ry, Ca * HARVARD UNIVERSITY LIBRARY OF THE Museum of Comparative Zoology Wes. aaa. 7D] UeRARY | NOV 29 1968 HARVARD UNIVERSITY hey Phylogeny and Evolution of Crustacea Proceedings of a conference held at Cambridge, Massachusetts March 6-8, 1962 Supported by the National Science Foundation Edited by H. B.. Whittington and W. D. I. Rolfe SPECIAL PUBLICATION MUSEUM OF COMPARATIVE ZOOLOGY CAMBRIDGE, MASS. OCTOBER, 1963 HARVARS UNIVERSITY PREFACE A Conference on Phylogeny and Evolution of Crustacea was held at the Museum of Comparative Zoology in March, 1962, at the invitation of the Director, Dr. Ernst Mayr. It arose out of a suggestion made by Dr. Martin F. Glaessner, who was visiting the United States as a guest of the American Geological Institute, and was made possible by grant G-22123 from the National Science Foundation. Attendance at the conference was limited to the zoologists and paleontologists listed below, in the hope that free and adequate discussion would take place around the table. Two preliminary reports (Geotimes, American Geological Institute, 1962, 6 (8): 20; Crustaceana, 1962, 4 (2): 163-166) have been published. This volume contains the majority of the papers presented, together with references to those presented but published elsewhere. Discussion of these papers occupied more than half the available time, and was recorded. Each participant was furnished with a complete transcription, and invited to condense and emend his contributions where necessary. The discussions given here are compiled from these responses or are our condensations of contributors’ remarks. As printed they total less than half their original length, but we believe we have retained the major points and play of ideas. Where important new material was added subsequently by the speaker, it is indicated. Throughout the volume the sign precedes the name of a fossil taxon. Publication would not have been possible without a further grant-in-aid; GN-144, from the National Science Foundation. We tender our sincere thanks to the Founda- tion, and to the participants for their prompt response to our queries. H. B. WHITTINGTON W. D. I. RoLFre ii PARTICIPANTS DorotHy E. Buiss, The American Museum of Natural History, Central Park West at 79th Street, New York 24, New York. E. L. BousFietp, National Museum of Canada, Natural History Branch, Ottawa, Ontario, Canada. T. E. Bowman, U.S. National Museum, Washington 25, D.C. H. K. Brooxs, Department of Geology, University of Florida, Gainesville, Florida. Ertk Dani, Department of Anatomy, Zoological Institute, University of Lund, Lund, Sweden. M. F. GLAEsSNER, Department of Geology, The University of Adelaide, Adelaide, South Australia. R. U. Goopinc, Department of Biology, Boston University, Boston 15, Massa- chusetts. ISABELLA GorDON, British Museum (Natural History), Cromwell Road, London S.W.7, England. Dora P. HENry, Department of Oceanography, University of Washington, Seattle 5, Washington. R. R. HeEsster, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts. A. G. Humes, Department of Biology, Boston University, Boston 15, Massachusetts. Orto Kinng, Biologische Anstalt Helgoland, Zentrale Hamburg-Altona, W. Germany. J. H. Locuueap, Department of Zoology, The University of Vermont, Burlington, Vermont. Sipni—e M. Manton, British Museum (Natural History), Cromwell Road, London S.W.7, England. Ernst Mayr, Museum of Comparative Zoology, Harvard University, Cambridge 38, Massachusetts. R. C. Moore, Department of Geology, University of Kansas, Lawrence, Kansas. W. A. Newman, Department of Zoology, University of California, Berkeley 4, Cali- fornia. H. B. Roserts, U.S. National Museum, Washington 25, D.C. W. D. I. Rotre, Museum of Comparative Zoology, Harvard University (now at Hunterian Museum, University of Glasgow, Glasgow, Scotland). H. L. Sanvers, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts. Rotr Srewrnc, Zoologisches Institut und Museum der Universitat, Kiel, Germany. Paut Tascu, Department of Geology, University of Wichita, Wichita 14, Kansas. T. H. Waterman, Department of Zoology, Yale University, New Haven, Connec- ticut. J. H. Wetsu, Biological Laboratories, Harvard University, Cambridge 38, Massa- chusetts. H. B. Wuirtincton, Museum of Comparative