Paleontological Contributions

THE UNIVERSITY OF KANSAS

PALEONTOLOGICAL CONTRIBUTIONS

December 16, 1981 Paper 106

LARGER BIVALVE ARTHROPODS FROM THE MIDDLE CAMBRIAN OF UTAH'

R. A. ROBISON and Ft—. C. RICHARDS

Department of Geology. University of Kansas, Lawrence, Kansas 66045, and Geological Survey of Canada, Calgary, Canada T2L 2A7

Abstract—Carapaces or valves of 12 species of larger bivalve arthropods and one species of an associated univalve arthropod are described from Middle Cambrian strata of northern and western Utah. Some of these bivalve species represent phyllocarid Malacostraca whereas others may represent either phyllocarids or Branchiopoda. Af- finity of the univalve species is less certain. One genus and five species of larger bi- valves are new, the new genus being Dicerocaris and the new species being Dicerocaris opisthoeces, Perspicaris? dziatus, Perspicaris? ellipsopelta, Tuzoia guntheri, and Tuzoia? pet erseni. Specimens of Branchiocaris? sp., Canadaspis cf. C. perfecta (Walcott), Dioxycaris argenta (Walcott), Proboscicaris agnosta Rolfe, Pseudoarc- tole pis sharpi Brooks and Caster, Tuzoia retifera Walcott, and an undetermined genus and species are illustrated and discussed. Three distinctive univalve carapaces are assigned to the new genus and new species Pahvantia hastata. Collectively, the specimens provide important information concerning the distribution and species diversity of a poorly known element of the Cambrian fauna.

EXTENSIVE SEARCHES and generous contribu- larger collection of such Cambrian fossils is tions by several persons during the past two that from the famous Burgess Shale in British decades have made possible the assembly of a Columbia. This paper includes descriptions of collection of more than 100 carapaces or disar- 12 larger bivalve arthropod species and one ticulated valves of larger bivalve arthropods probable univalve species. from Middle Cambrian strata in Utah (Fig. 1). Larger bivalve arthropods have been re- Because in life the exoskeletons of these ar- ported from few Cambrian formations in thropods were not mineralized, their preserva- North America. Several species, some of which tion is uncommon. To our knowledge, the only we consider to be synonyms, have been described from the Burgess Shale of Middle Cam-

'Manuscript received August 21, 1981 , brian age (Walcott, 1912a; Resser, 1929; 2 The University of Kansas Paleontological Contributions Paper 106

Rolfe, 1962; Simonetta and Delle Cave, 1975; bivalve carapaces. With only the carapace it Briggs, 1976, 1977, 1978). Distribution else- may be difficult, if not impossible, to distin- where in western North America, includes guish representatives of families of phyllocarid three species from the Eager Formation of Malacostraca from those of the Protocarididae, Early Cambrian age in British Columbia which show closest affinities with the Branchio- (Resser, 1929); one species from the Ophir poda (Briggs, 1976, p. 12). Because soft parts Shale (Walcott, 1886) and another from the and appendages are not preserved with cara- Wheeler Formation (Brooks and Caster, 1956), paces of any known bivalve arthropod from the both of Middle Cambrian age in Utah. In east- Cambrian of Utah, generic identifications for ern North America, six species have been de- some specimens are questionable. Also, the scribed from the Kinzers Formation in Pennsyl- family, order, and even class assignments are vania (Resser, 1929; Resser and Howell, 1938) uncertain for most. Nevertheless, we believe and two species from the Parker Shale in Ver- that these specimens warrant description mont (Resser and Howell, 1938), all of Early because they provide important information Cambrian age; one species from the Rogersville about the distribution and species diversity of a Shale of Middle Cambrian age in Tennessee poorly known element of the Cambrian fauna. (Resser, 1938); and two species from the Trilobites are the most abundant inverte- Nolichucky Formation of Late Cambrian age brate fossils in most Cambrian strata, partly at in Alabama (Resser, 1938). least because many trilobites possessed a well- calcified dorsal exoskeleton with a high poten- tial for preservation. Evidence from the excep- tionally preserved fauna of the Burgess Shale Wellsville • Bear River Range indicates that species of nontrilobite arthro- Mountains pods actually outnumber those of trilobites by more than two to one, and individuals of some species are far more common than those of tri- Salt Lake City f# lobites (Conway Morris, 1979; Conway Morris • Wasatch Range and Whittington, 1979; Whittington, 1980). Moreover, preliminary comparison of total numbers and proportions of normally preserved genera from different formations sug- • Drum Mountains gests that composition of the biota of the • House Range Burgess Shale, rather than being unusual, is probably representative of Middle Cambrian, low-latitude, open-shelf environments in North America (Conway Morris and Robison, 1982). UTAH The assemblages of rarely preserved, larger bivalve arthropods in the Middle Cambrian of Utah also bear strong resemblances to the assemblage in the Burgess Shale. This further supports the notion of normality for the Burgess biota. 0 km 100 Stratigraphy. — In general, lithostrati- graphic nomenclature in this paper conforms to the usage of Oriel and Armstrong (1971) for Fig. 1. Areas in Utah from which larger bivalve arthropods Utah and of Middle Cambrian age have been collected. northern Hintze and Robison (1975) for west-central Utah. The total observed stratigraphic ranges for From studies of Burgess Shale specimens all species of larger bivalve arthropods pres- that are preserved with soft parts and append- ently known from Cambrian strata in Utah are ages, Briggs (1976, 1977, 1978) has demon- plotted on Figure 2. Relevant biofacies rela- strated that morphologically very different tionships and biostratigraphic nomenclature arthropod bodies may possess quite similar have been discussed by Robison (1976, 1982). Robison Ce Richards Larger Bivalve Arthropods from the Middle Cambrian 3

tunate to find one such arthropod for every sev- Agnostoid eral thousand trilobites. Several other localities Interval- Species Ranges have yielded only one or a few specimens of a Zones single species. These less productive localities \ are described under the "Occurrence" of rele- Lc jopyge vant species in the Systematic Descriptions. All amnia localities are within the mountains or ranges Lejopyge plotted on Figure 1. calva 114. Medium-dark-gray, laminated, calcareous mudstone in the uppermost 3 m of ii the Wheeler Formation on a north-trend- Ptychagnostus ;' ing ridge crest; sec. 24 (unsurveyed), T. punctuosus F.) ,': 1 _.'''1 . 21 .....: . ., t; 16S., R. 13W.; about 700 m south of a ,3 a. Q. ù ; ii t.., r: 2 .:.2 4::.: .2 stock-watering trough that is supplied by

.g.,f,, --ti .L, ii water piped from Swasey Spring on the 0 east side of the House Range. t f,..t , t., z e . 4 ...... 4 4°., 4' 4 i-. Gc'h E- -F-I 115. Medium- to dark-gray, calcareous mud- -5 1 Z Pt ychagnostu s stone in the uppermost 30 m of the E i < a tavus II II II II II II li Wheeler Formation at its type locality in E R .. Wheeler Amphitheater; SE1/4 sec. 2, T. co t-. .tt 17 S., R. 13W.; central House Range. < Ptych. gibbus L.) 716. A 60-m unit of yellowish-gray to grayish. a.F, w orange calcareous mudstone in the upper _J I," 0 12 Marjum Formation; exposed in ledges t _0 Ptycliagnostus ••• .t-' along the south side of a large dry wash in praecurrens the SW 1/4NW 1/4 sec. 24 (unsurveyed), T. 17 S., R. 13W.; about 4 km due south of Wheeler Amphitheater in the central House Range. The stratigraphic position i n of this mudstone unit has not been pre- (..) cisely determined; however, its basal beds Perono psis . are probably about 300 m above the base bonnerensis of the Marjum Formation. 745. Brown-weathering, dark-gray, laminated, lime mudstone about 180 to 200 m above the base of the Marjum Forma- tion; SW 1/4SE1/4 sec. 13 (unsurveyed), T. 18 S., R. 13 W.; east side of Bird Can- unnamed yon, House Range.

Repositories. — Most of the specimens described or illustrated in this paper are in the Fig. 2. Observed stratigraphic ranges of larger bivalve ar- Invertebrate Paleontology, Univer- thropods from the Middle Cambrian of Utah. Also in- Museum of cluded is the probable arthropod Pahvantia hastata, new sity of Kansas (KUMIP). A few specimens, in- genus and new species. cluding the holotype of Dioxycaris argenta (Walcott), are in collections of the U.S. Na- tional Museum of Natural History (USNM), Localities. - The majority of specimens de- Washington, D.C. A few unillustrated speci- scribed here are from four localities. These are mens are in the collections of Brigham Young listed below and are referred to by number in University, Provo, Utah, and others are in the the section on Systematic Descriptions. Nontri private collection of Lloyd Gunther, Brigham lobite arthropods are not common at any of City, Utah. these localities. Usually a collector will be for- A cknowledgments. —Without the generous 4 The University of Kansas Paleontological Contributions—Paper 106 contributions and support of several persons, of the University of Kansas. Michael Frederick this study would be far less complete. Excep- aided with photography, R. B. Williams pre- tional contributions were made by Lloyd, pared the figures, and Virginia Ashlock, D. E. Metta, and Val Gunther of Brigham City, G. Briggs, K. E. Caster, and A. J. Rowell criti- Utah, and R. L. Harris of Delta, Utah. Other cally reviewed an early draft of the manuscript. important contributions of material were made This study was supported by National Science by L. F. Hintze and M. S. Petersen of Brigham Foundation grant EAR-7708689 and the Wal- Young University; R. F. Kohl of Willows, Cali- lace E. Pratt Research Fund, which is provided fornia; W. L. Stokes of the University of Utah; to the University of Kansas by the Exxon USA and D. C. Kopaska-Merkel and G. R. Vorwald Foundation.

