text-fig. 5). The lower boundary of the Pratt Ferry beds is charac
terized by a change within 1 0 cm to a dark gray, regularly bedded, very
coarse-grained, bioclastic grainstone. Bedding is at regular intervals
of 5 to 15 cm; however, the parting surfaces are way so that individual
beds pinch and swell and thus exhibit boudinage structures. The bedding. 28
EXPLANATION OF TEXT-FIGURE 5
Stratigraphie section at Pratt’s Ferry. Symbols : 1. Covered interval; 2. Mudstone, generally calcareous; 3. Shale, black shale, and calcareous shale; 4. Thin- to medium-bedded lime mudstone; 5. Thick-bedded lime mudstone; 6 . bentonite; 7. medium-bedded fossiliferous limestone (grainstone) with wavy bedding surfaces, i.e. Pratt Ferry beds; 8 . Massively bedded fossiliferous limestone (lime wackestone) with wavy bedding surfaces, i.e. Lenoir Limestone. 29
PRATT'S FERRY SECTION
Collections
>-5 2 .
3.
4 . Athens Shale 5.
6.
7.
Pratt Ferry H 8 Beds T 5 metres Lenoir 1
Text-figure 5 30 together with the coarseness of the bioclastic debris (3 to 10 mm), is the most useful feature for distinguishing the Pratt Ferry beds in the field. The bioclastic debris contains fragments of crinoids, brachio pods, trilobites, and bryozoans. In addition, chitinozoans and scoleco- donts are common in acid residues.
In thin sections, samples from the lower part of the Pratt Ferry beds consist predominantly of bioclastic debris (50 percent) and sparry calcite (50 percent). Large angular intraclasts were observed in a sample collected 0.75 m above the base of the Pratt Ferry beds. These intraclasts can be classified as bioclastic wackestone and closely resemble the Lenoir Limestone. Upward in the Pratt Ferry beds, the bioclastic content decreases to 20-30 percent and clasts become finer grained (less than 3 mm diam.). Sparry calcite also becomes finer grained, and approaches lime mud.
The contact between the Pratt Ferry beds and the overlying Athens
Shale is placed 8 m above the base of the Pratt's Ferry section. This
contact is quite gradational with the change in lithology occurring over a one metre interval. In the field it is marked by a change from
the way bedding surfaces of the Pratt Ferry beds to the smooth, planar bedding surfaces of the Athens Shale. In addition, large quantities of
clay are introduced into the section. The lowest 3 m of the Athens
Shale (the interval between 8 and 11 m above the base of the Pratt's
Ferry section) consists of dark gray lime mudstone. Bedding is regular
with partings occurring regularly at 5 to 10 cm intervals. The beds
pinch and swell and exhibit boundinage structure. The bioclastic debris 31 is fine grained (less than 3 mm) and constitutes less than 5 percent of the rock. Terrigenous mud constitutes as much as 20 percent of the rock by weight.
Between 11and 17 m above the base of the Athens Shale, the section is dominated by calcareous mudstone. The mudstone is tan to light gray and exhibits shaly weathering with parting surfaces at 1 to 3 cm inter vals. The carbonate content approaches 50 percent by weight. In thin section, the rock is faintly laminated by concentrations of clay. Beds of lime mudstone are rare in this interval, and a thin (less than 0 . 1 m) interval of shale occurs at 13 m. Fossils are scarce between 1 and 17 m. and they are primarily inarticulate brachiopods and trilobites. The first occurrence of graptolites at 12 m is rare specimens of Dicello graptus alabamensis and Dictyonema sp. However, at the shale horizon at
13 m, six species of graptolites occur in abundance. Between 13 and
17 m, graptolites are again scarce and represent D^. alabamensis and one specimen of Nemagraptus gracilis.
At 17 m above the base of the Pratt's Ferry section, the Athens
Shale takes on its characteristic shaly appearance and graptolites are extremely abundant. The interval between 17 and 25.8 m consists of two separate rock types, a calcareous shale and a lime mudstone. The cal careous shale composes approximately 60 to 75 percent of the stratigraphie thickness in the 17 to 25.8 m interval. It is light to medium gray and weathers as "slabs" as much as 0.5 m across. Graptolites cover the surfaces of these slabs. In thin section, the shale is very finely laminated by varying carbonate-clay concentrations. The 32 carbonate content of the shales approaches 60 percent in some lamina tions. The lime mudstones are light to medium gray, blocky weathering, unfossiliferous, regularly bedded, and bed thicknesses range from
5 to 10 cm. Characteristically, beds of lime mudstone pinch and swell.
The calcareous shale and lime mudstone occur in alternating stratigraphie intervals, which are 0.2 to 1.0 m thick. Above the 25.8 m level, the
Athens Shale is covered by a thick soil.
Pratt's Syncline Section
The Pratt's Syncline section is in creek beds in a small valley approximately 2.75 miles (4.4 km) southwest of the Pratt's Ferry section
(Text-figs. 2-3, 6 ). This valley is east of the Cahaba River and is accessible only from highway 25 by a dirt road. The Pratt's Syncline section is 146 m thick. It begins at a horizon 1.5 m below the top of the Lenoir Limestone and extends through the Pratt Ferry beds and the
Athens Shale to the base of the Devonian Frog Mountain Sandstone.
The lithology of the Lenoir Limestone and the Pratt Ferry beds resembles closely the lithology described above for those units at the
Pratt's Ferry section.
The boundary between the Pratt Ferry beds and the Athens Shale is in a covered interval, and the Athens Shale is first exposed 12 m above the base of the section. Betnjeen 12 and 33 m above the base of the section, the Athens Shale is calcareous shale and boudinage lime mud stone similar lithologically to those described in the Pratt’s Ferry
section. However, at Pratt's Syncline, the limestone dominates and the
shale generally occurs only as thin interbeds between beds of lime 33
EXPLANATION OF TEXT-FIGURE 6
Stratigraphie section at Pratt's Syncline. For symbols, see Text-figure 5. 34 PRATT'S SYNCLINE SECTION
Frog Mtn. Sst
Collections
Athens Shale
- T = 20 m etres Pratt Ferry Beds 1 Lenoir Text-figure 6 35 mudstone. Graptolite occurrences are rare and are restricted to the shale intervals. Twelve graptolite species occurring together in a shale interval 1 2 m above the base of the section mark the first occurrence of graptolites in the Pratt's Syncline section.
Between 33 and 146 m above the base of the Pratt's Syncline section, medium-to thick-bedded lime mudstone dominates. Again, shale beds with carbonate compositions as high as 50 percent by weight occur, generally as thin interbeds between beds of lime mudstone. The lime mudstone is light gray and in some instances occur as beds up to 50 cm thick. Significantly, individual beds do not show boudinage structure.
Most of the lime mudstone contains a high percentage (20 to 30 percent by weight) of terrigenous clay. Graptolites occur in both the calcareous shale and the lime mudstone. Although graptolites are rare in the lime mudstone, where they do occur, they are present in extremely large numbers concentrated in thin (5 to 10 cm) stratigraphie intervals.
Besides having a high clay content, the graptolitic lime mudstone is very finely laminated, and locally contains enough non-graptolitic bioclastic debris to be considered wackestone. In such instances, the bioclastic debris is very fine grained (less than 1 mm in size) and consists of fragments of bryozoans, brachiopods, and trilobites. The Athens Shale is unconformably overlain by the Devonian Frog Mountain Sandstone at 146 m above the base of the Pratt's Syncline section.
The entire section of the Athens Shale exposed at the Pratt's Ferry section appears to correlate lithologically with the lower part of the
Athens Shale at the Pratt's Syncline section. Additionally, thick- bedded lime mudstone was discovered in the woods approximately half a 36 mile CO. 8 km) east of the Pratt's Ferry section. The location of these beds suggests that the unexposed upper part of the Athens Shale at
Pratt's Ferry is lithologically similar to the upper part of the Athens
Shale at Pratt's Syncline. The lower few metres of the Athens Shale at
Pratt's Syncline are covered and a direct comparison can not be made of this part of the Athens Shale with the equivalent part at Pratt's Ferry.
However, on the basis of the stratigraphie distance above the top of the
Lenoir Limestone, the 12 to 33 m interval in the Pratt's Syncline sec tion probably correlates with the 14 to 25.8 m inteirval of the Pratt's
Ferry section. Within this interval, the Athens Shale at Pratt's Syn cline is more calcareous and less graptolitic. But in spite of these differences, which are only in degree, the equivalent intervals of the two sections can be correlated stratigraphically and they show the same graptolite fauna. Therefore, in the following biostratigraphic dis cussions, the two sections have been combined into one range chart
(Text-fig. 8 ) in order to show the range of graptolite species through the entire thickness of the Athens Shale in the areas of the Pratt's Fer ry and Pratt's Syncline sections.
Calera Section
The Calera section is in a shale quarry 3000 feet (0.9 km) north of highway 25, 3.3 miles (5.3 km) west of Calera in Shelby County, Ala bama (Text-fig. 4,7). The entire 75-metre thickness of the Athens Shale is exposed in this quarry from its basal contact with the Lenoir Lime stone to its top at the base of the Frog Mountain Sandstone (Text-, fig. 7). However, the upper part of the Athens Shale at the shale 37
EXPLANATION OF TEXT-FIGURE 7
Stratigraphie section at Calera. For symbols, see Text-figure 5. 38 CALERA SECTION Shale Quarry Highway 25 Frog Mtn. Sst.
Athens Shale
= } Collections
T 10 metres 1
Lenoir Text-figure 7 39
EXPLANATION OF TEXT-FIGURE 8
Stratigraphie distribution of graptolites in the combined Pratt’s Ferry-Pratt’s Syncline section. R Lenoir Pratt Ferry Beds AthensShsle ■Frog Mtn.Sst. ft I
"g Collections R Oicell. a/abamensis 00 Dictyonema sp.
Oicaulo.7 n. sp. A Dicell, bispiralis bispiraHs ooc Glosso. c /liatu si
Glypt. sp. of. G. leretiusculus •
Climaco. meridlonalis’
Crypt, marcidus'
Pseudocl. modestus
Nema. gracilis
Didymo. sp.
Glypt. euglyphus
Reteo. gainitzianus
Dicell, gurleyi gurleyh
Dicell, sextans
Dicrano. irregularis
Amphi. n. sp. A
Nemagraptid sp. A
P. serrus- ■P. anserinus Lepto. trentonensis
Ortho, sp * ■P- o 41
quarry is so deeply weathered that graptolites can not be collected.
Therefore, another section, representing the upper 29.5 m of the Athens
Shale, was measured in a road-cut on the south side of highway 25 direct
ly south of the shale quarry (Text-fig. 4, 7). Poorly preserved ;
graptolites can be collected from this section, which has served as a
graptolite locality for Ruedemann (1947) and Decker (1952). This road-
cut section can be stratigraphically tied into the shale quarry section by means of the overlying Frog Mountain Sandstone, and in the following biostratigraphic discussions, the shale quarry and road-cut sections
are referred to collectively as the Calera section.
The Lenoir Limestone at Calera is a bioclastic wackestone that
closely resembles the Lenoir Limestone at Pratt's Ferry. The basal
4.6 m of the overlying Athens Shale consist of alternating beds of black
shale and black boudinage limestones. The shale is calcareous (up to
2 0 percent by weight), and parting surfaces are covered with inarticulate brachiopods. The limestone occurs in beds 5 to 15 cm thick that pinch
and swell. In thin section, they can be classified as fossiliferous wackestone. The bioclastic debris consists dominately of fragments of
trilobites and brachiopods. The black color of the limestone appears
to be due to black clay filling voids and cracks within the rock.
Between 4.0 and 4.5 m above the base of the Calera section is a single
bed of fine-to medium-grained, bioclastic grainestone. Above this
bed, shale dominates in the section, graptolites occur in abundance,
inarticulate brachiopods are scarce, and the limestone beds that do oc
cur are very finely laminated lime mudstone with a high percentage 42
(30 to 50 percent by weight) of clay. The highest limestone bed is
20 m above the base of the Athens Shale, and except for 3-to 5-mm thick bentonite? beds, which occur at 7.5, 8 .8 , 12.2, 52, and 53.7 m, the rest of the Calera section consists of black, fissle shale in which nearly every parting surface is covered by graptolites.
Facies Relations and Environment of Deposition
As indicated above, the Athens Shale in Alabama consists of two
facies, an eastern shale facies at Calera and a western carbonate facies at Pratt's Ferry and Pratt's Syncline. Although the lithologie differen
ces appear to be great between these two areas, the two facies occur
adjacent to one another stratigraphically in each of the described
sections. The black color of the Calera Athens which is possibly due
to deposition in a reducing environment, tends to exaggerate the litho
logie differences. These differences are basically in the carbonate/
clay ratio, with the carbonate content increasing to the west and the clay
content increasing to the east.
The depositional environment of the Lenoir Limestone has been
interpreted by Stephenson et al. (1973), and Neumann (1976) has summarized
the tectonic and depositional history of the Middle Ordovician strata of the southern Appalachians. The lithologie and paléontologie evidence presented herein is consistent with these previous studies. The Lenoir
Limestone was deposited in very shallow water, in a shelf lagoon environ ment, whereas the Pratt Ferry beds and the Athens Shale were deposited 43 in progressively deeper-water shelf environments that were brought on by a westward transgressing sea. The bioclastic grainstone at the base of
the Pratt Ferry beds possibly indicates a zone of high energy between the shallow lagoonal basin of the Lenoir and the quiet, slightly deeper- water basin of the Athens Shale. Although the water depth at Calera was probably greater than at Pratt's Ferry, the absence of features indicating turbitity currents and the basic lithologie similarities between the base of the Athens Shale at the two sections suggest that water depth was not exceptionally greater at Calera. Clay was carried
into the basin of deposition from the east and lime mud was carried into
the basin from the west. BIOSTRATIGRAPHY
Graptolite Biostratigraphy In Alabama
Text figures 8 and 9 show the range of graptolite species within the
Athens Shale in the Pratt's Ferry, Pratt's Syncline, and Calera sections.
Because of geographic proximity and similarities in stratigraphy and faunal composition, data from the Pratt's Ferry section and the Pratt's
Syncline sections have been combined into one range chart (Text-fig. 8 ).
In the following discussions, reference to these two sections are often combined, in which cases the designation "Pratt's Ferry-Pratt's Syncline" is used.
In the Pratt's Ferry and Pratt's Syncline sections, the graptolite fauna is nearly uniform through almost the entire thickness of the
Athens Shale. Most of the species that make up the fauna first appear within a very limited stratigraphie interval close to the base of the
Athens Shale and most of these extend nearly to the top of the. Athens
Shale. These relatively long-ranging species, which numerically dominate the Pratt's Ferry-Pratt's Syncline section are: Didymograptus sp..
Reteograptus geinitzianus, Nemagraptus gracilis, Jicellograptus alaba- mensis, bispiralis bispiralis, D. gurleyi gurleyi, sextans,
Glossograptus ciliatus, Cryptograptus marcidus, Glyptograptus sp. cf. G^. teretinscuius, G. euglyphus, Climacograptus meridionalis, and
Pseudoclimacograptus modestus.
44 45
EXPLANATION OF TEXT-FIGURE 9
Stratigraphie distribution of graptolites in the Calera section. g Lenoir- Athens -— Frog Mtn rtI iH* - h 'ë Collections M (D Pseudocl. sp. cf. P. eurystomarn VO
Pseudocl. angulatus angulatus*
Lasio. sp.i
G lypt. sp. cf. G. teretiusculus*
Glypt. euglyphus*
Didymo. sp *
Crypto, marcidus*
Reteo. geinitzianus*
Glosso. cUiatus'
Dicell. genicuiatus
Dicell. gurleyi n. ssp. A
Nema. gracilis
A m phig. n. sp. B Dicell. bispiralis n. ssp. A "O Climaco. meridionalis
Dicell. sextans
Dicell. gurleyi gurleyi
Azygo. incurvusm Dicrano. irregularis ' P t e r o' sp.i Apoglosso. lyra* <- Pseudocl. modestus* P. s e rru s------+ — P. anserinus Lepto. trentonensis {• L— N. gracilis Z on e 3 0) 9^7 47 Leptograptus trentonensis, Orthograptus sp. , and Dicaulograptus? n. sp. A have limited stratigraphie ranges. The former two are restricted to the upper part of the Pratt's Syncline section,and the latter one is restricted to the lower part of the Pratt's Ferry and Pratt's Syncline sections. In addition, one specimen of Dicranograptus irregularis was collected from a horizon in the lower part of the Athens Shale near the top of the Pratt's Ferry section. Dictyonema sp., Amphig rap tus n. sp. A, and Nemagraptid sp. A have rather restricted stratigraphie ranges, but specimens are rare and the species appear to be of little stratigraphie value. The graptolite fauna of the Calera section, in contrast to that of the Pratt's Ferry and Pratt's Syncline sections, is not uniform throughout the Athens Shale (text-fig. 9). Between 4.5 and 9.0 m above the base of the section, the fauna consists of Pseudoclimaco graptus sp. cf. 2" eurystoma, 2* angulatus angulatus, Lasiograptus sp., Glyptograptus sp. cf. 2- teretiusculus, 2* euglyphus, Di dymog rap tus sp., Cryptograptus marcidus, Glossograptus ciliatus, and Reteograptus geinitzianus. Several of these species extend through most of the section, but some are restricted to the lower half of the section. Beginning at 9.0 m, there is a rapid appearance in the fauna of several additional species, primarily dicellograptids. In their order of appearance, these species are: Dicellograptus genicuiatus, D. gurleyi n. ssp. A, Nemagraptus gracilis, Amphigraptus n. sp. B, and Dicellograptus bispiralis n. ssp. A. Except for 2 - gracilis, all of these species have very limited stratigraphie ranges. Between 11.9 and 17.4 m, the following species, listed 48 in their order of appearance, appear: Climacograptus meridionalis, Dicellograptus gurleyi gurleyi, D. sextans, Azygograptus incurvus, and Dicranograptus irregularis. All these species range up to, or close to, the top of the Athens Shale at Calera. At 17.4 m, a single specimen of Pterograptus sp. was collected. Apoglossograptus lyra first appears at 28 m. Pseudoclimacograptus modestus first appears at 33 m, and Leptograptus trentonensis occurs only at the top of the section. Definite faunal differences within the Athens Shale exist between the Pratt's Ferry-Pratt's Syncline section and the Calera section. Dicellograptus alabamensis, D. bispiralis bispiralis, Dicaulograptus? n. sp. A, Orthograntus sp., Amphigraptus sp. A, and Nemagraptid sp. A are restricted to the Pratt's Ferry-Pratt's Syncline section, whereas Pseudoclimacograptus sp. cf. 2» eurystoma, 2* angulatus angulatus, Lasiograptus sp., Dicellograptus genicuiatus, D. gurleyi n. ssp. A, 2- bispiralis n. ssp. A, Amphigraptus n. sp. B, Azygograptus incurvus, Pterograptus sp., and Apoglossograptus lyra are restricted to the Calera section. Many of these faunal differ ences appear to be due to significant age differences between portions of the Athens Shale in the Pratt's Ferry-Pratt's Syncline section and the Athens Shale of the Calera section. Several of the species that appear together at the base of the Athens Shale in the Pratt's Ferry-Pratt's Syncline section are also present in the Calera section. However, at Calera these species occur together only in those parts of the section that are stratigraphically above the first appearance of Pseudoclimacograptus modestus. 49 Leptograptus trentonensis occurs only at the very top of the Calera section, whereas it ranges through the upper half of the Pratt's Syncline section. Orthograptus sp. occurs only at the very top of the Pratt's Syncline section, and it is absent fron the Calera section. In addition, certain morphologic features in Dicellograptus gurleyi gurleyi and sextans, which occur at both the Pratt's Ferry-Pratt's Syncline and Calera sections, vary stratigraphically (See Remarks under descriptions of those species), and a comparison of this variation in the two sections suggests that a horizon approximately 65 m above the base of the Athens Shale at Calera correlates with a horizon 15 to 20 m above the base of the Athens Shale at Pratt's Ferry and Pratt's Syncline. Thus, it appears that the base of the Athens Shale at Pratt's Ferry correlates with a horizon high up in the Athens Shale at Calera. It also appears that the upper part of the Athens Shale at Pratt's Syncline is younger than the top of the Athens Shale at Calera (Text-fig. 10). Additional evidence suggesting the diachronous nature of the Athens Shale is provided by conodonts. At Pratt's Ferry, Bergstrom (1971) placed the boundary between the conodont zones of Pygodus serrus and Pygodus anserinus within the Pratt Ferry beds 0.3 m below the base of the Athens Shale. I was able to locate this boundary in the Calera section within the interval between 4.0 and 5.2 m above the base of the Athens Shale. The different positions of this zonal boundary in the Pratt's Ferry- Pratt's Syncline and Calera sections also illustrate the diachronous and transgressive nature of the base of the Athens Shale. 50 EXPLANATION OF TEXT-FIGURE 10 Inferred relations between the Pratt's Ferry-Pratt's Syncline section and the Calera section. (1) connects levels representing the boundary between the zones of Pygodus serrus and Pygodus anserinus; (2) the levels of first appearance of Pseudoclimacograptus modestus ; (3) levels of similar morphology in Dicellograptus gurleyi gurleyi and D. sextans (See Text-figs. 27-30); (4) the level of the first appearance of Leptograptus trentonensis in the Calera section with the theoretically possible first appearance of that species at Pratt's Ferry-Pratt's Syncline section; and (5) the level of the first appearance of I., trentonensis in the Calera section with the observed first appearance of that species at Pratt's Ferry-Pratt's Syncline section- Wavy line represents erosional unconformity at top of Athens Shale. PRATT S FERRY BIrmlngl PRATT S SYNCLINE '6 mil*#i CALERA ATHENS SHALE Ui Text-figure 10 52 The boundary between the Pygodus serrus and Pygodus anserinus Zones is defined as the level at which 2- serrus evolved into 2' anserinus. It has been recognized throughout much of northwestern Europe and eastern North America (Bergstr8m, 1971). For these reasons, it has been used as a datum for construction of the correlation chart for the Pratt's Ferry-Pratt's Syncline and Calera sections (Text-fig. 10). Correlations of the first appearance of Pseudoclimaco graptus modestus and Leptograptus trentonensis, and the morphological variation in Dicellograptus gurleyi gurleyi and 2* sextans are also shown on this chart. Two lines connect the two possible first appearances of Leptograptus trentonensis. The higher one connects the obseryed first appearance. The lower one connects the theoretically possible lowest occurrence. Comparison of the rock thicknesses between the graptolite and conodont correlation lines indicates that the net rate of deposition at Calera was greater than at Pratt's Ferry- Pratt's Syncline, especially in light of the fact that the mud deposited in the Calera area is likely to have suffered much more compaction during diagenesis and lithification than the carbonate-rich muds in the Pratt's Ferry area. In addition, the correlation lines for L. trentonensis, together with the presence of Orthograptus sp. at the top of the Pratt's Syncline section and its absence at the Calera section, indicate that the Athens Shale at the Pratt's Ferry-Pratt's Syncline section ranges higher stratigraphically than at Calera. This difference is probably due to relatively greater erosion of the Athens Shale at Calera before deposition of the Devonian Frog Mountain Sandstone. 53 It might be argued that the gradual appearance of new species in the lower part of the Athens Shale at Calera is due to shifting environmental conditions (i.e. a transgressing shoreline). If this was true, then a gradual appearance of species similar to that of Calera (biserial graptolites followed by N. gracilis and dicellograp tids) should also be present in the Pratt's Ferry-Pratt's Syncline sections. However, this is definitely not the case in those sections. The first appearance of species at Calera occurs within an interval that is continuously graptolitic, and the first appearance of graptolite taxa may be due to local evolution or to migration into the Southern Appalachian basin. The species appearing at the base of the Athens Shale at the Pratt's Ferry-Pratt's Syncline section were probably present within the Southern Appalachian basin before deposition of the Athens Shale at those localities, and their appearance in the Pratt's Ferry-Pratt's Syncline sections is probably associated with the transgression that resulted in the deposition of the Athens Shale. In spite of slight faunal differences, differing rates of deposition, and somewhat imprecise correlations, the range charts for the Pratt's Ferry-Pratt's Syncline and Calera sections can be combined in such a way that they together illustrate a graptolite faunal succession that spans the time interval during which the Athens Shale was deposited at both the Calera and Pratt's Ferry- Pratt's Syncline sections (Text-fig. 11). The most prominent event within this succession occurs at a horizon 10.5 m above the base of the Athens Shale at Calera. Below this horizon, the graptolite fauna 54 EXPLANATION OF TEXT-FIGURE 11 Correlation of graptolite biostratigraphy of Athens Shale in Alabama with that at other localities. The left-hand column is a simplified biostratigraphic scheme of the Athens Shale based on the Pratt's Ferry-Pratt's Syncline section and the Calera section. (1) represents long-ranging biserial graptolites; (2) represents Nemagraptus gracilis. (3) represents species of Dicellograptus; (4) represents Leptograptus trentonensis; (5) represents Orthograptus sp. southern O slo, southern A la b a m a New York Texas Wales Australia Sweden N o rw ay S co tlan d gracilis b ico rn is g racilis g rac ilis g rac ilis gracilis gracilis g ra c ilis I 2 3 (9 t e r e t i t e r e t i t e r e ti t e r e ti te re ti usculus usculus usculus u sculus uscuius p_ J— p — m u rc h i - m u rc h i- murchisoni tereti soni son i usculus Text-figure 11 Ln 56 is uniform and consists dominantly of biserial graptolites. Above this horizon, several new species, mainly dicellograptids but including Nemagraptus gracilis, appear. The fauna above the 10.5-metre horizon in the Calera section, and the entire fauna from the Pratt's Ferry- Pratt's Syncline section is interpreted below as representing the Zone of N. gracilis. This fauna gradually appears through a very limited stratigraphie interval within the lower part of the Athens Shale at Calera. The 10.5-m horizon has been picked as the base of this fauna or as the base of the gracilis Zone. It can be defined on the first appearance of gracilis together with species of Dicellograptus. The fauna below the 10.5-m horizon is referred herein to the Zone of Glyptograptus sp. cf. G^. teretiusculus. Although all the species in this fauna extend upward into the N. gracilis Zone, below the 10.5-m horizon they occur to the exclusion of all other species. Concepts of the Zones of N. gracilis and G. cf. G^. teretiusculus Introduction Early Ordovician (Arenig through Llanvim) graptolite faunas display marked provincialism with regard to intercontinental correlations. Beginning with the Llandeilo and continuing through the Middle and Late Ordovician, graptolite faunas took on a more cosmopolitan aspect. As a result, the species Nemagraptus gracilis, and the gracilis Zone, have been recognized in, and used for correlation between, all the continents except Africa and Antarctica. However, there are few places in the world where the base of the N. gracilis Zone can be defined in a continuously graptolitiferous 57 sequence. In the "classical" European sequences of southern Sweden and southern Scotland and in the Hudson River Valley and the Marathon, Texas, areas of North America, the base of the N. gracilis Zone is not exposed, is marked by an unconformity, or occurs in structurally complex and inadequately sampled and/or poorly graptolitiferous sequences. The Calera section described above is exceptional because it is the only one known in North America in which the base of the N. gracilis Zone occurs in a continuously graptolitiferous sequence. However, in order to justify the use herein of the designation "N. gracilis Zone" and in order to correlate the base of this zone, as it is defined herein, with contemporaneous graptolitiferous sequences in other parts of the world, it is necessary to review the concepts on which the N. gracilis Zone and the underlying G^. teretiusculus Zone were previously based. This review is organized below in terms of geographic regions. Great Britain Lapworth (1880, p. 198) was the first to erect a "Zone of Coenograptus gracilis." Besides N. gracilis, Lapworth's zone was characterized by abundant dicranograptids; in fact, he referred to it as "the first of the Dicellograptidian zones," and suggested that it might also be called the Zone of Dicellograptus sextans. According to Lapworth, the typical development of this zone is in the lower portion of the Glenkiln Shales in the Southern Uplands of Scotland, but he also recognized it near Builth in southeastern Wales. Unfortunately, the base of Lapworth's N. gracilis Zone is not exposed in the Glenkiln Shales of Scotland, and Lapworth 58 considered the murchisoni Zone with its greatly different fauna and with its typical development in Wales to be the next underlying zone. As Skevington (1969) has described so adequately, Elies and Wood (1918) and Elies (1922, 1925, 1939) recognized an intervening zone, the Zone of Glyptograptus teretiusculus, between the murchisoni and gracilis Zones. This zone was characterized by an abundance of G^. teretiusculus, other species of Glyptograptus, and species of Climacograptus, Diplograptus, and Amplexograptus, and it was distinguished from the overlying gracilis Zone by an absence of species of Nemagraptus, Leptograptus, Dicellograptus, and Hallograptus. In the Builth area, where Elies (1939) described her zone, Lapworth (1880) reported that both of his zones of N. gracilis and murchisoni were represented. Exposures in this area are poor, there are large distances between sampled horizons, and the fauna that Elies (1939) actually collected and reported from her Zone of teretiusculus consisted of several species of Nemagraptus and Dicellograptus, as well as of the biserial graptolites that she reported in 1925. Thus, her (1939) teretiusculus Zone differed from Lapworth's gracilis Zone only by the absence of gracilis, and it was inconsistent with her earlier definition (Elies, 1925). Because of these ambiguities, Skevington (1969) doubted the existence of a teretiusculus Zone in Wales and proposed that the gracilis Zone should be taken as following directly on the murchisoni Zone and should be defined as characterized by the appearance of species of Nemagraptus and Dicellograptus to be joined later by Leptograptus and Dicranograptus. Southern Sweden Tullberg (1882) was the first to recognize a Zone of N. gracilis in Sweden. Below the gracilis Zone, Tullberg erected two additional zones, which are known in the recent literature as, in ascending order, the Zone of Glossograptus hincksi or Gymnograptus linarssoni and the Zone of Climacograptus putillus. Jaanusson and Strachan (1954), and Jaanusson (1960), included these two zones as subzones within a single Zone of Glyptograptus teretiusculus. The teretlusculus- gracilis sequence has been sampled very well in boring cores (Hede, 1951; Nilsson, I960, in prep.); however, in the key sections in Scania, Southern Sweden, an unconformity, or at least a break in the graptolite sequence that is represented by a bed of phosphorite, separates the teretiusculus and gracilis Zones. As it is defined in Sweden, the teretiusculus Zone, which overlies a Zone of jD. murchisoni, is remarkably similar to Elies (1925) concept of that zone because it contains an abundance of species representing Glyptograptus, Climacograptus, Diplograptus, and Amplexograptus. However, Dicellograptus, represented by D. vagus, Nemagraptus, represented by N. subtilis, and Dicranograptus, represented by D. irregularis, also range through most of the teretiusculus Zone in Southern Sweden. Above the overlying bed of phosphorite, gracilis appears with several species of Dicellograptus, Leptograptus, Hallograptus, Dicranograptus, and Corynoides, and thus marks the base of the Zone of N. gracilis. Oslo Region, Norway The fauna of the teretiusculus Zone as defined by Berry (1964) as a unit in the top most part of the Ogygiocaris Series is very 60 similar in terms of graptolite taxa at the specific level to that of the teretiusculus Zone of Southern Sweden. Although Nemagraptus subtilis is absent in the Oslo Region, Dicellograptus vagus and Dicranograptus irregularis range through most of the Oslo Region teretiusculus Zone. A graptolitic facies is not present right above strata of the Ogygiocaris Series, and thus the top of the teretiusculus Zone can not be determined. Australia The Australian graptolite sequence is characterized by "narrow bands of graptolitic shales .... separated by thick sequences of unfossiliferous rocks" (Webby, 1976). The teretiusculus Zone is known from a single horizon (loc. Ba 67 of Harris and Crawford, 1921), which contains a fauna rich in biserial graptolites. From a horizon 30 feet higher than the one representing the teretiusculus Zone, Thomas and Keble (1933) reported species of Dicellograptus and Dieranograptus together with biserial graptolites of the teretiusculus Zone. From another horizon (Evan Creek) approximately 5 miles (8km) from locality Ba 67, an assemblage has been reported that contains N. gracilis and Climacograptus bicornis. Thomas (1960) assigned the assemblage at loc. Ba 67 with Dicellograptus and Dicranograptus and the Evans Creek assemblage to the Zone of N. gracilis. He (1960, p. 62) based the gracilis Zone on the "incoming of Nemagraptus, Dicellograptus sextans, D. divaricatus, Dicranograptus nicholsoni, Dicr. zic-zac, Climacograptus bicornis, Hallograptus mucronatus, and the various Leptograpti. " 61 New York State Ruedemann (1947) defined the N. gracilis Zone on the basis of the fauna contained in the Mount Merino shale and chert member of the Normanskill Shale. So defined, the gracilis Zone contains an abundance of species and subspecies with many representing Dicellograptus. Dicranograptus, Leptograptus. Nemagraptus. Glyptograptus, Orthograptus, and Climacograptus. However, Ruedemann (1947, p. 72) failed "to find a continuous sequence in the thick but much folded and faulted Normanskill beds of the Hudson Valley... ." After his work on the Marathon area. Berry (1962) studied the Normanskill Shales and assigned them to the uppermost part of the gracilis Zone and to his (Berry, 1960) overlying Zone of C^. bicornis. Riva (1974) equated Berry's (1960, 1962) gracilis and bicornis Zones with a single Zone of N. gracilis. Marathon Region, Texas From the uppermost part of the Fort Pe'na Formation through the overlying Woods Hollow Shale, Berry (1960) established, in ascending order, the following zones: Glyptograptus sp. cf. teretiusculus, Nemagraptus gracilis, and Climacograptus bicornis. Unfortunately, the Marathon section is extensively faulted and exposures are poor. In addition. Berry (1960, p. 24) reports the graptolites to be poorly preserved. The teretiusculus-gracilis part of the sequence is pieced together from numerous geographically scattered collection localities and from three measured sections that are separated by as much as eight miles (13 km). The boundary between his teretiusculus 62 and gracilis Zones occurs in only one of his measured sections (Section XI) and, in that section, the lowest collection representing the gracilis Zone is stratigraphically 53 feet (16 m) above the high est collection representing the teretiuscuius Zone. Berry defined the G. sp. cf. iG. teretiusculus Zone as characterized by the widespread occurrence of G^. sp. cf. £. teretiusculus, Amplexograptus confertus, and Climacograptus riddellensis. Berry's gracilis zone is marked by the entrance of N. gracilis together with several species of Dicellograptus, Dicranograptus, Leptograptus, Hallograptus, and Corynoides. The biserial genera Climacograptus and Glyptograptus are represented by more species than are present in the underlying teretiusculus Zone. The presence of Climacograptus bicornis and species of Orthograptus distinguishes Berry's bicornis Zone from the underlying gracilis Zone. Correlation Southern Appalachians Besides the sections described herein, the only other section in the southern Appalachians that has so far proved to be of significance to the Alabama graptolite biostratigraphy is the one at Clifton Heights, one mile (1.6 km) south of Bristol, Tennessee (Text-fig. 1). From this section collections were made in the 1920's by geologists from Harvard University, and these collections were tied dotm stratigraphically. Decker (1952) identified the specimens in these collections, and the writer has had the opportunity to re-examine the 63 collections. According to the writer's identifications, these collections include the following taxa: 1) From a horizon 3 ft. above the base of the Athens Shale (Paperville Formation of Cooper and Cooper, 1946): Didymograptus sp. Glyptograptus sp. cf. £. teretiusculus G. euglyphus Pseudoclimacograptus sp. cf. 2- angulatus Glossograptus ciliatus Cryptograptus marcidus 2) From a horizon 6 ft. above the base of the Athens Shale: Didymograptus sp. Glyptograptus sp. cf. G^. teretiusculus G^. euglyphus Pseudoclimacograptus angulatus Glossograptus ciliatus Cryptograptus marcidus Reteograptus geinitzianus 3) From a horizon 10 ft. above the base of the Athens Shale : Didymograptus sp. Glyptograptus euglyphus Pseudoclimacograptus angulatus Glossograptus ciliatus Dicellograptus geniculatus 4) From a horizon 15 ft. above the base of the Athens Shale: Didymograptus sp. Glyptograptus euglyphus Cryptograptus marcidus 5) From a horizon 25 ft. above the base of the Athens Shale : Didymograptus sp. Glyptograptus euglyphus Glyptograptus sp. 64 6) From a horizon 50 ft. above the base of the Athens Shale: Didymograptus sp. Glyptograptus sp. cf. G^. teretiusculus G^. euglyphus Pseudoclimacograptus angulatus Glossograptus ciliatus Cryptograptus marcidus Reteograptus geinitzianus Nemagraptus gracilis Dicellograptus gurleyi gurleyi Dicranograptus sp. cf. D. irregularis 7) From a horizon 200 ft. above the base of the Athens Shale : Didymograptus sp. Glyptograptus euglyphus Dicellograptus alabamensis 2 " gurleyi gurleyi sextans A single specimen of Nemagraptus gracilis is present in the collection from 3 feet above the base of the Athens Shale. However, the shale chip on which this specimen is situated is a hard, black, fissle shale, whereas the other shale chips in this collection are light brown in color and crumble easily. I believe that this specimen of gracilis was collected from "float" and not from the designated horizon. The stratigraphie distribution of the graptolite fauna at the Clifton Heights section is similar to that at Calera. Biserial graptolites compose the fauna from the lower horizons, and nema- graptids and dicellograptids occur at higher horizons. In spite of the lack of detailed bed-by-bed collections from several measured sections throughout the southern Appalachians, the Clifton Heights section supports the biostratigraphic scheme established in Alabama. 65 Marathon Region, Texas The 10.5-metre horizon at Calera correlates with a horizon somewhere in the interval between the highest graptolite collection in Berry's (1960) G^. sp. cf. G^. teretiusculus Zone and the lowest collection in his N. gracilis Zone (Text-fig. 11). The fauna below the 10.5-metre horizon at Calera is similar to that reported for the G^. sp. cf. teretiusculus Zone, and the fauna above the 10.5-metre horizon is similar to that reported from the N. gracilis Zone. The high first occurrence of Leptograptus trentonensis and Orthograptus sp. and the absence of Corynoides and Hallograptus in the Athens Shale supports Berry's (1960) recognition of a Zone of C^. bicornis in the upper part of what other workers consider to be the Zone of ÎI. gracilis. New York State The graptolite fauna of the Normanskill Shale is much more varied than that of the Athens Shale. In addition to many of the species from the Athens Shale, it contains species of Dicranograptus with large biserial portions, Orthograptus calcaratus (several varieties), Climacograptus bicornis, C^. bicornis tridentatus, brevis strictus, Lasiograptus putillus, Corynoides pristinus, and C^. calicularis. Ruedemann (1947) suggested that the faunal differences between the Athens Shale and the Normanskill Shale represented faunal provincialism. The faunal differences are interpreted herein to represent temporal differences for the following reasons: 1) The species that are restricted to the lower parts of the Athens Shale do not occur in the Normanskill Shale. 66 2) The long-ranging species that occur in the Athens Shale also occur in the Normanskill Shale. 3) Leptograptus trentonensis and Orthograptus sp., which occur at only the highest horizons of the Athens Shale occur in abundance throughout the Normanskill Shale. Thus, the oldest beds in the Normanskill Shale probably correlate with a horizon very high in the Athens Shale (Text-fig. 11). The faunal differences between the Athens Shale and the Normanskill Shale suggest that the gracilis Zone can be subdivided into two subzones, a lower subzone that is presently un-named and an upper subzone to which the name C^. bicornis can be applied and which is characterized by the entrance of Leptograptus, Orthograptus, Corynoides, and climacograptids of the bicornis and brevis type. Southern Sweden The 10.5-metre horizon in the Athens Shale at Calera correlates well with the base of the N. gracilis Zone in Southern Sweden (Text- fig. 11). Because the bed of phosphorite in Sweden may mark an unconformity, or at least a short pause in the graptolite record, the top of the teretiusculus Zone in Sweden is probably not equivalent to the 10.5-metre horizon at Calera. The teretiusculus Zone fauna in Sweden differs primarily from the fauna below the 10.5-metre horizon at Calera by the presence of Nemagraptus subtilis, Dicellograptus vagus, and Dicranograptus irregularis. However, as discussed below (See SYSTEMATICS), IJ. subtilis is probably an ancestor of _N. gracilis, and Dicellograptus 67 vasus is related closely to jD. gurleyi n. ssp. A and 2» gurleyi gurleyi. Thus, although the earliest representatives of Nemagraptus and Dicellograptus evolved before the time represented by the base of the 2" gracilis Zone in Sweden and Alabama, they and/or their descendants did not migrate to the area of the southern Appalachians until the time represented by the basal most portion of the N. gracilis Zone. Several species at Calera, namely Azygograptus incurvus, Dicellograptus geniculatus, Dicranograptus irregularis, and Pterograptus sp., are found only in strata older than the gracilis Zone in Sweden. Their presence in gracilis Zone strata in the southern Appalachians indicates that these species were able to exist longer in the Southern Appalachians, than in Sweden. Oslo Region, Norway The top of the teretiusculus Zone in Norway probably correlates with a horizon well below the 10.5-metre horizon at Calera (Text-fig.11) Southern Scotland The fauna of the Glenkiln Shale, which contains dicranograptids with long biserial portions, Hallograptus, and Climacograptus bicornis probably correlates with strata no older than high portions of the Athens Shale at Calera (Text-fig. 11). Wales As defined by Skevington (1969), the base of the N. gracilis Zone is recognized by the entrance of species of Nemagraptus (not necessarily 2* gracilis) and Dicellograptus. Thus, the Welsh gracilis 6 8 Zone may correlate with a low portion of the teretiusculus Zone in Sweden and, in turn, with an interval well below the 10.5 metre horizon at Calera (Text-fig. 11). Toghill (1970b) reports a fauna from the Rendre Shales (Llandeilo) of the Mydrim area in southern Wales. This fauna includes several biserial graptolites together with Dicranograptus sp. (with a short biserial proximal end), Dicellograptus sp. cf. D. vagus, and Isograptus sp. Toghill considered this fauna to represent the teretiusculus Zone, and he suggested that the base of the gracilis Zone be placed at the first appearance of N. gracilis. As re-defined by Toghill, the base of the gracilis Zone would correlate with the 10.5-metre horizon at Calera (Text-fig. 11). Australia As presently defined on the first appearance of Dicellograptus and Dicranograptus, the base of the gracilis Zone in Australia correlates with a horizon within the teretiusculus Zone in Sweden and thus with a horizon below the 10.5-metre horizon at Calera (Text-fig. 11). Summary and Conclusions The 10.5-metre horizon in the section at Calera marks a major change in the graptolite fauna of the Athens Shale. Below this horizon the fauna consists primarily of long-ranging biserial graptolites. Above this horizon N. gracilis appears together with several species of Dicellograptus, and these species of Dicellograptus appear to represent an early radiation of that genus into a large 69 number of species. The fauna above the 10.5-metre horizon is very similar to faunas in Middle Ordovician shale throughout the Southern Appalachians, and it is also similar at both the generic and specific levels to faunas throughout the world that have been placed in the Zone of N. gracilis. Yet, it lacks many of the species and several of the genera that occur in what are here interpreted to be higher strata in the Zone of lî. gracilis. In addition, the fauna below the 10.5-metre horizon, which consists primarily of long-ranging species, but lacks species characteristic of the gracilis Zone, is similar to pre-gracilis Zone faunas that have been referred to the teretiusculus Zone in other parts of the world. Therefore, the 10.5-metre horizon in the Calera section is picked as the base of the N. gracilis Zone. Because this horizon occurs in a well-exposed, well-collected and continuously graptolitiferous sequence, it would be appropriate to select the Calera section as a stratotype for the base of the N. gracilis Zone in North America. This base of the gracilis Zone can be accurately correlated with the base of the gracilis Zone in Sweden and with the base of the gracilis Zone in Wales, as re-defined by Toghill (1970b). Faunal differences between the Athens Shale and the Normanskill Shale suggest that the N. gracilis Zone in North America can be subdivided into two zones or subzones with the term "C. bicornis" applied to the upper subdivision. SYSTEMATIC PALEONTOLOGY Classification Classification within the Order Graptoloidea has been the subject of extensive discussions in the recent graptolite literature, and except in certain instances the classification of Bulman (1970) is followed in the present study. Three notable departures from Bulman's classification in the present study are the inclusion of Corynoididae in Glossograptina, the transfer of Leptograptus from Nemagraptidae to Dicranograptidae, and the transfer of Reteograptus from Retiolitidae to Abrograptidae. The ba sis for each of these revisions is discussed under the family or genera of concern. Serious problems in classification are encountered in the diplograptids. As documented below, Climacograptus brevis probably evolved directly from Glyptograptus euglyphus, whereas C^. meridionalis is very closely related to Pseudoclimacograptus modes tus. These two relationships not only illus trate the polyphyletic origins of generic level taxa in the diplograptids but also raise serious questions about the significance of the diagnostic generic characters. Thecal characters serve as the diagnostic features for diplograptid genera, yet the entire range of variation in the thecal characters expressed by all the genera of the Diplograptidae is no greater than that described below in species within the single genus Dicellograp tus. The diplograptid material that is available in the present study is too limited to investigate in depth the taxonomic relationships within the Diplograptidae. Therefore, Bulman's (1970) classification is employed herein. 70 71 The study of isolated material has provided valuable information about astogeny (proximal-end development). More significantly, it has provided significant new information on thecal ontogeny. Together, astogeny and onto geny have been employed wherever possible to unravel taxonomic relations (i.e., in the Corynoididae and Leptograptus). In addition, isolated spec imens help to reveal the original shape of complex rhabdosomes, and in some instances (e.g. Nemagraptus gracilis) specimens that present very different rhabdosomal appearances when they are compressed on bedding surfaces, and which have been distinguished previously as distinct taxa are shown to represent only different preservational aspects of the rhabdo somes of one and the same species. Terminology The terminology employed in the present study is taken mainly from Bulman (1970). However, a definite need arose for additional terminology, most of which was derived from previous graptolite studies. Except for terms defined by Jaanusson (1960), these additional terms are defined below and, in some cases, illustrated in Text-figures 12 and 13. axillary angle: angle enclosed by dorsal margins of proximal parts of stipes. Measured in lateral view, geniculum: See Jaanusson (1960); -ar, adjective ending; -a, plural ending. genicular angle: angle defined, in lateral view, by distal part of infra- genicular wall and proximal part of supragenicular wall, infragenicular wall: See Jaanusson (1960). isolated: adjective designating specimens that have been removed from the enclosing matrix by means of acid treatment. 72 12 8 TEXT-FIGURE 12. Measured distances in dicellograptid specimens. Figures denote the following: (1) Thecal density. Number of thecae per number of millimeters, measured between genicula; proximal thecal density measured from th 2 geniculum. (2) Stipe width. Measured transversely at level of greatest curvature of supragenicular wall, which is interpreted as the level of the maximum stipe width. (3) Axillary angle. The angle enclosed by the dorsal margins of the proximal parts of the stipes as seen in lateral view. (4) Genicular angle. The angle, measured in lateral view, between the distal part of the infragenicular wall and the proximal part of the supragenicular wall. (5) Length of apertural excavation. (6) Width of apertural excavation. 2 (7) Stipe no. 2. The stipe that develops from th 1 . Directions relative to stipe no. 1: (8) distal, (9) proximal, (10) dorsal, (11) ventral, (12) dorso-distal, (13) ventro-distal, (14) ventro-proximal, (15) dorso-proximal. 73 TEXT-FIGURE 13: Measured distances in diplograptid specimens. Figures denote the following. (1) Thecal density. Number of thecae per number of millimeters, measured between ventral tips of thecal apertures ; proximal thecal density measured from th 2 aperture in non-geniculate thecae, from th 2 geniculum in geniculate thecae. (2) Stipe width. Measured transversely between ventral tips of apertures of corresponding thecae of opposing thecal series. (3) Width of apertural excavation. 74 longitudinal: indicates direction parallel to the axis of the sicula or stipe. Abbreviated long. 2 2 2 stipe no. 2: the stipe that is composed of th 1 , th 2 , ..., th n . supragenicular wall: See Jaanusson (1960). Thecal pair, pair of thecae: term used to refer to two thecae, each of which is at the same level on each of two stipes, 12 12 e.g. th 1 and th 1 , ..... th n and th n transverse: indicates direction that is perpendicular to the axis of the stipe or sicula. Abbreviated trans. th n, th n+1, th n-1: See Skevington (1965); Let th n represent any particular theca on a stipe, then th n-1 denotes the preceding theca, and th n+1 denotes the succeeding theca. Text-figure Abbreviations Numerous abbreviations are used in the text-figures in order to indicate particular morphologic features of specimens. These abbre viations are listed below. In addition, the abbreviations, PF, PS, C, and C-25, are used in the Text-figure explanations and Appendix A to denote samples. These abbreviations stand for the Pratt’s Ferry, Pratt's Syncline, Calera, and Calers (Highway 23) sections, respectively. Where used, they are followed by a number that indicates the strati graphie distance in metres that the sample is located above the base of the section, except for the Calera (Highway 25) section in which case the number indicates the stratigraphie distance in metres below the top of the section. 75 af Apertural flange al Apertural list avs Spine on anti-virgellar margin of sicular aperture 1 b Initial bud of th 1 cc Common canal cf Opening at base of cladium into aperture of "mother" thecae cl Cladium cli Connecting list cp Thin strip of periderm that restricts the right-lateral corner of the thecal aperture and serves as the base of a cladium. ct Cortical tissue g Geniculum gf Genicular flange is Interthecal septum iw Infragenicular wall li Sicular list structure Is Lateral spines: sls-sicular lateral spines; th Is-thecal lateral spines. Iw Lateral wall ms Metasicula msp Median septum mt Metatheca pb Prothecal base pbl List at aperture of prothecal base pf Prothecal fold 76 pi Parietal list pn Primary notch ps Prosicula pt Protheca 1 rf Resorption foramen in sicula for th 1 s Sicula sa Sicular aperture sg Septal groove si Septal list sp Periderm of median septum S t 1 Stipe n o . 1 S t 2 Stipe no. 2 sw Supragenicular wall ta Thecal aperture tr Transverse rod V Virgella vi Virgula vl Ventral list vp Ventral process Storage of Specimens The specimens illustrated in the present study are assigned the numbers OSU 32912 to OSU 33348, and they are stored in the Orton Geological Museum at The Ohio State University, Columbus, Ohio. Nu merous specimens stored at other museums are referred to in the text, generally by the museum number, which is prefixed as folloifs: AMNH, American Museum of Natural History, New York; USNM, United States 77 National Museum of Natural History, Washington, D.C.; NYSM, New York State Museum, Albany; YPM, Yale-Peabody Museum, New Haven; LO, Museum of the Palaeontological Institute, Lund University; RM, Swedish Museum of Natural History, Stockholm; PMO, Palaeontological Museum, Oslo. Class GRAPTOLITHINA Bronn, 1846 Order GRAPTOLOIDEA Lapworth, 1875 Suborder DIDYMOGRAPTINA Lapworth, 1880, emend. Bulman, 1970 Family DICHOGRAPTIDAE Lapworth, 1873 Genus Pterograptus Holm, 1881 Type Species : Pterograptus elegans Holm, 1881 Diagnosis: As in Bulman (1970). Pterograptus sp. (Text-figure 141) Material The available material consists of one specimen obtained from a horizon 17.4 meters above the base of the Calera section. Description ' The available specimen (Text-fig. 141) is fragmentary; the proximal end is missing. It consists of a 9 mm long primary stipe and eleven lateral branches, which are as long as 6 mm. Successive lateral branches project from alternating sides of the primary stipe at inter vals of 0.6 - 0.9 mm. The primary branch is 0.3 mm wide, and it is oriented with the ventral side down into the shale chip. As a result, 78 79 EXPLANATION OF TEXT-FIGURE 14 Text-figure 14 a-d. Didymograptus sp. Compressed, non-isolated specimens. a. Specimen with didymograptid proximal end. X3.6. C-5.4. OSU 32901. b. Distal stipe fragment. X7.2. PS-117. OSU 32902. c. Specimen with thamnograptid proximal end. X7.2. PF-17.3. OSU 32903. d. Specimen with janograptid proximal end. X7.2. PS-117. OSU 32904. 14 e-k. Azygograptus incurvus EkstrHm. Compressed, non-isolated specimens. e. Distal stipe fragment. Note shape of thecal apertures. Entire specimen X3.6, portion of stipe X15. C-28.3. OSU 32905. f. Note shape of virgella and curvature of stipe. X7.2. C-25-29.5. OSU 32906. g. Note curvature of stipe. X3.6. C-28.3.OSU 32907. h. Note direction of growth of stipe. X7.2. C-30. OSU 32908. i. Note length of sicula and curvature of stipe. X3.6. C-37. OSU 32909. j. Sicula and proximal end of stipe. Note curvature of virgella and origin of first theca. X30. C-39. OSU 32910. k. Sicula and first theca. Note the length of the first theca. X30. C-30. OSU 32911. 14 1. Pterograptus sp. Compressed, non-isolated specimen. Fragment of rhabdosome consisting of one primary stipe with lateral branches. Main stipe in dorsal aspect. X7.2. Portion of specimen. X15. C-17.4. OSU 32912. 80 th 1 Text-figure 14 81 only edges of the thecal apertures are visible projecting laterally out from under the stipe (Text-fig. 141). The maximum width of the primary stipe occurs immediately proximal to a lateral branch (Text- fig. 141) suggesting that the lateral branches project from the primary stipe at the level of thecal apertures on the primary stipe. The lateral branches present a lateral aspect, and the thecae are well displayed. They are of the dichograptid type with an apertural margin oriented perpendicular to the axis of the branch. The thecae number 5 in 3 mm and overlap almost half of their length. The free ventral walls are inclined at angles of less than 20 degrees to the axis of the lateral branch, and the lateral branches are 0.4 - 0.5 mm wide at the level of the thecal apertures. Remarks Although the specimen described above is incomplete, it can be assigned to the genus Pterograptus because of the dichograptid thecae and the alternating lateral branches. A specific assignment is not possible on the basis of the available specimen. Figured Specimen OSU 32912 82 Genus Didymograptus M'Coy ^ Sedgwich and M'Coy, 1851 Type Species: Didymograptus murchisoni Beck ^ Murchison, 1839 Diagnosis As in Bulman (1970) Didymograptus sp. (Text-figures 14a-d) Material More than 1000 non-isolated specimens were obtained from the Pratt’s Ferry, Pratt's Syncline, and Calera sections. Description Straight, uniserial stipe fragments possessing dichograptid thecae are abundant throughout the Athens Shale at Pratt’s Ferry, Pratt’s Syncline, and Calera. The vast majority of these specimens lack proximal ends and represent distal stipe fragments (Text-fig. 14b). These specimens are 1.5 - 1.8 mm wide. The thecae are of the simple dicho graptid type with ventral walls inclined at 40 degrees to the stipe axis and number 10 in 10 mm. Two specimens are preserved with didymograptid proximal ends (Text-fig. 14a). In these specimens the stipes are pendent and enclose an angle of 125 degrees. The stipe width increases from 0.5 mm at th 1 to 0.9 mm at th 5, and the thecae number 9 in 10 mm. These two speci mens are too poorly preserved to determine the detailed morphology of the proximal end. 83 Five specimens exhibit proximal ends in which two stipes with thecae oriented in opposite directions merge without an intervening sicula (Text-fig. 14d). The width and thecal densities of the opposing stipes differ sharply. These specimens resemble Janograptus; however, without the proximal ends, they are no different than the numerous distal stipe fragments described above. In seven specimens, a thin rod extends from the proximal end of the stipe. This rod has regularly distributed hair-like spines (Text-fig. 14c), and it resembles Thamnograptus. Thamnograptus was considered not to be a graptolite by Bulman (1970, p. V139). But here again, if the proximal end was not preserved, the stipe would resemble the numerous stipe fragments described above. Remarks The two specimens with didymograptid proximal ends are too poorly preserved to refer to a particular species. On the basis of their proximal ends, they are assigned to the genus Didymograptus. The specimens with Janograptus proximal ends might represent regenerated fragments, which had broken off from a didymograptid rhabdosome. The specimens with Thamnograptus proximal ends can not be readily explained. In this study, the abundant distal, uniserial stipe fragments with dichograptid thecae and the specimens with didymograptid, janograptid, and thamnograptid proximal ends are referred to collectively as Didymograptus sp. Figured specimens OSU 32901-OSÜ 32904. 84 Genus Azygograptus Nicholson, 1875 Type species: Azygograptus lapworthi Nicholson, 1875 Diagnosis: As in Bulman (1970). tl Azygograptus incurvus Ekstrom, 1937 (Text-figures 14e-k) It 1937 Azygograptus incurvus n. sp., Ekstrom, p. 33-34, PI. 6, figs. 17-20. Type data The holotype, which is stored in the Museum of the Palaeontological Institute, Lund University, Sweden, is the specimen figured by Ekstrom (1937, Plate 6, figure 17). It has not been examined by the writer. Diagnosis A species of Azygograptus with a strongly curved, semi-circular shaped stipe; stipe width 0.15 - 0.20 mm proximally increases distally to 0.6 mm; thecae number 10 in 10 mm proximally, 8 in 10 mm distally. Material More than 250 specimens were obtained from the Calera section. These specimens are preserved as carbon films, and while the outlines of the specimens are distinct, surface features such as interthecal septa can not be discerned. Description The largest available specimen without a proximal end (Text-fig. 14g) is 3 cm long, and the largest specimen with a proximal end (Text- fig. 14i) is 2 cm long. 85 The sicula is 1.6 mm long and 0.18 mm wide at its aperture. The sicular aperture is furnished with a 0.6 - 1.0 mm long virgella, which distally bends in a dorsal direction relative to the sicula (Text-figs. 14c-d). The first theca buds from the sicula at a point 0.3 mm above the sicular aperture. It grows downward along the virgella to a level 0.2 mm below the sicular aperture, where it bends and grows outward and slightly upward. The aperture of the first theca is 1.2 mm from the virgella. In lateral view, the stipe shows strong semi-circular curvature with the ventral wall concave. The upward direction of growth of the stipe, which commences with the first theca, continues for approximately 1.5 cm before the stipe bends and grows downward. Adjacent to the sicula, the stipe is 0.15 - 0.20 mm wide. Distally, it increases in width in such a way that it is 0.20 - 0.25 mm wide at 1 mm from the proximal end, 0.30 mm wide at 3 mm, 0.35 - 0.40 mm at 5 mm, and 0.60 mm at 10 mm. The thecae number 10 in 10 mm proximally and 8 in 10 mm distally. Their free ventral walls are parallel to the dorsal stipe margin. The available specimens are not preserved well enough to reveal the thecal length and thecal overlap. The apertural excavations are semi-circular in lateral view. They occupy a fourth to a third of the stipe width and less than an eighth of the length of the free ventral wall. 86 Remarks The specimens described above agree closely in all diagnostic features with the description and the illustrated specimens of Azygograptus incurvus Ekstrom. Figured specimens OSU 32905 - OSU 32911 87 Family ABROGRAPTIDAE Mu, 1958 Diagnosis Ehabdosome comprising two reclined uniserial or biserial, dipleural stipes; sicula completely sclerotized, but stipe periderm reduced to clathria and, in some instances, cortical membrane; development dicho- graptid'with one or two crossing canals. Discussion , Bulman (1970) assigned the genus Reteograptus to the family Retiolitidae and the subfamily Archiretiolitinae. Many of the genera within Archiretiolitinae are known from isolated material, and these genera show a diplograptid type of sicula in overall proportions T and morphology and very intricate and elaborated clathria. The sicula and the clathria of Reteograptus geinitzianus differ strongly from those of the other genera in Archiretiolitinae. The sicula of Reteograptus geinitzianus is nearly identical in appearance to the sicula of Dinemagraptus warkae KozXowski, 1952, which has been described on the basis of isolated material. The only difference is that the initial bud of th 1 is not sclerotized in D. warkae. Although the described specimens of Abrograptus are not well preserved, their siculae also closely resemble that of R. geinitzianus. The rhabdosome of Dinemagraptus consists of two uniserial, reclined stipes. Each stipe is represented by only a single ventral list to which ring-shaped apertural lists are attached. The manner in which the ventral lists of the stipes attach to the sicula and the apertural 88 lists attach, to the ventral lists of the stipes in Dinemagraptus (Kozïowski, 1952, figs. 1-2) resembles closely that of Reteograptus geinitzianus. The rhabdosome of Abrograptus also consists of two uniserial, reclined stipes, but each stipe consists of two ventral lists (Mu, 1958, fig. 1). The arrangement of the ventral lists and the apertural lists is nearly identical in appearance to that of Reteograptus geinitzianus. A circular list borders the initial bud of th 1^ in Abrograptus. Two lists representing the ventral lists of th 1^ and the first thecal series originate from the proximal and distal ends of the circular list, and 2 1 one of the lists for th 1 arises from one of the ventral lists of th I-*-. Thus, except for the absence of parietal and septal lists, the clathria of Abrograptus is essentially identical to that of R. geinitzianus. Mu and Qiao (1962) established the genus Parabrograptus. and Bulman (1970) synonymized it with Abrograptus. The specimens assigned to Parabro graptus are identical in structure to Abrograptus except that they exhibit two lists that extend distally from the sicula. These specimens differ structurally from R. geinitzianus only in the lack of parietal lists. Mu (1958, p. 265) recognized the similarities between Abrograptus and Reteograptus geinitzianus; however he considered the latter, which was not known at the time on the basis of isolated material, to have a type of proximal-end development that was much more advanced than the dichograptid type of Abrograptus. As described below and discussed 89 above, R. geinitzianus is so remarkably similar structurally to all the species brought to Abrograptidae that it must be closely related to them. Therefore, R. geinitzianus is transferred herein to the family Abrograptidae, and it appears to represent an independent line of evolution culminating in a biserial scandent rhabdosome with a loss of fusellar periderm. The described species of Reteograptus other than R. geinitzianus are too poorly known for their relationship to R. geinitzianus and Abrograptidae to be determined. However, the assignment of the genus Reteograptus to the family Abrograptidae is justified because R. geinitzianus is the type species of that genus. Genus Reteograptus Hall, 1859 Type species: Reteograptus geinitzianus Hall, 1859 Diagnosis Dipleural, biserial rhabdosome consisting of sclerotized sicula, clathria, and cortical periderm; dichograptid development. Reteograptus geinitzianus Hall, 1859 (Text-figures 15, 16) 1859 Reteograptus geinitzianus (n.s.). Hall, p. 518, figure. 1860 Reteograptus Barrandi (n.s.). Hall, p. 61, figure. 1876 Clathrograptus cuneiformis Lapworth, Lapworth, PI. 3, fig. 63. 1908 Retiograptus Geinitzianus Hall, Elies and Wood, p. 316-317, text-figs. 209a-c, PI. 34, figs. 7a-d. 90 EXPLANATION OF TEXT-FIGURE 15 Text-figure 15 a-f. Reteograptus geinitzianus Hall. Isolated specimens. a,b. Obverse and reverse aspects of specimen with cortical membrane covering clathria. X30. PS-109.2. OSU 32913. c,d. Obverse and reverse aspects of specimen consisting of only a clathria. Proximal end broken. Note arrangement of apertural, septal, parietal, and ventral lists. X30. PS-109.2. OSU 32914. e,f. Obverse and reverse aspects of specimen with proximal end. Note secondary thickening of sicula and pieces of cortical tissue attached to lists. X30. PS-109.2. OSU 32915. 91 t h i sa Text-figure 15 9 2 EXPLANATION OF TEXT-FIGURE 16 Text-figure 16 a-j. Reteograptus geinitzianus Hall. Isolated early growth stages. a,b. Reverse and obverse aspects of specimen consisting of sicula, initial bud of th 1^, circular list continuous with virgella, and initial part of ventral lists of th ll. X37.5.' PS-109.2. OSU 32916. c. Reverse aspect. Note development of ventral list of th 1 and appearance of ventral list of th 1^. X37.5. PS-102. OSU 32917. d,e. Obverse and reverse aspects. Note initial bud of th 1^. X37.5. PS-109.2. OSU 32918. f,g. Reverse and obverse aspects. Note the appearance of the first parietal list. X37.5. PS-76.4. OSU 32919. h,i. Obverse and reverse aspects. Note appearance of septal lists and arrangement of first parietal list. X37.5. PS-76.4. OSU 32920. j. Reverse aspect. Note arrangement of ventral, apertural, parietal, and septal lists and length of nema (virgula). X37.5. PS-109.2. OSU 32921. 93 --ms a VI si pi si thi thi 2 ' Text-figure 16 94 1908 Retiograptus geinitzianus Hall, Ruedemann, p. 463-467, text-figs. 444-448, Pl. 29, figs. 5-6; Pl. 31, figs. 9-17. 1947 Retiograptus geinitzianus Hall, Ruedemann, p. 459-460, Pl. 80, figs. 11-26. 1952 Retiograptus geinitzianus (Hall), Decker, Pl. 1, fig. 24; Pl. 2, fig. 86. 1960 Retiograptus geinitzianus Hall, p. 96, Pl. 15, fig. 3a. 1963 Retiograptus geinitzianus J. Hall, Ross and Berry, p. 158, Pl. 13, figs. 20-23. 1964 Retiograptus sp.. Berry, p. 163-164, Pl. 15, fig. 3. Type Data Althou^ Hall (1859) did not designate a holotype for Reteograptus geinitzianus, the specimen figured by him has come to be regarded as the type. This specimen is stored at the American Museumof Natural History, New York. Diagnosis A species of Reteograptus with sclerotized sicula, clathria, and cortical periderm; rhabdosome fusiform with rounded proximal end and parallel ventral walls; maximum width 1.8 - 2.0 mm, excluding apertural lists; clathria consisting of septal, ventral, parietal, and apertural lists; thecae of orthograptid? type, numbering 6 in 3 mm proximally and 4 in 3 mm distally. Material The available material represents the following states of preservation: 1) Isolated specimens. Approximately 100 specimens were obtained from the Pratt's Syncline and Pratt's Ferry sections. 95 Eighty-five specimens are proximal end fragments or early growth stages, and ten specimens are large rhabdosomes, which in most instances have a layer of cortical tissue. 2) Non-isolated specimens with periderm preserved. Thirty specimens were obtained from the Pratt's Ferry and Pratt's Syncline sections. 3) Carbon Films. One hundred and seventy specimens were obtained from the Calera section. Description The largest available isolated specimen (Text-fig. 15e-f) is 3 mm long, and the largest available non-isolated specimen is 10 mm long. The rhabdosome has a rectangular cross-section, and in a profile aspect it is fusiform with parallel ventral margins and a broad, rounded proximal end. Distal to th 2, the rhabdosome is 1.7 - 2.0 mm wide excluding the apertural lists and as much as 2.5 mm wide including the apertural lists. The rhabdosome consists of a sclerotized sicula, a clathria, and a layer of cortical tissue. The sicula (Text-figs. 16a-b, d-g) is conical and 0.5 - 0.65 mm long. The prosicula, which is 0.23 - 0.27 mm long and 0.09 - 0.12 mm wide at its aperture, accounts for almost half the length of the sicula. Longitudinal threads extend from the aperture of the prosicula to the base of the nema. In mose specimens representing early growth stages or proximal end-fragments, the nema is short; however, in a few large specimens the nema (virgula) extends distally through the entire length of the rhabdosome. 9 6 The metasicula, which accounts for only slightly more than half of the length of the sicula, is 0.28 - 0.36 mm long and is composed of transverse growth lines. From the aperture of the prosicula, the metasicula widens slightly to a maximum width of 0.13 - 0.15 mm before narrowing slightly to its aperture, which is 0.10 mm wide. The anti-virgellar wall of the sicula is longer than the virgellar wall, and as a result, the margin of the sicular aperture is oblique to the axis of the sicula. The virgella originates at a point 0.2 mm above the sicular aper ture, and it projects 0.2 mm beyond the sicular aperture. None of the available specimens reveals the structure of the virgella, i.e., whether or not it is formed by the downward curvature of growth lines on one side of the sicula. Its origin is at the level of th 1^ bud and the maximum width of the sicula. Besides the metasicula, the initial bud of th 1^ is the only other part of the rhabdosome composed of fusellar tissue. It originates at a point 0.2 mm above the sicular aperture, and it extends directly down ward along the sicula for a distance of 0.05 mm. The clathria (Text-figs. 15c-f) is composed of ventral, septal, parietal, and apertural lists. On each ventral side of the rhabdosome there are two ventral lists, vhich are connected at regular intervals by apertural lists. In profile aspect, each ventral list has a zig-zag course. The dorsally directed angles of the zig-zag course occur where the parietal lists merge with the ventral list, and the ventrally directed angles occur where the apertural lists join with the ventral 97 list. Proximally, the ventral lists of the first thecal series (Text- figs. 15e-f, 16h-j) are attached to the virgellar side of the sicula. The ventral lists of the second thecal series merge proximally at a point immediately below the th 1^ aperture, below which a single ventral list extends to and attaches to, the anti-virgellar side of the sicula. There is one septal list along the midline of each lateral wall of the rhabdosome. The septal lists originate from points near the apex of the sicula (Text-figs. 16h-j), and distally each has a zig-zag course in profile aspect. The angles of this zig-zag course occur where the parietal lists meet the septal lists. The septal lists are not connected directly to one another by any other lists. Parietal lists, which are perpendicular to the axis of the rhabdo some, are regularly distributed along the lateral walls of the rhabdosome, and they connect each septal list to the two ventral lists on the same lateral side of the rhabdosome. The parietal lists extend from the angles in a zig-zag course of the septal list to the dorsally directed angles in the zig-zag courses of the ventral lists. As a result, each lateral wall of the rhabdosome consists of two adjacent rows of hexagonal openings, which parallel the axis of the rhabdosome. Each opening is defined by the parietal lists on its distal and proximal sides, by the ventral lists on its ventral side, and by the septal list on its dorsal side. These openings are 0.4 - 0.5 mm wide (trans.) and 0.4 mm high (long). The apertures, as defined by the apertural lists, of the two thecal series are alternating, and as a result, the dorsally directed 98 angles in the zig-zag courses of the two ventral lists, the parietal lists on opposite sides of the septal list, and the two rows of hexagonal open ings are situated at alternating levels along the rhabdosome in profile aspect. The apertural lists are ring-shaped and oriented perpendicularly to the axis of the rhabdosome. They are situated at the ventrally-directed angles in the zig-zag course of the ventral lists, and they connect the two ventral lists of a single thecal series. The apertural lists of the opposing thecal series are alternating. The apertural lists are evenly distributed at 0.4 - 0.5 mm intervals along each ventral margin of the rhabdosome; thus, each apertural list corresponds to a hexagonal opening in the lateral wall of the rhabdosome. Short rhabdosomes are composed of only a sicula and a clathria. An additional component of large rhabdosomes is a layer of cortical tissue. In most of the large, non-isolated specimens, the proximal end of the rhabdosome is covered with cortical tissue, whereas the distal part of the rhabdosome lacks cortical tissue. It seems that the cortical tissue appears at a late stage in the astogeny of the rhabdosome. When it does appear, it first coats the proximal end of the rhabdosome, then spreads distally. In most of the large, isolated specimens, only pieces of the cortical periderm are preserved attached to the clathria (Text-fig. 15f). However, in a few specimens (Text-fig. 15a-b), the cortical tissue is well pre served. The cortical tissue is attached to the outer surface of the 99 lists that comprise the clathria, and in its maximum development, it covers the entire rhabdosome except for the sicular and thecal apertures. The addition of the cortical layer occurs in such a way that the free ventral walls of the thecae are straight, and for this reason, the thecae are referred to as being of the orthograptid type. Structure and Development of Proximal End Lists that have been broken show a blunt end, whereas in many specimens the distal end of the lists are pointed. These pointed ends of the lists are interpreted herein as representing "growing tips," and they allow the development of the proximal end to be determined. The earliest available growth stage (Text-fig. 16a-b) consists of the sicula, the initial bud of th 1^, and the initial part of the clathria. The initial part of the clathria is a circular list, one side of which appears to be continuous with the virgella. The initial bud of th 1^ is situated within the uppermost part of this circular list (Text-fig. 16d-e). With the growth of th 1^, two lists grow outward from the distal and proximal ends of the circular lists (Text-figs. 16a-c). These two lists are the ventral lists of th 1^ and the first thecal series (Compare Text-figs. 16c and 16j). Distally, at a distance of 0.5 - 0.6 mm from the sicula, the two ventral lists merge by the addition of an apertural list (Text-fig. 16d-e). With continued development (Text-fig. 16f-g) two lists appear 2 that represent the development of th 1 . One of these lists develops 1 0 0 from a point on the anti-virgellar side of the sicula near the sicular aperture, and it develops into the single ventral list of th 1 (Compare Text-figs. 16f-g, j). The other list develops from one of the ventral lists of th 1^, and it grows onliqueiy across the reverse side of the sicula. As it crosses the sicula, it joins with a septal list that appears on the reverse side of the apex of the sicula (Text-fig. 16h-i). With continued development (Text-fig. 16h-i), this list grows across to the apertural list of th 1 , thus, it is a parietal list for both th 1^ and th 1^. The direction of growth of the ventral and parietal lists of th 1^ suggests that th 1^ buds from the right side of th 1^, then grows across the reverse side of the sicula. When the parietal list of th 1^ joins the ventral list of th 1^, the lower part of the apertural list of th 1^ is formed. With continued development (Text-figs. 16h-j) the two ventral lists of the second 2 thecal series develop from th 1 apertural list and grow distally. Simultaneously, two septal lists begin growing from the apex of the sicula, and the two ventral lists of th 2^ develop. There are no 1 2 parietal lists for th 1 and th 1 on the obverse side of the sicula. In spite of the fact that there is no median septum, the timing of the appearance of the septal lists suggests that the two thecal series might be separated beyond the apex of the sicula. The origin of th 2^ can not be determined for certain, but the fact that the development of the ventral lists of th 2^ is approximately simultaneous with the development of the septal lists and the development of the 1 0 1 th 1^ aperture (Text-fig. 16h-i) suggests that th 2^ buds from th 1^. The development described and interpreted herein is of the dichograptid type. However, it is peculiar because of the upward direction of growth of all proximal thecae. Remarks The specimens described above agree closely with Hall’s (1859) figured specimen of Reteograptus geinitzianus. Several described specimens of R. geinitzianus (Hall, 1859; Ruedemann, 1908; Berry, 1964) show three rows of hexagonal openings on the lateral sides of the rhabdosome, and in some cases (Ruedemann, 1908), the rhabdosome has been interpreted as having a hexagonal cross-section. However, an examination of those specimens with three rows of openings, vdiich are preserved as carbon films compressed on shale surfaces, shows that one of the rows represents a ventral margin of the rhabdosome that has been rotated by compression into the plane of the lateral wall of the rhabdosome. Figured specimens OSU 32913 - OSU 32921 1 0 2 Family Nemagraptidae Lapworth, 1873 Diagnosis Ehabdosome consisting of two uniserial stipes having a primary angle of divergence of about 180 degrees; branches (if present) lateral, simple or compound, rarely paired. Thecae of Uemagraptus type, elongate, inclined at low angles, with distinct geniculum; proximal-end development of nemagraptid type. Apertural region of sicula projects downward beyond stipes. Discussion As defined by Bulman (1970), the family Nemagraptidae consists of the genera Nemagraptus Emmons, Amphigraptus Lapworth, Leptograptus Lapworth, Pleurograptus Nicholson, and Syndyograptus Ruedemann, which show a leptograptid type of theca and a leptograptid type of proximal- end development. Nemagraptids have not previously been described on the basis of isolated material, and the family Nemagraptidae is here revised in light of new evidence that is provided by the Alabama material described below. The leptograptid type of theca has long been considered to be intermediate between the dichograptid and dicellograptid types because of a gentle sigmoid curvature of the ventral wall and a low angle of inclination. However, as revealed by the Alabama material, Leptograptus has a type of theca that represents only a slight midification of the dicellograptid type by absence of introtorsion, and the thecae of Nemagraptus display a geniculum. 103 The leptograptid type of proximal-end development has not previously been described on the basis of isolated early growth stages. It was defined by Elies (1922) and Bulman (1970) on specimens that are compressed 2 on shale surfaces; and the number of crossing canals, the origin of th 1 , and the position of the dicalycal theca were not known. As described herein, Leptograptus has a diplograptid-dicranograptid type of proximal- end development, whereas Nemagraptus has three crossing canals, a 1 9 dicalycal th 2 , and a right-hand origin of th 1 . On the basis of its type of thecae and development, Leptograptus is here transferred to the family Dicranograptidae. The type of theca that is described and discussed below for IJ. gracilis is here termed the Nemagraptus type, and the type of proximal-end development that is described and discussed below for N. gracilis is here termed the nema graptid type. All the species of Nemagraptus investigated show nemagraptid type of thecae and proximal-end development. The proximal ends of all previously described representatives of Amphigraptus are not preserved in a way that the development can be determined. However, the Alabama material includes two new species of Amphigraptus, which show a Nema graptus type of theca and a nemagraptid type of proximal-end development. Strachan's (1969, fig. lb,c) illustrated specimens of Pleurograptus linearis (Carruthers), and Mu's (1963, Text-fig. 8a,b) illustrated specimens of ]P. lui Mu show a proximal-end morphology that resembles that of Leptograptus in the divergence of the stipes from the sicular 104 aperture. Elles and Wood (1903, p. 119) remark that "Pleurograptus is best regarded as a Leptograptus, from the main stipes of which numerous simple and compound secondary branches are given off at irregular intervals." This similarity suggests that Pleurograptus, along with Leptograptus, might be transferred to the Dicranograptidae. However, the thecal morphology and mode of branching are not well known for Pleurograptus because all the described specimens are preserved on shale surfaces, and pending further study, this genus is here retained in the Nemagraptidae because of its rhabdosomal shape. Although poorly known, the proximal-end morphology of Syndyograptus resembles that of vagus group dicellograptids because the sicula is in clined and the stipes diverge at the level of the siculular aperture (Ruedemann, 1908; p. 267, fig. 185). Tangyograptus Mu was regarded by Bulman (1970) as a junior subjective synonym of Syndyograptus. The orientation of the sicula and the first two thecae in Tangyograptus is very similar to that in elegans group dicellograptids. The thecal morphology is not well known in specimens of Syndyograptus and Tangyograptus ; however on the basis of the proximal-end morphology, Syndyograptus and Tangyo graptus can be regarded as multiramous dicranograptids associated with the vagus and elegans groups of Dicellograptus, respectively. Regarded as such, these genera should be transferred from Nemagraptidae to Dicranograptidae. The thecae of Nemagraptid sp. A, which is described below, show a modification of the Nemagraptus type by the introversion of the aperture 105 and the development of the genicular flange. The proximal-end development is of the nemagraptid type. Although the fragmentary nature of the available specimens do not allow the original shape of the rhabdosome to be determined, these specimens probably represent a new genus within the family Nemagraptidae. N^. subtilis Hadding, which lacks cladia and is the oldest known species in the family Nemagraptidae, may represent the "ancestral stock" from which all other species of Nemagraptidae evolved. Genus Nemagraptus Emmons, 1855 Type species: Nemagraptus gracilis (Hall, 1847) Diagnosis A genus of Nemagraptidae characterized by reclined stipes; thecal metacladia (if present) simple and unpaired. Discussion In the following species description, K[. gracilis, N. exilis, and their subspecies are interpreted as representing different preservational aspects of the same species. As revised, the genus Nemagraptus consists of the following species and subspecies: Nemagraptus gracilis (Hall, 1847) Nemagraptus subtilis Hadding, 1913 Nemagraptus explanatus explanatus (Lapworth, 1876) Nemagraptus explanatus pertenius (Lapworth, 1876) These species are essentially similar except for the habit of the stipes and the presence and position of thecal metacladia. 