Zoology, Harvard University, Cam- bridge 38, Massachusetts. PARTICIPANTS AND OBSERVERS Vv OBSERVERS IsaABEL CANET, Museum of Comparative Zoology, Harvard University, Cambridge 38, Massachusetts. Mary R. Dawson, representing the National Science Foundation, Washington 25, DC. ELISABETH DEICHMANN, Museum of Comparative Zoology, Harvard University, Cambridge 38, Massachusetts. JouHN S. GartH, University of Southern California, University Park, Los Angeles 7, California. RayMonp B. MANNING, The Marine Laboratory, University of Miami, Miami 49, Florida. E. L. Mixts, Department of Zoology, Yale University, New Haven, Connecticut. A. R. Ormiston, Museum of Comparative Zoology, Harvard University, Cambridge 38, Massachusetts. Harpinc B. Owre, The Marine Laboratory, University of Miami, Miami 49, Florida. A. R. PALMER, U.S. National Museum, Washington 25, D.C. A. J. PROVENZANO, JR., The Marine Laboratory, University of Miami, Miami 49, Florida. F. C. SHaw, Museum of Comparative Zoology, Harvard University, Cambridge 38, Massachusetts. INTRODUCTION It gives me great pleasure to welcome you on behalf of the Museum of Com- parative Zoology, an institution which has a proud tradition in the field of inver- tebrate zoology. It is our endeavor to continue this tradition, indeed to display even greater activity in this field than was possible in recent years. The subject matter of this conference—the classification, phylogeny, and evolution of the crustaceans—has made steady advances. This is due, on one hand, to an accumulation of new facts, owing either to a more refined morphological analysis, or to the discovery of new fossils, or even to the discovery of extraordinary new living types such as Hutchinsoniella. It must not be overlooked, on the other hand, that some of the advances are due to the emergence of new concepts. All these factors together have caused a revival of research in phylogeny so that phylogenetic studies are once more beginning to be in the center of interest of the evolutionist. In part this has been due to a shift in emphasis. The interest of the earlier evolutionists was almost exclusively in the determination of homologies and the reconstruction of common ancestors. The emphasis throughout was on those characters which various kinds of organisms had in common. The new emphasis that has been added in recent decades is based on the questions: Why do descendants from common ancestors become different? What are the causal factors that are responsible for evolutionary divergence, and what selection pressures and adaptive shifts are involved? Such an approach quite clearly depends on a sound classification if it is not to degenerate into sterile speculation. Particularly valuable in the development of phylogenetic thinking has been the elimination of a number of formerly dominant misconceptions. I shall single out a few of these unfounded generalizations: (1) That evolutionary trends always proceed from simple to complex, and that the simple structures or types are always ancestral and the complex ones derived. A consequence of this assumption is the postulate that a morphological series is by necessity a phylogenetic series. You are all familiar with numerous cases in which these assumptions are not true. (2) That an evolutionary trend in one line must be paralleled by a similar trend in all related lines. Actually, it happens not infrequently that of two related phyletic lines one increases in size or segmentation, while the other one shows a corresponding decrease. (3) That embryos or larvae pass through a series of stages which reflect or re- capitulate phylogeny. Here again we now know that the selection pressure is some- times stronger on the larval stages than on the phenotypes of the adults. The larvae in such cases are not only more different from each other than are the adults, but also may have more newly acquired structures. To interpret these as ancestral would be misleading. viii INTRODUCTION (4) That new types develop either by saltation or by a harmonious unfolding of an archetype. Actually, we know that the evolutionary rate of different organs and functional complexes can be exceedingly different, and that one key structure is usually far in advance of all others