SYSTEMATIC DESCRIPTIONS

Class MALACOSTRACA Latreille, 1806 Valves from the Wheeler Formation of Subclass PHYLLOCARIDA Packard, 1879 west-central Utah that were previously iden- Order CANADASPIDIDA Novozhilov tified as Canadaspis? sp. (Robison, 1971, p. in Orlov, 1960 797) are reassigned here to the new species Family CANADASPIDIDAE Novozhilov Perspicaris? dila tus. Valves of C. cf. C. perfecta in Orlov, 1960 closely resemble those of P.? dilatus, and differ- Genus CANADASPIS Novozhilov in Orlov, ences are reviewed in the discussion of P.? 1960 dila tus. Type species. —Hymenocaris perfecta Wal- Occurrence. --Four disarticulated valves cott, 1912a, p. 183. (KUMIP 144401, 153893-153895) are pre- The revised generic diagnosis of Briggs served in dark-gray, noncalcareous shale from (1978, p. 447) is followed here. an interval about 23 to 30 m below the top of the Spence Tongue of the Lead Bell Shale on CANADASPIS cf. C. PERFECTA the south side of Antimony Canyon. One artic- (Walcott, 1912a) ulated carapace (KUMIP 153896) is from an Plate!, figures 1-3 unknown stratigraphic level in the Spence For synonymy of C. perfecta, see Briggs Tongue in Miners Hollow. Both localities are (1978, p. 448-449). on the west side of the Wellsville Mountains Description.— Valves ovoid with rather (Fig. 1). All specimens are from the Peronopsis marked posteroventral expansion; hinge line bonnerensis Zone (Fig. 2). straight, antero- and posterodorsal processes mostly absent; antero- and posterodorsal angles Family PERSPICARIDIDAE Briggs, 1978 115 to 120°; free margins lacking spines. Genus PERSPICARIS B riggs, 1977 Adductor-muscle scar not evident. One valve in places showing narrow border. Observed maxi- Type species. — Canadas pis dictynna Simonetta mum length ranging from 28 to 39 mm, maxi- and Delle Cave, 1975, p. 12. mum height from 16 to 25 mm. The generic diagnosis of Briggs (1977, p. 597) Discussion. —Features of these valves, in- is followed here. cluding average size, agree well with those of C. PERSPICARIS? DILATUS, n. sp. perfecta described or illustrated by Briggs, (1978). Although the antero- and posterolat- Plate!, figure 4; Plate 2, figures 5-7 eral angles of C. perfecta were reported by Canadaspis? sp. Robison, 1971, p. 797. Briggs (p. 457) to be about 125° and 130°, re- Etymology. — Latin dilatus, expanded; in spectively, some of his illustrations (e.g., fig. reference to the moderate posteroventral ex- 87) show valves with posterodorsal angles as low pansion of the valves. as about 115°. Thus, the valves from Utah have Holotype. — Right valve, KUMIP 135128 dorsal angles within the range of those from (Pl. 1, fig. 4), from locality 115, House Range. British Columbia; however, average angle for Description. — Valves subtrapeziform with the few valves from Utah may be smaller. moderate posteroventral expansion; maximum Robison & Richards Larger Bivalve Arthropods from the Middle Cambrian 5 height usually about one-third length from (Fig. 2), as is probably also the Bloomington posterior margin. Hinge line normally straight. specimen. Antero- and posterodorsal processes usually weakly developed but seemingly broken or not preserved on some specimens; angles generally PERSPICARIS? ELLIPSOPELTA, n. sp. 100 0 and 110°, rarely ranging from between Plate 2, figures 1-4 about 95 to 115°. Small adductor-muscle scar commonly visible on interior surface of valve; Etymology. —Latin ellipsis, ellipse; pe/ta, situated anterodorsally with distance from shield. In reference to the elliptical shape of the hinge line almost twice that from anterior valves. margin. Free margins lacking spines. Observed Holotype. — Right valve, KUMIP 153902 maximum length of valve ranging from 29 to (Pl. 2, fig. 1), from locality 716. 55 mm, maximum height from 18 to 34 mm. Description.-- Valves elliptical in outline, Discussion.— Of the two previously de- usually with maximum height only slightly pos- scribed species of Perspicaris, valves of P.? terior from center, giving an aspect approach- dilatus most closely resemble those of P. dic- ing symmetry. Hinge line normally having tynna (Simonetta and Delle Cave, 1975) as slight dorsal convexity. Antero- and postero- described by Briggs (1977); however, valves of dorsal processes weak or absent; angles ranging P.? dilatus are almost four times greater in from about 110 to 115°. Hinge-line embay- average size. They also have a more anteriorly ments on articulated carapaces of subequal situated adductor-muscle scar, usually a better depth. Adductor-muscle scar not evident and developed posterodorsal process, and no evi- free margins lacking spines. Observed maxi- dence of an original border. Valves of P.? mum length of valve 36 to 71 mm, maximum dilatus are similar in average size to those of P. height 20 to 40 m. recondita Briggs (1977) but differ otherwise by Discussion. ---Valves of P.? ellipsopelta are having a straight hinge line, smaller antero- characterized by a near uniform convexity of and posterodorsal angles, a less strongly curved the ventral margin and normally a slight con- anterior margin, and an adductor-muscle scar vexity of the hinge line. They further differ positioned closer to the anterior margin. from those of P.? dilatus by usually having Valves of P.? dilatus may be difficult to dis- slightly smaller antero- and posterodorsal tinguish from those of Canadaspis cf. C. per- angles and a somewhat larger average size. fecta. Generally, however, maximum height Among more obvious differences, the valves relative to maximum length is less in P. of P.? ellipsopelta are on the average at least dilatus, and the antero- and posterodorsal five times larger than those of P. dictynna, angles are smaller. Although weak, the antero- have a convex rather than a straight hinge line, and posterodorsal processes are usually better and have better developed antero- and postero- developed in P. dilatus. Also, adductor- dorsal processes. Valves of P.? ellipsopelta dif- muscle scars are commonly seen on valves of P. fer from those of P. recondita by having dilatus but are not evident on valves of C. cf. C. smaller antero- and posterodorsal angles, less perfecta. strongly curved anterior and posterior margins, Occurrence.—Seven valves (KUMIP and a somewhat larger average size. 135128-135133, 153899) are preserved in Occurrence.—Twelve disarticulated valves medium-light-gray to dark-gray calcareous and three carapaces (KUMIP 153902-153910, mudstone from the upper 30 m of the Wheeler 153924-153929) are from locality 716 in the Formation at locality 115 in the House Range. Marjum Formation. Two additional carapaces Another nearly complete valve (KUM IP (KUMIP 159155, 159156) are in brown shale 134189) is in pale-brown noncalcareous mud- from about 60 m above the base of the Marjum stone from about 75 m above the base of the Formation in the SE1/4SE 1/4 sec. 4 (unsur- Hodges Shale Member of the Bloomington For- veyed), T. 18 S., R. 13 W. All specimens are mation on the south side of Baker Canyon in from the Ptychagnostus punctuosus Zone (Fig. the Wellsville Mountains. The Wheeler speci- 2) in the House Range (Fig. 1) and were col- mens are from the Ptychagnostus atavus Zone lected by Lloyd and Metta Gunther. 6 The University of Kansas Paleontological Contributions Paper 106

Class and order uncertain little meaning. For recent discussions of possi- Family PROTOCARIDIDAE Miller, 1889 ble affinities and classification, see Briggs Genus BRANCHIOCARIS Briggs, 1976 (1976, p. 14) and Glaessner (1979, p.24-25). BRANCHIOCARIS? sp. Plate!, figures 5,6 Genus DICEROCARIS, new Diagnosis— Valves subelliptical in outline Type with slight expansion in inferred anteroventral species.—Dicerocaris opisthoeces, n. sp. direction; maximum height slightly anterior Etymology.—Greek dikeros, from center. Hinge line straight or having two-horned and karis, shrimp; slight dorsal convexity. Antero- and postero- in reference to the pair of hornlike spines dorsal processes of moderate size; angles ap- on the anterodorsal parts of the carapace. proximately 115°. Adductor-muscle scars and Description.—Carapace surface reticulation not evident. Free margin bivalve with hinge line; each valve with a moderately nonspinose. Maximum length and height of large spine projecting upward one valve 56? and 38 mm, respectively; another and slightly forward from 30? and 20 mm. anterodorsal areas. Valves subelliptical with minor Discussion. Two nearly complete valves posteroventral expansion; anterodorsal process small; from the Drum Mountains differ in form from posterodorsal corner produced into large process projecting straight all others in the collections from Utah. With a back. Discussion. —Carapaces combination of moderate-sized hinge processes of Dicerocaris differ from those of all other genera of larger and relatively large maximum height com- bivalve arthropods by the presence of the two pared to maximum length, they most closely large, closely spaced spines on opposite sides of resemble species previously assigned to Bran- the anterior hinge line. The combination chiocaris and Tuzoia; however, they clearly dif- of small anterodorsal process and large postero- fer from representatives of Tuzoia by lacking dorsal process is also distinctive. marginal spines and surface reticulation. Thus, Of the the two specimens possibly represent Branchio- Cambrian genera with ovate or subelliptical valves, only caris, but the available material is inadequate Dicerocaris, Isoxys, and Silesicaris are always longest at for positive identification. the hinge line rather than lower Because adductor-muscle scars are not evi- on the valves. In addition to the anterodorsal spine small dent on the two valves, it is not possible to and anterodorsal determine anterior margins with certainty. process, valves of Dicerocaris further differ from those of Isoxys Evidence of such scars on future discoveries and Silesicaris by having could be taxonomically useful because Branch- relatively greater height with less broadly rounded iocaris seems to be the only described Cam- antero- and posteroventral margins. brian genus with short ovoid valves that expand In valve outline, representatives of Dicero- anteroventrally. caris resemble those of Dioxycaris, Pseudoarc- Occurrence.—Two disarticulated valves tolepis, and Tuzoia; however, in addition to differences in spine morphology, (KUMIP 153897, 153898) have been collected valves of Dice- rocaris have smaller by Lloyd Gunther from pale-yellowish-brown, anterodorsal processes. Dicerocaris is a monotypic genus that pres- calcareous mudstone of the uppermost 50 m of ently is known only from the late Middle Cam- the Wheeler Formation; NE 1/4 NW 1/4 sec. 20, brian (Ptychagnostus atavus T. 15 S., R. 10 W.; Drum Mountains (Fig. 1). Zone) of Utah. Both are from the Ptychagnostus atavus Zone (Fig. 2). DICEROCARIS OPISTHOECES, n. sp. Plate 3, figures 1-3 Etymology. Family uncertain —Greek opisthoekes, pointed in back; in reference to the large posterodorsal Until details of the soft parts and append- process. ages are known for such genera as Dicerocaris, Holotype. — Carapace preserved in lateral Dioxycaris, Isoxys, Pro boscicaris, Pseudoarcto- aspect with valves slightly offset and rotated, lepis, and Tuzoia, assignments to families have KUMIP 134939 (Pl. 3, figs. la, b); from local- Robison & Richards— Larger Bivalve Arthropods from the Middle Cambrian 7 ity 115, House Range. Occurrence. —One carapace and one disar-