106 Nemagraptus gracilis (Hall, 1847) (Text-figures 17-22) 1847 Grapolithus gracilis n.sp.. Hall, p. 274, PI. 74, fig. 6a-d. 1856 Nemagrapsus elegans (n.s.), Emmons, p. 109, PI. 1, fig. 6. 1859 Graptolithus gracilis Hall, Hall, p. 510, fig. 1-7. 1868 Coenograptus gracilis Hall, Hall, p. 179-180, fig. 17-19. 1868 Coenograptus surcularis Hall, Hall, p. 179, fig. 13-16. 1868 Helicograpsus gracilis Hall. Nicholson, p. 25, fig. 1. 1876 Coenograptus gracilis Hall, Lapworth, p. 5, PI. 3, fig. 65. 1876 Coenograptus surcularis Hall, Lapworth, PI. 3, fig. 64. 1876 Coenograptus nitidulus n. sp., Lapworth, PI. 3, fig. 66. 1877 Coenograptus gracilis Hall, Lapworth, p. 142, PI. 7, fig. 11. 1877 Coenograptus surcularis Hall, Lapworth, p. 143, PI. 7, fig. 12. 1896 Stephanograptus crassicaulis sp. nov., Gurley, p. 68. 1896 Stephanograptus exilis sp. nov., Lapworth in Gurley, p. 68-69. 1902 Coenograptus gracilis Hall, Ruedemann, p. 583, fig. 13. 1903 Nemagraptus gracilis (Hall), Elies and Wood, p. 127, PI. 19, fig. la-f. 1903 Nemagraptus gracilis var. surcularis (Hall), Elies and Wood, p. 129-130, PI. 19, figs. 2a-d. 1903 Nemagraptus gracilis var. remotus var. nov., Elies and Wood, p. 130-131, PI. 19, figs. 3a-h. 1903 Nemagraptusgracilis var. nitidulus (Lapworth) , Elies and Wood, p. 131-132, Pi. 19, figs. 4a-d. 1908 Nemagraptus gracilis (Hall), Ruedemann, p.277-282, Text-figs. 192-195; PI. 16, figs. 1-5. 107 EXPLANATION OF TEXT-FIGURE 17 Text-figure 17 a-k. Nemagraptus gracilis (Hall). Isolated specimens. a. Lateral aspect of stipe fragment showing a thecal aperture with a genicular flange. Note prothecal fold and orientation and thickness of growth lines on protheca and metatheca. X37.5. PS-126. OSU 32922. b. Lateral aspect of stipe fragment. Note the geniculum, the prothecal fold, and the absence of a genicular flange. X37.5. PS-117. OSU 32923. c,d. Dorsal and ventral aspects of stipe fragment. Note the growth-line pattern on the protheca and metatheca, position of primary notch, and presence of genicular flange. X37.5. PS-102. OSU 32924. e. Lateral aspect of stipe fragment. Note that growth-line pattern changes at genicular flange. X37.5. PS-109.2. OSU 32925. f. Lateral view of stipe fragment. Note length of theca and small amount of overlap. X37.5. PS 109.7. OSU 32926. g. Reverse aspect of proximal end. Note divergence of stipes from middle of sicula. X37.5. PS-117. OSU 32927. h. Lateral view of stipe fragment. Note orientation of growth lines. Dashed lines are growth lines on inter- thecal septum that are visible through lateral wall. X37.5. PS-109.2. OSU 32928. i,j. Reverse and obverse aspects of proximal end. Th 2^ broken. Note overlap of th 2^ and th 3 , and position and orientation of thecal metacladium. X37.5. PS-76.4. OSU 32929. k. Reverse aspect of proximal end. Stipe no. 2 broken off from specimen. Note curvature of th 2^ at th 1^ aperture. X37.5. PS-109.2. OSU 32930. 108 Text-figure 17 1 0 9 EXPLANATION OF TEXT-FIGURE 18 Text-figure 18 a-1. Nemagraptus gracilis (Hall). Isolated early growth stages. a. Sicula with initial bud of th 1^. Note resorption foramen high on metasicula. X37.5. PS-126. OSU 32931. b. Reverse aspect of sicula with th 1^ and initial part of th 1^. Note right-hand origin of th 1^ from th 1^ and divergence of th 1^ from midpoint of sicula. X37.5. OSU 32932. c. Reverse aspect of sicula with th 1 T and th 1 9 . Note lateral increase in width of th 1^ from that shown in Text-fig. 18b. X37.5. PS-109.2. OSU 32933. d. Reverse aspect of sicula with th 1^ and th 1^. Note process directed toward sicula aperture. X37.5. PS-109.2. OSU 32934. e. Reverse aspect of sicula with th 1 1 , th 1 2 , and opening for th 2^. Note left-hand origin of th 2^ from th 1^. X37.5. PS-126. OSU 32935. f. Reverse aspect of sicula with th 11 , th 12 , and opening for th 2^. X37.5. PS-109.2. OSU 32936. g. Reverse aspect of sicula with th 11 , th 12 , and opening for th 2^. X37.5. PS-109.2. OSU 32937. h. Reverse aspect of sicula with th 1 1 , th 1 2 , and th 2 1. Note horizontal direction of growth of th 2^. X37.5. PS-126. OSU 32938. 1 2 1 i. Reverse aspect of sicula with th 1 , th 1 , th 2 , and th 2^. Note that ventral margin of th 2^is continuous with ventral margin of th 2^. X37.5. PS-109.7. OSU 32939. 1 2 1 2 j. Reverse aspect of sicula with th 1 , th 1 » th 2 , and th 2 . Note directions of growth of th 2^ and th 2 . X37.5. PS-76.4. OSU 32940. 1 1 0 k. Reverse aspect of proximal end. Stipe no. 1 broken. Note position of th 2^ on dorsal wall of th 1 . X37.5. PS-109.2. OSU 32941. 1. Reverse aspect of proximal end. Stipe no. 1 broken. Note extent of th 2^ beyond th 1^. X37.5. PS-129. OSU 32942. Ill -rf th 1 th i ,th2 Text-figure 18 1 1 2 EXPLANATION OF TEXT-FIGURE 19 Text-figure 19 a-m. Nemagraptus gracilis (Hall). Lateral aspects of isolated stipe fragments showing thecal morphology. a. Type I theca. Note small thecal overlap, and length of free ventral wall. X37.5. PS-109.7. OSU 32943. b. Type I theca. Note extent of genicular flange. X37.5. PS-102. OSU 32944. c. Transitional type II-III theca. Note extent of genicular flange. X37.5. PS-109.2. OSU 32945. d. Type II theca. X37.5. PS-109.7. OSU 32946. e. Type II theca. X37.5. PS-117. OSU 32947. f. Transitional type I-II theca. Note extent of genicular flange. X37.5.' PS-126. OSU 32948. g. Type II theca. X37.5. PS-109.2. OSU 32949. h. Type II theca. X37.5. PS-109.2. OSU 32950. i. Type III theca. Note width of theca and overlap. X37.5. PS-109.2. OSU 32951. j. Type I theca. X37.5. PS-102. OSU 32952. k. Type I theca. X37.5. PS-102. OSU 32953. 1. Transitional type II-III theca. X37.5. PS-126. OSU 32954. m. Transitional type II-III theca. X37.5. PS-109.7. OSU 32955. 113 pn-_ / . / iw sw- iff g pf--/ Text-figure 19 114 EXPLANATION OF TEXT-FIGURE 20 Text-figure 20 a-g. Nemagraptus gracilis (Hall). Isolated stipe fragments showing development of thecal metacladia. a,b. Ventral and dorsal aspects of stipe fragment with cladium. Note relationship of cladium to thecal aperture and foramen at base of cladium opening into thecal aperture. X37.5. PS-129. OSU 32956. c,d. Ventral and dorsal aspects of stipe fragment. Stippled area is most proximal part of cladium. X37.5. PS-109.2. OSU 32957. e,f. Left- and right-lateral aspects of stipe fragment. Note thin strip of tissue extending across, and restricting the right-lateral side of, the thecal aperture. X37.5. PS-109.2. OSU 32958. g. Ventral aspect of stipe fragment with cladium. Note length of cladium, which does not include any thecal apertures. X37.5. PS-109.2. OSU 32959. h. Sketch showing proximal-end development in Nemagraptus gracilis (Hall), which is defined herein as nemagraptid type. X. Reconstruction of rhabdosome of N. gracilis (Hall). 115 I— u / Text-figure 20 116 EXPLANATION OF TEXT-FIGURE 21 Text-figure 21 a-f. Nemagraptus gracilis (Hall). Compressed, non-isolated specimens. a,b. Proximally, stipe no. 1 presenting ventral aspect, stipe no. 2 presenting dorsal aspect, a. X9, b. X18.8. PF-17.3. OSU 32960. c. Proximally, both stipe no. 1 and stipe no. 2 presenting dorsal aspects. X9. PF-17.3. OSU 32961. d. Stipe no. 1 presenting ventral aspect. Stipe no. 2 presenting dorsal aspect. X9. C-33. OSU 32962. e. Stipe no. 1 presenting right-lateral aspect. Proximally, stipe no. 2 presenting left-lateral aspect. X9. Pf-17.3. OSU 32963. f. Ascending stipe presenting left-lateral aspect proximally and dorsal aspect distally. Descending stipe presenting dorsal aspect. X9. PF-17.3. OSU 32964. Text-figure 21 118 EXPLANATION OF TEXT-FIGURE 22 Text-figure 22 a-h. Nemagraptus gracilis (Hall). Compressed, non-isolated specimens. a. Obverse aspect. Note that stipes present lateral aspects, and cladia are not visible. X18.8. PS-117. OSU 32965. b. Obverse aspect. Note that stipes present lateral aspects, and cladia are not visible. X18.8. PS-117. OSU 32966. c. Stipe fragment with cladium. Note that cladium arises from right-lateral side of thecal aperture on stipe and that the first thecal aperture on the cladium opens in a proximal direction relative to the stipe. X37.5. PS-119.5. OSU 32967. d. Stipe fragment. Note that stipe presents lateral aspect and cladia are not visible. X18.8. PS-129. OSU 32968. e. Stipe fragment with cladium. Note that stipe presents dorsal aspects and cladium arises from right-lateral side. X37.5. PF-16.4. OSU 32969. f. Stipe fragment with cladia. Note that where stipe presents dorsal aspect, cladia arise from right-lateral side. Where stipe presents lateral aspect, cladia are not visible. X9. PS-13. OSU 32970. g. Stipe fragment with cladium. X18.8. PS-132. OSU 32971. h. Obverse aspect of proximal end. Note that stipe no. 1 presents dorsal aspect, cladia project from right-lateral side of stipe no. 1, and thecal apertures on cladia open proximally relative to stipe no. 1. X18.8. PF-17.1. OSU 32972. 119 I Text-figure 22 1 2 0 1908 Nemagraptus gracilis var. surcularis (Hall), Ruedemann, p. 282-285, Text-figs. 196-197; Pl. 17, figs. 1, 2. 1908 Nemagraptus gracilis var. crassicaulis Gurley, Ruedemann, p. 285-286, Pl. 17, fig. 13. 1908 Nemagraptus gracilis var. distans nov., Ruedemann, p. 286 Pl. 16, fig. 7, 8. 1908 Nemagraptus gracilis var. approximates nov., Ruedemann, p. 287, Pl. 16, fig. 5, 6. 1908 Nemagraptus exilis (Lapworth), Ruedemann, p. 287-290, Text-figs. 202, 203; Pl. 17, figs. 3-9 1908 Nemagraptus exilis (Lapworth) var. linearis nov., Ruedemann, p. 290-291, Text-figs. 204, 205; Pl. 17, figs. 10-12 1913 Nemagraptus gracilis Hall var. remotus Elles and Wood, Hadding, p. 56-57, Pl. 4, figs. 24-29. 1926 Nemagraptus gracilis. Butts, p. 109, Pl. 23, fig. 4. 1933 Nemagraptus exilis (Lapw.), Sun, p. 11-12 Pl. 1, figs. 6a, b. 1933 Nemagraptus gracilis (Hall), Sun, p. 12-13, Pl. 2, figs. la-d. 1947 Nemagraptus gracilis (Hall), Ruedemann, p. 367-368, Pl.60 figs. 1-12 1947 Nemagraptus gracilis (Hall) var. approximates Ruedemann, Ruedemann, p. 368-369, Pl. 60, figs. 13-14 1947 Nemagraptus gracilis (Hall) var. crassicaulis (Gurley), Ruedemann, p. 369, Pl. 61, figs. 18-21 1947 Nemagraptus gracilis (Hall) var. distans Ruedemann, Ruedemann, p. 369, Pl. 60, figs. 15-16 1947 Nemagraptus gracilis (Hall) var. surcularis (Hall), Ruedemann, p. 370, Pl. 60, figs. 17-24 1947 Nemagraptus gracilis (Hall) cf. var. nitidulus (Lapworth), Ruedemann, p. 370-371, Pl. 61, figs. 15-17 1947 Nemagraptus exilis (Lapworth), Ruedemann, p. 371, Pl.61, figs. 1-9 1 2 1 1947 Nemagraptus exilis (Lapworth) var. linearis Ruedemann, Ruedemann, p. 372, Pl. 61, figs. 10-14. 1952 Nemagraptus gracilis (Hall), Decker, Pl. 1. fig. 50; Pl. 2, fig. 84. 1952 Nemagraptus gracilis var. dis tans Ruedemann, Decker, Pl. 2, fig. 81. 1952 Nemagraptus gracilis var. nitidulus Lapworth, Decker, Pl. 2, fig. 83. 1952 Nemagraptus gracilis var surcularis (Hall), Decker, Pl. 2, fig. 82. 1952 Nemagraptus exilis Lapworth, Decker, Pl. 1, fig. 51; Pl. 2, fig. 80. 1954 Genus? incisus n. sp., KozZowski, p. 434-437, figs. 7, 8 . 1960 Nemagraptus gracilis (Hall), Berry, p. 73, Pl. 15, fig. 13. 1960 Nemagraptus gracilis var. surcularis (Hall), Berry, p. 73. 1960 Nemagraptus exilis var. linearis Ruedemann, Berry, p. 72-73. 1960 Nemagraptus gracilis (Hall), Turner, p. 95, Pl. 5, fig. 6 . 1963 Nemagraptus sinicus Mu and Qiao sp. nov., Mu, p. 376-377, Text- figs. 5a-c. Type Data Hall (1847) did not designate a holotype among his syntypes. The syntypes are stored at the American Museum of Natural History, and all of them are numbered AMNH 1043/1. The most complete specimen, idiich is figured by Hall (1847, PI. 74, fig. 6b) is here designated the lectotype because it is the most complete rhabdosome; the remaining syntypes are distal stipe fragments. Diagnosis A species of Nemagraptus with the rhabdosome consisting of two curved stipes that define a S- or U-shape and of numerous cladia projecting from convex side of the stipes. Subapertural mesial spine on first theca of each stipe; metacladia on succeeding thecal apertures. Stipes increase 1 2 2 in width distally; 0.15 mm at th 1; maximum width distally 0.6 mm. Cladia increase in width distally; 0.07 mm at origin; maximum width distally 0.8 mm. Thecae long and slender; thecal length four to seven times greater than width; overlap a third to a sixth their length. Thecal density variable throughout rhabdosome; 5-10 thecae in 10 mm on stipes; 7-10 thecae in 10 mm on cladia. Geniculum with genicular flange of microfusselar tissue separates long supragenicular wall from short infra- genicular wall. Supragenicular wall slightly inclined to stipe or cladial axis. Thecal apertures simple and perpendicular to stipe or cladial axis. Prothecal folds on dorsal stipe margin. Material More than 600 generally fragmentary specimens are available for study. Half of these are isolated specimens from the Pratt's Syncline section. Most are compressed, but they commonly show growth lines. One hundred and twenty-five of the isolated specimens are proximal end fragments or early growth stages. Another 125 specimens are stipe fragments of which none is over 3 mm long or consists of more than three thecae. Fifty of the isolated specimens are short stipe fragments with projecting cladia. Over 100 of the available specimens from the Pratt's Ferry and Pratt's Syncline sections are compressed on shale surfaces. Although they are not well-preserved, these specimens occasionally show thecal morphology. Several of them are large rhabdosomes; however, most are distal stipe fragments and proximal ends. More than 250 specimens are preserved as 1 2 3 carbon films and rarely show thecal morphology. Large rhabdosomes are rare; the available specimens mostly are distal stipe fragments and proxi mal ends. Description The morphology and development of the proximal end and the thecae are described below in detail on the basis of isolated specimens. The thecae show a large amount of variation, and they are described in terms of three extreme variant forms. The exact position of these extreme forms within a rhabdosome can not be precisely determined on the basis of isolated material because of its fragmentary nature. The detailed thecal morphology is rarely exhibited in large rhabdosomes that are compressed on shale surfaces, but when it is preserved, the particular variant forms of the thecae can tentatively be positioned in the rhabdo some. There is a suggestion of morphological gradients of the thecae on both the stipes and the cladia as described by Thorsteinsson (1955) for Cyrtograptus. Isolated specimens, representing the proximal ends of Nemagraptus gracilis and Nemagraptid sp. A are very similar in both their morphology and their development. Both species occur together in three samples, and the available specimens of the two species can only be distinguished if the apertures of the first two thecae are developed. The various parts of the sicula appear to be larger in size for specimens of Nemagraptus gracilis than for specimens of Nemagraptid sp. A. Therefore, a quantitative differentiation of the two species was attempted 124 on specimens in which the proximal ends are well enough developed to show distinguishing thecal characteristics in order to establish a quantitative scheme for the differentiation of specimens in which the proximal ends are not well enough developed to show distinguishing thecal features. The measurements, which were taken on specimens from the same horizon, are shown in Table 1. They are generally larger for N. gracilis than for Nemagraptid sp. A. However, the ranges of variation for the two species overlap, and an individual proximal end without the distinguishing thecal characters would be nearly impossible to distinguish quantitatively. The descriptions of the morphology of the proximal ends of the two species are based on specimens that show the distinguishing thecal char acters. These specimens generally do not show the early development of the proximal end. Therefore, this development is described and illustrated for N. gracilis on the basis of specimens that show very large measurements of the sicula relative to the values in Table 1 because these specimens most likely represent N. gracilis■ Because the development is identical ■in the two species, it is only briefly discussed in the description of Nemagraptid sp. A. Isolated Specimens. Morphology of the Proximal End; The largest available specimen (Text-fig. 17i) has a 1.2 mm long stipe, which shows the initial part of th 3^ and the initial part of the cladia that originates from the aperture of th 2^. The first four thecae diverge from the sicula midway between the sicular aperture and the prosicula. The orientation of the initial portions of the stipes varies TABLE 1: Sicular measurements on specimens from locality PS-109.2, in mm. Nemagraptus gracilis Nemagraptid sp. A length of distance from distance from length of distance from distance from sicula horizontal part initial bud sicula horizontal part initial bud of th 1 ^ to of th 1 to of th 1 to of th 1^ to sicular apert. sicular apert. sicular apert. sicular apert. 0.91 0.19 0.57 0.76 0.14 0.63 0.18 0.80 0.17 0.53 0.27 0.53 0 . 1 1 0.49 0.87 0.24 0.55 0 . 2 1 0.56 0.84 0.21 0.56 0.74 0.14 0.49 0.31 0.69 0.20 0.55 0.91 0.28 0.60 mean= 0.77 mean= 0.16 mean= 0.54 0.87 0.28 0.62 range: 0.74- range: 0 .11- range: 0.49- 0 . 22 0.53 0.80 0.21 0.63 0.28 0.59 mean= 0.85 mean= 0.24 mean= 0.56 range: 0.70- range: 0.17- range : 0.45- 0.91 0.31 0.69 ro ui 126 from horizontal to reclined among the available specimens. The horizontal orientation (Text-fig. 17g) is most common and is found in nearly all early growth stages. The very reclined orientation (Text-fig. 17i) is common in proximal ends that are fragments from larger rhabdosomes, and it is generally associated with a bend in the first theca of each stipe, which appears to be due to préservâtional distortion. In a few specimens (Text-fig. 17k), a change from a horizontal to a reclined habit occurs at the first thecal aperture. This change in orientation appears to be caused by the direction of growth of the thecae and not by distortion. The apertures of the first two thecae are 0.5 - 0.6 mm from the sicula. They are identical to those described below for distal thecae, except for the presence of mesial spines located 0 . 1 mm below the ventral margin of the aperture. The stipes are 0.15 - 0.20 mm wide at the level of the subapertural spines on the first theca of each stipe. The aperture of the second theca on each stipe is 0.6 mm beyond that of the first theca. For two-thirds of this distance, the stipe consists only of the second theca, after which the third theca develops from the second. Less than 0.2 mm beyond the origin of the third theca, the aper ture of the second theca is attained. The first cladium projects from the right-lateral side of the second thecal aperture on each stipe, when the stipe is viewed dorsally (Text-fig. 17j). Development of the Proximal End. The sicula is 0.7 - 0.9 mm long. The cone-shaped prosicula occupies a third of this length and has a 0.10 - 0.12 mm wide aperture. Three or four longitudinal threads extend 127 from the aperture to the apex of the prosicula. A nema, which can be as much as 0.5 mm long, extends distally from the apex of the prosicula (Text- fig. 18f ). The metasicula is parallel-sided, and its aperture shows paired, lateral lappets and a 0.3 - 0.5 mm long virgella. Transverse growth lines on the metasicula bend downward as they approach the virgella (Text-fig. 18e). Th ll originates from a foramen high on the virgellar side of the metasicula and immediately below the aperture of the prosicula (Text- fig. 18a). It grows halfway down the metasicula, where it bends and grows outward 0.5 - 0.6 mm before developing an aperture (Text-figs. 18b-c, j)" None of the available growth stages show the initial development of th 1^. Where it is partly developed and can be traced backwards (Text- figs. 18b-c, h), th 1 merges with the right side of th 1 , and it appears to originate from th 1^ at a short distance below the th 1 ^ foramen. O As th 1 grows downward, it expands transversally until it fomns a hood at the level of the horizontally directed part of th 1^ (Text- figs. 18b-d). Th 2^ develops from the left-handed part of the opening 2 under the hood. Th 1 then grows 0.6 mm in a subhorizontal direction outward from the sicula, where it develops an aperture (Text-figs. 18d-g, k). The available growth stages do not reveal the exact manner in which th 2^ originates. Text-figure 18c shows a specimen in which the distal o 1 end of th 1 forms a broad hood and th 2 is not developed. In the 128 2 specimens shown in Text-figure 18e-g, th 1 is developed to at least the anti-virgellar side of the sicula, and it forms a complete tube. Compari son with the development of Dicellograptus suggests that the origin of th 2 ^ and the formation of the complete tube for th 1^ occurs by the extension of a process, which is marked by a "p" in Text-figure 18d, down onto the sicula from the free outer wall of th 1^. Two apertures would then be formed: a left-hand one from which th 2^ develops and a right-hand one from which th 1^ continues to develop. Th 2^ grows horizontally across the bend in th 1^ and onto the dorsal wall of th 1 ^, where it continues to grow beyond the th 1^ aperture (Text-figs. I8h-j). As with th 1^ and th 2^, the available growth stages do not reveal the exact manner in which th 2^ develops. In Text-figures 18 i and 2 18], th 2 appears as a horizontal tube whose lower margin is continuous with the lower margin of th 2^. When traced proximally, the upper margin of th 2^ appears to extend onto the earliest part of th 2 ^ (Text-fig. 18i). Thus, th 2^ is interpreted to originate from th 2^, 2 and its origin is right-handed. Th 2 grows horizontally across 2 9 the bend in th 1 and onto the dorsal wall of th 2 ^, where it continues 2 1 to grow out beyond the aperture of th 1 (Text-figs. 18i-l). Th 3 originates from the distal part of th 2^ (Text-fig. ISj), and th 3^ originates from the distal part of th 2 ^. Morphology of the thecae. The thecae can be characterized as having: 1) Subparallel and straight, dorsal and ventral thecal walls. 129 2) Indistinct genicula, which divide the free ventral walls into relatively short infragenicular walls and relatively long supra genicular walls. 3) Genicular flanges, which are probably composed of microfusellar tissue. 4) A small amount of overlap that results in relatively long prothecal segments in comparison to the metathecal segments. 5) Simple apertures with slightly raised lateral lappets. 6) Prothecal folds produced by the direction of growth of the initial part of each theca. The available specimens show a wide range of quantitative variation in the case of the length, width, and amount of overlap of the thecae- This variation is concentrated among the available specimens into three variant forms and transitional specimens are rare. Because the longest stipe fragments are composed of no more than three thecae, the exact position within the rhabdosome is only known on the basis of isolated specimens for the first variant form that is described below. Text- figure 19 shows thecae, which represent the three extreme variant forms, as well as transitional forms. Table 2 shows measurements taken on these specimens. Relative to the distance between thecal apertures, the thecae of Group I are very thin. They overlap for a sixth or less of their length, and, on the basis of the distance between thecal apertures, the TABLE 2: Measurements on figured thecae of N. gracilis* In mm. Text-fig. Maximum Minimum Distance Ratio of Thecal Thecal Amount of Thecal width width between distance length density. Overlap Group successive between Inferred, apertures apertures 5mra/10mm to max. 19j 0.11 0.07 1.0 9.0 1.34 5/10 1/6 19 a 0.14 0.07 1.2 8.6 ca. 1.45 4.2/8.4 ca. 1/7 Group I 19b 0.14 0.07 19k 0.20 0.10 1.2 6.0 1.44 4.2/8.4 1/6 19 f 0.20 0.09 1.1 5.5 1.26 4.5/9 1/5 191 0.28* 0.14 0.92 3.3* 1.36 5.4/11 1/4-1/5 I-II 19m 0.24 0.14 1.16 4.8 1.47 4.2/8.4 1/6 transitions 19h 0.28 U.18 19 g 0.24 0.14 1.0 4.2 1.33 5/10 1/4 Group II 19 d 0.28 0.18 1.12 4.0 1.56 4.5/9 1/4 19 e 0.24 0.14 1.0 4.2 1.40 5/10 1/4 II-III 0.24 0.87 2.8 1.32 5.7/11.5 1/3 19 c 0.31 transitions 191 0.42 0.2 1 0.87 2.1 1.36 5.7/11.5 1/3 Group III *Si)eclmen Is o1allquely con ipressed; lar; width may be due to d Istortlon, a nd the n itlo of dlst ance between apertures t thecae number 4 - 5 in 5 mm and 8 - 10 in 10 mm. The proximal parts of the stipes of all the available specimens with proximal ends always show Group I thecae. The thecae of Group II are wider in comparison to those of Group I. The ratio of the distance between successive thecal apertures to the maximum width is 4.2 or less. They overlap for a fourth of their length, and they number 4.5 — 5 in 5 mm and 9 — 10 in 10 mm. Group II thecae are most likely situated on distal portions of the stipes and cladia relative to Group I thecae. Transitional forms are present for Group I and Group II thecae. The significant quantitative differences between the two groups are the amount of overlap and the width, as expressed by the ratio of the distance between thecal apertures to the maximum width. Group III thecae are extremely wide and show a significantly shorter distance between successive thecal apertures in comparison to Group I and Group II thecae. Group II thecae number 5.5 — 6 in 6 mm and 11.5 in 10 mm; they overlap for a third their length. The position of Group III thecae within the rhabdosome is probably distal to that occupied by Group II thecae. In spite of quantitative variation, all the thecae show the same basic structures. The transverse cross-section of a theca is an ellipse with the longest axis in a dorso-ventral direction. The dorsal and ventral thecal walls are subparallel; the angle of divergence between them is generally less than 10 degrees. Each theca gradually increases in width distally until the succeeding theca develops. 132 The dorsal thecal (and stipe) wall is straight with regularly spaced undulations that represent prothecal folds. The position at which the prothecal fold originates is variable among the thecal groups described above, and it depends on the amount of thecal overlap. The position of the distal limit of the prothecal fold is consistently transverse to the geniculum. The ventral thecal wall is generally straight except in Group III thecae, which have slightly convex curvature. A geniculum located at the distal limit of the apertural excavation of the preceding theca and serving as a base for a genicular flange divides the free ventral wall into an extremely short infragenicular wall relative to the supragenicular wall. The ratio of the length of the supragenicular wall to the infragenicular wall is 20:1 in Group I thecae and 9:1 in Group III thecae. Almost all the available specimens are partly or fully compressed, and the geniculum is somewhat indistinct. Text-figure 17b shows a specimen in which the genicular flange is not present and the geniculum is distinct. The genicular flange projects a maximum of 0.1 mm from the geniculum. It appears as a dark band against the background of the thecal walls. In some specimens, e.g. that figured in Text-fig. 17b, it is not present. In others, e.g. those in Text-fig. 17a, f, it is relatively short and wedge-shaped in lateral view, and its lateral margins extend only a short distance onto the lateral stipe walls. At its largest, the genicular flange is a thin projection extending its maximum distance from the geniculum (Text-fig. 19d). Its lateral margins extend onto the lateral stipe walls and then bend proximally to the dorso-lateral comers of the 133 preceding thecal aperture. Together with the margins of the aperture, the genicular flange defines in lateral view a semi-circular apertural excavation. The absence of a genicular flange in specimens with the succeeding theca developed (Text-fig. 17b) indicates that the genicular flange is developed very late in the thecal ontogeny. It is superimposed on fusellar tissue, and it is gradually constructed. Growth lines are not visible on the genicular flange. Their dark color and their mode of development suggest that they are composed of microfusellar tissue as it is defined by Urbanek (1966, p. 306). The thecal aperture is relatively simple because it is not introverted or introtorted. Its apertural margins define a plane that is perpendicular to the longitudinal axis of the stipe. The mid-points on the lateral margins of the aperture are slightly raised to form gentle lappets. The apertural escavation, as delimited by the apertural margins and the infragenicular wall, occupies a half of the stipe width, and a twentieth and a ninth of the length of the free ventral wall in Group I and Group II thecae, respectively. Ontogeny of the thecae. Thecal ontogeny> which includes the develop ment of the prothecal fold, the infragenicular wall, and the interthecal septum, is revealed by several specimens that show growth lines. For its entire length, which is dependent on the amount of thecal overlap, the interthecal septum, as represented by a trace on the lateral stipe walls, is parallel to the longitudinal axis of the stipe. Several transparent specimens (Text-figs. 17c-d, h) show the base of the inter thecal septum. Proximal to the proximal end of the septum, there is a 134 circular opening through which th n+1 buds from th n. The margins of this opening appear as dark lines and may be secondarily thickened. They are continuous with the lateral margins of the interthecal septum, and in lateral view their traces extend in a dorso-proximal direction from the proximal end of the interthecal septum to the dorsal stipe wall, where they define the proximal end of the prothecal fold. Growth lines on the most distal part of the th n protheca are perpendicular to the dorsal stipe wall and attain their largest dorso- ventral length immediately proximal to the th n+1 prothecal fold. (Text-figs. 17e-f, h). They rapidly decrease in dorso-ventral length with the budding of th n+1, and they are of uniform length in the short metathecal segment. As revealed by the growth lines, the initial fuse Hi of th n+1 are relatively thin and are oriented in a dorso-proximal direction relative to the dorsal stipe wall (Text-figs. 17a, e, h). This orientation and thickness of the fuselli is maintained distally to the geniculum. The growth lines on the lateral walls of th n+1 are continuous with those on the interthecal septum between th n and th n+1 and with those on the th n+1 infragenicular wall (Text-fig. 17h). At the level of the th n+1 geniculum, the fuselli abruptly change to a thickness and orientation (thicker and perpendicular to the dorsal stipe wall ) that is uniform through the remaining ontogeny of the theca. If growth lines are disregarded, th n+1 appears to bud from the foramen that develops in the dorsal wall of th n and after the th n aperture is complete (Text-fig. 17c-d). The th n+1 protheca then 135 appears to grow along the dorsal wall of th n, bending ventrally over the th n aperture. However, the evidence from the growth lines indicates that the interthecal septum between th n and th n+1 and the th n+1 infragenicular wall are continuous with the lateral walls of th n+1 and develop with th n+1. The orientation of the growth lines on the lateral thecal walls indicates that the ventral wall of th n+1 (the interthecal septum and the infragenicular wall) develops sli^tly ahead of the dorsal wall. Although none of the available specimens show the stages in the initial development of th n+1 from th n, it is interpreted as follows: 1) Th n develops as a complete tube until th n+1 buds. 2) When th n+1 buds, it foirms a complete tube with its own ventral wall, which is the interthecal septum between th n and th n+1. A foramen, which forms as a primary notch, is left at the base of this ventral wall for the connection of living tissue. 3) The metatheca of th n continues to grow, but as a split tube without a dorsal wall. 4) Because the interthecal septum is interpreted to develop as a ventral wall of th n+1, th n+1 probably develops in advance of or at the same time as the th n metatheca. 5) Growth is at a constant rate until the geniculum is attained, at which there is an abrupt increase in the rate of growth as reflected by the increase in the thickness of the growth lines. 6) Growth of th n+1 continues at a constant rate, even through the budding of th n+2, until the aperture is developed. 136 7) Thecal ontogeny continues after the development of the th n+1 aperture, and it involves the gradual construction of the genicular flange on the th n+2 geniculum. The proximal end of the prothecal fold is associated with the budding of th n+1. Its distal end is delimited by a ventral bend in the dorsal thecal wall at the level of the geniculum. As revealed by the growth lines, a major change in the thecal ontogeny occurs at the geniculum, and it is intimately associated with the formation of the distal part of the apertural excavation of th n. Morphology and Development of Cladia. Secondary branches, which are present in many specimens (Text-figs. 20a-b,g), consistently project from the right-lateral side of the stipes when they are viewed dorsally. The mode of branching is lateral, and not dichotomous, because the secondary branches diverge perpendicularly from the stipe, which continues in its original direction of growth. The secondary branches originate from the right-lateral side of the thecal apertures on the stipes, and an opening at the base of the branch can be seen on the inner wall of the thecal aperture. Because of their origin, the secondary branches can be termed "thecal metacladia." The "mother" thecae of the cladia are dicalycal because they give rise to the succeeding theca on the stipe, as well as the first theca of the cladia. None of the available specimens are complete to the first aperture on the cladia. The largest available specimen shows a 1.4 mm long cladia, which has an initial width of 0.07 mm. Among the available specimens, the "mother" thecae of the cladia are always of Group I type. However, not all Group I thecae show cladia, and the development of the cladia is delayed relative to the development 137 of the stipe. The theca shown in Text-figure 20e-f is on a stipe fragment in which the succeeding theca is at least partly developed. The size and extend of the genicular flange suggests that the thecal aperture is fully developed. A thin, tapering band of dark tissue, possibly micro- fusellar, extends from the right-lateral margin of the th n+1 genicular flange to the midpoint of the right-lateral margin of the th n thecal aperture. This band forms an opening in the dorso-lateral comer of the right side of the aperture. It is from this opening that a cladia develops (Text-figs. 20a-d,g). Growth lines could not be seen on the proximal part of the cladia; however, its transparent, brown color is similar to that of the "mother" theca, and this suggests that it is composed of fusellar tissue. Non-isolated Specimens. Only 25 of the more than 350 available specimens are large rhabdosomes that have proximal ends, relatively long stipes, and several cladia. The largest specimen (Text-fig. 21a-b) has a 9.1 mm long stipe with six cladia. The rest of the specimens are proximal ends bearing relatively short stipes with or without cladia and distal stipe fragments with cladia. The large rhabdosomes, which are mainly from the Pratt's Ferry section, display the .rhabdosome shape well, but the thecae are generally poorly preserved. In some cases, the thecae are well preserved in small specimens that represent distal fragments and proximal ends. The rhabdosome consists of two stipes and numerous cladia. The stipes initially diverge horizontally from the middle of the sicula, and near the aperture of the first theca on each stipe, they bend drastically. 138 In some specimens (Text-figs. 21a,d), the stipes have concave curvature relative to the sicula, so that they converge and cross above the sicula. Some specimens (Text-fig. 21e) show less curvature, and the stipes are reclined and diverge distally. In other specimens (Text-figs. 21c,f), both stipes show concave curvature relative to the sicula, but one stipe is ascending and the other is descending, so that together they are S-shaped. Because the cladia always project from the right side of the thecae on the stipes, the orientation of the thecae on the stipes determines the orientation of the cladia and vice versa. For example, in the specimen shown in Text-figure 21a-b, the thecae on stipe no. 2 present a dorsal aspect, whereas those on stipe no. 1 present a ventral aspect. Stipe no. 2 has six cladia, which arise from th 2^ to th 7^, and it presents a dorsal aspect and a uniform width until th 6 , after which it widens abruptly. Th 7^ presents a lateral aspect. The proximal 2 end of the cladia that arises from th 7 is overlain by the stipe. The abrupt widening of the stipe is produced by a change in its orientation so that a lateral aspect is presented. Stipe no. 1 has three cladia, which arise from th 2^, th 3^, and th 4^. It presents a ventral aspect and a uniform width until th 4^, after which it gradually widens. More distal thecae present a lateral aspect. For the sake of symmetry, stipe no. 1 would be expected to have six cladia. However, distal to th 4^, the stipe widens and the thecae present a lateral aspect. In this orientation, any cladia would project upward from the shale surface, and the absence of the three distal cladia is most likely due to the preserva- tional orientation of the stipe within the enclosing rock matrix. 139 The specimen shown in Text-figure 21d is from Calera and shows the same shape of the rhabdosome as that illustrated in Text-figure 21a. The distribution of the cladia as related to the orientation of the thecae on the stipes is the same. The short length of stipe no. 1 is due to its position in a different plane from that of stipe no. 2, and it is a preservational phenomenon. The specimen shown in Text-figure 21e shows an erratic distribution of the cladia. Stipe no. 1 is of relatively uniform width, and the thecae present a lateral aspect. The proximal ends of the cladia arise from the dorsal part of the thecal apertures and extend across the lateral stipe wall. Th 2 presents a lateral aspect and the proximal end of the cladia that originates from it is overlain by the stipe. Although• thecal apertures can not be seen, the distal part of stipe no. 2 probably presents a ventral aspect because of the uniform orientation of the cladia that arise from th 3^, th 4^, and th 5^. Text-figures 21f and 21c show the most widely-described shape of the Nemagraptus gracilis rhabdosome. In Text-figure 21f, the ascending stipe presents a dorsal aspect except for the second theca which presents a lateral aspect. The proximal end of the cladia that arises from the second theca is overlain by the stipe. The descending stipe also presents a dorsal aspect until the most distal cladia is attained, after which it presents a lateral aspect. A fifth cladia that is oriented down into the rock matrix could be present on the descending stipe, thus equalizing the number of cladia on the ascending stipe. In Text-figure 21c, both the ascending and descending stipe again present dorsal aspects as far as the 140 most distal cladia, after which the lateral aspect is presented along with an increase in the stipe width. Specimens representing proximal ends with short stipes and no cladia are common (Text-figs. 22a-b). Their stipes are, as a rule, reclined. Thecae are in some cases well preserved and are of Group I type. Distal fragments with and without cladia are extremely common (Text-figs. 22c-g). When the stipe presents a lateral aspect (Text-figs. 22d,f), cladia are not seen. When the stipe presents a dorsal or ventral aspect (Text-figs. 22e-g), cladia are present and arise from the right-lateral side of the thecal apertures. Systematic quantitative measurements are hard to obtain from the available specimens because of: 1) their fragmentary nature, 2) their poor preservation, 3) the large variability in the size of the rhabdosomes, and 4) the variability in the preservational orientations. The sicula is 0.75 - 0.85 mm long. The first thecal aperture on each stipe is 0.5 - 0.6 mm from the sicula. It consistently presents a lateral aspect with a maximum width of 0.14 - 0.18 mm. The stipes twist distally in large rhabdosomes to present a dorsal or ventral aspect. In this case, the stipe width is generally uniform or gradually widens, and it ranges from 0.08 - 0.16 mm. Proximal stipe fragements that present a lateral aspect are 0.14 - 0.18 mm wide. The stipe width increases significantly to 0.24 - 0.28 mm in distal stipe fragments that show a change from a dorsal or ventral aspect to a lateral aspect. In the largest available rhabdosomes, this change occurs at the sixth to eighth theca of each stipe, which is several thecae distal to the youngest cladia. 141 The distance between successive thecal apertures is listed in Table 3 for those specimens with proximal ends that are figured herein. The distance between thecal apertures increases greatly from the second to the third theca on each stipe, after which the distance is uniform or irregularly varies. The thecal density on the stipes ranges from 4.5-5 thecae in 5 mm for proximal stipe fragments on the basis of the measure ments that are listed in Table 3. The aperture of the second theca on each stipe bears the most proximal cladium. Distal to the aperture of the second theca, cladia arise from every theca on the stipe up to the most distal theca that bears a cladium, beyond which the stipes extend distally for at least several thecae. The longest cladium measures 7.5 mm and arises from the aperture of the second theca (Text-fig. 21a). The length of the cladia decreases distally along the stipes. As mentioned previously, the cladia originate consistently from the fight-lateral side of the thecal apertures on the stipes; thus, their orientation is dependent on the orientation of the stipes. Generally, the cladia diverge perpendicularly from the convex margin of the stipes, when the stipes are curved. The cladia are also curved, but their curvature may be concave or convex relative to their ventral margin, which consistently faces the proximal part of the rhabdosome (Text-figs. 21c, 22c,e,f,h). The cladia almost always present a lateral aspect because of their orientation to the stipes. At their point of origin, the cladia are 0.06 - 0.11 mm wide. They gradually increase in width distally to a maximum observed value of 0.42 mm, although 0.24 - 0.28 mm is the range within which most values fall. Some TABLE 3: Distance in mm between successive thecal apertures in specimens of N. gracilis. Text-fig. or th n th n thecal Museum number th 1-2 th 2-3 th 3-4 th 4-5 th 5-6 th 6-7 th 7-8 max. min. density 21c 0.91 1.08 1.08 1.08 1.00 1.00 0.75 - - 4.5/5mm 21f 0.75 1.08 1.08 1.08 1.00 -- - - 4.5/5mm 21e 0.83 1.25 1.08 1.08 - - -- - 4.5/5mm 21a-b 0.70 0.92 1.00 1.30 - - - - - 4.5/5mm 21d 0.83 1.25 1.08 1.08 1.08 1.33 1.08 - - 4.5/5mm 22h 0.77 1.10 1.10 1.10 - - -- - 4.5/5mm 22b 0.88 1.24 ------ AMNH 30457; Hall (1847) PI. 74, fig. 6a 1.26 1.26 4/5mm AMNH 30458; Hall (1847) PI. 74, fig. 6b 1.05 1.26 4-5/5mm AMNH 30459; Hall (1847) PI. 74, fig. 6c-d 1.40 1.40 3.5/5mm LO 2450t, LO 2451t Hadding (1913) 1.26 1.40 3.5/5mm YPM 20351 Berry (1960) 1.12 1.26 4.5/5mm AMNH 1043/2 Hall (1859) fig. 3 1.40 1.70 3-3.5/5mm AMNH 1043/2 Hall (1859) fig. 6 1.40 1.40 1.54 1.61 1.47 1.19 1.82 3-4.2/5mm 143 cladia (Text-fig. 22g) attain their first thecal aperture and a wider width within a shorter distance from their origin than other cladia (Text-fig. 22c). The specimens illustrated in Text-figures 21c and 22h show that the more distally located a cladium is along a stipe, the closer its first thecal aperture is to the stipe. Although thecal apertures on the cladia are poorly preserved, there appears to be an increase in thecal density and in thecal width distally along each cladium (Text-figs. 22f-g). In addition, the proximal thecae on distal cladia within the rhabdosome resemble the distal thecae on more proximal cladia. The gradual distal increase in dorso-ventral width and distal decrease in the distance between successive thecal apertures along the stipes and cladia allow positioning within the rhabdosome of the three groups of thecae that are described on the basis of isolated specimens. Group I thecae compose the initial parts of the stipes and the initial thecae on the most proximal cladia. Group II thecae are situated distal to Group I thecae within the rhabdosome. For example, the first theca on the third cladium of the specimen illustrated in Text-figure 22h, and the first thecae on the fifth cladium of the specimen illustrated in Text-figure 21c, as well as the seventh to ninth thecae on stipe no. 1 of the specimen illustrated in Text-figure 21a-b, resemble Group II thecae. The thecae on the cladium of the specimen illustrated in Text-figure 22g resemble Group III thecae. This cladium is probably situated in a very distal part of a rhabdosome. 144 Remarks Development of Proximal End. A proximal-end development, which shows 2 a right-handed origin of th 1 , three crossing canals, and a dicalycal th 2^ (Text-fig. 20h) , has not previously been described, although it may be the type of development that Bulman (1970) postulated as the leptograptid type. This development, which is here termed the nemagraptid type because it is here first described in Nemagraptus gracilis, is intermediate structurally between the isograptid type and the diplograptid-dicranograptid type. It can be easily derived from the isograptid type by a delay in the position of the dicalycal theca from th 1^ to th 2^. The diplograptid- dicranograptid type, especially as it is displayed by Leptograptus and Dicellograptus (elegans group) with a horizontal direction of growth of the earliest thecae, can be derived easily from the nemagraptid type by a simple change from a right-handed origin of th 1^ to a left-handed origin. The origin of th 1^ high on the metasicula in the nemagraptid type of development is intermediate between a position in the prosicula that occurs in the late stages of the isograptid type of development and a position in the middle of the metasicula that '.s found in the diplograp tid-dicranograptid type. Morphology and Ontogeny of the Thecae. The thecae of Nemagraptus gracilis, which are characterized by a small amount of overlap, a large length to width ratio, a simple aperture, and a very short metatheca, resemble those of several species of dichograptids, in particular those specimens described by Skevington (1965) as Didymograptus cf. nov. sp a 145 aff. gracilis, Dichograptid sp. a and Dichograptid sp. d. The growth line pattern of the latter two species illustrated by Skevington (1965) is similar to that of N. gracilis. However, the presence of a geniculum and a genicular flange clearly distinguishes the thecae of IJ. gracilis. Although a geniculum and a genicular flange are found in repre sentatives of Dicranograptidae, they develop somewhat differently. The development of the geniculum occurs simultaneously with thecal budding in the dicranograptids, whereas the two are not associated in Nemagraptus gracilis. The genicular flange is an integral part of the infrgenicular wall in Dicellograptus gurleyi gurleyi, and probably in D^. geniculatus, and it is constructed of fusellar tissue. In N. gracilis, it is "secreted" onto the geniculum and composed of microfusellar tissue. The change in the rate of thecal growth as interpreted from the thickness of growth lines occurs simultaneously with the development of the geni culum in N. gracilis, whereas it occurs well in advance of the development of the geniculum in the dicranograptids. Development of Cladia. The branching in Nemagraptus gracilis is lateral, and the branches are thecal metacladia. The development of thecal metacladia has been described previously only by Thorsteinsson (1955) for Cyrtograptus. It differs from that of N. gracilis because it involves the formation of a pseudovirgula. Thecal Gradients. The progressive change in thecal form within the rhabdosome that has been described for Cyrtograptus (Thorsteinsson, 1955; Bulman, 1958) appears to be present also in Nemagraptus gracilis. Evidence for this is: 1) the variability of thecal morphology, 2) the distribution 146 of variant forms within the rhabdosome, 3) the delayed origin and growth of the cladia relative to the stipes, and 4) the variation in the morphology between the initial theca on each of the cladia within a single rhabdosome. Urbanek (1973) used Cyrtograptus to verify his hypothesis of morphophysio- logical gradients. The progressive change in thecal form in N. gracilis serves as an additional example. Original Shape of Bhabdosome. The stipes of specimens of Nemagraptus gracilis preserved on shale surfaces are in most cases U-shaped with the stipes reclined and crossing distally or S-shaped with one stipe ascending and the other descending. In some specimens, the two stipes do not lie in the same plane. Each stipe consistently shows curvature that is concave relative to the proximal end, and the cladia diverge from the convex side of this curvature. The stipes in undeformed rhabdosomes must have had such a position (Text-fig. 20i) that deformation on the depositional surface would cause the stipes to compress together to form a U-shape or to separate and form a S-shape. The fact that at least one stipe is reclined in either shape suggests that both stipes were reclined in the undeformed rhabdosome. The fact that both stipes show consistent curvature, and overlap distally in U-shaped rhabdosomes, indicates that each stipe formed a low, broad helicoidal spiral in the original rhabdosome. Each stipe would have spiraled in a clockwise direction if the rhabdosome were viewed from the proximal end. The cladia were probably arranged in such a way that they radiated outward in relation to the longitudinal axis of the rhabdosome. Relative to the stipes, they were oriented in such a way that the directions, in which all the thecal apertures of a particular cladium opened, defined 147 a plane that coincided with the longitudinal axis of the theca on the stipe from which that cladium originated- Comparison to Other Species and Subspecies. Specimens of Nemagraptus from the Pratt's Ferry section were first identified by Ruedemann (1908, 1947). In 1908, he (p. 11-12) reported N. gracilis surcularis. In 1947, he (p. 80) listed a new subspecies, N. gracilis sub tenuis, which he did not describe or define, and N. exilis. The Alabama specimens described above exhibit a large variation in rhabdosomal shapes (e.g. Text-figs. 21a, e-f; 22b), and an attempt to identify these specimens by their shapes, which is the method used by Ruedemann (1908, 1947), Elies and Wood (1903), and Berry (1960), results in the recognition of at least two species and several subspecies of Nemagraptus. The undeformed shape of the rhabdosome of Nemagraptus gracilis is complex, and it can be deformed and oriented in the rock to produce a wealth of different shapes. Deformation on the surface of deposition can produce either a S or Ü shape of the stipes. The orientation of the stipes on the bedding surface, whether the lateral or dorso-ventral aspect is presented, will determine the presence (observability) or absence (non observability) of cladia. Variation in the orientation of the stipe can result in a variation in the position of the cladia on the stipe; for example, the cladia may only be present on the distal or proximal extremi ties of the stipe. Variations in the orientation of the stipes and cladia can also produce significant variations in stipe width; for example, the dorso-ventral width of a theca is greater than the distance between its lateral walls. Because the development of cladia is delayed relative to 148 the development of the stipes, small rhabdosomes may not show any cladia. The progressive change in thecal form within the rhabdosome results in significant quantitative variation between proximal and distal parts of the rhabdosome. A thorough study of all the preservational aspects of the Alabama material has enabled me to confidently combine all specimens exhibiting this variation of rhabdosomal shapes into one species. However, a problem arises in relating the Alabama material to previously described nemagraptids. Nemagraptus gracilis, several subspecies, and closely related species, are all found within a restricted biostratigraphic interval (£. teretiu- scuius and N. gracilis Zones) on all the continents except Africa and Antarctica. The best available descriptions of these species and subspecies and the descriptions on which the following discussion is based are those of Hall (1847, 1859), Ruedemann (1908, 1947), Elies and Wood (1903), and Hadding (1913). All the taxa within Nemagraptus display the same morpho logical features of the proximal end, thecae, and lateral branches. The diagnostic features of the various taxa are: 1) the shape of the rhabdosome as defined by the shape of the stipes, and the position and orientation of cladia, 2) the width of the stipes and the cladia, and 3) the thecal density on the stipes and cladia (Table 4). The variation of these "diagnostic" features in the Alabama material suggests that many recognized taxa of Nemagraptus are nonspecific. Hall's (1847) original type specimens of Nemagraptus gracilis, which are referred to taxonomically as N. gracilis gracilis by Strachan (1971) but for the sake of brevity are referred to as N. gracilis in the following TABLE 4: Diagnostic features used by previous workers for taxa of Nemagraptus Nemagraptid taxon shape of StiLoe cla dia (source of diagnostic features) rhabdosome width thecae width thecae in mm in 10 mm in mm in 10 mm N. gracilis (Hall, Ruedemann, Elies and Wood) S 0.2-0.6 6-7 0.4-0.8 8 subsp, approximates (Ruedemann) S It 10 II 10 subsp. crassicaulis (Gurley, Ruedemann) ?S 0.5-0.75 ?6-7 0 . 4-1.0 6-8 subsp. distans (Ruedemann) S 0.2 5 0.2 7 subsp. nitidulus (Elies and Wood) U; only 1 cladia 0.3 7 proximally located on each stipe subsp. remotus (Elies and Wood, Hadding) S; cladia only on 0.2-0.3 10 (E+W) 0.4-0.6 10-12 distal part of 12 (Had.) stipes. subsp. surcularis (Ruedemann, Elies and U: rhabdosome is 0.5 or 7 0.5 8 Wood) small. less N. exilis (Gurley, Ruedemann) Stipes horizontal 0.17- 5-8 (Rued) 0.5 8-10 to S-shapedj cladia 0.5 12 (Gurl) only distally. subsp. linearis (Ruedemann) tl 0.3 or 7-9 less N. explanatus (Elies and Wood) Stipes horizontal; 0.6-0.8 8 1 or 2 cladia, distally located subsp. pertenuis (Elies and Wood) It tt 6-7 N. subtills (Hadding) Stipes slightly 0.2 9-11 reclines, cladia absent. 150 discussion, are distal stipe fragments with several cladia present on each. In a later description. Hall (1839) described specimens that showed: 1) a S-shape, 2) a U-shape, and 3) a U-shape with no cladia, which he interpreted as a young rhabdosome. In 1868, he suggested that the U-shaped rhabdosomes may represent a different species, which he called surculairis. Ruedemann (1908, 1947), believing that the S- and U-shapes of the rhabdosome were produced by differences in the angle of divergence of the stipes in undeformed rhabdosomes, separated the U-shaped rhabdosomes as the subspecies surcularis. Because N. gracilis and N. gracilis surcularis differ only in the shapes of their rhabdosomes, which can easily be produced by deformation in different directions of the same original rhabdosome, they are here considered to be nonspecific. The described specimens of the subspecies crassicaulis (Gurley, 1896; Ruedemann, 1908, 1947) are only distal stipe fragments that are essentially identical to Nemagraptus gracilis except for being more robust. Because the rhabdosome of N^. gracilis shows progressive quantitative variation in stipe width and because the ranges of variation of the stipe and cladial widths of the two taxa overlap (Table 4), the variety crassicaulis is here considered to be nonspecific with N. gracilis. Specimens previously referred to this variety probably represent a distal stipe fragment and/or a rhabdosome that has been secondarily thickened to a considerable degree by cortical tissue. Because the cladia in the Alabama material arise from each successive theca on a stipe, the density of cladia along a stipe reflects the density of thecae along that stipe. This makes it possible to determine the 151 thecal density and the distance between successive thecal apertures on the stipes in specimens in which the orientation of the stipe obscures the thecal apertures. As shown in Table 3, the distance between successive cladia, which reflects the distance between successive thecal apertures on the stipe, in Hall's (1847, 1859) and Ruedemann's (1908, 1947) figured specimens of liemagraptus gracilis (AIINH 30457, 30458, 30459, 1043/2) ranges from 1.05 to 1.82 mm, which indicates a variation in thecal density from 5.5 to 9.5 thecae in 10 mm. The specimen (AMNH 1043/2), figured by Hall (1859, fig. 6) also shows that in this one specimen there is a variation that ranges from 6 to 8.4 thecae in 10 mm. Ruedemann (1908, 1947) established the two subspecies approximatus and distans on the basis of extreme values of thecal density on the stipes, which are 5 and 10 thecae in 10 mm, respectively. Measurements taken on Ruedemann's (1908, 1947) illustrated specimens show that distances between successive cladia on the stipes range from 0.8 to 1.5 mm for approximatus and from 1.0 to 2.0 mm for distans. These values indicate thecal densities that are as low as 6.5 thecae in 10 mm for app roximatus and as high as 10 thecae in 10 mm for distans. Thus, because the ranges of thecal density in the subspecies approximatus and distans overlap considerably with the range of variation found in populations of typical Nemagraptus gracilis, the two subspecies are not distinguished here. These two subspecies apparently represent only extreme variant forms of gracilis. The subspecies nitidulus, which was defined by Lapworth (1876) and described by Elies and Wood (1903), was established on the basis of 152 specimens in which the rhabdosome has a S-shape and cladia are present only at the distal extremeties of the stipes. One of their illustrated specimens (Elies and Wood, 1903; PI. 19, fig. 3f) shows that the stipes are in lateral view where there are no cladia and in dorso-ventral view where cladia are present. Hadding's (1913) illustrated specimens of Nemagraptus gracilis remotus, which are pyritized and in full relief, display exceptionally well this same relationship between the orientation of the stipes and the presence or absence of cladia. Elies and Wood (1903) describe the subspecies remotus as having 10 thecae in 10 mm on the stipes, while Hadding (1913) reports 12 thecae in 10 mm. New measurements, taken on proximal stipe portions of Hadding's specimens (LO 2450t, LO 2451t), show 5 thecae in 5 mm. On distal stipe fragments the distance between cladia ranges from 1.26 to 1.40 mm, which indicates a thecal density that ranges from 7 to 8 thecae in 10 mm. On the basis of the new insights on the preservation of the rhabdosome and the new measurements, the sub species remotus is here considered to be nonspecific with gracilis. Lapworth (in Gurley, 1896) described Nemagraptus exilis on the basis of specimens in which the stipes are slightly reclined to horizontal, show no cladia, and distally increase in width from 0.17 to 0.50 mm. As it was originally defined, the types of this species resemble closely those Alabama specimens that are preserved in such a way that they present a lateral aspect, as well as one of Hadding's (1913; PI. 4, fig. 6) figured specimens (LO 2450t). Ruedemann (1908), who first illustrated exilis, revised the description to include S-shaped rhabdosomes with distally located cladia. An examination of Ruedemann's figured specimens 153 (NYSM 7316-7318) shows that on those stipe portions with no cladia, the stipes present a lateral aspect and on those stipe portions with cladia, the stipes present a dorso-ventral aspect. Lapworth (in Gurley, 1896), reports 12 thecae in 10 mm for N. exilis, whereas Ruedemann reports 5 to 8 thecae in 10 mm. N. exilis, expecially as it is figured by Ruedemann (1947; PI. 61, fig. 5), must here be considered conspecific with N. gracilis. The subspecies N. exilis linearis Ruedemann, which is distinguished by a small number of distally located cladia and nearly straight stipes, is also here considered to be conspecific with N. gracilis. Lapworth (1876) defined Nemagraptus explanatus and its subspecies pertenuis. He distinguished them by their long, straight stipes and the absence of cladia, except for one or two on the distal extremities of N. explanatus. It is tempting to also assign this species to N^. gracilis because the stipes of the illustrated specimens present a lateral aspect, they occur with N. gracilis, and their quantitative variation overlaps with that of N. gracilis. However, the reported length of the stipes, 6 cm or more, and the scarcity of cladia make it difficult to interpret N. explanatus as a different preservational aspect of N. gracilis. Therefore, it is here retained as a distinct species. Nemagraptus subtilis, as defined and illustrated by Hadding (1913), closely resembles specimens of N. gracilis in which the stipes present a lateral aspect. In southern Sweden, this species extends a full zone below the first appearance of N. gracilis (Hede, 1951). In the lower part of its range, specimens are common, and if specimens referred to 154 subtilis represented only a preservational aspect of specimens of N. gracilis, it would be expected that some specimens would show cladia. However, none do, and therefore N. subtilis is here considered to be a distinct species. Kozlowski (1954) described isolated stipe fragments, which he identi fied as Genus? incisus. Although the specimens are extremely fragmentary, the thecal morphology and the lateral branches are identical to those described in Alabama specimens of Nemagraptus gracilis. As here revised, the concept of Nemagraptus gracilis is now broadened to include several species and subspecies, which were previously established on what is inferred as different preservational aspects of specimens of the one and the same species. As can be seen in Table 4, the range of variation of the "diagnostic" quantitative features has been increased in the revised definition; for example, the range of variation in thecal density is increased from 6-7 to 5-10 thecae in 10 mm for the stipes and from 8 to 7-10 thecae in 10 mm for the cladia. As previously discussed, there are progressive quantitative changes between the proximal and distal portions of the rhabdosome, especially as reflected by the distance between successive cladia along a stipe, and the increase in the range of variation for the revised concept of Nemagraptus gracilis may merely reflect the discovery that it does exist. On the other hand, N. gracilis has a worldwide distribution, and geographic variants would be expected. The specimens from Alabama, for example, show a thecal density on the stipes that consistently falls at the high end of the range of variation for the revised gracilis, and this 155 thecal density is similar to that found in the subspecies approximatus. However, the variation for both typical specimens of approximates and those in the Alabama collections overlaps significantly with that found in Hall's and Ruedemann's specimens of 1^. gracilis, and the specimens of the subspecies approximatus occur at the same localities as N. gracilis, namely Glenmont and Normanskill in New York State. The Alabama material and the previously figured specimens of N. gracilis discussed here are not preserved in such a way that the progressive quantitative changes can be systematically studied throughout the rhabdosome. Without such a study, recognition of subspecies on the basis of slight differences in the ranges of variation of quantitative features seems premature. The diagnosis given above for Nemagraptus gracilis is for the species as it is revised herein. The Alabama material shows thecal densities that consistently are at the high end of the range of variation. Genicula and genicular flanges have not been reported previously for N. gracilis, as it is here revised. This most likely is due to the fact that previously described specimens are preserved as carbon films and pyritized internal molds. An examination of Berry's (1960) figured specimen (PYM 20351) shows that it does have genicula and genicular flanges. An examination of Hadding's (1913) figured specimens (Ui 2540t - 2451t) also shows the presence of genicula. Figured Specimens OSU 32922-OSU 32972 156 Genus Amphigraptus Lapworth, 1873 Type Species: Amphigraptus divergens (Hall, 1859) Diagnosis Rhabdosome consisting of two uniserial, horizontal to reclined, stipes and thecal metacladia. Cladia simple or compound, typically paired, but occasionally unpaired. Species Described Herein Amphigraptus n. sp. A Amphigraptus n. sp. B Discussion With the inclusion of the two following described species into Amphigraptus, the diagnosis of that genus given by Bulman (1970) has been expanded to include a greater variation in the shape of the rhabdosome, namely, the reflexed stipes of Amphigraptus n. sp. B and the distally restricted cladia of Amphigraptus n. sp. A. Most of the morphological features of this genus closely resemble those of Nemagraptus; however, Amphigraptus can be distinguished by the presence of paired cladia. 157 Amphigraptus n. sp. A (Text-figure 23a) Diagnosis A species of Amphigraptus with two horizontal, uniserial stipes from which curved thecal metacladia are produced distally and in pairs. Thecae of Nemagraptus type. Material The available material consists of a single specimen, which was collected from a horizon 76.4 metres above the base of the Pratt's Syncline section. It is compressed on a shale surface and poorly preserved, but a few thecae can be seen on the cladia. Description The sicula, which displays a virgella, is 0.84 mm long and 0.11 mm wide. Two straight stipes diverge horizontally from the midpoint of the sicula. Stipe no. 2 is 3 mm long, but it is too poorly preserved to show any thecae. Stipe no. 1 is 14.35 mm long, but the first 7 mm are only represented by short, discontinuous fragments of periderm. The aperture of th 1^ is 0.4 mm from the sicula, and at this point the stipe is 0.17 mm wide and presents a lateral aspect. For the rest of its length, stipe no. 1 shows a uniform width of 0.14 - 0.17 mm. At the base of one of the distal cladia, a small part of a thecal aperture can be detected, and it indicates that the stipe is oriented in such a way that it presents a dorsal aspect, which accounts for the uniform stipe width. The cladia are generally paired and curved distally. Where they can be observed, the thecal apertures are on the concave side of the 158 EXPLANATION OF TEXT-FIGURE 23 Text-figure 23 a. Amphigraptus n. sp. A. Distally stipe no. 1 presenting dorsal aspect. X9. PS-76.4. OSU 32973. 23 b-d- Amphigraptus n. sp. B. b. Obverse aspect of rhabdosome. X9. C-11.8. OSU 32974. c. Reverse aspect of rhabdosome. X9. C-11.8. OSU 32975. d. Reverse aspect of rhabdosome. Note orientation of paired cladia on stipe no. 2. X9. C-11.8. OSU 32976. un VO Text-figure 23 160 cladia. The first cladium originates on the stipe at a distance of 6.5 mm from the sicula. Because the specimen is poorly preserved at this point, the presence of a second cladium arising from the opposite side of the stipe can not be determined. The next point of cladial origin is 1.5 mm beyond the first, but here again only one cladium is present, and the absence of a second, opposing cladium may be due to the state of preservation of the specimen. For the remainder of its length, the stipe gives rise to paired cladia at regular 1.0 - 1.2 mm intervals. The relationship of the thecal aperture on the stipe to the base of the cladium adjacent to it suggests that all cladia arise from thecal apertures in the manner described for Nemagraptus gracilis. If, as in N. gracilis, cladia arise from every theca on the stipe that is distal to the theca that bears the most proximal cladium, then the thecal density on the stipe can be determined by the density of cladia along the stipe. Calculated in this way, the thecal density for the distal portion of stipe no. 1 is 8 - 10 thecae in 10 mm. In most instances, the cladia present a lateral aspect, and thus reveal the thecal morphology. The thecae resemble closely those of Nemagraptus gracilis in the following features: 1) The dorsal and ventral walls are subparallel and straight. 2) A genicular flange projects from a distinct geniculum. 3) The supragenicular wall is relatively long in comparison to the infragenicular wall. 4) The thecal overlap is small, approximately a fourth. 161 5) The aperture is perpendicular to the stipe axis and does not show introversion or introtorsion. The cladia are 0.07 mm wide at their origin, and they increase in width distally to a maximum of 0.28 mm. The distance between suc cessive thecal apertures is 0.77 ram, which indicates a thecal density of 13 thecae in 10 mm. Remarks The morphology of the proximal end of the specimens described above resembles closely that of Nemagraptus gracilis, thus suggesting a similar proximal-end development. The thecae are morphologically nearly identical to those of gracilis, and the orientation and position of the cladia relative to the stipes suggest that the cladia originated in a manner that is similar to that found in N. gracilis. However, the cladia are paired, which suggests that each theca on the stipe gave rise to three thecae, namely the succeeding theca on the stipe and the initial theca of a cladium from each of the lateral sides of its aperture. Although the proximal portions of the stipes of the described specimens are poorly preserved, careful observation strongly suggests that cladia are lacking proximally and are only present distally. The straight stipes and paired cladia indicate that the described speci mens belong to the genus Amphigraptus, but the arrangement of the cladia differs from that found in all other knovm species of the genus. The specimens described above undoubtably represent a new species, but it is not formally named here because of the paucity of available specimens. Figured Specimen OSU 32973 162 Amphlgraptus n. sp. B (Text-figures 23 b-d) Diagnosis A species of Amphigraptus with two reflexed stipes from which metacladia are given off singly from th 1 ^ and th 2 ^ and in pairs 2 from th 1 . Thecae of Nemagraptus type. Material The available material consists of 17 specimens, which were collected from the lower part of the Calera section. All the specimens are preserved as carbon films. Description The sicula, which displays both a nema and a virgella, is 1.05 mm long and 0.15 mm wide. Two stipes diverge horizontally in opposing directions from the midpoint of the sicula. At the first thecal aperture, each stipe bends upward in such a way that 1 ) the dorsal walls of the horizontal and ascending parts of stipe no. 1 enclose an angle of 140 to 145 degrees ; 2) the dorsal walls of the horizontal and ascending parts of stipe no. 2 enclose an angle of 115 to 120 degrees (a single exception is shown in Text-fig. 23d); and 3) the dorsal walls of the ascending parts of both stipe no. 1 and stipe no. 2 enclose an angle of 75 degrees. More distally, the stipes show relative to their ventral margins gentle concave curvature, and the angle that their dorsal margins enclose increases. Both stipes show a generally uniform width of 0.28 mm. 163 Gently curved branches, which can be termed thecal metacladia because of their origin, are given off in pairs from the aperture of 2 1 3. th 1 and singly from the apertures of th 1^ and th 2 . The paired 2 cladia appear to arise from the opposing, lateral sides of the th 1 aperture. The unpaired cladia appear to arise from the same lateral sides of the th 1 ^ and th 2 ^ apertures, but it can not be determined whether they arise from the right or left side. The cladia generally project downward relative to the stipes, except for one of che paired cladia in the specimen that is illustrated in Text-figure 23d, which projects upward. The largest available rhabdosome, which has a 14.5 mm long stipe and a 10 mm long cladium, shows no other additional cladia than those described above. The cladia are 0.28 mm wide at their origin; distally, they show a uniform width of 0.4 - 0.5 mm. The thecae are hard to discern because of the state of preserva tion and because the stipes and cladia generally are compressed obliquely. The few observable thecae show that 1) the apertural margin is perpendicular to the axis of the stipe or cladium; 2 ) the ventral thecal wall is subparallel to the axis of the stipe or cladium; and 3) a geniculum is present. Thus, the thecae resemble the Nemagraptus type. They number 3.5 - 4 in 3 mm on both the stipes and cladia. Remarks The morphology of the proximal end of the described specimens closely resembles that of Nemagrantus gracilis, which suggests a similar mode of development. The thecae appear to be of the Nemagraptus type, and the origin of the cladia from thecal apertures probably occurs in the same manner as that described for îî. gracilis. 164 The shape of the rhabdosome does not readily conform to any of the described genera of Nemagraptidae in the arrangement of the cladia and the orientation of the stipes, but it is most similar to that of Amphigraptus divergens radiatus Lapworth in which the cladia are restricted to the proximal end of the rhabdosome and are unpaired. Syndyograptus Ruedemann is defined as having reclined stipes, but its cladia are always paired and its proximal end, which is figured by Ruedemann (1908; p. 267, fig. 185). differs significantly from that of the specimens described above. Although the Alabama material most likely represents a new species of Amphigraptus, which is the earliest one knovm, it is not formally named here because the poor state of preservation does not adequately reveal the thecal morphology. Figured Specimens OSU 32974 - OSU 32976. 165 Nemagraptid sp. A gen. indet. et sp. nov. (Text-figures 24a-m) Diagnosis Rhabdosome with two reclined stipes and questionable cladia or secondary branches. Nemagraptid proximal-end development. Thecae of Nemagraptus type but modified by introverted aperture and genicular flange formed by fusellar tissue. Material The available specimens are all very fragmentary. They include 30 proximal ends and 28 distal stipe fragments, which were isolated from the rock matrix, and one distal stipe fragment, which is compressed on a shale surface. All specimens were collected from horizons 76.4, 109.2, and 109.7 metres above the base of the Pratt's Syncline section. Description The largest available specimen with a proximal end has a 1.6 mm long stipe that consists of three thecae (Text-fig. 24a). As discussed for Nemagraptus gracilis, the morphology and development of the proximal ends of gracilis and the specimens described here are so similar that 1 2 they can only be distinguished by the shapes of the th 1 and th 1 apertures and by the amount of overlap of th 2 and th 3. Text-figures 24a-b, e-f are proximal ends of the specimens described here. The sicula is 0.75 - 0.85 mm long and 0.20 mm wide. Its aperture is characterized by the presence of a virgella and lateral lappets. The development of the first four thecae is the same as that described for 166 EXPLANATION OF TEXT-FIGÜRE 24 Text-figure 24 a-m. Nemagraptid sp. A gen. indet. et sp. nov.. Isolated specimens, except that in Text-fig. 24 i, which is compressed and non-isolated. a. Reverse aspect of proximal end. Note th 4^ infra- genicular wall. Sicula resembles closely that of Nemagraptus gracilis, but note introverted thecal apertures and thecal overlap. X37.5. PS-109.2. OSU 32977. 1 O 1 b. Reverse aspect of sicula with th 1 , th 1 , th 2 , and th 2^. Proximal-end morphology resembles closely that of Nemagraptus gracilis. X37.5. PS-109.2. OSU 32978. c,d. Right- and left-lateral aspects of theca. Note change in growth-line pattern at geniculum. Genicular flange is composed of fusellar tissue. X37.5. PS-109.7. OSU 32979. e. Reverse aspect of proximal end. Note introverted thecal apertures and overlap of th 2^ and th 3^. X37.5. PS-109.2. OSU 32980. f. Reverse aspect of proximal end. Note curvature of stipes. X37.5. PS-109.2. OSU 32981. g. Stipe fragment. Note orientation of growth lines. X37.5. PS-109.7. OSU 32982. h. Stipe fragment. Note contrast in growth lines of metatheca and succeeding protheca. X37.5. PS-109.2. OSU 32983. i. Compressed, non-isolated specimen representing distal stipe fragment. X18.8. PS-117. OSU 32984. j. Distal stipe fragment. Note orientation of growth lines and that growth lines on infragenicular wall are continuous with those on the lateral wall. X37.5. PS-109.7. OSU 32985. k. Distal stipe fragment. X37.5. PS-109.7. OSU 32986. 167 l,m. Left- and right-lateral aspects of distal stipe fragment with base of possible cladium (secondary branch) projecting from dorsal stipe margin. X37.5. PS-109.2. OSU 32987. n V Ml-'- / \ MI.ZT 89T 169 Nemagraptus gracilis. Th 1 originates high on the metasicula. Th 1 has a right-handed origin from th 1^. There are three crossing canals. Th 2^ is dicalycal, and the first four thecae diverge from the middle of the sicula. Generally, the most proximal parts of the stipes that consist of the first four thecae of the rhabdosome show convex curvature relative to the ventral walls of the thecae. Distal to the first thecal aperture on each stipe, the stipes are straight. Thus, the reclined habit of the stipes is produced by curvature of the first four thecae, and this curvature aids in distinguishing proximal ends of the specimens described here from proximal ends of Nemagraptus gracilis in which the first four thecae generally are straight and horizontal. Th 1^ and th 1^ have subapertural mesial spines, which are also found in N. gracilis. However, 1 9 1 n the th 1 and th 1 apertures are introverted, and th 3 and th 3 1 2 originate within 0.2 - 0.3 mm of the th 1 and th 1 apertures, respectively, vdiich further aids in the differentiation of N. gracilis and the specimens described here. The largest available stipe fragment isolated is 1.8 mm long and has 3 thecal apertures (Text-fig. 241,m). This specimen shows what appears to be the base of a cladium or secondary branch projecting from the dorsal wall of the stipe. The specimen is not preserved well enough to determine if the "branch" originates from the dorsal or lateral wall of the stipe; however, it does not originate from the immediately under lying thecal aperture on the ventral margin of the stipe. A 3.4 mm long stipe, which is the largest available stipe fragment, is compressed on a shale surface and consists of three thecae (Text-fig. 24i). The 170 stipes are 0.17 - 0.20 mm wide at th 1 and 0.28 mm wide in the widest of the available distal fragments. Proximally, the distance between successive thecal apertures is 0.5 mm, which indicates that the thecae number 10 in 5 mm proximally. The maximum distance between successive thecal apertures among the available specimens in 1.13 mm. This measurement, which was taken on the specimen (Text-fig. 24i) that showed the largest width, indicates that there are 4.5 thecae in 5 mm distally. The thecae increase in length from 0.95 mm at th 3 to an observed maximum of 1.68 mm. They overlap for almost half their length, which results in prothecal and metathecal segments of nearly equal length. A distinct geniculum divides the free ventral wall into a relatively long supragenicular wall and a relatively short infragenicular wall. A genicular flange is absent in proximal thecae, but it is usually well developed in distal thecae. At its greatest development, the genicular flange barely extends onto the lateral stipe walls (Text-figs. 24c-d, k). It is composed of fusellar tissue, and in most specimens it bends downward (Text-fig. 24c-d) to form a "veil" around the geniculum or upward to form a "wall" around the geniculum. For most of its length, the supragenicular wall is straight to slightly convex and generally parallels the dorsal stipe wall. Distally, it shows very convex curvature, which results in an introverted aperture. Raised, angular lappets are present midway along each lateral margin of the aperture, and they are directed dorsally. Undulations, which represent prothecal folds, are present on the dorsal stipe margin midway between successive thecal apertures on the ventral stipe margin. 