during the morphological reorganization accom- panying any major adaptive shift. Mosaic evolution is far more frequent than arche- typal evolution. (5) That there are two sets of characters, phylogenetic characters and adaptive characters, and that by eliminating the unreliable adaptive characters, one is left with a residue of useful phylogenetic characters. It is true that some characters are more plastic than others, but no criteria are known by which such an a priori sorting of two such sets of characters would be possible. The elimination of these misconceptions makes it possible to arrive at sounder generalizations. I shall waste no more of your time with such introductory remarks, but turn the meeting back to your chairman. In concluding, I want to express my warmest wishes for a successful conference. Ernst Mayr, Director Museum of Comparative Zoology CONTENTS Page JORREENOD. = Gd '6'G:5.6S Gr. OS AI Gah bo AEN sep ue eine ne AR ili Parti clpantsmand mODSCVerSimecver racic cia euch cir ee eve cus here emer le eee ei ede teehee alee ade Aas eu seirelctledensts xeususyal sys iv Jiniroclnciiton— ang MEME Go ouoecdogouoodondouau Dob b bo GDo Dado soe OBO du Capon o Oo eoOREG vii I. Main evolutionary lines among Recent Crustacea (Figs. 1, 2)—Erik Dahl .......... 1 ee iscussionetollowingsWablism papenrsiyicce ceric rie cheicveraticn chee) lense ebeecstnebomiats, = 17 III. Adaptation, a primary mechanism of evolution (Figs. 3-9)—Otto Kinne ............ 27 IV. On the relationship of Dromiacea, Tymolinae and Raninidae to the Brachyura (Figs. IO Ika ks Gorka scosscckoceuuecanseducs podo es sudoo paces ooo abd GocmoeoS 51 V. The pericardial sacs of terrestrial Brachyura (Figs. 15-25)—Dorothy E. Bliss ....... 59 VI. Discussion following papers by Kinne, Gordon, and Bliss ..............-..--.2+00-- 79 VII. Studies in malacostracan morphology: results and problems (Figs. 26-42)—Rolf Sie- NRIADOV=E L—M Leptostraca 4 i L & Me X Mi Stomatopoda 4 7+1 LX Me x Mi Syncarida 2 7 L & Me x Mi Euphausiacea 2+2 7 L x Me X Mi Decapoda 4 7+1 L & Me X Mi Mysidacea 2+2 7 LX Me x Mi Amphipoda 2 5 L & Me X Mi Tsopoda 2+2 8—17 LX Me X Mi L, lamina ganglionaris; M, medulla; Me, medulla externa; Mi, medulla interna; x, chiasma. DAHL: RECENT CRUSTACEA 11 are aberrant in this respect with 5 and up to 17 retinular cells, respectively. The dif- ferences in the optic ganglia, however, are more far-reaching. As is well known, the Anostraca and Phyllopoda have only a lamina and one medulla without any chiasma between them. The Malacostraca, on the other hand, have a lamina and two medullae with chiasmata. It seems to have been generally and tacitly accepted that the malacostracan condition has been de- rived from the anostracan and phyllopodan one. However, there exists no evidence whatever, either paleontological or morpho- logical, that any such development has taken place. Obviously the malacostracan eye must have developed from simpler types of compound eyes, but there is noth- ing to tell us whether the ancestral forms had eyes of the anostracan type, or de- veloped along different lines. Indications of such a possibility are in fact to be found in the eyes of the Maxil- lopoda. Unfortunately, only a few of the Recent Maxillopoda have compound eyes available for comparison and a discussion of these problems has to be based only on the Branchiura and on the juvenile stages of the Cirripedia Thoracica. In these eyes we find 4 and 3 Semper’s cells, respectively, while the Branchiura have 4 plus 1 retin- ular cells, and the Cirripedia 6 retinular cells. Again, however, the optic ganglia are of greater interest. In both types the lamina ganglionaris is very poorly devel- oped, and it is indeed doubtful whether a typical lamina occurs at all. In the Bran- chiura there are in fact a number of cell nuclei in the interior of the eye, but they are not arranged as a typical lamina and it has not been possible to trace the de- tailed fibre arrangement. Detailed knowl- edge of the single medulla of the cirripede larva is not available; it has turned out to be a difficult object for this kind of study. Unpublished investigations by Mr. N. Madsen (verbal communications) have shown that the structure of the medulla in Argulus is highly