Description. — Hinge line straight in lateral ticulated valve (KUMIP 134939, 134940) were view. Maximum valve height behind midpoint collected by R. L. Harris and Lloyd Gunther, of valve. Anterodorsal angle approximately respectively, from the upper 30 m of the 125°, posterodorsal angle between 90 0 and Wheeler Formation at locality 115, House 95°. Length of hornlike spine near anterodor- Range. Another valve (KUMIP 153900) was sal corner of valve approximately one-third collected by G. R. Vorwald from 29 m below maximum valve height. Observed maximum the top of the Wheeler Formation in the length of valve below hinge line ranging from SEV4SW 1/1 sec. 21, T. 15 S., R. 10 W., Drum about 15 to 19 mm; maximum height ranging Mountains (Fig. 1). All three specimens are from about 11 to 15 mm. Observed maximum preserved in medium- to dark-gray calcareous length of hinge line about 23 mm. Adductor- mudstone. All are from the Ptychagnostus muscle scar and surface reticulation not evi- atavus Zone (Fig. 2). dent. Genus DIOXYCARIS GUrich, 1929 Discussion. — Valves of D. opisthoeces are characterized by the moderately large antero- Dioxvcaris Giirich, 1929, p. 36; Rolfe, 1969, p. dorsal spine, the large posterodorsal process, a Ii325; Briggs, 1976, p. 13. relatively long hinge line, and a relatively small Dyoxycaris Krestovnikov, 1961, p. 5. posterodorsal angle. No described species has a Type species. — Leperditia? argenta Wal- similar combination of characters. cott, 1886, p. 146-147. Although the valves assigned to D. opis- Diagnosis. — Carapace bivalve with hinge thoeces are relatively small, they do not seem to line. Valves subelliptical with slight posteroven- represent early instars of another species. As tral expansion; anterodorsal process of mod- demonstrated by a wide range in size, associ- erate size, posterodorsal process large; lateral ated specimens of Pseudoarctole pis sharpi show surfaces and free margins lacking spines. little change in valve form during ontogeny. If Discussion. — As demonstrated in this D. opisthoeces underwent similar ontogenetic paper, the carapaces of Dioxycaris development, larger valves could still be easily 1929, and Pseudoarctolepis Brooks and Caster, discriminated from all other species described 1956, are remarkably similar except that Di- here. oxycaris seems to lack a pair of prominent Because attachment of the anterodorsal lateral spines. Few specimens of Dioxycaris spine in D. opisthoeces is slightly abaxial from have been collected and all are poorly pre- the hinge line, the spine commonly remains served. Thus, it is possible that the absence of covered by a thin veneer of matrix. This usually lateral spines is the result of inadequate preser- requires careful removal before the spine can vation and the two genera may be synonyms. be seen. Nevertheless, until incomplete preservation of On the holotype carapace, the spine on the spines is actually demonstrated for Dioxycaris, left valve projects upward and slightly forward we provisionally recognize both names as being (PI. 3, fig. lb), apparently in normal position. valid. On the right valve (Pl. 3, fig. la), cuticle As presently defined, carapaces of Dioxy- around the base of the spine has been broken caris differ from those of Pseudoarctolepis by and the spine is rotated so that it projects the absence of a pair of large lateral spines. obliquely backwards. Spines and associated Dioxycaris differs from Branchiocaris, Canada- cuticle on the two single valves have been sim- spis, and Perspicaris by having larger terminal ilarly deformed (Pl. 3, fig. 2a, 3). This seem- hinge processes, particularly on the posterior ingly altered orientation of spines is such that it end. To date, D. argenta is the only species is difficult to explain as the result of tapho- that has been assigned to Dioxycaris. nomic compaction. Perhaps it resulted from escape activities during ecdysis. A random dis- DIOXYCARIS ARGENTA (Walcott), 1886 tribution of disarticulated trilobite sclerites in Plate 3, figures 4-6 matrix between valves of the holotype carapace suggests that the body was missing (possibly Leperditia? argenta Walcott, 1886, p. 146-147, because of ecdysis) at the time of burial. pl. 8, fig. 5. 8 The University of Kansas Paleontological Contributions—Paper 106

Dioxycaris argenta (Walcott) Giirich, 1929, p. is from a unit that is presently recognized as the 36, text-fig. 1.3; Briggs, 1976, p. 13-14, pl. lower member of the Ophir Shale. One of us 6, figs. 3, 4 (see for additional synonymy). (R.A.R.) has searched without success for addi- Tuzoi argenta (Walcott) Resser, 1938, P. 25. tional topotype specimens of D. argenta and concludes that Holotype. — Probable right valve, USNM the holotype locality now is 15401 (Pl. 3, fig. 4). probably covered by Utah Highway 152 on the north side of Big Cottonwood Diagnosis. -- Hinge line straight in lateral Canyon in the Wasatch Range (Fig. 1). The view. Maximum height of valve at about one- early mining town of Argenta third length from posterior margin. Anterodor- to which Walcott (1886, p. 147) referred, sal angle between 105° and 110°, posterodorsal and for which D. argenta is named, no longer exists. angle between 110° and 115°. Narrow border According to Walcott (1912b, adjacent to free margins probably adventitious, p. 189), the holotype was associated with resulting from taphonomic compaction. Ob- Bat hyuriscus pro- ductus ( Glossopleura producta; see Palmer, served maximum length of valve ranging up to 1954, p. 67) and other taxa. That fauna is 62 mm. Maximum height (30 mm) of longest presently considered to represent the Glosso- valve (Pl. 3, fig. 5) probably reduced by pleura Assemblage-zone of early Middle Cam- taphonomic wrinkling of ventral valve margin. brian age (the Glossopleura Zone is temporally Discussion. — Valves of Dioxycaris argenta equivalent to part of the Peronopsis bonneren- most closely resemble those of Pseudoarctole pis sis Zone, Fig. 2). Giirich (1929) and sharpi but lack a pair of large lateral spines. some subsequent authors erroneously indicated that These species also differ slightly in their antero- the holotype is from Lower Cambrian and posterodorsal angles, which average about strata, whereas Brooks and five degrees greater in D. argenta. The much Caster (1956) erroneously indicated that it is from higher percentage of disarticulated valves in D. the House Range (see summary by Briggs, 1976, p. argenta than in P. sharpi suggests a weaker 13). connection along the hinge. It also could indi- Additional specimens that we tentatively assign to D. argenta include three cate the deposition of remains of D. argenta valves under higher energy conditions; however, both (KUMIP 135145-135147) from about 3 m below species are commonly associated with articu- the top of the Spence Tongue of the Lead Bell Shale on the lated trilobites and eocrinoids, suggesting little south side of Antimony Can- yon and one valve (KUMIP 153911) difference in energy conditions at the times of from near burial. the base of the Ute Formation on the north side of Hansen Canyon; all collected by Lloyd and Valves here assigned to D. argenta vary Val Gunther from the west side of the Wells- moderately in maximum height relative to ville Mountains (Fig. 1), and all from the maximum length. The range of variation, how- Glossopleura Assemblage-zone. ever, is no greater than that shown by speci- Other specimens that possibly represent D. argenta are one mens of P. sharpi in which better preservation valve and some fragments (USNM 56513) allows the conclusion that much of the varia- from the Ute Formation (USNM locality 54a; Wal- tion resulted from taphonomic deformation of cott, 1912b, p. 151, 202) in Blacksmith Fork the ventral valve margin. Thus, lacking well- Canyon of the Bear preserved material, we can attach little tax- River Range (Fig. 1). onomic significance to shape of the ventral Genus PROBOSCICARIS Rolfe, 1962 margin or to ratios based on maximum height. Without explanation, Resser (1938, p. 25) Pro boscicaris Rolfe, 1962, p. 2; 1969, p. R327. reassigned D. argenta to Tuzoia. Valves of Type species. —Pro boscicaris agnosta Tuzoia can be readily differentiated from those Rolfe, 1969, p. 2-4. of Dioxycaris by such features as a lateral Diagnosis. —Carapace valves subtrape- caria and spines on the posterior margin. zoidal to subovate and nonspinose. Inferred Representatives of D. argenta lack these and anterodorsal area commonly produced into other features that might support Resser's tax- large spatulate process. onomic reassignment. Discussion. —Valves representing Pro bosci Occurrence. —The holotype of D argenta are unusually variable in lateral outline.-caris Robison & Richards Larger Bivalve Arthropods from the Middle Cambrian 9