171 Grooves, which represent the traces of the interthecal septa, extend along the lateral stipe walls from the bases of the apertural excavations to the prothecal folds. These grooves are oriented in a dorso-proximal direction relative to the axis of the stipe. At their distal ends, the grooves bend sharply in a dorsal direction. At their intersection with the dorsal stipe wall, they define the proximal limit of the prothecal folds. The initial part of the dorsal wall of a prothecal fold is oriented in a dorso-distal direction, and more distally it is oriented in a ventro- distal direction (Text-figs. 24a-h). This change in the direction of growth of the dorsal wall produces the prothecal fold. Distal to the prothecal fold, the dorsal wall is inclined to the interthecal septum in such a way that the dorso-ventral width of the theca increases distally. The greatest width is attained at the level of the budding of the succeeding theca. As revealed by growth lines, the fuselli on the lateral walls of the initial part of the protheca are relatively thin and are oriented at an angle to the stipe axis (Text-figs. 24c-d, h). The same thickness and orientation is maintained distally to the level of the geniculum, after which the fuselli are significantly thicker and oriented perpendicular to the stipe axis. This latter thickness and orientation is maintained distally through the rest of the protheca and the metatheca. The change in the thickness and orientation of the fuselli within a theca is associated with the development of the geniculum and the genicular flange. The growth line pattern on the interthecal septum can not be seen in any of the available specimens, and none of the 172 available specimens show the rate of growth of the th n+ 1 protheca relative to the th n metatheca (i.e., whether or not the th n metatheca grows in advance of the th n+1 protheca). Thus, the theca that secretes the inter thecal septum can not be definitely determined. The growth lines on the infragenicular wall are continuous with those on the lateral walls of the th n+ 1 protheca which indicates that the infragenicular wall forms as a ventral wall of the th n+1 protheca (Text-figs. 24a-b, g). The orientation of the growth lines indicates that the infragenicular wall developed in advance of the dorsal wall of the th n+1 protheca. In order to accomplish the change in the orientation of the growth lines at the geniculum the dorsal thecal wall must have grown distally which the ventral wall did not. This appears to have been accomplished by the addition of fuselli in such a way that their ventral margins, which form the infragenicular wall, grew in a ventral direction. In this way, the dorsal wall could grow distally until it was transverse to the ventral wall, after which additional fuselli and the resulting growth lines would be perpendicular to the stipe axis. The change in the direction of growth of the ventral thecal wall (from ventral to distal) is abrupt and produces the geniculum. If ventral growth of the infragenicular wall is excessive, which appears to be the case in distal thecae, a ventrally projecting flange is developed because the first fuselli oriented perpendicular to the stipe axis will not necessarily extend to the distal end of the infragenicular wall. The following interpretations of thecal ontogeny are based on the described growth line evidence and comparisons with similar features in Dicellograptus, Leptograptus, and Nemagraptus. 173 1) Th n+1 buds fromth n by meansof an opening that is represented by the most proximal and dorsally oriented part of the trace of the interthecal septum. This opening is probably a primary notch. 2) The protheca initially grows dorsally out of the primary notch, then bends and grows distally. This change in the direction of growth produces the prothecal fold. 3) The ventral wallof the th n+1protheca develops in advance of its dorsal wall. Although the thickness of the growth lines indicates a faster rate of growth for the th n metatheca, the orientation of the growth lines in the th n+ 1 protheca and the advanced growth of its ventral wall allows for the consideration that the interthecal septum developed as the ventral wall of the th n+1 protheca. The fact that the interthecal septum and the infragenicular wall are continuous and that the th n aperture is not introtorted also permits this interpretation. 4) A marked change in the thecal ontogeny occurs at the level of the geniculum. This is shown by the change in the thickness and orientation of the fuselli, which is associated with the development of the geniculum and the genicular flange. 5) Distally, th n+1 increases in dorso-ventral width until a maximum is attained at the origin of th n+2 from th n+1. More distally, the dorso-ventral width of the th n+ 1 metatheca gradually decreases until the aperture is developed. 1 7 4 Remarks The original rhabdosome probably consisted of two reclined stipes and an unknown number of secondary branches. However, the mature habit of the rhabdosome, which involves the distal shape and orientation of the stipes, the origin of the branches and whether or not they are cladia, and the distribution and orientation of the branches, can not be determined because of the fragmentary nature of the available specimens. The thecae of the described specimens are similar to those of Nemagraptus gracilis in the case of their overall shape, the presence of a geniculum, the development of the prothecae and the prothecal folds, and the pattern of the growth lines. The thecae described above differ from those of N. gracilis in their introverted aperture and the nature and development of their genicular flanges. The development of the genicular flange as an extension of the infragenicular wall in the described specimens is the same as in some dicraograptids, for example Dicellograptus gurleyi gurleyi. However, the thecae described above differ from those of the dicellograptid type in their overall proportions, their lack of introtorsion (although there also is no introtorsion in Leptograptus thecae), and the position of the change in their growth line pattern. These thecae represent in the genicular flange and intro verted aperture a slight modification of the Nemagraptus type of theca. The dicellograptid type of theca can be derived from them by change in their overall proportions and by an advance relative to the geniculum of the position of the change in the growth line pattern. 175 The specimens described above are assigned to the family Nemagraptidae because the proximal end morphology and development and the basic morphology and ontogeny of the thecae are very similar to that described for Nemagraptus gracilis. The specimens can not be given a generic assignment because the mature habit of the rhabdosome is unknown. The fact that the one questionable secondary branch does not originate at a thecal aperture, and thus is not a cladium, suggests that the described specimens represent a new genus. A shape of the thecae similar to that in this species has previously not been described for any other species. Because the thecae are described in some detail here, they can be used to distinguish the available specimens at the specific level. Figured Specimens OSU 32977-OSU 32987 176 Family DICRANOGRAPTIDAE Lapworth, 1873 Diagnosis Rhabdosome tmiserial or combined uni- and biserial, reclined or initially scandent, without branches except in centribrachiate lepto- graptids; thecae with conspicuously sigmoid curvature, in some species elaborated; development of diplograptid type (essentially as in Bulman, 1970), Discussion Leptograptus is included in the Dicranograptidae because well preserved specimens show a proximal-end development, thecal morphology, and thecal ontogeny that are similar to those shown by other genera of this family. In addition, Dicellograptus is subdivided into two groups that are differentiated by the direction of growth of the first four thecae. Genus Dicellograptus Hopkinson, 1871 Type species: Didymograpsus elegans, Carruthers, 1868 Diagnosis Rhabdosome with two reclined uniserial stipes; thecae with distinct geniculum; apertures introtorted and introverted in many specimens; infragenicular wall of th n+ 1 originates as apertural flange of th n. Prothecal folds present in well-preserved specimens of all species. Species and Subspecies Described Herein: Dicellograptus alabamensis Ruedemann, 1908 Dicellograptus bispiralis bispiralis (Ruedemann, 1947) Dicellograptus bispiralis n. ssp. A 177 Dicellograptus geniculatus Bulman, 1932 Dicellograptus gurleyi gurleyi Ruedemann, 1908 Dicellograptus gurleyi n. ssp. A Dicellograptus sextans (Hall, 1847) Discussion Except for Dicellograptus geniculatus, rhabdosome of all the species described below show an overall upward direction of growth of the first four thecae in such a way that the dorsal walls of the prothecae of th 2 ^ 2 and th 2 are in contact with the sicula. Additionally, the first two thecae grow down the sides of the sicula to the sicular aperture, then bend and grow outward and upward, thus defining a U-shape in lateral view. The sicula is inclined toward and in contact with the dorsal wall of stipe no. 2, except in the earliest representatives of D. bispiralis, sextans, and gurleyi. In contrast, several previously described, and generally stratigraphically younger species, of Dicellograptus, such as JD. forchammeri (Geinitz), complanatus Lapworth, and johnstrupi Hadding, erfiibit: 1 ) a primarily horizontal direction of growth of the first four thecae in such a way that the dorsal walls of the th 2^ and th 2 ^ prothecae are not in contact with the sicula; 2) a L-shape of the first two thecae because they grow down the sides of the sicula to the sicular aperture, then bend and grow directly outward; and 3) a sicula that generally is oriented upward symmetrically between the stipes. On the basis of proximal-end morphology, dicellograptids are here tentatively separated into two groups, which, after more detailed study, might be assigned formal taxonomic status. The stratigraphie distribution of all the described 178 species of Dicellograptus from North America and Europe are listed in Table 5, which also shows the group assignment of each species. The vagus group includes those species in which the first four thecae show a primarily upward direction of growth and the first two thecae are U-shaped. Stratigraphically older subspecies and variant forms exhibit a more upward-oriented sicula and, in some cases, a wider U-shape of the first two thecae. They can be related to typical vagus group species because of the existence of morphologically intermediate forms, and they provide evidence about the ancestry of the vagus group (Text-fig. 25). Except for anceps, which may indicate a polyphyletic origin of the group because of its isolated stratigraphie position and its upward-oriented sicula, species of the vagus group are restricted to the G^. teretiuscuius and N. gracilis Zones (Table 5, Text-fig. 25). In the orientation of the first four thecae, the vagus group resembles Dicranograptus well enough to have served as its direct ancestor (See Remarks under Dicranograptus). The elegans group includes those species in which the first four thecae show a primarily horizontal direction of growth and th 1 ^ and th 1 ^ are L-shaped. This group appears in the N. gracilis Zone and extends through the rest of the Ordovician (Table 5, Text-fig. 25). In the shape of the sicula and the orientation of the first four thecae, the elegans group closely resembles, and may thus be closely related to, Leptograptus. Dicellograptus geniculatus, in which th 1^ shows a broad U-shape and th 1 ^ shows a horizontal direction of growth, is not assigned to either group of Dicellograptus. This species is the oldest known dicellograptid, and it might represent the "ancestral stock" of Dicellograptus (Text-fig. 25). 179 EXPLANATION OF TABLE 5 Classification of Dicellograptus species in terms of the vagus and elegans groups. Sources of morphological data: 1-5, 20, Toghill (1970a); 4, 6-7, Skoglund (1963); 8-10, 12-13, 16-17, 23, 26, Elies and Wood (1904); 11, 15, Hadding (1915b); 14, Ruedemann (1947); 18-19, 21, 24, 28-29, my Alabama material; 22, Bulman (1932); 25, James (1965); and 27, Hadding (1913). TABLE 5: Distribution of Dlcellof.raptiis among vagus and elegans groups. sicula sicula L-shaped U-shaped Prothecal Rnerloa upward Inclined thecae thecae fold Group Zones 1.D. anceps (Nicholson) X X vagus? 1 2. ornntus ornatua Elies and Wood X X elegans anceps 3. ornatus minor Toghill X X elegans 1 4. compJnnntiis Lapworth X X X elegans complanatus 5. carruthers 1 Togiilll XX elegans T 6. iohnstriipl Hadding X XX elegans linearis 7. morrisL Hopkinson X X XT elegans 1 8. caduceus Lapworth X Î elegans T 9. elegans elegans (Carruthers) XX elegans 10. elegans rigens Lapworth in Elies and Wood X X elegans cllnganl 11. pi.rallls Lapworth X X elegans 1 12. forchammeri flexnosus Lapworth X X elegans 1 13. angulatus Elles and Woos X X elegans multiden» 14. mensiirans Ruedemann X X elegans T 15. forrhai-.miorl forcliammerl (Gelnltz) X X elegans 1 16. iilosiis I.apworth X X elegans gracilis 17. divaricates rlgldus Lapworth X X . vagus 1 18. gurleyi gurleyi Ruedemann X X X vagus 1 19. bispiralis bispiralis (Ruedemann) X X X vagus 20. rooffatensls (Carruthers) X X elegans 21. , alabamensis Ruedemann X X X vagus 1 22. geniculatus Bulman X XX X . 23. intortus Lapworth X X vagus 1 24. sextans (Hall) X X X X vagus 25. divaricates divaricates (Hall) X X X vagus 26. divaricates saloplensls Elies and Wood XX X vagus 27. vagus Hadding X X X vagus ceretlusculus 28. gurleyi n.ssp. A * X X X vagus 29. bispiralis n.ssp. A X XX vagus 0 5 C 181 EXPLANATION OF TEXT-FIGURE 25 Proposed phylogenetic relationships within Dicranograptidae. Abbreviations: Dg, Dicellograptus geniculatus; evg, early representatives of vagus group; vg, vagus group; Dicr, Dicranograptus; eg, elegans group; Da, Dicellograptus anceps; Lept, Leptograptus. 182 anceps complanatus V linearis clingani Lept Dicr multidens gracilis / Vevg teretiusculus Text-figure 25 183 Although the vagus and elegans groups appear to be distinct, and in spite of the fact that the former resembles Dicranograptus and the latter resembles Leptograptus, they are both still assigned to Dicellograptus because of similarities in thecal morphology, thecal ontogeny, and prothecal folds as described and interpreted below. The section at Calera is the only known continuously graptolitiferous sequence in which the eariest occurrence of Dicellograptus can be examined. In this sequence, D. geniculatus, with its upward-oriented sicula, appears before any other dicellograptid. Its appearance is quickly followed by that of subspecies or variant forms of vagus group species, which occupy a morphologically intermediate position in the sicular orientation and the shape of the first two thecae between geniculatus and typical vagus group species (Text-fig. 25). These taxa are 2» bispiralis n. ssp. A, D. sextans, and 2* gurleyi n. ssp. A. Slightly higher stratigraphically appear Dicranograptus and taxa of the vagus group in which the sicula is inclined, or in contact with, stipe no. 2 . Because of their higher stratigraphie position, Leptograptus, additional dicranograptids, and the elegans group may each have originated indepen dently from 2 - geniculatus and the vagus group. Description of Rhabdosomal Development The rhabdosomal development, which involves proximal ends, prothecal folds, interthecal septa, and infragenicular walls, is remarkably similar in the following four species of Dicellograptus for which isolated specimens are available: 2 * alabamensis, 2 * bispiralis bispiralis, 2- gurleyi gurleyi, and 2* sextans. Although available specimens of the 184 two newly described subspecies are preserved only as carbon films, rhabdosomal development of these two taxa is interpreted as agreeing closely with that described on the basis of the isolated specimens of the other taxa because of apparently close taxonomic affinities between these forms. Rhabdosomal development of 2» geniculatus differs from that of the other described species in development of the proximal end, in which the dorsal walls of the th 2 ^ and th 2 ^ prothecae are not in contact with the sicula. D. geniculatus is similar to the other described species in the development of prothecal folds, interthecal septa, and infragenicular walls. Development of Proximal End. The earliest growth stages (sicula + th 1^ + th 1^) of Dicellograptus alabamensis, ■ bispiralis, 2» gurleyi, and D. sextans are difficult to differentiate specifically because of great similarités in morphology and proximal-end development in these species. Some distinguishing morphological features are noted in each species description. The dicellograptid fauna of many graptolitiferous horizons, especially in the Pratt Syncline section, consists on only one species. The early growth stages obtained from such horizons are confi dently assigned to the same species as that represented by large rhabdo- somes at the same horizon. Pyritized internal molds. Pyritized internal molds readily reveal the development of the proximal end because they occupy the space within the original thecal and sicular walls. A pyritized internal mold of Dicellograptus gurleyi gurleyi was isolated and is illustrated in Text- figures 41h-i. This specimen shows a proximal-end development that is also characteristic of alabamensis, D. bispiralis bispiralis and D. sextans. 185 In these forms, th 1^ originates high up on the virgellar side of the sicula. It grows downward along the sicula to the sicular aperture where it bends and grows outward and upward to form a U-shaped tube. The origin of th 1 ^ is left-handed because it develops from the obverse side of th 1^. Th 1^ grows upward over and across the initial part of th 1^ to the reverse side of the sicula. It then grows obliquely down ward across the sicula to the anti-virgellar margin of the sicular aperture where it bends outward and upward to form a U-shaped tube. The origin of th 2^ is left-handed. It develops from the virgellar side of th l2 where th 1^ begins its downward direction of growth. Th 2^ grows toward the sicular aperture, then bends outward and upward filling in the space between the limbs of the U-shaped th 1^. The origin of th 2^ from th 2^ is right-handed. It develops from the anti-virgellar side of th 2^ where th 2 ^ changes from a downward to an outward direction of growth. Th 2^ grows horizontally across the downward-directed part of th 1^, and then it bends upward filling in the space between the limbs of the U-shaped th 1^. Th 3^ develops from th 2^, and th 3^ develops from th 2^. The proximal-end development that is revealed by the pyritized internal mold is typical of the streptoglastic diplograptid type as described by Bulman (1970). More detailed information regarding the development of dicellograptids can be obtained through the study of isolated early growth stages. This information includes the timing of budding and the origin of structures that are not preserved in pyritized internal molds. Early Growth Stages. 'Specimens representing early growth stages of Dicellograptus alabamensis, D. bispiralis bispiralis, D . gurleyi gurleyi, 186 and sextans are very numerous in my collections, but only a few specimens representing early growth stages of bispiralis bispiralis are available for study. Most of the early growth-stage specimens are compressed; however, fusellar structure is generally preserved. Speci mens preserved in full relief are not uncommon. Available specimens of some species, such as Dicellograptus alabamensis, represent many more stages of development than do available specimens of other species. Further, available specimens of some species represent certain stages of development for which I have no corresponding specimens as other species. Table 6 is a compilation of the illustrations of all the stages and the development of the proximal end that are represented by available specimens of D. alabamensis, bispiralis bispiralis, D. gurleyi gurleyi and D. sextans. These illustrations are arranged in order with regard to the development of the proximal end. Twenty-one stages of development are represented by the available specimens, and numbers that are affixed to these stages in Table 6 are used in the following description. The sicula has an overall length of 0.9 to 1.4 mm among the various species- The prosicula, which accounts for one-fourth to one-third of the length of the sicula, is conical. Longitudinal threads extend from the aperture of the prosicula to the base of the nema. A spiral thread is preserved in the prosicula of one specimen of Dicellograptus alabamensis (Text-fig. 34b). 187 TABLE 6 : Correlation of growth stages of Dicellograptus Text-fig. numbers for each species Stage of gurleyi blsplralls Development alabamensis gurleyi sextans blsplralls 1 34a -- - 2 34b - - - 3 34c - - - 4 34d 41a 45a 37a 5 34e 41b - - 6 - - 45b - 7 34f 41c 45c 37b 8 34g 41d,e - 37c 9 34h - - - 10 - 41f 45d - 1 1 - 41g 45e - 1 2 341 41h 45f - 13 34j,k - 45g 37d,e 14 341 411 - - 15 - 41j 45h - 16 - - 451 - 17 35a 42d - - 18 35b - 45 j,46a - 19 35d 42a,b - 2 0 35e 42c - 37f 2 1 - 42f - - 188 The metasicula develops by the addition of fusellar tissue at its aperture (Stages 1-3)• The virgella initially develops a short distance below the aperture of the prosicula by the curvature of the fuselli toward the metasicular aperture (Stage 1). The curvature of the fuselli increases with the development of the metasicula in such a way that the virgella assumes the shape of a spine (Stage 2). Before the metasicula is completely developed, a foramen forms immediately adjacent to and on the reverse side of the initial part of the virgella (Stage 3). This foramen, from which th 1 ^ will later bud, is 0 . 0 2 mm in diameter and circular. Its margins truncate the growth lines of the metasicula (Text-fig. 34c). When the metasicula is completely developed, its aperture is characterized by broad, lateral lappets, which consist of two or three fuselli and the virgella (Stage 4). The initial bud of th 1^ arises from the foramen in the metasicular wall. It grows down the side of the sicula along the virgella for a distance of 0.15 mm below the foramen (Stage 4), where th 1^ buds from th 1^ (Stage 5 and 6 ). This budding is accomplished by the development of two walls, one on each side of the aperture of the initial part of th 1^. These walls are composed of fuselli that are semi-circular shaped in lateral view and that are discontinuous with the growth lines of the initial part of th 1 . The walls grow outward, away from the sicular wall, until they meet above the virgella and the initial bud of th 1 ^ which results in the formation of two openings (Stage 7) . Th 1^ develops from the opening that is directed obliquely upward direction the reverse side of the sicula. Th 1^ continues to develop from the opening that is directed downward along the virgella. 189 Th 1^ grows down the virgella to the sicular aperture, where it bends outward and upward in the shape of a U (Stages 8—10). Th 2^ will later fill in the space between the limbs of the U. Th 1 ^ grows obliquely upward across th 1^ to the reverse side of the sicula. It then bends and grows obliquely downward across the reverse side of the sicula to the anti-virgellar margin of the sicular aperture where it grows outward and upward forming a U-shaped tube (Stages 10-15). Th 2^ will later fill in the space between the limbs of the U. The development of th 1^ proceeds much faster than the development of th 1%. Th 1^ develops a mesial spine when the growth direction of th 1^ changes from upward to downward (Stage 10). Th 1^ grows downward as a broad flap (Stage 11). When the aperture of th 1^ is completed, 2 a partition develops under the broad flap of th 1 . This partition separates the initial part of th 2 ^ from the virgellar side of the downward-directed flap of th 1^ (Stage 12). When th 1^ reaches the aperture of the sicula and begins to grow outward from the sicula, th 2 ^ has grown downward to a level that is transverse to the inner base of the U-shaped tube of th 1^ (Stage 13). A narrow process develops from the outer margin of the downward-directed aperture of th 2^ (Stage 14). This process grows downward, and then inward to the sicula, resulting in the formation of two horizontally directed openings (Stage 15). Th 2^ continues to develop from the opening that is directed toward th 1 . Th 2 develops from the opening that is directed 2 0 toward th 1 . At this stage of development, th 1^ is nearly .-.omplete to 190 the aperture, and a flange is growing upward over the aperture of th 1 ^ (Text-fig. 41j). This flange later becomes the infragenicular wall of th 2 I. Subsequent development of th 2^ continues with the filling in from the bottom upwards of the space between the limbs of the U-shaped th 1^. It is accomplished by the growth of two lateral walls which connect the downward-directed limb of th 1 ^ with the upward-directed limb of th 1 ^ and later with the infragenicular wall of th 2^ (Stages 16 and 17). The downward-directed part of th 1 ^ serves as the dorsal wall of th 2 ^. The upward-directed part of th 1^ and the infragenicular wall of th 2^ 1 2 serve as the ventral wall of th 2 . Th 2 will fill in the space between the limbs of the U-shaped tube of th 1^ in a similar manner. As the development of th 2^ proceeds past the distal end of the th 2^ infragenicular wall, fuselli, which form a closed circle in a transverse plane, are added. This change, in the manner in which th 2^ adds fuselli, results in the formation of a geniculum and the independent development by th 2 ^ of its own dorsal and ventral walls, which later become the interthecal septum and supragenicular wall, respectively (Stage 18). In D. gurleyi gurleyi (Text-fig. 42b) the distal end of the th 2^ infragenicular wall forms a narrow flange that projects from the geniculum of th 2 ^. During the initial development of the dorsal wall of the metatheca of th 2 ^, an opening is formed at the base of the dorsal wall where it intercepts the initial, downward-directed, part of th 1^. Th 3’ develops from this opening (Stage 19). 191 As th 3^ develops, the th 2^ infragenicular wall is complete, and the growth of the lateral walls of th 2 ^ fills in the space between the *> 2 limbs of the U-shaped th 1^. The downward-directed part of th 1 and the anti-virgellar side of the sicula serves as the dorsal wall of the metatheca of th 2^ (Stage 20). During budding, th 3^ grows upward out of its opening from th 2^ and then bends outward to grow along the dorsal wall of th 2^. Simultan eously, the aperture of th 2 ^ is completed, and the th 3^ infragenicular wall begins to develop (Stage 20) . The growth of th 3^ is accomplished by the addition of U—shaped fuselli. The bases of the U-shaped fuselli form the dorsal wall of the th 3^ protheca; the limbs of the U form the lateral walls. The dorsal wall of the th 2^ metatheca serves as the ventral wall of the th 3^ protheca. When th 3^ reaches the distal end of its infragenicular wall, th 4^ will bud and develop along with the metatheca of th 3^ in the same way that the th 3^ protheca and the th 2 ^ metatheca developed. The development of the metatheca of th 2^ along with the 9 1 1 protheca of th 3 is the same as for th 2 and th 3 , except that in many 2 specimens the initial bud of th 3 is in contact with the sicula (Stage 21). Development of Prothecal Fold. All of the described species and subspecies of Dicellograptus show prothecal folds on the dorsal stipe walls. These folds are similar externally in all the described specimens, and they are characterized as follows (^. alabamensis, Text-figs. 32 g & 33 h; D. bispiralis bispiralis, Text-fig. 36b; D. geniculatus. Text-fig. 38 a, b and fig. 6 (Jaanusson, 1964); D. gurleyi gurleyi, Text-figs. 39 c-f, j, 40a; D. gurleyi n. ssp. A, Text-fig. 36 d; and 2» sextans, Text-fig. 43 e): 192 1) The prothecal folds on the dorsal stipe margins are consistently located transverse to the genicula on the ventral stipe margins. 2) The dorsal wall of the th n protheca extends dis tally past the proximal end of the interthecal septum between th n and th n+1 ; thus, the dorsal wall of the th n protheca is continuous with the dorsal wall of the th n+1 prothecal fold. Ventral curvature of this wall forms the prothecal fold. 3) The distal limit of the th n+1 prothecal fold is marked by a constriction in the th n+ 1 protheca. 4) The th n+1 dorsal wall, which develops distally from the constriction, intercepts the dorsal wall of the prothecal fold at a high angle. Isolated specimens, especially early growth stages, show the development of prothecal folds to be similar in Dicellograptus alabamensis, D. bispiralis, D. gurleyi, and D. sextans. This development of the prothecal folds is intimately related with the budding of a protheca from a preceding protheca. The most proximal prothecal fold in the dicellograptid rhabdosome is at th 3^. It originates in early growth stages with the formation of an opening at the proximal end of the interthecal septum between th 2 ^ and th 3^, which is the dorsal wall of the th 2^ metatheca (Dicellograptus alabamensis, Text-fig. 35d; 2- sextans, Text-figs. 45j, 46a; and D. gurleyi gurleyi, Text-fig. 42e). The initial, downward directed part of th 1^ forms the proximal (or dorsal) margin of the opening. The proximal 193 end of the interthecal septum, which is bent upwards, forms the distal (or ventral) margin of the opening. Growth lines on the interthecal septum parallel the distal margin of the th 3^ opening (Text-fig. 45j). The lateral margins of the opening are formed by the ends of fuselli that form the lateral walls of th 2^ (Text-fig. 46a). The lateral and distal margins of the opening are strengthened by a list. The openings for the prothecae in distal thecae show the same morphology as in th 3 ^, except that the distal end of the dorsal wall of the th n protheca forms the dorsal margin of the opening for th n+ 1 protheca. The opening for the protheca may have formed by the absorption or the non-deposition of fuselli. Although there are no early growth stages that show the actual formation of the opening, non-deposit ion is favored because: 1 ) the opening is precisely located; 2 ) the distal (or ventral) margin of the opening is parallel to the growth lines on the interthecal septum; and 3) the proximal end of the interthecal septum bends upward. This opening is hereafter referred to as a "primary notch" because of its mode of formation. The development of the protheca is discontinuous. It begins with the formation of a prothecal base over the primary notch (Dicellograptus alabamensis, Text-fig. 35g; D. gurleyi gurleyi, Text-fig. 42e; and D. sextans, Text-fig. 46b). The prothecal base is equivalent to the prothecal fold in the described specimens. James (1965) original use of this term is essentially the same (see Remarks). The prothecal base, which has the shape of a hood, is composed of U-shaped fuselli. The limbs of the U-shaped fuselli form the lateral walls of the prothecal base. The bases 194 of the U-shaped fuselli form the dorsal wall of the prothecal base and are continuous with the dorsal wall of the preceeding protheca (^. alabamensis, Text-fig. 35g; and D, sextans, Text-fig. 46). Discontinuous growth in the protheca occurs when the prothecal base develops over and just past the primary notch. Externally, this discon tinuous growth is marked by a constriction of the protheca, which defines the distal end of the prothecal fold (Dicellograptus alabamensis, Text- fig. 35f; D. gurleyi, Text-fig. 39j ; and D. sextans. Text-fig. 46d). In ternally, this constriction is represented by a list, which probably develops at the aperture of the prothecal base at the time of discon tinuous growth (JD. gurleyi gurleyi, Text-figs. 39j and 40a; D. sextans, Text-fig. 46h) . The list is continuous with the lateral margins of the up turned proximal end of the interthecal septum. The trace of the inter thecal septum on the lateral stipe walls appears to extend proximally (or dorsally) beyond the distal (or ventral) margin of the primary notch (D. alabamensis, Text-fig. 35f; D. gurleyi gurleyi, Text-fig. 39j ; and D. sextans, Text-Fig. 46b) . Text-figure 35d of D^. alabamensis shows this to be only an apparent effect. Growth lines on the lateral walls of the protheca distal to the pro thecal base are oriented at a low angle to the interthecal septum (Dicellograptus alabamensis, Text-fig. 35f; D. gurleyi gurleyi, Text- fig. 42c; and ]D. sextans, Text-fig. 46e). Those lines proximal to the constriction and on the lateral walls of the prothecal base are oriented at a high angle to the interthecal septum (D. alabamensis, Text-fig. 35e,g; and D. gurleyi gurleyi, Text-figs. 42a-b, e-f. 195 Development of Thecae and Associated Structures. As interpreted above, the th n+ 1 protheca develops from a primary notch and not by absorption. Thus, the initial development of the th n+1 protheca is simultaneous with the initial development of the th n metatheca. However, the succeeding development of the th n metatheca is much more rapid than that of the th n+1 protheca (Dicellograptus alabamensis, Text-fig. 35e; D. gurleyi gurleyi, Text-fig. 42a-b; D. sextans. Text-fig. 46c; 2" bispiralis bispiralis, Text-fig. 37f). This phenomenon can be interpreted on the basis of growth line evidence. The growth lines of the prothecal folds (prothecal bases), which are oriented at a h i ^ angle to the interthecal septum, are closely spaced, whereas the growth lines of the more distal portions of the protheca, which are oriented at a low angle to the interthecal septum, are widely spaced (Dicellograptus alabamensis, Text-figs. 35e-d, g; D. gurleyi gurleyi, Text-figs. 42a-b, e-f; and D. sextans, Text-figs. 46e-g) . This change in the thickness and orientation of the growth lines occurs at the constriction that defines the distal limit of the prothecal fold (prothecal base). Additionally, the growth lines of the distal part of the protheca are of the same thickness as, and are oriented parallel to, the growth lines of the metatheca (^. alabamensis, Text-fig. 35g; and D. gurleyi gurleyi, Text-fig. 42f). The change in the growth lines correlates with the relative rates of development of the various parts of the theca as deduced from early 196 growth stages. After initial development, the th n metatheca develops much more rapidly than the th n+1 protheca. The th n metatheca is complete or nearly so before the th n+ 1 protheca has developed to the distal end of the prothecal fold (Dicellograptus alabamensis, Text-fig. 35e; jD. gurleyi gurleyi, Text-figs. 42a-b, e; and D. sextans, Text-fig. 45j, 46a-b). However, after development past the prothecal• fold, the rate of growth of the th n+ 1 protheca increases greatly in such a way that it appears to keep pace with the rate of growth of the th n aper- tural flange, which is the th n+1 infragenicular wall (D. alabamensis, Text-figs. 35f-g; D. gurleyi gurleyi. Text figs. 42c-f; and 2» sextans Text-figs. 46d-e). This rate of growth remains constant through the development of the th n+ 1 metatheca as evidenced by the uniform thick ness and orientation of growth lines. A similar pattern of thecal development is found in the proximal thecae. The initial bud of th 1 shows relatively thin growth lines (Dicellograptus alabamensis, Text-fig. 34d). The opening for th 1^ develops as a primary notch. After the formation of the th 1 primary notch, th 1^ develops rapidly as shown in early growth stages (D. alabamensis, Text-figs. 34e-h; D. gurleyi gurleyi,• Text-figs. 41e-f; and sextans ; Text-figs. 45b-e); whereas th 1^ grows relatively slowly. The presence of thick growth lines in that part of th 1^ that is distal to the initial bud supports this interpretation. Although growth lines can not be discerned in those parts of the proximal thecae that are 9 1 equivalent to the prothecal folds in distal thecae, th 1 , th 2 , and 197 2 th 2 appear to develop similarly because they bud from a primary notch and early growth stages show an abrupt increase in the rate of growth of th n after the primary notch for th n+ 1 develops. The development of the th n metatheca in advance of the th n+1 protheca implies that the interthecal septum initially forms as the dorsal wall of the th n metatheca. Growth lines on the interthecal septum support this interpretation because they are continuous with those on the lateral walls of the th n metatheca (Dicellograptus alabamensis, Text-fig. 35d; D. sextans, Text-fig. 45j). The interthecal septum between th n and th n+1 is continuous with the dorsal wall of the th n aperture, and it serves as the ventral wall of the th n+ 1 protheca. Early growth stages (Dicellograptus alabamensis, Text-figs. 35 c, d; jD. gurleyi gurleyi, Text-figs. 41 i-j, 42 a-d, f ; and sextans, Text- figs. 45 i-j, 45 a, e) show the development of the infragenicular wall. The th n+1 infragenicular wall, which is above the aperture of th n, develops after the th n aperture is complete and simultaneously with the development of the th n+1 prothecal base. Growth of the infragenicular wall originates on the dorsal wall of the metatheca, a short but distinct distance below the dorsal margin of the aperture. It grows in a ventro- distal direction to form a wall, which is semi-circular in lateral view and which projects over the thecal aperture. Its distal end forms a flange that projects from the geniculum in D. gurleyi. The infragenicular wall and the interthecal septum are separate structures. The dorsal wall of the metatheca (later to become the inter thecal septum) is completely developed before the infragenicular 198 wall develops. Because the th n+1 infragenicular wall develops well in advance of the th n+ 1 protheca, it is probably formed by the zooid of the underlying, th n aperture. Thus, the free ventral wall of an individual theca, consisting of the infra- and supragenicular walls, is formed by two separate zooids. The position of the origin of the infragenicular wall proximal to the dorsal margin of the thecal aperture results in an introtorted thecal aperture. The base of the apertural excavation is formed by two separate walls: the dorsal metathecal wall (distal end of the interthecal septum) and the infragenicular wall. Remarks Development of Proximal End. The proximal-end development of Dicellograptus alabamensis, bispiralis bispiralis, gurleyi gurleyi, and D. sextans resembles closely that of Dicranograptus nicholsoni Hopkinson (Bulman, 1944), Dicellograptus divaricates salopiensis Elies and Wood (Strachan, 1960), and Dicellograptus sp. (James, 1965). These forms are comparable in that: 1 ) th 1 ^ has a left-handed origin and an upward initial direction of growth; 2 ) th 2 ^ has a downward initial direction of growth; and 3) the first four thecae show an overall upward direction of growth. A similar proximal-end development is suggested for JD. gurleyi n. ssp. A and D^. bispiralis n. ssp. A because the first four thecae show an overall upward direction of growth with the dorsal walls of the th 2 ^ 2 and th 2 prothecae in contact with the sicula. The proximal-end development of Dicellograptus geniculatus (Bulman, 1932, 1947), Leptograptus? sp. ind. (Whittington, 1955), and species of 199 the elegans group such as D^. complanatus Lapworth (Skoglund, 1963) , differs from that described above by the primarily horizontal direction of growth of the first four thecae in such a way that their dorsal walls are not in contact with the sicula. The th 2^ and th 2^ prothecae of these latter species must be composed of U-shaped fuselli, which are oriented in such a way that the bases of the U-shaped fuselli form the dorsal 1 2 walls of the prothecae. In contrast, the th 2 and th 2 prothecae in vagus group species are in contact with the sicula and are composed of only two lateral walls of fuselli. Development of Prothecal Fold. James' (1965) Dicellograptus sp. is the only dicellograptid in which the detailed structure • of the prothecal fold was previously known. James' specimens are relatively few, yet they show structures that indicate a development similar to that described above. These similarities are: 1) The opening for the bud of the protheca develops as a primary notch. 2) The proximal end of the interthecal septum bends upward. 3) The prothecal fold is formed by a prothecal base. 4) The dorsal wall of the prothecal base is continuous with the dorsal wall of the preceding protheca. 5) Discontinuous growth occurs between development of the prothecal base and the rest of the protheca as evidenced by marked changes in thickness and orientation of growth lines. The prothecal folds of Dicellograptus sp. (James, 1965) are structurally more complex than those described above because: 2 0 0 1) The curvature of the proximal (or dorsal) end of the interthecal septum is so great that it bends in a dorso-ventral direction back over itself. This elevates the ventral wall of the prothecal base above the interthecal septum. 2) The dorsal wall of the th n+1 protheca intercepts the dorsal wall of the th n+ 1 prothecal base a slight distance proximal to its aperture. Thus, the dorsal walls of the th n+1 protheca and prothecal base overlap in such a way that the dorsal wall of the prothecal base projects a short distance into the interior of the protheca. Bulman (1969) discussed and re-figured James' (1965) specimens as Dicellograptus cf. divaricatus. On the basis of an additional specimen (fig. 4b), which he included in his discussion, Bulman shows the dorsal wall of the th n+ 1 protheca to cover almost completely or overlap the th n+1 prothecal base (fig. 2e). However, James' (1965, Text-fig. 11) longitudinal sections show such a relatively small amount of overlap that the prothecal fold is formed entirely by the prothecal base as in the specimens described above. Some specimens of jD. alabamensis (Text-fig. 33h) and D. gurleyi gurleyi (Text-figs. 40a) show externally what appears to be such an overlap of the dorsal walls of the prothecal base and the protheca in such a way that the dorsal wall of the prothecal base projects into the interior of the protheca. However, this overlap is not present internally in isolated specimens, and its appearance is probably an expression of the list, which forms at the aperture of the prothecal base, and which is restricted to the thecal walls. The external appearance of prothecal folds is described for Dicello graptus geniculatus by Bulman (1932) and Jaanusson (1964), for 2 0 1 2" complanatus Lapworth and 2» johnstrupi Hadding by Skoglund (1963), for D. vagus Hadding by Hadding (1913) and Berry (1960), and for Dicranograptus irregularis Hadding by Hadding (1913). In addition, prothecal folds are described below for Leptograptus trentonensis Ruedemann and Dicranograptus irregularis Hadding. Externally, the prothecal folds of these species are structurally comparable to those of Dicellograptus sp. (James, 1965) and the herein described species because: 1 ) the prothecal fold of th n+ 1 on the dorsal stipe margin is consistently transverse to the geniculum of th n on the ventral stipe margin; 2 ) the dorsal wall of the th n protheca is continuous with the dorsal margin of the th n+ 1 prothecal fold; 3) the distal limit of the th n+1 prothecal fold is marked by a constriction in the th n+ 1 protheca; 4) the th n+ 1 dorsal wall, which develops distally from the constriction, intercepts the dorsal wall of the prothecal fold at a high angle; and 5) the proximal (or dorsal) end of the interthecal septum between th n and th n+1 , which bends upward (or dorsally), originates within the curve that is formed by the dorsal wall of the prothecal fold of th n+1 . Development of Thecae and Related Structures. James (1965) does not discuss rates of growth of various parts of the thecae. However, her illustrations (Text-fig. 9) show that development of thecae is the same as that described above because: 1 ) fuselli on the prothecal base, which are relatively thin, are oriented at a high angle to the interthecal septum whereas those on the more distal parts of the protheca, which are relatively thick, are oriented at a low angle to the interthecal septum; 2 ) fuselli on the metatheca and the distal part of the protheca of the same theca are 2 02 parallel; and 3) although th n+1 originates from a primary notch, the th n metatheca develops far in advance of the th n+ 1 protheca. Skevington (1965) discussed the rate of thecal development in relation to the width of growth lines in Meandrograptus? geniculatus. He found that the change from a relatively slow to a relatively rapid rate of growth in thecae of this species occurs at the most proximal part of the metatheca and not at the distal limit of the prothecal folds. A similar arrangement of the fuselli is reported for Cryptograptus tricomis by Bulman (1944) and is described herein for specimens of C^. marcidus. In two species of Nemagraptidae described herein, the change in fusellar thickness occurs at the geniculum and within the prothecae. As described above, the infragenicular wall of th n+1 in the early growth stages develops independently of the initial bud of th n+1 , after the formation of the dorsal wall of the th n metatheca (the interthecal septum between th n and th n+1 ) and the aperture of th n are complete, and it is discontinuous with the interthecal septum between th n and th n+1. Thus, the development of the free ventral wall (supragenicular and infragenicular walls) of th n+1 reflects discontinuous growth. The infragenicular wall is interpreted as being formed by the zooid of th n, and the supragenicular wall is formed by the zooid of th n+1 . The distal end of the infragenicular wall of Dicellograptus gurleyi gurleyi projects from the geniculum as a flange. A somewhat similar structure is described for Dicellograptus sp. by James (1965). She interprets the genicular flange to be continuous with the infragenicular wall, which, in turn, she interprets to be continuous with the interthecal 203 septum (dorsal wall of metatheca). However, James' Text-figures 3 and 4 indicate that the dorsal wall of the metatheca (later to become the interthecal septum) is complete before the development of the infragenicular wall and they also show that the infragenicular wall initially develops proximal to the dorsal margin of the aperture. This results in an introtorted aperture as in the species described above. James (1965, p. 52, Text-figs. 3 and 4) reports that the infragenicular wall of th n+1 develops independently of, and ahead of, the th n+1 protheca. She believes that the zooid of th n is responsible for the formation of the th n+ 1 infragenicular wall. Bulman's (1944, Text-figs. 19G-19L) illustrations of Dicranograptus nicholsoni Hopkins on show that the infragenicular wall forms in the same manner as that in the species described above. Skoglund's (1963, Text-fig, lOA) illustration of Dicellograptus complanatus shows what appears to be the initiation of the infragenicular wall proximal to the dorsal margin of the th 1^ aperture. This, together with the introtorted nature of the thecal apertures, suggests that the infragenicular wall develops in the same manner as that in the species above. The introtorsion of the thecal apertures and/or the presence of genicular flanges suggest a similar development of the infragenicular wall in Dicellograptus geniculatus and 2» divaricatus salopiensis (Strachan, 1960), As described herein, specimens of Leptograptus trentonensis and Nemagraptus gracilis show that the th n infragenicular wall in these species is secreted by the th n zooid; whereas in specimens of Nemagraptid sp A, described herein, the th n+1 infragenicular wall is secreted by the th n zooid. 204 Dicellograptus alabamensis Ruedemann. 1908 Text-figs. 26, 31-35 1908 Dicellograptus moffatensis (Carruthers) var. alabamensis nov., Ruedemann, Pl. 310-312, text-figs. 234-236, Pl. 20, figs. 1, 2. 1908 Dicellograptus smithi sp. nov., Ruedemann, p. 313-315, text-figs 237, 238, Pl. 19, figs. 3-6. 1926 Dicellograptus moffatensis var. alabamensis.Butts, Pl. 23, fig. 1. 1947 Dicellograptus moffatensis (Carruthers) var. alabamensis Ruedemann, Ruedemann (partim), p. 385, Pl. 64, figs. 12-15 (non figs. 16). 1947 Dicellograptus smithi Ruedemann, Ruedemann, p. 388, Pl. 65, figs. 1-13. 1952 Dicellograptus moffatensis var. alabamensis Ruedemann, Decker, Pl. 1, fig. 22; Pl. 2, fig. 41. 1952 Dicellograptus smithi Ruedemann, Decker, Pl. 2, fig. 39. 1960 Dicellograptus moffatensis var. alabamensis Ruedemann, Berry, p. 76, Pl. 15, fig. 10. 1960 Dicellograptus smithi Ruedemann, Berry, p. 77, Pl. 15, fig. 3b. 71963 Dicellograptus cf. D. smithi Ruedemann, Rossand Berry, p. 107-108, Pl. 6 , fig. 20. Type Data Ruedemann (1908) did not designate a holotype for Dicellograptus moffatensis var. alabamensis. In his plate caption, he referred to several specimens in one figure (PI. 20, fig. 1) as cotypes. These specimens, which were also illustrated by Ruedemann (1947, PI. 64, fig. 12) , are stored in the Poor collection at the Geological Survey of Alabama, Tuscaloosa, Alabama. The writer has not had the opportunity to inspect these specimens in order to select a lectotype. 205 Diagnosis A species of Dicellograptus with axillary angle usually less than 30 degrees. Stipe divergence increases distally. Sicula inclined toward stipe no. 2. First two thecae U-shaped. Stipe width increases distally: 0.6 to 0.7 mm at th 2; 0.8 to 0.9 mm at th 5; constant thereafter. Proximally, 6.5 to 7.0 thecae in 5 mm; distally, 5 thecae in 5 mm. Theca with slightly obtuse genicular angle, slightly convex supragenicular wall, and slightly introtorted and introverted aperture. Mesial spines on first four to eight thecae of each stipe. Axillary web of cortical tissue in large rhabdosomes. Material More than 1000 specimens of Dicellograptus alabamensis were obtained from the Pratt's Ferry and Pratt's Syncline sections. This material represents the following states of preservation: 1) Isolated, compressed and uncompressed specimens. 150 specimens with proximal end and two or more pair of thecae; 165 early growth stages, generally with less than two pair of thecae; 1 2 pyritized specimens representing early growth stages; and 43 distal fragments. All of these specimens are from the Pratt's Ferry and Pratt's Syncline sections. The pyritized specimens are from a single locality that is 20.5 meters above the base of the examined portion of the Pratt's Ferry section. 2) Non-isolated and compressed specimens with periderm preserved. More than 500 specimens, mostly large rhabdosomes, from the Pratt's Ferry and Pratt's Syncline sections. Several hundred more specimens 206 are not included in this total. They occur in extreme abundance, completely covering a bedding surface that is 126 meters above the base of the Pratt's Syncline section. Description Isolated Specimens (Text-fig. 31-32). The largest available rhabdosome has a stipe 7.25 mm long. Proximally, stipes are parallel to subparallel and diverge at an angle of less than 30 degrees (Text-figs. 31a-f). In many specimens, stipes show torsion with clockwise rotation when the rhabdosome is viewed from the proximal end (Text-figs. 31g-i). The axil is acute, and the sicula is generally in contact with the dorsal wall of stipe no. 2 for its entire length (Text-fig. 31e-f). The axil in large rhabdosomes is occupied to varying degrees by a web of cortical tissue. This axillary web in the largest available rhabdosomes fills the axil to the level of the fourth pair of thecae (Text-fig. 31i). This gives the rhabdosome a dicranograptid appearance. The web develops upward from the base of the axil and connects the dorsal walls of the stipes. It is commonly as thick as the stipes (Text-fig. 31g-i, 32a-f). The web connects the lateral walls of the stipes in specimens that show torsion (Text-figs. 31g-i). Secondary thickening of the mesial spines of the proximal thecae accompanies development of the web. The proximal end is broad and rounded in lateral view and possesses three prominent spines, the virgella and the mesial spines of th 1 ^ and 2 th 1 . The first two thecae are U-shaped. The second pair of thecae has a dominantly upward direction of growth so that dorsal walls of prothecae 207 are in contact with the sicula. This enhances the biserial appearance of the proximal end. The stipes are 0.6 to 0.7 mm wide at th 2. Distally, they widen to 0.8 mm at th 7, and distal to this theca the width is relatively constant. The greater width in the proximal end occurs in large rhabdosomes that are secondarily thickened with cortical tissue. Distally, stipe width may be 1.1 mm in specimens that are fully compressed. Proximally, there are 6.5-7 thecae in 5 mm; distally, thecae number 4.5 in 5 mm. Thecae overlap for half their length. Proximally, thecae and supragenicular walls are 1.3 mm and 0.4 to 0.5 mm long, respectively. Distally, they lengthen to 2.2 mm and 0.9 mm, respectively. The theca shows a distinct geniculum, a convex suprageni cular wall, and an introverted and introtorted aperture (Text-fig. 32i-j). The genicular angle is 90 to 100 degrees in specimens preserved in full relief. The supragenicular wall is only slightly convex and the point of maximum dorso-ventral thecal width, which is produced by the curvature of the supragenicular wall, occurs transverse to the base of the apertural excavation. The maximum dorso-ventral width is located at three-fourths of the distance from the geniculum to the aperture in proximal thecae. Mesial spines, which project from the supragenicular wall at the point of maximum dorso-ventral thecal width, are present on the first four to eight thecae of each stipe. In distal thecae, the apex of the curvature of the supragenicular wall is half the distance between geniculum and aperture (Text-fig. 32h). The aperture is relatively circular in cross-section and opens in a dorso-distal direction. Rounded lappets are present on the ventro-lateral 208 margins. Viewed laterally, the apertural excavations have the shape of an inverted comma. They occupy a third to a half the stipe width and as much as half the length of the supragenicular wall. A list extends along the lateral margins of the infragenicular wall from the base of the apertural excavation to the geniculum. Well-preserved stipe fragments show gentle prothecal folds located on the dorsal stipe margins transverse to the genicula on the ventral stipe margins (Text-fig. 32g). In specimens that have a thick cortical layer, the prothecal folds cannot be observed, and the dorsal stipe margins are straight. Grooves, representing the traces of the interthecal septa on the lateral stipe walls, extend from the bases of the thecal excavations to the prothecal folds. Thecae are distorted in obliquely compressed specimens (Text-figs. 32k-l). This results in an apparent increase in stipe width. The supra genicular wall, which is less convex in lateral view, and the infragenicular wall are compressed so that they are in the same plane as the lateral stipe walls. As a result, genicula on one of the lateral stipe walls show 90-degree genicular angles ; whereas the genicula on the opposite stipe wall show very obtuse genicular angles. Non-isolated and Compressed Specimens with Periderm Preserved (Text-figs. 33a-h) . The largest available rhabdosome is 80 mm long. In such large specimens, the proximal end, including thecal spines, is secondarily thickened and the axil is filled by an axillary web (Text- fig. 33c-d, f). This results in a biserial appearance of the proximal end that may extend distally as far as the fifth pair of thecae. 209 Stipes show a uniform width of 0.9 to 1.1 mm because the narrower, proximal parts have been secondarily thickened. The cortical web is not present or is only slightly developed in smaller rhabdosomes (Text-figs. 33a-b, e, h, g). The axil is acute. The divergence of the stipes is rarely greater than 30 degrees. Stipes in some specimens are initially parallel and then diverge at th 4 to th 6 to enclose an angle of 30 to 45 degrees. Stipes are 0.6-0.7 mm wide at th 2 in specimens without the axillary web. They widen to 0.75-0.85 mm at th 5, and then esdiibit uniform size. A few non-isolated specimens are filled internally by pyrite and retain relief. This results in slightly narrower stipe widths of 0.55 mm at th 2 and 0.70 mm at th 5. Proximally, thecae number 6.5-7 in 5 mm and 13 in 10 mm. Distally, there are 5 thecae in 5 mm and 10 in 10 mm. Compression that is directed normal to the lateral stipe walls produced the same thecal morphology that is found in isolated, non-compressed specimens. Compression that is directed obliquely to the lateral stipe walls distorts the thecal morphology. Thus, the supragenicular wall is less convex, the geniculum shows either a 90 degree angle or an obtuse angle, and the width of the apertural excavation is significantly lesser or greater in lateral view. Early Growth Stages (Text-figs. 34-35). Th 1^ grows down the virgellar side of the sicula past the sicular aperture before bending and growing outward away from the sicula. As a result, the ventral wall of the horizontal part of th 1^ is 0.05-0.10 mm below the sicular aperture. This character proves useful in differentiating early growth stages of 2 1 0 Dicellograptus alabamensis from early growth stages of other dicellograptid species in which the ventral wall of th 1 ^ is at the same level as the sicular aperture. The range of variation in the length of the sicula, 1.05-1.20 mm, overlaps with that found in other dicellograptid species. With the development of th 2^, early growth stages of D. alabamensis can be recognized by the shape of the geniculum and supragenicular wall and the position of the mesial spine. Remarks Morphological Variation within Species. Two types of variation, which are not produced by preservation, occur among the specimens of Dicellograptus alabamensis. The variation in the development of the axillary web and the associated thickening of the proximal end is associated with the astogenic age of the rhabdosome. The variation in the number of proximal thecae that bear mesial spines is of a different nature. Mesial spines are present on the first four to eight thecae of each stipe. This variation in the number of thecae with spines is shown in Text-fig. 26 in which the relative percentage of each variant form is plotted stratigraphically. Although the data are poorly distributed, there appears to be a stratigraphie trend. Specimens from the Pratt Ferry section and the lower part of the Pratt Syncline section generally show mesial spines on the first five to six thecae of each stipe, whereas specimens from higher stratigraphie intervals generally show mesial spines on the first four thecae. This same type of variation is present in specimens of Dicellograptus gurleyi gurleyi and Pseudoclimacograptus 2 1 1 EXPLANATION OF TEXT-FIGURE 26 Stratigraphie variation in thecal spines in specimens of Dicellograptus alabamensis Ruedemann from the Pratt's Ferry- Pratt's Syncline section. 1. Spines on first four thecae of each stipe; 2. Spines on first five thecae of each stipe; 3. Spines on first six thecae of each stipe; 4. Spines on first seven thecae of each stipe; 5. Spines on first eight thecae of each stipe. Total number of specimens from 5 m intervals: 0-5 m: 22; 6-10 m: 57; 11-15 m: 33; 16-20 m: 15; 61-65 m: 11; 66-70 m: 2; 96-100 m: 2; 116-120 m: 60; all other intervals: 0. 212 140 120*3 ^ 100 w cCO V SI < 'o 80 V M CO Si o1 CD > o m 60 Q)(0 0) E c S 4 0 c CD CO O 20 ». • ? , > 2 . . * 3 0 100 percent of total number of specimens in 5 m Text-figure 26 213 modestus. For 2» modes tus this variation is speculated to be the result of genetic variation, which also appears to be the cause of variation in 2* alabamensis. This genetic variation may be ecologically influenced because those specimens of 2 « alabamensis and 2 * gurleyi with the greater number of thecal spines occur together within a restricted stratigraphies! interval. Comparison to Previously Described Species. Ruedemann (1908, 1947) defined Dicellograptus moffatensis alabamensis and D. smithi on the basis of specimens collected from the Pratt Ferry section. He distinguished these species primarily on the basis of features of the proximal end: presence or absence of axillary web, divergence of stipes, and number of thecae with mesial spines, all other features of the two species being similar. The preceding description shows that Ruedemann's distinguishing characters are actually related to the astogenetic age of the rhabdosome. Therefore, the two species are considered nonspecific. Ruedemann (1908, p. 311) placed his subspecies alabamensis in the species Dicellograptus moffatensis because of similarités in "the general form of the rhabdosome and of the thecae " Figured specimens of 2* moffatensis, a species which is not known from North America, are poorly preserved and poorly illustrated, and the detailed thecal structure cannot be discerned. Specimens of 2* moffatensis, which are described by Toghill' (1970a), occur in the Pleurograptus linearis Zone and show a horizontal 214 direction of growth of the first two thecae and a maximum stipe width of 1.5 mm. Therefore, the subspecies alabamensis is herein raised to specific status. Figured Specimens OSU 32988-OSU 33027 215 Dicellograptus bispiralis bispiralis (Ruedemann, 1947) Text-figs. 36a,b,e, 37 1947 Dicranograptus furcatus (Hall) var. bispiralis n.var., Ruedemann, p. 390. 1947 Dicellograptus furcatus var. bispiralis n.var., Ruedemann, Pl. 65, figs. 46-47. 1952 Dicranograptus furcatus var. bispiralis Ruedemann, Decker, Pl. 2, fig. 48. Type Data Ruedemann (1947) designated as a holotype for Dicranograptus furcatus var. bispiralis the specimen that he illustrated in Plate 65, figures 46-47. This specimen, which has not been assigned a museum number, is stored at the U. S. National Museum, Washington, D.C., in U. S. G. S. collection number 400 0, Butts. The writer has examined this specimen, which agrees with Ruedemann 's (1947) description and illustration. Diagnosis A species of Dicellograptus whose rhabdosomes have helical stipes and prominent, laterally projecting lappets on thecal apertures. Stipe width increases distally: 0.5 mm at th 2; 0.75 to 0.85 mm at th 5; more distally, constant. Proximally, 5.5 thecae in 3 mm; distally 4 in 3 mm and 7.5 in 5 mm. Theca with obtuse genicular angle, convex supragenicular wall, introverted and introtorted aperture, and prominent mesial spine. Material More than 200 specimens of Dicellograptus bispiralis bispiralis were obtained from the Pratt's Ferry and Pratt's Syncline sections. This material represents the following states of preservation: 216 1 ) isolated and generally compressed specimens: 25 specimens with proximal end and two or more pairs of thecae; 30 early growth stages; and 43 distal fragments. All these specimens are from the Pratt's Syncline section. 2 ) non-isolated and compressed specimens with periderm commonly preserved: More than 100 specimens, mostly large rhabdosomes, from the Pratt's Ferry and Pratt's Syncline sections. Description Isolated Specimens (Text-figs. 36e). The most conspicuous feature of the rhabdosome is the shape of the stipes. As indicated in the description of compressed and non-isolated specimens, below, a pattern of stipe divergence, followed by convergence and crossing, results in the formation of a series of distally enlarging loops. This pattern, together with torsion of the stipes, supports the interpretation that the original shape of each stipe is a helicoidal spiral (Text-fig. 36a). In spite of compression and breakage, isolated specimens support the interpretation of helicoidal-spiral shaped stipes. As shown in Text- fig. 36e, the stipes show clockwise rotation when the rhabdosome is viewed from the proximal end. Torsion of the stipes is associated with this rotation. The torsion is accomplished by the gradual clockwise rotation about the stipe of successive thecae, when the stipe is viewed from the proximal end, so that the free ventral wall of each theca is on the outside of the helicoidal spiral (Text-fig. 36a-b, e). The proximal end of the rhabdosome is rounded in lateral view and possesses three prominent spines, which are the virgella and the mesial X 2 spines on th 1 and th 1 . The first pair of thecae are 217 U-shaped. The axil is rounded. The sicula is tilted toward stipe no. 2. For the greater part of its length, the sicula is in contact with the dorsal wall of stipe no. 2 . The stipes are 0.50 - 0.55 mm wide at th 2, and increase in width distally to 0.75 - 0.85 mm at th 5, after which the width is constant. Torsion and compression of the stipes results in a mayimitm observed stipe width of 1.0 mm. Proximally, there are 5.5 thecae in 3 mm. Dis tally, the thecae number 4in3mm, 7 - 7.5 in 5 mm, and 12 - 13 in 10 mm. The thecae overlap for half their length, which increases from 1.0 mm at th 3 to 1.3 mm at th 7, to a maximum observed length of 1.7 mm at th n. An individual theca originates at the level of the aperture of the second preceding theca. At the point of origin of each theca, a prominent prothecal fold is present at the dorsal stipe margin. Each theca esdiibits a geniculum with an obtuse genicular angle. The geni- culum is usually obscured in compressed specimens because of the con vexity of the supragenicular wall. The supragenicular wall is very convex with the point of maximum dorso-ventral thecal width occurring at a point transverse to the proximal end of the apertural excavation. A prominent mseial spine, which is 0.4 mm long on proximal thecae and up to 1 . 0 mm long on distal thecae, is present at the point of maximum dorso-ventral thecal width. The supragenicular wall is 0.5 mm long proximally and gradually increases to a maximum length of 0.65 mm distally. The thecal aperture, of thecae that are distal to th 4, is intro torted and extremely introverted. It opens in a dorso-proximal direction 2 1 8 toward the base of thecal excavation. The lateral margins of the aperture form conspicuous lappets. These lappets project up to 0.15 mm outward (laterally) from the aperture and are semi-circular in cross-section (Text-fig. 36b, e). The morphology of the thecal apertures gradually changes in a distal direction along the stipes. At the first theca on each stipe, the aperture is only slightly introverted and opens in a distal direction. Lappets, which are represented by rounded projections, are situated at the ventro lateral comers of the aperture margin. At the second theca, the supra genicular wall is more introverted than at the first theca and the aper ture opens in a dorso-distal direction. The lappets are slightly convex on their lateral surfaces. At the third theca, the supragenicular wall is even more introverted and the aperture opens in a dorsal direction. The lappets are semi-circular in cross-section and project 0.05 mm laterally. At the fourth theca, the aperture opens in a dorso-proximal direction, and the lappets project 0.15 mm laterally. Distal to the fourth theca, the thecal morphology remains constant. The thecal excavations are 0.3 mm long and 0.2 mm wide at th 2. They increase in size distally to a length of 0.4 mm and a width of 0.2 mm. Grooves, representing the traces of the interthecal septa, are present on the lateral stipe walls and extend from the prothecal folds to the bases of the apertural excavations. Non-isolated and Compressed Specimens With Periderm Preserved. The rhabdosomes are conspicuous on shale surfaces because of the spiral appearance of the stipes. In many specimens, the proximal end is com pressed in such a way that it appears to be biserial. The stipes 219 diverge from the proximal end, then converge and cross each other. This pattern continues distally and results in the formation of a series of distally enlarging loops. The largest rhabdosome found, which is 20 mm long, consists of three loops. The convex supragenicular wall and the prominent prothecal folds give the stipes a "beaded" appearance. The thecal spines and apertural lappets are easily observed in well-preserved specimens. The torsion of the stipes can be detected in flattened specimens by the orientation of the thecae in relation to the stipes. At the maximum amount of stipe divergence, the thecae are located on the outside of the rhabdosome, and they present a lateral aspect. As the stipes converge and diverge, the stipe will overlie the thecal apertures on one stipe; on the other stipe, the thecal apertures will be exposed on the stipe. The supra- jacent stipe at one stipe intersection will be subjacent at the next stipe intersection. This feature reflects the original, helicoidal- spiral organization of the rhabdosome. Early Growth Stages (Text fig. 37) . Specimens representing early growth stages are poorly preserved, being compressed and heavily carbon ized. They rarely show fuselli. The available specimens represent very few stages in the development of the proximal end. Specimens repre senting early growth stages (sicula + th 1^ + th 1^) of Dicellograptus bispiralis bispiralis can be distinguished from similar specimens of the other described dicellograptid species by the size of the sicula, which is 1.2 to 1.4 mm long. 2 2 0 Remarks The described specimens are identical to those described and illus trated by Ruedemann (1947) as Dicranograptus furcatus bispiralis. In his plate caption, Ruedemann referred to his illustrated specimens as Dicellograptus furcatus bispiralis. Ruedemann (1947) reported a biserial proximal end for his specimens. However, the specimens described above, which are from the type locality of the subspecies, show that the proximal end is not biserial, although it may appear so in compressed, non-isolated specimens. Hall's (1847) syntypes of Graptolithus furcatus (AMNH 30455 and 30456) show biserial proximal ends and simple apertural lappers. Therefore, Dicranograptus furcatus (Hall) var. bispiralis Ruedemann is here raised to specific status and referred to as Dicel lograptus bispiralis bispiralis (Ruedemann) . Figured Specimens OSU 33030, OSU 33033-OSU 33038. 2 2 1 Dicellograptus bispiralis (Ruedemann) n.ssp. A Text-figs. 36c-d, f-h Diagnosis A subspecies of Dicellograptus bispiralis with sicula oriented symmetrically upward between stipes. Stipes show convex curvature. Stipe width increases distally: 0.3 mm at th 2; 0.5 - 0.6 mm at th 4. Thecae number 4 - 5 in 3 mm. First two thecae subhorizontal. Thecae with obtuse genicular angle, convex supragenicular wall, introverted and introtorted aperture, and mesial spines. Prominent prothecal folds on dorsal stipe margin. Material Forty-nine specimens, which were collected from an interval between 11.1 and 15.7 meters above the base of the Athens at Calera. These specimens are preserved as carbon films. Detailed features of the thecae can rarely be seen. The largest available rhabdosome has a stipe 7 mm long. Description (Text-figs. 36c-d, f-h) The sicula, which is 1.0 mm long, is oriented symmetrically upward between the stipes. The first two thecae diverge horizontally from the sicula and show upward curvature near the apertures. The orien tation of the sicula and the first two thecae, together with the shape of the stipes, characterize the proximal end. The stipes show curvature that is convex relative to the ventral stipe margin. As a result, the stipes diverge at an angle of 45 to 60 2 2 2 degrees for the first three to five thecae on each stipe. They then converge until they cross at the sixth to tenth theca on each stipe. In the largest available specimen (Text-fig. 36h) , the stipes continue diverging after crossing. The stipe width increases distally. It is 0.3 - 0.4 nm at th 2 and 0.5 - 0.6 mm at th 4. There are four to five thecae in 3 mm proximally. The thecae, which generally are poorly preserved and usually present only a lateral aspect, show a very convex curvature of the supragenicular wall. A mesial spine projects from the supragenicular wall at the point of the maximum dorso-ventral thecal width which is transverse to the base of the apertural excavation. A generally indistinct geni culum with an obtuse genicular angle and an introverted and introtorted aperture are associated with the convex supragenicular wall. The aper ture generally shows prominent pointed lappets; however, one of the available specimens (Text-fig. 36f-g) is compressed obliquely and shows lateral lappets that are similar in appearance to those described above for Dicellograptus bispiralis bispiralis. Prothecal folds characterize the dorsal stipe wall there they are located transverse to the genicula on the ventral stipe wall. The apertural excavations occupy a third to a half the stipe width and half the length of the supragenicular wall. Remarks • The specimens described above resemble D. bispiralis bispiralis in the shape of their stipes, but they differ slightly from bispiralis bispiralis in stipe width, thecal density, and sicular orientation. 223 Because the morphology of the apertural lappets that can be determined on one theca on one of the available specimens is extremely characteristic of 2 » bispiralis, the specimens described above are regarded as being related closely to D. bispiralis bispiralis. The specimens described above do not occur together with 2 - bispiralis bispiralis. and they probably occupy a lower stratigraphie position in relation to 2 » bispiralis bispiralis; therefore they are assigned herein to sub specific status. Figured Specimens OSU 33028, OSU 33029, OSU 33031, OSU 33032 224 Dicellograptus geniculatus Bulman, 1932 (Text-figs. 38a-b) 1932 Dicellograptus geniculatus n.sp., Bulman, p. 19-21, text-fig. 6 , Pl. 1, figs. 9 - 13. Type Data The holotype of Dicellograptus geniculatus is the specimen figured by Bulman (1932, text-fig. 6 ; PI. 1, figs. 9, 10). It is stored in the Holm collection (specimen number 2762) at the Swedish Museum of Natural History. This specimen was examined by the writer, and it agrees with Bulman's description and illustrations. Diagnosis A species of Dicellograptus with the stipes diverging at 120 degrees angle. Sicula erect between the stipes; first pair of thecae horizontal. Stipes uniformly 0.3 mm wide. Thecae number 3 in 2 mm proximally. Material The available material consists of approximately 100 specimens that are preserved as carbon films on black shale surfaces. The thecal apertures distal to the first thecal pair can rarely be seen. The largest specimen has an 18 mm long stipe. All of the specimens were collected from an interval between 8.5 and 11.8 meters above the base of the Athens Shale at Calera. Description (Text-figs. 38a-b) Ths sicula, which is 1.0 mm long, is oriented right upward symmetrically between the two stipes. Th 1^ has a wide D-shape, 225' whereas th diverges horizontally from the sicula. Both thecae ex hibit slight upward curvature near the apertures. At the second pair of thecae, the stipes bend upward to enclose an angle that ranges from 100 to 160 degrees. The proximal end of the rhabdosome is very broad 1 o in lateral view; the distance between the apertures of th 1 and th 1 averages 1.5 mm. The stipes are thin (0.3 mm) and of uniform width. Proximally there are 1.5 thecae in 1 mm and 3 thecae in 2 mm. The thecae exhibit a distinct geniculum, very slight curvature of the supragenicular wall in the aperture region, and introtorsion and slight introversion of the aperture. A flange projects from the geniculum. The apertural excavations occupy one-half of the stipe width and one-fourth to one- fifth of the length of the supragenicular wall. Subapertural, mesial spines are present on the first theca of each stipe. Prothecal folds are present on the dorsal stipe walls at the level of the thecal aper tures. Remarks The described specimens agree closely with Dicellograptus genic ulatus Bulman in all of the features discussed above except for thecal density. However, in the described specimens, the thecal density was measured in the proximal part of the rhabdosome, whereas in Bullman's specimens, it was measured on distal fragments. Bullman's specimens are isolated and preserved in full relief, whereas the described speci mens are compressed, carbon films. The morphological features that 226 support the assignment of the described specimens to Bulman's species are: 1) the upright sicula; 2) the U-shape of th 1^ and the L-shape of th 1^; 3) the shape of the supragenicular wall; 4) the thinness of the stipes; and 5) the presence of genicular flanges. The rhabdosome of JD. geniculatus resembles many leptograptids in the thinness of the stipes, the upright orientation of the sicula, and the horizontal orientation of the stipes. Thecal morphology favors a dicellograptid assignment. The prothecal folds, as described by Jaanusson (1964, fig. 6 ), closely resemble those described in other dicellograptid species. The introtorted aperture and the genicular flange suggest a development of the infragenicular thecal wall that is similar to that found in other dicellograptids. Figured Specimens OSU 33039 - OSU 33040 227 Dicellograptus gurleyi gurleyi Ruedemann, 1908 (Text-figs. 27-28, 39-42) 1896 Dicellograptus gurleyi Lapworth sp. nov., Gurley, p. 70-71. 1908 Dicellograptus gurleyi Lapworth, Ruedemann, p. 303-306, figs. 223-228, PI. 19, figs. 7-9. 1947 Dicellograptus gurleyi Lapworth, Ruedemann (Partim), p. 382-383, PI. 63, figs. 21-28 (not figs. 29-33). 1947 Dicellograptus gurleyi var. exilis n. var., Ruedemann, p. 383, PI. 63, figs. 34-36. ?1952 Dicellograptus gurleyi Lapworth, Decker, PI. 1, fig. 20; PI. 2, fig. 35. ?1952 Dicellograptus gurleyi var. exilis Ruedemann, Decker, PI. 1, ?1960 Dicellograptus gurleyi Lapworth, Berry, p. 75, PI. 16, fig. 9. 1960 Dicellograp tus gurleyi var. exilis Ruedemann, Berry, p. 75. 1960 Dicellograptus sextans var. exilis Elies and Wood, Berry, p. 77, PI. 15, fig. 11. 1963 Dicellograptus gurleyi Lapworth, Ross and Berry, p. 105-106, PI. 6 , figs. 14, 16, and 18. Type Data Dicellograptus gurleyi was originally described and illustrated in a manuscript by Lapworth. His specimens were obtained from the Normanskill Shale at Stockport, New York, however today these speci mens are lost. Gurley (1896) printed Lapworth's description, but he gave no illustrations. The first published description of 0. gurleyi with illustrations is that of Ruedemann (1908). Ruedemann published Lapworth's manuscript drawings, figured additional specimens from the Stockport locality, and referred to Lapworth as the author of the species. 228 Because Lapworth's specimens are not available and his manuscript was not published, Ruedemann, who published the first valid description of the species, is considered herein to be the author of Dicellograptus gurleyi gurleyi. The lectotype has yet to be selected among Ruedemann*s figured specimens that are stored at the New York State Museum, Albany (specimen numbers NYSM 7040-7047) and the U. S. National Museum, Washington, D. C. (specimen numbers unknown) . An examination of the specimens stored in Albany shows them to differ in stipe width from that figure reported by Ruedemann (See Remarks). Diagnosis A species of Dicellograptus with the stipes diverging at 60 to 90 degrees'angle. Sicula inclined at varying degrees toward stipe no. 2. First two thecae U-shaped. Prominent mesial spines on first two thecae and occasionally on variable number of successive thecae. Stipe width uniform, 0.30 to 0.55 mm, depending on state of preservation. Proxi mally, 7 to 7.5 thecae in 5 mm. Theca with distinct geniculum, geni cular flange, 90 degree genicular angle, convex supragenicular wall, and introverted aperture. Apertural excavations occupying a third of stipe width and a fourth of length of supragenicular wall. Prominent prothecal folds on dorsal stipe margin. Material Several hundred specimens of Dicellograptus gurleyi gurleyi were obtained from the Pratt's Ferry, Pratt's Syncline, and Calera sections. This material represents the six different states of preservation listed below. 229 1 ) isolated and uncompressed specimens, two large rhabdosomes with proximal ends and ten distal fragments from a horizon 90 meters above the base of the Athens Shale at Pratt's Syncline. 2 ) isolated and compressed specimens. six large rhabdosomes with proximal ends and approximately 1 0 0 growth stages with less than four thecae from the upper one-half of the Pratt's Syncline section. 3) isolated, compressed and uncompressed, pyritized internal molds. 25 specimens with proximal end and more than four pair of thecae, 25 proximal ends with less than four pair of thecae, and ten distal fragments with more than 10 thecae from a hosizon 107.5 meters above the base of the Athens Shale at Pratt's Syncline. 4) non-isolated and compressed specimens with periderm preserved. Approximately 100 specimens from the entire Pratt's Ferry and Pratt's Syncline sections. 5) non-isolated and uncompressed, pyritized internal molds, commonly with periderm preserved; more than 50 specimens from the Pratt's Ferry section and the lower 17 meters of the Athens Shale at Pratt's Syncline. 6 ) carbon films, more than 200 specimens from the Calera section. Description. Isolated, Compressed and Uncompressed Specimens (Text-figs. 39a-g). The largest available specimen has five thecae on one stipe. The axil is rounded, and the diverging stipes enclose an angle of 60 to 90 degrees. In lateral view, the rhabdosome has a rounded proximal end produced by 230 the U-shape of the first two thecae. Three prominent spines characterize the proximal end: the downward-directed virgella, which attains a length of 0.5 mm, and horizontally-directed mesial spines (0.5 mm long) on the first two thecae. The sicula, which is 1.0 mm long, is inclined toward the dorsal wall of stipe no. 2. Througjiout their length, the stipes show a uniform width of 0.35 mm in uncompressed specimens and 0.40 - 0.45 mm in fully compressed specimens. The theca number 4.5 in 3 mm and 7 - 7.5 in 5 mm proximally. Prothecal folds, which are prominent features on the dorsal stipe margin, are consistently located transverse to the genicula on the ventral stipe margin. The thecae are 1.5 mm long and overlap for half of their length. An individual theca develops from the prothecal fold that is transverse to the second preceding thecal aperture. Each theca exhibits a dis tinct geniculum with a genicular angle of 90 degrees. A flange projects 0.15 - 0.40 mm from the geniculum. The supragenicular wall is convex in such a way that the maximum dorso-ventral thecal width occurs two- thirds the distance from the geniculum to the aperture. Thecal spines are only present on the first two thecae of all the isolated, non- pyritized specimens. The thecal aperture is a circular tube that is introtorted and introverted. The aperture opens in a dorso-distal direction with the plane of the aperture forming an angle of 45 degrees with the dorsal margin of the stipe. The lateral margins of the aperture are slightly raised to form lappets. In lateral view, the apertural excavations 231 have the shape of an inverted comma and occupy a third of the stipe width and a fourth the length of the supragenicular wall. The lateral margins of the excavations are rimmed by lists which extend from the dorso-lateral comers of the aperture to the base of the infragenicular wall, then along the lateral margins of the infragenicular wall to the geniculum. Grooves, representing the traces of the interthecal septa, are present on the lateral stipe walls and extend from the prothecal folds to the bases of the apertural excavations. Isolated, Compressed and Uncompressed, Pyritized Internal Molds. The largest available rhabdosome, which is 6.3 mm long, has eight thecae on one stipe. The stipes consistently diverge at an angle of 60 or 90 degrees. Distally, the stipes begin to converge attaining the appearange of plier-handles. In those specimens with an axillary angle of 60 degrees, the sicula is tilted toward, but not completely in contact with, the dorsal wall of stipe no. 2. The prothecal fold, from which th 3 develops, is in contact with the initial bud of th 1 . In those specimens with a 90 degree axillary angle, the sicula is usually in contact along its entire length with the dorsal margin of stipe no. 2 , and the prothecal fold, from which th 3^ develops, is not in contact with the initial bud of th 1 ^. The stipes show a uniform width throughout their length. Among the specimens studied, this width ranges from 0.35 mm in uncompressed specimens to 0.45 mm in compressed specimens. The thecae number 4.5 in 3 mm and 7 - 7.5 in 5 mm, proximally. Mesial spines are not present on any of the thecae because these specimens are internal molds. 