666 AMERICAN JOURNAL OF BOTANY [Vol. 40

FRENCH, R. C. 1953. The effect of growth regulators and stimulating substances and related chemicals. Amer, allied compounds on growth and respiration of maize Jour. Bot. 36: 368-378. tissues. Thesis. Purdue Univ. Lafayette, Indiana. NANCE, J. F., AND L. W. CUNNINGHAM. 19.50. Acetalde­ LARDY, H. A., AND H. WELLMAN. 1952. Oxidative phos­ hyde accumulation in excised wheat roots induced by phorylations: Role of inorganic phosphate and ac­ growth substances. Science 112: 170-172. ceptor systems in control of metabolic rates. Jour. BioI. Chern. 19.5: 21.5-224. RHODES, A., AND R. DE B. ASHWOIlTH. 19.52. Mode of action ---, ---, AND C. FELDOTT. 19.52. Mechanism of ac­ of growth regulators in . Nature 169: 76-77. tion of dinitrophenol and thyroxine. Abstracts, Spring YAMAKI, T. 1949. Einfluss der Wuchstoffe auf den At­ Meeting, Div. BioI. Chem., Amer. Chern. Soc. mungsvorgang der Koleoptile von Avena sativa. II. MITCHELL, J. E., R. H. BURRIS, AND A. J. RIKER. 1949. Mitteilung iiber Pflanzliche Wirkstoffe. Acta Phyto­ Inhibition of respiration in plant tissues by callus chim. 1.5: 1.51-167.

STATUS OF AN ASIATIC MEMBER OF THE REGARDED AS A 'LIVING FOSSIL'l Richard A. Scott ~

THE DESCRIPTION of a living representative of a sis, Chevalier also inexplicably referred to it as genus known previously only from the fossil record indochinensis in the same paper. Later in is a rare event of considerable interest to both neo­ the same year Kuang (1941) independently pro­ and paleobotanists. The extent of this interest may posed another new genus, Rhamphocarya, whose be measured in part by the large number of papers single species, R. integrifoliolata, is also based upon concerned with Metasequoia glyptostroboides since material of the plant under discussion. the discovery of this 'living fossil' was reported in Hjelmqvist (1948), overlooking the earlier de­ 1948. Recently Dr. Hsen·Hsu Hu (1952), who had scriptions by both Dodeand Chevalier, transferred a part in the discovery of the modern Metasequoia, the species from Rhamphocarya to Carya, creating has concluded that a member of the family the new combination C. integrifoliolata (Kuang) now found in China and Indo-China is a living Hjelmqvist. W. Y. Chun had previously referred species of Juglandicarya, a genus previously known the species to Carya as C. tsiangii, but this combina­ only from fossil occurring in the Eocene tion was never published (Manning and Hjelm. London Clay formation of England. The writer, qvist, 1951). who recently examined the material of Juglandi­ Leroy (1950) concluded that the species consti­ carya in the collections of the British Museum [Nat­ tutes a distinct genus and, with full knowledge of ural History), believes that this assignment is in the pertinent literature, proposed the new com­ error. The following discussion reviews the avail­ bination sinensis (Dode) Leroy as able information on the living and fossil species as its name. Leroy's reason for taking the species out a basis for the contention that the species are not of Carya. was based chiefly upon the vascular struc­ congeneric. ture of the (1951a, b). He reported that the Unlike Metasequoia glyptostroboides, this Recent vascular strands extend from the base to the apex species has been known to botanists for a number within the inner wall of the fruit rather than with­ of years. However, a lack of reference to the perti­ in the primary partition as is the case for both nent literature, in part unavoidable, by authors Carya. and luglans. dealing with this species has led to much confusion Manning and Hjelmqvist (1951), after examin­ with regard to its systematic position and nomen­ ing all available herbarium material but without clature. knowledge of Leroy's work, reaffirmed Hjelmqvist's Fruits of the species were described as Carya earlier opinion that this Asiatic species belongs in sinensis by Dode (1912), but his description was Carya; As Leroy had done earlier, they linked overlooked until 1950. Meanwhile the species was Dode's description of Carya sinensis to the material again described from other material by Chevalier at hand. After rejecting the possibility that the spe­ (1941), who placed it in a new genus, Annamo­ cies is closely related to Juglandicarya, they con­ carya. Although naming this plant A. indochinen- c~uded that its name should remain as Carya. sinen­ SIS. 1 Received for publication April 2.5, 19.53. The first intimation that this Recent species might The author wishes to thank Mr. W. N. Edwards of the British Museum (Natural History) for his courtesy in per­ be a 'living fossil' was made by Chevalier when mitting study of the material of Juglandicarya in the Mu­ he described it as Annamocarya indochinensis. He seum's collections. This paper was written while the au­ considered that its fruits showed features suggest­ thor was a Rackham postdoctoral fellow at the University ing both Juglans and Carya, and that the species of Michigan. 2 Present address: The Biological Laboratories, Harvard might represent a form ancestral to these two gen­ University. era. Merrill (1948), using the name Rhampho- November, 1953] SCOTT-ASIATIC MEMBER OF JUGLANDACEAE 667