Nevertheless, the general form, commonly Discussion. - A single specimen from Utah characterized by a prominent spatulate pro- seemingly represents a carapace that has cess, is unique among the larger bivalve arthro- undergone selective dissolution, leaving a thin pod genera of Cambrian age. internal mold. It has a maximum length of 59 Three species of Pro boscicaris have previ- mm and a maximum height of 29 mm. Al- ously been described. All are from the Burgess though relatively small, it has an outline that Shale. P. agnosta Rolfe, 1962, is based on two fits well within the range of variation illustrated particularly large valves with relatively small by Rolfe (1962, text-fig. 1) for valves of P. anterodorsal processes, whereas P. obtusa Si- agnosta from the type locality in British Co- monetta and Delle Cave, 1975, is based on a lumbia. Compared to the several valves illus- single valve with a "wider," more obtuse pos- trated by Rolfe, the specimen from Utah most terodorsal region and "obsolete - anterodorsal closely resembles USNM 139875, differing corner. Rolfe (1962, p. 6) mentioned the possi- mainly by its slightly convex rather than bility that "P. ingens is simply an older instar of slightly concave hinge line and a more angular P. agnosta," but further noted that "such radi- posterodorsal corner. We do not consider these cal changes in shape are not common except in minor morphological differences to be suffi- early ontogeny." Because no combination of cient to warrant taxonomic separation. characters readily distinguishes the type speci- An unusual color pattern renders the Utah mens of P. ingens and P. obtusa from the specimen difficult to photograph. The spatu- several variable specimens that have been late anterior end is light gray with a bleached assigned to P. agnosta (Rolfe, 1962, fig. 1), we aspect. A large central area is light brown and consider all of these to represent a variable a crescentic area at the posterior end is grayish population of a single species. black. The matrix is a pale yellowish-brown, To date, Proboscicaris is known from Can- calcareous mudstone. We suggest that the dark ada (British Columbia) and the United States posterior deposit may represent carbonaceous (Utah). It has an observed stratigraphic range residue from decomposition of body tissues. through part of the lower Middle Cambrian Occurrence.- One specimen (KUMIP (Ptychagnostus praecurrens to Ptychagnostus 153901) was collected by Lloyd Gunther from a tavus zones). the upper 30 m of the Wheeler Formation in the SE VI SE1/4NE 1/4 sec. 35 (unsurveyed), T. 17 S., R. 13 W. in the central House Range (Fig. PROBOSCICARIS AGNOSTA Rolfe, 1962 1). Plate 4, figure 1 The discovery of P. agnosta in Utah extends the observed stratigraphic range of the species Proboscicaris agnosta Rolfe, 1962, p. 2-4, Pl. the Ptychagnostus praecurrens 1, figs. 1, 2, text-fig. 1; 1969, Fig. 150.7; upward from Simonetta and Delle Cave, 1975, p. 8, Pl. Zone to the Ptychagnostus atavus Zone (Fig. 2). observed geographic range 6, fig. 1, Pl. 37, figs. 1, 5. It also extends the to Utah. Proboscicaris ingens Rolfe, 1962, p. 4-6, Pl. 1, from British Columbia fig. 3, text-fig. 2; Simonetta and Delle Genus PSEUDOARCTOLEPIS Cave, 1975, p. 8. Brooks and Caster, 1956 Pro boscica ris obtusa Simonetta and Delle Cave, 1975, p. 8-9, PI. 6, fig. 2, PI. 47, fig. Pseudoarctolepis Brooks and Caster, 1956, p. 5. 10; Rolfe, 1969, p. R327. Diagnosis. - Valves elongate, usually with Type species. - Pseudoarctole pis sharpi large, spatulate anterodorsal process. Hinge Brooks and Caster, 1956, p. 11-12. line slightly concave to slightly convex. Poste- Diagnosis. -Carapace bivalve with hinge rodorsal angle obtuse and usually rounded. line. Valves subelliptical with slight posteroven- Ventral margin concave in anterior part and tral expansion; large, long, posteriorly curved convex in posterior part. Posterior margin spine projecting from anteromedian region; commonly with minor indentation in middle or anterodorsal process of moderate size, postero- lower part. Surface rarely bearing patches of dorsal process large; free margins lacking fine reticulate ridges. spines. Rostrum absent. 10 The University of Kansas Paleontological Contributions—Paper 106

Discussion. —The most distinctive feature tween 115° and 120°. Adductor-muscle scar of Pseudoarctolepis is the pair of prominent not evident on laterally compressed valves but lateral spines. These give the carapace a super- site coinciding with small deformed area in ficial resemblance to the cephalothoracic der- anterodorsal region on some dorsoventrally mal armor of some Devonian arthrodire fish; compressed valves (see Pl. 5, fig. 6). Large, however, composition, structure and various posteriorly curved spine attached just above morphologic details indicate an arthropod af- midheight of valve and about one-third of finity. length from anterior margin; length of spine Previously, the taxonomic concept of Pseu- ranging from about one-fourth to one-third doarctolepis was based solely on the holotype of that of valve. Surface of valves smooth except P. sharpi, which is an articulated carapace that for wrinkles and folds seemingly caused by is preserved in parallel aspect. Previous authors taphonomic compaction. have failed to recognize that approximately the Discussion. — Specimens of P. sharpi are lower half of each valve is folded beneath the preserved as dark carbonaceous films in calcar- exposed dorsal carapace. This fact resulted in eous mudstone. All show effects of deformation an erronneous restoration of original valve related to compaction, which suggests that the form, which is corrected in the following review original carapace was thin and not mineral- of P. sharpi. ized. Lack of decalcification in associated The possibility that Pseudoarctolepis is a trilobites and echinoderms further supports junior synonym of Dioxyca ris is reviewed in the this suggestion. discussion of Dioxycaris. When an articulated carapace of P. sharpi came to rest on the seafloor, the projecting lateral spines normally would have prevented the carapace from resting on its sides. This PSEUDOARCTOLEPIS SHARPI seemingly is why most articulated carapaces are Brooks and Caster, 1956 preserved in parallel aspect. Also, as the result Plate 4, figures 3-5; Plate 5, figures 1-6; Plate 6, of dorsoventral compaction figures 1, 2 in mud, approximately the lower half of each valve is usually Pseudoarctolepis sharpi Brooks and Caster, folded underneath or crumpled beneath the 1956, p. 11-12, P1.2 and text- fig. 1.1,2. upper half. This latter condition in the bob- type was not recognized by Brooks and Caster Material. —The holotype, which has been (1956, text-fig. 1.2), who illustrated a restora- contributed to the University of Kansas tion in lateral view of a relatively long, antero- Museum of Invertebrate Paleontology (KUMIP ventrally expanded valve with the large lateral 159141; Pl. 5, fig. 6) by Prof. W. L. Stokes, is spine attached near the ventral margin. New an articulated carapace preserved in parallel specimens, several with nonfolded valves (e.g., aspect. Collecting by several persons during the Pl. 5, figs. 2-4), demonstrate a relatively past two decades has produced an additional greater valve height than originally described. 23 articulated carapaces preserved in parallel Moreover, the lateral spine is actually attached aspect, 3 carapaces preserved in lateral aspect, at about midheight on the valve (Pl. 4, fig. 3; and 16 disarticulated valves preserved in lateral PI. 5, fig. 3) and the valve expands slightly in a aspect. These additional specimens are now in posteroventral direction (Pl. 5, figs. 3,4). collections at the University of Kansas (KUMIP Style of carapace deformation in P. sharpi 135134-135142, 144383-144385, 144396- provides additional information about mor- 144398, 153912-153916, 159142-159153), phology as well as clues to the attitude of cara- Smithsonian Institution (USNM 159918), and paces at the time of their burial. For example, Brigham Young University (unnumbered speci- noticeably more surface area of either one or mens), and in the private collection of Lloyd the other valve is exposed in articulated cara- Gunther. paces that are preserved in parallel aspect Diagnosis. —Hinge line straight or slightly (e.g., Pl. 6, fig. 1). This probably resulted from convex upward in lateral view. Maximum valve the carapaces being tilted on the seafloor at the height at about one-third length from posterior time of burial, each resting on the ventral mar- margin. Anterodorsal angle usually between gin and spine tip of either one or the other of 110° and 115°, posterodorsal angle usually be- the valves (Fig. 3A). Vertical compaction sub- Robison & Richards—Larger Bivalve Arthropods from the Middle Cambrian 11 sequently caused some parts of the carapace to valve margins. On some specimens (Pl. 5, fig. be flattened while other parts were folded un- 1), these wrinkles may weakly demarcate a rim derneath or crumpled in place (Fig. 3B). These or borderlike area; however, the original sur- differences in area of exposure also indicate face was probably smooth and all such features that the large lateral spines projected outward probably represent nothing more than defor- at approximately right angles to the valve sur- mation by compaction of thin, convex valves. faces (Fig. 3A). Otherwise, if the spines pro- Amount of surface wrinkling near the ventral jected downward, the resting carapace would margin of disarticulated valves is particularly have had less tilt and a more symmetrical pres- variable and has variably altered the original ervation; if the spines projected upward, the valve height as well as the position of the spine asymmetry would have been more pronounced. base relative to the ventral margin (cf. Pl. 5, figs. 1-4). Thus, for most valves, original length to height ratios are difficult, if not impossible, hinge to determine. The zone of wrinkling also prob-