232 Non-isolated and Compressed specimens with Periderm Preserved (Text-figs. 40b, e, f). In fully compressed specimens, the convexity of the supragenicular wall and the curvature and prominence of the pro thecal folds are significantly greater than in specimens preserved in partial or full relief. The axillary angle is generally 60 or 90 degrees. Associated with these two axillary angles, as described above, are the orientation of the sicula and the position of the th 3^ pro thecal fold. The largest specimen shows a stipe 11.5 mm long. The stipe width is uniform and ranges from 0.40 to 0.55 mm among the avail able specimens. The thecae number 4.5 in 3 mm and 7 in 5 mm, proximally. Distally, there are 6 thecae in 5 mm. A flange projects from the geni culum. Prominent mesial spines are present on the first two thecae. Spines are absent from subsequent thecae. Non-isolated and Uncompressed, Pyritized Internal Molds, Commonly with Periderm Preserved (Text-figs. 40a, d,i, 1) . These specimens were not isolated but were studied on shale surfaces. They are pyritized internal molds that show partial or full relief and are commonly covered by periderm. The stipes of the largest specimen are 30 mm long. The stipes diverge at angles ranging from 45 to 90 degrees. The stipe width is uniform and it ranges from 0.28 to 0.40 mm in the available specimens. The thecae number 4.5 in 3 mm and 7 - 7.5 in 5 mm, proximally. Distally, there are 6 thecae in 5 mm. A flange projects from the geniculum. The presence of spines is variable in the available specimens. Approximately 25% of the specimens bear spines only on the first two thecae. The first two, three, or four 233 thecae bear mesial spines in the majority of the specimens. Less than 1 0 % of the specimens esdiibit spines on every thecae. Carbon Films (Text-figs. 40c, h). The largest available specimen possesses an 8 mm long stipe. Generally, the axillary angle ranges from 60 to 90 degrees. However, axillary angles as low as 25 degrees are common. The first pair of thecae bear prominent mesial spines. Spines frequently are present on a variable number of subsequent thecae. The stipe width is uniform and ranges from 0.40 to 0.50 mm. There are 4.5 thecae in 3 mm and 7.5 thecae in 5 mm, proximally. A flange projects from the geniculum. Early Growth Stages (Text-figs. 41, 42). Early growth stages (sicula + th 1^ + th 1^) of Dicellograptus gurleyi gurleyi are ex tremely difficult to differentiate from those of 2 » sextans on the basis of morphology. In both species, the sicula is approximately 1.0 mm long; the first two thecae show the same orientation and curvature; and the ventral wall of the horizontally directed part of th 1 ^ is at the level of the sicular aperture. The early growth stages could be distinguished in those graptolitiferous intervals in which the dicel lograptid fauna, as represented by large rhabdosomes, is monospecific. Growth stages of Dicellograptus gurleyi gurleyi with th 2^ can be easily recognized by the angle of the geniculum, the presence of genicular flanges, and the curvature of the supragenicular wall. Remarks Variations within Species. A great deal of variation occurs among the specimens of Dicellograptus gurleyi gurleyi described above. Some 2 3 4 of this variation, such as in the stipe width (Table 7 ) and the cur vature of the supragenicular wall and the prothecal folds, is directly associated with the state of preservation. In uncompressed specimens, the stipes are 0.35 mm wide. In fully compressed specimens, the stipe width may be as large as 0.50 mm, and the curvature of the supra genicular walls and the prothecal folds can be greatly exaggerated. Some variation is inherent to the species as defined above. An example is the axillary angle (60 or 90 degrees) and the associated features. The low angle of divergence reported for some specimens that are preserved as compressed carbon films is interpreted as being due to the state of preservation. A third type of variation is exemplified by the thecal spines. In those specimens obtained from the lower part of the Athens Shale at Pratt’s Ferry and Pratt's Syncline and from the entire range of Dicellograptus gurleyi gurleyi at Calera, there is a great deal of variation in the presence of spines in distal thecae. In many specimens, spines are present only on the first two thecae. In an equally large number of specimens, the first two, three, or four thecae of each stipe bear spines. And, in a few specimens, spines are present on every theca. This variation in the number of thecae with spines is shorn in Text-figures 27-28. Specimens from the lower stratigraphie intervals (e.g., the lower part of the Calera section) show spines on only the first, and rarely the second pair of thecae. Specimens from higher stratigraphie intervals (e.g., upper part of the Calera section and 235 TABLE 7: Variation in stipe width of D^. gurleyi gurleyi in relation to state of preservation. State of Preservation Stipe Width Th/Smm prox. isolated, full relief 0.35 ram 4.5 isolated, fully compressed 0.38 mm 4.5 non-isolated, compressed, periderm preseirved 0.42-0.45 mm 4.5-4.75 non-isolated, full relief 0.35 mm 4.5-4.75 non-isolated, fully compressed, periderm poorly preserved, spines on all thecae 0.28 ram* 4.75 isolated, pyritized internal molds full relief 0.35 ram 4.5 fully compressed 0.45 mm 4.5 compressed, carbon films 0.40-0.50 mm 4.5 *These specimens are discussed at the end of the Remarks section. 236 EXPLANATION OF TEXT-FIGURE 27 Stratigraphie variation in thecal spines in specimens of Dicellograptus gurleyi gurleyi Ruedemann from the Calera section. 1. Spines on first pair of thecae; 2. Spines on first two pairs of thecae; 3. Spines on first three to five pairs of thecae; 4. Spines on all thecae. Total number of specimens from 5 m intervals: 11-15 m: 23; 16-20 m: 109; 21-25 m: 35; 26-30 m: 57; 31-35 m: 32; 36-40 m: 63; 41-45 m: 25; 46-50 m: 47; 51-55 m: 12; 56-60 m: 12; 61-65 m: 0; 66-70 m: 0; 71-75 m: 3. I f t I r-hH* 'g n a> to •n J Distance in metres above base of Athens Shale o to to /to ■ D (71 O O LZZ 238 EXPLANATION OF TEXT-FIGURE 28 Stratigraphie variation in thecal spines in specimens of Dicellograptus gurleyi gurleyi Ruedemann from the Pratt's Ferry- Pratt's Syncline section. Symbols used as in Text-figure 27. Total nvmber of specimens from 5 m intervals: 0-5 m: 2; 6-10 m: 15; 11-15 m: 140; 16-20 m: 35; 66-70 m: 17; 91-95 m: 5; 96-100 m: 62; 106-110 m: 121; 116-120 m: 5; 121-125 m: 3; all other intervals: 0. 239 140 120 w 100 S 80 60 5 40 20 #2 100 percent of total number of specimens in 5 m Text-figure 28 240 lower part of the Pratt's Ferry and Pratt's Syncline sections) commonly show spines on additional thecae. In the highest strati graphie intervals (e.g., upper part of the Pratt's Syncline section), specimens show spines on only the first pair of thecae. The number of thecae that bear spines can be used to correlate between the Pratt's Ferry-Pratt's Syncline and Calera sections. As discussed for alabamensis. the variation of thecal spines may be genetically controlled and ecologically influenced. Constant Features within Species. In spite of the variation pro duced by the factors listed above, all the described specimens are included in Dicellograptus gurleyi gurleyi because of certain consistent features. Many of the features are qualitative, such as the presence, position, and shape of prothecal folds, the curvature of the supra- genicular wall, the presence of genicular flanges, and an introverted and introtorted aperture with lateral lappets. The one quantitative feature common to all the states of preservation and variants is the thecal density. Comparison to Previously Described Specimens of D. gurleyi. Lapworth's (1890, Ms.) specimens of Dicellograptus gurleyi gurleyi, which were first illustrated by Ruedemann (1908, figs. 223, 226), are very similar in appearance to the specimens described above. Lapworth's description (1890, Ms.) of 2" gurleyi, which was published by Gurley (1896), reports an initial stipe width of 0.5 mm and a distal stipe width of 1.25 mm. However, figure 223 of Ruedemann (1908), which is reproduced from Lapworth's original manuscript, shows a specimen with a uniform stipe 241 width of 0.5 mm. The stipe width of 1.25 mm may be in error or may be based on fully compressed or highly distorted specimens. The specimens described above agree closely with Ruedemann’s (1908, 1947), descriptions and illustrations of Dicellograptus gurleyi except for those from the Viola Limestone (1947, PI. 63, figs. 29-33). The Oklahoma specimens were first described and illustrated by Ruedemann and Decker (1934). They differ from the above-described specimens in that the sicula is upright between the two stipes and the first two thecae are horizontal in their direction of growth. For these reasons, Ruedemann’s (1947, PI. 63, figs. 29-33) and Ruedemann and Decker’s (1934, p. 310-311, PI. 41, figs. 3, 3a, 4, 4a, 5, 5a, 6, and 6a) specimens are here excluded from 2- gurleyi. Measurements were made by the author on Ruedemann’s figured specimens (NYSM 7040-7047), which are preserved as carbon films. The maximum stipe width is 0.5 mm, not 0.7 - 0.8 mm as reported by Ruedemann (1908, 1947). These specimens are illustrated in text-figures 45, 46, 206, 228 and Plate 19, figures 7, 8, and 9 of his 1908 Memoir and in Plate 63, figures 22 and 27 of his 1947 Memoir. Ruedemann's (1947) type specimens of Dicellograptus gurleyi exilis which are stored in collection 440 0 at the U. S. National Museum, Washington, D. C., are similar in all respects to D. gurleyi gurleyi except for a smaller stipe width (0.3 to 0.4 mm). The type locality for these specimens is the section at Pratt's Ferry, and Ruedemann's illustrations (Plate 63, figs. 34-36) are of specimens in the collections of the uses (loc. 440 0, Butts Coll.). These specimens are comparable 242 to those described above from the Pratt's Ferry section. As such, they owe their smaller width to their state of preservation, and they are here included in D. gurleyi gurleyi. Decker (1952) reports Dicellograptus gurleyi and gurleyi exilis from several localities throughout the Southern Appalachians. His collections from Pratt's Ferry, Simpson Springs, and Calera in Alabama and from near Bristol, Tennessee were studied by the author. Some specimens that Decker identified as gurleyi or D. gurleyi exilis are comparable to the specimens described above. However, some of his specimens are too poorly preserved to be identified to generic or specific level, represent other species of Dicellograptus, e.g., sextans, or even other genera, e.g., Azygograptus. Some specimens that Decker identified as other species, such as D. sextans exilis are actually representatives of 2- gurleyi gurleyi. Because the specimens illustrated by Decker (1952) cannot be identified, it is doubtful whether these should be included with 2» gurleyi gurleyi in the synonymy above. Berry's (1960) figured specimen of Dicellograptus gurleyi agrees closely with the specimens described above. Berry (1960) describes 2- gurleyi exilis. No reference specimen could be found in his collection; however, on the basis of his description, his specimens are comparable to those from the Pratt's Ferry section. The specimen of sextans var. exilis Elies and Wood (YPM 20349) that is illustrated by Berry (1960, pi. 15, fig. 11) is herein grouped with gurleyi gurleyi because it shows: 1) a slightly convex supragenicular wall, 2) a 90 degree genicular angle, 3) a genicular flange, 4) a uniform stipe width of 0.4 mm, and 243 5) 4.5 thecae in 3 mm proxlmally. Specimens referred to as Di. gurleyi by Ross and Berry (1963) are very similar to the Alabama specimens. Comparison to Other Species of Dicellograptus. The Alabama speci mens of Dicellograptus gurleyi gurleyi that have an axillary angle of 60 degrees bear a close resemblance to D. vagus Hadding, 1913. Among their specimens, Toghill (1970b) and Hadding (1913) report two distinct forms of 2" cf. vagus and D. vagus, respectively; 1) those in which the axillary angle is 30 to 60 degrees, the sicula is inclined toward stipe no. 2, and the first pair of thecae are U-shaped (Text-fig. 40a), and 2) those in which the axillary angle is 80 to 90 degrees, the sicula is symmetrically placed between the two stipes, and the first pair of thecae are horizontal in their direction of growth (Text-fig. 40k). Only the form with the smaller axillary angle is represented among Berry's (1964) specimens of 2- vagus (Text-fig. 40j), and it is this form which bears a striking resemblance to those specimens of 2* gurleyi gurleyi with an axillary angle of 60 degrees. The forms of Dicellograptus gurleyi gurleyi and those specimens of 2- vagus with the smaller axillary angle are similar in the following respects: 1) the sicula is inclined toward the dorsal wall of stipe no. 2; 2) the first pair of thecae are U-shaped; 3) the sicula serves as the dorsal wall of the th 2^ and th 2^ prothecae; 4) the stipe width is uniform and 0.5 mm; 5) the thecal density is 4.5 thecae in 3 mm proximally; 6) the aperture is introtorted and introverted; and 7) ex ternally, the prothecal folds are structurally similar. 244 Dicellograptus vagus differs from gurleyi by the generally smaller axillary angle, the sigmoid curvature of the supragenicular wall, and the absence of genicular flanges. However, in those specimens of vagus with the smaller axillary angle, the angle at which the stipes diverge increases to 60 degrees after the first four to six pair of thecae. In Toghill's (1970b) specimens of cf. vagus that show the smaller axillary angle, the divergence of the stipes is uniform and closely resembles that in gurleyi gurleyi. Text-figure 40h illustrates a specimen of 2" gurleyi gurleyi from Calera that is preserved as a carbon film. The divergence of the stipes is remarkably similar to that in vagus. The differences in the curvature of the supragenicular wall has been con sidered as a diagnostic feature. However, it should be used with caution becanse in some specimens of D. gurleyi gurleyi, the supragenicular wall of some thecae occasionally shows a sigmoid curvature, e,g., th 3^ and th 8^ of Text-figure 401 and th 6^ of Text-figure 40i. Also, in ex ternal molds of vagus, the thecae often show convex, not sigmoid, curvature, e.g., in Text-figure 40j (drawn from specimens illustrated by Berry, 1964, PI. 10, fig. 4). Toghill's (1970b, fig. 2d-2f and 2h-2k) specimens could be interpreted as showing convex curvature of the supragenicular wall. The specimens of Dicellograptus gurleyi gurleyi with a 90 degree angle show a sicula completely embedded in the dorsal wall of stipe no. 2 and a general U-shape to the first two thecae of the rhabdosome. 245 Thus, they are not closely similar to the specimens of vagus in which the axillary angle is 90 degrees. Specimens, which are described below as Dicellograptus gurleyi n.ssp. A, are somewhat similar to those specimens of vagus (Hadding, 1913) and cf. vagus (Toghill, 1970b) in which the axillary angle is large and the first pair of thecae are horizontal. Dicellograptus sextans (Hall) occurs together with the Alabama speci mens of p. gurleyi gurleyi. It can be recognized by its stipe width, which increases distally, and the greater convexity of the supragenicular walls, p. divaricatus (Hall) and its subspecies bear superficial resem blance to p. gurleyi gurleyi. However, the stipes do not show the upward initial direction of growth that is so characteristic of p. gurleyi gurleyi, and the supragenicular wall is straight. Fifteen specimens, collected from the Pratt's Ferry section and the equivalent part of the Pratt's Syncline section, show only 0.28 mm wide stipes. These specimens show spines on every theca, but some specimens of Dicellograptus gurleyi gurleyi with wider stipe widths also show spines on every theca. The narrow specimens present a rhabdosomal appear ance that is very different from specimens with wider stipe widths. In fact, these narrow specimens closely resemble p. sextans (Hall) var. perexilis Ruedemann (1908, 1947) in rhabdosomal appearance. The narrow specimens are not here recognized as a species separate from p. gurleyi gurleyi because they are poorly preserved and few in number. The thecal structures, which are rarely preserved, are similar to those found in p. gurleyi gurleyi. Ruedemann's (1908, 1947) type specimens of p. sextans perexilis (NYSM 7056, 7057) also show the same thecal structures. 246 Figured Specimens OSÜ 33044 - OSÜ 33073 247 Dicellograptus gurleyi Ruedemann n .ssp. A (Text-figs. 38c-e) Diagnosis A subspecies of Dicellograptus gurleyi with sicula upright and symmetrically located between stipes. Axillary angle 60 degrees. Th 1^ U-shaped; th 1^ subhorizontal. Stipes of uniform width, 0.4 - 0.5 mm. Proximally, 4 thecae in 3 mm. Theca with distinct geniculum, genicular flange, convex supragenicular wall, and introverted and introtorted aperture. Prominent prothecal folds on dorsal stipe margin. Material The available material consists of 40 specimens that are preserved as carbon films on a black shale matrix. All of the specimens were obtained from an interval between 10.5 and 11 meters above the base of the Athens Shale at Calera. Description (Text fi^s. 38c-e) The largest available rhabdosome has a 9 mm long stipe. The axillary angle is in most specimens 60 degrees, but may range as low as 30 degrees in single specimens. The axil is acute. The sicula which is 1.1 mm long, is upright within the axil and symmetrically located between the two stipes. Th 1^ is U-shaped. Th 1^ extends obliquely upward away from the sicula. The first two thecae and, in a few specimens, subsequent thecae bear mesial spines. The stipe width is uniform and varies from 0.4 - J.5 mm in different specimens. There are 4 thecae in 3 mm proximally and 3 to 3.5 thecae 248 in 3 mm distally. The thecae exhibit a distinct geniculum, a convex supragenicular wall, and an introtorted and introverted aperture. A genicular flange is present in only a few specimens. Its rarity is probably due to the poor state of preservation. The thecal excavations occupy a half of the stipe width and a fourth of the length of the supragenicular wall. Prominent prothecal folds on the dorsal stipe wall are located transverse to the genicula on the ventral stipe wall. Remarks The available specimens are identical in appearance to Dicellograptus gurleyi gurleyi except for the upright position of the sicula between the two stipes and the relatively straight shape of the distal part of th 1 ^. gurleyi n.ssp. A could possible be included with the other forms referred to ]D. gurleyi gurleyi. However, it is here distinguished as a separate subspecies because apart from its morphological characteristics, it is restricted to a limited stratigraphie interval; which is immediately below the first appearance of typical specimens of gurleyi gurleyi at Calera. Hadding (1913; PI. 4, fig. 18) describes and illustrates a similar form, as referred to as Dicellograptus vagus, in which the sicula is upright and th 1^ is relatively straight. Text-figure 40k is a drawing of one of Hadding's syntypes (LG 2438t), which primarily differs from D. gurleyi n.ssp. A by the sigmoid curvature of the supragenicular wall. Toghill (1970b, fig. 2d-2f, 2k) describes and illustrates specimens of Dicellograptus cf. vagus *at are similar in appearance to D. gurleyi n.ssp. A. His specimens show an upright sicula, a somewhat horizontal 249 direction of growth for th and th 1^, a stipe width of 0.5 mm, and 4.5 thecae in 3 mm, proximally. However, Toghill's specimens are so poorly preserved that thecal characters can not be adequately compared with those of D. gurleyi n.ssp. A. Figured Specimens OSU 33041 - OSU 33043 250 Dicellograptus sextans (Hall, 1847) (Text-figs. 29-30, 43-46) 1847 Graptolithus sextans n.sp.. Hall, p. 273, Pl. 74, fig. 3a-e. 1904 Dicellograptus sextans (Hall), Elles and Wood, Part 4, p. 153 - 154, Pl. 21, figs. la-e. 1904 Dicellograptus sextans (Hall) var. exilis var. nov., Elles and Wood, Part 4, p. 155, fig. 97, Pl. 21, figs. 2a-d. 1908 Dicellograptus sextans (Hall), Ruedemann, p. 306-308, fig. 229, 230, Pl. 19, fig. 1. 1908 Dicellograptus sextans var. exilis Elles and Wood, Ruedemann, p. 309-310, fig. 231. 1908 Dicellograptus sextans var. tortus nov., Ruedemann, p. 309-310, fig. 232. 1913 Dicellograptus sextans Hall var. exilis Elles and Wood, Hadding, p. 55-56, text-fig. 20a-e. 1947 Dicellograptus sextans (Hall), Ruedemann, p. 386-387, Pl. 64, figs. 28-31. 1947 Dicellograptus sextans (Hall) var. exilis Elles and Wood, Ruedemann, p. 387, Pl. 64, figs. 32, 33. 1947 Dicellograptus sextans (Hall) var. tortus Ruedemann, Ruedemann, Pl. 64, fig. 36. ?1952 Dicellograptus sextans (Hall), Decker, Pl. 1, fig. 21; Pl. 2, fig. 30. 31952 Dicellograptus sextans var. exilis Ruedemann, Decker, Pl. 2, fig. 33. 31960 Dicellograptus sextans(Hall), Berry, p. 76-77. non 1960 Dicellograptus sextans var. exilis Elles and Wood, Berry, p. 77, Pl. 15, fig. 11. 1963 Dicellograptus sextans J. Hall, Ross and Berry, p. 106-107, Pl. 6 , figs. 1 0 , 1 1 , and 2 2 . 1963 Dicellograptus sextans var. exilis Elles and Wood, Rossand Berry, p. 107, Pl. 6 , figs. 7, 15. 251 Type Data Hall's (1847) syntypes of Graptolithus sextans are stored at the American Museum of Natural History, New York. These specimens have been examined by the writer, and specimen numbered AMNH 30454 is here designa ted as a lectotype. The lectotype, which is illustrated by Hall (1847; PI. 74, fig. 3c) has an axillary angle of 60 degrees. The longest stipe has 9 thecae and widens from 0.53 mm at th 2 to 0.7 mm at th 5. The thecae number 4.5 in 3 mm proximally and resemble closely those described below. They show a very convex supragenicular wall, a mesial spine, and an introtorted and very introverted aperture with prominent apertural lappets. Undulations, representing prothecal folds, are present on the dorsal stipe margin. Diagnosis A species of Dicellograptus with 60 to 70 degree axillary angle. Sicula upright between stipes or inclined at varying degrees toward stipe no. 2. First two thecae U-shaped. Stipe width increases dis tally; 0.40 - 0.55 mm at th 2; 1.0 mm maximum distal width. Proximally, 7 thecae in 5 mm; distally 5.5 thecae in 5 mm. Theca with obtuse geni cular angle, very convex supragenicular wall, introtorted and introverted aperture, and mesial spine. Prominent prothecal folds on dorsal stipe margin. Material Several hundred specimens of Dicellograptus sextans were obtained from the Pratt's Ferry, Pratt's Syncline, and Calera sections. This material represents the following states of preservation: 252 1 ) isolated, compressed and uncompressed specimens. 45 specimens with proximal end and two or more pair of thecae; more than 80 early growth stages with less than two pair of thecae; and 19 distal fragments, mostly consisting of three to five thecae. All of these specimens are from the upper half of the Pratt's Syncline section. 2 ) non-isolated and compressed specimens with periderm preserved. 27 specimens, mostly large rhabdosomes, from the Pratt's Ferry and Pratt's Syncline sections. 3) carbon films, approximately 175 specimens from the Calera section. Description; Isolated specimens (Text-figs. 43a-e);The largest available rhabdosome has a 4.6 mm long stipe that consists of seven thecae. The stipes con sistently diverge at an angle of 60 to 70 degrees. The axillary angle may rarely be as small as 30 degrees or as large as 90 degrees. The proximal end of the rhabdosome has a characteristic lateral view. The orientation of the stipes produces a V-shaped rhabdosome. The U-shaped of the first two thecae add a broad appearance to the base of this V. The sicula is 1.0 mm long. It is either inclined toward, and partly in contact with, the dorsal wall of stipe no. 2 in such a way that the prosicula remains free within the axil, or it is in contact along its entire length with the dorsal wall of stipe no. 2 (See Remarks). The sicula is rarely upright between the two stipes. The th 2^ and th 2% prothecae use the sicula as their dorsal walls. Mesial spines on th 1^ 253 and th are 0.3 mm long. The virgella attains a length of 0.6 mm. These three spines add to the characteristic appearance of the proximal end. The stipes have the appearance of a string of beads because of the prominent curvature of the prothecal folds and the supragenicular walls on the dorsal and ventral stipe walls, respectively. The width of the stipes increases distally from 0.40 mm at th 2 to 0.70 ram at th 7, which is the most distal theca on the largest available rhabdosome. There are 4.5 to 5.0 theca in 3 mm proximally. The thecae number 3.5 in 3 mm distally. Prothecal folds, which are prominent features on the dorsal stipe margin, are consistently located transverse to the genicula on the ventral stipe margin. The thecae are 1.0 mm long at th 3. Distally, they increase in length to a maximum of 1.4 mm at th 5, after which their length is constant. They overlap for half their length. An individual theca orig inates at the prothecal fold that is transverse to the geniculum of the preceding theca. There is a gradational change in the shape of the thecae along the stipe. The proximal thecae show the following characters: 1) the supra genicular wall is very convex; 2 ) the geniculum is distinct, but the genicular angle is obtuse (150 - 180 degrees); 3) a prominent mesial spine, which is 0.3 mm long, projects from the supragenicular wall at the point of the maximum dorso-ventral thecal width; 4) the mesial spine is located transverse to the base of the thecal excavation and is situated two-third of the distance from the geniculum to the aperture; 254 and 5) the aperture, which is introtorted and introverted, opens in a dorso-distal direction. The distal thecae (Text-fig. 43d) differ from the proximal ones as follows: 1 ) the supragenicular wall is less convex; 2) the geniculum is more distinct with a genicular angle of 130 to 150 degrees; 3) the mesial spine and the point of maximum dorso-ventral thecal width which are transverse to the base of the thecal excavation, are situated at one-half of the distance from the geniculum to the aperture; and 4) the aperture, which is introtorted but slightly less introverted, opens in a distal direction. The shape of the proximal thecae grada- tionally changes to that of the distal thecae between th 5 and th 8 . The thecal apertures are simple tubes with an ellipsoid cross- section. The long axis of the ellipse extends between the lateral walls. Prominent pointed lappets are situated on the dorso-lateral comers of the aperture. Viewed laterally, the apertural excavations have the shape of an inverted comma. They occupy a third of the stipe width and over two-third of the length of the supragenicular wall. The lateral edges of the apertural excavations are rimmed by lists. The lists extend from the dorso-lateral margins of the aperture to the base of the apertural excavation, and then follow the lateral margins of the infragenicular wall to the geniculum. In those specimens in which the genicular angle approaches 180 degrees, the geniculum is recognized by the presence of the list. Grooves, which represent the traces of the interthecal septa, extend along the lateral stipe walls from the pro thecal folds to the bases of the apertural excavations. 255 Non-isolated and Compressed Specimens, with Periderm Preserved (Text-figs. 43f-g, 44b-c). The largest available rhabdosome has a 45 mm long stipe. The stipes diverge at angles ranging from 50 to 90 degrees. However, the axillary angle is generally between 60 and 70 degrees. In lateral view, the base of the axil is broad and rounded. The sicula is inclined toward the dorsal wall of stipe no. 2. Among the available specimens, there are two distinct forms based on the inclination of the sicula. In 70% of the specimens, the sicula is in contact along its entire length with stipe no. 2 (Text- fig. 43g). In the remaining 30%, only the metasicula is in contact with stipe no. 2 (Text-fig. 44b), and the prosicula remains free within the axil (See Remarks). The stipe width increases distally. It is 0.4 - 0.5 mm at th 2; 0.7 - 0.8 mm at th 5; 0.8 - 0.9 mm at th 10; and 1.0 mm thereafter. Proximally, there are 4.5 thecae in 3 mm, 7 thecae in 5 mm, and 13 - 14 thecae in 10 mm. Distally, the thecae number 5 in 5 mm and 10 in 10 mm. Specimens that are compressed in a direction normal to the lateral stipe walls show the same thecal structure as isolated specimens. Specimens that are obliquely compressed show distorted thecal structures. In such specimens, the genicula are indistinct, and can only be detected by the presence of lists which border the apertural excavations. Also, the curvature of the supragenicular walls and the prothecal folds is reduced. The apertures, when viewed ventrally, are broad, and mesial spines are commonly not visible. 256 Carbon Films (Text-figs. 44a, d-g). The largest available rhabdosome has a 5.5 mm long stipe. In lateral view, the rhabdosome is V-shaped. The stipes diverge at an angle of 60 degrees. The proximal end, in lateral view, is broad because of the U-shape of the first pair of thecae. The following three orientations of the sicula are recognized among the available specimens : 1 ) upright between the two stipes (Text-figs. 44a-f) ; 2) partly in contact with the dorsal wall of stipe no 2 , in such a way that the prosicula remains free within the axil (Text-fig. 44e) ; and 3) in contact for its entire length with the dorsal wall of stipe no. 2 (Text-figs. 44d,g; See Remarks). The stipes show a "beaded" appearance, which is produced by the curvature of the prothecal folds and the supragenicular walls. The stipe width increases distally from 0.40 mm at th 2 to 0.60 mm at th 5. There are 4.5 to 5.0 thecae in 3 mm proximally. The following thecal structures are visible on the relatively better preserved specimens: pointed apertural lappets, very convex supragenicular walls, and obtuse genicular angles. Early Growth Stages (Text-figs. 45-46). More than 80 early growth stages of Dicellograptus sextans are available for study. More than one-half of these are preserved in partial or full relief. The majority of the specimens show breakage, and growth lines are somewhat hard to discern. As discussed under Dicellograptus gurleyi gurleyi, early growth stages (sicula + th 1 ^ + th 1 ^) of sextans are extremely difficult to differentiate from those of gurleyi. Growth stages with th 2^ 257 can be easily recognized by: 1 ) the convexity of supragenicular walls; 2) the absence of genicular flanges; 3) the genicular angle ; and 4) the shape of the aperture, especially the lappets. Remarks Variation within Species. Three forms of Dicellograptus sextans can be recognized on the basis of the orientation of the sicula. This variation is stratigraphically related as shown in Text-figures 29-30. In the stratigraphically oldest specimens of 2- sextans at Calera, the sicula is consistently oriented upright between the stipes. Specimens, in which the sicula is either tilted toward or in complete contact with the dorsal wall of stipe no. 2, appear at higher horizons in the Athens shale. Specimens with inclined orientations of the sicula increase in relative abundance upward through the Athens Shale. Specimens with the upright sicula decrease in relative abundance in higher horizons of the Athens and successively they become extremely rare. This variation, expressed by a stratigraphie change in the relative abundance of the different forms, is interpreted as repre senting evolution within Dicellograptus sextans. This variation is here regarded as being intraspecific and genetically controlled because: 1 ) all three forms can generally be found in the same popu lation, as represented on a single bedding plane; 2 ) the variation within populations varies with time; and 3) the different forms are identical in all other morphological features. Evidently, the varia tion is of stratigraphical value and can be used to correlate between the Pratt's Ferry-Pratt's Syncline and Calera sections. 258 EXPLANATION OF TEXT-FIGURE 29 Stratigraphie variation in sicular orientation in specimens of Dicellograptus sextans (Hall) from the Calera section. C. Sicula in complete contact with stipe no. 2; P/M. Sicula in partial contact with stipe no. 2 , prosicula remains free within axil; U. Sicula oriented symmetrically upward between stipes. Total number of specimens from 10 m intervals: 0-10 m: 0; 11-20 m: 33; 21-30 m: 0; 31-40 m: 29; 41-50 m: 48; 51-60 m: 33; 61-70 m: 7; 71-80 m: 9. s fT hhI ?H* S N3 VO Distance in metres above base of Athens Shale ro o> 00 o o o o T3 (D ■0 N O (D 3 s 3 C 3 o- » C/) % o 3 CD (O3 O 3 é c -0 o o 6GZ 260 EXPLANATION OF TEXT-FIGURE 30 Stratigraphie variation in sicular orientation in specimens of Dicellograptus sextans (Hall) from the Pratt’s Ferry-Pratt's Syncline section. Symbols used as in Text-figure 29. Total number of specimens from 10 m intervals: 0-10 m: 13; 11-20 m: 2; 61-70 m: 13; 91-100 m: 46; 116-120 m: 5; all other intervals: 0. 261 140 120 c 100 80 P / M f 60 40 20 P/MiC P/M 0 100 percent of total number of specimens in 10 m Text-figure 30 262 Comparison to Previously Described Specimens of D. sextans. Hall's C1847) syntypes of Graptolithus sextans n.sp. resemble closely the Alabama specimens. Measurements taken on Hall's syntypes (AMNH 30453, 30454, and 30455) show the stipe width to increase distally from 0.4 mm at th 2 to 0.6 to 0.7 mm at th 10. Elies and Wood's (1904) descriptions and illustrations of Dicellograptus sextans agree closely with the above-described specimens. They defined sextans var. exilis on specimens in which the stipe width is half of that found in typical specimens of D. sextans. The repeated recognition of this variety is questioned here. Measurements of Elies and Wood's syntypes of the variety exilis show that the stipes increase in width from 0.4 - 0.6 mm proximally to 0.75 mm at th 5 (B. Erdtmann, personal communication). These values of the width closely agree with those of the Alabama specimens. Ruedemann (1908) described and illustrated specimens of Dicellograptus sextans. He describes these specimens as having a uniform stipe width of 0.8 mm. However, my examination of his figured specimens (NYSM 7053, 7054, 7055, and 10461) shows that the stipe width increases from 0.5 m m at th 2 to 0.7 mm distally. Thus, Ruedemann's specimens are similar to the Alabama specimens. Ruedemann (1908, 1947) describes Dicellograptus sextans var. exilis. He uses one of Hall's syntypes (Hall, 1847, PI. 47, fig. 3a; AMNH 30453) as the type specimen of this variety. However, my examination of this specimen shows it to be the same as the Alabama specimens described above, Ruedemann's (1908, 1947) sextans, and the other syntypes of 263 Hall's (1847) Graptolithus sextans. Ruedemann (1908, 1947) also defines the variety D. sextans var. tortus on the basis of its stipe width, which increases distably to a maximum of that of D. sextans ■ The holotype (NYSM 7058), by monotypy, also closely agrees with the Alabama specimens. Hall's (1847) syntypes of Graptolithus sextans, and Ruedemann's (1908, 1947) figured specimens of 2- sextans. Hadding (1913) describes and illustrates specimens of Dicellograptus sextans exilis. The stipes in his specimens (LO 2441t-2445t) increase in width distally from 0.5 to 0.75 mm. Thus, his specimens agree closely with the Alabama specimens. Decker (1952) reports Dicellograptus sextans and jD. sextans var. exilis from several localities throughout the Southern Appalachians. As discussed under 2- Rurleyi gurleyi, it is doubtful that specimens illustrated by Decker (1952) should be included with the specimens described above because of his questionable identifications and the poor quality of his illustrations. Berry (1960) reports Dicellograptus sextans and sextans var. exilis from the Marathon Region, Texas. He does not illustrate sextans, his description is very brief, and there.are no specimens of this form in his collections. Therefore, his specimens can not be adequately compared to the Alabama specimens. Berry's (1960) 2 . sextans var. exilis is here referred to 2 - gurleyi gurleyi. Ross and Berry's (1963) specimens of Dicellograptus sextans agree closely with the Alabama specimens. Their specimens of 2* sextans 264 var. exilis fall within the range of variation in stipe width exhibited by Alabama specimens. Comparison to Other Species of Dicellograptus. On the basis of the development of the proximal end. Dicell^graptus sextans can be grouped with D. gurleyi, D. divaricatus Salopians is, and D. vagus. It can be distinguished easily from these forms by the following characteristics: 1) the axillary angle is consistently between 60 and 70 degrees; 2) the genicular angle is obtuse; 3) the prothecal folds and suprageni- cular walls are very convex and give the stipe a "beaded" appearance; 4) mesial spines are present on every theca; and 5) the stipe width increases distally. Figured Specimens OSU 33074 - OSU 33104 265 EXPLANATION OF TEXT-FIGURE 31 Text-figure 31 a-i. Dicellograptus alabamensis Ruedemann. Isolated specimens. a,b. Reverse and obverse aspects of specimen with axillary web. Note subparallel stipes. X16.7. PS-126. OSU 32988. c,d. Reverse and obverse aspects of specimen with subparallel stipes proximally. Note sicula embedded in dorsal wall of stipe no. 2. X16.7. PS-126. OSU 32989. e.f. Reverse and obverse aspects of specimen with large axillary angle. Note sicula embedded in dorsal wall of stipe no. 2. X16.7. PS-127.5. OSU 32990. g. Reverse aspect of specimen with axillary web. Note torsion of stipes. X 8 . PS-109.2. OSU 32991. h. Reverse aspect of specimen with axillary web. Note torsion of stipes. X 8 . PS-109.2. OSU 32992. i. Reverse aspect of specimen with axillary web. Note torsion of stipes. X16.7. PS-109.2. OSU 32993. 266 g f Text-figure 31 267 EXPLANATION OF TEXT-FIGURE 32 Text-figure 32 a-1. Dicellograptus alabamensis Ruedemann. Isolated specimens. a,b. Reverse and obverse aspects of specimen with axillary web. X16.7. PS-126. OSU 32994. c,d. Reverse and obverse aspects of specimen with small axillary web. X16.7. PS-126. OSU 32995. e,f. Reverse and obverse aspects of specimen with very small axillary web. X16.7. PS-126. OSU 32996. g. Reverse aspect of specimen without axillary web. Note prothecal folds. X16.7. PS-126. OSU 32997. h. Lateral aspect of distal stipe fragment. X 8 . PS-109.2. OSU 32998. i,j. Left- and right-lateral aspects of stipe fragment. Note prothecal folds. X33.3. PS-109.2. OSU 32999. k,l. Left- and right-lateral aspects of obliquely compressed stipe fragment. Compare thecal morphology in two aspects. X8 . PS-76.4. OSU 33000. 268 Text-figure 32 269 EXPLANATION OF TEXT-FIGURE 33 Text-figure 33 a-h. Dicellograptus alabamensis Ruedemann. Non-isolated, partly to fully compressed specimens. a. Reverse aspect of specimen with large axillary angle. X4. PS-132. OSU 33001. b. Reverse aspect of specimen with small axillary angle. X4. PS-132. OSU 33002. c. Obverse aspect of specimen with small axillary angle. Dorsal margins of stipes in contact proximally. X 8 . PF-17. OSU 33003. d. Specimen with large axillary web. X4. PS-132. OSU 33004. e. Reverse aspect of specimen with large axillary angle. X4. PS-132. OSU 33005. f. Specimen with axillary web. X4. PF-20.5. OSU 33006. g. Obverse aspect of specimen with large axillary angle. X4. PS-127.5. OSU 33007. h. Pyritized specimen preserved in partial relief and representing section through rhabdosome parallel to plane of rhabdosome. Note prothecal folds, interthecal septa, sicula in contact with stipe no. 2. X16.7. PF-17. OSU 33008. 270 Text-figure 33 271 EXPLANATION OF TEXT-FIGURE 34 Text-figure 34 a-I. Dicellograptus alabamensis Ruedemann. Isolated specimens representing early growth stages. a. Sicula. Note curvature of growth lines. X33.3. PS-126. OSU 33009. b. Reverse aspect of sicula. Note spiral thread in prosicula. X33.3. PS-126. OSU 33010. c. Sicula. Note position of resorption foramen. X33.3. PS-126. OSU 330II. d. Reverse aspect of sicula. Note initial bud of th I^. X33.3. PS-126. OSU 33012. e. Reverse aspect of sicula with th I^. Note budding of th l2 from th I^. X33.3. PS-109.2. OSU 33013. f. Reverse aspect of sicula tjith th I^ and opening for th I^. X33.3. PS-126. OSU 33014. g. Reverse aspect of sicula with th I^ and opening for th l2. X33.3. PS-109.2. OSU 33015. I 9 h. Reverse aspect of sicula with th I and th I . Note th I^ growing downward. X33.3. PS-126. OSU 33016. I 9 1. Reverse aspect of siculawith th I and th I . Note that thI aperture is complete, and th Iis oriented downward as a wide hood. X33.3. PS-126. OSU 33017. 1 2 I j. Reverse aspect of sicula with th I , th I , and th 2 . X33.3. PS-126. OSU 33018. 1 2 I k. Reverse aspect of sicula with th I , th I , and th 2 . X33.3. PS-126. OSU 33019. 1 2 1 I. Reverse aspect of sicula with th I , th I , and th 2 . Note downward-directed process at aperture of th 2^ representing budding of th 2^. X33.3. PS-126. OSU 33020. 272 — a r f - - th2'> th i t h 2 Text-figure 34 273 EXPLANATION OF TEXT-FIGURE 35 Text-figure 35 a-g. Dicellograptus alabamensis Ruedemann. Isolated specimens representing growth stages. a. Reverse aspect of specimen with th 2^ developed to geniculum. X33.3. PS-126. OSU 33021. 2 b. Reverse aspect. Note that infragenicular wall of th 2 develops as an apertural flange for th 1^. X33.3. PS-126. OSU 33022. c. Lateral aspect of stipe fragment. Note development of protheca and genicular flange. X33.3. PS-109.2. OSU 33023. d. Distal aspect of obverse side of specimen. Note growth lines on dorsal wall of th 2 ^, parimary notch for th 3^. and apertural flange for th 1 . X33.3. PS-126. OSU 33024. 2 e. Reverse aspect of specimen showing development of th 2 and th 3^. X33.3. PS-126. OSU 33025. f. Obverse aspect of specimen showing growth lines in distal part of th 3^ metatheca. X33.3. PS-109.7. OSU 33026. g. Reverse aspect of specimen showing development of th 4^ and th 3^. Note dotted line marking margins of inter thecal septum between th 3^ and th 4^. X33.3. PS-126. OSU 33027. 274 IW iw.'\ Text-figure 35 275 EXPLANATION OF TEXT-FIGURE 36 Text-figure 36 a,b. Dicellograptus bispiralis bispiralis (Ruedemann). Reconstruction of rhabdosome and segment of stipe, a. X4.5, b. X9. 36 c,d. Dicellograptus bispiralis n. ssp. A. Compressed, non-isolated specimens. c. Note orientation of sicula. X9. C-11.1. OSU 33028. d. Note orientation of sicula. X9. C-11.1. OSU 33029. 36 e. Dicellograptus bispiralis bispiralis (Ruedemann). Isolated specimen. Note shape of stipes. X18.8. PS-76.4. OSU 33030. 36 f,g. Dicellograptus bispiralis n. ssp. A. Compressed, non-isolated specimen. Note orientation of stipes and thecal morphology, f. X18.8, g. X37.5. C-14.8. OSU 33031. 36 h. Dicellograptus bispiralis n. ssp. A. Compressed, non-isolated specimen. Note shape of stipes. X18.8. C-12.7. OSU 33032. 276 \ Text-figure 36 277 EXPLANATION OF TEXT-FIGURE 37 Text-figure 37 a-f. Dicellograptus bispiralis bispiralis (Ruedemann). Isolated specimens representing early growth stages. a. Reverse aspect of sicula with initial bud of th 1^. X25. PS-76.4. OSU 33033. b. Reverse aspect of sicula with th 1^ and initial bud of th l2. X25. PS-76.4. OSU 33034. c. Reverse aspect of sicula with th 1^ and th 1^. X25. PS-76.4. OSU 33035. d. Reverse aspect of sicula with th 1 , th 1 , and th 2 . X25. PS-76.4. OSU 33036. 1 9 1 e. Reverse aspect of sicula with th 1 , th 1 , and th 2 . X25. PS-76.4. OSU 33037. f. Reverse aspect of specimen m t h prothecal base of th 3^. X25. PS-76.4. OSU 33038. 278 l b y 1/ \ - ^ V Text-figure 37 279 EXPLANATION OF TEXT-FIGURE 38 Text-figure 38 a,b. Dicellograptus geniculatus Bulman. Non-isolated, compressed specimens. a. Reverse aspect. Note shape of stipes. X9. C-9. OSU 33039. 1 9 b. Obverse aspect. Note shapes of th 1 and th 1 . X18.8. C-9. OSU 33040. 38 c-e. Dicellograptus gurleyi n. ssp. A. Non-isolated, compressed specimens. c. Reverse aspect. Note orientation of sicula and morphology of thecae. X9. C-11.8. OSU 33041. d. Obverse aspect. Note orientation of sicula. X9. C-10.5. OSU 33042. e. Reverse aspect. Note thecal morphology including prothecal folds. X18.8. C-10.5. OSU 33043. 280 Text-figure 38 281 EXPLANATION OF TEXT-FIGURE 39 Text-figure 39 a-j. Dicellograptus gurleyi gurleyi Ruedemann. Isolated specimens- a,b. Reverse and obverse aspects of fully compressed specimen. X16.7. PS-109.2. OSU 33044. c,d. Reverse and obverse aspects of specimen preserved in full relief. Compare thecal morphology with compressed specimen in Text-fig. 39 a,b. X16.7. PS-102. OSU 33045. e,f. Reverse and obverse aspects of specimen preserved in full relief. X16.7. PS-102. OSU 33046. g. Lateral aspect of distal stipe fragment. Note genicular flange. X33.3. PS-102. OSU 33047. h,i. Obverse and reverse aspects of pyritized internal mold of proximal end. Note course of th 1^. X33.3. PS-119.5. OSU 33048. j. Reverse aspect. Note prothecal folds (bases) and genicular flange. X33.3. PS-109.2. OSU 33049. 282 thl2 , W\i>\n Ah2^ R thl2 Text-figure 39 283 EXPLANATION OF TEXT-FIGURE 40 Text-figure 40 a-f, h-i, 1. Dicellograptus gurleyi gurleyi Ruedemann. Non-isolated, compressed specimen. a. Lateral aspect of pyritized distal stipe fragment. Note relationship between distal end of dorsal wall of protheca and proximal end of interthecal septum. X16.7. PF-21.3. OSU 33050. b. Reverse aspect of obliquely compressed specimen. X 8. PS-132. OSU 33051. c. Obverse aspect of specimen with large axillary angle. X8. C-12.7. OSU 33052. d. Obverse aspect of specimen with mesial spine on every theca. Note thecal morphology. X 8. PF-21.3. OSU 33053. e. Reverse aspect of normally compressed specimen. Note genicular flanges. X 8. PS-117. OSU 33054. f. Obverse aspect of normally compressed specimen. X 8. PS-117. OSU 33055. h. Obverse aspect of specimen with small axillary angle. Note thecal morphology. X 8. C-19.9. OSU 33056. i. Reverse aspect. X 8. PF-21.3. OSU 33057. 1. Reverse aspect. X 8. PF-21.3. OSU 33058. 40 g, j-k. Dicellograptus vagus Hadding. Specimens are pyritized and in full relief. g. Obverse aspect of specimen figured by Hadding (1913, Plate 4, fig. 15). Note curvature of supragenicular walls. X8. LO 2435t. j . Obverse aspect of specimen figured by Berry (1964, Plate 10, fig. 4). Note curvature of supragenicular walls. X8. PMO 69858. 284 k. Obverse aspect of specimen figured by Hadding (1913, Plate 4, fig. 18). X16.7. LO 2438t. 285 i Text-figure 40 286 EXPLANATION OF TEXT-FIGURE 41 Text-figure 41 a-j. Dicellograptus gurleyi gurleyi Ruedemann. Isolated specimens representing early growth stages. a. Sicula and initial bud of th 1^. Dotted circle denotes position of resorption foramen in metasicula. X37.5. PS-109.2. OSU 33059. b. Reverse aspect of sicula with initial bud of th 1^ and lateral walls that will form opening for th 1^. X37.5. PS-109.2. OSU 33060. c. Reverse aspect of sicula with th 1 and opening for th l2. X37.5. PS-109.2. OSU 33061. 1 9 d. Reverse aspect of sicula with th 1 and th1-^ opening. X37.5. PS-109.2. OSU 33062. e. Reverse aspect of sicula with th 11 and th 1 2 opening. X37.5. PS-109.2. OSU 33063. 1 2 f. Reverse aspect of sicula with th 1 and th 1 . X37.5. PS-109.2. OSU 33064. g. Reverse aspect of sicula with th 1^ and downward-directed th l2. X37.5. PS-109.2. OSU 33065. h. Reverse aspect of sicula with th 11 and th 1 2. Note that downward-directed process restricts the th 1 aperture in such a way that openingfor th 2^ is formed. X37.5. PS-109.2. OSU32066. i. Reverse aspect of sicula with th 11 , th 12 , th 1 2 , and opening for th 2^. Note that opening for th 2^ is formed by a downward-directed process across the th 2 aperture. X37.5. PS-109.2. OSU 32067. j. Reverse aspect of specimen with th 21 and th 2 2 . Note development of th1^ apertural flange. X37.5. PS-109.2. OSU 32068. 287 f th2l __af Text-figure 41 288 EXPLANATION OF TEXT-FIGURE 42 Text-figure 42 a-f. Dicellograptus gurleyi gurleyi Ruedemann. Isolated specimens representing growth stages. a,b. Obverse and reverse aspects of specimen with partly developed th 2^ and th 3^. Note relationship of lateral and infragenicular walls of th 2^. X37.5. PS-129. OSU 33069. c. Reverse aspect of specimen showing development of th 3 protheca. Note development of th 2l apertural flange, and compare growth-line pattern on th 3^ with that in Text-fig. 42 a,b. X37.5. PS-109.2. OSU 33070. d. Reverse aspect of specimen with partly developed th 2^ and th 2^. X37.5. PS-109.2. OSU 33071. e. Obverse aspect of specimen with partly developed th 3^ and th 2^. Note that prothecal base of th 3^ only partly covers primary notch at proximal end of interthecal septum, which is turned upward. X37.5. PS-109.2. OSU 33072. f. Obverse aspect of specimen with partly developed th 3^ and th 3 . Note th 2^ apertural flange. X37.5. PS-109.2. OSU 33073. 289 pb n t h 2 p t af-iw. 2 thS^pt 3f th2 th2 af-iw af af-iw Text-figure 42 290 EXPLANATION OF TEXT-FIGURE 43 Text-figure 43 a-g. Dicellograptus sextans (Hall). a. Obverse aspect of isolated specimen. Note orientation of sicula. X1S.8. PS-102. OSU 33074. b. Reverse aspect of isolated specimen. Note orientation of sicula. X18.8. PS-102. OSU 33075. c. Reverse aspect of isolated specimen. Note orientation of sicula. X18.8. PS-102. OSU 33076. d. Lateral aspect of isolated specimen representing distal stipe fragment. Note thecal morphology, particularly prothecal folds, convex supragenicular wall, and lateral lappets. X18.8. PS-102. OSU 33077. e. Reverse aspect of isolated, obliquely compressed specimen. X18.8. PS-76.4. OSU 33078. f. Lateral aspect of non-isolated, compressed specimen representing distal stipe fragment. X9. PF-17.3. OSU 33079. g. Reverse aspect of non-isolated, compressed specimen. Note distal increase in stipe width. X9. PF-17.3. OSU 33080. 291 Text-figure 43 292 EXPLANATION OF TEXT-FIGURE 44 Text-figure 44 a-g. Dicellograptus sextans (Hall). Non-isolated, compressed specimens. a. Reverse aspect. Note orientation of sicula and "beaded" appearance of stipes. X9. C-12.7. OSU 33081. b. Obverse aspect. Note orientation of sicula. X9. PF-18.7. OSU 33082. c. Obverse aspect. X9. PS-132. OSU 33083. d. Obverse aspect. Note orientation of sicula. X9. C-50. OSU 33084. e. Obverse aspect. X9. C-33. OSU 33085. f. Obverse aspect. Note orientation of siculaand "beaded" appearance of stipes. X18.8. C-34. OSU 33086. g. Reverse aspect. X18.8. C-25-16.5. OSU 33087. 293 Text-figure 44 294 EXPLANATION OF TEXT-FIGURE 45 Text-figure 45 a-j. Dicellograptus sextans (Hall). Isolated specimens representing early growth stages. a. Reverse aspect of sicula with th 1^ bud. X37.5. PS-102. OSU 33088. 2 Reverse as,aspect of specimen showing budding of th 1 from th 1^. X37.5. PS-102. OSU 33089. c. Reverse aspect of specimen showing growth of th 1^ and th l2. X37.5. PS-102. OSU 33090. 1 2 d. Reverse aspect of specimen with th 1 and th 1 . X37.5. PS-102. OSU 33091. e. Reverse aspect of specimen with th 1^ and initial doifnward-directed part of th 1^. X37.5. PS-102. OSU 33092. f. Reverse aspect of specimen with th 1 1 and th 1 2. X37.5. PS-102. OSU 33093 12 1 Reverse aspect of :specimen with th 1 , th 1 , and th 2 . X37.5. PS-102. OSU 33094. 2 h. Reverse aspect of specimen with opening for th 2 on right side of th 2^. X37.5. PS-102. OSU 33095. i. Reverse asoect of specimen showing upward growth of th 2^. X37.5. PS-102. OSU 33096. j. Reverse aspect of specimen with th 2^ nearly complete to aperture. Note primary notch for th 3^, growth lines on dorsal and lateral wall of th 2^, and apertural flange of th 1^. X37.5. PS-102. OSU 33097. 295 2 thi t h 2'* thi th2'pt / af-iw th2 Text-figure 45 296 EXPLANATION OF TEXT-FIGURE 46 Text-figure 46 a-h. Dicellograptus sextans (Hall). Isolated specimens showing various stages of development. 2 a. Reverse aspect. Note apertural flange for th 1 . X37.5. PS-76.4. OSU 33098. b. Reverse aspect of specimen showing growth of th 3^ prothecal base over primary notch. X37.5. PS-76.4. OSU 33099. c. Reverse aspect of specimen showing constriction that separates prothecal base from rest of protheca. X37.5. PS-102. OSU 33100. d. Reverse aspect of specimen showing growth of protheca. X37.5. PS-109.7. OSU 33101. e. Reverse aspect of specimen showing growth lines in distal part of protheca. X37.5. PS-102. OSU 33102. f,g. Reverse aspect of specimen showing th 4^ prothecal base, f. X37.5, g. X18.8. PS-109.2. OSU 33103. h. Reverse aspect of specimen showing list that rims distal margin of prothecal base. X37.5. PS-102. OSU 33104. 297 Text-figure 46 298 Genus Leptograptus Lapworth, 1873 Type Species: Graptolithus flaccidus Hall, 1865 Diagnosis Ehabdosome with two horizontal to slightly reclined uniserial stipes; thecae with distinct geniculum, apertures slightly introverted, but not introtorted; th n infragenicular wall secreted by zooid of th n. Discussion Leptograptus has previously been described only on the basis of specimens preserved as carbon films on shale surfaces. Besides its general rhabdosomal shape, Bulman (1970) distinguished the genus by its leptograptid type of proximal-end development and its leptograptid type of thecae. Elies (1922) defined the leptograptid type of proximal- end development by the presence of two crossing canals and the hori zontal direction of growth of the earliest thecae. Bulman (1970) allowed for the possibility of three crossing canals and a diplograptid 2 type of proximal-end development, if th 1 was discovered to have a left-handed origin. The leptograptid type of thecae, as defined by Elies (1922) and Bulman (1970), shows gentle sigmoid curvature of the ventral wall. The following description of 1^. trentonensis, which is based primarily on isolated specimens, together with intrepretations of Whittington's (1955) figured specimens of Leptograptus? sp. ind., adds substantial new information on the detailed morphology and development of Leptograptus. 299 As described below, the proximal-end development of 1^. trentonensis 2 shows a left-handed origin of th 1 and three crossing canals; thus, it represents the streptoblastic diplograptid type of Bulman (1970). The thecae show a geniculum and an introverted aperture, features that are commonly distorted or lost in compressed specimens. Additionally, prothecal folds are associated with thecal budding, and both the infra- and supragenicular walls of a particular theca are formed by the zooid of that theca. Except for stipe width and thecal density, Whittington's (1955) figured specimens of Leptograptus? sp. ind. agree closely with I^. trentonensis ; the similarities are: 1) The sicula has a relatively high length to width ratio , as well as apertural lappets (See Whittington's figure 3). 2) Th 1^ originates 0.5 mm (approximally a third of the length of the sicula) above the sicular aperture. 3) Whittington's figures 4 and 5 show a left-handed origin of th 1^ and the presence of three crossing canals. 4) The thecal apertures show lateral lappets, slight introversion, but no introtorsion (See Whittington's figure 5). 5) A geniculum divides the free ventral wall of each theca into infra- and supragenicular walls. 6) As shown in Whittington's figures 1 and 2, the th 2^ protheca is nearly complete, yet no apertural flange is present over th 1^. Thus, the th 2^ infragenicular wall is probably formed by the th 2^ zooid. This situation most likely exists for all of the thecae 300 and accounts for the absence of introtorsion of the apertures. 7) Gentle prothecal folds are present on the dorsal stipe margin at levels transverse to the genicula on the dorsal stipe margin. 8) Mesial spines on proximal thecae are situated near the apertures. The similarities between Leptograptus? sp. ind. and jL. trentonensis strongly indicate a close affinity of these two species and these features may be diagnostic of all leptograptids. Strachan (1960, p. 53) states that the thecal form of Whittington's (1955) specimens of Leptograptus? sp. ind. is that characteristic of Dicellograptus, and thus not of the leptograptid type. However, comparison of fully compressed specimens of L. trentonensis with those preserved in full relief shows that the leptograptid type of thecal form is a distortion of the original dicellograptid thecal form. The isolated specimens of Leptograptus show many similarities to Dicellograptus. The siculae are similar, especially in the presence of lateral lappets on the sicular aperture. The length of the sicula (1.5 mm or greater) is comparable to that found in the elegans group of dicellograptids, whereas it is half again as large as that found in the vagus group of dicellograptids. The proximal-end development of Leptograptus with a left-handed origin of th 1^ and three crossing canals is similar to that found in the dicellograptids, especially the elegans group in which the th 2^ and th 2^ prothecae are not in contact with the sicula and are thus interpreted as being composed of U-shaped fuselli. The first four thecae in the elegans group show a direction of growth that is 301 intermediate between the horizontally directed thecae of Leptograptus and the upward-directed thecae of the vagus group. Except for the absence of introtorsion and the development of the infragenicular wall, the thecae of Leptograptus are similar to those of Dicellograptus in the following features: 1) A geniculum which divides the free ventral wall into infra- and supragenicular walls. 2) An introverted aperture with lateral lappets. 3) Prothecal folds which a) are located at the level of the geni cula b) develop with thecal budding, c) involve the formation of a primary notch followed by the development of a prothecal base over the primary notch, and d) are defined dis tally by a change in the direction of growth of the dorsal wall of the protheca. 4) A thecal ontogeny interpreted on the thickness of fuselli, which shows a slow rate of growth for the proximal end of the pro theca (primarily the prothecal base) followed by a rapid rate of growth for the remaining part of the protheca and the metatheca. As described for the vagus group of dicellograptids and interpreted for the elegans group, the infragenicular wall of a particular theca is initially formed as an apertural flange by the zooid of the preceding theca. This results in an introtorted aperture. As described below for Leptograptus trentonensis, the infragenicular wall of a particular theca is formed by the zooid of that theca, and the aperture is not introtorted. In the development of prothecae in both Leptograptus and Dicellograptus, the fuselli are U-shaped and together with the interthecal septum (dorsal wall of the preceding metatheca) they form a closed tube. As development 302 procédés past the aperture in Leptograptus, the zooid begins to form its own ventral wall, which is the infragenicular wall. In Dicellograptus. this is delayed until the theca attains the distal end of the apertural flange over the preceding theca, in which case the initially formed part of the ventral wall corresponds to the supragenicular wall. It appears that if a wall is already present, a zooid will not duplicate it. In Leptograptus, a ventral wall is needed when the aperture of the pre ceding theca is attained, and it is thus developed, whereas in Dicello graptus, it is not needed until the distal end of the apertural flange is attained. A dorsal wall is not needed and thus not developed in the th 2^ and th 2^ prothecae of the vagus group of dicellograptids because with the direction of growth of the prothecae the sicula can readily serve as the dorsal wall. Thus, the major difference between Dicello graptus and Leptograptus thecae (and the zooids) is the ability to develop an apertural flange in the dicellograptids. If that ability is absent, a thecal form described for Leptograptus would probably result. Although it develops differently in the leptograptids and dicello graptids, the geniculum is always located: 1) transverse to a prothecal fold, 2) at the level of the maximum dorso-ventral width of the theca, and 3) at the level of asexual budding of the succeeding theca. Im mediately distal to the geniculum, there is an abrupt and significant decrease in the dorso-ventral width of the theca, and it is at this level that the succeeding protheca develops as a separate entity. Thus, 303 the presence and location of the geniculum appear to be controlled by thecal budding which is similar in both the leptograptids and dicello graptids. Bulman (1947) discusses the horizontal versus the reclined habit of the stipes in Leptograptus and Dicellograptus, respectively, and tries to relate this to the orientation of the first pair of thecae. For example, in the vagus group of dicellograptids the distal part of th 1^ shows upward growth, and the stipes are very reclined. In Leptograptus, th 1^ is horizontal, and the stipes are horizontal. Bulman. (1947, p. ix) partly discounts the influence of the orientation of the earliest thecae on the habit of the stipes by referring to dicellograptids of the elegans group in which the first thecae are horizontal and the stipes are reclined. Dicellograptus complanatus Lapworth (Skoglund, 1963, text-fig. lOB), D. geniculatus Bulman . (1932, text-fig. 6), and 2» forchammeri (Geinitz), as described by Hopkinson (1871, PI. 1, fig. 1), are examples of this. However, as interpreted herein (See Dicellograptus), these species most likely possess apertural flanges, which are upward- oriented features, and it is at the base of the apertural flanges (in fragenicular wall) on the first pair of thecae that the stipes of those species initiate their inclined habit. Thus, it appears that upward- directed features on the earliest theca, whether they consist of the entire theca (e.g., the U-shaped theca in the vagus group) or. of a part of a theca (e.g., the apertural flange in the elegans group), are closely associated with, and may be the cause of, the upward or reclined habit of the stipes in dicellograptids. 304 The elegans group of dicellograptids shows very close similarities to the leptograptids. Considering the interpretations discussed above, the only significant difference is the ability of the dicellograptid zooid to develop an apertural flange, which in turn might produce reclined stipes and introtorsion of the thecal apertures. The vagus group differs to a greater degree from the leptograptids by the U- shape of the first pair of thecae, which together with the apertural flange, may cause the very reclined habit of the stipes, as well as the introtorsion of the thecal apertures. Except for Leptograptus trentonensis and Whittington's (1955) Leptograptus sp. ind., no other species of Leptograptus is known on the basis of isolated specimens. However, all other species of Leptograptus (for example, those described by Elies and Wood, 1903 and Ruedemann, 1947) closely resemble in appearance the compressed, non-isolated specimens of L. trentonensis. These resemblences, which include the morphology of the proximal end and the width, thecal density, and orientation of the stipes, suggest that the proximal-end development, thecal morphology, and thecal ontogeny are similar in all species of Leptograptus. Therefore, the genus Leptograptus is transferred herein from the Nemagraptidae to the Dicranograptidae because its morphology and development, as deter mined from isolated specimens of 1^. trentonensis and Leptograptus sp. ind., are nearly identical to those of Dicellograptus and are in sharp contrast to those of Nemagraptidae, as described herein. 305 Leptograptus trentonensis Ruedemann, 1908 (Text-figures 47-49) 1908 Leptograptus flaccidus (Hall) mut. trentonensis nov., Ruedemann, p. 261-262, text-figs. 172-175, Pl. 14, figs. 6, 7. 1934 Leptograptus flaccidus (Hall) mut. trentonensis Ruedemann, Ruedemann and Decker, p. 306-307, pl. 40, figs. 7, 8. 1947 Leptograptus flaccidus (Hall) mut. trentonensis Ruedemann, Ruedemann, p. 366-367, pl. 59, figs. 14-17. Î1952 Leptograptus flaccidus mut. trentonensis Ruedemann, Decker, pl. 2, fig. 64. 71960 Leptograptus flaccidus mut. trentonensis Ruedemann, Berry, p. 72, pl. 15, fig. 7. Type Data Ruedemann (1908) did not designate a holotype among his syntypes of Leptograptus flaccidus mut trentonensis. Except for one specimen (USNM 23778) stored at the U. S. National Museum, Washington, D. C., these syntypes are stored at the New York State Museum, Albany, and are numbered NYSM 7275-7278. These specimens have been examined by the author and found to be too poorly preserved to show detailed thecal morphology. However, they do agree with Ruedemann's (1908) description and illustrations. A lectotype has not yet been selected from among the syntypes. Diagnosis A species of Leptograptus with proximally convex curvature and distally concave curvature of stipes. Sicula 1.4 to 1.6 mm long. Stipes increase in width from 0.30 - 0.35 mm at th 2 to 1.05 mm dis tally. Thecae number 10 - 11 in 10 mm, proximally, and 8 - 9 in 10 mm, distally. Theca with straight supragenicular wall, distinct geniculum. 306 EXPLANATION OF TEXT-FIGURE 47 Text-figure 47 a-1. Leptograptus trentonensis Ruedemann. a. Reverse aspect of compressed, isolated specimen. X18.8. PS-109.7. OSU 33105. b,c. Non-isolated, compressed specimen, b. X18.8, c. X4.5. PS-117. OSU 33106. d. Reverse aspect of non-isolated, compressed specimen. X9. PS-117. OSU 33107. e. Lateral aspect of isolated, compressed specimen representing distal stipe fragment. X18.8. PS-109.2. OSU 33108. f. Lateral aspect of isolated, compressed specimen representing distal stipe fragment. X18.8. PS-109.2. OSU 33109. g. Lateral aspect of isolated, compressed specimen representing distal stipe fragment. X18.8. PS-109.2. OSU 33110. h,i. Pyritized internal mold of proximal end. Note left-hand origin of th 1^ from th 1^. X37.5. PS-119.5. OSU 33111. 307 thi? th 2 Text-figure 47 308 EXPLANATION OF TEXT-FIGURE 48 Text-figure 48 a-i. Leptograptus trentonensis Ruedemann. Isolated specimens showing rhabdosomal development. a,b. Reverse and obverse aspects of specimen showing budding of th 1^ from th 1^. X37.5. PS-109.2. OSU 33112. c. Reverse aspect of sicula with th 1^ and initial downward-directed part of th 1^. X37.5. PS-109.2. OSU 33113. 1 2 d. Reverse aspect of sicula with th 1 and th 1 . X37.5. PS-109.2. OSU 33114. e. Reverse aspect of sicula with th 1^ and th 1^. X37.5. PS-109.2. OSU 33115. f. Reverse aspect of sicula with th 1 1 , th 12 , and th 2 1 . Note left-hand origin of th 2^ from th 1^. X37.5. PS-129. OSU 33116. g. Reverse aspect of specimen showingth 2^ budding right-handedly from th 2^. X37.5.PS-129. OSU 33117. 10 1 h. Reverse aspect of sicula with th 1 , th 1 , and th 2 . X37.5. PS-109.2. OSU 33118. i. Reverse aspect of specimen with primary notch for th 3^. X37.5.PS-109.2. OSU 33119. thi th 2 th I 4h 1 ,th I thi th2 th2‘ LO S Text-figure 48 310 EXPLANATION OF TEXT-FIGURE 49 Text-figure 40 a-k. Leptograptus trentonensis Ruedemann. Isolated specimens showing rhabdosomal development. a. Lateral aspect of stipe fragment. Note that growth lines on interthecal septum are continuous with those on lateral walls. X37.5. PS-109.2. OSU 33120. b. Lateral aspect of compressed stipe fragment showing growth lines on infragenicular wall and prothecal base. X37.5. PS-109.2. OSU 33121. c. Obverse aspect of specimen showing growth lines on infragenicular and lateral walls of th 2^. X37.5. PS-109.2. OSU 33122. d. Lateral aspect of stipe fragment. Note upturned proximal end of interthecal septum. X37.5. PS-109.2. OSU 33123. e,f. Right- and left-lateral aspects showing growth lines on interthecal septum, infragenicular wall’, and lateral walls. X37.5. PS-109.2. OSU 33124. g. Obverse aspect of specimen showing primary notch of th 3 2 . X37.5. PS-109.2. OSU 33125. h. Lateral- aspect of stipe fragment. Note growth lines on interthecal septum. X37.5. PS-109.2. OSU 33126. i. Reverse aspect of specimen showing primary notch of th 22. X37.5. PS-109.2. OSU 33127. j. Reverse aspect of specimen showing growth of th 3^. X37.5. PS-109.2. OSU 33128. k. Reverse aspect of specimen showing growth of th 3^. Note absence of th 2^ genicular flange. X37.5. PS-109.7. OSU 33129. pn IS w 312 and slightly introverted aperture. Prothecal folds on dorsal stipe margin. Material Because of its stratigraphie range, all of the available specimens of Leptograptus trentonensis were collected from the upper part of the Pratt's Syncline section, except for one specimen from the hig&est locality in the Calera section. One hundred twenty-three specimens were isolated, and of these 72 are proximal-end fragments and early growth stages, 49 are distal stipe fragments, and 2 are pyritized, internal molds of proximal ends. Only a few specimens show relief; the majority are fully compressed. Most early growth stages show fusellar structure. One hundred and twenty specimens were studied on shale surfaces. These specimens, generally large rhabdosomes, are fully compressed but otherwise well-preserved. Description Isolated Specimens (Text-figs. 47a, e-g, 49k). the largest avail able rhabdosome with a proximal end has a 3 mm long stipe with 4 thecae. The sicula is 1.4 to 1.6 mm long and is oriented upright between the stipes. The first theca originates 0.5 to 0.6 mm above the sicular aperture. The first two thecae diverge horizontally from the proximal end at the level of the sicular aperture. Beginning with the second pair of thecae, the stipes show convex curvature with respect to the ventral stipe wall. The stipes distally widen, being 0.30 - 0.40 mm at th 2 and 0.77 mm in the widest, isolated, distal fragment. 313 The thecae number 3 in 3 mm proximally; in the largest available distal fragment (Text-f ig. 47g), there are two theca in 2 mm. The thecae, which overlap for half of their length, are 1.5 mm long proxi mally, and 2.1 mm long in the largest, available distal fragment. Each theca shows a geniculum, a slightly introverted aperture, and a pro thecal fold. The geniculum is distinct in specimens that are preserved in full relief and in specimens that are fully compressed, if the compression is normal to the lateral stipe walls. However, if compression is oblique to the lateral stipe walls, the geniculum becomes indistinct and/or assumes a very obtuse genicular angle because the infragenicular wall is flattened into the same plane as one of the later walls of the stipe. In distal thecae, the geniculum is difficult to discern because the genicular angle is very obtuse and compression easily obscures it (Text-fig. 47e-g). The supragenicular wall is straight for most of its length and inclined at 10 to 15 degrees angle to the dorsal stipe wall. The aperture, which shows no introtorsion, is slightly introverted so that it opens in a dorso-distal direction. As a result of the introversion, the supragenicular wall shows slight convex curvature in its very distal portion. The point of the maximum dorso-ventral width is at a level approximately transverse to the base of the apertural excavation, and a mesial spine projects from this point in at least the first two to four thecae on each stipe. In fully compressed specimens, the introversion is reduced so that the aperture opens in a distal direction. In distal 314 thecae (Text-fig. 47g), introversion is minimal; the supragenicular wall is relatively straight; and the aperture opens in a distal direction. The margin of the aperture shows paired, lateral lappets in the proximal thecae (Text-fig. 49k). In distal thecae and secondarily thickened, proximal thecae, the ventral margin of the aperture is generally thickened in such a way that a single lappet extends from the middle of one lateral margin around the ventral margin to the middle of the opposite lateral margin. In proximal thecae, especially those preserved in full relief, the apertural excavations have the shape of an inverted comma, which occupies a third of the stipe width and a third to a fourth of the length of the supragenicular wall. In distal thecae the large genicular angles result in nearly straight free ventral walls. As a result, the apertural excavations appear in lateral view as "saw teeth," which occupy a fourth of the stipe width and a fifth of the length of the supragenicular wall. The lateral margins of the apertural excavations are strengthened by a list. In fully com pressed thecae and in distal thecae, the apertural excavation can be distinguished by the presence of the list. Prothecal folds are commonly observed on proximal stipe portions. They are located on the dorsal stipe margin at levels transverse to the genicula on the ventral margin of the stipe. Grooves, representing traces of interthecal septa, extend along the lateral walls of the stipe from the bases of the apertural excavations to the prothecal folds. Non-isolated Specimens (Text-fig. 47b-d)- The largest available specimen shows a 16.5 mm long stipe. The sicula, which bears a nema. 315 is oriented upright between the stipes. The first two thecae are strai^t and diverge horizontally away from the sicula. Initially, the stipes show gentle, convex curvature and diverge at an angle of 140 to 170 degrees. Distally, by means of concave curvature, they become horizontal in relation to the vertical sicula. The stipes increase in width distally. They are 0.30 - 0.35 mm wide at th 2, 0.40 - 0.45 mm wide at th 5, and 0.50 - 0.60 mm wide at th 10. The width of the most distal, available stipe fragment is 1.05 mm. Proximally, the thecae number 3.5 in 3 mm, 5 - 5.5 in 5 mm, and 10 - 11 in 10 mm. Distally, there are 2.5 thecae in 3 mm, 4 - 4.5 in 5 mm, and 8 - 9 in 10 mm. The thecae appear similar to those of compressed, isolated specimens. The genicula are occasionally distinct in proximal thecae. In distal thecae, the genicula are indistinct; the free ventral walls are straight or show slight sigmoid curvature. Grooves, representing traces of the interthecal septa, are parallel to the supragenicular walls and are inclined at angles of 10 to 15 degrees to the dorsal stipe wall. These grooves terminate proximally at the level of the prothecal folds on the dorsal wall of the stipe. - Development of Proximal End. Pyritized, internal molds. Text-figures 47h-i show a pyritized, internal mold of the proximal end. Th 1^ buds 0.75 mm above the sicular aperture, then continues down the side of the sicula to the sicular aperture, where it bends and continues directly outward 2 for 0.6 mm. Th 1 originates on the obverse (left-handed) side of th 1^. It grows obliquely upward across the 316 initial part of th 1^ to the reverse side of the sicula, where it bends and grows obliquely down to the anti-virgellar margin of the aperture of the sicula. Then, it extends directly outward from the sicula for 0.75 mm. The origin of th 2^ is left-handed. It develops from the virgel- lar side of th 1^ where th 1^ begins its downward direction of growth. Th 2^ grows down th 1^, then bends and grows outward along the dorsal wall of th 1^. The origin of th 2^ from th 2^ is right-handed. It develops from the anti-virgellar side of th 2^ where th 2^ changes from a downward to an outward direction of growth. Th 2 extends hori zontally across the downward-directed part of th 1^, and th 3^ develops from th 2^. The proximal-end development of L. trentonensis, in which th 1^ has a left-handed origin and there are three crossing canals, is the streptoblastic diplograptid type as described by Bulman (197Ù) similar to that found in Dicellograptus. More detailed information, which includes the timing of budding and the origin of structures that are not preserved in pyritized, internal molds, can be obtained through the study of isolated, early growth stages. Early growth stages (Text-figs. 48, 49i-j). The sicula has an over all length of 1.4 - 1.6 mm. The prosicula, which accounts for a fifth of the length of the sicula, is conical with an apertural width of 0.10 - 0.15 mm. Longitudinal threads can be observed in a few specimens ex tending from the aperture to the base of the nema (Text-fig. 48a,b). The metasicula is parallel-sided and relatively long and narrow. Its aperture, which is 0.20 - 0.25 mm wide, exhibits broad lateral lappets, which consist of two to three fuselli. 317 In the earliest available growth stage (Text-fig. 48a,b), th 1^ and th 1 2 are partly developed. Th 1 1 originates from the virgellar side of the metasicula 0.5 - 0.6 mm above the sicular aperture. The initial part of th 1^ consists of a tube that is 0.15 mm long. The arrangement of the walls about the initial part of th 1^ and the orien- 2 2 tation of the opening for th 1 suggest that the budding of th 1 occurs in a manner similar to that described for Dicellograptus. 2 1 After the budding of th 1 , th 1 grows down the virgella to the level of the sicular aperture where it bends and grows directly outward from the sicula (Text-figs. 48c-e) . Simultaneously, th 1^ grows hori zontally across the initial part of th 1^ to the reverse side of the sicula, where it begins growing downward as a broad flap (Text-fig. 48e). As shown in Text figure 48h, the broad flap of th 1 divides into two separate tubes, th 1^ and 2^. The development of th 1^ proceeds in advance of that of th 2^. Th 1^ grows obliquely downward to the anti-virgellar margin of the sicula, where it bends and grows directly outward from the sicula (Text. figs. 4Sf-h). Before th 1^ is complete to the mesial spine, th 2^ grows down to the level of the dorsal wall of the horizontal part of th 1^. Th 2^ then buds from the anti-virgellar side of th 2^(text-figs. 49g, i) , and 1 19 th 2 grows out along the dorsal wall of th 1 , while th 2 grows out 2 1 along the dorsal wall of th 1 (Text-fig. 49i) . However, th 2 develops 2 in advance of th 2 . Most of its metathecal portion and the initial bud of th 3^ are complete before th 2^ grows out on the dorsal wall of 318 th 1“ (Text-fig. 481). Th 3^ buds from th 2^ and develops along the dorsal wall of the th 2^ metatheca (Text-fig. 49k). Development of Prothecal Folds ■ Prothecal folds are conspicuous in isolated proximal ends. They can rarely be seen in distal fragments, which usually show a thick development of cortical tissue and are compressed. However, they have been observed in well-preserved, non isolated specimens (Text-figs. 47b-d). The prothecal folds are gently curved in lateral view and always occur at a level transverse to the genicula. As represented by a groove on the lateral stipe wall, the interthecal septum originates at a prothecal fold. Several of the available, isolated specimens show the manner in which the prothecal folds develop. The prothecal segment of th n develops by the addition of U-shaped fuselli (Text-fig. 49e-f). The bases of the U's form the dorsal wall of the protheca, and the U's open onto the dorsal wall of the th n-1 metatheca. As the n develops past the dorsal margin of the th n-1 aperture, it increases in size in a dorso-ventral direction; fuselli are added, which form a complete circle; and th n develops its own ventral wall, which is its infragenicular wall. With the addition of more fuselli, th n increases in size in a dorso-ventral direction, until the distal end of the infragenicular wall is attained. Then, with growth of the n past the distal end of the infragenicular wall, U-shaped fuselli are added with the U opening in a dorsal direction (Text-fig. 49a). This results in the formation 319 of an opening in the dorsal wall for th n+1, and the size of th n metatheca abruptly decreases in a dorso-ventral direction (Text-figs- 49b, e-f) . After the addition of two to four U-shaped fuselli, devel opment of the th n metatheca continues with the addition of fuselli that form a complete circle. Thus, th n again becomes a complete tube; it develops its own dorsal wall, which later becomes the interthecal septum between th n and th n+1; and an opening is left for the budding of th n+1 (Text-figs. 49a, e-f, h) . The opening for the initial bud of th n+1 is a primary notch because of the manner in which it forms. Its proximal margin parallels the growth lines on the dorsal wall of the th n protheca. Its distal margin parallels the growth lines on the interthecal septum (Text- figs. 49a, h) . Its lateral margins intersect the fuselli on the lateral walls of th n (Text-figs. 49a, e-h). All of the margins of the primary notch appear to be strengthened by a list. Commonly, the proximal end of the interthecal septum (dorsal wall of th n metatheca) is upturned (Text-fig. 49d). The th n+1 protheca develops from the th n+1 primary notch by the addition of U-shaped fuselli (Text-fig. 49b). The bases of the U ’s are continuous with the dorsal wall of the th n protheca. The limbs of the U ’s are continuous with the lateralmargins of the primary notch. With the continual addition of U-shaped fuselli, the th n+1 protheca grows over the primary notch and along the dorsal wall of the th n metatheca. 320 The proximal end of the dorsal wall of the th n+1 protheca is continuous with the distal end of the dorsal wall of the th n protheca, and both walls are oriented in a dorso-distal direction in relation to the longitudinal axis of the stipe. At a short, but significant, distance distal to its proximal end, the dorsal wall of the th n+1 protheca bends so that it is parallel to the longitudinal axis of the stipe. Thus, a prothecal fold is formed, and it consists of the distal end of the dorsal wall of the th n protheca and the proximal end of the dorsal wall of the th n+1 protheca. It appears to be caused by the asexual budding of th n+1 from th n, which results in a dorso-ventral expansion of the distal end of the th n protheca, and a change in the direction of growth of the th n+1 zooid from initially upward out of the primary notch to outward along the dorsal wall of th n metatheca. The part of the th n+1 protheca that is proximal to the change in the direction of growth is equivalent to the prothecal base in Dicello graptus. The development of prothecal folds here described is nearly identical to that found in Dicellograptus. The similarities involve: 1) the formation of the primary notch, 2) the growth of a prothecal base over the primary notch, and 3) a change in the direction of growth of the dorsal wall, which defines the distal limit of the prothecal fold. Development of Thecae and Associated Structures. As shown above, the th n+1 protheca develops from a primary notch, and its initial development is simultaneous with the initial development of the th n metatheca. However, there is a delay in the development of the th n+1 protheca, while the th n metatheca develops far in advance 321 (Text-fig. 481, 49a, g-i) . This can be explained by growth line evidence. Growth lines in the most proximal portion of the protheca are relatively thin whereas growth lines in medial and distal parts of the protheca are relatively thick. Although the boundary between these two sets of growth lines can not be discerned in any of the available specimens, it appears to be at the distal end of the prothecal base (compete Text-figs. 49i, j). The growth lines of the distal part of the protheca show the same thickness and orientation as those of the metatheca. The development of the list on the margins of the primary notch may represent a pause or discontinuity in the growth of th n+l. This is followed by thin growth lines, which suggest a relatively slow rate of growth, while the th n metatheca is growing relatively rapidly as suggested by its thick growth lines. There may be a second pause or discontinuity in the growth of th n+1 as represented by the change in the thickness of the growth lines that occurs near the distal end of the prothecal base. By this time th n develops to its aperture (Text- fig. 49j). The thin growth lines and suggested discontinuities appear to explain the slow rate of development of the th n+1 protheca relative to the th n metatheca. With the addition of thick fuselli, the rate of growth of the th n+1 protheca significantly increases, and it continues through the remaining development of the theca. The development of the th n metatheca in advance of the th n+1 protheca strongly suggests that the interthecal septum between th n 322 and th n+1 initially forms as the dorsal wall of the th n metatheca. Growth lines on the interthecal septum support this interpretation because they are continuous with those on the lateral walls of the th n metatheca (Text-figs. 49a, h, j). The interthecal septum between th n and th n+1 is continuous with the dorsal margin of the th n aperture, and it serves as the ventral wall of the th n+1 protheca. Growth lines on the th n infragenicular wall are continuous with those on the lateral walls of the th n protheca (Text-figs. 49b-c, e-f). This, together with the absence of any apertural flange, strongly suggests that the th n infragenicular wall is formed by the th n zooid. Early growth stages also support this interpretation. Text-figure 49k shows that th 2^ is complete to the aperture with no trace of an apertural flange, and the th 3^ protheca is developed near the th 2^ aperture. Text-figure 49c shows th 2^ developing past the th 1^ aperture. The th 2^ infragenicular wall is composed of the same fuselli that form the lateral and dorsal walls of th 2^. As a result of this manner of development, the dorsal margin of the th n aperture is in contact with the base of the th n+1 infragenicular wall. The th n aperture will thus show no introtorsion. Remarks The specimens described here agree closely with Leptograptus flaccidus trentonensis as described by Ruedemann (1908, 1947) and Ruedemann and Decker (1934). Decker (1952) illustrates only a stipe fragment; thus, his identification is questionable. Berry's (1960) figured 323 specimen (YPM 20345) of L. flaccidus trentonensis is poorly preserved. Thecal morphology can not be discerned, yet the rhabdosome has the same general appearance as that described above. The type specimens of L. flaccidus and L. flaccidus trentonensis have been examined and found to be poorly preserved; thecal morphology can rarely be discerned. The subspecies trentonensis is here raised to specific status because its detailed morphology is now known on the basis of excellently preserved specimens and it occurs five zones below the typical I^. flaccidus. Figured Specimens OSU 33105 - OSU 33129 324 Cenus Dicranograptiis Hall, 1865 Type Species: Graptolithus ramosus Hall, 1848 Diagnosis Rhabdosome with biserial proximal end which divides distally to two reclined uniserial stipes. Remarks The close affinities of Dicranograptus and Dicellograptus of the vagus type have been known since at- least the time of Hopkins on (1870). In his review of the genus, Hopkinson (1870) includes the species sextans in Dicranograptus and comments on its great similarity to JD. formosus, which shows an extremely short, biserial proximal end. Re flecting on the close similarities of Hopkinson's two species and reporting intermediate forms, Elies and Wood (1904) assigned all of the variant forms to Dicellograptus sextans. In the following species description, it is argued that Dicranograptus irregularis is closely related to dicellograptids of the vagus group. These dicellograptids approach Dicranograptus in the shape of the proximal end. The first four thecae sho%f a dominantly upward direction of growth; the dorsal walls of the th 2 1 and th 2 2 prothecae are in contact with the sicula; and the initial bud of th 3^ is commonly in contact with the sicula. The only significant difference between D. irregularis and Dicellograptus gurleyi is that the dorsal walls of the th 3^ and th 3^ prothecae are in contact with the sicula. The specimen illustrated in Text-figure 50c appears to represent an intermediate form (not neces sarily transitional) between the two species, and it is very similar to JD. gurleyi ssp. A. 325 Dicranograptus ni choisoni which is described by Bulman (1944), is the only dicranograptid in which the proximal-end development has been known. It is identical to that found in the dicellograptids described above. The similar development of the infragenicular wall in D^. nicholsoni and in the dicellograptids described above strongly suggests comparable ontogeny of the zooids of the two genera. The position of the dicalycal theca in Dicranograptus remains a problem. Although he did not have a sufficient number of early growth stages to show its position, Bulman (1944) believed that th 2^ is the dicalycal theca in D^. nicholsoni and that a complete medium septum in the biserial portion developed with the third pair of thecae. It is suggested below that th 3^ is the dicalycal theca in D. irregularis because there is no evidence of a medium septum. This delay in the position of the dicalycal theca can easily explain the change from biserial to uniserial stipes in irregularis. However, in species with a medium septum, the change from the biserial to the uniserial portions is still unexplained. Dicranograptus irregularis Hadding, 1913 (Text-figure 50) 1913 Dicranograptus irregularis, n.sp., Hadding, p. 52, PI. 4, figs. 1-12. ?1960 Dicranograptus brevicaulis Elies and Wood, Berry, p. 77, PI. 15, fig- 2 . 