complicated and shows an arrangement of fibres which might in- dicate a functional splitting into two parts. No definite proof of the existence of any chiasma has, however, come forth, and the general structure of the medulla is rather unlike that found in Malacostraca. The finer structure of the compound eye found in some ostracods is comparatively poorly known and this applies especially to the most interesting part in this con- nection, the optic ganglia. As is seen from this brief summary the evidence now available shows that we have two well defined and rather different types of compound eyes, one in the Branchiopoda and one in the Malacostraca. Further, there are indications to show that in the Maxillopoda we meet a third type of eye differing from both the others. Finally, the nauplius eye of anostracans and phyllopods is similar both with respect to general structure and to the highly complicated mode of development. They seem to differ profoundly from the nauplius eyes of most other crustacean groups, at least from those of the Malacostraca and the Maxillopoda. Other Organ Systems My personal knowledge of other organ systems is too superficial to permit a criti- cal evaluation. Undoubtedly, however, use- ful information could be obtained in various cases, as shown e.g. by Siewing, 1956, and Mayrat, 1959, with respect to the circu- latory system of Malacostraca. With re- spect to the vascular system we have also another interesting difference between the Anostraca and the Phyllopoda, on the one 12 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 hand, and other Crustacea, on the other hand, in that heart function in the two former groups is myogenic while among other Crustacea it is nervously controlled (Krijgsman, 1952; Lagerspetz, 1962). There are also indications that valuable in- formation would be gained from a com- parative study of the finer structure of the excretory organs, of vitellogenesis, etc. IV. CONCLUSIONS The Main Crustacean Groups It is interesting to note that the brief summaries of external and internal morpho- logical evidence of crustacean relationships which were made above, give, on the whole, an impression of mutual corroboration. The view that the main groups previously dis- tinguished are natural units is strengthened by this new evidence. One is repeatedly struck by the great degree of similarity in anostracan and phyllopod organisation in various organ systems. I find it very difficult to accept the conclusions drawn by Preuss (1951), according to which the two groups have very little in common. The general struc- ture of the mouth parts, the close similarity between the metanauplius larvae which are different from those of all other Crustacea (cf. Sanders, present volume), the struc- ture of the alimentary canal, the compound and nauplius eyes, and the mode of func- tion of the heart, all seem to point to a fairly close relationship. In my tentative systems of 1956 I re- ferred the Anostraca and Phyllopoda to- gether with the Cephalocarida to a new group of higher rank, the Gnathostraca. This seemed justified from the point of view of the evidence then available. The later results obtained by Sanders and par- ticularly those reported in the present volume have, however, convinced me that the Cephalocarida are more suitably re- garded as a separate group retaining many features of ancestral crustacean forms, but, on the other hand, with various highly ad- vanced adaptations, especially in head morphology, due to the peculiar mode of life. If the Cephalocarida are removed, how- ever, the justification for the concept Gnathostraca disappears and we can again group the Anostraca and the Phyllopoda together under the time-honored name Branchiopoda. The new evidence now available seems, on the other hand, to strengthen the valid- ity of the concept Maxillopoda. To the evidence recorded by Dahl (1956b), Birsh- tein (1960), and Siewing (1960) can be , added the general similarity with respect to the alimentary canal and possibly also the X-organ of some of the forms con- cerned. A grouping together of the Mysta- cocarida, the Copepoda, the Branchiura, and the Cirripedia under the heading Maxillopoda thus appears justified. The Ostracoda remain difficult. Some