lateral spine ably indicates that the lower part of the valve left valve originally had a greater adaxial curvature than did parts above (Fig. 3A). right volve Two small, broken bulges near the anterior margin of the holotype carapace were considered by Brooks and Caster (1956, p. 10, pl. 2) the position of compound A possibly to "represent eyes or more likely, the places of origin of an hinge wrinkled carapace adductor carapace muscle." Such features are present on other carapaces of P. sharpi preserved in parallel aspect, but are not evident on wrinkled carapace valves preserved in lateral aspect. We concur Fig. 3. Effects of compaction on the carapace of Pseudo- with the suggestion that they most likely repre- arc:ale/n.5 sharp:, shown in section through lateral spines. sent insertion sites for adductor muscles, and A, Carapace resting on tip of left spine and ventral margin we have found no evidence for an ocular func- of left valve. B, After compaction in mud, upper left and tion. Moreover, the sites seem to have been dif- lower right parts of valves are crumpled and narrower than ferentially deformed only during dorsoventral upper right and lower left parts. compaction of valves. Although the carapace is produced anteri- Where the large lateral spine attached to orly into a prominent hingeline process, there is the carapace wall, the spine has a broad, sub- no evidence of a separate, movable rostral plate triangular cross section with an obtuse apex on as in some of the Phyllocarida (Rolfe, 1969). the dorsal surface (Pl. 4, fig. 3). Additional An even larger posterior hingeline process is evidence for a dorsal ridge on these spines has present, and in some specimens these processes not been observed; however, such a ridge may are medially torn or cleft (e.g., Pl. 6, fig. 1). As have been obliterated by taphonomic compac- noted by Brooks and Caster (1956), this medial tion. cleft indicates an axis of weakness and articula- An indication of original valve convexity is tion rather than a caria. provided by articulated carapaces of P. sharpi Available valves of P. sharpi range from 20 that are preserved in parallel aspect with the to approximately 110 mm in maximum length lower parts of their valves folded underneath and provide some information about ontogeny. the upper parts (e.g.. Pl. 5, figs. 5, 6). Maxi- Only one fragmentary valve exceeds 75 mm in mum width, which may have been slightly ac- length and it is appreciably larger than all centuated by compaction, is at the position of others in the collection. Modal length is be- the lateral spines. The anterior decrease in tween 50 and 55 mm (Fig. 4). The smallest width is more pronounced than is the posterior specimens (Pl. 4, figs. 2, 4) differ little in decrease. lateral valve outline from those in the modal Single valves preserved in lateral aspect size range. Lateral spines, however, show a niso- commonly show a zone of wrinkling near the metric growth, increasing from about a third of 12 The University of Kansas Paleontological Contributions—Paper 106

the valve length in small specimens (PI. 4, fig. pattern of fine ridges; mesh finer on lateral 2) to more than half in larger specimens (Pl. 4, ridge and near valve margins. fig. 5a, b). These lateral spines may have func- Discussion. —Valves of Tuzoia differ from tioned as stabilizers and may indicate a swim- those of most other larger bivalve genera of ming mode of life. The anisometric increase in Cambrian age by development of marginal spine length was probably coordinated with spines and conspicuous surface reticulation. In volumetric increase in the body. general outline, excluding spines, valves of P. sharpi remains the only species referable Tuzoia closely resemble those of Canadaspis to Pseudoarctole pis. but have larger terminal hinge processes. The Occurrence. — All specimens of P. sharpi terminal hinge processes of Tuzoia closely were collected from either the upper 40 m of resemble those of Dioxycaris and Pseudoarcto- the Wheeler Formation in the House Range or lepis, but otherwise the posteroventral margin the upper 80 m of the Wheeler Formation in is usually more expanded in Tuzoia. the Drum Mountains (Fig. 1). The specimens The thin cuticle of Tuzoia commonly shows are from several localities, the most productive evidence of taphonomic deformation. Also, being 114 and 115. All are from the Ptychag- where several valves are available from a single nostus atavus Zone (Fig. 2). locality, the number of marginal spines tends to vary. For some specimens, careful prepara- tion has demonstrated that spines may remain 10 - buried in matrix. In other specimens, spines 2 have probably been broken off, but this may be difficult, if not impossible, to determine. Out- n=35 a. line of valves and spacing of spines may be 5 o significantly modified by compression, the principal amount usually depending on orientation of the specimen at the time of its burial. o Compression may also obliterate the fine surface reticulation. Some change in spine mor- 20 30 40 50 60 70 80 90 100 110 phology may have occurred during ontogeny. maximum length in mm. Thus, all of these possibilities deserve evalua- Fig. 4. Observed size distribution in Pseudoarctolepis tion in discrimination of species of Tuzoia. sharpi Spine morphology of the dorsal margin is especially difficult to determine for valves of Genus TUZOIA Walcott, 1912 Tuzoia. Where carapaces are preserved in par- Tuzoia Walcott, 1912a, p. 187; Henriksen, allel aspect, some show that the hinge line 1928, p. 15; Resser, 1929, p. 6-7; Shimer separates closely spaced pairs of spines (e.g., and Shrock, 1944, p. 655; Brooks and Resser, 1929, Pl. 7, fig. 3). On bedding sur- Caster, 1956, p. 13; Opik, 1968, p. 10; faces where disarticulated valves are mostly Rolfe, 1969, p. R328; Simonetta and Delle preserved in lateral aspect, some or all of the Cave, 1975, p. 7-8; Briggs, 1977, p. 618; dorsal spines may remain covered by a thin ve- Glaessner, 1979, p. 23. neer of matrix and their presence may easily go undetected. Consequently, particular care Type species. — Tuzoia retifera Walcott, should be exercised in using dorsal spines for 1912a, p. 187, Pl. 33, fig. 2. taxonomic discrimination. Diagnosis. — Carapace bivalve with hinge In most species of Tuzoia, a lateral ridge line. Valves ovoid, usually with marked pos- extends nearly the full length of each valve, teroventral expansion; lateral ridge commonly generally in a central position and parallel to extending almost full length of valve at about the hinge. Some authors (Resser, 1929, p. 6; midheight, rarely vestigial; terminal hinge Raymond, 1935, p. 228; Rolfe, 1969, p. R328) processes medium to large. Spines variable; have interpreted this ridge to be an original none to several along dorsal margin, one to feature of the valve. Because of a somewhat several along posterior margin, rare on ventral variable position, other authors (Simonetta and margin. Surface normally covered by reticulate Delle Cave, 1975, p. 8; Briggs, 1977, p. 620) Robison & Richards—Larger Bivalve Arthropods from the Middle Cambrian 13 have argued that the ridge is a secondary fea- Gunther, who have generously contributed ture resulting from compression. After review many of the specimens on which this study is of this matter, we conclude that the lateral based. ridge is a primary feature and that its position Holotype.—Left valve, KUMIP 153917 (Pl. is variable because of secondary compression. 7, fig. 2), from locality 716, House Range. A primary origin for the ridge is indicated by a Description. —Valve commonly with single, major difference in average diameter of reticu- short, broad-based, vertical spine on dorsal lae on the ridge and on adjacent valve surfaces, margin, approximately one-fifth length from the latter generally being 5 to 10 times greater. anterior end; hinge line, in lateral view, Moreover, reticulae on the ridge rarely show straight to slightly convex behind spine, offset evidence of differential deformation. In most to slightly lower level in front of spine. Antero- species, the largest marginal spine is on a direct dorsal process moderate to large, usually point- line with the lateral ridge, which suggests an ing straight forward; posterodorsal process long original functional relationship. In valves and commonly directed upward and backward where the lateral ridge is markedly closer to at about 45° angle from hinge line. Anterodor- either the dorsal or ventral margin, the surface sal angle usually between 110° and 115°, between commonly shows abnormal wrinkling posterodorsal angle usually between 108° and parallel to the ridge (e.g., Resser, 1929, Pl. 3, 113°. Maximum valve height about two-fifths fig. 1) or lengthwise stretching of reticulae length from posterior margin. Lateral ridge (e.g.. Resser, 1929, P1.2, fig. 1). Both of these vestigial; seemingly confined to anterior half of features indicate secondary differential de- valve, about one-third way down from dorsal crease in height within that part of the valve margin; indicated mainly by finer pattern of and a corresponding secondary displacement in surface reticulation. Posterior margin with 3 or position of the lateral ridge. 4 spines of short to moderate length. Observed Henriksen (1928, p. 13) suggested that maximum length of valve, exclusive of spines, specimens of Anomalocaris may represent a to 125 mm and maximum height to 99 mm. segmented body belonging with carapaces of Discussion. —Most of the better preserved either Tuzoia or Carnarvonia. In support of valves of T. guntheri show only three posterior Henriksen's idea, Resser (1929, p. 6) noted that marginal spines (PI. 7, fig. 1). The highest and "three of the four localities yielding Tuzoia second largest of these is situated at about mid- have thus far also furnished Anomalocaris." height on the margin. The smallest spine is in Briggs (1979) has recently given a detailed the lowest position, whereas the largest and analysis of Anomalocaris and reinterpreted the longest spine is intermediate in position. specimens as ambulatory appendages of an ex- Rarely, a tiny fourth spine is present about ceptionally large arthropod, probably exceed- midway along the upper half of the posterior ing one meter in length. A similar conclusion margin (Pl. 7, fig. 2). was recorded by Bergstrom (1979, p. 10-11), The lateral ridge on valves of T. guntheri is who interpreted the specimens of A no malocaris indicated by only a weak line of finer reticulae to be possible limbs of a large, uniramian ar- on the anterior half of better preserved valves thropod. (Pl. 7, fig. 2). As this is the stratigraphically Among the Cambrian genera of larger bi- youngest species of Tuzoia, and earlier species valve arthropods. Tuzoia is one of the most with short lateral ridges are unknown, a late common and most widely distributed. It pres- phyletic reduction of ridge length may be indi- ently has an observed stratigraphic range from cated. the upper Lower Cambrian to the upper Mid- In valve outline and surface reticulation, T. dle Cambrian. Geographically, it is known guntheri most closely resembles T. retifera from Australia (South Australia), Canada Walcott (1912a) and T. polleni Resser (1929); (British Columbia), China (Manchuria), and however, T. guntheri differs from both by hav- the United States (Pennsylvania, Utah, Ver- ing only a vestigial lateral ridge, which is mont). situated higher on the valve. Consistent differences in spine morphology are also evident. GUNTHER', n. sp. TUZOIA The number of marginal spines in T. guntheri 9, figure 2 Plate 7, figures 1, 2; Plate 8, figures 4, 5; Plate is seemingly less variable than in T. retifera. Etymology. -- After Lloyd, Metta, and Val Further, in T. guntheri the largest and longest 14 The University of Kansas Paleontological Contributions-Paper 106