1964 Dicranograptus irregularis Hadding, Berry, p. 122-123, PI. 11, figs. 1-4. ?l970b Dicranograptus sp., Toghill, p. 126-127, fig. 3a,b. 326 EXPLANATION OF TEXT-FIGURE 50 Text-figure 50 a-d. Dicranograptus irregularis Hadding. Compressed, non isolated specimens. a. Obverse aspect. Note thecal morphology. X9. C-30. OSU 33130. b. Reverse aspect. Note prothecal folds on dorsal stipe margins. X9. PF-22.5. OSU 33131. c. Specimen with apex of prosicula visible in axil. X18.8. C-14.8. OSU 33132. d. Reverse aspect of specimen with long nema. X18.8. C-32. OSU 33133. 50 e. Dicranograptus brevicaulis Elies and Wood. Specimen figured by Berry (1960, PI. 15, fig. 2). X18.8. YPM 20341. 327 Text-figure 50 328 Type Data Hadding's (1913) syntypes of Dicranograptus irregularis are stored at the Museum of the Palaeontological Institute, Lund University, Sweden. These specimens are well preserved. They are pyritized and retain relief. Hadding (1913) did not designate a holotype among his syntypes. The specimen illustrated by Hadding (1913) in Plate 4, figure 12 is here selected as the lectotype because it is well preserved and well illus trated. Diagnosis A species of Dicranograptus with 1.0 - 1.25 mm long biserial proximal end. Uniserial stipes diverge at 20 to 60 degrees angle. Uniform stipe width 0.45 - 0.55 mm. Proximally, 4.5-5 thecae in 3 mm and 8 in 5 mm. Theca with distinct geniculum, convex supragenicular wall, and introtorted and introverted aperture. Prothecal folds on dorsal stipe margin. Material The available material consists of 39 specimens, which are pre served as carbon films. One specimen was obtained from the Pratt’s Ferry section; the rest were collected from the Calera section. The largest available specimen has a 5.8 mm long stipe consisting of 9.5 thecae. Description (Text-figs. 50a-d) The biserial proximal end, which consists of at least two thecae on each stipe, is 1.0 - 1.25 mm long and 1.0 - 1.1 mm wide. The shape of this proximal end is remarkably similar to that found in 329 dicellograptids of the vagus group. The sicula is 1-2 mm long and shows a nema, which may be as long as 1.8 mm, as well as a virgella. The first pair of thecae are distinctly U-shaped and bear mesial spines. The thecae are alternating, and there is no evidence of a medium septum in the biserial portion of the rhabdosome. This suggests that, except for a delay in the position of the dicalycal theca, the proximal-end development is identical to that found in dicellograptids of the vagus group. In the case of most of the available specimens (Text- figs. 50a-b, d) , th 3^ would be the dicalycal theca. Th 2^ could conceivably be the dicalycal theca in two specimens (Text-fig. 50c), in which the biserial portion is slightly shorter and the tip of the prosicula is free within the axil. The uniserial stipes, which enclose an angle of 20 to 60 degrees, show a uniform stipe width that ranges from 0.45 - 0.55 mm among the available specimens. As measured distally from the th 2 geniculum, there are 4.5-5 thecae in 3 mm and 8 thecae in 5 mm. The thecae are 1.2 - 1.4 mm long and overlap for half their length. They show distinct genicula, convex supragenicular walls, introverted and introtorted apertures, and mesial spines. Compression has produced a large variation in the genicular angles, however, they are generally 90 degrees. The apertural excavations, which have the shape of an inverted comma, occupy a third of the stipe width and a fourth of the length of the supragenicular wall. Prominent prothecal folds are present on the dorsal stipe margins at levels transverse to the 330 apertural excavations on the ventral stipe margins. Grooves, repre senting traces of the interthecal septa, can rarely be seen on the lateral stipe walls extending from the bases of the apertural excava tions on the prothecal folds. Remarks The specimens described above agree closely with Dicranograptus irregularis as described and illustrated by Hadding (1913) and Berry (1964). The gross outline of Toghill's (1970b) specimen of ?Dicranograptus sp., which is too poorly preserved to show thecal morphology, is similar to that of the Alabama specimens. Dicranograptus brevicaulis Elies and Wood and 2» furcatus (Hall) are other species with a relatively short biserial portion- D. brevi caulis , as described by Elies and Wood (1904), differs from the speci mens described above by the greater number of thecae in the biserial portion. Berry's (1960) illustrated specimen (YPM 20341) of 2- brevi- caulis, which is shown in Text-figure 50e, bears a strong resemblance to the Alabama specimens. The stipes initially diverge at the third pair of thecae, and the thecal morphology is remarkably similar to that found in the specimens described above. Berry's specimen is obliquely compressed, which may account for the stipe width (0.63 mm) that is slightly wider than in the Alabama specimens. 2- furcatus shows a greater number of thecae in the biserial portion and a greater degree of curvature of the supragenicular wall. Dicranograptus irregularis, as described above, shows strong similarities to the vagus group of dicellograptids. In fact, Hadding 331 (1913, p. 52) comments on the similarities of the thecae of Dicrano graptus irregularis and Dicellograptus vagus, which is difficult to distinguish from D. gurleyi. Among the Alabama specimens, the similar ities in the thecae of I), gurleyi and Dicranograptus irregularis include: 1 ) the shape of the geniculum, 2 ) the curvature of the supragenicular wall, 3) the shape, introversion, and introtortion of the aperture, 4) the length and overlap of the thecae, 5) the size of the apertural ex cavations, and 6 ) the position of the prothecal folds. The stipes of the two species show comparable thecal densities and widths, but, as described above, it is the proximal ends that show the greatest similarity. Dicellograptus gurleyi ssp. A can easily be confused with Dicrano graptus irregularis. The length of the biserial portion is the diag nostic feature. Dicranograptus irregularis is the earliest known member of the genus. It first appears in Alabama, Wales, and Southern Sweden in strata that is equivalent to the Swedish graptolite Zone of Glyptograptus teretiuscuius. This is within the same stratigraphie interval, in which several dicellograptids of the vagus group first appear. The evolution of B. irregularis from this group of dicellograptids by delay of the dicalycal theca from th 2^ to th 3^ seems highly possible. Figured Specimens OSU 33130-OSÜ 33133 GRAPTOLITES OF lliE MIDDLE ORDOVICIAN ATHENS SHAI.E, ALABAÎLA VOLUME II Presented in Partial Fulfillment of the Requirements fo: the Degree Doctor of Philosophy in the Graduate School of The Ohio State University by Stanley C. Finney, B.S., M.S. ***** The Ohio State University 1977 Reading Committee: Approved by Professor S. M. BergstrBn, Chairman Professor W. C. Sweet Professor J. W. Collinson f i o . J l . 7 Adviser j Department of (teology and Mineralogy TABLE OF CONTENTS VOLUME II Page LIST OF TABLES ...... xv LIST OF TEXT-FIGURES ...... xvi Suborder Glossograptina ...... 332 Family Glossograptidae ...... 354 Genus Glossograptus Emmons, 1855 ...... 358 Genus Apoglossograptus n. gen...... 391 Family Cryptograptidae Hadding,1915 ...... 404 Genus Cryptograptus Lapworth,1880 ...... 404 Family Corynoididae Bulman, 1944 ...... 437 Suborder Diplograptina Lapworth, 1880, emend. Bulman, 1970...... 442 Family Diplograptidae Lapworth, 1873 ...... 442 Genus Climacograptus Hall, 1865 ...... 444 Genus Pseudoclimacograptus Pribyl, 1947 ...... 469 Genus Glyptograptus Lapworth, 1873 ...... 503 Genus Orthograptus Lapworth, 1873 ...... 555 Family Lasiograptidae Lapworth, 1879 ...... 557 Genus Lasiograptus Lapworth, 1873 ...... 557 Family Dicaulograptidae Bulman, 1970 ...... 562 Genus Dicaulograptus Rickards and Bulman, 1965 ...... 562 REFERENCES ..... 569 APPENDICES Appendix A. Distribution of graptolites in the sections investigated ...... 579 xiv LIST OF TABLES Table Page 8 . Distance in mm between "dorsal" spines and proximal end in specimens of G. ciliatus ...... 383 9. Numerical analysis of 2- and 3-spined forms of 2' modestus ...... 496 10. Distribution of graptolites in the Pratt’s Ferry Section ...... 581 11. Distribution of graptolites in the Pratt's Syncline Section ...... 582 12. Distribution of graptolites in the Calera Section ...... 583 XV LIST OF TEXT-FIGURES Text-figure Page 51. Proposed phylogenetic scheme for "isograptids," Glossograptidae, Cryptograptidae, and Corynoididae ...... 334 52. Stratigraphie ranges of genera and families in the proposed phylogenetic scheme shoîcn in Text-figure 5 1 ...... 346 53. Isolated specimens representing early growth stages of Glossograptus ciliatus Emmons ...... 360 54. Isolated specimens representing early growth stages of Glossograptus ciliatus Emmons ...... 362 55. Isolated specimens representing early growth stages of Glossograptus ciliatus Emmons ...... 364 56. Isolated and non-isolated specimens of Glossograptus ciliatus Emmons ...... 366 57. Non-isolated specimens of Glossograptus ciliatus Emmons ...... 368 58. Non-isolated specimens of Apoglossograptus lyra (Ruedemann)...... 395 59. Non-isolated specimens representing early growth stages of Apoglossograptus lyra (Ruedemann) ..... 397 60. Isolated specimens of Cryptograptus marcidus (Hall) .... 409 61. Isolated specimens of Cryptograptus marcidus (Hall) .... 4 H 62. Isolated specimens representing early growth stages of Cryptograptus marcidus (Hall) ...... 413 63. Non-isolated specimens of Cryptograptus marcidus (Hall) ...... 415 64. Isolated and non-isolated specimens of Climacograptus meridionalis Ruedemann ...... 446 XVI Text-figure Page 65. Isolated specimens representing early growth stages of Climacograptus meridionalis Ruedemann ...... 448 6 6 . Isolated specimens representing early growth stages of Climacograptus meridionalis Ruedemann ...... 450 67. Isolated specimens representing early growth stages of Climacograptus meridionalis Ruedemann ...... 452 6 8 . Non-isolated specimens of Pseudoclimacograptus sp. cf. 2- eurystoma Jaanusson, Orthograptus sp., and Pseudoclimacograptus angulatus angulatus Bulman ... 471 69. Isolated and non-isolated specimens of Pseudoclimacograptus modestus (Ruedemann) ...... 477 70. Isolated specimens representing early growth stages of Pseudoclimacograptus modestus (Ruedemann) ..... 479 71. Isolated specimens representing early growth stages of Pseudoclimacograptus modestus (Ruedemann) ..... 481 72. Isolated and non-isolated specimens of Glyptograptus euglyphus (Lapworth) ...... 505 73. Isolated specimens representing early growth stages of Glyptograptus euglyphus (Lapworth) ...... 508 74. Isolated specimens representing early growth stages of Glyptograptus euglyphus (Lapworth) ...... 510 75. Isolated specimens showing development of rhabdosome in Glyptograptus euglyphus (Lapworth) ...... 512 76. Isolated and non-isolated specimens of Glyptograptus sp. cf. 2* teretiusculus (Hisinger) .. ;...... 533 77. Isolated specimens representing early growth stages of Glyptograptus sp. cf. G. teretiusculus (Hisinger) ...... 535 xvii Text-figure Page 78. Isolated specimens representing early growth stages of Glyptograptus sp. cf. JG. teretiusculus (Hisinger) ...... 537 79. Isolated specimens showing development of rhabdosome in Glyptograptus sp. cf. G^. teretiusculus (Hisinger) ...... 539 80. Isolated and non-isolated specimens of Dicaulograptus? n. sp. A and Lasiograptus sp...... 559 xviix 332 Suborder GLOSSOGRAPTINA Jaanusson, 1960 Diagnosis Biserial, monopleural, axonophorous graptoloids with pericalycal proximal end. Proximal-end development of isograptid type with slight modifications in some species. Thecae basically orthograptid with characteristic ventral apertural process. Exceptions to parts of diagnosis : Corynoididae with arrested development and Apoglossograptus wi.th reclined, uniserial stipes in distal part of rhabdosome. Discussion Except for some unknown dichograptid, an ancestor for the suborder Glossograptina has been difficult to postulate because a primitive dichograptid-type of proximal-end development had been reported for Glossograptus by Whittington and Rickards (1969) and for Cryptograptus by Bulman (1938, 1944). New information, concerning not only the proximal-end development but also the thecal morphology and ontogeny, from the three species that are described below, namely Glossograptus ciliatus Emmons, Apoglossograptus lyra (Ruedemann), and Cryptograptus marcidus (Hall), and re-interpretations of previously described species,indicate that (See text-fig. 51): 1) Glossograptus arose directly from Apiograptus Cooper and McLaurin, which in turn evolved from Isograptus Moberg via Maeandrograptus Moberg and Pseudisograptus Beavis. 2) The ancestry of the other genera of Glossograptidae can be found in Glossograptus. 333 EXPLANATION OF TEXT-FIGURE 51 Proposed phylogenetic scheme for "isograptids," Glossograptidae, Cryptograptidae, and Corynoididae. British graptolite zones used for biostratigraphic subdivisions. A. Isograptus, represented by 1. gibberulus. Illustration after Bulman (1932, PI. 8 , fig. 4); B. Maeandrograp tus, represented by M. geniculatus. Illustration idealized after Skevington (1965, fig. 54b); C. Pseudisograptus, represented by 2» manubriatus. Illustration after Bulman (1968, fig. 1-2); D. Apiograptus, represented by A. crudus. Illustration idealized after Cooper and McLaurin (1974, text-fig. 2g) , proximal- end development interpreted herein; E. Glossograptus, represented by 2 - holmi (lower) and ciliatus (upper). 2 * holmi simplified after Whittington and Rickards (1969 , text-fig. 6a), proximal-end development interpreted herein; F. Paraglossograptus, represented by 2* proteus. Illustration simplified after Whittington and Rickards (1969, text-fig. 6b), proximal-end development as interpreted herein; G. Cryptograptus, represented by 2* marcidus (lower) and 2 - tricomis (upper). 2 - tricornis illustration after Bulman (1970, fig. 90-5a), proximal-end development as interpreted herein; H. Lonchograptus, represented by L. ovatus. Illustration after Bulman (1970, fig. 90-3b) ; I. Apoglossograptus, represented by A. lyra; J. Nanograptus, represented by N. lapworthi. Illustration after Hadding (1915a, PI. 6 , fig. 8b); K. Corynoides, represented by 2- calicularis. Illustration after Bulman (1970, fig. 86-1); L. Corynites, represented by 2* divnoviensis. Illustration after Bulman (1970, fig. 86-2). clingani , 1 m ultidens y gracilis teretiusculus m urchisoni bifidus hirundo gibberulus Text-figure 51 335 3) The Cryptograptidae evolved from an early glossograptid or independently from Apiograptus. 4) The Corynoididae evolved directly from Glossograptus. Before discussing these phylogenetic relations, a summary of the morphologic, astogenetic, and ontogenetic characters of the taxa shown in text-figure 51 is appropriate. 1. Isograptus Moberg. The development and morphology of this genus are best known for the species gibberulus (Nicholson), which is described by Bulman (1932) on the basis of well-preserved specimens. The proximal end has an isograptid type of development, a platycalycal arrangement and an entirely dotmward direction of growth of th 1 ^ and O th 1 in such a way that the midline of the rhabdosome passes between the sicular and th 1 ^ apertures. All the thecae within a rhabdosome have straight metathecal segments that are inclined at high angles to the dorsal stipe margin and exhibit prominent ventral apertural processes. The most distal thecae in the specimens illustrated by Bulman (1932, PI. 8 , figs. 3, 4) are only partly developed and suggest a thecal ontogeny that includes: a) the growth of the metatheca as an incomplete tube with the dorsal wall of the preceding metatheca serving as its ventral wall; b) the development by the metatheca of its own ventral wall when its lateral walls are at, or are close to, the dorsal margin of the preceding metathecal aperture; and c) the development of the ventral wall in advance of the lateral walls and as a dis tally tapering process, which becomes the ventral apertural process. The proximal-end morphology and development and the thecal morphology of other isograptids, excluding Pseudisograptus Beavis, 336 resemble closely those of gibberulus. 2. Maeand rograptus Moberg. The genus Maeandrograp tus is regarded herein as consisting of the species M. schmalenseei Moberg, M. mobergi Tarnquist, M. geniculatus Skevington, M. tau Harris, and M. aggestus Harris. leptograptoides Monsen, a species that Jaanusson (1964) and Skevington (1965) assigned to Maeandrograp tus primarily on the basis of prothecal folding, is excluded herein from Maeandrograptus because Bulman (1969) was able to report significant differences between the manners of prothecal folding in M. geniculatus and leptograptoides. Skevington (1965) described well-preserved specimens of M. geniculatus that show growth lines. These specimens have an isograptid type of proximal-end development and a platycalycal arrangement. Although th 1^ grows downward along the sicula for most of its length, its aperture is directed outward away from the sicular aperture. The distal parts of the proximal thecae also show outward curvature away from the sicula. On distal stipe fragments the metathecal segments are straight and oriented at low to intermediate angles to the dorsal stipe wall. All thecal apertures are provided with ventral apertural processes. Several of Skevington's (1965) specimens show growth lines indicating a metathecal ontogeny that is identical to that interpreted above for Isograptus'. In addition, an abrupt increase in the thickness of individual fuselli.(a decrease in growth-line density) occurs at the contact between the prothecal and metathecal segments of each theca. Maeandrograptus schmalenseei, which is described by Bulman (1932) differs from M. geniculatus in the greater curvature of the proximal 337 thecae. The proximal end has an isograptid type of development and a platycalycal arrangement. The thecae are oriented at a low angle to the dorsal stipe margin and bear ventral apertural processes. Two other species of Maeandrograp tus, M. tau and M. aggestus, are known only from compressed specimens, but both show the thecae to be furnished with ventral apertural processes. Both species show the characteristic manubrium of Pseudisograptus, which might represent a concentration of early thecae about the sicula. 3. Pseudisograptus Beavis. Bulman (1968) described specimens of Isograptus manubriatus (T. S. Hall) from Texas that are preserved in almost full relief, and Skevington (1968) assigned this species to Maeandrograptus. However, after a detailed study of 2- manubriatus, Beavis (1972) established a new genus, Pseudisograptus, to encompass the manubriate isograptids j[. manubriatus (T. S. Hall), 1C. hastatus Harris, and I. dumosus Harris. Bulman (1968) showed that the proximal-end development of Pseudisograptus manubriatus is of the isograptid type. Th 1^ grows down along the sicula for most of its length, turning outward away from the sicula at its aperture. The first six thecae of each stipe develop rapidly in the region of the sicula and show strong semi- circular curvature in lateral view. It is significant that th 4 to th 6 ^ appear to originate on the obverse side of the sicula, thus partly enclosing the sicula and simulating a pericalycal arrangement. In distal parts of the rhabdosome, the metathecae are inclined at low to intermediate angles to the dorsal stipe margin, and all the thecae show ventral apertural processes. One specimen of P^. manubriatus that 3 3 8 is figured by Cooper (1973, text-fig. 22o) shows distal thecae that may be partly developed and represent a "growing tip". If this is true, the morphological similarities to Bulman's (1932) specimens of Isograptus gibberulus suggest a thecal ontogeny that is similar to that interpreted above for I. gibberulus. 4. Apiograptus Cooper and McLaurin. This genus was erected to include Glossograptus? crudus Harris and Thomas, Glossograptus? crudus var. gisbomensis Harris and Thomas, and Paracardiograptus abnormis Yao (Cooper and McLaurin, 1974). Although the only known specimens representing this genus are preserved as compressed films on shale surfaces, they show, in outline, a proximal end that is similar to Pseudisograptus manubriatus and stipes that are scandent. Specimens representing early growth stages show a semi-circular curvature of the earliest thecae in such a way that the apertures of the thecae open outward away from the sicula whereas the proximal parts of the thecae are concentrated about the sicula to form a manubrium. Cooper and McLaurin (1974) interpret the morphological similarities between the proximal ends of Pseudisograptus and Apiograptus to reflect similar patterns of development. More distally, as described by Cooper and McLaurin (1974), the stipes change sharply to scandent growth, and as a result, the stipes are in contact first with the manubrium then with each other. They partly overlap each other laterally suggesting an incipient monopleural arrangement. In distal parts of the rhabdosome, the metathecae are straight and inclined at low to intermediate angles to the dorsal stipe margin, and all thecae are provided with ven tral apertural processes. The nema extends distally between, and 339 beyond, the stipes as a virgula. 5. Glossograptidae Lapworth. All the genera of Glossograptidae have a pericalycal proximal end with a variable number of the proximal thecae showing strong semi-circular curvature. In those cases in which the proximal end is well-known, its development is of the isograptid type. An exception is Glossograptus holmi Bulman as described by Whittington and Rickards (1969); however, their interpretations of a "primitive" dichograptid type of development is open to debate. The stipes in all genera of the Glossograptidae have a monopleural arrange ment except in distal parts of rhabdosomes of Apoglossograptus. In distal parts of the rhabdosome, the thecae are oriented at inter mediate angles to the dorsal stipe margin. All thecae in all the genera of the Glossograptidae have ventral apertural processes, and the thecal ontogeny, as interpreted below for ciliatus. is similar to that interpreted above for Isograptus gibberulus. 6 . Cryptograptidae Hadding, emend. Bulman. Cryptograptus, which is the only genus in the Cryptograptidae, shows a pericalycal arrangement of the proximal end with most of the proximal thecae showing strong semi-circular curvature. The proximal-end development is of the isograptid type, as described below for C^. marcidus, or represents a slight modification, of the isograptid type as re-interpreted below for C^. tricomis. The stipes have a monopleural arrangement, and in distal parts of the rhabdosome, the thecae are oriented at intermediate angles to the dorsal stipe margin. Except for tricornis, all species of Cryptograptus show ventral apertural processes. The thecae also show a short free ventral wall, however 340 growth-line patterns indicate that the thecal ontogeny is similar to that interpreted above for Isograptus gibberulus and nearly identical to that described by Skevington (1965) for Maeandrograptus geniculatus. 7. Corynoididae Bulman. Although rhabdosomes representing Corynoides Nicholson consist of only three thecae, the proximal-end development, thecal morphology, and thecal ontogeny is remarkably similar to that of Glossograptus (See Discussion below under Corynoididae). Corynites Kozowski differs from Corynoides by its elaborate apertural processes. General trends and constant features can be discerned among the genera and families summarized above and suggest the evolutionary scheme shown in text-figure 51. Beavis (1972) and Cooper and McLaurin (1974) have postulated and substantiated in detail the evolutionary lineage of Isograptus-Maeandrograptus-Pseudisograptus- Apiograptus. The common features that unite these genera are an isograptid type of proximal-end development, a prosicular origin of th 1 ^, and an orthograptid type of theca with a ventral apertural process. The evolutionary trend from Isograptus to Apiograptus is suggested by progressive changes in the morphology of the proximal end. These changes are: 1) An increase in the number of proximal thecae that originate in the sicular region. 2) An increase in the degree of semi-circular curvature of the proximal thecae, as well as an increase in the number of proximal thecae that show semi-circular curvature. 3) A change from a platycalycal arrangement to a simulated pericalycal arrangement. 341 4) A change from two reclined uniserial stipes to an incipient monopleural arrangement of the two thecal series. Apiograptus has been suggested as a morphological intermediate between Isograptus and Glossograptus (Harris and Thomas, 1935 and Thomas, 1960). As discussed above, the isograptid origin of Apiograptus is well documented. However, because Whittington and RickardsC 1969) reported a "primitive" dichograptid type of proximal- end development for jG. holmi Bulman, the speculation of an Apiograptus origin for Glossograptus has been considered to be premature (Cooper and McLaurin, 1974). Recent discoveries reported below show that ciliatus Emmons has an isograptid type of proximal-end development, and Whittington and Rickards' (1969) figured specimens of G^. holmi Bulman are re-interpreted as representing a proximal-end development that is, or that can be easily derived from, the isograptid type. This isograptid development of Glossograptus, together with general morphological features of Apiograptus (the incipient monopleural and pericalycal stipe arrangement and long, straight thecae with ventral apertural processes), strongly suggest that Glossograptus evolved from Apiograptus. A trend towards a biserial rhabdosome with a pericalycal, monopleural stipe arrangement characterizes the Isograptus-Apiograptus lineage and culminates with Glossograptus. However, structural modifications of the Apiograptus rhabdosome that are not present in the Isograptus-Apiograptus lineage must have occurred if Glossograptus arose from Apiograptus. These modifications involve the curvature and growth direction of the most proximal thecae and determine the 342 arrangement and orientation of the thecal series. In Pseudisograptus manubriatus, the simulated pericalycal stipe arrangement is accomplished by the curvature around to the obverse side of the sicula of portions of stipe no. 1 (e.g. th 4^ - th 6^). The morphological resemblance of early growth stages suggests a similar proximal-end arrangement in Apiograptus. In addition, a comparison of Cooper and McLaurin's (1974) text-figures 2b, 2d, and 2g indicate that the adjacent thecal series are in contact along their right-lateral walls in Apiograptus. In contrast, the monopleural arrangement of the stipes in Glossograptus is initiated with the first pair of thecae and the pericalycal arrangement is accomplished by growth around to the obverse side of the sicula of the two most proximal thecae of the second thecal series. As a result, the two thecal series are in contact along their left-lateral walls. The modifications that are required to derive the structural plan of Glossograptus from that of Apiograptus can be accomplished solely by changes in the curvature and direction of growth of the first two thecae. In Apiograptus, the first two thecae show semi-circular curvature that is convex relative to the ventral thecal wall. Thus, the initial stipe arrangement is platycalycal, and a pericalycal, monopleural stipe arrangement can only be accomplished by the bending around to the obverse side of the sicula of more distal parts of one of the stipes, in this case stipe no. 1. In Glossograptus, the first two thecae are straight and grow in such a way that they are situated on opposite sides of the sicula, thus initiating a pericalycal arrangement. The growth of the second pair of thecae with their ventrally convex curvature results in a monopleural stipe arrangement. 343 Thus, in spite of what appears to be major structural differences, the morphological gap between Apiograptus and Glossograptus is crossed by a simple straightening of the first two thecae. This allows the second thecal series (stipe no. 2 ) to develop on the obverse side of the sicula, and a biserial rhabdosome is attained in a more simple manner. If continued, the trend in the morphological changes that produce the simulated pericalycal and incipient monopleural rhabdosome in Apiograptus would probably not culminate in a truly pericalycal, monopleural rhabdosome because of the initial platycalycal arrangement. This initial platycalycal arrangement in the basic rhabdosomal struc ture of Apiograptus is an inherited feature of the Isograptus- Pseudisograptus lineage and results in a concentration of thecae on the reverse side of the sicula, thus producing an asymmetrical rhabdosome. However, by a simple change in the curvature and growth direction of the most proximal thecae that results in an initial pericalycal arrangement, a truly symmetrical, biserial rhabdosome is attained in Glossograp tus. The contrasts between the diversities, stratigraphie ranges, and geographic distributions of the Apiograptus and Glossograptus structural plans (text-figs. 51) strongly suggest that there is selective advantage of the latter over the former. The basic structural plan of Glossograptus consists of a pericalycal, monopleural stipe arrangement that is initiated with the first pair of thecae and is associated with an isograptid, or slight modification of the isograptid, type of proximal-end development. As discussed below (See Discussion under Glossograptidae, Cryptograptidae, and Corynoididae), the structural plan of Glossograptus characterizes all the other genera of the Glossograptidae, as well as the Crypto- 344 graptidae and the Corynoididae. Thus, as indicated in text-figure 51, Lonchograptus, Nanograptus, and Apoglossograptus are interpreted as arising from Glossograptus by relatively simple structural modifica tions (See Discussion below under Glossograptidae). Paraglossograptus differs from Glossograptus to a greater extent than the other genera of the Glossograptidae by its metathecal origin of th 1^. However, the similarities between Glossograptus and Paraglossograptus in thecal morphology and rhabdosome shape suggest a close affinity of these two genera. Likewise, Cryptograptus differs from Glossograptus by the metasicular origin of th 1^. In addition, the thecae of Cryptograptus have short free ventral walls. Yet, because the basic structural plan and the thecal ontogeny of Cryptograptus are basically the same as those of Glossograptus, a glossograptid origin, probably associated closely with the origin of Paraglossograptus, is favored herein for Cryptograptus. As discussed below (See Discussion below under Corynoididae) Corynoides can be easily derived from Glossograptus by means of arrested development at an early stage in the astogeny of the rhabdosome. Corynites, in turn, is interpreted as having evolved from Corynoides by an elaboration of the sicular apertural process. The evolutionary scheme proposed above and shown in text-figure 51 is based on the degree of morphological resemblance among the various taxa. As a test for this proposed scheme, the stratigraphie ranges of the various taxa were compiled and are shown in text-figure 52. The range of each taxa is shown individually for each of four areas North America, Britain, Baltoscandia, and Australia. The ranges for the taxa in North America are taken from Berry (1960), 345 EXPLANATION OF TEXT-FIGURE 52 Stratigraphie ranges of genera and families in the proposed phylogenetic scheme shown in Text-figure 51. Sources of data for stratigraphie ranges and correlation of zonal schemes are discussed in text (See Discussion under Glossograptina). Britain Baltoscandia A ustralia manitoutineniiy y^vgmSTus linearis C SDinritlUt O ruedtmanni D hians clingani C amancanus C. baragwanathi bicornis ef teretiusculus ?terêtiusculus Da m urchisoni bitidus bi/idus hirundo hirundoé gibberulus Ca nitidus Ch North America Britain 7 fruticosusOl Baltoscandia D.balticus A ustralia T. fruttcosua{4) w Text-figure 52 •p- O' 347 Riva (1974), and this report. The ranges in Britain are taken from Jackson (1962), Elies and Wood (1901-1918), and Toghill (1970a); for Baltoscandia, from EkstrBm (1937), Hede (1951), Erdtmann (1965), Tjernvik (1960), and Nilsson (unpublished data); and for Australia, from Thomas (1960). The correlation of the North American and European zonations is that of BergstrBm and Cooper (1973) for the Lower Ordovician and Riva (1974) for the upper Ordovician. The correlation of the Australian zonation to those of North America and Europe is based on Berry (1960) and Strachan (1972). As shown in text-figures 51 and 52, the stratigraphie ranges of the genera involved in the Isograptus-Apiograptus lineage overlap, and the first appearance of each genus in the lineage is slightly higher than that of its immediate ancestor and slightly lower than that of its immediate descendant. The critical part of the evolutionary scheme, which involves the first appearances of Glossograptus and Cryptograptus, is confused by questionable identifications and the lack of bed-by-bed collections from measured sections. In Baltoscandia, North America, and Australia, Glossograptus and Cryptograptus first appear at the base of the D. bifidus Zone, in terms of the British zonation. Monsen (1937) reports Glossograptus? sp. from the zone of Phyllograptus angustifolius elongates at Slemmestad in Norway. However, I have recently examined a large number of graptolites that were collected by Prof. Stig BergstrWm from the Slemmestad locality and are housed in the Orton Geological Museum. On the basis of this experience, I believe that Monsen's specimens, as they are illustrated in her 348 Plate 11, figure 1, represent the genus Phyllograptus. In England, Glossograptus and Cryptograptus are reported from the hirundo Zone (Jackson, 1962), but the exact horizon of their appearance within the hirundo Zone is not known. A first appearance of Glossograptus and Cryptograptus at the base of the hirundo Zone would not be consistent with their proposed origin from Apiograptus, if the correlation of the Australian and British sequences is correct as it is shown in text-figure 52. However, this correlation is open to some question, e.g. Harris and Thomas (1938) correlate the base of the hirundo Zone with the top of the Yapeenian. Thus, the first appearances of Cryptograptus and Glossograptus are placed in text- figure 51 at a level above the first appearance of Apiograptus because the exact levels at which Glossograptus and Cryp tograp tus first appear in the British hirundo Zone are not known, the correlation of the hirundo Zone with the Yapeenian is somewhat questionable, and Cryp tograp tus and Glossograptus first appear above Apiograptus in Australia. Paraglossograptus first appears in the upper part of the hirundo Zone (Rickards, 1972), which is consistent with its origin being closely related to that of Cryptograptus and Glossograptus. The stratigraphie ranges of Nanograptus, Lonchograptus, Apoglossograptus, and the Corynoididae are consistent with their proposed origin from Glossograptus. The first appearances of the Corynoididae seem to be quite variable geographically; this may be due to taxonomic mis- identifications (e.g. early growth stages of Glossograptus identified as Corynoides), exposure of graptolite sequences (e.g. the base of the gracilis Zone is not exposed in the south of Scotland), and lumping 349 of stratigraphically separate collections (e.g. the stratigraphie ranges of species in Thomas (1960) begin and end at zonal boundaries). The stratigraphie data together with the morphological inter pretations strongly support the proposed evolutionary scheme that is shovm in text-figure 51. The one debatable part of this scheme involves the first appearance of Glossograptus and Cryptograptus in the British sequence relative to the first appearance of Apiograptus in the Australian sequence. However, the stratigraphie ranges of these three genera in Australia are definitely consistent with the pro posed evolutionary scheme. This scheme can be summarized as follows: 1. During the time represented by the upper part of the gibberulus Zone, Maeandrograptus evolved from Isograptus by means of ventrally convex curvature of the proximal theca. 2. At the base of the hirundo Zone, Pseudisograptus evolved from Maeandrograp tus by an increase in the curvature of the proximal thecae, which was accompanied by, and probably caused, the curvature around the sicula of that part of stipe no. 1 consisting of th 2^ to th 6^. This resulted in a simulated pericalycal arrangement of the stipes about the sicula. 3. In the upper part of the hirundo Zone, Apiograptus arose from Pseudisograptus by a convergence of the reclined stipes into a biserial, and partially monopleural, arrangement. This convergence of the stipes increased the simulated pericalycal arrangement of the proximal end; however, it produced an asymmetrical rhabdosome because of the concentration of thecae on the reverse side of the sicula. 350 4. By means of a simple change in the growth direction and curvature of the proximal thecae, a truly symmetrical, biserial rhabdosome with a pericalycal, monopleural stipe arrangement evolved from Apiograptus in the uppermost part of the hirundo Zone. This new structural plan quickly developed along three separate lines represented by the genera Glossograptus, Paraglossograptus, and Cryp to grap tus. Whether or not each of these genera developed independently from Apiograptus is uncertain. Their similarities suggest that they evolved one from another. 5. The glossograptid lineage persisted until the multidens Zone and experienced some structural variability that is repre sented by the number of proximal thecae with semi-circular curvature. Several offshoots developed from the Glossograptus lineage by simple one-step structural modifications. These are: a) Lonchograptus with its single pair of long, stout spines in the murchisoni Zone, b) Nanograptus by means of arrested development at the base of the teretiusculus Zone, c) Apoglossograptus by the divergence of the two thecal series to form two reclined, uniserial stipes in the gracilis Zone, and d) the Corynoididae by means of arrested development in the gracilis Zone. 6 . The paraglossograptid lineage persisted until high in the gracilis Zone, but it did not give rise to any offshoots. It originated from the glossograptid lineage by an abrupt change to a metasicular origin of th 1 ^, by semi-circular 351 curvature of the first pair of thecae, and by the development of a lacinia. However, it retained a thecal morphology that resembles closely that of the glossograptid lineage. 7. The cryptograptid lineage persisted until the linearis Zone, but it did not give rise to any offshoots. Because of the metasicular origin of th 1 ^, the origin of the cryptograptid lineage was possibly associated with the origin of the paraglossograptid lineage. The straightness of the first pair of thecae and the basic thecal morphology in the earliest representatives of the cryptograptid lineage suggest that it evolved from an ancestor with a glossograptid structural plan. However, the short free ventral wall that characterizes the thecae indicates a deviation from the glossograptid and paraglossograptid lineages. A significant feature of the entire evolutionary scheme proposed herein is the similarity of the thecal morphology and ontogeny, as well as it can be determined, among all the various taxa. The thecae can be characterized morphologically as straight, long, and narrow and of the orthograptid type with ventral apertural processes. Except for structural modifications (e.g. the free ventral wall) in Cryptograptus, the variations in the thecae among the various taxa seems to be entirely quantitative, involving length to width ratios, curvature, angle of divergence between protheca and metatheca, inclination to stipe axis, amount of overlap, density, length to width ratio of apertural process, and inclination of apertural process 352 to thecal axis. What little is known of the thecal ontogeny seems to confirm the constancy of the thecae throughout the proposed evolutionary scheme. The thecal ontogeny is best known on the basis of growth lines for Skevington's (1965) Maeandrograptus geniculatus. In this species there is an abrupt change in growth line density at the boundary of the metasicula and prosicula. This change has been interpreted as representing an abrupt increase in the rate of development of the theca. The metatheca develops without its o\m ventral wall until its lateral walls are at, or near, the aperture of the preceding theca, then it develops its own ventral wall. The growth lines on this ventral wall are continuous with those on the lateral walls, yet they project distally in advance of the corresponding growth lines on the lateral thecal walls (Skevington, 1965; text-fig. 58). Thus, the ventral wall develops as a distally tapering process in advance of the lateral and dorsal walls of the thecae. When fully developed, the ventral wall forms a ventral apertural process. Thecal ontogenies, essentially similar to that summarized above for Maeandrograptus, are reported on the basis of growth line evidence for Glossograptus (See Thecal Ontogeny below under ciliatus), Cryptograptus (See Morphology and Development of Thecae below under C. marcidus), and Corynoides (See Discussion below under Corynoididae). Similar thecal ontogenies are also interpreted on the basis of partly developed thecae on "growing tips" of specimens representing Isograptus (See above), Pseudisograptus (See above), and Apoglossograptus (See Development of Thecae below under A. lyra). 353 Recent ultrastructural studies on graptolites suggest strongly that the periderm is secreted within an epithelial évagination (Urbanek, 1976). Thus, the ontogeny of a theca that is revealed by growth-line evidence is probably associated closely with the ontogeny of the zooid that secreted the theca. This association lends itself to two further speculations. First, the development of the ventral wall in advance of the other walls of the theca and its distally-tapering shape may be related to some particular "soft part" of the zooid. It might be argued that the ventral apertural process is merely a spine, but all the other spines that are present in taxa with ventral apertural processes are of an entirely different nature, structurally, morphologically, and ontogenetically, than the apertural processes (e.g. compare the development of lateral spines and apertural processes that are described below for G^. ciliatus). Second, the constancy of the thecal ontogeny throughout the proposed evolutionary scheme suggests that, except for Cryptograptus, the evolutionary changes involved primarily the morphology of the colony (rhabdosome) and not the morphology of the zooids (thecae). These evolutionary changes are consistent with a planktonic mode of life, which in the case of graptolites probably placed the greatest selective pressure on the shape of the colony. The contrasts that are noted above in the diversities, geographic distributions, and stratigraphie distributions, between the asymmetrical Apiograptus structural plan and the symmetrical Glossograptus structural plan, together with the constancy of the thecal morphology, supports the above speculations on the nature of the evolutionary changes within the proposed phylogenetic scheme. Family GLOSSOGRAPTIDAE Lapworth, 1873 Diagnosis As in Bulman (1970). Discussion As defined by Bulman (1970), the family Glossograptidae consists of the genera Glossograptus Emmons, Lonchograptus Tullberg, Nanograptus Hadding, and Paraglossograptus HsU. A new genus Apoglossograptus, which is described below, is assigned herein to Glossograptidae. The characters of Glossograptus are discussed below on the basis of well-preserved isolated specimens. Lonchograptus, which has been described only on the basis of compressed specimens on shale surfaces, is remarkably similar in outline to Glossograptus. The sole difference between the two genera seems to be the single pair of long, stout spines of Lonchograptus. Superficially, Nanograptus differs from Glossograptus by its small rhabdosome that consists of no more than 5 pairs of thecae- N. lapworthi, which is described by Hadding (1915a) on the basis of pyritized specimens preserved in relief, shows that: 1 ) th 1^ originates on the right-lateral side of the sicula and grows directly down along the sicula (Hadding, 1915a; PI. 6 , fig. 1); 2) th 1^ originates right-handedly from th 1^ (Hadding, 1915a; PI. 6 , fig. 2-3); 3) th 2^ and th 2^ originate from th 1^ (Hadding, 1915a; p. 329), 2 thus th 1 is the dicalycal theca, a'nd the proximal-end development 354 355 is the isograptid type; 4) the arrangement of the proximal end is pericalycal and monopleural; and 5) the most proximal thecae show semi-circular curvature, whereas the distal thecae are relatively straight and open distally. The thecal apertures in Nanograptus phylloides Elies and Wood are provided with ventral processes. The absence of ventral apertural processes in specimens of N. lapworthi may be due to the preservation of the specimens as pyritized internal molds. In spite of its small size and small number of thecae, Nanograptus is very similar to Glossograptus. Paraglossograptus is described on the basis of isolated specimens by Whittington and Rickards (1969), and the relations between species of this genus are discussed by Rickards (1972). Paraglossograptus differs from Glossograptus by the upward curvature of the first two thecae, the presence of a lacinia along the entire length of the sicula, and the origin of th 1^ in the metatheca. In spite of these differences, Paraglossograptus is remarkably similar 2 to Glossograptus. The right-hand origin of th 1 indicates an isograptid type of proximal-end development or a slight modification of that type of development depending on the position of the dicalycal thecae, which is presently unknown. The shape of the ventral apertural processes resembles closely that of Glossograp tus. As discussed below, Apoglossograptus has two reclined, uniserial stipes. However, the stipes show a pericalycal, monopleural arrangement in proximal parts of the rhabdosome. The proximal-end development and the thecal morphology of Apoglossograptus indicate strong affinities of this genus to Glossograptus. 356 As it is constituted herein, Glossograptidae shows a large range of variation in the general appearance of the rhabdosome. Basically, however, all the genera of Glossograptidae have: 1) an isograptid type of proxima 1-end development or possibly a slight modification of this type of development; 2) a pericalycal arrangement of the proximal end; 3) a monopleural arrangement of the stipes; 4) semi-circular shaped proximal thecae and relatively straight distal thecae; and 5) simple orthograptid thecae with spoon-shaped ventral apertural processes. On the basis of these features, a common ancestry is postulated here for all the genera of the Glossograptidae. Intra-familial relationships can be postulated on the basis of the degree of morphological similarity. As discussed below, Apoglossograptus probably evolved directly from Glossograptus ciliatus by a divergence of the thecal series. Lonchograptus can be derived easily from Glossograptus by the development of one pair of stout spines. Nanograptus with its definite isograptid type of proximal-end development could easily have evolved from Glossograptus by arrested development at the fifth pair of thecae. Paraglossograptus is similar to Cryptograptus, especially C^. tricornis, in the metasicular origin of th 1^ and the curvature of the first pair of thecae. However, as discussed below (See Discussion under Cryptograptidae), the curvature of the first pair of thecae in C^. tricornis seems to be confined to only that species in the genus, and in earlier representatives of Cryptograptus (e.g. £. marcidus), the first pair of thecae are straight and directed downward. Paraglossograptus is considered here to be closely related to, and most likely arose from. 357 Glossograptus because of the strong similarities in thecal morphology. The metasicular origin of th 1^ suggests a close relationship between the earliest representatives of Paraglossograptus and Cryptograptus. 358 Genus Glossograptus Emmons, 1855 Type Species: Glossograptus ciliatus Emmons, 1856 Diagnosis Rhabdosome with apertural, "dorsal", and in some cases lateral spines. Discussion On the basis of well-preserved specimens representing early growth stages, the proximal-end development of Glossograptus ciliatus is shown below to be of the isograptid type. The isolated specimens of G^. holmi Bulman that were interpreted by Whittington and Rickards (1969) to show a "primitive" dichograptid type of development are re-interpreted below as showing a right-hand origin of th 1^ and possibly a dicalycal th 1 . Although the presence of "dorsal" and lateral spines is variable among the different species of Glossograptus, all species show prominent ventral apertural processes, which are spoon-shaped in well- preserved specimens of G^. holmi and G^. ciliatus. Glossograptus ciliatus Emmons, 1856 (Text-figures 53-57) 1856 Glossograptus ciliatus (n.s.), Emmons, p. 108, PI. 1, fig. 25. 1856 Glossograptus setaceus (n.s.), Emmons, p. 236, PI. 1, fig. 20. 1859 Graptolithus spinulosus (n.s.). Hall, p. 517, fig. 9-13. 359 EXPLANATION OF TEXT-FIGURE 53 Text-figure 53 a-n. Glossograptus ciliatus Emmons. Isolated specimens representing early growth stages. a. Right-lateral aspect of prosicula and proximal end of metasicula. Note position of resorption foramen for th 1^ in prosicula'. X37.5. PS-76.4. OSU 33134. b. Ventral aspect of prosicula and proximal end of meta sicula with initial bud of th 1^. Dotted line indicates location of resorption foramen. X37.5. PS-126. OSU 33135. c,d. Ventral and dorsal aspects of specimen showing growth of th 1^ as incomplete tube. X37.5. PS-109.7. OSU 33136. e. Ventral aspect of distal end of metasicula. Note that growth lines bend downward at virgella. X37.5. PS-102. OSU 33137. f. Left-lateral aspect of distal end of metasicula and th 1^. Sicular aperture with virgella and lateral spines, and th 1^ with ventral process. X37.5. PS-109.2. OSU 33138. g,h. Dorsal and left-lateral aspects of proximal end of sicula showing th 1^, th 2l, and th 2^. X37.5. PS-102. OSU 33139. i,j. Left- and right-lateral aspects of middle part of sicula with th 1^, th th 2^, and th 2 . Note that th 2^ and th 2^ bud from th 1^. X37.5. PS-127.5. OSU 33140. k,l. Left- and right-lateral aspects of middle part of sicula with th 1^, th 1^, th 2^, and th 2 . Note that th 2^ and th 2^ bud from th 1^. X37.5. PS-109.2. OSU 33141. m,n. Dorsal and ventral aspects of sicula with th 1^ and th 1^. Note right-hand origin of th 1^ from th 1^ and lateral spines on sicular aperture. X37.5. PS-Î09.2. OSU 33142. , p s th2 ,f h |l l-'rf — s th 2 \ ms th - - th ' thi! th2' th 2 th l' J h \ ~i th2' th 2 ms th2 th 2 th I » s --Î----S-- w o\ o Text-figure 53 361 EXPLANATION OF TEXT-FIGURE 54 Text-figure 54 a-f. Glossograptus ciliatus Emmons. Isolated specimens representing early growth stages- a,b. Ventral and dorsal aspects (relative to sicula) of specimen showing th 3 budding from th 2^ and th 3^ budding from th 2^. X37.5. PS-117. OSU 33143. c,d. Dorsal and ventral aspects (relative to sicula) of specimen showing growth of th 1 , th 1^, th 2^, and th 2^. X37.5. PS-109.2. OSU 33144. e,f. Ventral and dorsal aspects (relative to sicula) of specimen showing growth of th 2^, th 2^, th 3^, and th 3^. X37.5. PS-109.2. OSU 33145. 'th 2 ' " t h I ' -t h I' vp th l^v th 2%_ th th t^vp LO ON Text-figure 54 to 3 6 3 EXPLANATION OF TEXT-FIGURE 55 Text-figure 55 a-f. Glossograptus ciliatus Emmons. Isolated specimens representing growth stages. a,b. Right- and left-lateral aspects of specimen with th 1^ - th 4 and th 1 - th 3 . Note growth of theca around rhabdosome. X37.5. PS-117. OSU 33146. 2 p c,d. Dorsal and ventral aspects of specimen with th 1 - th 5 and th 1^ - th 5 . Note upward direction of growth of distal parts of th 3 - th 5. X37.5. PS-109.2. OSU 33147. e,f. Left- and right-lateral aspects of specimen that is complete through th 6^. X37.5. PS-109.2. OSU 33148. 354 «h,( Text-figure 55 365 EXPLANATION OF TEXT-FIGURE 56 Text-figure 56 a-g. Glossograptus ciliatus Emmons. a,b. Dorsal and ventral aspects (relative to sicula) of ^ isolated specimen showing apertures of sicula, th 1 , and th 1^. Note relationship of growth lines to lateral spines. X37.5. PS-109.7. OSU 33149. c,d. Lateral aspects of isolated distal stipe fragment representing a "growing tip." Note relationship of thecae to virgula. X37.5. PS-109.2. OSU 33150. e. Isolated specimen representing distal stipe fragment. Note shape of ventral apertural processes. X37.5. PS-117. OSU 33151. f. Sketch of proximal-end development. g. Biprofile aspect of non-isolated, compressed specimen. Note change in orientation of spines along rhabdosome. X4.5. PF-19. OSU 33152. 366 w th (n + lr th(n+l)' X Text-figure 56 367 EXPLANATION OF TEXT-FIGUKE 57 Text-figure 57 a-k. Glossograptus ciliatus Emmons. Non-isolated, compressed specimens. a. Dorsal or ventral aspect. Note length of spines on sicular? aperture. X4.5. PF-25. OSU 33153. b. Intermediate aspect showing ventral apertural processes and "dorsal" spines along both ventral margins of specimen. X4.5. PF-19. OSU 33154. c. Intermediate aspect showing only "dorsal" spines along ventral margins of specimen. X4.5. PF-19. OSU 33155. d. Biprofile aspect showing ventral apertural processes along both ventral margins. X4.5. PF-25. OSU 33156. e. Dorsal aspect of early growth stage consisting of sicula, th 1^, and partly developed th 1^. X18.8. PF-17. OSU 33157. f. Sub-biprofile aspect showing ventral apertural processes along one ventral margin and "dorsal" spines along the other ventral margin. X4.5. PF-25.5. OSU 33158. g. Scalariform aspect showing "dorsal" spines along both margins. X9. PF-20.5. OSU 33159. h. Scalariform aspect. Note overlap of dorsal apertural margin by ventral apertural process of succeeding thecal aperture. X9. PF-25. OSU 33160. i. Young rhabdosome. X18.8. PS-29. OSU 33161. j. Intermediate aspect. X4.5. PF-19. OSU 33162. k. Intermediate aspect. X4.5. PF-19. OSU 33163. 368 Text-figure 57 369 1908 Glossograptus ciliatus Eimnons, Ruedemann, p. 379-383, Text-figs. 324-335, Pl. 26, figs. 1-5; Pl. 27, figs. 1-4. 1931 Glossograptus ciliatus Emmons var. douglasi (Lapworth MS), Bulman, p. 75, Pl. 12, figs. 11-13. 71934 Glossograptus ciliatus Emmons, Ruedemann and Decker, p. 317-318, Pl. 42, fig. 8. 1947 Glossograptus ciliatus Emmons, Ruedemann, p. 449-450, Pl. 77, figs. 27-40. 1947 Glossograptus ciliatus (Emmons) var. antennatus n. var., Ruedemann, p. 450-451, Pl. 77, figs. 41-44. 1947 Corynoides tricornis n. sp., Ruedemann, p. 362, Pl. 58, figs. 61-64. 71952 Glossograptus ciliatus Emmons, Decker, Pl. 1, fig. 43; Pl. 2, fig. 70. 71952 Glossograptus ciliatus var. antennatus Ruedemann, Decker, Pl. 2, fig. 71. 1952 Corynoides tricornis Ruedemann, Decker, Pl. 1, fig. 16; Pl. 2, figs. 9, 9a. 1955 Glossograptus ciliatus Emmons, Harris and Thomas, p. 40-41, text figs. 40-44. 1960 Glossograptus ciliatus Emmons, Turner, p. 90-91, Pl. 5, fig. 5; Pl. 8, fig. 8. 1960 Glossograptus ciliatus Emmons var. douglasi Bulman, Turner, p. 91, Pl. 8, fig. 3. 1960 Corynoides tricornis Ruedemann, Berry, p. 69. 1960 Glossograptus ciliatus Emmons7, Berry, p. 71. Type Data The location of Emmon's (1856) figured specimen of Glossograptus ciliatus, which is the holotype by monotypy, is unknown. It is most likely lost. The selection of a neotype must await a critical examination of Hall's (1859) figured specimens of Graptolithus spinulosus, which is synonymized above with ciliatus, and Ruedemann's (1908) figured specimens of G^. ciliatus. Diagnosis A species of Glossograptus with rhabdosome 1.2 mm wide in scalariform view and 3.2 mm wide in biprofile view. Sicula, th 1^ and th 1^ with paired lateral spines. All thecae with spoon-shaped, ventral processes similar to sicular virgella. "Dorsal" spines on every second theca proximally, at greater intervals distally. Plane defined by apertures of sicula and first four thecae at 45 degree angle to plane defined by all subsequent thecal apertures. Thecae inclined at 40-50 degrees to rhabdosome axis; number 6.5 - 8 in 5 mm proximally, and 5.5 - 6.5 in 5 mm distally. Material More than 100 isolated specimens are available for study. Half of these are early growth stages; six specimens show five to seven pairs of thecae; and more than 50 specimens are proximal thecal fragments. Most of the available isolated specimens are compressed and carbonized, but a few specimens show partial relief and early growth stages generally are transparent. Two hundred and seventy specimens, mostly large rhabdosomes, are available for study as carbonized films on shale surfaces. A few of these specimens retain slight relief. Half of the specimens from the Pratt's Ferry and Pratt's Syncline sections show in some cases detailed morphology that contrasts with the light-colored rock matrix. The rest of the specimens are from Calera, and detailed morphology is difficult to study because the specimens blend with the black back ground of the shale. Description Isolated Specimens. Because the largest available specimen, which consists of seven pairs of thecae, is less than 3 mm long, information that is obtained from the isolated specimens is restricted to the most proximal end of the rhabdosome and to early growth stages. Morphology of the Proximal End. The largest available isolated specimens are those shown in Text-figure 55 e-f. These specimens are partly to fully compressed, but they suggest that the rhabdosome originally was oval to circular in cross-section. Viewed laterally the rhabdosome is somewhat fusiform in shape, but the maximum width of 1.16 mm is located proximally at the level of the third pair of thecae, and the rhabdosome tapers distally. 1 2 The distal portions of th 3 and th 3 are horizontal, and their ventral walls form prominent features on the rhabdosome. Below these walls, the rhabdosome consists of the distal portions of the sicula and the first four thecae, which are oriented directly and obliquely downward, and their numerous spines. Above the ventral walls of the distal parts of the third pair of thecae, the rhabdosome shows a rather uniform structure. When traced proximally along the thecae, the ventral walls of the third pair of thecae change from a horizontal orientation to a vertical orientation. The point at which this change in orientation occurs forms the base of the common canal of each thecal series. The common canals are prominent features of the rhabdosome, and the common canal of one thecal series is located adjacent to the thecal apertures of the opposing thecal series. The second pair of thecae open in an obliquely downward direction, whereas all subsequent thecal apertures open in an 372 obliquely upward or distal direction. The apertures of the sicula and the first four thecae open in directions that define a single plane with the axis of the rhabdosome. All subsequent thecae of both thecal series also open in directions that form a single plane with the axis of the rhabdosome, but this plane does not coincide with that of the most proximal thecae. Instead, the plane of the distal thecae is approximately 30 degrees in a clockwise direction, when the rhabdosome is viewed proximally, from the plane of the proximal thecae (Text-figs. 55 c-f). This feature is discussed further in the description of the development of the proximal end. The thecae are oval tubes 0.4 - 0.5 mm in diameter, and proximally they show a sinuous direction of growth. Beginning with th 3, they number 2.3 in 1 mm and 4 in 2 mm. The thecal apertures are relatively simple, and except for those of th 1 1 and th 1 2 , they show only a single, ventrally-located process that is identical in appearance to the sicular virgella and is 0.5 mm long in the largest available specimen. The distal ends of the ventral processes are spoon-shaped and tilted downward (Text-fig. 56e), and the lateral edges of the processes, which merge proximally with the margins of the thecal aperture (Text-fig. 54 c-d, 56 c-d) , are turned upward, giving the process a wide U-shape in cross-section. Growth lines on the distal part of the ventral wall (Text-fig. 54 c-d, 56 a-d) bend toward the ventral process and indicate a somewhat similar mode of development to that of the sicular virgella. In addition to the ventral processes, the apertures of th 1 1 and th 1 2 show paired, horizontal, lateral spines and a gentle, dorsal lappet, and thus 373 they are identical in appearance to the sicular aperture (Text-figs. 54 c-d, 56 a-b) . In the development of the sicula (See Description below), the lateral spines develop only when the aperture is complete (Compare Text-figs. 53e and 53 m-n) , and the growth lines do not bend toward the spines as they do for the virgella (Text-figs. 53 m-n, 56 a-b). These facts suggest that the spines are not composed of fusellar tissue and they may be composed of microfusellar tissue as reported by Whittington and Rickards (1969) for Glossograptus holmi Bulman. Development of Proximal End. The sicula is easily recognized by its elongate, conical shape and its three apertural spines, including the virgella (Text-fig. 53 m-n). It ranges in length from 2.0 - 2.5 mm and averages 2.2 mm among the available specimens. The prosicula, which is poorly preserved, has been recognized in very few specimens and then only by a slight color difference. It is cone- shaped, less than 0.3 mm long, and accounts for less than a sixth of the length of the sicula (Text-figs. 53a, m-n). The metasicula, which is composed of transversely oriented fuselli, is 0.15 mm wide at its contact with the prosicula and gradually widens to a maximum width of 0.3 mm immediately above the apertural spines. The aperture of the sicula shows a downward-directed virgella, two horizontally-directed lateral spines, and a gentle lappet that is located on the anti- virgellar margin (Text-fig. 53 m-n). The virgella originates within 0.6 mm of the apex of the sicula (Text-figs. 53 b, e). Along its entire length, the transverse growth lines of the metasicula bend downward as they approach the virgella (Text-figs. 53e, m-n). At an 374 early stage of development (Text-fig. 53e), the virgella is a spine, whereas in more developed rhabdosomes, the distal end of the virgella is spoon-shaped with the lateral edges turned upward to form a wide U-shape (Text-figs. 54 c-d, 55f) . The lateral edges of the virgella appear as dark lines, which may represent secondary deposits of cortical tissue along the margins of the virgella. The lateral spines are developed late in the ontogeny of the sicula (Compare text-figs. 53 e, m). Because growth lines do not bend toward the spines as they do for the virgella, the spines are probably not composed of fusellar tissue. The lateral spines are consistently spine-shaped (Text-figs. 53m, 55e). They are up to 1.0 mm long among the available isolated specimens, and the virgella is slightly shorter. Th 1^ originates from a foramen in the right-lateral wall of the prosicula (Text-figs. 53a, m). By the addition of U-shaped fuselli (the sicula serves as the ventral wall), th 1^ grows downward and over to the virgellar side of the sicula (Text-figs. 53 b-d, n). At 1 2 a high level on the metasicula, th 1 gives rise to th 1 from its right-hand side (Text-fig. 53m), after which it continues growing downward along the virgella. Some time late in its development, probably after the stage shown in Text-figure 53m, th 1^ becomes a complete tube by developing a ventral wall, which is in contact with the virgellar wall of the sicula. The formation of a ventral wall appears to be associated ifith the development of a ventral process at the level of the sicular aperture. This process, which is essentially identical to the sicular virgella, diverges from the sicular virgella at an angle of about 50 degrees (Text-fig. 53f). 1 375 Growth lines on the ventral wall of th 1 hand toward the ventral process, thus indicating that the ventral process of th 1^ forms in the same manner as the sicular virgella (Text-fig. 56a, b). When it is completely developed, the aperture of th 1^ is located slightly above, and is nearly identical to, the sicular aperture (Text-figs. 54c-f, 56a-b). It shows a ventral process with a spoon-shaped distal end, two horizontal lateral spines, and a gentle dorsal lappet (Text-fig. 54e, 56a-b) . As with the lateral spines of the sicula, those on th 1^ form at a late stage in its development (compare text- figs. 53f and 56a), and the fuselli of the thecal walls do not bend toward the spine. Thus, the lateral spines of th 1 and the virgella are probably of a similar construction. Cortical tissue begins to fill in the angle between the virgella and the ventral process of th 1^ at an early stage in the development of the rhabdosome (Text- figs. 53c-d, 55e). After budding from th 1 1 , th 1 2 grows horizontally across to the anti-virgellar side of the sicula where it gives rise to th 2^ and th 2 (Text-figs. 53g-j, n). It then grows doim the anti-virgellar side of the sicula as a split tube without a ventral wall until it develops an aperture that is transverse to the th 1^ aperture. As 1 2 with th 1 , the aperture of th 1 is identical to that of the sicula, showing a ventral process with a spoon-shaped distal end, two horizontal lateral spines, and a slight lappet (Text-figs. 55f, 56a-b). The specimens shown in Text-figures 53 g-h and 53 i-j are compressed in different directions that are at approximately 45 degrees 376 to each other. They both show that th gives rise to th 2^ and th 2^ shortly after its origin from th 1^. Th 2^ develops from the left- hand side of th 1^, and it is oriented obliquely downward toward th 1^ and the virgellar side of the sicula (Text-figs. 53 i-1). Th 2^ 2 originates from the right-hand side of th 1 at the same level that th 2^ originates. Initially, it is oriented obliquely downward across the left-lateral wall of the sicula, which is the opposite side of the sicula that th 2^ is growing along (Text-figs. 53 i-1). Within a short distance of its origin, th 2^ gives rise to th 3^ from its right-hand side (Text-fig. 54b), then th 2^ grows obliquely doxmward across the sicula and the distal part of th 1^ (Text- figs. 54 b-c). Proximally, th 2^ is in contact with the left-lateral 2 1 wall of th 1 . Distally, it wraps around the sicula and th 1 in such a way that the aperture comes to lie immediately above, and on the dorsal wall of, the aperture of th 1^ (Text-fig. 54 c,e). Because of the sinuous direction of growth, the aperture of th 2^ opens in a direction slightly to the left of the direction in which th 1^ opens, and the right side of the ventral wall of th 2^ rests on the proximal part of the right-lateral spine of the th 1^ aperture. The th 2^ aperture bears only a single ventral process, which is spoon-shaped and identical to the sicular virgella and the ventral processes of th 1^ and th 1^. o 2 Within a short distance of its origin, th 2 gives rise to th 3 from its right-hand side, then bends to grow downward and backward 2 across the sicula and th 1 in such a way that the aperture comes to lie immediately above, and on the dorsal wall of, the aperture of 377 2 th 1 (Text-figs. 54 a, d, f). In the orientation of its aperture, 2 2 1 1 th 2 shows the same relation to th 1 as th 2 shows to th 1 . In 2 addition, th 2 has only a single spoon-shaped ventral process. Th 3 1 and th 3 2 show directions of growth that are similar to 1 2 1 those of th 2 and th 2 , respectively. After its origin, th 3 2 bends downward and grows along the left-lateral side of th 2 . Distally, it wraps around the rhabdosome by growing horizontally across, and above the aperture of, th 2^ (Text-figs. 54 b, f; 55 a,e). The aperture of th 3^, which displays a single spoon-shaped ventral process, opens immediately above, and in a direction to the left of the th 2^ aperture (Text-figs. 55 d-e). Th 3^ initially grows downward along the left-lateral side of th 2^ before bending and 2 growing horizontally across th 2 (Text-figs. 54 a, e, 55 a-b, f). Its aperture, which also shows a single spoon-shaped ventral process, opens immediately above, and in a direction to the left of, the aperture of th 2^. Each theca of the fourth thecal pair show a downward initial direction of growth for only a short distance before bending upward and growing around the rhabdosome (Text-figs. 55 a-f). The apertures of th 4 1 and th 4 2 open immediately above, and in the same direction 1 2 as, the apertures of th 3 and th 3 , respectively. Beginning with the fifth pair of thecae, all subsequent thecae show an initial upward direction of growth (Text-figs. 55 c-f). The type of proximal-end development described above is the 2 isograptid type because it shows a right-hand origin of th 1 , two 2 crossing canals, and a dicalycal th 1 (Text-fig. 56f). A monopleural 378 arrangement of the thecal series is initiated with the first two thecae because they are on opposite sides of the sicula. However, 2 because th 1 is the dicalycal theca, the two thecal series do not separate until the second pair of thecae. When the rhabdosome is viewed from the proximal end, th 2^ has a clockwise direction of 2 growth. Th 2 initially shows a counter-clockwise direction of growth, but distally for most of its length it bends back to grow in a clock wise direction. All subsequent thecae show clockwise directions of growth. The apertures of the sicula and the first four thecae open in directions that fall within the same plane (Text-fig. 54 e-f). However, all subsequent thecae open in directions that define a plane that is different to that defined by the proximal thecae (Text-figs. 55 e-f, 56 c-d). Such an arrangement can be explained by the length and direction of growth of the thecae. Assuming, for example, that th 2 and th 3 are of the same length, the fact that th 3 buds from th 2 indicates that th 3 would extend beyond th 2 if it grew in the same direction. Because th 3 grows around, and above the aperture of th 2, its length would necessitate that it extend completely across th 2, and thus it would open in a direction that is significantly different from that of th 2. In spite of this arrangement, all the apertures of a thecal series are always on the same side of the rhabdosome. Thecal Ontogeny. Text-figures 56 c-e show specimens that represent portions of the rhabdosome that are distal to the prosicula. The distal end of both specimens is a "growing tip", and the virgula 379 extends a significant distance beyond the most distal theca. The left-lateral wall of each theca is in contact with the virgula. None of the available specimens show growth lines on the protheca because of the presence of a thick cortical tissue. As described for th 1^, the development of the ventral thecal wall appears to be delayed until late in the thecal ontogeny when its development is associated with the formation of a ventral process, as revealed by the curvature of growth lines. There are no thecal growth stages that show the exact formation of the ventral wall, but text- figures 53 k-1, 54 a-f, 55 a-d, 56 c-d show that none of the thecae has a ventral wall during most of its development. This suggests that the ventral wall and the ventral process may develop in all the thecae as it does in th 1^. Non-Isolated Specimens. Several distinctive preservational aspects of the rhabdosome, which differ in shape, size, and spines, occur among the available material, and the oval to circular cross- section of the rhabdosome most likely permits this variation. In spite of the differences, all the aspects show: 1) a rounded proximal end with downward and laterally projecting spines, 2) parallel sides, and 3) a broad distal end from which a virgula may extend for several centimeters. In biprofile view (Text-fig. 56g) , the width of the rhabdosome, excluding spines, increases from 1.0 - 1.3 mm at the proximal end to a maximum of 2.3 - 3.2 mm at a distance of 6 - 8 mm from the proximal end, after which the rhabdosome is parallel-sided or gradually tapers distally. The lesser of the two values for the maximum width are from specimens that appear to retain partial relief 380 and that show the distal ends of the thecae along both lateral edges (Text-fig. 57d). The greater of the two values are from specimens that appear to be fully compressed and that only show the ventral apertural processes along both lateral edges (Text-fig. 56g). In scalariform view (Text-fig. 57g) , the rhabdosome is parallel sided, and the width is a uniform 1.2 mm. In intermediate views (Text-figs. 57 b-c, j-k), the rhabdosome is generally parallel-sided with the maximum width of 2.3 - 2.9 mm occurring within 5 mm of the proximal end. The thecae number 6.5 - 8 in 5 mm proximally, and 5.5 - 6.5 in 5 Iran distally. As revealed by the interthecal septa (Text-fig. 56g), the thecae are up to 4.0 mm long, 0.4 - 0.6 mm wide, and overlap for almost their entire length. They are inclined 40-50 degrees to the axis of the rhabdosome, and their apertures are perpendicular to their long axis. Apertural processes, which are 2.0 - 2.5 mm long, arise from the ventral margin of every thecal aperture (Text-figs. 56g, 57d). "Dorsal" spines, which are up to 2 ram long, arise somewhat irregularly at intervals of every two thecae in the proximal part of the rhabdosome and at greater intervals in distal parts of the rhabdosome. In Glossograptus holmi Bulman and G^. hincksi (Hopkinson) the distribution and orientation of spines were a matter of debate for some time (Hadding, 1913, 1915a; Bulman, 1931; Lemon and Cranswick, 1956), and this debate was finally settled by the discovery of isolated specimens (Jaanusson, 1960; Whittington and Rickards, 1969). Isolated specimens of G. ciliatus have been described herein for the 381 first time, but unfortunately, the available specimens, which represent relatively young colonies, lack dorsal spines. However, the following descriptions of the various preservational aspects, especially the scalariform aspect, explain the arrangement of the spines in the available specimens. In biprofile view (Text-figs. 56g, 57d), only the ventral apertural processes are visible, and they are evenly distributed along the margins of the rhabdosome. Proximally, they are inclined toward the proximal end of the rhabdosome; dis tally, they are inclined toward the distal end of the rhabdosome, and for most of the length of the rhabdosome they project horizontally from its margins. This distribution of the processes corresponds to that in Text-figure 7a of Lemon and Cranswick (1956). In scalariform view (Text-figs. 57 g-h), the ventral margins of the apertures are turned downward and indicate the location of the ventral apertural processes. However, these processes do not extend out of the margins of the rhabdosome. Instead, a different set of spines, which arise from the lateral margins of the thecal apertures, extends out of the margins of the rhabdosome. Isolated specimens show that in scalariform view, the apertures of one thecal series occupy the left side of the rhabdosome and the dorsal wall (common canal) of the other thecal series occupies the right side of the rhabdosome. On the basis of the orientation of the thecal series and observations of the nature of the spines in isolated specimens of G. holmi (Whittington and Rickards, 1969) and G^. hincksii (Jaanusson, 1960), the set of spines that extend out beyond the 382 margins of the rhabdosome are interpreted to arise from, or at least proximally to be in contact with, the dorsal wall of a thecal series. The distribution of these "dorsal" spines is very irregular among the specimens that present a scalariform aspect which may be a preserva tional feature as suggested by Text-figure 7e of Lemon and Cranswick (1956). In intermediate aspects (Text-figs. 57 b,j; Lemon and Cranswick, 1956, text-fig. 7c) "dorsal" spines and ventral apertural processes project from both sides of the rhabdosome. In sub—biprofile aspects (Text-figs. 57f; Lemon and Cranswick, 1956, text-fig. 7b), only "dorsal" spines project from one side of the rhabdosome, and only ventral apertural processes project from the opposite side. The "dorsal" spines are best preserved in specimens that present a subscalariform view (Text-figs. 57 c,k; Lemon and Cranswick, 1956, text-fig. 7d). Although the apertures of one thecal series should be present on the surface of the specimens that present this aspect, none of the available specimens is well-enough preserved to show the thecal apertures. In order to gain additional knowledge on the distribution of the "dorsal" spines, the distances between the proximal end of the rhabdosome and each spine along one side of the rhabdosome were measured on fourteen specimens (Table 8). These measurements are not very accurate because of varying degrees of preservational distortion of the available specimens, but they do show that the spines occur at regular intervals, which suggest that they are "dorsal" spines and not a preservational phenomenon of the ventral apertural processes. Comparison of the values shown in Table 8 TABLE 8: Distance in mm between "dorsal" spines and proximal end in specimens of ciliatus. Specimen first second third fourth fifth sixth seventh eighth Number spine spine spine spine spine spine spine spine OSU 33200 0.93 1.62 2.52 3.30 3.89 4.67 6.47 - OSU 33162 1.16 1.80 2.55 3.28 4.44 5.65 6.87 - OSU 33201 .093 1.68 2.38 3.10 3.83 - - - OSU 33202 0.87 1.62 2.32 3.05 3.83 --- OSU 33203 1.16 1.80 2.58 3.39 4.26 5.13 6.79 - OSU 33204 0.93 1.51 2.15 2.90 3.65 4.41 -- OSU 33205 1.01 1.60 2.47 3.10 4.26 - - - OSU 33155 0.96 1.74 2.44 3.30 4.32 5.48 - - OSU 33206 0.87 1.57 2.15 2.81 3.97 5.13 - - OSU 33207 0.87 1.80 2.47 3.48 4.26 5.42 6.87 - OSU 33208 0.87 1.65 2.26 2.96 4.26 --- OSU 33209 1.07 1.68 2.47 3.28 4.44 5.71 6.93 - OSU 33163 1.04 1.68 2.47 3.25 4.47 5.68 6.90 9.22 OSU 33210 0.96 1.68 2.32 2.84 4.00 5.10 - - mean 0.97 1.67 2.40 3.15 4.13 5.24 6.81 9.22 Distance from preceeding spine 0.70 0.73 0.75 0.97 1.09 1.57 2.41 384 with the observed thecal densities indicates that the "dorsal" spines arise in proximal parts of the rhabdosome at intervals of one spine for every 1.5 - 2 thecae. Distally, the intervals between spines appear to increase to two or three thecae. The measurements in Table 8 also indicate that the first two, and possibly three, "dorsal" spines on one side of the rhabdosome occur before th 7, which is the most distal theca in the largest available isolated specimen. However, none of the available isolated specimens, not even the largest, show "dorsal" spines. This suggests that the development of the "dorsal" spines is significantly delayed relative to the development of the rhabdosome. Whittington and Rickards (1969) recognized the same development of dorsal spines in Glossograptus holmi and reported that they are composed of microfusellar tissue. The specimens illustrated in Text-figures 57 a, e, i show additional aspects of the Alabama material. Text-figure 57a shows a specimen with 6 mm long proximal spines. These spines are lateral spines of either the sicula or one of the proximal thecae. As described for the isolated specimens, these spines have been interpreted as being composed of microfusellar tissue. Therefore, their extreme length can not be used as a diagnostic feature of the species. Text-figure 57e shows an early growth stage, which is compressed on a shale surface. It is composed of the sicula, which is 2.4 mm long, th 1 1 , and the initial bud of th 1 2. Text-figure 57i shows a more developed growth stage, which is compressed on a shale surface and is probably composed of at least seven pairs of thecae. Remarks 385 Development of Proximal End. As described above, the early growth stages of Glossograptus ciliatus show an isograptid type of 2 proximal-end development with a right-hand origin of th 1 , two 2 crossing canals, and a dicalycal th 1 . This development is very different from that interpreted for holmi by Whittington and 2 1 Rickards (1969) , in which th 1 has a left-hand origin and th 1 is the dicalycal theca. However, their interpretation was not based on early growth stages but on young colonies and fragments of the proximal end, which might be interpreted to show a proximal-end development similar to that described above. Text-figure 3b of Whittington and Rickards (1969) shows that th 1^ extends proximally onto the right-lateral wall of the sicula. The 2 position shown for the proximal end of th 1 and the visibility of the left-lateral wall of the sicula between the proximal parts of th 1^ ■? and th 1 indicate a right-hand origin of th 12 from th 1 1 . Whittington and Rickards (1969) do not showthe origins of th 2^, but o 1 1 th 2 arises from the right-hand side of th 2 , and th 2 can be 2 interpreted to arise from the left-hand side of th 1 . Confirmation of this re-interpretation of the proximal-end development of G^. holmi Bulman must, however, await the examination of Whittington and Rickards' (1969) figured specimens. The proximal-end development of Glossograptus described by Mu and Zhan (1966) is not based on early growth stages but interpreted from the morphology of the proximal ends of large rhabdosomes. Their specimens of G^. sinicus and G^. kep ingens is are very similar to the oog specimens of G. ciliatus described above and this suggests a similar mode of development. Mu and Zhan's (1966) interpretation that the proximal two thecal apertures of one series are associated with the more distal apertures of the other thecal series can not be determined on the basis of their available specimens, and on the basis of my observations in specimens of ciliatus, it seems likely that they have misinterpreted their specimens. Because the first four thecae grow directly downward along the sicula. Mu and Zhan (1966) suggest that Glossograptus is initially dipleural and becomes monopleural only with the development of the third pair of thecae. However, as described for ciliatus, all the apertures of a thecal series are closely related on the rhabdosome, and the separation of the apertures of the two thecal series occurs with the development of the first pair of thecae. The first four thecae are thus not "back-to- back" as is found in a dipleural arrangement. Relations to Other Species in the Genus. The Alabama specimens can be distinguished from Glossograptus holmi Bulman by lack of lateral bulges in the proximal end of the rhabdosome, by absence of right-lateral spines (if the dorsal spines in the described specimens actually arise from the lateral margin of the aperture and not from the dorsal wall of the adjacent thecal series, they can be designated as left-lateral spines), by absence of a lacinia, and by downward- directed apertures of the second pair of thecae. In addition, the thecae of G^. holmi differ from the simpler thecae of G. ciliatus by the presence of a dorsal hood of microfusellar tissue and a transverse ridge at the base of the ventral apertural process. Because 387 quantitative characters such as thecal density, maximum width, and thecal inclination, are closely similar in the two species, poorly preserved specimens that are compressed on shale surfaces can only be differentiated if several different preservational aspects are present in order to determine the distribution of spines. In Glossograptus holmi, all the thecal apertures open within the same plane, and the lateral bulges in the proximal end are probably formed to accomodate the length of the thecae (Whittington and Rickards, 1969; p. 803-804). The rotation of the plane in which the thecal apertures open at the third pair of thecae in the Alabama material is interpreted as accomodating the extra length of the thecae, and this may explain the absence of lateral bulges in those specimens. Glossograptus hincksii (Hopkinson), as it is defined by Hopkinson (1872), shows 10 thecae in 10 mm. Elies and Wood (1908) describe G^. hincksii as having 16 thecae in 10 mm proximally, and 10 in 10 mm distally. They (p. 311) state that its affinities are with G. ciliatus, only differing from that species by more closely spaced thecae. Ruedemann (1908, 1947) reported 11 thecae in 10 mm in his description of ciliatus, which included specimens from Pratt's Ferry. The specimens from Pratt's Ferry described above show 16 thecae in 10 mm proximally and 10 thecae in 10 mm distally. Hopkinson's (1872; PI. 12, fig. 9) and Elies and Wood's (1908; text-fig. 205d) illustrations that show a scalariform view can be interpreted as showing "dorsal" (or left-lateral apertural) spines, but not right-lateral spines. In southern Scotland G. hincksii 388 occurs in the Nemagraptus gracilis and Climacograptus peltifer Zones, and this stratigraphie range is similar to that of G. ciliatus. 