indica- tions of possible affinities to the Maxil- lopoda exist but they are rather uncertain and the possible relations to the Cephalo- carida can in any case not be very close. The Ostracoda are in my opinion better retained as a separate group of rather doubtful affinities. The Malacostraca also constitute a well defined group although with some indica- tion of affinities with the Maxillopoda and possibly also the Cephalocarida. As an outcome of these discussions, the Class Crustacea could be subdivided into five main groups, as shown on page 13. In my tentative system of 1956, the main subunits below the class level were called subclasses, and the next category orders. As pointed out to me by Dr. I. DAHL: RECENT CRUSTACEA 13 Class Cohorts Subclasses Crustacea Name Definition Anostraca Branchiopoda Calman, 1909 Phyllopoda Ostracoda Calman, 1909 — Cephalocarida Sanders, 1955 _- Reece Maxillopoda DH, Toye ove Branchiura {Cirripedta Malacostraca Calman, 1909 — Gordon, the arrangement then suggested caused difficulties, especially with respect to the Malacostraca. Further, it is clear that if the group Maxillopoda is given subclass rank, the rank of the Cirripedia will have to be lower, and that will again upset the well arranged system within that group. I am indebted to Professor E. Mayr for the suggestion that the highest units should be called cohorts and placed at the level between class and subclass. This sug- gestion has been followed here. In this way the Ostracoda and Malacos- traca, not to mention the Cephalocarida, have no subunits at the subclass level, but this appears to be a lesser evil than the up- setting of longstanding systems. It has to be left to the specialists on the respective groups finally to decide whether any of the lower units now recognized should be ele- vated to subclass rank. Considering the high degree of mutual independence of the groups now recognized as subclasses, doubts arise, however, whether this would not up- set the scale of values expressed in the system. Evolutionary Lines In the case of some of the crustacean groups, especially the Malacostraca, but also the Branchiopoda and the Cirripedia Thoracica, so much morphological and/or paleontological evidence is available that the systematics within the groups can be said to approach at least partly a natural system built on the main evolutionary lines. Numerous attempts have also been made to construct phylogenetic trees demonstra- ting the evolution of the Crustacea as a whole. Unfortunately, however, the most important features of these phylogenetic trees have to be conjectural, for at present we possess no actual evidence demonstra- ting any case of a group at subclass or higher level being derived from another group. We have reasons to believe that e.g. the Mystacocarida and the Copepoda have been derived from common ancestors which already showed many maxillopodan features, but we have no evidence what- ever to show how and at which level of organisation the lines separated. Similarly, it is reasonable to assume that the Mala- costraca and the Maxillopoda are closer to each other phylogenetically than either is to the Branchiopoda. Also, evidence is forthcoming which makes it probable that the Cephalocarida are at least in many re- spects fairly close to ancestral crustacean 14 types. But any attempt to derive any one of these groups from the other must be based on guesswork. A reasonable degree of conjecture must inevitably play a certain part in practi- cally all phylogenetic considerations, and especially on such dealing with the groups of higher rank, for evidence is rarely, if ever, complete. Nevertheless, it seems im- portant to make a clear distinction between MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 what is fact and what is conjecture in phylogeny. This seems the only way to lessen the influence of preconceived ideas, an influence felt only too often in discus- sions of this kind. Figures 1 and 2 have been constructed in order to demonstrate the present writer’s view on the main evolutionary lines among the Crustacea. The horizontal line cutting across Figure 1 represents the present limit Malacostraca Maxillopoda Cephalocarida Branchiopoda Ostracoda Copepoda Branchiura