spine is low on the posterior margin and well by being less smoothly curved. Also, the maxi- below the level of the lateral ridge, whereas in mum height is near the center of the valve T. retifera the largest marginal spine is situated rather than being well back on the posterior at about midheight, directly behind the lateral half. The cuticle is not well enough preserved ridge. Valves of T. guntheri further differ from to determine the presence or absence of origi- those of T. polleni by having a less spinose dor- nal surface reticulation. sal margin. Among Tuzoia-like species, only T.? dun- Occurrence. -Three valves (KUMIP bani Resser (1929, Pl. 7, fig. 1) has marginal 153917, 153919, 153920), including the spines as large as those of T. peterseni, but the holotype, have been collected from locality form and arrangement of spines differ greatly 716. Twenty-two valves and three carapaces in the two species. (KUMIP 134188, 134790-134798, 144390- Occurrence. -Upper Wheeler Formation 144395, 144399, 144400, 144982, 144987- (Ptychagnostus atatrus Zone) near Marjum 144989, 153918, 153921) have come from lo- Pass, House Range. cality 745. An additional valve (KUMIP 159154), the largest known, is from about 60 m TUZOIA RETIFERA Walcott, 1912 above the base of the Marjum Formation in the Plate 8, figure 3 SE VISE1/4 sec. 4 (unsurveyed), T. 18 S., R. 13 Tuzoia retifera Walcott, 1912a, p. 187, Pl. 33, W. All specimens are preserved in brown- fig. 2; Resser, 1929, p. 7-8, Pl. 1, figs. 1, 2, weathering, laminated, lime mudstone in the Pl. 4, fig. 3; Giirich, 1929, p. 42, text-fig. Marjum Formation of the House Range (Fig. 2.4; Shimer and Shrock, 1944, Pl. 278, figs. 1). All are from the Ptychagnostus punctuosus 15, 16; Simonetta and Delle Cave, 1975, p. Zone (Fig. 2), and most were collected by either 8, Pl. 6, fig. 8, Pl. 46, figs. 3-5, Pl. 47, fig. Lloyd Gunther, Metta Gunther, Val Gunther, 1, Pl. 48, figs. 1-4, Pl. 49, fig. 4. or R. L. Harris. Tuzoia burgessensis Resser, 1929, p. 8, Pl. 2, fig. 1, Pl. 3, fig. 1; Brooks and Caster, TUZOIA? PETERSENI, n. sp. 1956, text-fig. 1.5; Rolfe, 1969, Fig. 152.6. Tuzoia canadensis Plate 9, figure 1 Resser, 1929, p. 8, Pl. 2, figs. 2, 3. Etymology. -After Prof. M. S. Petersen of New material. - One left valve (part and Brigham Young University, who collected the counterpart), KUMIP 153918. only known specimen of this unusual species. Diagnosis. -Hinge line usually straight. Holotype.-Right? valve, KUMIP 134799. Maximum valve height at about one-third Description. - One relatively small, sub- length from posterior margin. Anterodorsal polygonal valve; anterior margin missing. Dor- angle usually between 100° and 110°, postero- sal margin slightly convex and thickened, dorsal angle usually between 110° and 115°. perhaps by curling; small, vertical spine at Lateral ridge extending almost full length of about midlength. Posterodorsal angle about valve at about midheight. Multiple small spines 115 0 . Posterodorsal corner produced into present in some along dorsal margin. Number giant, slightly curved, oblique spine; length ap- and size of spines on posterior margin variable; proximately equal to maximum valve height. largest spine normally situated directly behind Two similar but successively shorter, irregu- end of lateral ridge; second largest spine usu- larly spaced spines arising from upper half of ally posteroventral, third largest spine usually posterior margin. Maximum valve height, ex- above level of lateral ridge; additional smaller clusive of spines, 10.3 mm. spines may be present, may extend onto ventral Discussion. -Although not well preserved, margin, and may grade to serrations. this extraordinary valve possibly warrants as- Discussion.-Simonetta and Delle Cave signment to a new genus. It is characterized by (1975, p. 8) tentatively considered T. burges- three unusually large spines on the posterior sensis Resser and T. canadensis Resser to be margin and a smaller spine on the dorsal mar- junior synonyms of T. retifera Walcott. We ac- gin. Placement of these spines is somewhat sim- cept the synonymy of these species and con- ilar to that in species of Tuzoia; however, the clude that the characters used by Resser (1929) posterior and ventral outlines of the valve differ to define his two species are mainly attributable Robison Ce Richards—Larger Bivalve Arthropods from the Middle Cambrian 15 to taphonomic compression. Specimens that elongate, subfusiform valves. Margins of valves were assigned to T. burgessensis (Resser, 1929, divided into three parts of unequal length; Pl. 2, fig. 1; Pl. 3, fig. 1) are longer than usual longest part broadly curved, intermediate and but show lengthwise wrinkles, which indicate a shortest parts nearly straight. Anterior and postmortem decrease in relative height. One of posterior ends acutely pointed. Surface having these (Resser, 1929, Pl. 2, fig. 1) also tapers microgranular texture. posteroventrally, seemingly because of post- Discussion. —Two partly superimposed mortem crumpling and gathering of cuticle specimens (Pl. 9, fig. 3) appear to represent toward the posterior end of the lateral ridge. laterally compressed carapaces of an undeter- This same specimen is the basis for line draw- mined arthropod. The moderate-brown cuticle ings (Brooks and Caster, 1956, text-fig. L5; of each specimen is folded along a nearly Rolfe, 1969, fig. 152.6) that convey a mislead- straight line, forming a margin of intermediate ing image of the normal valve outline in length. Space between the folded valves is filled Tuzoia, which expands anteroventrally. Speci- with very light-gray calcareous mudstone, mens that were assigned to T. canadensis have which thickens away from the folded margin. additional minor marginal spines that differ in The long, free margin is broadly curved. The number even in the two specimens illustrated short margin, oblique to the folded margin, is by Resser (1929, Pl. 2, figs. 2, 3). Hence, we at- inferred to be posterodorsal. This short margin tribute this difference in spinosity to intraspe- may have bounded one side of a deep axial em- cific variation. bayment of the carapace. A single left valve from Utah fits within the The possibility that these two specimens are range of morphologic variation shown by speci- folded and dorsoventrally compressed valves of mens from the type locality in British Colum- a single carapace has been considered. Such a bia. Compression has nearly obliterated the possibility seems unlikely, however, because of long lateral ridge on this valve but the feature is the unusual form. Also, a hinge margin is not still discernible when viewed under oblique obvious. light. Bases of a few small spines are evident The Wheeler specimens bear superficial re- along the dorsal margin. Spines on the poste- semblance to carapaces belonging to several rior margin are not well preserved; however, members of the phyllocarid orders Leptostraca the base of the large spine (on unillustrated and Archaeostraca. For example, the general counterpart) behind the lateral ridge and most form and relative lengths of the tripartite of the large posteroventral spine are evident. In margins are remarkably similar to those of the places the fine reticulate ridges on the valve extant leptostracan Nebalia bipes (Rolfe, 1969, surface are underlain by delicate deposits of fig. 120). There is also a lesser resemblance to pyrite. such fossil archaeostracans as Callizoe bohe- Occurrence.— The valve from Utah is pre- mica (Rolfe, 1969, fig. 142.2). Nevertheless, served in dark-gray, noncalcareous shale. It the Wheeler specimens differ from all of these was collected by Lloyd and Val Gunther from by haying much more acute anterior and poste- an unknown stratigraphic level in the Spence rior ends. At this time, the specimens cannot be Tongue of the Lead Bell Shale near the mouth assigned with confidence to any described of Donation Canyon on the west side of the genus. Wellsville Mountains (Fig. 1). Occurrence. Wheeler Formation (Ptych- The discovery of T. retifera in Utah extends agnostus atavus Zone) at locality 115, House the observed stratigraphic range of the species Range. Associated on the same bedding surface downward from the Ptychagnostus praecurrens are specimens of the agnostoid trilobite Pero- Interval-zone to the upper Peronopsis bonne- nopsis interstricta and numerous small shells of rensis Interval-zone (Fig. 2). It also extends the a Helcionella-like mollusk. The specimens were observed geographic range from British Co- collected by L. F. Hintze. lumbia to Utah. Phylum ARTHROPODA? Genus and species undetermined Class, order, family uncertain Plate 9, figure 3 Genus PAHVANTIA, new Description.— Probable carapace with Type species. — Pahvantia hastata, n. sp. 1 6 The University of Kansas Paleontological Contributions Paper 106