2» hincksii (Hopkinson) may therefore be nonspecific with G. ciliatus, in which case 2- hincksii would be a junior synonym. However, a decision on this matter is delayed until type specimens of G. hincksii can be investigated. The specimens of Glossograptus hincksii that are reported from Sweden by Hadding (1913, 1915a) show both left- and right-lateral spines on the thecal apertures and are restricted to the lower subzone of the Glyptograptus teretiusculus Zone. They show only 9 to 10 thecae in 10 mm and do not appear to be nonspecific with the described specimens of 2* ciliatus or the British specimens of 2* hincksii. However, the assignment of a new name to the Swedish specimens is delayed until the type specimens can be investigated. Berry (1964) reports Glossograptus hincksii from the 2* teretisculus Zone in the Oslo Region. His specimens show thecal densities that are similar to those reported for 2* ciliatus and G. hincksii (non Hadding, 1913). However, Berry (1964) describes these specimens as showing both right- and left-lateral apertural spines, and the specific assignment of these specimens appears to be somewhat questionable on the basis of the preceding discussion. Mu and Zhan (1966) describe two new species of Glossograptus, G. sinicus and 2* kepingensis, that resemble closely the specimens described above in their general size and orientation of proximal thecae. 2* sinicus differs from the Alabama material by the low angle of inclination of the proximal thecae. 2- kepingensis is slightly narrower than the Alabama specimens. 389 Relations to Taxa Listed in Synonymy. Although he does not give any quantitative characters, the described specimens are closely similar in appearance to the illustrations of Emmon's (1856) specimens of ciliatus and G^. setaceus. Ruedemann. (1908, 1947) reports a thecal density of "11 in 10 mm with little variation." On the basis of an examination of Ruedemann's reference specimens, his G. ciliatus is considered to be nonspecific with the Alabama material. The discrepancy in the thecal densities is probably due to Ruedemann's measurements being located distally on the rhabdosome. Ruedemann (1947) erected the subspecies antennatus on the basis of specimens that showed long proximal spines. As discussed above, the length of these spines is not considered to be of taxonomic value. Ruedemann's (1947) type specimens of Corynoides tricomis are very similar in appearance to the earliest growth stages of the specimens described above. Therefore, C^. tricornis is regarded herein as a junior synonym of G^. ciliatus. Because of differences to Ruedemann's (1908) figures of thecal density for Glossograptus ciliatus, Bulman (1931) differentiated his specimens of G^. ciliatus as the subspecies douglasi, which show 7-8 thecae in 5 mm. Bulman's specimens are small rhabdosomes; his measurements are on the proximal part of the rhabdosome; and thus his specimens are here considered to be nonspecific with the Alabama material. Ruedemann and Decker (1934) and Decker (1952) describe and/or illustrate Glossograptus ciliatus and the subspecies antennatus, as well as Corynoides tricornis. Their specimens, however, are poorly 390 preserved, and are here only tentatively assigned to G. ciliatus. Berry (1960) reports ciliatus? and Corynoides tricomis from the Marathon region; however, he does not illustrate his specimens and his descriptions are very brief. Harris and Thomas (1955) describe specimens of Glossograptus ciliatus from Australia, and Turner (1960) describes specimens of ciliatus and G^. ciliatus douglasi from South America. These Australian and South American specimens agree closely with the Alabama material. Figured Specimens OSU 33134 - OSU 33163. 391 Genus Apoglossograptus n. gen. Type Species: Isograptus lyra Ruedemann, 1947 Derivation of Name Apo- from Greek for away, from, separate - referring to the interpreted direct evolution of this genus from Glossograptus. Diagnosis Same as for species described herein. Discussion Apoglossograptus, as represented solely by the species A. lyra Ruedemann, is remarkably similar to Glossograptus, but differs by the distal divergence of the two thecal series in such a way that they form two reclined, uniserial stipes. The similarities are: 1) A proximal-end development with an origin of th 1^ high on the 2 right side of the sicula, a right-hand origin of th 1 , and 2 possibly a dicalycal th 1 . 2) A primarily downward direction of growth of the first two thecae with the th 1^ aperture directed downward. The ventral process of th 1^ and the sicular virgella are in contact at their bases. 3) A clockwise rotation in the direction of growth of the thecae in such a way that their apertures open outward and upward. In Apoglossograptus this rotation originates with the second pair of thecae. 392 4) A pericalycal proximal end and a monopleural arrangement of the thecal series in the proximal end. 5) A simple orthograptid type of theca with a single, spoon shaped, ventral process, a large amount of overlap, and an intermediate angle (30 - 60 degrees) of inclination. Although the stipes diverge distally in Apoglossograptus, the point at which the divergence occurs is somewhat variable. In those specimens in which the divergence is located most distally, the part of the nema that is nearest the sicula is enclosed between the lateral walls of the two thecal series resulting in an axonophorous condition. 2 The orientation of the th 1 aperture in Apoglossograptus, which opens in a subhorizontal direction, differs from the downward 2 direction in which the th 1 aperture is oriented in Glossograptus. As shown in the range chart for the Calera section (Text-fig. 9), the stratigraphie ranges of Glossograptus ciliatus and Apoglossograptus lyra overlap, and ciliatus decreases in abundance as A. lyra increases. A. lyra is known only from the Athens Shale, and it has been reported from as far north as Bristol, Tennessee (Decker, 1952). In contrast, G^. ciliatus has a nearly worldwide distribution. Because of the overlapping stratigraphie and geographic ranges and the morphological similarities, it seems highly probable that A. lyra arose directly from ciliatus by means of a distal divergence of the two thecal series. Although,the point at which stipe divergence occurs is somewhat variable in Apoglossograptus, no specimens are known that show a really intermediate position between the scandent stipes of Glossograptus and the reclined stipes of Apoglossograptus. 393 Thus, the evolution of Apoglossograptus probably occurred by an abrupt crossing of a morphological discontinuity. Apoglossograptus lyra (Ruedemann, 1947) (Text-figures 58-59) 1947 Isograptus lyra n. sp. , Ruedemann (partim), p. 353-354, PI. 57, figs. 43-46 (non fig. 47). 1947 Dicellograptus moffatensis (Carruthers) var. alabamensis Ruedemann, Ruedemann (partim), PI. 64, fig. 16 (non figs. 12-15). 1952 Isograptus lyra Ruedemann, Decker, PI. 2, fig. 77. Type Data The holotype of Apoglossograptus lyra is the specimen illustrated by Ruedemann (Ruedemann, 1947, PI. 57, figs. 43-44) as Isograptus lyra. This specimen (NYSM 10493) , which has been examined by the writer, is stored at the New York State Museum, Albany and agrees with Ruedemann's (1947) description. Diagnosis Rhabdosome with biserial, monopleural, axonophorous proximal end from which two reclined, uniserial stipes diverge distally. Stipe divergence 30 - 90 degrees, complete at about level of apex of 1 2 1 prosicula. Th 1 and th 1 in complete contact with sicula; th 1 2 aperture opens downward; th 1 aperture opens subhorizontally; apertures of all subsequent thecae open obliquely upward. Maximum uniserial stipe width 2.0 mm. Thecae inclined at 30 degrees to stipe axis, numbering 5.5 - 7 in 5 mm proximally and 4.5 - 5.5 in 5 mm distally. 3 9 4 EXPLANATION OF TEXT-FIGURE 58 Text-figure 58 a-e. Apoglossograptus lyra (Ruedemann). Non-isolated specimens. a. Lateral aspect. Note nema free within axil. X4.5. C-25-28. OSU 33164. b. Lateral aspect of specimen that retains partial relief. Note that one stipe overlies the other in the proximal end. X9. C-35. OSU 33165. c. Lateral aspect. Note apex of prosicula visible in axil. X4.5. C-51.5. OSU 33166. d. Lateral aspect of young rhabdosome. Note that sicula is between the two thecal series. X9. C-45. OSU 33167. e. Lateral aspect. Note sicula situated between two stipes. X18.8. C-45. OSU 33168. 395 th 2' Text-figure 58 396 EXPLANATION OF TEXT-FIGURE 59 Text-figure 59 a-d. Apoglossograptus lyra (Ruedemann). Non-isolated specimens representing early growth stages. a. Ventral aspect of sicula with th 1^. X18.8. C-34. OSU 33169. b. Dorsal aspect of sicula with th 1 1 and th 1 2 . Note right-hand origin of th 1^ from th 1^. X18.8. C-34. OSU 33170. c. Dorsal aspect of sicula with th 1^, th 1^, ?th 2^, and ?th 2^. X18.8. C-34. OSU 33171. d. Lateral aspect of specimen. Note growth of thecae "x" and "y" in clockwise direction (in proximal aspect) around rhabdosome. X18.8. C-34. OSU 33172. th iL.i th I X-- th I^-----1 -t h I* a w VO Text-figure 59 • v j 398 Material The available material consists of 42 specimens that are preserved as carbon films on shale surfaces. A few of the specimens are early growth stages; several show partial relief in the proximal end. Although preserved as carbon films, definite thecal boundaries are visible on several of the specimens that represent early growth stages. Description Morphology of Rhabdosome. Although the stipes show a reclined habit, the most characteristic feature of the available specimens is the monopleural arrangement of the stipes in the proximal end of the rhabdosome. In the available specimens that show partial relief (Text-fig. 58b), one stipe, when traced proximally, overlies the other stipe. Text-figures 58d and 58e show opposite lateral aspects of young rhabdosomes (the terms obverse and reverse are meaningless here). In both aspects, the sicula can be seen to overlie one stipe and to underlie the other. Although the stipes are not completely in "side-to-side" contact, as they are in Glossograptus, parts of the second and third, and in some cases up to seventh (Text-fig. 58b), thecal pair are laterally in contact. The proximal end has a pericaly cal arrangement, and the arrangement of the stipes in the proximal end can be considered as monopleural. The divergence of the stipes is variable ; the angle between them ranges from 30 to 90 degrees among the available specimens. Distally, the stipes may be straight and show uniform divergence (Text-fig. 58a), or they may show convex curvature relative to the ventral margin, and 399 converge and be subparallel distally. A nema, up to several millimeters long and situated between the stipes, is commonly present in the available specimens. In a few specimens, stipe divergence occurs at a low level on the rhabdosome. As a result, the apex of the prosicula is visible in the axil (Text-fig. 58c). The largest available rhabdosome (Text-fig. 58a) has 23 mm long stipes. The stipe width, excluding apertural processes, averages 1.7 mm and ranges from 1.5 to 2.0 mm in the available specimens. The thecae number 5.5 - 7 in 5 mm proximally and 4.5 - 5.5 in 5 mm distally. They are basically of the orthograptid type, but a 1.0 mm long process projects from the ventral margin of the aperture. In well-preserved specimens (Text-figs. 58 d-e), these processes show a rounded distal end, which bends downward relative to the thecal aperture. The first two thecal apertures are in contact with the sicula, and the th 1^ aperture opens directly downward, whereas 2 the th 1 aperture opens subhorizontally. Succeeding proximal thecae initially grow downward, then bend in such a way that their apertures open in a distal direction. Distal thecae, which are up to 3.0 mm long and 0.5 mm wide, are straight, overlap for three-fourths of their length, and are inclined at less than 30 degrees to the dorsal stipe margin. Development of Proximal End. The sicula ranges in length from 4.0 to 5.5 mm in the available specimens. For most of its length, it is very narrow, but near the aperture it rapidly widens (Text- fig. 59a). The available specimens are too poorly preserved to reveal the detailed morphology of the aperture, but one specimen (Text- 400 fig. 59b) shows the dorsal (anti-virgellar) margin of the aperture, the bases of lateral spines, and a broad virgella, which extends 1.0 - 1.5 mm below the aperture. Th 1^ originates near the apex of the sicula from its right side (Text-figs. 59 a-b). This high position suggests a prosicular origin. Th 1^ grows downward and quickly around the sicula in such a way that its aperture is in contact with the virgellar margin of the sicular aperture (Text-fig. 59c). None of the available specimens reveals the aperture of th 1^. However, if the thecae and their associated ventral processes are traced proximally in Text-figure 58 d-e, the two adjacent, and most proximal processes, must represent the sicular virgella and the ventral process of th 1^. These two processes can only be interpreted as being in contact at their bases which indicates that the virgellar wall of the sicula and the ventral wall of th 1^ are in "back-to-back" contact. 2 Th 1 , as it is shown in Text-figure 59b, is interpreted as originating from the right-hand side of th 1^. It then grows across to, and then down, the anti-virgellar wall of the sicula (Text-fig. 59c). Two pieces of carbon film on the specimen shown in Text-figure 59c lie at different levels within the rock. The piece that is designated as ?th 2^ overlies, and cross-cuts, the margins of 1 2 2 th 1 and th 1 . The piece that is designated as ?th 2 is overlain, and its boundaries are cut across, by the margins of the sicula and th 1 . If these interpretations are correct, they could indicate left- and right-hand origins of th 2 1 and th 2 2 , respectively, from th 1^. 401 The most proximal end of the specimen illustrated in Text- figure 59d is too poorly preserved for the determination of thecal designations. It does suggest that the first six thecae develop rapidly with their origins concentrated high above the sicular aper ture. After growing directly downward, the most distal theca, which is designated by "x", bends and grows across the rhabdosome in a clockwise direction (rhabdosome viewed proximally) to open obliquely upwards on the opposite side of the rhabdosome. The theca that is designated by a "y" grows directly downward and is overlain distally by the aperture of the "x" theca. It probably bends and extends behind the rhabdosome appearing on the opposite side. Although they are somewhat poorly preserved, e.g. pieces of periderm have flaked off, the specimens illustrated in text-figures 58 d-e suggest a similar clockwise rotation of proximal thecae about the rhabdosome. Development of Thecae. The most distal theca on stipe no. 1 of the specimen illustrated in text-figure 58 d-e is only partly developed. It is growing along the dorsal wall of the preceding theca. Its gross morphology and its relationship to the preceding theca are suggestive of a thecal ontogeny that is similar to that of Isograptus, Maeandrograptus, Glossograptus, and Cryptograptus. This ontogeny involves the growth of th n+1 as a split tube without its own ventral wall (the dorsal wall of th n serves as the ventral wall of th n+1) up to a point close to its aperture. When th n+1 develops its own ventral wall, the growth of the ventral wall is associated closely with the development of a ventral apertureal process. The lateral margins of the ventral process of th 5 (Text-fig. 58d) appear to 402 merge with the margins of the thecal aperture. This feature is also suggestive of the glossograptid type of thecal ontogeny. Remarks It can definitely be stated that the specimens described above show; 1) a monopleural arrangement of the stipes proximally; 2) a pericalycal arrangement of the proximal end; 3) an origin of th 1 1 high . on the sicula; - 4) a right-hand origin of th 1 0 from 1 1 o th 1 ; 5) a downward direction of growth of th 1 and th 1 in contact with the sicula and with a downward — opening aperture of th 1^; 6) an early concentration in the budding of proximal thecae; 7) clockwise rotation beginning with the second pair of thecae in such a way that their apertures, and all subsequent thecal apertures, open outward and upward; and 8) orthograptid type of thecae with ventral processes. Several specimens, representing early growth stages, are well enough preserved to show the margins of the sicula and the thecae. However, because they do not show growth lines, some of the observa tions that are made on the early growth stages are somewhat inter pretive and open to debate. These debatable observations involve the origin of th 2 1 and th 2 2 . In spite of this, the morphology and the development of the proximal end of the described specimens is remarkably similar to that described for Glossograptus ciliatus, and differ primarily in the divergence of the stipes and the subhorizontal 2 orientation of the aperture of th 1 . This strongly suggests an isograptid type of proximal-end development for the described specimens. 403 Ruedemann's (1947) type specimens of Isograptus lyra were collected from the road cut that represents the uppermost part of the Calera section described herein. He (p. 353) described the shape of the stipes as "...reclined to such a degree that they become parallel then bending out somewhat and finally reapproaching again, thus assuming a somewhat lyra-shaped outline.". Only two of the available specimens show this lyra-shape. However, all other characters are identical, and the specimens described above were collected from the type locality of Isograptus lyra Ruedemann. One of Ruedemann's (1947; PI. 57, fig. 47) figured specimens, which is from the Glenogle Shale of British Columbia, was referred with doubt to lyra. The thecae in the proximal end of the rhabdosome of this specimen show an orientation that is found in Isograptus, but neither in the other figured specimens of 1. lyra nor in my collections. One figured specimen that Ruedemann (1947; PI. 64, fig. 16) reports as Dicellograp tus moffatensis alabamensis is nearly identical in appearance to the proximal end of the specimens described above. Figured Specimens OSU 33164 - OSU 33172 404 Family CRYPTOGRAPTIDAE Hadding, 1915 a, emend. Bulman, 1970 Diagnosis Characters of genus. Genus Cryptograptus Lapworth, 1880 Type Species : Cryptograptus tricornis (Carruthers, 1859) Diagnosis As in Bulman (1970). Discussion Bulman (1938, 1944) was the first to describe isolated specimens of Cryptograptus. He reported a "primitive" dichograptid type of proximal-end development for £. tricornis (Carruthers). In this type of development, th 1^ is the dicalycal theca and th 1^ originates left-handedly from th 1^. As described below, C^. marcidus has an isograptid type of proximal-end development, and by homologizing the sicular list structures and crossing-canal growth lines of C^. tricornis and C^. marcidus, C^. tricomis is re-interpreted below to show a right- handed origin of th 1^. By an advance in the position of the dicalycal theca from th 1^ to th 1^, the proximal-end development of C^. tricornis represents a slight modification of, and can be easily derived from, the proximal-end development of C^. marcidus. Differences in the orientation of the first two thecae (directly downward in C^. marcidus, outward in C^. tricornis) and the side of the sicula on which the first thecal series opens (virgellar side in C^. marcidus, anti-virgellar side in C. tricornis) are associated with the different positions of the dicalycal theca; however, in both species the initial growth of the 405 second pair of thecae is characteristically spiral shaped. Although the two species differ in the presence of apertural processes, both show the same basic thecal morphology and, as determined by growth line evidence, the same thecal ontogeny. Other species of Cryptograptus are known only as compressed films on shale surfaces. C^. hopkinsoni (Nicholson), which is only figured in a biprofile aspect by Elies and Wood (1908) has ventral apertural processes and resembles C^. marcidus. iC. antennarius (Hall) and C^. minimus Ruedemann, as illustrated by Ruedemann (1947), have ventral apertural processes, and the first two thecae in iC. antennarius may be oriented directly downward (Ruedemann, 1947; PI. 76, fig. 3). The Cryptograptidae can be distinguished easily from the Glossograptidae on the basis of features that are described in well- preserved specimens. The characteristic features of the Cryptograptidae are: 1) The relatively shorter sicula with its list structure. 2) The origin of th 1^ in the metasicula, although Paraglossograptus also shows a metasicular origin of th 1^. 3) The initial spiral direction of growth of the second pair of thecae. 4) The absence of thecal spines other than the ventral apertural processes. 5) The presence of a short free ventral wall. 6) A sharp change at the prothecal-metathecalcontact in the direction of growth of each distal theca. 4 0 6 In spite of these differences, the Cryptograptidae and Glossograptidae show many similarities in their basic structural plan. These similar ities are: 1) An isograptid type of proximal-end development with some species of both families showing slight modifications as represented by the position of the dicalycal theca. 2) A pericalycal arrangement of the proximal end and strong semi-circular curvature of several of the proximal thecae. 3) A lame Hi form process, not a spine, for a virgella. 4) A rapid proliferation of the first six pairs of thecae of the rhabdosome together with a rapid change in their orientation (downward to upward) and curvature (strongly curved to relatively straight). 5) A monopleural arrangement of the stipes. 6) Simple orthograptid thecae that exhibit ventral apertural processes, in most species, and similar thecal ontogenies. In accordance with Bulman (1970), the difference between Cryptograptidae and Glossograptidae are considered herein to be at the familial rank. Because of the similarities in their basic structural plan, Cryptograptidae and Glossograptidae are grouped together in the suborder Glossograptina and most likely share a common ancestry. 407 Cryptograptus marcidus (Hall, 1859) (Text-figures 60-63) 1859 Graptolithus marcidus (n.s.). Hall, p. 514-515, figs. 1-3. 1860 Graptolithus marcidus Hall, Hall, p. 58-59, figs. 1-3. 1880 Cryptograptus tricomis var. Schaeferi, Lapworth, PI. 5, fig. 28a-b, 1908 Cryptograptus tricomis (Carruthers), Ruedemann (partim), text-fig. 410 (non text-figs. 411-417; PI. 28, fig. 104). 1908 Cryptograptus tricomis (Carruthers) var. Schaferi Lapworth, Elies and Wood, p. 299, text-figs. 201a-b, PI. 32, figs. 13a-c. 1913 Cryptograptus lanceolatus n. sp., Hadding, p. 40, PI. 2, figs. 10-12. 1913 Cryptograptus tricomis Carr., Hadding, p. 40-41, PI. 2 figs. 13, 14. 1915a Cryptograptus lanceolatus Hdg., Hadding, p. 324-325, PI. 6, fig. 16. 1931 Cryptograptus tricomis var. schaferi Lapworth, Bulman, p. 65-66, text-fig. 31, PI. 6, figs. 1-5; PI. 7, fig. 3. 1933 Cryptograptus schaferi (Lapworth) var. latus nov., Bulman, p. 352, PI. 33, figs. 8a-c. 1935 Cryptograptus schaferi (Lapworth), Harris and Thomas, p. 304, fig. 3, nos. 11, 12. 1937 Cryptograptus tricomis (Carruthers), Ekstrom, p. 39, PI. 9, figs. 1-5. 1937 Cryptograptus tricomis var. longispinus nov. Ekstrom, p. 40, PI. 8, fig. 13. 1937 Cryptograptus lanceolatus Hadding, Ekstrom, p. 40, PI. 8 fig. 11-12. 1947 Cryptograptus tricomis (Carruthers), Ruedemann (partim), PI. 76, Fig. 23 (non PI. 76, figs. 24-33). ?1947 Cryptograptus tricomis (Carruthers) var. schaferi Lapworth, Ruedemann, p. 447, PI. 76, fig. 48. 408 EXPLANATION OF TEXT-FIGURE 60 Text-figure 60 a-i. Cryptograptus marcidus (Hall). Isolated specimens. a. Scalariform aspect showing ventral apertural processes. X18.8. PS-109.2. OSU 33173. b. Biprofile aspect showing downward-directed ventral apertural processes. X18.8. PS-109.7. OSU 33174. c. Dorsal aspect showing apertural spines and processes of sicula, th 1^, th 1^, and th 2^. X37.5. PS-117. OSU 33175. d. Distal stipe fragment showing growth-line pattern on lateral walls of prothecae and metathecae. X37.5. PS-117. OSU 33176. e,f. Right- and left-lateral aspects of distal stipe fragment. Note apertures of one thecal series adjacent to prothecae of the other thecal series. X37.5. PS-109.2. OSU 33177. g,h. Left- and right-lateral aspects of proximal end of rhabdosome. Note horizontal orientation of th 2, relationship of virgella to th 1^ ventral process, and position of lateral spines on sicular aperture. X37.5. PS-102. OSU 33178. i. Fragment showing relationship of virgella to ventral processes of th 1^ and th 1^. X37.5. PS-109.2. OSU 33179. 409 -îhZ^vp fh 1* vp th n+l / th n+lpt. thl'vp th rv p Text-figure 60 410 EXPLANATION OF TEXT-FIGURE 61 Text-figures 61 a-g. Cryptograptus marcidus (Hall). Isolated specimens. a,b. Ventral and dorsal aspects of specimen consisting of parts of sicula, th 1^, and th 2^. Note that growth lines on ventral process of th 2^ are continuous with those on lateral walls of th 2^, whereas growth lines on dorsal wall of th 1^ (interthecal septum) are continuous with those on lateral walls of th 1^. X37.5. PS-109.2. OSU 33180. c,d. Lateral aspects of fragment of theca showing growth lines on interthecal septum and lateral walls of theca. X37.5. PS-109.2. OSU 33181. e. Ventral aspect of theca showing growth lines that make up short free ventral wall above ventral apertural process. X37.5. PS-117. OSU 33182. f,g. Lateral aspects of stipe fragment. Note growth-line patterns on interthecal septa and lateral thecal walls. X37.5. PS-109.2. OSU 33183. 61 h. Sketch of proximal-end development of Cryptograptus tricomis based on Bulman (1944). Right-hand origin of th 1^ as interpreted herein. Note that second thecal series opens on virgellar side of sicula. 61 i. Sketch of proximal-end development of Cryptograptus marcidus. Note that second thecal series opens on anti- virgellar side of sicula. 411 th n+lpt thi'lw _ th n mt i'&llhZ'vp th,n mt -th rvp prth n pt ''th l'v p b U -v Text-figure 61 412 EXPLANATION OF TEXT-FIGURE 62 Text-figure 62 a-j. Cryptograptus marcidus (Hall). Isolated specimens representing early growth stages. a. Prosicula and proximal end of metasicula. Note spiral thread. X37.5. PS-117. OSU 33184. b. Ventral aspect of prosicula and metasicula. Note formation of virgella by curvature of fuselli. X37.5. PS-117. OSU 33185. c. Specimen with very deteriorated periderm. Note longitudinal lists. X37.5. PS-109.2. OSU 33186. d. List structure that rims sicula. X37.5. PS-109.2. OSU 33187. e,f. Proximal end of sicula showing th 1 1 and th 1 2 . Note right-hand origin and obliquely upward direction of growth of th l”. X37.5. PS-109.2. OSU 33188. g,h. Fragment showing middle part of sicula with th 1^ and th l2. X37.5. PS-109.2. OSU 33189. i,j. Ventral and dorsal aspects of sicula with initial bud of th 1^. X37.5. PS-109.2. OSU 33190. k.l. Ventral and dorsal aspects of compressed, fragmentary specimen. Note origin of th 2^ and th 22 from th 1 . X37.5. PS-109.2. OSU 33191. m,n Ventral and dorsal aspects of sicula, th 1 1 , and th 1 7 . Periderm very degenerated. X37.5. PS-117. OSU 33192. û- th |2- th 2 7 Text-figure 62 414 EXPLANATION OF TEXT-FIGURE 63 Text-figures 63 a-g. Cryptograptus marcidus (Hall). Non-isolated, compressed specimens. a. Sub-scalariform aspect showing ventral apertural processes. X7.2. C-22.5. OSU 33193. b. Scalariform aspect. Note that ventral apertural processes obscure dorsal margin of preceding thecal aperture. X7.2. C-44.5. OSU 33194. c. Scalariform section through young rhabdosome preserved in partial relief. X15. PS-12. OSU 33195. d. Scalariform section through rhabdosome preserved in partial relief. Note overlap of virgella and th 1 ventral apertural process. X7.2. PS-117. OSU 33196. e. Biprofile aspect. Note change in orientation of inter thecal septa. X7.2. PS-127.5. OSU 33197. f. Biprofile aspect. Note two downward-directed spines (processes) at proximal end. XI.2. PS-109.7. OSU 33198. g. Biprofile aspect of specimen with 3- (or possibly 2-) vaned virgular structure. Note characteristic downward- directed processes on proximal end. X I . 2. C-28.3. OSU 33199. 415 Text-figure 63 416 1952 Cryptograptus tricomis (Carruthers), Decker, Pl. 1, figs. 13,13a; Pl. 2, fig. 27; Pl. 3A, fig. 10. 1956 Cryptograptus tricornis schaferi Lapworth, Lemon and Cranswick, p. 19, text-figs. 4j-k. non 1960 Cryptograptus schaferi (Lapworth), Berry p. 69-70, Pl. 12, figs. 7,8. 1960 Cryptograptus tricornis (Carruthers), Berry, p. 70, Pl. 15, fig. 9. ?1963 Cryptograptus schaferi Lapworth, Ross and Berry, p. 96-97, Pl. 5, figs. 28,29. ?1963 Cryptograptus tricornis (Carruthers), Ross and Berry, p. 97-98, Pl. 5, fig. 27. 1964 Cryptograptus tricomis (Carruthers), Berry, p. 117, Pl. 9, figs. 1, 2a. 1964 Cryptograptus tricornis var. schaeferi Lapworth, Berry, p. 117- 118, Pl. 11, figs. 7-8. 1970 Cryptograptus tricomis schaeferi Lapworth, Skevington, p. 418- 422, text-figs. 6a-h, 7a-d. Type Data Hall's (1859) syntypes of Cryptograptus marcidus are stored at the American Museum of Natural History, New York. Hall did not designate a holotype among his syntypes. One of the specimens figured by Hall (1859, fig. 2) is selected herein as the lectotype. Diagnosis A species of Cryptograptus with isograptid pfoximal-end develop ment and downward-directed th 1^ and th 1^. Stipe width uniform; 0.7 - 1.7 mm in scalariform view; 1.3 - 2.0 mm in biprofile view including apertural processes; 0.6 - 1.2 mm in biprofile view excluding apertural processes. Thecae with downward-directed ventral apertural processes; number 5.5 - 8 in 5 mm proximally, 5 - 5.5 in 5 mm dis tally. Proximally, 417 metathecal axis perpendicular to rhabdosome axis; distally, metathecal axis inclined at 50 - 60 degree angle. Material Approximately 50 isolated specimens are available for study. Except for a few early growth stages, all the specimens are fragments of larger rhabdosomes broken during acid treatment. However, most of them retain some relief and show growth lines. More than 700 specimens that are compressed on shale surfaces are available for study. Half of these are from the Pratt's Ferry and Pratt's Syncline sections, are well-preserved, and commonly retain some relief. The remaining speci mens are from the Calera section and preserved as carbon films on shale surfaces. Description Isolated Specimens. Morphology of Rhabdosome. The isolated specimens are mostly proximal ends and distal segments, which are fragments of large rhabdosomes and generally consist of less than four pairs of thecae. The fragmentary nature of the specimens is probably related to the structure of their periderm, which is reported by several authors (Lapworth, 1880; Ruedemann, 1908; Elies and Wood, 1908; Bulman, 1944) to be thin in other species of the same genus. Even though they were isolated by the same techniques, specimens in my collections represent ing other species than the present one do not have the same degree of transparency. Thick cortical deposits are generally restricted to the most proximal ends of the rhabdosome of the specimens described herein. 418 and growth lines are widely spaced. These factors most likely account for the thin, fragmentary, very transparent nature of the periderm. Text-figures 60 a,g-h show scalariform and biprofile views of isolated proximal ends. The sicula and the first two thecae open directly downward and the aperture of each is furnished with a blunt virgella or ventral process, which is generally less than 1 mm long. In addition, paired apertural spines project laterally from the aper ture of the sicula (Text-fig. 60 c, 62 i-j). Beginning with the second pair of thecae, the two thecal series are in contact laterally; the distal parts of the thecae are horizontal; and the thecal apertures, which are furnished with short, blunt, downward-directed, ventral processes, open in a ventral direction. Thus, the rhabdosome in scalariform view shows a proximal end that is characterized by broad, blunt, downward-directed processes that can be superimposed to appear as a single process and two horizontal spines. Beginning with the second thecal aperture, the rhabdosome shows on the left a row of thecal apertures and on the right the dorsal wall of the adjacent thecal series. The dorsal margin of each thecal aperture is hidden by the downward- directed ventral process of the succeeding theca. In biprofile view, the proximal end of the rhabdosome shows two downward-directed spines representing the lateral edge of the th 1^ apertural process and the combined virgella and th 1^ apertural process. Beginning with the second pair of thecae, the rhabdosome shows the lateral wall of the thecal series whose apertures open to the right and generally only the apertural processes of the other thecal series along its left ventral 419 margin. The downward-directed ventral processes give the false impres sion that the thecal apertures are also directed downward. In distal fragments, the arrangement of the thecal series and the various aspects of the rhabdosome are similar to those described in proximal-end fragments, except for the orientation of the thecae. The second and third pair of thecae, which are the most distal thecae in the largest available specimen with a proximal end, show an initial downward direction of growth, and distally they bend to grow and open in a direction that is perpendicular to the axis of the rhabdosome. In distal fragments (text-figs. 60 a-f, 61 f-g) the thecae initially grow upward (distally) before bending and growing obliquely outward. As a result, their distal portions are oriented at a 50 - 60 degree angle to the axis of the rhabdosome, and their apertures open in a ventro-distal direction. However, ventral apertural processes, which are oriented obliquely downward, give the false appearance that the apertures open downward. In contrast to the thin periderm composing most of the rhabdosome, the virgula is always represented by a 0.05 mm wide rod of black, glossy tissue. As a result, the virgula can be seen, and its course can be traced, through the thecal walls of transparant specimens. Development of Proximal End. The sicula has the shape of a rapidly expanding cone and is 1.3 mm long. The prosicula accounts for a sixth of the length of the sicula and shows a spiral thread and longitudinal rods (text-figs. 62 a-b) . The metasicula widens from 0.10 ■ 0.15 mm at its contact with the prosicula to 0.4 mm at its aperture. It shows transverse growth lines that overlap to form zig-zag sutures 420 (text-fig. 62 i-j). Proximally, the growth lines are closely spaced and number 10 in 0.10 mm. Distally, the density of the growth lines decreases to 4 in 0.10 mm. Approximately 0.6 mm above the sicular aperture the growth lines begin to bend downward. This curvature is greatest for each growth line at the zig-zag suture. It increases dis tally and is responsible for the formation of the virgella. The virgella is almost 1.0 mm long and 0.10 mm wide, and its distal tip is blunt. When the sicular aperture is complete, it is furnished with subhorizon tal paired lateral spines, which are up to 0.5mm long in early growth stages. Because these spines are black, developed late in sicular ontogeny, and at angles to the growth lines of the sicula, they are probably composed of microfusellar or cortical tissue. At a late stage in the development of the sicula, the margins of its aperture and virgella are strengthened by a list structure (Text- fig. 62 d) . Two rods arise from this list structure near the lateral spines and extend proximally to the nema (text-fig. 62 c) . In many of the isolated specimens, the periderm of the sicula appears to have de generated and/or been destroyed because it is either absent or represent ed by nearly structureless tissue (text-figs. 62 c-d, m-n) . In such specimens, the sicula is recognized by the list structure. The first theca originates from a foramen that develops in the virgellar (ventral) wall of the sicula immediately above the point at which the virgella initially develops (text-fig 62 i). It immediately gives rise to th 1^ from its right-hand side, then grows down the virgellar side of the sicula (text-figs. 62 g-h, k-1). Ifhen it is 421 completely developed, th 1^ is oriented directly downward, and it is in contact with the virgellar wall of the sicula for its entire length. Its aperture, which is slightly above the sicular aperture, is provided with a wide process that is in contact with the virgella for its entire length and extends distally a distance equal to approximately one-half the length of the virgella (text figs. 60 g-i). This ventral process shows growth-line patterns that indicate a development similar to that of the virgella (text-fig. 60 i). As th 1 9 develops from the right-hand side of th 1 1 , it grows transversely towards the anti-virgellar side of the sicula and expands in such a way that it appears to grow upward to a level on the sicula above that of th 1^ (text-figs. 62 e-h) . Growth lines on the proximal part of th 1^ are roughly parallel to the longitudinal axis of the sicula. After reaching the antivirgellar side of the sicula, th 1^ grows directly downward along the sicula for its entire length (text- figs. 62 k-n). When th 1^ is complete, its aperture is at the same level as, and on the opposite side of the sicula from, the th 1^ 2 aperture. The aperture of th 1 displays a 0.5 mm long, wide, ventral process. Growth lines on this process suggest a development that is similar to that of the virgella (text-fig. 62i). Further growth stages are represented by a single, partly- compressed, fragmentary specimen that is illustrated in text-figure 62 k-1. This specimen shows segments of the sicula, th 1 , and th 1 . One theca shows growth lines on its proximal part that are diagonally oriented with respect to the sicular axis. These growth lines indicate 4 2 2 O that the theca on which they occur is th 1 . Distally, two, partly- developed, primary notches occur at the same level on th 1^. The 2 primary notch arising from the left-hand side of th 1 opens onto the dorsal wall of th 1^, while the primary notch arising from the right- hand side of th 1^ opens toward the right-lateral side of the sicula and th 1^. Evidently, the two primary notches that arise from th 1^ rep resent the second thecal pair in the rhabdosome. Because of lack of additional growth stages, thecal series designations for these two thecae can be made only by comparison with the specimen illustrated in text-figure 60 g-h in which distal parts of th 2^ and th 2^ show downward growth followed by clockwise (when the rhabdosome is viewed proximally) rotation and horizontal growth across the distal portions of the sicula and the first two thecae. Thus, the primary notch on the left-hand side of th 1^ must be th 2^, and the primary notch on the right-hand side of th 1 2 must be th 2 2 . In order for a reversed designation to agree with the thecal arrangement shown in text-figure 60 g-h, a counter-clockwise direction of growth of the thecae and/or a significantly greater thecal length would be required. The initial orientation of the primary notches, when compared to the orientation of the distal ends of the second pair of thecae indicates a very sinuous direction of growth in which the thecae initially grow horizontally, then bend downward and backward to grow around to the side of the proximal end that is in the opposite direction from which the primary notch is directed. 423 With the development of the second pair of thecae, the two thecal series become separated. Th 3^ and th 3^ would thus develop from th 2^ O and th 2 , respectively. As illustrated in text-figure 60 g-h, the third pair of thecae show an initial downward direction of growth fol lowed by clockwise rotation. The right-hand origin of the 1^ from th 1^, the dicalycal th 1^, and the two crossing canals represent an isograptid type of proximal-end development (text-fig. 61 i). With the clockwise direction of growth of the second and third pairs of thecae, the prox imal end has a pericalycal arrangement. The monopleural arrangement of the two thecal series is initiated with the development of th 1^ and th 1^ because these first two thecae are separated by the sicula. Morphology and Development of Thecae. The thecae, which have oval cross-sections in uncompressed specimens and elliptical cross-sections in compressed specimens, are composed of prothecal and metathecal seg ments of approximately equal length (text-figs. 60 d, 61 a-b). In distal parts of the rhabdosome, the protheca is parallel to the axis of the rhabdosome. Successive prothecae are aligned end-to-end and con stitute the common canal, which occupies the left side of the lateral wall and the right side of the ventral wall of the rhabdosome. The metathecal segments are inclined 50 - 60 degrees to the axis of the rhabdosome, and each metatheca is in contact with the preceding and succeeding metatheca of the thecal series. In biprofile view (text-figs. 60 d), the metathecae of one thecal series occupy the right side of the rhabdosome and open to the right. The thecal apertures have semi-circular shaped apertures that are 424 accentuated by subhorizontal to downward-directed apertural processes. In scalariform view (text-fig. 61 e), the thecal apertures occupy the left side of the rhabdosome's ventral wall and are adjacent to the prothecae (common canal) of the other thecal series. Distal thecae are approximately 0.5 mm wide in a dorso-vental direction, and their prothecal and metathecal segments are each ap proximately 0.5 mm long. The thecae overlap for at least three-fourths their length. The apertural excavations, which are delimited by the apertural processes, are 0.7 - 0.9 mm long and up to 0.5 mm wide in bi profile view, depending on the length of the apertural processes. The trace of the interthecal septum is visible on the lateral side of the rhabdosome (text-fig. 60 d) . It originates at the budding of a protheca and extends to the dorsal margin of the metathecal aperture. A wall that is continuous with the interthecal septum shows growth lines that are continuous with those on both the lateral walls of the preceding (underlying) metatheca (text-figs. 61, a-b, f-g). Such an arrangement indicates that the interthecal septum develops as the dorsal wall of a metatheca. The growth lines on the wall that is continuous with, and extends distally beyond, the interthecal septum are continuous with those on the lateral walls of the succeeding metatheca (text-fig. 61 a-b) . This wall forms the "floor" (ventral wall) of the aperture process, which also shows growth lines that are continuous with those of the succeeding metatheca (text-figs. 61 c-g). Such an arrangement indicates that the apertural process, including its ventral wall, develops as a ventral structure of the immediately succeeding metathecal aperture. A few 425 growth lines are present between the ventral margin of the aperture and its associated apertural process (text-figs. 61 d-e). These growth lines, which are continuous with those on the lateral walls of the metatheca, represent a short, free ventral wall or a "supragenicular wall". The growth line evidence indicates that a metatheca initially develops without a ventral wall but with a dorsal wall, which will be come the overlying interthecal septum, and two lateral walls. The dorsal wall of the preceding metatheca serves as the ventral wall. As it approaches the dorsal apertural margin of the preceding thecal aperture, the metatheca begins to develop its own ventral wall. The growth lines on this ventral wall are continuous with those on the lateral walls of the developing metatheca; however, they are far in advance of corresponding growth lines on the metatheca (text-fig. 61 a). The ventral wall is pointed and its development is similar to that of the virgella. After the ventral wall is complete, a few additional fuselli are added to the metatheca. These fuselli build up the thick ness of the ventral wall to form a blunt apertural process and construct a short free ventral wall immediately above, and proximal to the distal tip of, the ventral apertural process (text-figs. 61 e-g). As a final stage in the development of the metatheca, dark, possibly cortical, tissue fills the triangular area within the ventral margin of the thecal aperture and coats the ventral apertural process (text-figs. 60 a, d-f). 426 Growth lines are much more closely spaced on the protheca than on the metatheca (text-fig. 60 d) . Those on the distal part of one protheca show the same orientation as those on the proximal part of the succeed ing metatheca. Because of this orientation and the fact that succeeding prothecae are aligned end-to-end, it is difficult to distinguish the boundary between successive prothecae. The growth lines of the metatheca are at an angle to those of the protheca. None of the available specimens show the relationship of the initial growth lines on the metatheca to the growth lines on the protheca. However, the thecal ontogeny can be summarized as follows: 1) The th n protheca grows distally and slowly relative to the adjacent th n-1 metatheca as revealed by the density of growth lines. It does not develop its own ventral wall; the dorsal wall of the th n-1 metatheca (th n - th n-1 inter thecal septum) serves as its ventral wall. 2) The th n metatheca develops simultaneously with the th n+l protheca. As the th n+l protheca buds and grows distally from the th n protheca, the th n metatheca develops and grows in a ventro-distal direction from the th n protheca. The rate of growth of th n also abruptly increases as suggested by the abrupt change to widely spaced growth lines. 3) The th n methatheca employs the dorsal wall of the th n-1 metatheca as its ventral wall. The th n metatheca develops its own lateral and dorsal walls. 427 4) The th n metatheca develops its oim ventral wall when its lateral walls are at, or are close to, the dorsal margin of the th n-1 aperture. This ventral wall develops in advance of the lateral walls and as a distally tapering process. 5) After the ventral wall of th n has developed out and over the th n-1 aperture, it forms a subhorizontal to downward- directed apertural process. Continual growth of the th n metatheca involves the formation of a short free ventral wall that rests on the apertural process at a point proximal to its distal end. The fuselli that compose the free ventral wall are continuous with those in the lateral walls of the metathecal apertures and represent the completion of the metathecal aperture. Non-isolated Specimens. The largest available specimen is 18.9 mm long and the virgula with a two- (or possibly three-) vaned structure that is continuous with the virgula extends an additional 10.3 mm beyond the distal end of the rhabdosome (text-fig. 63 g) . The avail able specimens present a variety of aspects that depend on their orientation and relief. In most of the available specimens the course of the virgula can be traced through the rhabdosome. This fact, to gether with the lack of dark, glossy periderm, suggests that the periderm is thin. In specimens that present a biprofile aspect, the rhabdosome has a uniform width and shows apertural processes along both its ventral margins (text-figs 63 e-g). The proximal end is characterized by two spines that are oriented directly downward. The longer of the two spines, which may be up to 1.5 mm long, represents the virgella and 428 the apertural process of th 1^. Distally, the ventral apertural pro cesses may extend up to 0.5 mm outward from the ventral margins of the rhabdosome. Proximally, the apertural processes extend directly downward. Distally, they may attain a more subhorizontal orientation. In lateral view, the proximal end of the rhabdosome, excluding spines, is broad and rounded. The width of the rhabdosome gradually increases to a maximum of 1.3 - 2.0 mm, including apertural processes, and 0.6 - 1.2 mm, excluding apertural processes, at a distance ofabout 2 mmfrom the proximal end.The maximum width occurs at the level where the thecae are horizontal (text-fig. 63 e). Distal to this level the rhabdosome narrows gradually or is of uniform width. Fully compressed specimens are significantly wider than specimens that retain some relief. Where they are preserved, the interthecal septa (or their traces) show that the metatheca of the first three thecal pairs are perpendicular to the axis of the rhabdosome, and more distally the metathecae are inclined at 50 - 60 degree angles to the rhabdosome (text-fig. 63 e) . As measured from the ventral process of th 2, the thecae number 5.5 - 8 in 5 mm and 10 - 12 in 10 mm. Distally, the thecae number 5-5.5 in 5 mm. The scalariform aspect (text-figs. 63 b-d) is very different from the biprofile aspect. The proximal end shows a wide, distally- tapering, downward-directed spine, which represents the superimposed virgella and apertural processes of th 1^ and th 1^. In well-preserved specimens (text-fig. 63 d), the overlap of the virgella and the apertural processes can be discerned. The proximal end is also furnished 429 with paired, subhorizontal, lateral spines, which are generally 1.0 mm long but may rarely be as long as 2.7 mm. Distally, the rhabdosome is parallel-sided and 0.7 - 1.7 mm wide, depending on the relief of the specimen. In fully-compressed specimens (text-fig. 63 b), a single row of thecal apertures is present in the center of the rhabdosome. The ventral margin of the thecal aperture extends downward and obscures the dorsal margin of the preceding thecal aperture. Specimens that retain relief often present a view of a plane within the rhabdosome (text-fig. 63 d). These specimens show sections through both thecal series, as well as the outline of the sicula. Additional aspects among the available specimens are those that are intermediate between the scalariform and biprofile views (text-fig. 63 a) and those that represent early growth stages (fig. 63 c). REMARKS Bulman (1944) describes isolated uncompressed specimens of Cryptograptus tricornis (Carruthers) that resemble the specimens described above in many general characters such as, 1) the sicular shape and ornamentation; 2) the thecal shape, orientation, and arrange ment; and 3) the growth line pattern. However, Bulman's specimens differ from those described above in several notable characters. These differences are: 1) The proximal-end development in which th 1^ is the dicalycal theca and th 1^ has a left-hand origin. 2) A slightly different structure of the downward-directed sicular ornamentation. 430 3) The horizontal direction of growth of the distal parts of th 1^ and th 1^. 4) The difference in growth-line patterns between the lateral metathecal walls. 5) The absence of ventral processes on the thecal apertures. Bulman's (1944) interpretation of the left-hand origin of th 1^ is based on "Traces of what may represent growth-lines on the crossing canal ... on very few specimens ...". However, these growth lines, as they are shown in Bulman's text-figure 15, are oriented exactly like those on the crossing canal in the specimens described above. The high position on the sicula of the initial part of th 1^ in Bulman's speci mens corresponds to the high position of the initial part of th 1^ in the Alabama specimens. Additionally, th 1^ in Bulman's specimens originates high on the virgellar side of the sicula (see following discussion on the structure of the sicula in Bulman's specimens). In the specimens described above th 1^ originates in a similar location. Thus, it seems most likely that the thecal designations of Bulman's (1944) figured specimens should be reversed, and such an arrangement requires th 1^ to have a right-hand origin from th 1^ (text-fig. 61 h) . Bulman's (1944) specimens differ from the Alabama specimens in the definite dicalycal nature of th 1^, but the initial spiral growth that Bulman describes for th 2^ and th 2^ is also present in the Alabama specimens. Many of the siculae described in Bulman's specimens are represent ed only by a list structure. However, several of his specimens (PI. 2, 431 figs. 1-3) show pieces of periderm attached to the list structure, which suggests that the periderm has degenerated and/or'been destroyed. This interpretation is similar to that for the Alabama specimens. Although the list structure of the Alabama specimens differs slightly from that of Bulman's specimens, the two can be homologized. In the Alabama specimens, the margins of the sicular aperture and the virgella are parts of the list structure. If the same is true for Bulman's specimens, then 1) his transverse cross-bar (p. 31) can only represent the anti-virgellar margin of the sicular aperture, 2) the Y-shaped sicular spine on the opposite side of the sicular aperture must represent the virgella, and 3) th 1^ as re-interpreted above (Bulman's th 1^) originates on the virgellar side of the sicula. Because the apertures of th 1^ and th 1^ are not directed downward as in the Alabama specimens, the list structure of Bulman's specimens lacks the downward—directed apertural processes of those thecae. However, Bulman's specimens do show a Y-shaped sicular spine on the anti-virgellar margin of the sicula. This spine may be analogous to the th 1^ apertural process in the Alabama specimens. The arrangment of the lateral sicular spines and the longitudinal sicular rods in Bulman's specimens agrees closely with that in the Alabama specimens and further supports the interpretations of homology and analogy. A major difference between Bulman's specimens and the specimens described above is the orientation of the first pair of thecae. In Bulman's specimens, the first two thecae grow down the sicula, then rotate in a clockwise direction around the sicula in such a way that 432 th (as re-interpreted above) opens in a horizontal direction on the anti-virgellar side on the sicula (Text-fig. 61h; Bulman, 1944, Text- fig. 15). All the thecae in a particular thecal series open on the same side of the sicula in Bulman's specimens and in the Alabama speci mens; however, the first thecal series in Bulman's specimens open on the anti-virgellar side of the sicula, whereas the Alabama specimens it opens on the virgellar side of the sicula. Bulman (1944; p. 35) reports a metathecal growth-line pattern in which the growth lines on the free ventral wall are discontinuous with those of the inner metathecal wall. Instead, he suggests that they are continuous with the underlying interthecal septum. Such an inter pretation is difficult to visualize and conflicts with his interpretation (p. 35) that the growth lines of the inner thecal wall are continuous with the overlying median septum, which is a term that Bulman may have mistakenly used. Several of Bulman's specimens (PI. 2, figs. 4-6) show the same growth-line pattern on both the inner and outer metathecal walls. This pattern is also present on the interthecal septum in his illustrations, and I can only interpret this growth-line pattern to be similar to that in the specimens described above. As a result, Bulman's specimens as well as those described above do not have a "diplograptid" type of median septum. Instead, the inner thecal walls of the two thecal series are in lateral contact and form a double wall that encloses the virgule, in a similar manner as shown for Glossograptus by Jaanusson (1960, text-fig. 5). 433 Except for the absence of the ventral apertural processes, the thecae of Bulman's specimens are remarkably similar in appearance, even in the details of growth lines, to those of the Alabama specimens. The difference between the thecae involves basically a blunt, spine-shaped, downward-directed structure that is composed of fusellar tissue. It is consistently found on all thecae of all the Alabama specimens, and its absence from all the thecae of Bulman's specimens indicates that it is of diagnostic value. In the isolated state, Bulman's specimens and the Alabama specimens are easy to differentiate on the basis of 1) the downward- directed processes of the proximal end, 2) the orientation of the first two thecae, and 3) the presence or absence of ventral apertural proces ses. However, as compressed specimens on shale surfaces, the two forms are somewhat difficult to differentiate because their quantitative features (stipe width, thecal density) are similar. It is likely that a scalariform aspect of Bulman's specimens would show a rhabdosome of uniform width with a single downward-directed sicular process; paired, lateral spines; and a single row of thecal apertures. This appearance resembles closely that of the Alabama specimens; however, well-preserved specimens can show the nature of the dorsal margin of the thecal aperture (whether it is obscured by the margin of the succeeding thecal aperture as in the Alabama specimens) and possibly the details of the proximal end (whether the first thecae open downward and whether a th 1 apertural process is superimposed on the virgella as in the Alabama specimens. The biprofile aspects are easily differentiated because of 434 the doxmward-diracted apertural processes that are a prominent feature of the Alabama specimens. Except for Bulman's (1944) description of Cryptograptus tricornis, all previous descriptions and illustrations of Cryptograptus are based on generally poorly preserved specimens compressed on shale surfaces. As a result, many authors assign specimens that resemble closely those described above to the species C^. tricornis or its subspecies schaeferi. The first instance of this involves Hall's (1859) Graptolithus marcidus, which shows in biprofile view downward-directed apertural processes and which shows in scalariform view the dorsal margin of each thecal aperture obscured by a wide process that projects from the ventral margin of the succeeding thecal aperture. Lapworth (1880, p. 174) interpreted Hall's (1859) specimens of Graptolithus marcidus as a préservâtional aspect of Cryptograptus tricornis (Carruthers). Lapworth (1880; PI. 5, fig. 28) illustrated, in biprofile view, specimens that resemble Hall's (1859) specimens of G^. marcidus by their downward-directed apertural processes, and although he neither describes these specimens nor includes them in his descrip tion of JC. tricornis (Carruthers), he refers to them in his plate explanation as C^. tricornis var. Schaeferi. Elies and Wood (1908) first described the subspecies schaeferi. Their illustrations of schaeferi show only biprofile aspects, in which the rhabdosome has downward-directed apertural processes, but they synonymize Hall's G^. marcidus with Cryptograptus tricornis (Carruthers). Ruedemann (1908, 1947) also assigns Hall's specimens of G^. marcidus to C^. tricornis 435 (Carruthers), and he (1947) describes and illustrates a specimen as C^. tricornis schaferi that differs greatly in rhabdosomal shape from C^. tricornis schaeferi Lapworth. Skevington (1970) describes specimens that agree closely with the Alabama specimens. He synonymizes them with many previously described specimens of Cryptograptus tricornis schaeferi and realizes that records of C^. tricornis schaeferi are based on specimens preserved only in bi profile view. This recognition of the differences in notable; however, Skevington (1970) considers the differences to be of subspecific value, whereas the knowledge of isolated specimens described above suggests that the differences are of specific value. Additionally, all the described specimens of schaeferi (Bulman, 1931, 1933; Harris and Thomas, 1935; Lemon and Cranswick, 1956; Berry, 1964, and Ross and Berry, 1963), as well as the Alabama specimens, resemble closely Hall's (1859) Graptolithus marcidus, and because of priority they are designated here in as Cryptograptus marcidus (Hall). Many previously described specimens of Cryptograptus tricornis (Carruthers) are based on specimens that are preserved only in scalariform view and occur with specimens that show a marcidus type of biprofile view. These are here assigned to C^. marcidus (Hall) . They include specimens described by Hadding (1913), Ekstrom (1937), Decker (1952), Berry (1960 and 1964) and Ross and Berry (1963). As previously report ed, well-preserved specimens of C^. marcidus that present a scalariform aspect can be distinguished from C^. tricornis by the overlap of the dorsal thecal margin by the ventral process of the succeeding thecal 4 3 6 aperture. Hadding’s (1913); PI. 2, figs. 13, 14) figured specimens of C^. tricornis serve as an example of this. Hadding's (1913, 1915 a) figured specimens of Cryptograptus lanceolatus agree closely with the specimens described above and are here considered to be nonspecific with £. marcidus (Hall). Berry's (1960; PI. 12, figs. 7, 8) figured specimen (YPM 20332) of £. schaferi differs greatly from C^. marcidus, especially in the shape of its proximal end, and it may not even be a cryptograptid. Figured Specimens' OSU 33173-OSU 33198. 437 Family CORYNOIDIDAE Bulman, 1944 Diagnosis As in Bulman (1970). Discussion Since the time that Hopkinson and Lapworth (1875) assigned Nicholson's (1867) genus Corynoides to a separate section, Corynoidea, within their suborder Rhabdophora, the systematic position of the Corynoididae, consisting of the genera Corynoides Nicholson and Corynites Kozowski, within the Graptoloidea has been uncertain and generally regarded as isolated. On the basis of morphological similarities between Corynoides and the proximal end of Isograptus gibberulus, Bulman (1944-47) and Strachan (1949) suggested that Corynoides arose from an isograptid ancestor by means of arrested development at an early astogenetic stage. Kozïowski (1953) has speculated, primarily on the structure of the prosicula, a dendroid affinity. Although Jaanusson (1960) regarded the general characters of Corynoididae to be of a graptoloid nature, he erected a separate suborder, Corynoidina, for the Corynoididae because of their sharp morphological discontinuity with the other graptoloids. After reviewing earlier discussions on the systematic position of the Corynoididae, discussing its characteristic features, and describing well-preserved isolated specimens of Maeandrograptus geniculatus, Skevington (1965) favored a "derivation of the Corynoididae from a true graptoloid such as M. ?geniculatus," and found "insufficient 438 evidence for establishing a taxon of higher rank to accomodate this one family." However, he conceded that the Corynoididae could not have arisen directly from M. geniculatus because a considerable stratigraphie gap separates the two taxa. In 1970, Bulman (p. V105) re-interpreted the proximal-end development as being of the isograptid type and suggested strongly an isograptid ancestry for the Corynoididae, while conceding that a considerable time gap separates the last isograptids and earliest species of Corynoides. New information on the astogeny and ontogeny of Glossograptus, which is presented herein on the basis of early growth stages of G^. cilia tus, suggests that the Corynoididae did evolve from Isograptus via Maeandrograptus, Pseudisograptus, Apiograptus, and Glossograptus. The morphologic, astogenetic, and ontogenetic similarities between early growth stages of Glossograptus and certain corynoidid species are striking. In fact, Corynoides tricornis Ruedemann is shown herein to be an early growth stage of G^. ciliatus. Bulman (1944, 1947) described Corynoides cf. cur tus and C^. cf. gracilis on the basis of isolated specimens. As illustrated by Bulman, the rhabdosome of C^. cf. curtus consists of the sicula, th 1, and a short th 2, whereas in C^. cf. gracilis the rhabdosome has two long thecae and a short third theca. The similarities of these two corynoidid species to early growth stages of G^. ciliatus, the only species of Glossograptus for which early growth stages are known, are listed below in the order of rhabdosome astogeny and thecal ontogeny. 1. The prosicula is very small, accounting for only a small fraction of the length of the sicula. 439 2. The metasicula is composed of transverse growth lines. Its virgella is formed by the downward curvature of the growth lines along the zig-zag suture on one side of the sicula, and this curvature, which originates high on the sicula, increases distally. Bulman (1947, p. 73) notes that in distal portions of the sicula, an individual growth line becomes discontinuous by developing sicular and virgellar portions that pinch out laterally. Although the Alabama specimens of G^. ciliatus are not preserved well enough to show this feature, growth lines on Alabama specimens of Cryptograptus marcidus, which probably shares a common ancestry with Glossograptus, show this feature. 3. The virgella, especially at maturity, is a lamellariform process with its lateral margins turned upward forming a wide U-shaped cross-section. At maturity, the lateral edges of the virgella are dark black, which suggests the presence of cortical deposits. 4. Th 1^ originates from a foramen in the right-lateral wall of the sicula and develops by the addition of U-shaped fuselli, that is, without its o \ m ventral wall. Early in its development, th 1^ grows around to (in a clockwise direction when the rhabdosome is viewed proximally) then down and in contact with the virgellar wall of the sicula. 5. As th 1 approaches, but before it reaches, the level of the sicular aperture, a process is produced from the base of the sicular virgella. This process is similar in appearance to 440 the sicular virgella; it diverges from the sicular virgella; and it eventually serves as a ventral apertural process of th 1^. 6. After th 1^ grows around to the virgellar side of the sicula and before it grows halfway down the sicula, th 1^ (Bulman's th 2) originates right-handedly from th 1^. The origin of th 1^ is by means of a primary notch. Bulman’s (1947, p. 74) original interpretation of a resorption origin of th 1^ was later considered to be a misinterpretation (Bulman, 1970; p. V105). 2 7. Th 1 grows down the anti-virgellar side of the sicula and develops its aperture on the opposite side of the sicula from 19 1 th I-*-. Th 1 , as with th 1 , is provided with a ventral apertural process that resembles closely in appearance the sicular virgella. 8. The third theca in Corynoides cf. gracilis originates left- 2 handedly from, and high on, th 1 . This theca, which is Bulman's (1947) th 3, is extremely short and probably homologous to th 2^ in G^. ciliatus. Bulman (1944) also described isolated specimens of Corynoides cf. calicularis, the rhabdosome of which consists of three theca. Only a small number of specimens were available for study, but they seem to bestructurally similar to C^. cf. gracilis. The only significant difference between early growth stages of G^. ciliatus and Bulman's specimens of Corynoides is the presence of paired lateral apertural spines on the sicula and proximal thecae of 441 ciliatus. The distribution of these spines among the different species of Glossograptus is variable; they are composed of microfusellar, or possibly cortical, tissue; therefore, they do not seem to represent a significant structural difference between Corynoides and Glossograptus. On the basis of the similarities listed above, Corynoides shows close affinities to, and could have easily evolved directly from, Glossograptus by means of arrested development at an early astogenetic stage. Corynites, species of which Kozowski (1953, 1956) has described on the basis of isolated specimens, is characterized by very elaborate sicular apertural flanges, which argue against a direct glossograptid ancestry of the genus. However, Corynites has the same basic structural plan as Corynoides, and it is likely that Corynites evolved from Corynoides by an elaboration of the sicular apertural process. Suborder DIPLOGRAPTINA Lapworth, 1880 emend. Bulman, 1970 Diagnosis As in Bulman (1970). Family DIPLOGRAPTIDAE Lapworth, 1873 Diagnosis As in Bulman (1970). Discussion The generic classification for the diplograptids used in the present study is that of Bulman (1970). The diagnostic characters of the genera are based on thecal morphology, and as Bulman (1970) admits, it is likely that most diplograptid genera are polyphyletic. Two instances of polyphyly are documented in the species described below. The rhabdosomal development, which includes the development of the proximal end and median septum and the thecal ontogeny, of Climacograptus meridionalis resembles closely that of Pseudoclimacograptus modestus. and the rhabdosomal development of Glyptograptus euglyphus resembles closely that of Climacograptus b revis. The differences between Climacograptus meridionalis and Pseudoclimacograptus modestus are in the shapes of the supragenicular walls and the median septum. These characters are the diagnostic generic features employed by Pribyl (1947), Jaanusson (1960), and Bulman (1970). Yet, the great similarities in all other morphological features suggest that meridionalis and P. modestus are very closely related. 4 4 2 443 Glyptograptus euglyphus differs from Climacograptus brevis by the restriction of geniculate thecae to the proximal end of the rhabdosome. These geniculate thecae are present throughout the rhabdosome of C^. brevis. Additionally, the rhabdosome of C^. brevis is much shorter than the rhabdosome of G^. euglyphus. As discussed below, jC. brevis could have been easily derived from G^. euglyphus by two biological processes, namely a change in the morphophysiological gradient and arrested development. In spite of the strong evidence of polyphyly, the generic classifi cation of Bulman (1970) is used in the present study because the available material is too limited to examine in detail the taxonomic relationships within the Diplograptidae. 444 Genus Climacograptus Hall, 1865 Type species; Climacograptus bicornis (Hall, 1847) Diagnosis As in Bulman (1970). Climacograptus meridionalis Ruedemann, 1947 (Text-figures 64-67) 1932 Climaco grap tus sp., Bulman (partim), p. 16-17, PI. 3, figs. 24-25, 28-32 (non PI. 3, figs. 21-23, 26-27). 1947 Climacograptus modestus Ruedemann var. meridionalis n. var. Ruedemann, p. 433, PI. 73, figs. 47-48, 1952 Climacograptus modestus var. meridionalis Ruedemann, Decker, PI. 1, fig. 9; PI. 2, fig. 20. 21960 Climcograptus modestus var. meridionalis Ruedemann, Berry, p. 81. 1963 Climacograptus skagensis n. sp., Jaanusson and Skoglund, p. 355-356, fig, 5D, E. Type Data The holotype is the specimen illustrated by Ruedemann (1947; PI. 73, figs. 47-48), and it is stored at the United States National Museum, Washington, D. C. (specimen number USNM 102757). The holotype was collected from the Athens Shale at the Clifton Park section, Bristol, Tennessee (USNM locality 321 L9 Kirk and Messier) . The holotype was examined by the writer and found to agree with Ruedemann's (1947) description. Diagnosis A species of Climacograptus with 6.5 - 7.5 thecae in 5 mm proximally and 4.5 - 5.5 thecae in 5 mm distally; rhabdosome width 0.70 - 0.95 mm (depending on degree of compression) at level of th 2, increasing distally 444 Genus Climacograptus Hall, 1865 Type species: Climacograptus bicornis (Hall, 1847) Diagnosis As in Bulman (1970). Climacograptus meridionalis Ruedemann, 1947 (Text-figures 64-67) 1932 Climacograptus sp., Bulman (partim), p. 16-17, PI. 3, figs. 24-25, 28-32 (non PI. 3, figs. 21-23, 26-27). 1947 Climacograptus modestus Ruedemann var. meridionalis n. var. Ruedemann, p. 433, PI. 73, figs. 47-48, 1952 Climacograptus modestus var. meridionalis Ruedemann, Decker, PI. 1, fig. 9; PI. 2, fig. 20. ?1960 Climcograptus modestus var. meridionalis Ruedemann, Berry, p. 81. 1963 Climacograptus skagensis n. sp., Jaanusson and Skoglund, p. 355-356, fig. 5D, E. Type Data The holotype is the specimen illustrated by Ruedemann (1947; PI. 73, figs. 47-48), and it is stored at the United States National Museum, Washington, D. C. (specimen number USNl-I 102757). The holotype was collected from the Athens Shale at the Clifton Park section, Bristol, Tennessee (USNM locality 321 L9 Kirk and Messier) . The holotype was examined by the writer and found to agree with Ruedemann' s (1947) description. Diagnosis A species of Climacograptus with 6.5 - 7.5 thecae in 5 mm proximally and 4.5 - 5.5 thecae in 5 mm distally; rhabdosome width 0.70 - 0.95 mm (depending on degree of compression) at level of th 2, increasing distally 445 EXPLANATION OF TEXT-FIGURE 64 Text-figure 64 a-j. Climacograptus meridionalis Ruedemann. a. Non-isolated, compressed specimen. X9. PF-17. OSU 33211. b,c. Ventral and lateral aspects of isolated specimen showing structure of median septum. Transverse rods attached to lateral walls of rhabdosome at points marked by dotted circles. X37.5. PS-102. OSU 33212. d. Obverse aspect of isolated, compressed specimen. X18.8. PS-109.2. OSU 33213. e. Reverse aspect of isolated specimen preserved in full relief. Periderm of median septum absent. Note arrangement of virgula, transverse rod, connecting list, and interthecal septum. X18.8. PS-102. OSU 33214. f. Isolated specimen of proximal end with virgellar process. X18.8. PS-109.7. OSU 33215. g. Obverse aspect of non-isolated, compressed specimen. X9. PF-17.3. OSU 33216. h. Reverse aspect of isolated, partly compressed specimen. X18.8. PS-109.2. OSU 33217. i. Reverse aspect of isolated specimen preserved in full relief. X37.5. PS-102. OSU 33218. j. Lateral aspect of isolated, partly compressed specimen. X18.8. PS-109.2. OSU 33219. 446 msp tr Text-figure 64 447 EXPLANATION OF TEXT-FIGURE 65 Text-figure 65 a-m. Climacograptus meridionalis Ruedemann. Isolated specimens representing early growth stages. a,b. Reverse and obverse aspects of sicula with initial bud of th 1^ and opening for th 1^ partly formed. X37.5. PS-109.2. OSU 33220. c. Reverse aspect of sicula with th 1^ and horizontally directed th 1^. X37.5. PS-109.2. OSU 33221. d. Reverse aspect of sicula with th 1^ and downward-directed th 1^. X37.5. PS-109.2. OSU 33222. e. Reverse aspect of sicula with th 1^, th 1^, and initial bud of th 2l. X37.5. PS-109.2. OSU 33223. f. Reverse aspect of sicula with th 1 17 , th 1 , and th 2 1 . X37.5. PS-109.2. OSU 33224. g. Reverse aspect showing horizontal and upward growth of th 2^. X37.5. PS-109.2. OSU 33225. h,i. Reverse and obverse aspects showing growth of th 2^ and th 2^. Note that th 2^ does not have left-lateral wall. X37.5. PS-109.2. OSU 33226. j,k. Reverse and obverse aspects showing growth of th 2^ and th 2^. Note transverse list on th 2^, growth lines on th 2^ infragenicular wall continuous with those on right-lateral wall, and left-lateral wall present for th 2^. X37.5. PS-109.2. OSU 33227. l,m. Reverse and obverse aspects showing growth of th 2^ and th 2^. Note th 2^ lacks a right-lateral wall. X37.5. PS-109.2. OSU 33228. 00 Text-figure 65 449 EXPLANATION OF TEXT-FIGURE 66 Text-figure 66 a-h. Climacograptus meridionalis Ruedemann. Isolated specimens representing early growth stages. a,b. Obverse and reverse aspects of specimen showing growth of th 2^ and th 2 . Note that th 2^ has right-lateral wall and earliest part of inf ragenicular wall. X37.5. PS-109.2. OSU 33229. c. Obverse aspect of specimen showing that growth lines of inf ragenicular wall and right-lateral wall of th 2^ are continuous. X37.5. PS-109.2. OSU 33230. d,e. Obverse and reverse aspects of specimen showing th 2^ and th 2^. Note extent of left-lateral wall of th 2^. X37.5. PS-109.2. OSU 33231. e,f,g. Obverse, obverse (viewed distally), and reverse aspects of specimen. Note transverse rod and closely spaced fuselli indicating budding of th 3^ from th 2^. Note that th 2^ is provided with its own dorsal wall. X37.5. PS-109.2. OSU 33232. a th 2f th 2 th 2‘ Text-figure 66 451 EXPLANATION OF TEXT-FIGURE 67 Text-figure 67 a-h. Climacograptus meridionalis Ruedemann. Isolated specimens representing growth stages. a,b. Obverse and reverse aspects of specimen showing development of left-lateral wall of th 3^. X37.5. PS-102. OSU 33233. c. Obverse aspect of specimen showing growth of left-lateral and infragenicular walls of th 3^. X37.5. PS-109.2. OSU 33234. d. Obverse aspect of specimen showing growth of th 3^ and th 2^. Note dorsal wall of th 2^ metatheca. X37.5. PS-109.2. OSU 33235. e. Reverse aspect of specimen showing that transverse rod is continuous with growth lines in lateral wall. X37.5. PS-109.2. OSU 33236. f. Reverse aspect of specimen showing th 3 . X37.5. PS-109.7. OSU 33237. g,h. Reverse and obverse aspects of specimen shotting development of th 4 . X37.5. PS-109.2. OSU 33238. th 3.' A th22 -th 3 th 22 . th3' ' msp ■p- Ln Text-figure 67 ho 453 to 1.2 - 1.5 mm at tenth to fifteenth thecal pair; thereafter rhabdosome parallel-sided or slightly narrowing distally; virgella with sheath-like process in garantie? specimens. Material The available material consists of more than 550 specimens that represent the following states of preservation: 1) Isolated specimens. Approximately 130 specimens were obtained from the Pratt's Syncline section of which 80 represent early growth stages, 40 are young rhabdosomes or fragments of large rhabdosomes with proximal ends and more than two pairs of thecae, and 15 are distal fragments. Many of the early growth stages retain some degree of relief and show growth lines. Most of the larger rhabdosomes are compressed and have a glossy black surface that probably represents a thick cortical layer. Growth lines are visible on only the most distal thecae in the large rhabdosomes. 2) Non-isolated, compressed specimens with periderm preserved. Approximately 85 specimens were obtained from the Pratt's Ferry section and 20 were obtained from the Pratt's Syncline section. 3) Carbon Films. More than 300 specimens were obtained from the Calera section. Description Isolated Specimens. The largest available specimen (Text-fig. 64e) is 3.6 mm long and consists of 5 pairs of thecae. The cross-section of the rhabdosome is oval in uncompressed specimens. In profile aspect. 