Anostraca Mystacocarida Cirripedia Phyllopoda {es \ \ | y \ / / aN \ \ | | / / / ‘\ \ \ \ \ | Tee / SS Nae \\inert|\ \ Uo / \ Nira aera | He oo / \ hey, y, xX lay / \ SY / / / \ \ es, / \ \ | Ll 2 »~ \ | iy Re \ ! ly \ S ly \ \ 1] \ \ an XN Fic. 1. Main evolutionary lines within Crustacea. The horizontal line is meant to indicate the level of present actual knowledge; the lines below it indicate the present writer’s views on the somewhat earlier evolutionary stages. Malacostraca Cephal Maxillopoda————__, | | boy sf | / I/ Ostracoda ocarida Branchiopoda ? / / Fic. 2. Relationships between cohorts of Crustacea, as seen in the horizontal plane of the Recent. DAHL: RECENT CRUSTACEA 15 of actual knowledge, the area below is open to speculation, and the dotted lines there indicate my own opinion of the main direction in which some evolutionary lines may have gone. Admittedly this diagram is a very poor substitute for a phylogenetic tree. It may be regarded as a retrograde step, but also as a step towards increased realism. ACKNOWLEDGEMENTS I am greatly indebted to Professor E. Mayr, Professor H. B. Whittington, and Dr. W. D. I. Rolfe for all their kindness and help before, during, and after the Harvard conference. I also wish to thank my collaborators, Mr. R. Elofsson, Mr. N. Madsen, and Mr. T. Kauri for permission to use their unpublished material. My participation in the conference was made possible by the National Science Foun- dation to which institution I wish to tender my respectful thanks. REFERENCES BEKLEMISHEV, V. N. 1952. Osnovy sraviteljnoi anatomii bespozvono¢énych. Moscow. BirsHTeEIN, J. A., et al. 1960. Nadklass Crustaceo- morpha. Jn Orlov, Yu. A. ed., Osnovy paleon- tologii, trilobitoobraznye i rakoobraznye, pp. 201-454. Moskva. CALMAN, W. T. 1909. Crustacea. Im Sir Ray Lan- kester: A Treatise on Zoology 7(3): 346 pp. London. CarLisLeE, D. B. anp W. J. Pitman. 1961. Dia- pause, neurosecretion and hormones in Cope- poda. Nature 190:825. Daut, E. 1953. Frontal organs in free-living copepods. K. Fysiogr. Sallsk. i Lund Forh. 23: 32-38, 4 figs. 1956a. On the differentiation of the topography of the crustacean head. Acta Zoo- logica 37:123-192. . 1956b. Some crustacean relationships. In Wingstrand, Karl Georg (ed.), Bertil Han- strom: Zoological papers in honour of his sixty-fifth birthday, November 20th, 1956, Lund Zool. Instit., pp. 138-147. . 1957. Embryology of X Organs in Crangon allmanni. Nature, 179:482. . 1958. The ontogeny and comparative anatomy of some protocerebral sense organs in notostracan phyllopods. Quart. J. Micr. Sci. 100 :445-462. HAnstrOM, B. 1928. Vergleichende Anatomie des Nervensystems der wirbellosen Tiere. Berlin, 628 p., 650 figs. Kauri, T. 1962. On the frontal filaments and nauplius eye in Balanus. Crustaceana 4:131-142. KriycsMAN, B. J. 1952. Contractile and pace- maker mechanisms of the heart of arthropods. Biol. Rev. 27:320-346. LAGERSPETZ, K. 1962. Heart mechanism in a con- chostracan, Limnadia lenticularis (L.). Nature 194:992. Lanc, K. 1948. Monographie der Harpacticiden. 2 vols., 1682 pp., Lund. Linper, F. 1941. Contributions to the morphology and the taxonomy of the Branchiopoda Anos- traca. Zool. Bidr. fr. Uppsala 20:103-302. . 1945. Affinities within the Branchio- poda, with notes on some dubious fossils. Ark. f. Zool. 37:1-28. Manton, S. M. 1953. Locomotory habits and the evolution of the larger arthropodan groups. Symposia, Soc. Exp. Biol., VII, Evolution, pp. 339-376. Mayrat, A. 1959. Anatomie compareé et evolution du systeme artériel des Malacostracés. Proc. XVth Int. Congr. Zool. London, pp. 340-343. Preuss, G. 1951. Die Verwandschaft der Anos- traca und Phyllopoda. Zool. Anz. 147:49-64. ReMane, A. 1956. Die Grundlagen des natiirlichen Systems, der vergleichende Anatomie und der Phylogenetik. Leipzig. 364 pp. Sanpers, H. L. 1955. The Cephalocarida, a new subclass of Crustacea from Long Island Sound. Proc. Nat. Acad. Sci. 41:61-66. . 1957. The Cephalocarida and crustacean phylogeny. Syst. Zool. 6:112-129. Srewrnc, R. 1956. Untersuchungen zur Morphol- ogie der Malacostraca (Crustacea). Zool. Jahrb. (Anat.) 75, 39-176. . 1960. Neuere Ergebnisse der Verwand- schaftsforschung bei den Crustaceen. Wiss. Zeitschr. Univ. Rostock Math.