Etymology. —After the Pahvant Indians of spearhead, although the inferred anterior and western Utah. posterior ends are reversed. Description. --Carapace univalve or ob- Holotype. —Relatively small but seemingly scurely hinged bivalve, elongate, bilaterally complete or nearly complete carapace, KUMIP symmetrical, and shaped like a spearhead. 134878 (Pl. 8, fig. 2), from locality 115, House Maximum width slightly behind prominent, Range. rounded shoulders and about one-third dis- Description. — Because this is a monotypic tance from inferred anterior end. Laterally genus, description of the species is the same constricted, shanklike process nearly parallel as that of the genus. The holotype is 18 mm wide sided; width less than one-half that of main and 40 mm long. One fragmentary paratype is subtriangular part. Inferred posterior end 42 mm wide and was probably acutely pointed. about 85 mm long, the other was probably about 40 mm Discussion. —In form, three specimens rep- wide and about 90 mm long. resenting this new genus are unlike the exo- Discussion. — A feature that may prove to skeletal remains of any known arthropod. In be of taxonomic importance is the terminal details of color, texture, thickness, and mode of margin of the anterior process, which has preservation, however, they seem to be com- an angular cleft successively flanked by a pair of posed of cuticular material similar to that of small lobes, a pair of shallow rounded indenta- associated larger bivalve arthropods. Hence, tions, and a pair of small marginal spikes. the genus is questionably assigned to the Ar- Another such feature is a pair of weak exsagit- thropoda but further assignment to class, or- tal lines that commence at small pointed pro- der, and family is deferred. jections or slight flection points midway out on The relative size, general form, and lack of the shoulders and extend straight back across segmentation suggest that the specimens are the outer edges of the main ovate part. These cephalothoracic carapaces. If so, bilateral sym- lines may mark the inner edges of subjacent metry indicates a complete or nearly complete doublures or infolded carapace; however, these carapace rather than disarticulated, hinged possibilities seem unlikely because of the valves. Inferred anterior and posterior ends are straightness of the lines and the roundness designated with much uncertainty. of the lateral margins where folding would have Pahvantia is a monotypic genus that pres- occurred. ently is known only from the late Middle Cam- Occurrence. —Three specimens of brian (Ptychagnostus atavus Zone) of Utah. P. hastata (KUMIP 134187, 134878, 134879) were collected by R. L. PAHVANTIA HASTATA, n. sp. Harris from a quarry in the Wheeler Formation at locality 115, House Plate 7, figure 3; Plate 8, figures 1.2 Range (Fig. 1). They are preserved in dark- Etymology. —Latin hastatus, spear-shaped; gray calcareous mudstone of the Ptychagnostus in reference to the general similarity to a atavus Zone (Fig. 2). Robison Ce Richards-Larger Bivalve Arthropods from the Middle Cambrian 17

REFERENCES

of Paleontology: Ar- Bergstrom, Jan, 1979, Morphology of fossil arthropods as a Orlov, Yu. A. (ed.), 1960, Principles [in Russian]: 515 guide to phylogenetic relationships: p. 3-56, in A. P. thropods, trilobites and crustaceans Gupta (ed.), Arthropod phylogeny, Van Nostrand p., Gos. Nauchno-Techn. lzdvo (Moscow). Mid- Reinhold Co. (New York). Palmer, A. R., 1954, An appraisal of the Great Basin U.S. Briggs, D. E. G., 1976, The arthropod Branchiocaris n. dle Cambrian trilobites described before 1900: gen., Middle Cambrian, Burgess Shale, British Colum- Geol. Surv. Prof. Pap. 264-D, p. 55-86. other mid-Cam- bia: Geol. Surv. Can. Bull. 264, 29 p. Raymond, P. E., 1935, Leanchodia and Zool. Harv. , 1977. Bivalved arthropods from the Cambrian brian Arthropoda: Bull. Mus. Comp. Burgess Shale of British Columbia: Palaeontology, v. Univ., v. 76, no. 6, p. 205-230. Lower and Middle Cambrian 20, p. 595-621. Resser, C. E., 1929, New Natl. Mus., v.76, no. 9, p. 1-18. , 1978, The morphology, mode of life, and affinities Crustacea: Proc. U.S. Fourth contribution to nomenclature of of Canadas pis perfecta (Crustacea: Phyllocarida), Mid- , 1938, Misc. Collect., v. 97, no. dle Cambrian, Burgess Shale, British Columbia: Cambrian fossils: Smithson. Philos. Trans. R. Soc. London, Ser. B, v. 281, p. 10, p. 1-43. Lower Cambrian 439-487. , and Howell, B. F., 1938, Am. , 1979, Anomalocaris, the largest known Cambrian Olenellus Zone of the Appalachians: Geol. Soc. arthropod: Palaeontology, v.22, p.631-664. Bull., v.49, p. 195-248. Middle Cambrian trilo- Brooks, H. K., and Caster, K. E., 1956, Pseudoarctolepis Robison, R. A., 1971, Additional sharp!, n. gen., n. sp. (Phyllocarida), from the Wheeler bites from the Wheeler Shale of Utah: J. Paleontol., v. Shale (Middle Cambrian) of Utah: J. Paleontol., v. 30, 45, p. 796-804. trilobite biostratigraphy p. 9-14. , 1976, Middle Cambrian Brigham Young Univ. Geol. Stud., Conway Morris, Simon, 1979, The Burgess Shale (Middle of the Great Basin: Cambrian) fauna: Annu. Rev. Ecol. Syst., v. 10, p. v.23, p-93-109. trilobites 327-349. , 1982, Some Middle Cambrian agnostoid (in , and Robison, R. A., 1982, The enigmatic from western North America: J. Paleontol., v. 56 medusoid Peytoia and a comparison of some Cambrian press). biotas: J. Paleontol., v. 56 (in press). Rolfe, W. D. I., 1962, Two new arthropod carapaces from of Canada: Bre- , and Whittington, H. B., 1979, The animals of the the Burgess Shale (Middle Cambrian) Burgess Shale: Sci. Am., v.241, p. 122-133. viora, v. 160, p. 1-9. R. C. Moore Glaessner, M. F., 1979, Lower Cambrian Crustacea and , 1969, Phyllocarida: p. R296-R331, in Part R. annelid worms from Kangaroo Island, South Australia: (ed.), Treatise on Invertebrate Paleontology, America & Univer- Alcheringa, v. 3, p.21.31. Arthropoda 4: Geological Society of Giirich, Georg, 1929, Silesicaris von Leipe und die sity of Kansas (Boulder and Lawrence). fossils of Phyllokariden iiberhaupt: Mitt. Mineral.-geol. Staat- Shimer, H. W., and Shrock, R. R., 1944, Index North America: 837 p., J. Wiley & Sons (New York). sinst. Hamburg, v. 11, p-21-90. Cam- Henriksen, K. L., 1928, Critical notes upon some Cam- Simonetta, A. M., and Delle Cave, L., 1975, The brian arthropods described by Charles D. Walcott: brian nontrilobite arthropods from the Burgess Shale of mor- Vidensk. Medd. Dan. Naturhist. Foren., v.86, p. 1-20. British Columbia: A study of their comparative evolutionary significance: Pa- Hintze, L. F., and Robison, R. A., 1975, Middle Cam- phology, taxonomy and brian stratigraphy of the House, Wah Wah, and adja- laeontogr. Ital. v. 69 (n.s.39), p. 1-37. to the studies cent ranges in western Utah: Geol. Soc. Am. Bull., v. Walcott, C. D., 1886, Second contribution America: U.S. Geol. 86, p. 881-891. on the Cambrian faunas of North Krestovnikov, V. N., 1961, New phyllocarid crustaceans Surv. Bull. 30, 369 p. and paleontology II; from the Paleozoic of the Russian platform, Urals, , 1912a, Cambrian geology Timan, and Donetz basin [in Russian]: Akad. Nauk Middle Cambrian Branchiopoda, Malacostraca, Trilo- SSSR, Geol. Inst., Tr., v. 52, p. 1-67. bita, and Merostomata: Smithson. Misc. Collect., v. Opik, A. A., 1968, Ordian (Cambrian) Crustacea 57, no. 6, p. 145-228. Bradoriida of Australia: Aust. Bur. Miner. Resour. , 19126, Cambrian Brachiopoda:. U.S. Geol. Surv. Geol. Geophys. Bull. 103,45 p. Monogr. 51, 2 v. Oriel, S. S., and Armstrong, F. C., 1971, Uppermost Pre- Whittington, H. B., 1980, The significance of the fauna of Colum- cambrian and lowest Cambrian rocks in southeastern the Burgess Shale, Middle Cambrian, British Idaho: U.S. Geol. Surv. Prof. Pap. 394, 52 p. bia: Proc. Geol. Assoc. Can., v.91, p. 127-148. 18 The University of Kansas Paleontological Contributions Paper 106