carya, regarded the plant asa living representative TABLE 1. Summary of opinions regarding the possible af­ of the fossil genus Caryojuglans Kirchheimer. The finities of the species of [uglandicarya Reid and monotypic Caryojuglans, described from fruits Chandler. found in European brown-coal beds, is regarded by Kirchheimer (1938) as an intermediate between Nearest Modern Relative luglansand Coryo; Miidler (1939) believes that its Reid and Manning and Juglandicarya Chandler Kirchheimer Hjelmqvist single species should be assigned to luglans, and Species (1933) (1951) (1951) Leroy (1952) assigns the species to Carya. The J. lubbocki [uglans fruits of luglans and Carya have several features Engelhardtia , in common and, particularly in the fossil state, are J. cantia Juglans Engelhardtia, sometimes difficult to distinguish. Their character­ J. depressa Pterocarya, Pterocarya Juglans istics have been discussed by Miidler and by Kirch­ Juglans rupestris heimer (1951). Both Kirchheimer (1951) and J. crassa Specimens disintegrated in storage Manning and Hjelmqvist (1951), after examining fruits of the Recent 'Rhamphocarya', have rejected the possibility that it isa species of Caryo juglans. able features of the species of luglandicarya are Kirchheimer has also expressed the opinion that the for the most part general ones typical for the wal­ Asiatic species does not belong to luglandicarya. family and not necessarily limited to anyone Hu's (1952) transfer of the problematic species modern genus. Reid and Chandler's diagnosis of to the fossil genus luglandicarya is the most recent­ I. cantia is representative (1933, p. 142) : ly published opinion regarding its affinities. Of the "Endocarp globular, smooth, and without ex­ papers cited above, Hu referred only to Kuang's ternal nodulations, dehiscing into equal valves, one­ work. It should be noted, however, that because of loculed, one-seeded; walls thick, without cavities. the length of the period during which Hu's paper erect, orthotropous, conforming to the shape was in press, he could not have seen any of the of the locule, simple above, two-lobed below, each papers published during 1951. Hu considered the lobe being slightly emarginate at the base. Diam­ earliest previous name for the species to be Rham­ eter of endocarp about 12 mm." phocarya integrijoliolata and created the new com­ The similarities upon which Hu bases the sup­ bination luglandicarya integriloliolata (Kuang) Hu posed congeneric relationship between 'Rhampho­ to designate it. carya' and luglandicarya are stated in this quota­ The net result of this involved sequence is that tion (1952, p. 264) : since it was first described this one juglandaceous 'Rhamphocarya has a smooth, globular to ellip­ species has been associated with six generic names: soid, one-loculedand one-seeded endocarp with Carya, luglans, Annamocarya, Rhamphocarya; thick wall without cavities. Its seed is erect, ortho­ Caryojuglans, and luglandicarya. There are cur­ tropous, conforming to the shape of the locule, sim­ rently three independent proposals regarding its ple or emarginate above, deeply two-lobed below, designation in the literature: Annamocarya sinen­ the lobes again being shallowly two-lobed by a sec­ sis (Dode) Leroy, Carya sinensis Dode, and the ondary septum. Its contours are smooth. These one to be discussed here, luglandicarya integri­ characteristics are similar to those of luglandi­ joliolata (Kuang) Hu. carya; only the size of the endocarp is much The genus luglandicarya was founded by Reid larger." and Chandler (1933, p. 140) to contain "Fruits This description of the fruit of 'Rhamphocarya' which, although clearly referable to the Juglanda­ is obviously similar to the diagnosis of luglandi­ ceae, are of doubtful generic relationship both to carya cantia and is also like that of I. lubbocki. Of living genera and to one another." It includes four the species of luglandicarya, Hu found these two to