454 the rhabdosome is somewhat parallel-sided, and the proximal end is broad and rounded (Text-fig. 64i). The first two thecae are U-shaped, and each bears a subhorizontal, subapertural spine that attains a maximum length of 0.5 mm. The proximal end also exhibits a third prominent spine, the virgella, which rarely exceeds 1.0 mm in length except in special instances (See description below of sheath-like process). The sicula CSee description below under Early Growth Stages) is visible to the level of the th 1^ aperture on the obverse side of the sicula. The aperture of the 1^ is situated at a higher level above the sicular aperture than the aperture of th 1^. This alternation of the apertures of the opposing thecal series is repeated throughout the length of the rhabdosome. At the level of the first and second pairs of thecae, the rhabdosome is 0.7 - 0.8 mm wide in specimens preserved in full relief and 0.85 - 8.95 mm wide in fully compressed specimens. Distally, the width of the rhabdosome increases gradually to 0.9 mm and 1.2 mm at the fifth to sixth pairs of thecae in uncompressed and compressed specimens, respectively. A maximum observed width of 1.5 mm was measured on a fully compressed distal fragment. Proximally, there are 4.5 thecae in 3 mm, and distally, the thecae number 3.5 - 4 in 3 mm. The thecae are of the climacograptid type with straight supragenicular walls, distinct genicula, and relatively shallow apertural excavations. The thecae are 0.9 - 1.0 mm long in proximal parts of the rhabdosome and 1.3 mm long in distal parts of the rhabdosome. This distal increase in thecal length is accompanied by a distal increase in the length of the supragenicular wall from 0.40 mm proximally to 0.75 455 mm distally. Throughout the rhabdosome, the thecae overlap for half their length. The distal part of the infragenicular wall is perpendicular to the axis of the rhabdosome, and the supragenicular wall parallels the axis of the rhabdosome. Together these two walls define a 90 degree genicular angle in profile aspect, and the geniculum is situated directly above the ventral margin of the preceding thecal aperture. The apertural excavations, which have a semicircular shape in profile aspect and are rimmed by lists, are consistently 0.15 mm high (long.) and occupy a fifth and a sixth of the length of the free ventral wall in proximal and distal thecae, respectively. In specimens preserved in full relief (Text-figs. 64e, i), the apertural excavations are 0.10 mm wide (a sixth of the rhabdosome width) proximally and 0.20 - 0.25 mm wide (between a third and a fourth of the rhabdosome width)distally. In compressed specimens (Text-figs. 64h, j), the apertural excavations have a greater width (up to 0.45 mm) and appear to lose the semicircular appearance that characterizes apertural excavations in uncompressed specimens. However, the increase in the width of the apertural excavations in compressed specimens is associated with an increase in rhabdosome width, and the width of the apertural excavation occupies the same fraction of the rhabdosome width as it does in uncompressed specimens. In most of the compressed specimens, the rhabdosome is obliquely distorted relative to the lateral walls of the rhabdosome. As a result, the apertural excavations of one thecal series are much wider in profile aspect than those of the opposing thecal series. However, a pair of thecal apertures, representing corresponding thecae from each of the two 456 thecal series, occupies an equal fraction of the rhabdosome width (between a half and two-thirds) as a pair of thecal apertures in specimens that are distorted in a direction perpendicular to the lateral walls of the rhabdosome (Compare Text-figs. 64h and 64d). Additionally, in obliquely compressed specimens, the supragenicular walls are distorted into the plane of the lateral walls of the rhabdosome in such a way that the supragenicular walls of the thecal series that presents the wider thecal apertures are visible in profile aspect. A groove, representing the trace of the median septum, begins on both the obverse and reverse walls of the rhabdosome at points that are slightly above the level of the th 2^ aperture. The median septal groove has an undulating course until the level of the th 3^ aperture, after which its course is straight. Grooves, representing the traces of internal lists, extend perpendicularly outward from the median septal groove and merge distally with the proximal ends of longitudinally oriented grooves representing the traces of the interthecal septa. The structure of the median septum, which is described in detail below (See Development of Median Septum), consists of the virgula, paired transverse rods, and fusellar tissue. The median septum is complete, and growth lines in the fusellar tissue are obliquely oriented and converse distally toward the virgula (Text-fig. 64b, c). The paired transverse rods arise from the virgula at regular intervals and extend outward to the lateral rhabdosomal walls (Text-figs. b-c, e). Connecting lists extend transversely along the interior lateral walls of the rhabdosome from the tips of the transverse rods to the lateral margins of the proximal ends of the interthecal septa (Text-fig. 64e). 457 In a few specimens, a sheath-like process projects from the proximal- end of the rhabdosome (Text-fig. 64f). At its maximum observed size, this process is 2.2 mm long and 0.4 mm wide. The process is tube-shaped; proximally, its walls are continuous with the margins of the sicular aperture and the virgella is embedded in the wall of the process. The process has a glossy black appearance that is similar to cortical tissue. In transmitted light, faint lines, possible growth lines, are visible on the walls of the process. These lines, which number 10 in 0.10 mm, parallel the aperture of the process in the specimen illustrated in Text-figure 64f. From the virgella they extend around the process and back to the virgella. A single "growth line" begins and ends at different points along the virgella, and a plane defined by a single "growth line" is slightly oblique to a plane oriented perpendicularly to the axis of the process. The proximal end, including the spines on the first two thecae, is greatly thickened by glossy black cortical tissue in specimens with the sheath-like process. This secondary thickening of the proximal end suggests that the sheath-like process develops in gerontic rhabdosomes. Non-isolated Specimens with periderm preserved and Carbon Films. The largest available rhabdosome is 14 mm long and has a sheath-like virgellar process. The largest rhabdosome without a virgellar process is 11 mm long and consists of 14 pairs of thecae. A 1.4 mm long virgella extends from the proximal end of the rhabdosome, and a 2.8 mm long virgula projects from the distal end. The non-isolated specimens (Text-figs. 64a, g) present a profile aspect that agrees closely with that described above for compressed, isolated specimens. Because the non-isolated specimens are larger and less fragmentary than the isolated specimens, measurements 458 can be taken for greater lengths of the rhabdosome. The rhabdosome is 0.70-0.95 mm wide at the second pair of thecae. Distally, it gradually increases in width to a maximum of 1.2-1.5 mm at the tenth to fifteenth pair of thecae, after which it is parallel-sided or narrowing slightly distally. Proximally, the thecae number 3.5-4.5 in 3 mm, 6.5-7.5 in 5 mm, and 14 in 10 mm. Distally, the thecae number 4.5—5.5 in 5 mm. Early Growth Stages and Bhabdosomal Development. Morphology of Sicula. The sicula is conical and widens rapidly toward its aperture (Text-figs. 65b, c) . Its length ranges from 0.9 to 1.1 mm among the available specimens, and its width at the aperture averages 0.25 mm. The prosicula, which accounts for a third of the length of the sicula, is generally broken or consists of two or three thick longitudinal threads that extend from the metasicula and converge into the nema (virgula). The metasicula is composed of transverse growth lines, which bend downward to form the virgella along one side of the sicula. Except for the virgella, the sicular aperture has no ornamentation. Development of Proximal End. As shown in Text-figure 65b-d, the development of the initial bud of th 1 1 and the origin of th 1 2 from th 11 is very similar to that described below in detail for Dicellograptus. 1 The initial bud of th 1 originates halfway down the metasicula adjacent to, and on the right side of, the virgella. It grows only a short distance (0.10 mm) downward along the sicula before two walls, which are discontinuous with the initial bud, begin developing on the sides of the aperture of the initial bud. The left wall (relative to th 1^) grows upward over the aperture of the initial bud to meet the wall on the right 459 side in such a way that two openings are formed. Th 1 continues to develop from the downward-directed opening. It grows downward along the virgella to the level of the sicular aperture, where it bends arid grows outward and upward, thus forming a U-shaped 2 tube (Text-figs. 65b-d). The opening for the 1 is directed obliquely upward, and th 1^ initially grows upward to the reverse side of the sicula, then it bends and grows downward to the anti-virgellar margin of the sicula, where it again bends and grows outward and upward, thus forming a S-shaped tube (Text-figs. 65a-i). After th 1 has started its downward direction of growth and th 1 is complete to its aperture (Compare Text-figs. 65d and 65 e) , th 2^ develops from the dorsal wall of th 1^. Th 2^ grows downward as a broad hood (Text-fig. 65f), from which two openings develop (Text-fig. 65g), one of which represents th 2 . The available growth stages do not reveal 2 the exact manner in which th 2 forms. However, the similarities of these early growth stages to those of Pseudoclimacograptus modestus and Dicellograptus, which are described below, suggest a similar manner of o 1 formation of th 2 , which involves the growth of the th 2 hood down onto the sicula as a narrow process in such a way that two horizontally directed openings are formed, one on each side of the process. With the budding of th 2^, the direction of growth of th 2^ changes abruptly from downward to oubfard. The horizontal (transverse) growth is peculiar because it involves initially only the formation of a right-lateral wall (Text-fig. 65g). This wall is composed of fuselli that extend from the previously developed, horizontally directed aperture 460 of th 2^ to the right-handed margin of the dorsal wall of th 1^. In this way, th 2^ fills in the right-lateral side of the space between the 1 downward- and upward-directed limbs of th 1 (Text-figs. 65g-i). As th 2^ continues to develop, it bends and grows upward between the downward- and upward-directed parts of th 1^. Wlien the 2^ reaches the level of the aperture of th 1^, it still consists of only a right-lateral wall, which extends from the right-lateral margin of the dorsal wall of the upward-directed part of th 1^ to the reverse wall of the downward- directed parts of th 1 1 and th 1 2 . As th 2 1 grows above the aperture of th 1^, it begins to develop an infragenicular wall (Text-figs. 65h-k). The growth lines on this infragenicular wall are continuous with those on the right-lateral wall. At a slightly later stage in its development, but before the infragenicular wall is complete (Text-fig. 65j-k) , the O left-lateral wall of th 1 begins to develop by the addition of fuselli that fill in the left-lateral side of the U-shaped space formed by the limbs of th 1^. Simultaneously, the dorsal margin of the right-lateral wall of th 2^ grows outward away from its contact with the downward-directed parts of th 1^ and th 2^, and in doing so, a rod develops from the dorsal margin of the distal end of the right-lateral wall of th 2^. Growth lines extend from the right-lateral wall into the base of this rod, but the rod itself does not show growth lines. The rod connects the right-lateral wall of th 2^ to the upward-directed initial part of th 1^, and an opening is left below it. The growth of the left-lateral wall of th 2^ catches up with that of the ri^t-lateral wall at about the stage when the infragenicular wall is complete (Text-fig. 65 l,m). As th 2^ develops 461 past the geniculum, it develops a dorsal wall, and thus it is a complete tube (Text-figs. 66c-h). Th 2^ continues to grow to'a'level'slightly above the top of the metasicula, where it forms an aperture (Text-fig. 67a-b). The development of th 2^ is similar to that of th 2^. It grows at first horizontally and then upward along the dorsal wall of th 2^. In doing so, it only forms a left-lateral wall, which widens greatly as it grows upward (Text-figs. 65g-m). When the left-lateral wall reaches the level of the aperture of th 2^, the right-lateral and infragenicular walls begin to de-^elop (Text-fig. 66a-b) , and the growth lines on the infra genicular wall are continuous with those on the left-lateral wall. The right-lateral wall not only fills in the space between the sicula and the upward-directed part of th 1^, but also extends onto the obverse side of the sicula (Text-fig. 66a, b). In addition, the dorsal margin of the left-lateral wall of th 2^ extends across the reverse side of the sicula to the dorsal margin of the right-lateral wall of th 2^ (Text- fig. 66a, b). 2 When th 2 reaches the distal end of the infragenicular wall, its left-lateral wall still extends completely across the reverse side of the sicula to the dorsal margin of the right-lateral wall of th 2^, and it thus fills in the space between the sicula and the th 2^ metatheca (Text-fig. 66d, e). The protheca of th 3^ develops in this space, and the extent of the left-lateral wall of th 2^ indicates that th 3^ buds from th 2^. With the development of th 2^ past the infragenicular wall (Text-figs. 66f-h, 67a-e), a transverse rod appears between the sicula and the left- lateral wall of th 2^. Thin fuselli, which are closely spaced relative 462 2 to the thick fuselli of th 2 , extend from the transverse rod to the dorsal wall of the th 2^ metatheca, and these fuselli represent a right-lateral wall of the th 3^ protheca that is discontinuous with the left-lateral wall of th 2^. Thus, as with the preceding thecae, the th 3^ protheca initially develops with only one lateral wall. The th 3^ prothecae continues its development with the formation of a left-lateral wall and an infragenicular wall (Text-fig. 67a-c). The left-lateral wall extends from the left-lateral margin of the dorsal wall of the th 1^ metatheca to the dorsal margin of the right-lateral wall of th 2^ on the obverse side of the sicula. The dorsal margin of the right-lateral wall of th 2^ is in contact with the sicula and initiates the median septum on the obverse side of the sicula (Text-figs. 67a-b, d) . As th 3^ grows past the aperture of th 2^, it develops an infra genicular wall (Text-figs. 67a-b, d). Simultaneously, the manner of growth of th 2^ changes abruptly, and an opening is left for the development of o n the th 3 protheca between the sicula and the th 2 metatheca (Text-fig. 67d). The type of proximal-end development described above is the streptoblastic diplograptid type. There are four crossing canals (th 1^, th 2^, th 2^, and th 3^), and th 2^, from which both th 3^ and th 3^ 9 originate, is the dicalycal theca. Th 2 has a downward initial direction of growth. With the development of the third pair of thecae, the two thecal series are completely separated by the development of the median septum. Development of Median Septum. The initial development of the median septum occurs at an early stage in the development of the proximal end. 463 On the obverse side of the rhabdosome, the median septum begins at the point where the dorsal margins of the right-lateral wall of th 2^ and the left-lateral wall of th 3^ meet (Text-figs. 67a, b, d). Here, the median septum is formed by that portion of the right lateral wall of th 2^ that bends inward in such a way that its margin rests against the sicula, thus forming on the obverse side of the sicula a short dorsal wall for th 2^. At a later stage in the development of the proximal end, this wall serves also as the dorsal wall of the th 3^ protheca on the obverse side of the sicula. The development of the transverse rod that connects the right-lateral 1 O wall of th 2 to the initial upward-directed part of th 1 probably represents the initiation of the median septum on the reverse side of the sicula (Text-fig. 65j-k). The next stage in the development of the median septum on the reverse side is represented by the formation of a transverse rod that is associated with the budding of th 3^ (Text-fig. 12 1 66f-h). The budding of th 3 from th 2 occurs between the th 2 and th 3^ transverse rods; thus, no fusellar septal membrane connects the th 2^ and th 3^ transverse rods, and the initial part of the median septum on the reverse side of the sicula is cryptoseptate. With the development of the median septum above the th 3^ transverse bar, the median septum probably becomes a complete septum because a groove that represents the trace of the median septum appears on the external wall of the rhabdosome (Text-fig. 67g, h). Below the level of the apex of the sicula, the median septum consists of two walls, one on each side of the sicula. Distally, beyond the apex 464 of the sicula, these two walls join and become one at the base of the virgula. The median septum serves as the dorsal wall of all the prothecae of the rhabdosome, and it is composed of fusellar tissue. The growth lines on the median septum are obliquely oriented and converge distally toward the virgula. In a few instances where they are well preserved (Text-figs. 67f-h), the growth lines on the most distal part of the median septum appear to be continuous with those on the lateral walls of the most distal protheca, thus indicating that the median septum develops as, and is composed of, the dorsal walls of the prothecae. Because the prothecae of the opposing thecal series are alternating, the median septum must be composed longitudinally of segments of equal length, and 'each of these segments must be formed as the dorsal wall of the prothecae that is on the opposite side of the median septum from the prothecae that forms the immediately adjacent, preceding and succeeding segments. The transverse rods are situated at the contacts between the segments of the median septum. The transverse rods occur in pairs and connect the virgula to the obverse and reverse walls of the rhabdosome. The distal tip of each transverse rod meets the lateral wall of the rhabdosome at the dorsal comer of the most proximal margin of the lateral wall of a protheca (Text-figs. 64b-c, e). In the undulating part of the median septum, the points at which the transverse rods meet the lateral walls of the rhabdosome coincide with the apices of the undulations of the median septal groove. Thus, in undulating parts of the median septum, successive pairs of transverse rods extend in opposing, slightly oblique directions from the virgula. 465 In the part of the rhabdosome with an undulating median septum, a distal increase in the dorso-ventral width of a protheca occurs in such a way that the dorsal wall of the protheca (the median septum) grows to a position behind the virgula. When the succeeding protheca, which is in the opposing thecal series, develops, its dorsal wall, which is the next segment of the median septum, grows back across the virgula, thus producing an undulating median septum. In those parts of the rhabdosome where the median septum is straight, the initial portions of the prothecae that are responsible for the construction of the median septum do not show a distal increase in their dorso-ventral width; instead, their dorsal and ventral walls (the median septum and interthecal septum, respectively) are parallel. Thecal Ontogeny. As discussed above, the development of the th 2^, 9 1 th 2 , and th 3 prothecae is peculiar in that it begins with the formation of only one lateral wall. Specimens that show different stages in the development of more distal thecae are scarce, but in those that are available (Text-figs. 57f-h), both lateral walls of each protheca seem to develop simultaneously. The peculiar ontogeny of the most proximal prothecae is most likely related to their development adjacent to the sicula. The most proximal prothecae (th 2^ - th 3^) develop initially with only lateral walls, whereas the distal prothecae also develop a dorsal wall (the median septum). The dorsal wall of the preceding metatheca serves as the ventral wall of the prothecae. The protheca begins to develop its own ventral wall (the infragenicular wall) when it grows past the aperture of the preceding metatheca, and simultaneously the 466 density of growth lines gradually decreases (Text-fig. 67d). In the proximal part of the protheca the growth lines are closely spaced, whereas distal to the level at which the infragenicular wall initially develops, the growth lines are more widely spaced. The change in the growth-line density is not abrupt, but it does occur within a distance equal to less than half the height of the infragenicular wall. At a late stage in the development of the infragenicular wall of the th n^ protheca, the th n^ protheca of the opposing thecal series begins to develop, and the development of the median septum (dorsal wall of protheca) is taken over by the th n^ protheca. The th n® protheca then continues its development without its own dorsal wall. The dorsal wall of the th n^ protheca, which is the median septum, now serves as the dorsal wall of th n^. At a short distance past the geniculum, th (n+1)^ buds from th n^. New fuselli added to th n^ only extend dorsally halfway along the lateral walls before bending inward to form a dorsal wall for the th n^ metatheca (Text-fig. 67d). This dorsal wall is the interthecal septum between th (n+1)^ and th n^. The last fusellus of the th n protheca extends dorsally to the lateral margins of the median septum. Between the interthecal septum and the medain septum, the portion of the distal edge of the most distal fusellus of the th n^ protheca forms not only the base of the lateral wall of the th (n+1)^ protheca but also the list that connects the distal end of the transverse rod of the median septum to the lateral edge of the proximal end of the interthecal septum (Compare Text-figs. 64b-c, e and 67 d). After the budding of th (n+1)^, th n^ grows as a complete tube until its aperture is developed. The final stage of thecal ontogeny is the development of a list that rims the 467 ventral and lateral edges of the infragenicular wall and the thecal aperture. Remarks The proximal-end development of the specimens described above is of the streptoblastic diplograptid type. The budding and early growth of the first four thecae resemble closely those in the development of Dicranograptus nicholsoni Hopkinson, which is described by Bulman, (1944), and the dicellograptids of the vagus group, which are described above. An especially close similarity is the early ontogeny of th 2 and th 2 , which involves the formation on only lateral walls, but Dicranograptus and Dicellograptus differ from the specimens described above in that both the right- and left—lateral walls develop simultaneously. Addition ally, the position of the dicalycal theca and the thecal ontogeny of the described specimens differs greatly from that of the dicellograptids described above. Except for the position of the dicalycal theca, the proximal-end development of the specimens described above resembles superficially the development of Pseudoclimacograptus eurystoma Jaanusson (=Bulman's (1932) Climacograptus scharenbergi). However, the detailed proximal- end development of eurystoma is not well known, and a close comparison with the specimens described above is not possible. The structure of the zig-zag median septum in P. eurystoma is remarkably similar to the much straighter median septum in the specimens described above. The specimens described above agree closely with the holotype and the original description of Ruedemann’s (1947) Climacograptus modestus var. medionalis. Specimens that are described below and referred to as 468 Pseudoclimacograptus modestus (Ruedemann) show a rhabdosomal development (proximal-end development, median septum development, and thecal ontogeny) that resembles closely the rhabdosomal development of the specimens described above. However, the specimens described above and the specimens assigned below to 2" modestus differ by the shapes of the median septum and supragenicular walls. These differences are considered to be of specific rank, and the specimens described above, as well as the holotype of 2* modestus var. meridionalis, are referred to here as Climacograptus meridional is Ruedemann. The generic assignment is based on the shape of the supragenicular wall (See discussion above under Diplograptidae). Several, but not all, of the specimens described by Bulman (1932) as Climacograptus sp., together with some more recently isolated specimens were assigned by Jaanusson and Skoglund (1963) to a new species, Climacograptus skagensis. These specimens are preserved in full relief and represent early growth stages and short proximal-end fragments. The qualitative and quantitative morphological features of these specimens and the uncompressed specimens described above are extremely similar, and C, skagensis Jaanusson and Skoglund and £. meridionalis Ruedemann are considered to be conspecific herein. In profile aspect, 2* meridionalis can be distinguished from other superficially similar climacograptids such as jC* antiquus Lapworth, 2* caudatus Lapworth, and 2* parvus Hall by its relatively broad proximal end, parallel ventral margins, small width, high thecal density, and shallow apertural excavations. Figured Specimens OSU 33211 - OSU 33238 469 Genus Pseudoclimacograptus Pribyl, 1947 Type Species: Pseudoclimacograptus scharenbergi(Lapworth 1947) Diagnosis As in Bulman(1970). Species Described Herein Pseudoclimacograptus angulatus angulatus (Bulman, 1953) modestus (Ruedemann, 1908) ]P. sp. cf. jP. eurystoma Jaanusson, 1960 Pseudoclimacograptus angulatus angulatus(Bulman, 1953) (Text-figures 68e-f, i-j) 1953 Climacograptus scharenbergi var. angulatus nov., Bulman, p. 511- 512, text-fig. 2; PI. 1, fig. 8. 1964 Climacograptus angulatus Bulman, Berry, p. 125-128, PI. 13, Fig.4. Type Data The holotype is stored in the Holm collection, specimen number 2547, at the Swedish Museum of Natural History, Stockholm. Although the holotype has not been examined by the writer, Bulman's (1953) description is based on this single specimen, and Jaanusson(1960, p. 331), having examined the holotype, emended Bulman's (1953) descrip tion by reporting a wider proximal end for the holotype. 470 EXPLANATION OF TEXT-FIGURE 68 Text-figure 68 a-c. Pseudoclimacograptus sp. cf. 2» eurystoma Jaanusson. Non-isolated, compressed specimens. a. Reverse aspect. X15. C-27.5. OSU 33239. b. Profile aspect of distal stipe fragment. X15. C-6.6. OSU 33240. c. Reverse aspect. X15. C-4.6. OSU 33241. 68 d,g,h. Orthograptus sp. Non-isolated, compressed specimens. d. Profile aspect. X3.6. PS-132. OSU 33242. g. Profile aspect. X7.2. PS-132. OSU 33243. h. Profile aspect. X15. PS-132. OSU 33244. 68 e,f,i,j. Pseudoclimacograptus angulatus angulatus Bulman. Non-isolated, compressed specimens. e. Obverse aspect. X3.6. C-5.4. OSU 33245. f. Obverse aspect. X7.2. C-33. OSU 33246. i. Obverse aspect. X15. C-34. OSU 33247. j. Obverse aspect. X15. C-39. OSU 33248. 471 h ) Text-figure 68 472 Diagnosis The nominal subspecies of Pseudoclimacograptus angulatus with 4.5-5 thecae in 3 mm proximally and 10 thecae in 10 mm distally; rhabdosome width 0.70-0.85 mm proximally, maximum width 1.2-1.6 mm. Material More than 1000 specimens, which are preserved as carbon films, were obtained from the lower two-third of the Calera section. Description The largest available rhabdosome is 3.1 cm long, but most of the rhabdosomes are between 1.5 and 2.5 cm long.The proximal end is wide and rounded in profile view (Text-fig. 2). The first two thecae are U-shaped; the distal parts of their ventral walls are parallel to the rhabdosomal axis; and they bear short (0.2-0.5 mm long), hori zontal, subapertural spines. A stout virgella extends as much as 6.4 mm beyond the sicular aperture. In profile aspect, the rhabdosome is relatively parallel-sided, but it does widen distally from a width of 0.70-0.85 mm at th 2. In most specimens, the rhabdosome width increases to a maximum of 1.2—1.4 mm at the seventh to eighth pairs of thecae, and more distally this width is maintained or decreases slightly to 1.1 mm at the most distal pair of thecae. In those rhabdosomes with the greatest length, the width increases to a maximum of 1.4-1.6 mm at the tenth pair of thecae, after which it decreases slightly or is uniform. Beginning with th 2, the thecae number 4.5-5 in 3 mm and 7-7.5 in 5 mm. Distally, there are 5-5.5 thecae in 5 mm and 10-10.5 thecae 473 in 10 mm. The thecae are of the climacograptid type with inclined apertural excavations and convex supragenicular walls. Because of the introverted thecal aperture and the proximally inclined, distally horizontal infragenicular wall, the apertural excavation is comma- shaped. In specimens that are compressed in a direction normal to the lateral walls, the apertural excavations of the opposing thecal series are of equal size. They are 0.4 mm wide (trans.) and occupy a third of the rhabdosome width. In obliquely compressed specimens, the apertural excavations of corresponding thecal apertures of the opposing thecal series are of unequal width (e.g. a half and a sixth of the rhabdosome width); however together, the corresponding apertural excavations occupy two—third of the width of the rhabdosome, as in normally compressed specimens. The apertural excavations are 0.1-0.2 ram high (long.), occupying a fourth to a sixth of the length of the supragenicular wall, and they are rimmed by lists. The supra genicular wall is usually slightly convex; however in many instances probably due to preservation and distortion, it is straight. The supragenicular wall is 0.6-0.9 mm long in distal parts of the rhab dosome. It is 0.4 mm long at th 2. The thecae, which range in length from 1.0 to 2.0 mm between proximal and distal parts of the rhabdo some, overlap for half their length. Because the specimens are poorly preserved, the trace of the median septum on the lateral rhabdosome wall can only be followed for a short distance in the proximal part of a few specimens (Text- fig. 68i). Where it is observed, the trace of the median septum 474 has a zig-zag course. Remarks The described specimens, especially the shorter, narrower ones, agree closely with the holotype of Pseudoclimacograptus angulatus angulatus (Bulman). The original description of the nominal subspecies of P. angulatus is based on a single specimen, and thus it does not provide any information on variation within the taxum. Besides describing two new subspecies of P. angulatus, Berry(1964) assigned many of his specimens to the nominal subspecies of 2» angulatus. Al though Berry's specimens were not from the type locality of 2» angulatus angulatus (Nordal Brunsgatan, Oslo), they were from closely surrounding localities (Oslo region), and many of them were from the type horizon (4au2)• The specimens that Berry assigned to 2* angulatus angulatus show a range of variation that is comparable to that of the Alabama specimens described above. The Alabama specimens differ from Berry's (1964) subspecies magnus and marcidus in rhabdosome width and from Jaanusson's (1960) subspecies sebyensis in thecal density. Figured Specimens OSU 33245-OSU 33248. 475 Pseudoclimacograptus modestus (Ruedemann, 1908) (Text-figures 69-71) 1908 Climacograptus modestus sp. nov., Ruedemann, p. 432-433, text-figs. 400-403; Pl. 28, fig. 30. 1908 Climacograptus putillus(Hall) mut eximius nov., Ruedemann (partim), p. 420, text-figs. 378, 379, 380? (non text-figs. 381-384, PI 28, fig. 16). 1931 Climacograptus cf. modestus Ruedemann, Bulman, p. 51, text-fig. 22, Pl. 5, fig. 12?, 13. 1947 Climacograptus modestus Ruedemann, Ruedemann, p. 432, Pl. 73, figs. 32-46. 1947 Climacograptus eximius Ruedemann, Ruedemann(partim) , p. 435, Pl. 72, figs. 2, 3 (non figs. 1, 4-15). 1948 Climacograptus modestus Ruedemann, Bulman, p. 222-223, text-fig. la,b. 71952 Climacograptus modestus Ruedemann, Decker, text-fig. 32, Pl. 1, fig. 8; Pl. 2, fig. 19; Pl. 3A, fig. 6. ?1960 Climacograptus scharenbergi cf. var. stencstoma Bulman, Berry, p. 83, Pl. 15, fig. 6. 1974 Pseudoclimacograptus modestus (Ruedemann), Riva, p. 24-26, text-fig. 8a, b. ?1974 Pseudoclimacograptus scharenbergi stenostoma (Bulman), Riva, p. 26-27, text-fig. 8i. Type Data In his original description of Pseudoclimacograptus modestus, Ruedemann(1908) illustrated four specimens in text-figures and a 476 EXPLANATION OF TEXT-FIGURE 69 Text-figure 69 a-p. Pseudoclimacograptus modestus (Ruedemann). a-n. Isolated specimens, o-p. Non-isolated specimens. a. Reverse aspect of fragment from proximal end of gerontic specimen. Note cortical deposits on thecal spines and virgella. X15. PS-102. OSU 33249. b. Obverse aspect of gerontic specimen. X15. PS-102. OSU 33250. c,d. Ventral and lateral aspects of specimen preserved in full relief. Note thecal apertures. X15. PS-102. OSU 33251. e,f. Obverse and reverse aspects of compressed specimen. Note spines on th 2^ and shape of thecal apertures. X15. PS-76.4. OSU 33252. g. Reverse aspect of compressed specimen. Distal end of th 2^ spine broken. X15. PS-76.4. OSU 33253. h. Lateral aspect of distal stipe fragment. X15. PS-76.4. OSU 33254. i,j. Reverse and obverse aspects of compressed specimens. X15. PS-76.4. OSU 33255. k,l. Obverse and reverse aspects of compressed specimen. X15. PS-76.6. OSU 23256. m,n. Obverse and reverse aspects of compressed specimen. X15. PS-76.4. OSU 23257. o. Reverse aspect of non-isolated specimen. X7.2.PS-76.4. OSU 23258. p. Obverse aspect of non-isolated specimen with well- preserved septal grooves. X15. PS-76.4. OSU 23259. 477 Text-figure 69 478 EXPLANATION OF TEXT-FIGURE 70 Text-figure 70 a-k. Pseudoclimacograptus modestus (Ruedemann). Isolated specimens representing early growth stages. a. Reverse aspect of sicula with initial^bud of th 1^ and walls that will form opening for th 1^. X37.5. PS-102. OSU 33260. b. Reverse aspect showing th 2 1 budding from th 1 2 . X37.5. PS-126. OSU 33261. c. Reverse aspect showing th 2^ dividing distally into two separate openings for th 2^ and th 2^. Prosicula represented by three rods. X37.5. PS-109.2. OSU 33262. d,e. Reverse and obverse aspects of specimen showing development of th 2^ and th 2^. Note th 2^ has no right-lateral wall. X37.5. PS-102. OSU 33263. f,g. Reverse and obverse aspects of specimen showing development of th 2^ and th 2^. X37.5. PS-102. OSU 33264. h,i. Reverse and obverse aspects of specimen with complete th 2^. Note extent of left-lateral wall of th 2 and transverse rod. X37.5. PS-126. OSU 33265. j,k. Reverse and obverse aspects of specimen slightly more developed than that shoxm in Text-figs. 70 h,i. X37.5. PS-102. OSU 33266. - - - t h I --ms ■p- -vi Text-figure 70 VO 480 EXPLANATION OF TEXT-FIGURE 71 Text-figure 71 a-f. Pseudoclimacograptus modestus (Ruedemann). Isolated specimens showing development of rhabdosome. 2 a. Reverse aspect of specimen showing development of th 2 metatheca and th 3 protheca. X37.5. PS-126. OSU 33267. b,c. Obverse and reverse aspects of specimen showing growth of th 2^ metatheca and th 3^ protheca. Note that prosicula consists of longitudinal rods. X37.5. PS-76.4. OSU 33268. d,e. Reverse and obverse aspects of specimen showing develop ment of th 4^ metatheca and th 3^ protheca. Note that growth lines on dorsal parts of lateral walls of th 3% protheca appear to be continuous with those on median septum. X37.5. PS-76.4. OSU 33269. f. Lateral aspect of specimen showing development of th n^ metatheca and th (n+l)^ protheca. Note that growth lines on median septum are comparable to those on lateral wall of protheca. X37.5. PS-126. OSU 33270. 71 g. Sketch showing measured distances for numerical analysis in Table 9. Numbers on Text-figure correspond to those in Table 9. 481 ih3 th3 vL Text-figure 71 482 fifth specimen on a plate. In 1947, Ruedemann re-illustrated these same specimens, and in his plate explanation (PI. 73), he referred to one of the specimens as a holotype and to the others as paratypes. These specimens, which are stored in the American Museum of Natural History, New York, have the following museum and illustration numbers: holotype: AMNH 6926, Ruedemann (1908; text-fig. 403; PI. 28, fig. 10), Ruedemann (1947; PI. 73, figs. 32-33) paratypes: AMNH 6923, Ruedemann (1908; text-fig. 400), Ruede mann (1947; PI. 73, fig. 34) AMNH 6924, Ruedemann (1908; text-fig. 401), Ruede mann (1947; PI. 73, fig. 36) AMNH 6924, Ruedemann (1908; text-fig. 402), Ruedemann (1947; PI. 73, fig. 35). The holotype and paratypes have been examined by the writer, and they agree with Ruedemann's (1908) description. Diagnosis A species of Pseudoclimacograptus with 8-9 thecae in 5 mm proximally and 7.5-8 in 5 mm distally; proximal end broad and rounded in profile view, 0.6-0.9 mm wide; rhabdosome width increases distally to 0.9-1.2 mm at sixth to seventh pair of thecae, thereafter rhabdo some parallel-sided or slightly tapers distally; apertural excavations transverse and narrow; zig-zag median septal groove. Material Approximately 900 specimens are available for study. This material represents the following states of preservation: 1) Isolated Specimens. 330 specimens were obtained from the Pratt's Syncline and Pratt's Ferry sections of which 185 483 represent early growth stages, 110 are young rhabdosomes or fragments of large rhabdosomes with proximal ends and more than two pairs of thecae, and 35 are distal fragments. Most of these specimens are compressed and heavily carbonized, but a few specimens do retain some relief and show growth lines. Although most of the early growth stages are compressed and heavily carbonized, the mode of development of the proximal end can be determined by noting the positions of the open ends of the most distal thecae at various stages of develop ment. In many specimens, growth lines are visible on these distal thecae even though other parts of the specimens are heavily carbonized. 2) Non-isolated, compressed specimens with periderm preserved. More than 550 specimens were collected from the Pratt's Ferry and Pratt's Syncline sections, and more than 95 percent of these specimens are from a horizon that is 76.4 meters above the base of the Pratt's Syncline section. 3) Carbon Film. 38 specimens preserved as carbon film were ob tained from the section at Calera. Description Isolated Specimens. The largest available rhabdosome with a proximal end is 3.8 mm long and consists of 7 pairs of thecae, and the largest rhabdosome without a proximal end is 6 mm long and consists of 9 pairs of thecae. The cross-section of specimens preserved in full relief is oval with somewhat straight lateral walls. In profile 484 aspect, the rhabdosome has a fusiform shape with a broad proximal end (Text-figs. 69e-n). The first two thecae are U-shaped, and each 1 bears a horizontal, subapertural spine. In some specimens, th 2 also bears a horizontal, subapertural spine (See discussion of 3- spined forms below under REMARKS) , but this spine is shorter than 1 those of the first two thecae (0.15-0.20 mm for the 2 ; 0.20-0.50 mm 1 2 for th 1 and th 1 ). The sicula is visible up to the level of the 2 th 1 aperture on the obverse side of the rhabdosome. Its virgella often reaches, but rarely exceeds, a length of 1.0 mm. The aperture 2 of th 1 is situated at a slightly higher level above the aperture of 1 th 1 , and this alternation of the thecal apertures of the opposing thecal series is repeated throughout the length of the rhabdosome. In large, heavily carbonized(gerontic) specimens, a membrane is commonly developed at the sicular aperture (Text-figs. 69a-b). This membrane, which appears to be composed of cortical tissue, initially forms in the angle between the sicular aperture and the virgella. In the maximum development observed, its outer edge extends from the anti-virgellar margin of the sicular aperture to a point midway along the virgella. Maximum development of this membrane is commonly accompanied by the presence of membranes on the thecal spines, which 1) are fin-shaped because they taper in both distal and proximal directions along the spines; 2) are oriented in a plane that coincides with the rhabdosome axis; and 3) appear to be composed of cortical tissue. The rhabdosome is 0.6-0.9 mm wide at the level of the first pair 485 of thecal apertures. Distally, the width increases to a maximum of 0.9-1.2 mm at the level of the sixth or seventh pair of thecae. Thereafter, the rhabdosome is parallel-sided for three to five pairs of thecae before narrowing gradually to a width of 0.9-1.0 mm at the distal end. In most instances, specimens preserved in full relief are slightly narrower than compressed specimens. Proximally, the thecae number 2-2.5 in 1 mm, 5.3-6 in 3 mm, and 8-9 in 5 mm. Distally, the thecal density decreases slightly to 5 in 3 mm. The thecae display a convex supragenicular wall, a distinct geniculum, a slightly introverted aperture, and a narrow, slit-like apertural excavation that is rimmed by a thick list. In distal parts of the rhabdosome, the thecae are 1.2 mm long and overlap for a fourth of their length. The supragenicular wall, which is distinctly convex in both compressed and uncompressed specimens, is 0.3-0.4 ram long at the second thecal pair. Distally, the length of the supragenicular wall increases to 0.5 mm at the sixth to seventh pairs of thecae, after which it is constant. The distal end of the infragenicular wall is perpendicular to the rhabdosomal axis, and this orientation together with that of the supragenicular wall produces a distinct geniculum that overhangs considerably the ventral margin of the pre ceding thecal aperture. The apertural excavations in fully compressed specimens are trans verse and slit-like. The width (trans.) of the apertural excavations (0.25-0.30 mm) is a fourth of the rhabdosome width, and the height (long.), which is 0.05-0.15, is a sixth to a seventh of the length of 485 the free ventral wall. In sharp contract to the condition compressed specimens, the apertural excavations in specimens preserved in full relief (Text-fig. 69c-d) are in most instances hardly visible because the aperture is introverted and the lateral margins of each infrageni cular wall extend downward and overlap the lateral and ventral margins of the preceding thecal aperture. This relationship does not complete ly seal off the thecal apertures because the lateral and ventral margins of the infragenicular wall are situated at a slightly greater distance away from the thecal axis than the lateral and ventral mar gins of the underlying thecal aperture. Additionally, there is a small excavation (or lip) on the mid-ventral margin of the geniculum that allows a small part of the apertural margin to be visible in a ventral aspect. The appearance of the apertural excavations in a profile aspect is the most significant difference between compressed and non-compressed specimens. On the lateral walls of the rhabdosome, a zig-zag groove represents the trace of the median septum. This groove, which extends to the distal end of the rhabdosome, begins at the level of the aper tures of the second pair of thecae on both the obverse and reverse sides of the rhabdosome. Each angle along the zig-zag course of the median septal groove is situated at the level of a thecal aperture on the opposite side of the rhabdosome to which the angle points. A horizontal groove representing the trace of an internal connecting list extends transversely a short distance outward from the angle of the zig-zag median septam groove and merges with the proximal end of a groove that represents the trace of the interthecal septum. 487 Non-isolated Specimens and Carbon Films. The largest available specimen is 9.2 mm long and consists of 14 pairs of thecae. These specimens agree closely with, and show the same quantitative and quali tative features as, the isolated, compressed specimens described above (Text-figs. 69o,p). Early Growth Stages and Rhabdosomal Development. Morphology of Sicula. The sicula is short and broad, widening rapidly to its aperture. Its length ranges from 0.75 to 0.90 mm among the available specimens, and its width at the aperture averages 0.2 mm. The metasicula, which accounts for two-third of the length of the sicula, ranges in length from 0.50 to 0.65 mm among the available specimens. In most specimens, the prosicula consists of three (rarely one, two, or four) thick longitudinal threads, which arise from the meta sicula and converge into the nema (Text-figs. 70a-c). The intervening periderm is rarely present. When it is, it usually occurs only as traces attached to the proximal end of the metasicula. In many speci mens, there are no thick longitudinal threads, and the apex of the prosicula is broken. In rare specimens, the prosicula is complete, in which case faint longitudinal threads extend from the aperture to the apex of the prosicula. The variation in the appearance of the prosicula might be attri buted to the state of preservation of the material or to the laboratory techniques involved in isolating the specimens. However, in several early growth stages, development has proceeded to a point where the 488 prosicula is only represented by thick longitudinal rods (Text-fig. 71 b,c). This suggests that in many instances breakage of the prosicula occurred while the organism was living and at an early stage in its astogeny. The thick longitudinal threads, which connect the metasicula to the nema (virgula), were probably formed by the organism after breakage of the prosicula. The metasicula is composed of transverse growth lines, which bend downward to form the virgella along one side of the sicula (Text-figs. 70a). Except for the virgella, the sicular aperture lacks ornamentation. Development of Proximal End. The available specimens do not represent all the detailed stages in the development of the proximal end. However, the growth stages available are remarkably similar to corresponding growth stages described above for specimens of Climacograptus meridionalis, and they indicate that the specimens referred to here as Pseudoclimacograptus modestus have the same proximal-end development as those referred to above as C^. meridionalis. The similarities of each of the growth stages referred to here as P^. modestus to corresponding growth stages of C^. meridionalis are briefly discussed below. 1 Text-figure 70a: This specimen shows that the origin of th 1 2 1 and the budding of th 1 from th 1 are identical to that of C^. meridionalis (Text-fig. 65a-b). 2 Text-figure 70b: This specimen shows that the growth of th 1 1 and the budding of th 2 resemble closely that of C^. meridionalis (Text-fig. 65e). 489 2 Text-figure 70c: This specimen shows that the budding of th 2 1 from th 2 is similar to that of C^. meridionalis (Text-fig. 65g). 1 2 It also indicates that the horizontally directed th 2 and th 2 apertures form by the growth of a process down onto the sicula from the outer wall of the hood that represents the initial 1 downward-directed part of th 2 . Text-figure 70d-e: This specimen shows that the growth directions 1 2 of th 2 and th 2 are similar to those of meridionalis (Text-fig. 65 1-m); however, none of the available growth stages show the timing of the development of the left-and right-lateral 1 2 walls of th 1 . Th 2 does not have nearly as great a lateral 2 extent as th 2 does in £. meridionalis. Text-figure 70f-g: This specimen is only slightly more advanced than that of Text-figure 70d-e. It does show that the left- 2 lateral wall of th 2 develops in advance of the right-lateral wall as in meridionalis (Text-fig. 65 1-m). Text-figure 70h-k: Except for the introverted thecal aperture, these specimens are very similar to specimens of C^. meridionalis (Text-figs. 65c-e) that represent corresponding stages of develop- 2 ment. The left-lateral wall of th 2 extends dorsally across the reverse side of the sicula to the dorsal margin of the 1 right lateral wall of th 2 , as in C^. meridionalis. The lack of a septal membrane below the transverse list and the extent of 2 the left-lateral wall of th 2 indicate it is the dicalycal thecae. Text-figure 71a-c: The growth of th 3 and the restriction of 2 the th 2 metatheca, which leaves a space for the development 490 2 2 of th 3 between the th 2 metatheca and the virgula, occur similarly in corresponding growth stages of C^. meridionalis (Text-figs. 67a-b,d). Additionally, the right-lateral wall of 1 th 3 develops in advance of the left-lateral wall. The specimens referred to here as IP. modestus have a streptoblastic diplograptid type of proximal-end development. There are four cross- 2 12 1 2 ing canals (th 1 , th 2 , th 2 , and th 3 ), and th 2 is the dicalycal theca. The two thecal series become separated with the development of the third pair of thecae. Development of Median Septum. Although the median septum has a zig-zag course in the specimens described above, it is very similar to that described for C lima cog rap tu s meridionalis. It consists of the virgula, paired transverse rods, and a fusellar membrane. The trans verse rods are regularly spaced and meet the lateral rhabdosomal walls at the angles in the zig-zag median septal groove (Text-fig. 71f). The horizontal grooves on the external surface of the rhabdosome walls indicate that lists extend along the interior surface of the rhabdosome walls from the tips of the transverse rods to the lateral edges of the proximal margins of the interthecal septa. The fusellar membrane of the median septum shows growth lines that are obliquely oriented and converge distally toward the virgula. The initial development of the median septum resembles closely that described above for Climacograptus meridionalis. On the obverse side of the rhabdosome, the median septum initially develops between 1 2 the th 3 protheca and the th 2 metatheca (Compare Text-figs. 491 71bc and 71d-e). It is formed by the dorsal part of the right-lateral 2 wall of th 1 , which bends inward in such a way that its dorsal edge is in contact with the sicula. Thus, the initial part of the median septum is a complete septum on the obverse side of the rhabdosome. In Ç. meridionalis the first transverse rod of the median septum on 1 the reverse side of the sicula forms with th 2 . Such a rod is not present in the specimens described above. However, the formation of 1 the transverse rod that is associated with the budding of th 3 (Text-fig. 70h-i) is very similar to a similarly located rod in C^. meridionalis and marks the initiation of a complete median septum on the reverse side of the rhabdosome in the specimens described above. The median septum is composed of segments of equal size which are separated by the transverse rods. Successive segments develop as the dorsal walls of prothecae situated on opposite sides of the median septum. In some specimens (Text-fig. 7Id,e), the dorsal wall of a protheca (the median septum) develops slightly in advance of its lateral walls. The zig-zag course of the median septum is associa ted with a distal increase in the dorso-ventral width of a protheca, which is accommodated for by growth of the median septum in an ob liquely distal direction. The alternating manner in which prothecae of the opposing thecal series develop results in alternating directions of growth obliquely across the rhabdosome of successive segments of the median septum, producing in a zig-zag course. Thecal Ontogeny. The specimens described above and referred to as Pseudoclimacograptus modestus show features that indicate that 492 their thecal ontogeny agrees closely with that described for Clima cograptus meridionalis. In proximal thecae, the lateral prothecal wall on the reverse side of the rhabdosome develops in advance of the lateral wall on the obverse side of the rhabdosome. The growth lines on the infragenicular wall are continuous with those on the lateral walls of the same theca (Text-fig. 70j-k). A change from closely spaced growth lines (thin fuselli) to widely spaced growth lines (thick fuselli) occurs at the level of the preceding thecal aperture. The geniculum occurs at a level in the distal part of the protheca. The th n+1 protheca buds from the th n theca at a level distal to the th n geniculum. The space between the th n metatheca and the virgula in which the th n+1 protheca develops is formed by a decrease in the dorsal extent of the fuselli that constitute the lateral walls of th n (Text-figs. 71b-e) . The most distal fusellus of the th n protheca extends dorsally to the lateral margins of the median septum. Be tween the interthecal septum and the median septum, the distal edge of this fusellus forms the list that connects the tips of the transverse rods of the median septum to the lateral edges of the proximal end of the interthecal septum. 2 In parts of the rhabdosome distal to th 2 , the initial growth of the protheca involves the formation of two lateral walls and a dorsal wall (the median septum). The dorsal wall of the preceding theca (the interthecal septum) serves as the ventral wall of the protheca. As the protheca develops past the aperture of the prece ding metatheca, it develops its own ventral wall (the infragenicular 493 wall) , but it no longer develops its own dorsal wall. The development of the median septum is taken over by the protheca of the opposing the cal series, resulting in a sharp change in the direction of growth of the median septum. With the budding of the succeeding protheca, the theca again de velops its own dorsal wall, this being the interthecal septum. The theca then grows as a complete tube and eventually develops an aper ture. The final stage in the thecal ontogeny is the formation of the list that borders the thecal excavation. This occurs after the suc ceeding protheca has grown past the aperture. Remarks Numerical Analysis of 2- and 3-spined Forms. As previously men- 1 2 tioned, 2-spined (spines on th 1 and th 1 ) and 3-spined (spines on 12 1 th 1 , th 1 , and th 2 ) forms are present among the available speci mens. Five possibilities are available to explain the existence of these two forms. These explanations are as follows: 1) the two forms represent separate species; 2) the two forms are an expression of continuous variation within a population represented by the Alabama material; 3) the two forms are an expression of discontinuous varia tion within my collections, but in nature the variation is continuous; 4) the two forms are an expression of discontinuous variation resulting from mutation; and 5) the two forms are an expression of discontinuous variation resulting from genetic polymorphism. Among 50 specimens isolated from one sample collected from a sin gle horizon 76.4 meters above the base of the Pratt’s Syncline section, 43 are 2-spined forms, and 7 are 3-spined forms. The high frequency 494 (14%) of the 3-spined forms eliminates mutation (possibility number 4) as an explanation for the presence of the 3-spined form. Among the specimens representing the 3-spined form, the length of the third spine ranges from 0.14 to 0.20 mm and averages 0.17 mm. Because this variation in the length does not range downward to 0 mm, it can be stated that there is not a gradual appearance of the third spine, and the difference between the 2- and 3-spined forms is one of discontinu ous variation. Thus, possibility number 2 does not explain the existence of the 3-spined form. Possibility number 3 does not seem to be an adequate explanation for the existence of the two forms because among the several hundred Alabama specimens collected from many dif ferent horizons and localities throughout the Athens Shale, the 3-spined form occurs only at the horizon situated 76.4 meters above the base of the Pratt's Syncline section, and there it occurs in sufficient abundance to display continuous variation if such variation did indeed occur in nature. In order to test possibilities number 1 and 5, a detailed numeri cal analysis was carried out on the sample of 50 specimens mentioned above. The results of this analysis are tabulated in Table 9, and the manner in which the morphological features of the specimens were measured are shown in Text-figure 71g. On the basis of the calculated means and standard deviations, it appears highly probable that the 2- and 3-spined forms represent genetic polymorphs within one species (possibility number 5). Jaanusson (1973) and Urbanek and Jaanusson (1974) have discussed the phenomenon of genetic polymorphism, which is phenotypically 495 EXPLANATION OF TABLE 9 Numerical analysis of 2- and 3-spined forms of Pseudoclimaco graptus modestus (Ruedemann). Text-figure 71 g shows measured distances listed in Table 9. Numbers in Text-figure 71 g correspond to those in Table 9. Symbols used in Table 9 are as follows: m, mean; s, standard deviation; [ ] , range of variation; N, number of specimens in sample. 496 TABLE 9 2-spined form 3-spined form Measurements in mm. m s [■] N m s r 1 N Length of Sicula 0.84 0.1378 0.60- 9 0.86 - - 1 0. 8 7 1. Width of Sicular Aperture 0.21 0.0316 0.16- 38 0.21 0 . 0 0 7 1 0 .20- 7 0.33 0.21 2. Stipe Width th ll-th 1^ 0 . 7 1 0.0447 0.62- 39 0.67 0.0293 0.62- 7 0.87 0 . 7 1 th 2l-th 22 0 . 7 4 0 . 0 6 3 2 0.57 30 0.70 0.0412 0 .54- 6 0.87 0 . 7 1 th 3^-th 32 0 . 8 0 0 . 0 7 0 7 0 .69- 23 0 . 7 7 010693 0.71- 6 0.90 0 . 9 0 th 4^-th 4^ 0 . 8 9 0 . 6 3 2 0 .82- 15 0 . 8 9 0.0911 0.83- 4 0.99 1.02 th Sl-th 5^ 0.92 0.0775 0.84— 6 0 1.05 3. Thecal Density t h / l m m 2 . 2 4 0 . 1 4 8 3 2.0- 2 8 2 . 3 3 0 . 2 0 5 9 2.2- 6 2.5 2 . 5 t h / 2 m m 4.04 0.1449 3.6- 13 4 . 2 1 0 . 1 8 4 9 4 . 0 - 4 4.2 4 . 5 t h / 3 m m 5 . 5 7 0 . 2 8 9 8 5.3- 0 6.0 4. Longitudinal distances sicula-th l2 0 . 4 1 0 - 0 3 1 6 0 . 3 4 - 41 0 . 4 1 0 . 0 4 3 6 0 . 3 5 - 7 0.49 0 . 4 6 s i c u l a - t h 32 1 . 3 3 0.0775 1.18- 26 1 . 3 0 0.1233 1.17- 5 1.52 1 . 3 9 sicula-th 5^ 2 . 4 5 0 . 1 0 4 9 2.29- 6 —— 0 2.57 th 1^ -th 3^ 0 . 9 2 0 . 0 5 4 8 0.84- 26 0 . 9 0 0 . 0 7 3 5 0 . 8 1 - 5 1.02 0 . 9 9 th 1% -th 5^ 2 . 0 4 0.1000 1.89- 6 0 2.14 5. Length (long.) of T h e c a l A p e r t u r e th 1^ 0 . 0 7 0.0200 0 .03- 40 0 . 0 7 0.0100 0 . 0 6 - 7 0.12 0 . 0 9 t h 3^ 0 . 0 8 0.0200 0 .07- 19 0 . 0 9 0 . 0 1 7 3 0 . 0 7 - 4 0.13 0.10 t h 52 0 . 0 8 0.0100 0.07- 4 0 0.09 6. Width (trans.) of Thecal A p e r t u r e th 1^ 0.12 0 . 0 4 4 7 0 .05- 40 0.11 0 . 0 1 4 1 0 . 0 9 - 7 0.22 0 . 1 3 th 3^ 0 . 1 9 0 . 0 5 7 4 0.09- 19 0 . 1 7 0 . 0 5 2 0 0.12- 4 0.30 0.22 th 5-^ 0 . 1 8 0 . 0 8 9 4 0.09- 4 0 0.27 7. L e n g t h o f s u p r a g e n i c u l a r w a l l th 2^ 0 . 3 5 0 . 0 3 1 6 0.28- 38 0 . 3 1 0 . 0 4 8 0 0 . 2 4 - 7 0.42 0 . 3 b th 4I 0 . 4 3 0 . 0 4 4 7 0.33- 18 0.41 0.0412 0.36- 5 0 . 5 1 0 . 4 7 497 expressed by morphological discontinuities, and its significance in the evolution of graptolites. In the case of the specimens described above, the evolutionary significance of the morphological discontinuity and implied genetic polymorphism of the third thecal spine is negligible because this feature is restricted to populations within a very limited interval of the stratigraphie range of the species in Alabama. Similar appearances and disappearances of morphological discontinuities invol ving thecal spines occur in other graptoloid species described herein (e.g. Dicellograptus gurleyi and alabamensis), and because the pop ulations with the discontinuous variation are restricted generally to those parts of the Pratt's Ferry and Pratt's Syncline sections with the greatest shale to carbonate ratio, it seems likely that the phenotypic expression of the inferred genetic polymorphism is environmentally influenced. Rhabdosomal Development. As discussed above, the rhabdosomal development of the specimens referred to here as Pseudoclimacograptus modestus is very similar to that of Climacograptus meridionalis. P. modestus differs from C^. meridionalis by the consistent occurrence throughout the rhabdosome of the distal increase in the dorso-ventral width of each protheca, which produces the zig-zag median septum. This difference involves basically a slight change in thecal ontogeny. A similar change in thecal ontogeny, which is of a comparable degree, occurs between those portions of a rhabdosome of C^. meridionalis with an undulating and a straight median septum. Thus, by a slight change in thecal ontogeny and an increase in the number of thecae in 498 the rhabdosome that show this ontogeny, the median septum of modestus can be derived easily from the median septum of C^. meridionalis. Relationship to Other Species. Besides slight differences in quantitative features, the only significant morphological differences between Pseudoclimacograptus modestus and Climacograptus meridionalis are the convex supragenicular wall and the slightly introverted thecal aperture. Considering these differences and the great number of similarities, 2» modestus is here considered to be a species that is distinct from, yet closely related to, meridionalis. These two species are assigned to two different genera because of the differen ces in the shapes of their supragenicular walls (See Discussion under Diplograptidae). The specimc -s described above agree closely with the holotype and paratypes of Pseudoclimacograptus modestus, as well as with Ruedemann's (1908, 1947) descriptions. Measurements were taken on the specimen of Climacograptus cf. modestus that is figured by Bulman (1931; Text-fig. 22, PI. 5, fig. 13). This specimen agrees closely with the specimens described above. Bulman's (1948) descriptions and illustrations of 2- modestus also agree with the Alabama specimens. According to Riva (1974), two specimens figured by Ruedemann (1908, text-figs. 378-379) as C. putillus mut. eximius and refigured as 2- eximius (Ruedemann, 1947; PI. 72, figs. 2-3) should be assigned to 2- modestus. The figures of these specimens in Ruedemann (1908, 1947) have a strong resemblance to the Alabama specimens. Pseudoclimacograptus scharenbergi stenostoma (Bulman, 1947) 499 exhibits a strong resemblance to 2» modestus, agreeing in rhabdosome width, thecal density and overall appearance. Riva (1974) differentia ted these two species by the more slit-like apertural excavations and the greater curvature of the supragenicular walls in P. scharenbergi stenostoma. However, when this distinction was made, 2* modestus was known only from fully compressed specimens, which show wide thecal excavations, and 2* scharenbergi stenstoma was originally de scribed on the basis of specimens preserved in full relief. Several of the specimens of 2* modestus described above are preserved in full relief, and these specimens cannot be distinguished by the shapes of the supragenicular walls and apertural excavations from Bulman's (1947) specimens of 2- scharenbergi stenostoma. 2* scharenbergi stenostoma can be distinguished from P. modestus by its proximal-end 2 development, which shows a very sinuous direction of growth of th 1 , and by the horizontal lists that extend outward from the angles of the median septum but do not merge with the proximal ends of the interthecal septa. Thus, these two species can only be distinguished on the basis of exceptionally well-preserved specimens that show the proximal-end development and/or the septal grooves. Specimens assigned to 2' scharenbergi stenostoma by Riva (1974) and Berry (1960) are fully compressed and poorly preserved, and they could easily represent 2» modestus. Pseudoclimacograptus eurystoma Jaanusson (=Climacograptus scharenbergi Lapworth of Bulman, 1932) resembles the specimens described above in the structure of the median septum, the arrangement 500 of the septal grooves, the density of the thecae, and the width of the 1 rhabdosome. The two can be distinguished by the dicalycal th 2 , the inclined infragenicular wall, and the wide thecal excavations in P. eurystoma. Figured Specimens OSU 33249-OSU 33270. 501 Pseudoclimacograptus sp. cf. eurystoma Jaanusson, 1960 (Text-figures 68a-c) Type Data The holotype Pseudoclimacograptus eurystoma is the rhabdosome figured by Bulman (1932, Pl. 1, fig. 19). It is stored in the Swedish Museum of Natural History, Stockholm. Diagnosis A species of Pseudoclimacograptus with apertural excavations oc cupying half the length of the free ventral wall; pseudoclimacograptid thecae number 5.5 in 3 mm; rhabdosome attaining maximum width of 1.2-1.3 mm at th 3. Material The available material consists of six specimens obtained from the Calera section. These specimens are poorly preserved, being represented by carbon films on black shale surfaces. Description The largest available specimen (Text-fig. 68c) is 4 mm long and consists of seven pairs of thecae. In profile aspect, the proximal end is wide and rounded. The first two thecae are U-shaped with their apertures opening distally. A short, 0.5 mm long virgella is the only spine on the proximal end. 502 At the first pair of thecal apertures, the rhabdosome is 0.8-1.0 mm wide. Distally, it increases in width to 1.2-1.3 mm at th 3, after which it is parallel-sided. The thecae, which number 5.5 in 3 mm, are of the pseudoclimaco graptid type and are characterized by large apertural excavations. The thecae exhibit a distinct geniculum and a convex supragenicular wall, which is 0.2-0.3 mm long. In some instances, possibly due to compres- sional distortion, the supragenicular wall is inclined. In proximal thecae, the aperture is horizontal (trans.), but in distal thecae it is slightly inclined. The apertural excavations are 0.2-0.3 mm long and wide, thus they occupy a fourth of the rhabdosome width and half of the length of the free ventral wall. Each apertural excavation is rimmed by a list. The lateral walls of the rhabdosome are too poorly preserved in the available specimens to reveal the presence of a median septal groove. Remarks ■ The specimens described above are assigned to the genus Pseudoclimacograptus because of the convex supragenicular wall. The general shape of the rhabdosome and the thecal characteristics, pri marily the large apertural excavation and short supragenicular wall, resemble those of eurystoma Jaanusson. Because jP. eurystoma has been described only for uncompressed isolated specimens, a close com parison with the specimens described above is difficult. Figured Specimens OSU 33239-OSU 33241. 503 Genus Glyptograptus Lapworth, 1873 Type Species: Glyptograptus tamariscus(Nicholson, 1868a) Diagnosis As in Bulman (1970). Species Described Herein Glyptograptus euglyphus(Lapworth, 1880) G^. sp. cf. G. teretiusculus(Hisinger, 1840) Glyptograptus euglyphus(Lapworth, 1880) (Text-figures 72-75) 1877 Diplograptus dentatus Brongniart, Lapworth, p. 132, PI. 6, fig. 13. 1880 Diplograptus (Glyptograptus) euglyphus sp. nov., Lapworth, p. 166- 167, PI. 4, figs. 14a-e. 1907 Diplograptus (Glyptograptus) teretiusculus(Hisinger) var. euglyphus Lapworth, Elies and Wood, p. 252, text-fig. 172; PI. 31, figs.2a-d. 1908 Diplograptus (Glyptograptus) euglyphus Lapworth, Ruedemann, p. 369-370, text-figs. 315-316; PI. 25, figs. 21-23. ? 1935 Diplograptus (Glyptograptus) cf. euglyphus (Lapworth) , Harris and Thomas, p. 297-298, fig. 3, nos. 39-41. 1947 Diplograptus (Glyptograptus) euglyphus Lapworth, Ruedemann(partim), p. 405-406, PI. 69, figs. 46-50(non PI. 69, figs. 55-59). 1952 Diplograptus (Glyptograptus) euglyphus Lapworth, Decker, PI. 1, fig. 33; PI. 2, fig. 56; PI. 3A, fig. 19, 19a. 1960 Glyptograptus teretiusculus var. euglyphus(Lapworth), Berry, p. 88, PI. 15, fig. 8. 504 EXPLANATION OF TEXT-FIGURE 72 Text-figure 72 a-c. Glyptograptus euglyphus (Lapworth). a-f. Isolated specimens, g-i. Non-isolated specimens. a,b. Reverse and obverse aspects of uncompressed specimen. X18.8. PS-109.2. OSU 33271. c. Reverse aspect of partly compressed specimen. X18.8. PS-126. OSU 33272. d. Reverse aspect of partly compressed specimen. X18.8. PS-109.2. OSU 33273. e. Profile aspect of partly compressed specimen. X18.8. PS-109.2. OSU 33274. f. Profile aspect of partly compressed specimen. Note change in growth-line thickness at distal end of protheca. X37.5. PS-109.2. OSU 33275. g. Obverse aspect. X9. PS-77.5. OSU 33276. h. Obverse aspect. X4.5. PF-25. OSU 33277. i. Reverse aspect. X4.5. PF-17.1. OSU 33278. 505 e Text-figure 72 506 EXPLANATION OF TEXT-FIGURE 73 Text-figure 73 a-r. Glyptograptus euglyphus (Lapworth). Isolated specimens representing early growth stages. a. Obverse aspect of sicula showing resorption foramen and initial bud of th 1^. X37.5. PS-109.2. OSU 33279. b. Reverse aspect of sicula with initial bud of th 1^. X37.5. PS-109.2. OSU 33280. c. Reverse aspect of sicula with th 1^. X37.5. PS-109.2. OSU 33281. d. Reverse aspect of sicula with th 1^. Note lateral expansion of th 1^ and curvature of growth lines. X37.5. PS-109.2. OSU 33282. e. Ventral aspect of sicula with th 1^. Note lists bordering2 downward-directed opening of th 1^ and opening of th 11^ directed to reverse] side of sicula. X37.5. PS-109.2. OSU 33283. f. Reverse aspect of sicula showing opening in th 1^ for th l2. X37.5. PS-109.2. OSU 33284. g. Reverse aspect of sicula and th 1^. Note growth of wall on obverse side of th 1^. X37.5. PS-109.2. OSU 33285. h. Reverse aspect of sicula with th 1^ growing upward. X37.5. PS-109.2. OSU 33286. i. Reverse aspect of sicula with th 1^ growing upward. X37.5. PS-109.2. OSU 33287. j . Reverse aspect of sicula with th 1^ growing upward. X37.5. PS-109.2. OSU 33288. k. Reverse aspect of sicula with th 1^ growing upward and partly enclosing downward-directed part. X37.5. PS-129. OSU 33289. l,m. Ventral and reverse aspects of sicula with upward- directed part of th 1^ partly enclosing downward-directed part. Note that lateral wall of th 1 on reverse side of 507 sicula does not join sicular wall. An opening is left for th 1^. X37.5. PS-129. OSU 33290. Text-figure n. Reverse aspectof sicula with th 1 . Note that upward- directed part of th 1^ completely encloses downward-directed part except for an opening for th 1^. X37.5. PS-109.2. OSU 33291. o. Reverse aspect showing opening for th 1^ in lateral wall of th 1^. X37.5. PS-102. OSU 33292. p. Reverse aspect showing earliest growth of th 2^. X37.5. PS-109.2. OSU 33293. q. Reverse aspect showing growth of th 1^. X37.5. PS-109.2. OSU 33294. 2 r. Reverse aspect showing growth of th 1 . X37.5. PS-109.2. OSU 33295. -ms ^ - t h I — Ln Text-figure 73 8 509 EXPLANATION OF TEXT-FIGURE 74 Text-figure 74 a-n. Glyptograptus euglyphus (Lapworth). Isolated specimens representing early growth stages. a,b. Reverse and obverse aspects showing growth of th 1^ and th 1^. X37.5. PS-1261 OSU 33296. c,d. Obverse and reverse aspects showing development of th 1 1 and th 1^. X37.5. PS-109.2. OSU 33297. e,f. Obverse and reverse aspects showing th 1^, th 1^, and initiation of th 2^. Note that intercalary fuselli mark initiation of th 2^ and are continuous %fith infragenicular wall. X37.5. PS-109.2. OSU 33298. g,h. Ventral and distal-obverse aspects showing dorsal wall of th 1^ (interthecal septum). X37.5. PS-102. OSU 33299. i,j. Obverse and reverse aspects showing growth of th 2^. Th 1^ dorsal wall (interthecal septum) probably marks budding of th 2^. X37.5. PS-126. OSU 33300. k. Obverse aspect showing sicula enclosed by lateral walls of th l2 and th 2^. X37.5. PS-126. OSU 33301. l,m. Obverse and reverse aspects showing development of th 2^. X37.5. PS-109.2. OSU 33302. n. Lateral aspect of distal fragment of rhabdosome. Note that growth lines on median septum compare in thickness with those on most distal protheca. X37.5. PS-109.2. OSU 33303. Vi I-* Text-figure 74 o 511 EXPLANATION OF TEXT-FIGURE 75 Text-figure 75 a-j. Glyptograptus euglyphus (Lapworth). Isolated specimens showing development of rhabdosome. a. Obverse aspect of specimen shown in detail in Text-figs. 75 b-c. X18.8. PS-109.2. OSU 33304. b. Obverse aspect showing development of th 3^ metatheca and th 3^ protheca. Note that thickness of fuselli on median septum are comparable to those on lateral walls of th 3^ protheca. X37.5. PS-109.2. OSU 33304. c. Distal-ventral aspect showing interthecal septum between th 2^ and th 3^. X37.5. PS-109.2. OSU 33304. d. Obverse aspect of fragment of rhabdosome showing development of th 2^ and th 3^. Note that th 3^ initially develops on obverse side of rhabdosome. X37.5. PS-109.2. OSU 33305. e,f. Reverse and obverse aspects showing development of th 2^ and th 3^. X37.5. PS-117. OSU 33306. g,h. Reverse and obverse aspects of specimen showing develop ment of th 2 metatheca and th 3 protheca. Note that intercalary fuselli denote budding of th 3^ from th 2^. X37.5. PS-109.2. OSU 33307. 1,]' Reverse and obverse aspects of specimen showing develop ment of th 2 metatheca and th 3 . Note that th 3^ occupies entire width between lateral walls of rhabdosome. X37.5. PS-109.2. OSU 33308. msp m,sp th 3/ 1 .is jk 22 th 3'pt-- > ..... ,th3! Ui ro Text-figure 75 513 1963 Glyptograptus euglyphus (Lapworth), Ross and Berry, p. 140, Pl. 10, figs. 27-28; Pl. 11, figs. 3-4. 1964 Glyptograptus euglyphus (Lapworth), Berry, p. 146-147, Pl. 15, figs. 1-2. Type Data Strachan (1971) reports that Lapworth's (1880) type specimens can not be traced, and for this reason a lectotype can not be selected. Several of the specimens figured by Elies and Wood (1907, PI. 31, figs. 2a-c) are stored at the Geological Survey of Scotland and the Sedgwick Museum. The selection of a neotype must await an examination of these specimens. Diagnosis A species of Glyptograptus with parallel-sided rhabdosome tapering distally to narrow, pointed proximal end; rhabdosome width 0.60 - 0.75 mm at th 1, increases distally to maximum of 2.2 - 2.5 mm between tenth and fifteenth pairs of thecae, thereafter parallel-sided. Proximal thecae of glyptograptid type with geniculum, number 5.5 - 6.5 in 5 mm, free ventral wall inclined less than 20 degrees; distal thecae of glyptograptid type without geniculum, number 4 - 5 in 5 mm, free ventral wall inclined 40 degrees. Median septum complete, straight, 1 extends proximally to level of th 2 aperture. Material More than 900 specimens are available for study and represent the following states of preservation: 1) Isolated specimens. Approximately 250 specimens were ob tained from the Pratt's Syncline section of which 150 514 represent early growth stages, 60 are young rhabdosomes or fragments of large rhabdosomes with proximal end and two or more pairs of thecae, and 45 are distal fragments. The early growth stages are very well preserved in full relief and show growth lines. Larger specimens are generally com pressed and have a thick cortical layer. 2) Non-isolated, compressed specimens with periderm preserved. 35 specimens were obtained from the Pratt’s Syncline and Pratt’s Ferry sections. 3) Carbon films. More than 600 specimens were obtained from the Calera section. Description Isolated Specimens. The largest available specimen with a proximal end is 4.9 mm long and consists of six pairs of thecae (Text-fig. 72c). In cross-section, specimens preserved in full relief are oval with straight lateral walls. In biprofile aspect, the ventral walls of the rhabdosome taper proximally to a narrow proximal end, and thecal apertures of the opposing thecal series occur at alternating levels along the rhabdosome (Text-figs. 72a-d). The narrow proximal end is produced by the shape of the first 1 two thecae. The distal, upward-directed part of th 1 is in com- 2 pie te contact with the virgellar side of the sicula. Th 1 extends obliquely across the reverse side of the sicula from the base of the • 1 upward-directed part of th 1 to a level on the anti-virgellar side 1 of the sicula slightly above the aperture of th 1 . Only the 515 anti-virgellar margin of the sicular aperture is visible in the re verse aspect. In the obverse aspect, the sicula is visible up to the 2 level of the th 1 aperture. The virgella averages 0.7 mm in length, but in a few specimens it is almost 1.4 mm long. The base of the virgella is in contact with the base of the upward-directed part of th 1 1 2 1 . The free ventral walls of th 1 and th 1 are inclined at angles of less than 20 degrees to the axis of the rhabdosome, and the rhab dosome is 0.60-0.75 mm wide at the level of the apertures of the first thecal pair. Distally, the rhabdosome gradually increases in width to 1.0-1.3 mm at the fifth pair of thecae. The rhabdosome fragment with the greatest width (Text-fig. 72e), which is probably from the proximal end of the rhabdosome judging by the shape of the thecae, attains a ' maximum width of 1.6 mm. The thecae of the second to sixth pairs of thecae are basically of the glyptograptid type, yet at the distal end of the apertural excavation the inclination of the free ventral wall decreases in such a way that a geniculum with a very obtuse genicular angle (140-170 degrees) is formed. At th 2, the free ventral wall is 0.5-0.6 mm long, and its supragenicular portion is 0.35-0.40 mm long. At th 5 the free ventral wall is 0.7-0.8 mm long, and its supragenicular portion is 0.3 mm long. Thus, distally in the proximal part of the rhabdosome, the portion of the free ventral wall that is a supra genicular wall decreases, and the distal continuation of this trend, which occur in the large, non-isolated rhabdosomes described below. 516 results in a loss of supragenicular wall. In the distal parts of the rhabdosome, the free ventral walls of the thecae assume a typical glypto graptid shape. The supragenicular walls in the proximal thecae are inclined to the axis of the rhabdosome at angles less than 20 de grees, whereas the infragenicular walls are inclined at angles of 40 • degrees. With the distal decrease and elimination of the supra genicular wall, the inclination of the free ventral wall, as a whole, increases. Proximally the thecae are 0.6 mm long and overlap a third to a half their length. At th 5, the thecae are 1.2-1.4 mm long and overlap a third their length. The thecae number 4-5 in 3 mm as 2 measured from the aperture of th 1 . The thecal apertures are simple, without ornamentation, and their margins define a plane that is perpendicular to the axis of the rhabdosome in uncompressed specimens. The apertural excavations of the first two thecae have a semi-circular shape in profile view. They are 0.2 mm wide (trans.) and 0.2 - 0.3 mm high (long.), and they occupy a fourth to a third of the width of the rhabdosome and less than a third of the length of the free ventral wall. The margins of the apertures are thickened by lists, which are continuous with lists along the lateral margins of the infragenicular wall. The lists on the infragenicular walls are much thinner, and thus less distinct, than the lists on the margins of the aperture. The position of the lists on the free ventral wall is useful for determining the distal limit of the apertural excavations in distal thecae. Distally, as the length of the infragenicular wall assumes a 517 greater portion of the length of the free ventral wall, the apertural excavation takes on a notch shape with one side of the notch represent ing the margin of the aperture. At th 5, the apertural excavations are 0.4-0.5 mm high (long.) and 0.4-0.5 mm wide (trans.), and they occupy between a third and a half of the width of the rhabdosome and half of the length of the free ventral wall. The larger, isolated specimens have thick cortical layers, which tend to obscure the septal grooves on the reverse and obverse walls of the rhabdosome. However, in some specimens (Text-figs. 72a-c), a straight median septal groove can be traced proximally to 1 the level of the th 2 aperture, and the Interthecal septal grooves, which are inclined 30 - 40 degrees to the axis of the rhabdosome, can be traced proximally from the dorso-lateral corners of the aper tural excavations toward the median septal groove. In rare instances 1 2 (th 3 and th 3 in the specimen shown in Text-fig. 72a-b), transverse grooves connect the proximal ends of the interthecal septal grooves to the median septal groove. In the one available transparent distal fragment (Text-fig. 72f), the only feature that can be regarded as a transverse groove is the contact between the most distal fusellus of a protheca and the most proximal fusellus of the succeed ing protheca of the same thecal series. Non-isolated, compressed Specimens with periderm preserved and Carbon Films. The largest available specimen (Text-fig. 72i) is 3.7 cm long and consists of 35 pairs of thecae. At the level of the first pair of thecal apertures, the rhabdosome is 0.60-0.75 mm wide. 518 Distally, the rhabdosome width increases to 1.2-1.9 mm (average 1.7 mm, smaller widths are from specimens that retain some relief) at the tenth pair of thecae, which is situated 8.5-9 mm from the proximal end. The rhabdosome continues to widen, but more slowly, to its max imum width of 2.2-2.5 mm at the fifteenth pair of thecae, which is situated approximately 14 mm from the proximal end. Thereafter, the rhabdosome is parallel-sided. As described above in isolated specimens, the free ventral walls of the proximal thecae are geniculate, and the supragenicular portions are inclined less than 20 degrees to the axis of the rhabdosome. The large, non-isolated specimens (Text-figs. 72g-i) show that the portion of the free ventral wall occupied by a supragenicular wall gradually decreases distally, and between the tenth and fifteenth pairs of thecae, where the rhabdosome attains its maximum width, the free ventral walls of the thecae assume a typical glyptograptid sigmoid shape. In fact, distal fragments of the rhabdosome can not be dis tinguished from distal fragments of the Alabama specimens referred to below as Glyptograptus sp. cf. teretiusculus. Corresponding apertures of thecae of the opposing thecal series are alternating in the proximal part of the rhabdosome, but in distal parts of the rhabdosome, the corresponding apertures are situated at equal distances from the proximal end. There are 5.5-6.5 thecae in the most proximal 5 mm of the rhabdosome and 10-11 in the most proximal 10 mm. In distal parts of the rhabdosome, there are 4.5 thecae in 5 mm and 8-10 in 10 mm. 519 The large, uncompressed specimens also show that the thecae and thecal apertures increase in size to the tenth to fifteenth pairs of thecae, after which they are of uniform size. In distal parts of the rhabdosome, the thecae are 2.0-2.5 mm long and overlap for a third their length, and the free ventral wall, which is inclined at 30-40 degrees to the axis of the rhabdosome, is 1.3-1.6 mm long. The thecal apertures, in the distal thecae, are straight and perpendicular to the axis of the rhabdosome, but because of preservational and distortional phenomena, the apertures are commonly everted, introverted, and/or denticulate. The apertural excavations in distal thecae, which are ill-defined because of the slightly sigmoid free ventral wall, occupy a fourth to a third of the width of the rhabdosome. Early Growth Stages and Rhabdosomal Development. Morphology of Sicula. The sicula is relatively short and broad, and the width of the sicular aperture (0.30 mm) is equal to approxi mately half the length of the sicula (0.65-0.80 mm) and approximately two-thirds the length of the metasicula (0.45-0.55 mm). The prosicula (Text-fig. 741-m) has an apertural width of 0.10-0.12 mm and displays longitudinal threads. The apex of the prosicula is situated approxi- 2 mately at the level of the th 1 thecal aperture (Text-fig. 75e-f). The metasicula is composed of transverse growth lines, which in the distal half of the metasicula bend downward to form the virgella on one side of the sicula. Except for the virgella, the metasicular aperture lacks ornamentation. 520 Development of Proximal End. The proximal-end development that is shown in Text-figures 73-75 is so similar to the proximal-end development of Climacograptus brevis brevis Elies and Wood (described by Bulman, 1947) and brevis mutabilis Strachan, 1960 that it can be considered to be identical. This type of proximal-end develop ment is represented here by a more complete series of growth stages than what was avilable to Bulman (1947) and Strachan (1960). For this reason, a detailed description of the proximal-end development is justified. 1 The budding of th 1 begins with the formation of an absorption foramen in the metasicula immediately to the right of the proximal 1 end of the virgella (Text-fig. 73a). Th 1 develops from this foramen and grows directly down the sicular wall along the virgella (Text- 1 figs. 73b-c). Within a short distance of its origin, th 1 expands laterally to form a hood (Text-figs. 73d-e) . The left side of the hood is in contact with the obverse wall of the sicula immediately to the left of the virgella, whereas the right side of the hood 2 is not in contact with the sicula, and an opening, from which th 1 1 will later bud, is formed on the right side of th 1 . 1 Th 1 continues to grow downward to near the level of the metasicular aperture, where two lists are formed (Text-figs. 73e-f). One list, which forms on the margin of the downward-directed 1 opening of th 1 , is in contact with the sicula on both its right and left sides, thus it forms the distal margin of the opening 1 2 in th 1 through which th 1 buds. The other list rims the margins 521 2 of the opening for the th 1 bud. 1 The downward growth of th 1 continues past the list by the addition of fuselli on only the right-lateral wall (which is on the obverse side of the sicula, Text-fig. 73g). These fuselli are obliquely oriented and extend from the aperture that is rimmed by the 1 list to the virgella. With continued growth, th 1 develops a free ventral wall, and its aperture is directed upward (Text-fig. 73h). The growth lines on this ventral wall are continuous with those on the right-lateral wall. 1 The upward growth of th 1 continues in such a way that the upward- 1 directed part of th 1 encloses the downward-directed part (Text- 1 figs. 73i-o). However, as th 1 grows back over the level at which 2 the th 1 opening is situated, its left-lateral wall does not extend onto the reverse side of the sicula. Instead, it extends only to the 2 list that borders the th 1 opening (Text-figs. 731-o). Above the 2 level of the th 1 opening, the right- and left-lateral walls of th 1 1 are in contact with the reverse and obverse sides of the sicula 1 respectively, and the upward-directed part of th 1 completely en closes the virgellar side of the sicula (Text-fig. 74b). In specimens preserved in full relief (Text-fig. 74g-h), the dorsal portions of the 1 lateral walls of the upward-directed part of th 1 actually extend inward toward the sicula, thus forming two walls that together with 1 the virgellar side of the sicula constitute the dorsal wall of th 1 . These two walls are obliquely oriented with respect to the sicular axis, and distally they grow outward from the sicula and merge to form 522 a single dorsal wall (Text-fig. 74e-f). The growth lines on this single dorsal wall are continuous with those on the lateral walls of 1 1 th 1 . However, after the aperture of th 1 is completed, its dorsal 1 wall continues to develop as the infragenicular wall of th 2 . When 1 it is completed, the aperture of th 1 is situated at the level of the prosicula-metasicula contact. 