EXPLANATION OF PLATES

Plate 1 indent surfaces of the carapace. la, Interior view of FIGURE most of right valve, which is overlain by small 1-3. Canadaspis cf. C. perfecta (Walcott), all from the anteroventral part of left valve; anterodorsal cuticle Spence Tongue of the Lead Bell Shale, Wellsville of right valve is broken and spine has been rotated Mountains. 1. Left valve, x 1.5, KUMIP obliquely backwards. 1 b, Exterior view of posterior 153893. 2. Right? valve, x 2.0, KUMIP part of right valve and interior view of anterior part 153894. 3a,b. Counterparts of right valve, of left valve, which seemingly has anterodorsal spine X 1.5 and 2.0, KUMIP 144401. preserved in normal position. 2a,b. Counter- parts of 9. Perspicaris7 dilatus, n. sp. Right valve, holotype, right valve with secondarily misoriented anterodorsal from the Wheeler Formation, House Range, x 1.3, spine; from locality 115, House Range; KUMIP 135128. Subcircular muscle scar (arrow) KUMIP 134940. 3. Mostly exposed right valve with secondarily shows near anterior margin. Irregular, linear mark- misoriented anterodorsal spine; from the SEI/4SEV4SW V4 ings were probably made by a burrowing animal sec. 21, T. 15 S., R. 10 W., Drum Mountains; KUMIP 153900. after burial of valve. Impressions of small, disartic- 4-6. Dioxycaris argenta (Walcott). ulated trilobite sclerites are also evident. 4. Holotype from the lower member of the Ophir Shale, 5,6. Branchiocaris sp., both from the Wheeler USNM locality 30a in the Wasatch Range, Formation, Drum Mountains. 5. Left valve?, X 1.0, USNM 15401. 5. Right? from X 1.3, KUMIP 153897. 6 Incomplete left? valve the Spence Tongue valve, x 2.0, KUMIP 153898. of the Lead Bell Shale, Wellsville Mountains, x 1.0, KUMIP 135145. 6. Left? valve from the Ute Plate 2 Formation, Wellsville Mountains, x 1.3, KUMIP 153911. FIGURE 1 -4. Perspicaris? ellipsopetta, n. sp.; all from the Marjum Formation, House Range. /. Right Plate 4 valve, holotype, x 1.0, KUMIP 153902. 2. Left FIGU RE valve, x 1.5, KUMIP 153903. 3. Nearly com- Proboscicaris agnosta Rolfe. la, b. Counter- plete carapace preserved in parallel aspect, anterior parts of a probable thin internal mold from the toward top, x 2.0, KUMIP 153904. Slight dorsal Wheeler Formation, House Range, both x 1.3, convexity of hinge line together with compression KUMIP 153901 after burial has resulted in slight overlap of dorsal 2-5. Pseudoarctolepi,s sharpi Brooks and Caster; all from valve margins in central part of hinge line. Max- the Wheeler Formation, House Range. 2. imum height of left valve has been reduced by Dorsoventrally compressed juvenile carapace with crumpling in area along ventral margin. 4. Left upper parts of each valve flattened and lower parts valve, x 1.3, KUMIP 153905. folded inward, causing the lower parts to be hidden 5.7. Perspicaris? ditatus, n. sp.; all from the Wheeler from view; lateral spines are relatively short; X 2.0, Formation, House Range; all x 1.5. 5. Right KUMIP 135139. 3. Enlarged interior view of valve, KUMIP 135131. 6. Right valve, KUMIP lateral spine base showing broad, triangular cross 135130. 7. Left valve, KUMIP 135129. section; same valve as PI. 5, fig. 3; x 5.0, KUMIP 153912. 4. Juvenile carapace with left valve par- Plate 3 tially flattened parallel to bedding surface and right FIGURE valve dorsoventrally crumpled; x 2.5, KUMIP 1-3. Dicerocarts opisthoeces, n. gen. and n. sp.; all from 135135. 5a,b. Counterparts of dorsoventrally the Wheeler Formation and all x 2.5. 1. compressed mature carapace with relatively long Holotype carapace preserved in lateral aspect with lateral spines; both x 1.0, KUMIP 153913. Separa- valves slightly offset and rotated; from locality 115, tion along different surfaces has exposed the House Range; KUMIP 134939. Disarticulated trilo- inward-folded, anterior, lower part of the right bite sclerites with coatings of cone-in-cone calcite valve (upper right, 5a; upper left, 56). A more ex- Robison & Richards Larger Bivalve Arthropods from the Middle Cambrian 19

tensive surface exposes much of the upper left valve ated trilobites are Elrathta kingii. 1. Carapace, and the posterior upper part of the right valve. The KUMIP 135134. 2. Carapace. KUMIP 135138. posterior end of the carapace has suffered dorsoventral crumpling. Plate 7 FIGURE Plate 5 1,2. Tuzoia guntheri, n. sp., both from the Marjum FIGURE Formation, House Range. I. Right valve, partly 1-6. Pseudoarctolepis sharp: Brooks and Caster, all from exfoliated; x 1.0, KUMIP 144390. 2. Holotype the Wheeler Formation, House Range. I. left valve; x 1.5, KUMIP 153917. Exterior of right valve with area below lateral spine 3. Pahtuntia hastata, n. sp. Incomplete carapace from strongly crumpled; dark spots on adjacent bedding the Wheeler Formation, House Range; x 1.0, surface may represent remains of the alga Mora nia,' KUMIP 134879. X 1.0, KUMIP 153914. 2. Interior of left valve, which is overlain by small part of anterior right Plate 8 valve; x 1.3, KUMIP 153915. 3. Interior of right valve showing lateral spine base at about 0.3 FIGURE length from anterior margin (compare Pl. 4, fig. 3); 1,2. Pahvantia hastata, n. sp., both from the Wheeler specimen is impressed against sclerites of the trilo- Formation, House Range. I. Incomplete carapace; inferred left anterior part still covered with bites Elrathia kingii and Peronopsis interstricta; x 1.0 . KUMIP 153912. 4. Exterior of right matrix; x 1.0, KUMIP 134187. 2. Holotype carapace, . valve with minor crumpling of anteroventral region; x 2.0 KUMIP 134878. 3. Tuzota retifera Walcott. Left valve with broken slightly elevated base of lateral spine about 0.3 posterior margin; from Spence Tongue of the Lead length from anterior margin; dark spots on adjacent Bell Shale, Wellsville Mountains; x 1.3, KUMIP bedding surface may represent Morattia, and the broad, lighter colored trace that extends from top 153918. 4,5. Tuzoia guntheri, n. sp., both from the Marjum to bottom of the figure may represent a burrow or Formation, House Range, and both x 1.0. Concen- feeding trail of an unknown animal; x 1.0, KUMIP tric wrinkles in the central part of each valve are the 153916. - 5. Dorsoventrally compressed carapace result of compression. 4. Interior view of left with parting surface exposing internal anterodorsal valve; anterior end is partly exfoliated; KUMIP and exterior posteroventral parts of folded valves; 5. Mold of left valve, associated with Etrathia king::; x 1.3, KUMIP 134797. KUMIP 144982. 135142. 6. Holotype carapace preserved in parallel aspect with lower parts of valves folded underneath; small pair of broken bulges near anterior Plate 9 margin may be sites where adductor muscles in- FIGURE serted; x 1.0, KUMIP 159141. I. Tuzoia? peterseni, n. sp. Holotype valve from the Wheeler Formation, House Range; x 3.0, KUMIP Plate 6 134799. FIGURE 2. Tuzota guntheri, n. sp. Carapace preserved in 1,2. Pseudoarctolepts sharp: Brooks and Caster, both parallel aspect; from the Marjum Formation, House from the Wheeler Formation, House Range, and Range; height of left valve has been reduced by both x 1.0. These two carapaces, each preserved in crumpling of lower part; x 0.8 . KUMIP 134798. parallel aspect, differ markedly in anterior outline; 3. Genus and species undetermined. Two probable however, the difference is attributed to variation in carapaces; small circular objects on surface are style of dorsoventral compaction rather than to sclerites of the agnostoid Peronopsis interstricla; original variation in carapace morphology. Associ- X 2.0, KUMIP 134188. The University of Kansas Paleontological Contributions Paper 106, Plate 1 Robison & Richards—Larger Bivalve Arthropods The University of Kansas Paleontological Contributions Robison & Richards—Larger Bivalve Arthropods Paper 106, Plate 2 The University of Kansas Paleontological Contributions Paper 106, Plate 3 Robison & Richards—Larger Bivalve Arthropods The University of Kansas Paleontological Contributions Robison & Richards—Larger Bivalve Arthropods Paper 106, Plate 4

5a 5b The University of Kansas Paleontological Contributions Paper 106, Plate 5 Robison & Richards Larger Bivalve Arthropods The University of Kansas Paleontological Contributions Robison & Richards— Larger Bivalve Arthropods Paper 106, Plate 6 The University of Kansas Paleontological Contributions Paper 106, Plate 7 Ro bison & Richards—Larger Bivalve Arthropods

3 The University of Kansas Paleontological Contributions Robison & Richards—Larger Bivalve Arthropods Paper 106, Plate 8

1 The University of Kansas Paleontological Contributions Paper 106, Plate 9 Robison & Richards—Larger Bivalve Arthropods