published species, one of which, I. crassa (Bower­ agree most closely with the modern species. The bank) Reid and Chandler, was based on material correspondence which he points out is not, how­ which had disintegrated in storage before the lat­ ever, adequate evidence for congeneric relation­ ter authors transferred it to luglandicarya. The ship. Hu's characterization of 'Rhamphocarya' is original description is not adequate for compari­ a generalized one which is also applicable in most sons. Opinions regarding the possible affinities of respects to fruits of other genera in the walnut fam­ the other three species are shown in table 1. These ily, for example, luglans and Carya. The only fea­ diverse opinions bear out the conclusion of Reid ture in which the agreement might indicate close and Chandler that the species assigned to lug­ relationship to luglandicarya is the stated lack of landiearya do not constitute a single natural genus. cavities in the wall of the modern nut. Wall cavi­ It is often necessary for paleobotanists to erect ties, absent in I uglandicarya., are usually lacking in genera of this sort; unfortunately, botanists who Carya but do occur in at least three species; they work chiefly with modern plants do not always rec­ are always present in luglans although sometimes ognize the element of artificiality inherent in them. greatly reduced. This similarity between 'Rham­ As would be expected where relationships can be phocarya' and luglandicarya is not a reliable one, established clearly only to family level, the observ- however, for Manning and Hjelmqvist (1951) have 668 AMERICAN JOURNAL OF BOTANY [Vol. 40 described and illustrated cavities in the wall of the separate genus. Further investigation is needed be­ nut near the apex in the modern species. fore this question can be settled finally. Both the Except to note the much larger size of the fruit anatomy of the secondary xylem (Heimsch and of the Asiatic species, Hu did not consider the dif­ Wetmore, 1939) and the morphology of the pollen ferences between it and the species of Juglandi­ (Heimsch, 1944) are useful in delimiting the gen­ carya. Internal ridges, typical for Carya, are pres­ era in the Juglandaceae and for the most part sup­ ent on the nutshell of the modern form but are ab­ port the relationships suggested by Manning (1938) sent from Juglandicarya. As Manning and Hjelm­ on the basis of his study of the floral morphology qvist have pointed out, the seed of the modern of the family. Thus it appears that when suffi­ species is compressed in the plane parallel to the cient material becomes available to permit applica­ primary partition rather than at right angles to this tion of the varied approaches of modern plane as are the of J. cantia and J. lubbocki. to this enigmatic Asiatic species, it may become The primary embryo lobes of J. cantia are entire possible to establish its correct affinities. Mean­ at the base; J. lubbocki has a secondary partition while, it seems advisable to refer to the plant as which is as wide as the primary partition. The Carya sinensis, in the genus to which it was orig­ modern nut is prominently apiculate, but the nuts inally assigned and which Manning and Hjelmqvist of these two Juglandicarya species are rounded. (1951) have lately reaffirmed as correct. One thing The resemblances of the Asiatic species under seems certain, however; the species is not a 'living consideration to the two species of Juglandicarya fossil' in the sense of being a closely related sur­ are only general ones, and there are significant vivor of those members of the walnut familv that differences between them. Through the courtesy formed a part of the Paleotropical assemblage of Miss M. E. J. Chandler and Mr. W. N. Edwards, known to have flourished during the Eocene in the Keeper of Geology, British Museum (Natural His­ region of present-day England. tory) , the writer was permitted to consult the manu­ script of Miss Chandler's forthcoming publication SUMMARY in which additional species of Juglandicarya are described. None of these new species appears to Dr. Hsen-Hsu Hu has concluded that a member be any more closely related to 'Rhamphocarya' of the walnut family now found in China and Indo­ than are J. cantia and 1. lubbocki. No substantial China is a living species of Juglandicarya, a genus evidence for the relationship of 'Rhamphocarva' to based on fossil fruits from the Eocene London Clay the fossil genus has been presented by Hu, and his formation of England. The genus luglandicarya transfer -of the modern form to Juglandicarya is was