2 The growth of th 1 commences when the upward-directed part of 1 th 1 reaches the proximal end (relative to the downward-directed 1 2 part of th 1 ) of the th 1 opening (Text-fig. 73p), The initial fuselli, which are obliquely oriented with respect to the sicular axis, extend from the margin of the sicular aperture adjacent to the 2 virgella across the distal end of the th 1 opening to the dorsal 1 2 margin of the right-lateral wall of th 1 . Th 1 grows obliquely upward across the reverse wall of the sicula in such a way that its right-lateral wall grows around to the obverse side of the sicula 1 and its left-lateral wall overlaps the right-lateral wall of th 1 2 (Text-fig. 73q-r, 74a-d, g-h). As th 1 approaches the level of the 1 1 aperture, intercalary fuselli, which represent the budding of th 1 2 2 , appear on the dorsal portion of the left-lateral wall of th 1 2 1 (Text-figs. 74e-f, i-j). When th 1 attains the level of the th 1 2 aperture, a wall that connects the lateral walls of th 1 develops (Text-fig. 74i-j). The growth lines on this wall are discontinuous 2 with those of the th 1 lateral walls. Initially, this wall probably 2 2 represents the interthecal septum between th 1 and th 2 , and with 2 continued growth it will form the th 2 infragenicular wall. 523 1 The budding of th 2 , which is denoted by the appearance of 2 intercalary fuselli in the left-lateral wall of th 1 , occurs simul taneously with the development of a wall on the obverse side of the sicula (Text-figs. 74i-k). This wall extends from the dorsal wall of 1 2 th 1 to the dorsal margin of the right-lateral wall of th 1 , and with its development, the sicula becomes completely enclosed. As re- 1 vealed by growth lines, th 2 also develops its own ventral wall, which is continuous with and forms an extension of, the dorsal wall of th 1 1 (Text-fig. 74i-j). The development of the dorsal wall of the th 2 2 1 metatheca described above indicates the budding of th 2 from th 1 2 (Compare Text-figs. 74i-j and 75a-b). 1 2 Advanced growth stages that show the growth of th 2 and th 2 1 2 and the budding of th 3 and th 3 are scarce, but those that are available (Text-fig. 75) are transparent enough to reveal growth 1 lines. These specimens show that th 2 grows slightly in advance of 2 1 th 2 (Text-figs. 75e-f). When the th 2 metatheca is slightly more than half the distance to its aperture, it develops its own dorsal wall (Text-fig. 75i-j). The growth lines on the initial part of this 1 1 dorsal wall (the interthecal septum between th 2 and th 3 ) are 1 continuous with those on the lateral walls of th 2 (Text-fig. 75d) . At a slightly higher level on the rhabdosome, intercalary fuselli 2 appear on the dorsal parts of the lateral walls of th 2 (Text-figs. 1 2 75c, i-j), and this represents the budding of th 3 from th 2 . 1 Th 3 initially develops on the reverse side of the rhabdosome (Text-fig. 75d) , but it quickly expands to the full width of the 524 rhabdosome (Text-fig. 75i-j). As revealed by growth lines (Text-figs. 75c, i-j), the median septum develops with, and as the dorsal wall of, 1 th 3 . Accordingly, it appears slightly earlier on the reverse side of the rhabdosome than on the obverse side. 1 1 As th 3 grows past the aperture of th 2 , it develops its own ventral wall (the infragenicular wall), which is continuous with the 1 1 interthecal septum between th 2 and th 3 (Text-fig. 75i-j). Simul- 2 taneously, th 3 develops (Text-fig. 75a-c). Because of the presence 2 2 of a complete median septum, th 3 must bud from th 2 . With its 2 2 development, th 3 becomes separated from th 2 by an interthecal septum, which according to the growth line pattern forms as a dorsal 2 wall of th 2 (Text-fig. 75a-c). Growth line evidence also indicates 2 that with the budding of th 3 , the development of the median septum 2 changes in such a way that it now forms as a dorsal wall of the th 3 protheca. With the development of the third pair of thecae, the two thecal series become permanently separated by the median septum, and the thecae in each thecal series bud successively one from another. The proximal-end development with the entirely upward direction of growth 2 of th 1 is of the prosoblastic diplograptid type (Bulman, 1970). The proximal-end development is also characterized by the presence of four 2 crossing canals and a dicalycal th 2 . Development of Median Septum. The median septum in the available specimens is a simple structure consisting of a fusellar septal mem brane and the virgula. It has no transverse rods nor any lists 525 connecting the median septum to the interthecal septa. As described above, the initial development of the median septum is simultaneous 1 with the budding of th 3 , and the growth line pattern on this initial part of the median septum is similar to that on the lateral wall of 1 th 3 (Text-fig. 75d). On the lateral walls of a protheca, the initial (most proximal) growth lines are closely spaced, whereas more distally 2 they are relatively widely spaced. When the th 3 protheca initially develops (Text-fig. 75a-c), the growth lines on its lateral walls and on the median septum are closely spaced, whereas those on the lateral 1 walls of the adjacent, further developed portion of th 3 are widely spaced. This same relationship is also found in more distal thecae (Text-fig. 74n), and it indicates that the median septum develops in successive segments that are formed by prothecae on alternating sides of the median septum. No specimen is available that..represents the non-alternating distal thecae (described above under non-isolated specimens) , and thus the development of the median septum in those portions of the rhabdosome can not be determined. Thecal Ontogeny. The budding of a new protheca (th n) occurs when th n-1 has grown past the preceding thecal aperture (th n-2). It is commonly, but not always, preceded by the appearance of inter calary fuselli on the dorsal parts of the lateral walls of th n-1. Its development is fully realized when th n-1 develops a dorsal wall (the interthecal septum) in such a way that a space, which th n simultaneously fills, is left between the dorsal wall of the th n-1 metatheca and the median septum. 526 a The protheca th n initially develops its own lateral walls and its own dorsal wall, which is the median septum. The dorsal wall of a the preceding metatheca, th (n-1) serves as its ventral wall. The initial fuselli of the prothecal walls are thin, being represented by closely spaced growth lines, and they resemble in their thickness the preceding intercalary fuselli. As the prothecal walls develop, the fuselli gradually increase in thickness (Text-fig. 74n). At approxi mately the level of the preceding metathecal aperture, the fuselli at tain their maximum thickness, which is maintained to the thecal aperture. a At this same level, the protheca (th n ) no longer continues to devel op its o \ m dorsal wall (the median septum) because a new protheca b (th n ) of the adjacent thecal series takes over that function. How- a ever, the protheca (th n ) does begin to develop its own ventral wall (its infragenicular wall), which is continuous with the dorsal wall a (interthecal septum) of the preceding metatheca, th (n-1) . The protheca continues its growth using the median septum as its dorsal wall until it is at a level that is transverse to a thecal aperture b a of the opposing thecal series, th(n-l) . There, th n gives rise to a the th (n+1) protheca, and continues its growth towards the aperture as a metatheca developing its own dorsal wall (interthecal septum) and its own ventral wall (the free ventral wall). With the construction of a list around the margins of the apertural excavation, the thecal ontogeny is complete. It is not possible in any of the available specimens to determine the precise number of growth lines that constitute every proximal 527 theca, as Bulman (1945, 1947) has done for several diplograptid species. 1 However, in ithe specimens described above, the growth lines on th 1 2 2 are more closely spaced than those of th 1 ; in turn those on th 1 1 are more closely spaced than those on th 2 ; and 10 growth lines 1 comprise the portion of the th 2 free ventral wall that is distal to the apertural excavation (Text-fig. 75i-j). This distal decrease in growth line density also occurs in Climacograptus brevis as described by Bulman (1947). Remarks As mentioned above (See TYPE DATA), Lapworth’s (1880) holotype of Glyptograptus euglyphus can not be found, and thus a detailed comparison of Lapworth’s material to the specimens described above is not possible. However, there are no significant discrepancies between Lapworth's (1880) illustrations, his brief description, and the Alabama specimens figured and described here. The Alabama specimens agree closely with Elies and Wood’s (1907) description and illustrations of Diplograptus (G^,) teretiusculus var. euglyphus Lapworth. Because they illustrate (Text-fig. 172, PI. 31, fig. 2d) specimens from Lapworth's collections, it can be reasonably assumed that they examined Lapworth’s (1880) figured specimen. The Alabama specimens agree closely with Ruedemann's (1908, 1947) figured specimens of Diplograptus (G.) euglyphus, except for his specimens from Maine and Washington (Ruedemann, 1947; PI. 69, figs. 55-59). These specimens show a relatively wide proximal end and an 1 2 outward direction of growth for th 1 and th 1 ; thus, they differ 528 considerably from the Alabama specimens and all other described speci mens of Glyptograptus euglyphus. The Alabama specimens are similar to Harris and Thomas' (1935) figured specimens of Diplograptus (G.) cf. euglyphus. However, these Australian specimens are poorly preserved and poorly illustrated making an exact comparison difficult. Decker’s (1952) figured specimens of D. (G^.) euglyphus are similar to the Alabama specimens; however, the writer has examined many of the specimens on which Decker's (1952) fau- nal lists are based and has found them to often be distal fragments of many other species of Glyptograptus and Orthograptus. Berry's (1960, 1964) Texas and Norwegian specimens of £. euglyphus agree closely with the Alabama specimens. Glyptograptus euglyphus, as described and figured herein, is a most peculiar species. It is similar to a few other, primarily Llandoverian, species of Glyptograptus (e.g. G. tamariscus and G. 1 2 persculptus), in which the upward-directed parts of th 1 and th i are in close contact with the sicula and the proximal thecae are geniculate, but in these same features it differs greatly from other Ordovician glyptograptids (e.g. dentatus and £. teretiusculus). In fact, the species that G. euglyphus most closely resembles are Climacograptus brevis brevis Elies and Wood and G^. brevis mutabilis Strachan, which have been described on the basis of isolated material by Bulman (1947) and Strachan (1960), respectively. The proximal part of the rhabdosome of Glyptograptus euglyphus 529 is similar morphologically to the entire rhabdosome of Climacograptus brevis (includes both C^. brevis brevis and C^. brevis mutabilis in the following discussion), differing only in the inclination and length of the supragenicular wall. The rhabdosomal development of C^. brevis, which is unique among climacograptids, is identical to that of G. euglyphus. The similarities between (C. brevis and £. euglyphus suggest a close affinity of the two species. Although their stratigraphie ranges overlap, G^. euglyphus first appears in the murchisoni Zone (Berry, 1964), whereas C^. brevis does not appear until the gracilis Zone (Ragnar Nilsson, personal communication). Thus, it is postulated here that brevis evolved from G. euglyphus. The changes necessary for this evolution are an arrested development of growth when the rhab dosome was relatively short (1.5-2.0 mm long) and a decrease in the inclination of the supragenicular wall. The supragenicular walls in the most proximal thecae of Glyptograptus euglyphus resemble closely the supragenicular walls that are present throughout a rhabdosome of Climacograptus brevis ■ G^. euglyphus differs from C^. brevis by the progressive distal reduction of the supragenicular wall (or by the progressive proximal increase of the supragenicular wall). Because the thecae in mature parts of the rhabdosome of G^. euglyphus, are typically glyptograptid, which is the feature that the generic assignment is based on, the geniculate thecae represent a proximally progressive type of structural change within the rhabdosome. The distal spread of these geniculate thecae through the rhabdosome, which the evolution of C^. brevis from 530 iG. euglyphus involves, can be easily explained by Urbanek's (1973) theory of morphophysiological gradients within graptoloid colonies. Thus, by means of two biological processes, arrested development and a change in the morphophysiological gradient, G^. euglyphus could easily give rise to jC. brevis. Figured Specimens OSU 33271-OSU 33308. 531 Glyptograptus sp. cf. teretiusculus (Hisinger, 1840) (Text-figures 76-79) cf. 1840 Prionotus teretiusculus, Hisinger, p. 5, PI. 38, fig. 4. 1859 Graptolithus angustifoliusÇn.s.), Hall, p. 515, figure. cf. 1868a Diplograpsus angustifolius Hall, Nicholson, p. 525, PI. 19, figs. 8-9. cf. 1876 Diplograptus angustifolius Hall, Lapworth, PI. 2, fig. 35. cf. 1877 Diplograptus angustifolius Hall, Lapworth, p. 132, PI. 6, fig. 11. cf. 1882 Diplograptus teretiusculus (Hisinger), Tullberg, pi. 18-19, PI. 2, figs. 1-7. 1907 Diplograptus (Glyptograptus) teretiusculus(Hisinger), Elies and Wood, pi. 250-252, text-figs. 171a-d; PI. 30, figs.lla-b. 1908 Diplograptus (Glyptograptus) angustifolius(Hall) Ruedemann, p. 366-369, text-figs. 311-314, PI. 25, figs. 19-20. 1908 Diplograptus foliaceus var. alabamensis nov., Ruedemann, p. 352, PI. 25, fig. 3. cf. 1913 Diplograptus teretiusculus His., Hadding, p. 43-45, PI. 2, figs. 15-20. 1926 Diplograptus foliacens. Butts, PI. 23, figs. 7-8, 11-12. 1947 Diplograptus (Glyptograptus) teretiusculus(Hisinger), Ruede mann, p. 408-409, PI. 69, figs. 34-35. 1947 Diplograptus (Orthograptus) calcaratus var. alabamensis Ruedemann, Rudemann, p. 399, PI. 68, fig. 20. 1952 Diplograptus (Glyptograptus) teretiusculus(Hisinger) Decker, PI. 2, fig. 59; PI. 3A, fig. 18. 532 EXPLANATION OF TEXT-FIGURE 76 Text-figure 76 a-o. Glyp tog rap tu s sp. cf. G^. teretiusculus (Hisinger). a-i, m-o. Isolated specimens, j-1. Non-isolated specimens. a. Reverse aspect of specimen preserved in full relief. X15. PS-109.2. OSU 33309. b,c. Reverse and distal aspects of uncompressed distal stipe fragment shomng transverse rods at proximal end of median septum. Dashed circle marks position of trans verse rod on obverse side of sicula; dotted circle marks position of transverse rod on reverse side of sicula. X30. PS-102. OSU 33310. d,e. Obverse and reverse aspects of fully compressed specimen. X15. PS-76.4. OSU 33311. f,g. Obverse and reverse aspects of partly compressed specimen. X15. PS-109.2. OSU 33312. h,i. Reverse and obverse aspects of partly compressed specimen. X15. PS-109.7. OSU 33313. j . Reverse aspect of non-isolated specimen. X3.6. PF-17.3. OSU 33314. k. Reverse aspect of non-isolated specimen. X3.6. PF-17. OSU 33315. 1. Obverse aspect of non-isolated specimen. X3.6. PF-17.1. OSU 33316. m. Ventro-lateral aspect of obliquely compressed distal fragment of rhabdosome. X15. PS-109.2. OSU 33317. n. Lateral aspect of fully compressed distal fragment of rhabdosome. X15. PS-109.2. OSU 33318. o. Lateral aspect of specimen broken to reveal interior of rhabdosome. Note poorly preserved median septum, consisting of only septal membrane and virgula. X30. PS-109.2. OSU 33319. 533 ,--IW Text-figure 76 534 EXPLANATION OF TEXT-FIGURE 77 Text-figure 77 a-k. Glyptograptus sp. cf. G^. teretiusculus (Hisinger). Isolated specimens representing early growth stages. a. Obverse aspect of sicula with initial bud of th 1^. X37.5. PS-109.2. OSU 33320. b,c. Reverse and obverse aspects of sicula with initial bud of th 1^ and initial development of th 1^. X37.5. PS-117. OSU 33321. d. Reverse aspect of sicula with th 1^ and horizontally directed opening for th 1^. X37.5. PS-102. OSU 33322. e. Reverse aspect of sicula with th 1 and th 1 . X37.5. PS-109.2. OSU 33323. f,g. Reverse and obverse aspects of sicula with th 1^ and downward-directed th 1 . X37.5. PS-109.2. OSU 33324. h. Reverse aspect of specimen with downward-directed process from th 1^ forming two apertures, one for th 1^ and one for th 2l. X37.5. PS-117. OSU 33325. i. Reverse aspect of specimen showing horizontally directed parts of th 2^ and th 1^. X37.5. PS-126. OSU 33326. j,k. Reverse and obverse aspects of specimen showing growth of th 1 and th 2^. Note that th 2^ fills U-shaped space between limbs of th 1 on reverse side of specimen, but leaves an opening for th 2^, X37.5. PS-117. OSU 33327. Ui w U l Text-figure 77 536 EXPLANATION OF TEXT-FIGURE 78 Text-figure 78 a-j. Glyptograptus sp. cf. G^. teretiusculus (Hisinger). Isolated specimens representing early growth stages. 2 a. Reverse aspect of specimen showing opening for th 2 . X37.5. PS-126. OSU 33328. b,c. Obverse and reverse aspects of specimen showing develop ment of th 2^. Note that growth lines on infragenicular and left-lateral walls of th 2^ are continuous. X37.5. PS-117. OSU 33329. d,e. Obverse and reverse aspects of specimen showing development of th 2^ and th 2^. X37.5. PS-109.2. OSU 33330. f. Reverse aspect of specimen showing upward growth of th 22. X37.5. PS-117. OSU 33331. g,h. Obverse and reverse aspects of specimen showing develop ment of th 2^ and th 22, X37.5. PS-109.2. OSU 33332. i,j. Obverse and reverse aspects showing growth of th 2^ and th 22. Note dorsal wall of th 2^ and lateral extent of th 22. X37.5. PS-109.2. OSU 33333. fh 2i^ th2 th 2l thZ th ' 2 th 2 th 2 Olw Text-figure 78 538 EXPLANATION OF TEXT-FIGURE 79. Text-figure 79 a-j. Glyptograptus sp. cf. G^. teretiusculus (Hisinger). Isolated specimens showing development of rhabdosome. a,b. Reverse and obverse aspects of specimen showing development of th 3^ protheca. Note prominent list marking dorsal margin of th 3 aperture. X18.8. PS-109.2. OSU 33334. c,d. Obverse and reverse aspects of specimen showing development of th 22. X37.5. PS-109.2. OSU 33335. e,f. Obverse and reverse aspects of specimen showing development of th 2^ metatheca and th 3^ protheca. Intercalary fuselli denote budding of th 3^ from th 2^. Prominent list forms dorsal margin of th 2^ aperture. X37.5. PS-117. OSU 33336. g,h. Reverse and obverse aspects of specimen showing development of th 2^ and th 3^. Note median septum on reverse side of sicula. X37.5. PS-109.2. OSU 33337. i,j. Reverse and obverse aspects of specimen showing development of th 4^ protheca. X37.5. PS-109.2. OSU 33338. th 3 ' Ln LO Text-figure 79 VO 540 1952 Diplograptus (Orthograptus) calcaratus var. alabamensis Ruedemann, Decker, Pl. 1, fig. 40; Pl. 2, fig. 63. ?1960 Glyptograptus teretiusculus (Hisinger) , Berry, p. 87-88, Pl. 14, figs. 3-5, 8; Pl. 16, fig. 5. ?1960 Orthograptus calcaratus var. alabamensis(Ruedemann), Berry, p. 89-90. I960 Glyptograptus cf. teretiusculus (Hisinger, 1840), Jaanusson, p. 322, Pl. 3, figs. 10-11. 1963 Glyptograptus cf. G. teretiusculus (Hisinger), Ross and Berry, p. 142, Pl. 10, fig. 19; Pl. 11, fig. 5. 1964 Glyptograptus teretiusculus(Hisinger, 1840), Berry, p. 147-151, Pl. 15, figs. 4-6. Type Data The lectotype of Glyptograptus teretiusculus (Hisinger) was designated by Tullberg (1882, Pl. 2, fig. 1) and is stored in the Swedish Museum of Natural History, Stockholm. This specimen, which was examined by the writer, is a barely visible carbon film on a surface of black shale, and it represents only a distal fragment of a rhabdosome. Berry (1964) reports that this specimen is 7 mm long, consists of 6 pairs of thecae, and has a maximum width of 2.5 mm. Diagnosis A species of Glyptograptus with 6-7 thecae in 5 mm proximally and 4.5-5 distally; rhabdosome width gradually increases distally, 0.9-1.2 mm at th 1, maximum width of 2.0-2.3 mm at fifteenth pair of thecae; first two thecae with subapertural spines; sicular aperture with virgel- la and two short, anti-virgellar spines; median septum complete, straight. Material The available material consists of more than 850 specimens that represent the following states of preservation: 541 1) Isolated specimens. Approximately 250 specimens were obtained from the Pratt's Syncline section of which 130 are early growth stages, 80 are young rhabdosomes or fragments of large rhabdo- somes with proximal ends and more than two pairs of thecae, and 15 are distal fragments. Most of the early growth stages are partly or completely compressed, yet most of them show growth lines at least in the youngest (most distal) thecae. Most of the large rhabdosomes are compressed and have a thick, black glossy,cortical layer 2) Non-isolated specimens with periderm preserved. More than 200 specimens were obtained from the Pratt's Ferry and Pratt's Syncline sections. Many of these specimens are in partial relief. 3) Carbon Films. More than 400 specimens were obtained from the Calera section. Description Isolated specimens. The largest available specimen with a proxi mal end is 5 mm long and consists of six pairs of thecae (Text-figs. 76f-i). Specimens preserved in full relief have an oval cross-section. In profile aspect, the ventral walls of the rhabdosome are sub-parallel, the proximal end is wide, and the thecae are alternating. The sicula is visible only on the obverse side of the rhabdosome 2 to a level slightly above the aperture of th 1 . Its aperture is fur nished with a 0.7 mm long virgella and two 0.2 mm long spines, which 1 are situated on the anti-virgellar side of the aperture. Th 1 ex- 542 tends downward along the virgella to a level below the sicular aperture then bends outward and upward in a U-shape. Its aperture is at a level 2 only slightly above that of the sicular aperture. Th 1 is also U- 1 1 shaped, and its aperture is at a level above that of th 1 . Both th 1 2 and th 1 are furnished with 0.5 mm long subapertural spines. The rhabdosome is 0.9-1.1 mm wide at the level of th 1 with the width being somewhat greater in compressed specimens than in uncompressed specimens. Distally, the width increases to 1.1-1.4 mm at the fifth pair of thecae. The widest available distal fragment of a rhabdosome is 3.2 mm wide. Among the available isolated specimens, the thecae number 3 in 2 mm and 4.5-5 in 3 mm proximally, and 3 in 3 mm and 4.5-5 in 5 mm distally. The thecae are of the glyptograptid type with sigmoid curvature of the free ventral wall. In thecae in proximal parts of the rhabdo some, the proximal end of the free ventral wall is concave, whereas the distal part, which is twice as long as the proximal part, is con vex (Text-figs. 76d-g, o). A geniculum with a very obtuse genicular angle is present where the convex and concave portions of the free ventral wall meet, and in proximal thecae this geniculum defines the distal limit of the apertural excavation. In distal portions of the rhabdosome, the free ventral walls are only slightly sigmoidal (Text- fig. 76n) , and in some instances they may be straight. Proximally (at th 3), the thecae are 0.8 mm long; at th 5, they are 1.2 mm long; and in the widest distal fragment, they are 1.7 mm long. Throughout the entire length of the rhabdosome the thecae consistently overlap 543 for a third their length, and they are inclined at 30 degrees to the axis of the rhabdosome. In compressed specimens, the free ventral wall is often distorted in such a way that it nearly parallels the axis of the rhabdosome. In profile aspect, each apertural excavation appears as a V-shaped notch, in which one side of the notch that is defined by the apertural margin, is perpendicular to the axis of the rhabdosome. The width of the thecal aperture is 0.2 mm in uncompressed specimens and as much as 0.3 mm in compressed specimens. Distally, the width of an aperture increases. The apertures in the widest available distal fragment are 0.9 mm wide. Thus throughout the entire length of the rhabdosome, the apertural excavations occupy less than a third the width of the rhabdosome. In profile aspect, the ventral part of the lateral margin of the aperture is oriented in a dorso-distal direction, whereas the dorsal part is oriented in a ventro-distal direction. The dorsal and ventral parts of each lateral apertural margin meet in an obtuse angle forming a lappet. In proximal thecae, these lappets are at the midpoint of the lateral apertur .1 margins (Text-figs. 76a, d-i). In distal thecae, the lappets are close to the ventral apertural margin (Text-figs. . 76m-n). 1 Beginning at the level of the th 2 aperture, a straight median septal groove extends the entire length of the rhabdosome. Interthe- cal septal grooves extend from the dorso-lateral corners of the 544 apertural excavations toward the median septal groove; however, they do not join with the median septal groove but stop short of it. The median septum consists of the virgula, septal periderm, and only two transverse rods. The transverse rods, which extend between the sicula and the lateral walls of the rhabdosome are only at the proximal end of the median septum (Text-fig. 76b-c). The transverse 2 rod on the obverse side of the sicula is at the level of the th 2 aperture, whereas the transverse rod on the reverse side of the 1 sicula is at the level of the th 2 aperture. Thus, the proximal end of the median septum is at a different level on opposite sides of the sicula. Distal to the transverse rods, the median septum consists only of a virgula and septal periderm (Text-fig. 76o). The septal membrane is not preserved well enough in the available specimens to reveal fusellar structure. The membrane seems to be weak because in most specimens that are broken in such a way that the interior of the rhabdosome is revealed, the median septum is only represented by pieces of periderm attached to the two transverse rods or to the lateral walls of the rhabdosome. Non-isolated Specimens and Carbon Films. The longest available specimen is 9.5 cm long and lacks a proximal end. The morphology of the non-isolated specimens (Text-figs. 76j-1) resembles closely that described above for isolated specimens. However, in many cases, sediment covers portions of the thecae and/or the specimens are com pressed and distorted in such ways that the apertures appear to be introverted, everted, and/or denticulate and the free ventral walls appear to be straight, sigmoidal, and/or geniculate. 545 In contrast to the isolated specimens, which represent only short fragments of rhabdosomes, the non-isolated specimens are large and can be systematically measured. In many of the very large specimens, the virgella approaches 10 mm in length. At the first pair of thecal aper tures, the rhabdosome is 0.9-1.2 mm wide. Distally, it increases in width to 1.3-1.6 mm at th 5, 1.6-2.2 mm at th 10, and 2.0-2.3 mm at th 15. Most of the available specimens are 2.5-3.5 cm long, and after the fifteenth pair of thecae (approximately 13 mm from the proximal end), the rhabdosome is parallel-sided. However, in some of . the larger specimens, the width continues to increase very gradually to as much as 2.9 mm at a distance of 4.5 cm from the proximal end. Proxi mally, the thecae number 4.5-5 in 3 mm, 6-7 in 5 mm, and 11.5-12 in 10 mm. Distally, the thecae number 4.5-5 in 5 mm and 9-10 in 10 mm. Early-Growth Stages and Rhabdosomal Development. Morphology of Sicula. The sicula is conical and 1.0-1.3 mm long. The prosicula accounts for a fourth of the length of the sicula, ex hibits longitudinal threads (Text-fig. 77f-g), and a nema (Text-fig. 77h), and is 0.1 mm wide at its aperture. From its contact with the prosicula, the metasicula widens to 0.3 mm at its aperture. It is composed of transverse growth lines, and its aperture is furnished with a virgella and two anti-virgellar spines (Text-fig. 77a). The virgella forms by the downward curvature of growth lines along one side of the sicula, and its proximal end is at a point two-third of the distance from the sicular aperture to the prosicula. Two 0.2 mm long spines are present on the anti-virgellar margin of the sicular 546 aperture, and the dorsal margin of the aperture, which is situated between the anti-virgellar spines, is concave in a dorsal aspect. The anti-virgellar spines are composed of fusellar tissue (Text-fig. • 77a), and judging from the growth line pattern, they develop when the metasicula is nearly complete. Development of Proximal End. The proximal-end development of the specimens described herein resembles closely that of Glyptograptus dentatus (Brongniart), which has been described by Bulman (1963) arid Skevington (1965). Their descriptions are based on a small number of specimens, and the large number of early growth stages that are avail able in the present study permit a more detailed description of the proximal-end development. 1 2 The initial development of th 1 and the budding of th 1 are similar to that described in other diplograptids and in the dicranograp- 2 teods of the present study (Text-figs. 77a-d). After th 1 buds, th 1 1 grows downward along the virgella to a level below the sicular aper-r ture, where it bends and grows outward and upward forming a U-shape 1 (Text-figs. 77d-h). When th 1 is complete, its aperture is slightly above the sicular aperture. 2 Th 1 initially grows in an obliquely upward direction to the re verse side of the sicula, where it bends and grows downward (Text-figs. 2 77d-g). As th 1 grows downward, it expands to form a hood, and from the midpoint on the margin of this hood, a process grows downward and inward to the sicula (Text-fig. 77h). This process consists of fusellar tissue, and when it meets the sicula, two apertures are formed. 1 Th 2 develops from the aperture on the left side of this process. 2 Th 1 continues developing from the aperture on the right side of the 547 process. It grows across the reverse side of the sicula, then bends and grows upward forming a U-shaped tube (Text-figs. 77h-k, 78a-f). 2 When th 1 is complete, its aperture is significantly higher than the 1 aperture of th 1 . 2 Before th 1 begins its upward direction of growth, growth of th 1 1 2 commences. Th 2 initially grows in a slightly downward direction toward the space between the downward-and upward-oriented limbs of th 1 1 (Text-figs. 77i-k, 78a). 1 Distally, the subhorizontally oriented part of th 2 meets the right-lateral margin of the dorsal wall of the upward-directed limb 1 of th 1 , and it forms the right-lateral wall of the upward-directed 1 1 1 part of th 2 . The relationship between th 2 and th 1 is such that 2 an opening, from which th 2 will develop, is formed below the sub- 1 horizontally directed part of th 2 . The dorsal wall of the base of 1 2 the U-shaped part of th 1 also borders the opening for th 2 . 1 Growth of th 2 continues with the development of a left-lateral and a ventral wall. Growth lines of these two walls are continuous 1 and indicate that at this stage the development of th 2 consists only of the growth of a left-lateral wall and a ventral wall (Text- fig. 78b-c). However, when the left-lateral and ventral walls reach the level of the upper margin of the horizontally directed part of 1 th 2 , growth of the right-lateral wall again commences (Text-fig. 2 78d-e) . Also at this stage, th 2 begins growing obliquely upward from 1 its opening at the base of th 2 . 1 Th 2 continues to grow directly upward with the sicula serving as 2 1 its dorsal wall until the level at which th 1 originated from th 1 548 1 (Text-fig. 781—j). At this level, th 2 begins to develop its own 1 1 dorsal wall (the interthecal septum between th 2 and th 3 ) in such a way that a space is left between this dorsal wall and the sicula 1 (Text-fig. 78i—j). At a later stage, th 3 will occupy this space. 1 When th 2 completes its aperture, its dorsal apertural margin is thickened in such a way that it forms a prominent list (Text-fig. 79e-f). 2 The upward growth of th 2 is delayed relative to the growth of 1 2 th 2 (Text-figs. 78d-j, 79a-f). Th 2 grows upward as a narrow tube 2 until it reaches the dorsal wall of th 1 , after whch it expands laterally in such a way that it extends to the dorsal margin of the 1 right-lateral wall of th 2 (Text-fig. 78i-j). With continued growth, 2 2 th 2 grows past the aperture of th 1 and begins to form a ventral wall as well as a right-lateral wall that is in contact with the ob verse side of the sicula (Text-fig. 79c-d). At this stage of 2 development, the left-lateral wall of th 2 extends dorsally to the 1 right-lateral margin of the dorsal wall of th 2 . Thus, the dorsal part 2 of the left-lateral wall of th 2 forms the right-lateral wall of the 1 initial part of th 3 . With continued growth, the right-lateral wall 2 of th 2 expands laterally, growing further around on the obverse side 2 of the sicula. The left-lateral wall of th 2 , which remains in con- 1 tact with th 2 , develops intercalary fuselli in its dorsal part (Text-fig. 79e-f). The appearance of these intercalary fuselli denotes 1 2 the budding of th 3 from th 2 , and it is accompanied by the develop- 1 ment of a left-lateral wall for th 3 . 549 1 The right-lateral wall of th 3 does not become distinct from the 2 1 left-lateral wall of th 2 until th 3 reaches the level of the 1 aperture of th 2 (Text-fig. 79g-h). At this point, a transverse rod, which serves as the proximal end of the median septum on the reverse side of the sicula, develops between the sicula and the left-lateral 2 wall of th 2 . The appearance of the transverse rod is accompanied 1 by 1) the development of a ventral wall by th 3 , which is continuous 1 with the dorsal margin of the th 2 aperture; 2) by the growth of the 1 dorsal margin of the left-lateral wall of th 3 in such a way that it comes in contact with the dorsal margin of the right-lateral wall of th 2 2 ; and 3) by an abrupt decrease in the dorsal extent of the lateral 2 2 walls of th 2 in such a way that a space is left between th 2 and 2 2 the sicula. Th 3 will later occupy the space between th 2 and the sicula, and the formation of this space is probably simultaneous with 2 2 2 the budding of th 3 from th 2 . The metatheca of th 2 does not have a dorsal wall. Specimens that represent more advanced growth stages are rare. 2 Those available (Text-figs. 79a-b, i-j) indicate that th 3 buds 2 1 1 from th 2 , and th 4 buds from th 3 . Judging from the position of the transverse rod and most proximal appearance of the median septal groove, the medial septum initially develops on the obverse side of 2 the sicula at a level slightly above the th 2 aperture (Text-figs. 76b-c, 79i-j). Thus, its initial appearance may be associated with 2 the budding of th 3 . 550 12 2 With the budding of th 3 and th 3 from th 2 , the two thecal series become separated. The proximal-end development with four 2 crossing canals, a dicalycal th 2 , and primarily upward directions 1 2 of growth of th 2 and th 2 is of the prosoblastic diplograptid type. Development of Median Septum. As described above, the median septum initially develops with the appearance of transverse rods at different levels on opposite sides of the sicula. The available growth stages do not reveal the subsequent stages of development. The position of the median septum relative to the thecae indicates that the median septum serves as the dorsal wall of the prothecae and sug gests that it is formed as the dorsal wall of successive prothecae. Thecal Ontogeny. The ontogeny of the most proximal five thecae is affected greatly by their relationship to the sicula and their mode of budding. However, more distally the thecal ontogeny probably occurs in a regular manner. Growth stages that show the thecal ontogeny of dis tal thecae are rare, and the available specimens (Text-figs. 79a-b, i-j) suggest the following stages in the thecal ontogeny; 1) Initially the th n protheca develops lateral walls and possi bly a dorsal wall (the median septum) . It has no ventral wall because the th n-1 metatheca does not develop a dorsal wall. However, the th n-1 metatheca does develop a dorsal apertural margin that is represented by a prominent list. 2) When the th n protheca reaches the list that represents the dorsal apertural margin of th n-1, it begins to develop its 551 own ventral wall. Whether or not it still develops its own dorsal wall (the median septum) can not be determined in the available specimens. 3) After the budding of th n+1, th n develops without a dorsal wall but with ventral and lateral walls. Growth lines on the lateral walls of the initial part of the protheca are closely spaced, whereas growth lines on the walls of more distal parts of the protheca and the metatheca are widely spaced. This suggests that the rate of growth of a theca is initially slow, but at some level in the distal part of the protheca, the rate of growth increases. Remarks Rhabdosomal Development. The proximal-end development of the spec imens described herein resembles closely that described for Glyptograptus dentatus by Bulman (1963) and Skevington (1965). Additional similari ties between G^. dentatus and the specimens described herein are as follows: 1) The median septum initially appears on the obverse side of the 1 sicula at the level of the proximal part of the th 3 protheca. 2) The metathecae lack dorsal walls; the interthecal septa are nearly non-existent; and the ventral wall of a metatheca be gins at a list that is the dorsal margin of the preceding thecal aperture. 3) The fuselli on the most proximal parts of the lateral walls of a protheca are closely spaced, whereas those on more distal parts of the protheca and on the metatheca are widely spaced. 552 4) The metathecal portion of each theca has a glyptograptid curvature with a weakly developed geniculum situated low on the ventral wall, and the thecal aperture has paired lateral lappets. These similarities suggest that, besides the proximal-end development, the development of the median septum and the thecal ontogeny of the specimens described above are very similar to those of G^. dentatus • Taxonomic Assignment. The specimens described above resemble closely, and are considered herein to be nonspecific with, the holotype (NYSM7145) of Orthograptus calcaratus alabamensis (Ruedemann, 1908, 1947). However, in rhabdosome shape, proximal-end characteristics, and thecal form, the specimens described above also agree closely with Glyptograptus teretiusculus(Hisinger) and with Diplograptus angustifol- ius (Hall) , which has been synonymized with G^. teretiusculus. As it is reported in the literature, Glyptograptus teretiusculus has a long stratigraphie range and a worldwide geographic range. Throughout its stratigraphie and geographic ranges, the over-all morphologic features of G^. teretiusculus are constant; however, quan titative variation is present. This quantitative variation has never been systematically examined, and in some instance (i.e. Hadding, 1913) the concept of the species has been expanded to include all this quantitative variation. In some instances, this variation may be re lated to the maturity or astogenetic age of the rhabdosome. For exam ple, Ruedemann's (1908, 1947) and Hall's (1859) figured specimens of G. teretiusculus, which are narrower than the Alabama specimens, are also 553 shorter, whereas the thecal densities are comparable. In other in stances (Hadding, 1913; Berry, 1960, 1964), the variation may be infraspecific because the greatest reported variation is among speci mens from the same locality. Yet, because of its range, geographic and temporal variant forms (possibly subspecies) of G^. teretiusculus would be expected. A systematic examination of the quantitative varia tion in teretiusculus, which is necessary to determine the signi ficance of the variation, is outside the scope of the present study. Therefore, the specimens described above are placed in open nomenclature. Most of the previously reported specimens of Glyptograptus teretiusculus are poorly preserved and/or not described precisely enough for detailed comparison with the Alabama specimens. However, Jaanusson (1960) described isolated specimens, which he referred to as G^. cf. teretiusculus. Jaanusson’s specimens differ slightly from the Alabama specimens in thecal density in distal parts of the rhabdo some (5.5-6 in 5 mm in Jaanusson's specimens; 4.5-5 in 5 mm in Alabama specimens) and in the lack of anti-virgellar spines. The fact that Jaanusson’s specimens are shorter than the Alabama specimens could account for the differences in thecal density. The taxonomic significance of the presence or absence of anti-virgellar spines is not known; at the most, Jaanusson's specimens and the Alabama specimens might be geographic variant forms. The Alabama specimens have a larger maximum width than Hall’s (1859) and Ruedemann’s (1908, 1947) specimens of Glyptograptus teretiusculus, which were previously assigned to Diplograptus angustifolius (Hall). 554 However, their specimens are usually no more than 3 cm long, which may account for the differences in maximum width. Additionally, one of the several specimens that are on the shale chip with Hall’s (1859) syntype of Diplograptus angustifolius (AMNH 1033/1) shows anti- virgellar spines on the sicula, Figured Specimens OSU 33309-OSU 33338, 555 Genus Orthograptus Lapworth, 1873 Type Species: Orthograptus quadrimucronatus(Hall, 1865) Diagnosis As in Bulman (1970). Orthograptus sp. (Text-figure 68d, g-h) Diagnosis A species of Orthograptus with rhabdosome 1.0 mm wide at proximal end, widening rapidly to 4 mm at a distance of 10 mm from the proximal end, thereafter of uniform width; thecae numbering 10 in 10 mm distally with apertural margin perpendicular to rhabdosome axis and free ventral wall inclined at 20 to 30 degrees. Material The available material consists of five poorly preserved specimens from the two stratigraphically highest collections in the Pratt's Syncline section. These specimens are fully compressed on shale sur faces, and the periderm is heavily carbonized. Description The largest specimen is 3.5 cm long, but it lacks much of the proximal end. In fact, the most proximal part of the rhabdosome is not 556 preserved in any of the available specimens. The most proximal theca in the specimen illustrated in Text-figure 68g bears a spine, and this 1 theca might be th 2 . The rhabdosome, which is 1.0 mm wide at the level of the most proximal theca in Text-figure 68g widens rapidly to a width of 4.0 mm at a distance of 10 mm from the proximal end. Thereafter, the rhabdosome width is constant. In proximal parts of the rhabdosome, the thecae are alternating. However, distally, the corresponding thecal apertures of the two thecal series are situated at equal distances from the proximal end. The thecae, which number 10 in 10 mm in distal parts of the rhabdosome, have a straight apertural margin that is perpendicular to the axis of the rhabdosome, although in some instances probably due to the state of preservation, the aperture appears to be slightly everted or intro verted. The free ventral wall is inclined 20 to 30 degrees to the axis of the rhabdosome and increases in length from 0.5-0.6 mm in proximal thecae to 1.5 mm in distal thecae. In distal parts of the rhabdosome, the thecae overlap half their length. Remarks The specimens described above resemble Orthograptus calcaratus, particularly the subspecies calcaratus acutus Elies and Wood, in the rapid widening of the rhabdosome, the rhabdosome width, the thecal density, and the thecal inclination. However, these specimens are not assigned herein to that species or subspecies because they lack the proximal end of the rhabdosome. Figured Specimens OSU 33342-OSU 33344. 557 Family lASIOGRAPTIDAE Lapworth, 1879 Diagnosis As in Bulman (1970). Genus Lasiograptus Lapworth, 1873 Type Species: Lasiograptus cestatus Lapworth, 1873 Diagnosis As in Bulman (1970) Lasiograptus sp. (Text-figure 80 i-k) Diagnosis Rhabdosome parallel-sided, 0.9-1.2 mm wide; thecae of lasiograptid type with genicular spines, number 6 -8 in 3 mm proximally; apertural excavations occupy a third of rhabdosome width and a fourth to a third of free ventral wall. Material The available material, which was obtained from the Calera section, consists of 34 specimens that are preserved as carbon films. Description The largest available specimen with a proximal end is 3 mm long and consists of 6 pairs of thecae. In profile aspect, the rhabdosome 558 EXPLANATION OF TEXT-FIGURE 80 Text-figure 80 a-h. Dicaulograptus? n. sp. A. a-e, h. Isolated specimens, f-g. Non-isolated specimens, all specimens fully compressed. a,b. Reverse and obverse aspects. Note torsion of stipes, distal separation of stipes, and thecal morphology. X18.8. PF-25. OSU 33339. c. Fragment of rhabdosome showing thecal aperture. X37.5. PF-22.5. OSU 33340. d. Reverse aspect of sicula xdLth th 1^ and th 1^. X37.5. PF-22. OSU 33341. e. Ventral aspect of proximal end of rhabdosome. Note convex ventral wall and apertural and mesial spines. X37.5. PF-29. OSU 33342. f. Obverse aspect. X18.8. PS-29. OSU 33343. g. Obverse aspect. X18.8. PS-29. OSU 33344. h. Distal stipe fragment. Note two separate stipes. X37.5. PF-22. OSU 33345. 80 i-k. Lasiograptus sp. Non-isolated, compressed specimens. i. Reverse aspect. X18.8. C-32. OSU 33346. j. Obverse aspect. X18.8. C-4.6. OSU 33347. k. Obverse aspect. Note downward growth of th 1^ along virgella. X18.8. C-32. OSU 33348. 559 Text-figure 80 560 is parallel-sided and 0.9-1.2 mm wide. It narrows slightly at the two most proximal pairs of thecae, and it is 0.-75-0.90 mm wide at the level of the first pair of thecal apertures. 1 In profile aspect, the proximal end is wide. Both th 1 and th 2 1 are U-shaped and exhibit mesial spines. A characteristic feature of the proximal end is the distance below the sicular aperture to 1 1 which th 1 grows before it bends upward. The aperture of th 1 is situated only a short distance above the level of the sicular aperture. Because of the poor state of preservation of the specimens, the thecal morphology is difficult to evaluate. The thecae number 6-8 in 3 mm proximally. The proximal part of the free ventral wall between the mesial spine and the apertural excavation is horizontal (trans.) or highly inclined in a ventro-distal direction. The spine on each theca is interpreted herein to project from the geniculum, in which case the supragenicular wall is short and inclined in a dorso-distal direction. The thecae can thus be referred to as the lasiograptid type. The thecal apertures may be introverted. The apertural excava tions are slit-like and in most instances inclined in a ventro-distal direction. They are 0.2-0.4 mm wide (trans.) and 0.15-0.20 mm high (long.), thus they occupy up to a third of the rhabdosome width and a fourth to a third of the length of the free ventral wall. The specimens are not preserved well enough to reveal the presence of lists or septal grooves. Remarks The specimens described above are referred to the genus 561 Lasiograptus because of thecal characters, but they are too poorly preserved to justify a specific assignment. Superficially, the thecae in the specimens described above resemble those in specimens referred to as Lasiograptus sp. by Berry (1964). However, Berry's specimens differ from those described herein in most quantitative features. The specimens described above resemble I., pusillus Ruedemann, 1947 in the small size of the rhabdosome and the high thecal density, but again, the Alabama specimens are too poorly preserved for a close comparison. Figured Specimens OSU 33346-OSU 33348. 562 Family DICAULOGRAPTIDAE Bulman, 1970 Diagnosis As in Bulman (1970). Discussion As defined by Rickards and Bulman (1965) , the family Dicaulograptidae and the genus Dicaulograptus are monotypic, being based on a single species, hystrix. The specimens described below are very similar in thecal characteristics and proximal-end morphology to D. hystrix and could possibly be assigned to the genus Dicaulograptus; alternatively} because of the peculiar shape of the rhabdosome, a new genus within Dicaulograptidae could be established for these specimens. Either action would necessitate a revision of the familial or generic diagnoses. Because the specimens described below are not assigned a formal specific name, no revisions are made herein of Bulman*s (1970) diagnoses of Dicaulograptidae and Dicaulograptus. Genus Dicaulograptus Rickards and Bulman, 1965 Type Species: Dicaulograptus hystrix(Bulman, 1932) Diagnosis As in Bulman (1970). 563 Dicaulograptus? n. sp. A (Text-figures 80a-h) Diagnosis Rhabdosome with 0.7-0.8 mm wide biserial dipleural proximal portion and two spiral-shaped uniserial stipes distally; thecae number 2 in 1 mm, of dicranograptid? type with strongly introverted and introtorted aperture, furnished with mesial spine and paired apertural spines. Material The available material represents the following states of preser vation: 1) Isolated specimen. 40 specimens were obtained from the Pratt's Ferry section and 22 specimens were obtained from the Pratt's Syncline section. These specimens are poorly preserved. They are fully compressed, heavily carbonized, and very fragmentary. 2) Non-isolated specimens with periderm preserved. Approxi mately 140 specimens were obtained from the Pratt's Ferry and Pratt's Syncline sections with 95 percent of the speci mens coming from intervals 29 meters above the base of the Pratt's Syncline section and 25 meters above the base of Pratt's Ferry section. The preservation of these specimens is as poor as that of the isolated specimens. Description The largest available rhabdosome (Text-fig. 80g), which is not isolated, is 2.8 mm long and consists of 6 pairs of thecae. The 564 largest available isolated rhabdosome is 2 mm long and consists of 4 pairs of thecae. Proximally, the rhabdosome is biserial with the thecal series in a dipleural arrangement. Distally, beginning at the third pair of thecae, the thecal series begin to rotate in a clockwise direction (when the rhabdosome is viewed proximally) around the axis of the rhabdosome (Text-figs. 80a-b, f-g), and more distally they separate in such a way that they form two uniserial stipes (Text-figs. 80a-b, h). The biserial portion of the rhabdosome is 0.7-0.8 mm wide. More, distally, the width is difficult to measure because of the spiral shape of the stipes, however it does not seem to increase significantly. In most specimens, the uniserial stipes are in contact with one another, but in a few specimens, representing distal stipe fragments (Text-fig. 80h), the stipes are oriented in such a way relative to one another that they probably were not in contact in their original shape. The proximal end of the rhabdosome in profile aspect is wide and 1 2 truncated. Th 1 and th 1 grow directly outward away from the sicula at the level of the sicular aperture in such a way that their mesial spines are horizontal. The virgella is 0.6 mm long, and the mesial spines of the first two thecae are 0.4 mm long. The thecae superficially appear to be of the dicranograptid type with very introverted and introtorted apertures. However, there is no well defined geniculum, and there appears to be no introversion of the aperture. From the preceding thecal aperture, the free ventral wall extends in a ventro-distal direction to the mesial spine. In 565 2 1 some specimens (e.g. th 2 and th 4 in the specimen illustrated in Text-fig. 80a-b) there is a bend in the free ventral wall at the distal end of the apertural excavation of the preceding thecal aperture. It is doubtful that this bend is a geniculum. Distal to the mesial spine, the free ventral wall is oriented in a dorso-ventral to dorso-distal direction, and here again the bend in the free ventral wall can not really be considered a geniculum. The thecal apertures 1 are introtorted (Text-fig. 80c and th 3 of specimen in Text-fig. 80a-b) , however they are not introverted. The distal part of the free wall forms a "lid" over the aperture. Without this "lid" the aperture would open in a distal direction, but because of the "lid" the aperture opens in a dorsal direction. In addition, paired spines extend laterally from the ventro-lateral comers of the aperture where the lateral margins of the aperture merge with the lateral margins of the dorso-ventrally oriented part of the free ventral wall. The thecae, which number 2 in 1 mm in proximal parts of the rhabdosome, are 0.6-0.8 mm long and overlap less than a third their length. The free ventral wall is inclined at 40 degrees to the axis of the rhabdosome in the biserial portion of the rhabdosome. The mesial and apertural spines of the thecae are as long as 0.6 mm and 0.5 mm, respectively. In the biserial portion of the rhabdosome, the apertural excavations, which are oriented in ventro-distal directions, occupy less than a fifth of the rhabdosome width and a fourth of the length of the free ventral wall. One of the available specimens (Text-fig. 80d) is an early growth 566 1 stage. Although it is poorly preserved, th 2 appears to originate 1 from the left side of th 1 . This feature and the overall appearance of the specimen suggest that the proximal-end development is of the diplograptid type. The sicula is more than 0.8 mm long, and its aper ture, which exhibits no ornamentation except for the virgella, is 0.2 mm wide. Remarks , The generic assignment of the specimens described above is diffi cult to determine. The rhabdosome shape suggests an assignment to Dicranograptus, yet the thecal characteristics differ greatly from those of known dicranograptids. Although the stipes are uniserial distally, in most instances they are in contact with one another. The spiral shape of the uniserial stipes and the superficial shape of the thecae are suggestive of Dicellograptus bispiralis bispiralis. However, the dorsal wall of the introtorted part of the theca in D. bispiralis is strongly introverted. The thecae in D^. bispiralis display a distinct geniculum, and the overall proportions of the rhabdosome of D^. bispiralis are much larger than those of the speci mens described above. It seems possible, though highly unlikely, that the species represented by the specimens described above may have evolved from D^. bispiralis. The necessary changes for this would involve a decrease in the introversion of the thecal aperture, a modification of the lateral apertural lappets into lateral apertural spines, and a merging of the two stipes into a biserial, dipleural arrangement in the proximal part of the rhabdosome. The specimens described above most closely resemble specimens of 567 Dicaulograptus hystrix (Bulman). This species was established by Bulman (1932) as Lasiograptus hystrix. Rickards and Bulman (1965) established the monotypic genus Dicaulograptus on the basis of this species. D. hystrix is very similar to the specimens described above in its overall proportions (rhabdosome width and thecal density) and appearance (shape of apertural excavations, free ventral wall, intro torted thecal aperture, and thecal spines). hystrix differs from the specimens described above as follows: 1) The rhabdosome is entirely biserial. 2) The sicular aperture is furnished with anti-virgellar spines. 3) The distal-lateral corners of the strongly introverted distal part of the free ventral wall are fused with the dorso-lateral comers of the aperture, and there are no paired apertural 1 2 spines. However, th 1 and th 1 do have paired apertural spines, and the shape of these thecae resembles closely the shape of the first two thecae in the specimens described above. 4) Paired spines arise from the lateral margins of the proximal end of the free ventral wall. In spite of these differences, the overall rhabdosomal and thecal characters of the specimens described above resemble closely those of 2" hystrix and suggest a close affinity of the two species. However, because the diagnostic generic character in graptoloid classi fication has traditionally been the shape of the rhabdosome, an assignment of the specimens described above to Dicaulograptus is questionable. The available specimens doubtlessly represent a new species, but because of their poor preservation and fragmentary 568 nature, the erection of a new species and a new genus seems premature. Figured Specimens OSU 33339-OSU 33345. REFERENCES Beavis, F. G., 1972, The Maniibriate Isograptids: Geol. Mag., v. 109, p. 193-204. Bergstrom, S. M., 1971, Conodont biostratigraphy of the Middle and Upper Ordovician of Europe and eastern North America. In Symposium on Conodont Biostratigraphy, Sweet, W. C., and Bergstrom, S. M., (Eds.): Geol. Soc. Am. Mem. v. 127, p. 83-161. Bergstrom, S. M. and Cooper, R.. A., 1973, Didymograptus bifidus and the trans-Atlantic correlation of the Lower Ordovician: Lethaia, V. 6, p. 313-340. Bergstrom, S. M. and Drahovzal, J. A., 1972, Facies relations in the Middle Ordovician of the Southern Appalachians in Alabama (abstr.): Geol. Soc. Am. Abstr. with Progr., v. 2, no. 2, p. 60-61. Berry, W. B. N., 1960, Graptolite Faunas of the Marathon Region, West Texas: Texas Univ., Bur. Econ. Geol., Pub. 6005, 179 pp. Berry, W. B. N., 1962, Stratigraphy, Zonation, and Age of Schaghticoke, Deepkill, and Normanskill Shales, Eastern New York: Geol. Soc. Amer. Bull, v. 73, p. 695-718. Berry, W. B. N., 1964, The Middle Ordovician of the Oslo Region, Norway. No. 16. Graptolites of the Ogygiocaris Series: Norsk Geol. Tidskr., V. 44, p. 61-170. Bulman, 0. M. B., 1931, South American graptolites with special reference to the Nordenskibld Collection: Ark. Zool., v. 22A, p. 1-111. Bulman, 0. M. B., 1932, On the graptolites prepared by Holm. 4. New Species of Didymograptus and Dicellograptus from bland, with remarks on the development of Dicellograptus : Ark. Zool., v. 24A, no. 9, p. 15-21. Bulman, 0. M. B., 1933, Graptolites from the Quitari Area: Quart. Jour. Geol. Soc. London, v. 89, p. 348-356. Bulman, 0. M. B., 1938, The proximal end of Cryptograptus: Geol. Mag., V. 75, p. 539-543. 569 570 Bulman, G. M. B., 1944-1947, A monograph of the Caradoc (Balclatchie) graptolites from limestones in Laggan Bum, Ayrshire: Palaeontog. Soc. Monogr. Part 1, 1944, p. 1-42; Part 2, 1945, p. 43-58; Part 3, 1947, p. i-xi, 59-78. Bulman, 0. M. B . , 1948, Some Shropshire Ordovician Graptolites: Geol. Mag., V. 85, p. 222-228. Bulman, 0. M. B., 1953, Some graptolites from the Ogygiocaris Series of the Oslo district: Ark. f. Mineral, och Geol., v. 1, no. 17, p. 509-518. Bulman, 0. M. B. , 1958a, Patterns of colonial development in graptolites: Linnean Soc. London (Zool.), Jour., v. 44, p. 24-32. Bulman, 0. M. B., 1963, On Glyptograptus dentatus (Brongniart) and some allied species: Palaeont., v. 6, pp. 665-689. Bulman, 0. M. B., 1968, The mode of development of Isograptus manubriatus (T. S. Hall): Geol. Mag., v. 105, p. 211-215. Bulman, 0. M. B., 1969, "Prothecal folds" and the origin of Dicellograptus: In Campbell, K. S. (ed.). Stratigraphy and palaeontology: Essays in honour of Dorothy Hill, p. 1-16, Australian Natl. Univ. Press (Canberra). Bulman, 0. M. B., 1970, Graptolithina. With sections on Enteropneusta and Pterobranchia, In Teichert, C. (Ed.), Treatise on Invertebrate Paleontology, Part V, (2nd Ed.), i-xxxiii, 163 pp., Geol. Soc. Amer, and Univ. Kansas Press. Butts, Charles, 1926, The Paleozonic Rocks. ^ Geology of Alabama, Geol. Surv. Alabama, Sp. Report no. 14, p. 41-230. Carruthers, W., 1859, On the Graptolites from the Silurian Shales of Dumfriesshire, with a description of three new species: Ann. Mag. Nat. History, ser. 3, v. 3, p. 23-26. Cooper, B. N., and Cooper, G. A., 1946, Lower Middle Ordovician stratigraphy of the Shenandoah Valley, Virginia: Geol. Soc. Amer. Bull., v. 57, p. 25-114. Cooper, G. A., 1956, Chazyan and related brachiopods: Smithsonian misc. Coll., V. 127, Pt. 1, i-xvi, 1-1024; Pt. 2, p. 1025-1245. Cooper, R. A., 1973, Taxonomy and Evolution of Isograptus Moberg in Australiasia: Palaeont. v. 16, p. 45-115. 571 Cooper, R. A. and McLaurin, A. N. , 1974, Apiograptus gen. nov. and the origin of the biserial graptoloid rhabdosome; Palaeont. Assoc. Spec. Papers in Palaeont., no. 13, pp. 75-85. Decker, C. E., 1952, Stratigraphie Significance of Graptolites of Athens Shale: Am. Assoc. Petrol. Geol. Bull,, v. 36, p. 1-145. Drahovzal, J. A. and Neathery, T. L., 1971, Middle and Upper Ordovician stratigraphy of the Alabama Appalachians. In The Middle and Upper Ordovician of the Alabama Appalachians, Drahyozal, J. A. and Neathery, T. L. (Eds.): Ala. Geol. Soc., Ninth Ann. Field Trip, Guidebook, p. 1-62. EkstrSm, G., 1937, Upper Didymograptus Shale in Scania: Sver. Geol. Unders. Afh. Ser. C, no. 403, p. 3-53. Elies, G. L., 1922, The graptolite faunas of the British Isles: Geol. Assoc., Proc., V. 33, p. 168-200. Elies, G. L., 1925, The characteristic assemblages of the graptolite zones of the British Isles: Geol. Mag., v. 62, p. 337-347. Elies, G. L., 1939, The stratigraphy and faunal succession in the Ordovician rocks of the Builth-Llandrindod inlier, Radnorshire: Quart. Jour. Geol. Soc. Lond., v. 95, p. 383-445. Elies, G. L. and Wood, E. M. R., 1901-1918, A monograph of British Grapto lites, Ed. by C. Lapworth: Palaeontog. Soc. Monogr. Part 1, 1901, p. 1-54; Part 2, 1902, p. i-xxviii, 55-102,; Part 3, 1903, p. xxix-lii, 103-134; Part 4, 1904, p. liii-lxxii, 135-180; Part 5, 1906, p. Ixxiii-xcvi, 181-216; Part 6, 1907, p. xcvii-cxx, 217-272; Part 7, 1908, p. cxxi-cxlviii, 273-358; Part 8, 1911, p. 359-414; Part 9, 1913, p. 415-486; Part 10, 1914, p. 487-526; Part 11, 1918, p. a-m, cxlix-clxxi» 527-539. Emmons, E., 1855, American Geology, v. 1, 194 pp., (Albany, N.Y.). Emmons, E., 1856, American Geology, v. 2, 251 pp., (Albany, N.Y.). Erdtmann, B.-D., 1965, Outline Stratigraphy of Graptolite-Bearing 3b (Lower Ordovician) Strata in the Oslo Region, Norway: Norsk Geol. Tidsskr., v. 45, p. 481-547. Gurley, R. R., 1896, North American graptolites: Jour. Geology, v. 4, p. 63-102, 291-311. Hadding, A., 1913, Undre Dicellograptusskiffem i Skane: Acta Univ. Lund., N.F., avd. 2, v. 9, no. 15, pp. 1-92. 572 Hadding, A., 1915a, Om Glossograptus, Cryptograptus, och tvenne dem narstâende graptolitslakten: Geol. For, Stockholm Fôrh., v. 37, pp. 303-36. Hadding, A., 1915b, Der mittlere Dicellograptus-schiefer auf Bomholm: Acta Univ. Lund., N.B., avd. 2,v. 11, no. 4, pp. 1-40. Hall, J., 1847, Descriptions of the organic remains of the lower division of the New York system: Palaeontology of New York, v. 1., Albany, N.Y. Hall, J., 1859, Notes upon the genus Graptmlithus: Paleontology of New York, V. 3 (suppl.), p. 495-529. Albany, N.Y. Hall, J., 1860, Notices of new forms of the genus Graptolithus and allied genera : N.Y. State Cab. Nat. Hist., Ann. Hep. 13, p. 55-64. Hall, J., 1865, Graptolites of the Quebec Group: Geol. Surv. Canada, Canad. Organic Remains, dec. 2, p. 1-151. Hall, J., 1868, Introduction to the study of the Graptolitidae: N.Y. State Cab. Nat. History, 20th Kept., p. 169-240. Albany, N.Y. Harris, W. J., and Crawford, W., 1921, The relationships of the sedimentary rocks of the Bisbome district, Victoria: Roy. Soc. Viet., Proc. N. S., V. 33, p. 39-78. Harris, W. J., and Thomas, D. E., 1935, Victorian Graptolites (new series). Part III: Roy. Soc. Victoria, Proc., v. 47, p. 288-313. Harris, W. J., and Thomas, D. E., 1938, Revised classification and cor relation of the Ordovician graptolite beds of Victoria: Victoria Dept. Mines, Mining and Geol. Jour., v. 1, p. 69-72. Harris, W. J., and Thomas, D. E., 1955, Victorian graptolites, new series. Part XIII. Graptolites from the Wellington River, Part 1: Victoria Dept, of Mines, Mining and Geol. Jour., v. 5, no. 6, p. 35-47. Hayes, C. W., 1894, Kingston folio, Tennessee: U.S. Geol. Survey, Geol. Atlas, Folio 95, 6 p. Hede, J. E., 1951, Boring through Middle Ordovician-Upper Cambrian strata in the Fagelsâng District, Scania (Sweden). 1. Succession en countered in the Boring: Lunds Univ. Arsskrift, N.F. Avd. 2, v. 46, no. 7, 78 pp. Hisinger, W., 1840, Lethaea Suecica sen Petrificata Sueciae: Supplementum secundum. Stockholm. 573 Holm, G., 1881, Bidrag till kannedomen om Skandinaviens Graptoliter. I, II: K. Svenska Vetenskaps. Akad., Forhandl., v. 38, no. 4, p. 71-84; no. 9, p. 45-52. Hopkinson, J., 1870, On the structure and affinities of the genus Dicranograptus : Geol. Mag., v. 7, p. 353-359. Hopkinson, J., 1871, On Dicellograpsus, a new genus of graptolites: Geol. Mag., V. 8, p. 20-26. Hopkinson, J., 1872, On some new species of Graptolites from the South of Scotland: Geol. Mag., v. 9, p. 501-509. Hopkinson, J. and Lapworth, C., 1875, Descriptions of the graptolites of the Arenig and Llandeilo rocks of St. David's: Geol. Soc. London, Quart, Jour., v. 31, p. 631-672. Jaanusson, V., 1960, Graptoloids from the Ontikan and Viruan (Ordov.) limestones of Estonia and Sweden: Geol. Inst. Uppsala, Bull., v. 38, pp. 290-366. Jaanusson, V., 1964, Two Multiramous Graptoloids from the Lower Didymograptus Shale of Scandinavia: Geol. Foren. Stockholm Forh., V. 86, pp. 413-432. Jaanusson, V., 1973, Morphological discontinuities in the evolution of graptolite colonies. In Boardman, R. S. et. al. (eds.). Animal Colonies: Their Development and Function Through Time. Dowden, Hutchinson, and Ross, Stroudsburg, PA., p. 515-521. Jaanusson, V., and Skoglund, R., 1963, Graptoloids from the Viruan (Ordovician) Dalby and Skagen Limestones of Vas ter got land : Geol. Foren. Stockholm Forhandl., v. 85, p. 341-357. Jaanusson, V. and Strachan, I., 1954, Correlation of the Scandinavian Middle Ordovician with the graptolite succession: Geol. Foren. Stockholm Forhandl., v. 76, p. 684-696. Jackson, D. E., 1962, Graptolite zones in the Skiddaw Group in Cumberland, England: Jour. Paleont., v. 36, p. 300-313. James, J., 1965, The development of a dicellograptid from the Balclatchie Shales of Laggan Bum: Palaeont., v. 8, pp. 41-53. Kozïowski, R., 1952, Sur un remarquable Graptolithe ordovicien : Acta. Geol. Polonica, v. 2 (1951), p. 86-93 (Polish text p. 291-299). 574 Kozïowski, R., 1933, Badania nad nowym gatunkiem z rodzaju Corynoides (Graptolithina): Acta. Geol. Polonica, v. 3, pt. 2, p. 193-209. KozXowski, R., 1954, Sur la structure de certains Dichograptides :Acta Geol. Polonica, v. 4, p. 118-135 (Polish Text, p. 423-444). KozXowski, R., 1956, Nouvelles observations sur les Corynoididae (Graptol ithina): Acta Paleont. Polonica, v. 1, no. 4, p. 259-269. Lapworth, C., 1873, On an improved classification of the Rhabdophora: Geol. Mag., v. 10, p. 500-504, 555-560. Lapworth, C., 1876, The Silurian System in the South of Scotland. In Armstrong, J. et. al.. Catalogue of Western Scottish Fossils, p. 1-28. Lapworth, C., 1877, On the graptolites of County Down: Belfast Nat. Field Club, Proc., V. 1, p. 125-144. Lapworth, C., 1879-1880, On the geological distribution of the Rhabdophora: Ann. and Mag. Nat. History, serv. 5, v. 3, p. 245-257, 449-455; V. 4, p. 333-341, 423-431; v. 5, p. 45-62, 273-285, 359-369; v. 6, p. 16-29, 185-207, (v. 3-4, 1879; v. 5^6, 1880). Lapworth, C., 1880, On new British graptolites: Ann. and Mag. Nat. History, ser. 5, v. 5, p. 149-177. Lemon, R. R. H., and Cranswick, J. S., 1956, Graptolites from Huacar, Peru: Mus. Hist. Nat. "Javier Prado," Lima, Publ. Ser. C, Geol., no. 5, p. 1-30. Monsen, A., 1937, Die Graptolithenfauna im Unteren Didymograptusschiefer (Phyllograptusschiefer) Norwegens: Norsk, geol. Tidsskr., v. 16, p. 57-266. Mu, A. T., 1958, Abrograptus, a new graptolite genus from the Hulo Shale; Acta Palaeont. Sinica, v. 6, p. 259-265. Mu, A. T., 1963, On the complication of graptolite rhrbdosome: Acta Palaeont. Sinica, v. 11, p. 346-377. Mu, A. T. and Qiao, X. D., 1962, New materials ofAbrograptidae: Acta Palaeont. Sinica, v. 10, p. 1-8. Mu, A. T. and Zhan, S. G., 1966, On the probable development and systematic position of Glossograptus: Acta Palaeont. Sinica, v. 14, p. 99-109. Neuman, R. B., 1955, Middle Ordovician rocks of the Tellico-Sevier belt, eastern Tennessee: U. S. Geol. Surv. Prof. Paper 274-F, p. 141-175. 575 Neuman, R- B., 1976, Ordovician of the Eastern United States, p. 195-207. In Bassett, M. G. (ed.). The Ordovician System: proceedings of a Palaeontological Association symposium, Birmingham, September 1974, 696 pp.. University of Wales Press and National Museum of Wales, Cardiff. ' Nicholson, H. A., 1867, On some fossils from the Lower Silurian rocks of the south of Scotland: Geol. Mag., v. 4, p. 107-113. Nicholson, H. A., 1868, Notes on Helicograpsus, a new genus of graptolites: Ann. and Mag. Nat. History, ser. 4, v. 2, p. 23-26. Nicholson, H. A., 1868a, On the graptolites of the Coniston Flags; with notes on the British species of the genus Graptolites: Quart. Jour. Geol. Soc., London, v. 24, p. 521-545. Nicholson, H. A., 1875, On a new genus and some new species of graptolites from the Skiddaw Slates: Ann. and Mag. Nat. History, ser. 4, v. 16, p. 269-273. Nilsson, R. , 1960, A preliminary report on a boring through Middle Ordovician strata in western Scania (Sweden): Geol. Foren. Stockholm Forhandl., v. 82, p. 218-226. Pribyl, A., 1947, Classification of the genus Climacograptus Hall, 1865: Acad. Tchèque Sci., Bui. Intematil., v. 48, no. 2, p. 1-12. Rickards, R. B., 1972, The Ordovician Graptolite Genus Paraglossograptus Mu: Geol. Mag., v. 109, p. 99-113. Rickards, R. B., and Bulman, 0. M. B., 1965, The development of Lasiograptus harknessi (Nicholson 1867): Palaeont., v. 8, p. 272- 280. Riva, J., 1974, A revision of some Ordovician graptolites of eastern North America: Palaeont., v. 17, p. 1-40. Ross, R. J., Jr., and Berry, W. B. N., 1963, Ordovician Graptolites of the Basin Ranges in California, Nevada, Utah, and Idaho: U.S. Geol. Surv., Bull. no. 1134, 177 pp. Ruedemann, R., 1902, Graptolite facies of the Beekmantown formation in Rensselaer County : N.Y. State Mus., Bull. 52, p. 546-575. Ruedemann, R., 1908, Graptolites of New York, Part I, N.Y. State Museum, Mem. 11, p. 457-583. 576 Ruedemann, R., 1947, Graptolites of North America, Geol. Soc. Amer., Memoir, no. 19, 652 pp. Ruedemann, R., and Decker, G., 1934, The Graptolites of the Viola Lime stone: Jour. Paleont., v. 8, p. 303-327. Safford, J. M . , and Killebrew, J. B., 1876, The elementary geology of Tennessee. 225 p., Nashville. Skevington, D., 1965, Graptolites from the Ontikan Limestones (Ordovician) of bland, Sweden, II. Graptoloidea and Graptovennida: Geol. Inst. Uppsala, Bull., v. 43, no. 3, p. 1-74. Skevington, D., 1968, The affinities of Oncograptus, Cardiograptus. and allied graptolites from the Lower Ordovician: Lethaia, v. 1, p. 311- 324. Skevington, D., 1969, The classification of the Ordovician System in Wales. In A. Wood (ed.). The Pre-Cambrian and Lower Paleozoic Rocks of Wales, p. 161-179. University of Wales Press, Cardiff, 461 pp. Skevington, D., 1970, A Lower L Ianvi m Graptolite Fauna from the Skiddaw Slates, Westmorland: Yorkshire Geol. Soc., Proc., v. 37, part 4, no. 18, p. 395-444. Skoglund, R., 1961, Kinnegraptus, a new graptolite genus from the the Lower Didymograptus Shale of VastergStland. central Sweden: Geol. Inst. Uppsala, Bull., v. 40, p. 389-400. Skoglund, R., 1963, Uppermost Viruan and Lower Harjuan (Ordovician) Stratigraphy of Vastergotland, and Lower Harjuan graptolite faunas of Central Sweden: Geol. Inst. Uppsala, Bull., v. 42, no. 45, pp. 1-55. Stephenson, J. P., Walker, K. R., and McLau^lin, R. E., 1973, The Lenoir Formation-back reef supratidal, intertidal, and subtidal shelf lagoon facies. In Geology of Knox County, Tennessee: Tenn. Div. Geol., Bull., V. 70, p. 122-126. Strachan, I., 1949, On the genus Corynoides Nicholson : Geol. Mag., v. 86, p. 153-160. Strachan, I., 1960, Graptolites from the Ludibundes Beds (Middle Ordovician) of Tvaren, Sweden: Geol. Inst. Uppsala, Bull., v. 38, p. 47-68. 577 Strachan, I., 1969, A redescription of W. Carruthers' type graptolites; Brit. Mus. Nat. Hist. (Geol), Bull., v. 17, no. 4, p. 183-206. Strachan, I., 1971, A synoptic supplement to "A Monograph of British Graptolites by Miss G. L. Elies and Miss E. M. R. Wood": Palaeontogr. Soc. Monogr., 130 pp. Strachan, I., 1972, Correlation of British and Australian graptolite zones. In Williams et. al., A Correlation of Ordovician Rocks in the British Isles. Geol. Soc. London, Spec. Report no. 3, Scottish Academic Press Ltd., Edinburgh. Sun, Y. C., 1933, Ordovician and Silurian Graptolites from China: Palaeont. Sinica ser. B, v. 14, no. 1, p. 1-52. Sweet, W. C., and Bergstrom, S. M., 1962, Conodonts from the Pratt Ferry Formation (Middle Ordovician) of Alabama: Jour. Paleont. v. 36, p. 1214-1252. Thomas, D. E., 1960, Zonal distribution of Australian graptolites: Roy. Soc. N.S.W., Proc., v. 94, p. 1-58. Thomas, D. E., and Keble, R. A., 1933, The Ordovician and Silurian rocks of the Bulla-Sunbury area, and discussion of the sequence of the Melbourne area. Roy. Soc. Viet., Proc., v. 45, p. 33-84. Thorsteinsson, R., 1955, The mode of cladial generation in Cyrtograptus: Geol. Mag., v. 92, p. 37-49. Tjemvik, T. E., 1960, The Lower Didymograptus Shales of the Flagabro drilling core: Geol. Foren. Stockholm Forh., v. 82, p. 203-217. Toghill, P. , 1970a, Highest Ordovician (Hartfell Shales) graptolite faunas from the Moffat Area, South Scotland: Brit. Mus. nat. Hist. (Geol.), Bull., v. 19, no. 1, pp. 1-26. Toghill, P., 1970b, A fauna from the Rendre Shales (Llandeilo) of the Mydrim area, Carmarthenshire: Geol. Soc. Lond., Proc., no. 1663, pp. 121-129. Tullberg, S. A., 1882 , Skanes graptoliter. I. Allman ofversigt Bfver de siluriska bildningarne i Skâne och jemforelse med ofriga kânda saratidiga aflagringar: Sver. Geol. U.ders., ser. C, no. 50, Stockholm. 578 Turner, J. C. M., 1960, Faunas graptoliticas de América del sur: Âsoc. Geol. Argentina, Revista, v. 14, p. 1-180. Urbanek, A., 1966, On the morphology and evolution of the Cucullograptinae (Monograptidae, Graptolithina): Acta Palaeont. Polonica, v. 11, p. 291-544. Urbanek, A., 1973, Organization and Evolution of Graptolite Colonies. Boardman, F.. S. et. al., (eds.). In Animal Colonies, Dowden, Hutchinson, and Ross, Inc., Stroudsburg, Pa., p. 441-514. Urbanek, A., 1976, The problem of graptolite affinities in the light of ultrastructural studies on peridermal derivatives in Pterobranchs: Acta Palaeont. Polonica, v. 21, no. 1, p. 1-36. Urbanek, A., and Jannusson, V., 1974, Genetic Polymorphism as Evidence of Outbreeding in Graptoloids. In Rickards, R. B., Jackson, D. E., and Hughes, C. P., Graptolite Studies in Honour of 0. M. B. Bulman, Palaeont. Assoc., Spec. Paper, no. 13, p. 15-17. Webby, B. D., 1976, The Ordovician System in South-eastern Australia. ^ Bassett, M. G. (ed.), 1976, The Ordovician System: proceedings of a Palaeontological Association symposium, Birmin^am, September, 1974, p. 417-446. University of Wales Press and National Museum of Wales, Cardiff. Whittington, H. B., 1955, Additional new Ordovician graptolites and a diitinozoan from Oklahoma. Jour. Paleont., v. 29, pp. 837-851. Whittington, H. B., and Rickards, R. B., 1969, Development of Glossograptus and Skiagraptus, Ordovician Graptoloids from Newfoundland: Jour. Paleont., v. 43, pp. 800-817. APPENDIX A Distribution of Graptolites in the Sections Investigated Collection numbers (C.) are listed in stratigraphically ascending order in the left column of Tables 10-12. As mentioned above (See Text- figure abbreviations), the collection designations PF, PS, C, and C-25 refer to the Pratt's Ferry, Pratt's Syncline, Calera-quarry, and Calera-highway 25 roadcut sections, respectively. These designations are followed by a number that indicates in metres the stratigraphie distance that the collection is located above the base of the section, except for the Calera-highway 25 section in which case the number indicates the distance below the top of the section. Although data for the Calera-highway 25 section are included in Table 12 with data for the Calera-quarry section, the Calera-highway 25 data are placed at the end of the table. The species present in the study area are designated by numbers (sp.) across the top of Tables 10-12. These numbers corres pond to the following species: 1. Dictvonema sp; 2. Pterograptus sp. 579 580 3. Didvmograptus sp. 4. Azveograptus incurvus EkstrBm 5. Reteograptus geinitzianus Hall 6. Nemagraptus gracilis (Hall) 7. Amphigraptus n. sp. A 8. Amphigraptus n. sp. B 9. Nemagraptid sp. A gen. indet. et sp. nov. 10. Dicellograptus alabamensis Ruedemann 11. Dicellograptus bispiralis bispiralis (Ruedemann) 12. Dieelloeraptus bispiralis n. ssp. A 13. Dicellograptus geniculatus Bulman 14. Dicellograptus gurleyi gurlevi Ruedemann 15. Dicellograptus gurlevi n. ssp. A 16. Dicellograptus sextans (Hall) 17. Leptograptus trentonensis Ruedemann 18. Dicranograptus irregularis Hadding 19. Glossograptus ciliatus Emmons 20. Apoglossograptus Ivra (Ruedemann) 21. Crvptograptus marcidus (Hall) 22. Climacograptus meridionalis Ruedemann 23. Pseudoclimacograptus angulatus angulatus (Bulman) 24. Pseudoclimacograptus modestus (Ruedemann) 25. Pseudoclimacograptus sp. cf. P. eurvstoma Jaanusson 26. Glyptograptus euglyphus (Lapworth) 27. Glypto gr aptus sp. cf. G. teretiusculus (Hisinger) 28. Orthograptus sp. 29. Lasiograptus sp. 30. Dicaulograptus? n. sp. A Rare, common and abundant species occurrences are designated by r, c, and a, respectively. Although systematic sampling was attempted in this study, the sizes of the collections, in terms of number of speci mens, collecting time expended, and area of bedding surface sampled, vary considerably. Accordingly, determinations of abundance are subjective. A rare occurrence is one in which a particular species is represented by two or less specimens in a single collection. An abundant occurrence is one in which a particular species is represented by more than half of the specimens in a single collection. A common occurrence is one that is intermediate between rare and abundant. Table 10. Distribution of graptolites in the Pratt's Ferry Section. sp. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 C. PF-12 0 PF-13 a r - r - - r - PF-13.5 PF-16.4 PF-17 -- r - r r -- - a a - - c - r - - r - r c - r - c c - - r PF-17.1 - - r - r r -- - a a - - r ------r r - r - c c - - - PF-17.3 - - r - r r - - - c a -- c - c -- r - c a - - - c c --- PF-17.7 - - - - r c - - - a a ------c r --- - c -- - PF-18.7 - - a - r - - - - c c - - c - r -- r - c - - r - c c - -- PF-19 r - a - c r -- - r a - - r - r - - a - c - - c - - r --- PF-19.2 -- a - r r - - - c r - - r - r - - a - c -- c -- r - - - PF-19.9 -- c - r r -- - a -- - r ------a - •• -- r c - - - PF-20.5 - - r - c r - - - a r -- r - r - - r - a ----- r -- - PF-21.3 - - r - r -- - - r r - - a - r - - r - - - - r -- c -- - PF-21.7 - - r - r r - - - r r r PF-22 - - r -- c - - - r c r PF-22.5 - - c - c c - - - r c - - r - r - r - - r r - r - - a - - r PF-22.9 PF-23.8 PF-24.4 -- a -- r - -- r r -- r -- - - r - a c - a - - c - - - PF-25 - -- - - c -- - - r - - a - c -- a - - c - - - r r -- c PF-25.5 Ln 00 Table 11. Distribution of graptolites in the Pratt's Syncline Section. sp. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 C. PS-12 a 5 c a r r c r r r c r PS-13 a PS-17 - - - - - r -- - a a c r - --- r - - r PS-29 r - a - r r - - - a - - - c - - - - r -- c - r - r c -- a PS-76.4 - - c - r r r - r a c - - r - c c - r - r r - a - - c --- PS-77.5 r - r - - r -- - c ------r - - - c - - - - r c --- PS-102 - - - - c c --- r - - - c - a -- c - c c - a - r c - - - PS-109.2 - - c - a c - - r a r - - r - c a - a - a c - r - a c - - - PS-109.7 - - - - c r - - r a - • - - c - c c - r - a r - r - r c - - - PS-117 -- a - r r --- r - - - a -- a - c - a c - c - c r - - - PS-119.5 - - - - r r PS-126 -- - - r r --- a --- r - - - - r - r r - r - c c - - - PS-127.5 -- - -- r - -- a - - - r - - c - r - c r - r - r - - - - PS-129 --- - r r - - - c - -- r - c a - r - c c - r - a r r - - PS-132 r - r c - - - r -- - - r r c -- L n 0 0 r o Table 12. Distribution of graptolites in the Calera Section. sp. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 le 17 18 19 20 21 22 23 24 25 26 27 28 29 31 c. C-4.6 r a c r C-5.1 -- c - a ------r - a - a - - a c - c - C-5.4 c a - - c c - r - C-5.5 C-5.8 C-6.3 -- c - c ------c - a - a - - c r - - - C-6.6 - - c - c ------r - a - a - r a r - - - C-7.2 r •• c - a - - a r --- C-7.6 - - r - c ------r - c - a - - a - -- - C-8.0 - - c - c ------c - c - a - - a c - - - C-8.5 - - c - c ------a -- -- - r - a - c - - a r - r - C-9.0 - - c - c ------a - - -- - r - c - a - r a r - r - C-9.4 - - c - r ------a - - --- r - c - a - - a r - - - C-9.9 - - c - r ------a ------r - a - - a r --- C-10.5 - - c - c r ------c - a - -- c - c - c - - a r --- C-11.1 - - c - c r - r - - - r c - c -- - r - c - a - - a c - - - C-11.8 - - c - c c - r --- c c - a -- - a - r r a -- a r - - - C-12.7 - - a r r r -- - - - r - c - c - - r - c c a -- c c - -- C-13.0 -- c r c r -- -- - r - c - - -- c - - r a - - c c --- C-13.9 - - c c a r a r - - c - r r - - - a - - c r --- C-14.8 - - c r a c - r -- - a - c - c - c r - r - a - - r r - - - C-15.7 - - c r r c - - - -- r - c - V - - - - c - a - - c r - - - C-16.2 - - c - c c c r - - - C-16.8 -- c c c a ------a - r - - - - r r a - - a c - - - C-17.4 - r r - - r ------c - r - - r -- a c -- a c - -- C-18.0 - -- c r r r r - -- C-18.3 -- r r c ------a - r -- - - c - r - - a c - - - C.18.5 c c r - - a r - - - C-19.2 - - r r - c c r - -- C-19.9 - - r r c r r a a - - a c - - - U i C-22.5 ----- r ------c - r - - -- c - c - - a r - - - 00 co Table 12. Distribution of graptolites in the Calera Section, continued. sp. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 C-23.5 G-24.5 -- r r r r C-26.0 - - - r r r c r -- - a a - - - C-26.5 - - r c - r c c -- - a r - -- C-27.5 -• - r r r r ------r -- -- r - r - a - - c c - r - C-28.3 -- c c - r r r c a - - c c - -- C-29.0 - - c r r r c a - - a r - - - C-30.0 - - - a r c r c c r - a c - - - C-31.0 - - - - - a ------r - r - c - - - a c - - c r - -- C-32.0 - - r c - a c r - - c - - c - C-33.0 - - - c - r ------c - r - - -- r a r - - c r - r - C-34.0 -- r c r a ------c - r - - r r c r c - - c r - r - C-35.0 -- r c r r ------c - r -- r r r r c - - r r - -- C-36.0 - - r c r c c r r r c c - - c r - r - C-37.0 -- c a r a r r r r - - r c - r - C-38.3 - -- r r c r r - -- C-39.0 -- r a r c ------a - r -- r - - r r - - c c - r - C-40.0 - -- c r r ------a - r -- -- c a -- - r a - r - C-41.0 - - c - r r ------c - c -- - c c c r - - c c - r - C-42.0 C-44.5 - - - c r c r - r r - r - c a -- - C-45.0 -- r c r a ------a - r - - - c r a r r - - c - -- C-45.5 --- a r a ------a - r - --- r c - r - - r -- - C-46.0 - - --- a ------■ - c - r - ---- c - r -- c - - - C-46.5 -- r - r c ------c - r - -- r - a - - - c r - -- 0-47.0 a a --- c c --- C-47.7 - - - - r r c a - c - - a - -- C-48.2 -- - - r r r r r a c --- r a - - - 0-49.0 - -- r - r ------a - r - r - r - a - r - r c - - - 0-50.0 -- - - r ------c - a - r - - r c - r - c c - - - 0-50.4 ------c- -- -- rra----a---r r a - --- a --- <-n 00 Table 12. Distribution of graptolites in the Calera Section, continued. sp. 1 2 3 4 5 6 7 8 9 1011 121314 15 16 17 1819 20 21 22 2324 25 26 27 28 29 30 c. C-52.0 C-53.0 c a -- -- a - - - C-53.8 - - r - - r ------a - c - r c a - r - c a -- - C-57.1 c a - a -- a - - - C-57.9 - - r r •• r r c c - r - r a - - - C-59.5 C-25-29.5 --- r r r c c - - - r c - r - C-25-28.0 -- c r r c c c --- C-25-26.0 -- c c r c ------r - c - - - - r - - r - a r - - - C-25-16.5 -- c r ------a - a -- c c - r - - a - - - C-25-11.8 -- - a - r ------a - - - - r r - r - - c --- C-25-8.0 - - r a ------r - r - - - - a c - - - - r --- C-25-0.7 - - r c r c r - r - r a - c - r - - - % Ln