founded to include fossil juglandaceous fruits unjustified. identifiable to family only, and its species are of In addition to the question of affinity there are doubtful relationship to one another. The features other objections to the use of the name Juglandi­ which Hu found to be common to the living species carya integrifoliolata (Kuang) Hu for the species and Juglandicarya are for the most part general under discussion. It has been established that this ones also shared by other genera in the family. He species was first described by Dode as Carya sinen­ ignored-important differences that exist between the sis; hence Hu's combination is illegitimate. Arti­ modern and the fossil fruits. In the opinion of the cle 68 of the new International Code of Botanical present writer, these differences are sufficient to Nomenclature, which came into effect after Hu preclude their generic identity. There are also no­ submitted his manuscript for publication, provides menclatorial objections to the use of the name pro­ that when a taxon of Recent plants, except algae, posed by Hu. This modern Asiatic species cannot and a taxon of the same rank of fossil or subfossil be considered to be a 'living fossil' in the sense of plants are united, the correct name of the former being a closely related survivor of any of the ex­ taxon must be accepted even if it is antedated by tinct species included in the genus Juglandicarya. the latter. Thus use of the name Juglandicarya, DEPARTMENT OF BOTANY, based on fossil material, is now illegitimate for UNIVERSITY OF MICHIGAN, modern plants. If the relationship pointed out by ANN ARBOR, MICHIGAN Hu were correct, it would be necessary to transfer the related species of Juglandicarya. to the genus to LITERATURE CITED which the modern Asiatic species belongs. The underlying wisdom of this Article is well illus­ CHEVALIER, A. 1941. Variabilite et hybridite chez les noy­ trated in this case, for the transfer of a living spe­ ers. Notes sur des [uglans peu cormus, sur I'Annamo. cies to Juglandicarya would place the species in a carya et un Carya d'Indochine. Rev. Inter. Bot. Appl. 21: 477-509. taxon whose members are ef uncertain generic DODE, L. A. 1912. Deux genres nouveaux pour la Chine. relationship to one another. Bull. Soc. Dendrol. France. 24: 58-61. If this modern species is not related to Juglandi­ HEIMSCH, c., JR. 1944. Aljaroa pollen and generic rela. carya, what are its affinities? Two opinions remain: tionships in the Juglandaceae, (Abstract) Amer. Jour. Manning and Hjelmqvist consider it to be a species Bot. 31: 3s. of Carya, while Leroy believes that it constitutes a ~-. AND R. H. WETMORE. 19,19. The significance of November, 1953] A VERS-BIOSYSTEMATIC STUDIES IN ASTER 669

anatomy in the taxonomy of the J uglandaceae, . --. 1951b. Morphologie vegetale.e-Contre la theorie Amer. Jour. Bot. 26: 651-660. generalisee des carpelles-sporophylles, II. Phylogenie H,TELMQVIST, H. 1948. Studies on the floral morphology structurale de la placentation dans le groupe Juglans­ and phylogeny of the Amentiferae. Bot. Notiser. Carya (J uglandaceae}, Compt. Rend. Acad. Sci. Paris Suppl. 2: 1-171. 232: 1007-1009. Hu, H. H. 1952. On a living species of Juglandicarya found ---. 1952. Debat sur Annamocarya. Documents sur les in south Yunnan. The Palaeobot. 1: 263-265. Carya de I'ere tertiare. Rev. Inter. Bot. Appl, 355-356: KIRCH HEIMER, F. 1938. Ein Beitrag zur Kenntnis der Alt­ 289-290. tertiarflora des Harzvorlandes. Planta 27: 615-644. MADLER, K. 1939. Die pliozane Flora von Frankfurt am --. 1951. Uber Antweileria und andere Gattungen der Main. Abhandl. Senckb. Naturf. Ges. Frankfurt 446: Juglandaceen. Planta 39: 527-541. 1·202. KUANG, K. Z. 1941. Genus novum Juglandacearum ex MANNING, W. E. 1938. The morphology of the of austro-orientali Yunnan. Icon. Flor. Sinicae 1: 1-3. the Juglandaceae, 1. The inflorescence. Amer. Jour. LEROY, J. F. 1950. Note sur les noyers (Carya et Annamo­ Bot. 25: 407-419. carya) sauvages d'Indochine. Rev. Inter. Bot. Appl. --, AND H. Hn:LMQvIST. 1951. Annamocarya, Rham­ 333-334: 425-428. phocarya, and Carya sinensis. Bot. Notiser 4: 319-330. --. 1951a. Morphologie vegetale.v-Contre la theorie MERRILL, E. D. 1948. A living Metasequoia in China. Sci­ generalisee des carpelles-sporophylles, 1. Une struc­ ence 107: \4.0. ture singulariere d'axe invagine et de placentation REID, ELEANOR M. and MAllJOllIE E. J- CHANDLER. 1933. caulinaire chez Annomocarya A. Chev, (Juglundaceae) . The London Clay flora. British Museum (Natural Compt, Rend. Acad. Sci. Paris. 232: 432-434. History). London.

BIOSYSTEMATIC STUDIES IN ASTER. 1. CROSSING RELATIONSHIPS IN THE HETEROPHYLLI 1 Charlotte J. Avel'S

l'HE COMPLEXITY of species pattern in the genus The species under consideration here are mor­ -ister has long been noted. Much of the taxonomic phologically similar in many respects. Varieties difficulty has been ascribed to the blurring of spe­ have been largely ignored in the present treatment. cies boundaries by hybridization. The only previous Many of the varieties now recognized are merely experimental hybridization studies of asters were the result of hybridization and local ecological vari­ conducted by Wetmore and Delisle (1939) with ation and their taxonomic recognition would bur­ Aster novae-angliae and A. ericoides (multiflorus). den the literature without elucidating the genetic Several floristic studies of asters have been made relationships among the species. recently (Shinners, 1941, 1945; Rosendahl and The heterophyllous asters are perennial herbs Cronquist, 1949) but the group under considera­ which are most frequently found in small popula­ tion here has never been subjected to cytogenetic tions in shaded woodlands. The species occur in analysis. eastern North America except for the boreal A. The present study concerns a group of nine close­ ciliolatus Lindl, which extends from coast to coast. ly-related aster species generally known as the While several of these taxa are allopatric, many HETEROPHYLLI. It is the purpose of this study are sympatric over a large part of their range. to examine the concept of hybridization as a prom­ ·CYTOLoGY.-Determinations of chromosome inent cause for the taxonomic difficulty of the group numbers from pollen mother cells were made using and to examine those factors which may be op­ the acetocarrnine squash technique after prelimi­ erative in directing the evolutionary pattern of nary fixation in chloroform-ethanol-acetic acid, these species. The present paper is designed pri­ 4:3:1. The basic chromosome number in this group marily to present a preliminary survey of the cross­ is 9, with diploid, tetraploid, and octaploid species ing relationships among seven of the heterophyllous represented. No multivalents were apparent in ex­ species. The evolutionary consequences of the spe­ aminations of wild material of the parent species. ciation pattern will be discussed elsewhere (Avers, Table 1 shows the chromosome numbers of the 19.'>3) . HETEROPHYLLI, which are here reported for the ] Received for publication April 30, 1953. first time. Part of a thesis submitted to the Graduate School of In­ Meiotic behavior in the hybrids was examined diana University in 1953 in partial fulfillment of the re­ quirements for the degree of Doctor of Philosophy. The and all F] hybrids evinced regularity of chromo­ writer wishes to express her appreciation to Dr. Charles some pairing although a small percentage of pol. B. Heiser for his advice and criticisms throughout the len mother cells contained two to four unpaired course of this investigation. A portion of this work was chromosomes, Hybrids between homoploid spe­ carried out with the aid of the Kathryn McHale Fellowship cies showed no evidence of multivalent associa­ granted by the American Association of University Women. Present address: Botany Department, Connecticut Col­ tions. lege, New London, Connecticut. Pollen fertility was determined by relative stain-