-Nat. Reihe, 9:343-358. Tiecs, O. W. anp S. M. Manton. 1958. The evo- lution of the Arthropoda. Biol. Rev. 33:255- 337, 18 figs. vin (eS PHYLOGENY AND EVOLUTION OF CRUSTACEA Museum oF CoMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 II Discussion Following Dahl’s Paper HESSLER: My studies of crustacean abdominal musculature agree with Dahl’s ideas on the primitiveness of the caridoid facies. Syncarids, mysids, euphausiids and decapods (Daniel) all possess a compli- cated, basically ‘‘caridoid” abdominal mus- culature. The non-mysid peracarids, on the other hand, have a simple abdominal mus- culature in which a few straight fibers run from segment to segment. Assuming the mysids are primitive, this is a secondary reduction. In the abdomen of stomatopods the dor- sal longitudinal muscles are of the caridoid type. The ventral muscles are a pair of relatively small, spiral bundles in which the fibers make a complete revolution every three segments, attaching in every segment: essentially the form of the ventral caridoid musculature. This suggests that the sto- matopods reflect the precursor condition to the caridoid musculature. Siewing has sug- gested that the stomatopods branched off from the main malacostracan line prior to the appearance of the caridoid facies. Dahl’s findings on the position of the go- nads coincide with the conclusions derived from musculature: in higher peracarids the abdominal muscles are secondarily re- duced; in the stomatopods their simple de- velopment is primary. The stomatopodan condition is probably derived from that of a leptostracan-like precursor where muscles go straight from 17 segment to segment without a spiral. Such is the condition found in cephalocarids and branchiopods. Thus it seems possible to derive the caridoid musculature from an early entomostracan condition. GLAESSNER: We heard that the prim- itive forms are filter-feeders with the mouth curved backwards. How does that fit in, if we look back to a possible origin of the crustaceans? Does not this condition have to be secondary in the ultimate origin of the Crustacea as such? DAHL: The ancestors of Crustacea may have had a terminal mouth, but at the crustacean level we must suppose that appendages have been drawn into the feeding mechanism. I can’t see how they could have fed after they acquired an exo- skeleton, without the use of appendages. We have to assume, on the crustacean level, that filter-feeding, or at least feeding by means of food transported along the ventral side is primitive. MANTON: We have little evidence of the derivations of the various types of crustacean filter-feeding. If glandular se- cretions are used in food collection, the labrum is turned backwards, as in many branchiopods. If the collection is more mechanical, by brush setae, etc., as in mysids and copepods, there is no need for the labrum to be turned so far backwards. Is the presence of spiral muscles asso- ciated with the capability of an “escape” 18 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, or other sudden movement, such as the flapping of the abdomen under the thorax? Such muscles also occur in Petrobius, an expert jumper. HESSLER: MacDonald (1927, J. Mar. Biol. Assoc. Plymouth, 14:753-794) says that Meganyctiphanes does not flex its ab- domen as an escape reaction. MANNING: I have observed Recent stomatopods, and they do not flap the abdomen. Their swimming mechanism uses the pleopods or they move along with their walking legs. The loose articulation of the abdomen enables them to turn around in their burrow, but they don’t use it as an escape reaction in the same sense that the shrimps do. HESSLER: When you disturb the water in front of stomatopods, they turn around almost instantaneously. The motion does involve flexure of the abdomen. MANNING: But they don’t move backward in the same sense that a lobster would, with a rapid tail flap first. They roll themselves and turn and _ retreat perhaps. HESSLER: Perhaps this is an inter- mediate stage. LOCHHEAD: Getting back to the role of filter-feeding in the origin of crustaceans, it is worth noting that the notostracan, Triops, can feed on small particles, despite a complete lack of filter setae. Even in the absence of mud, Triops can capture small numbers of 3 ut flagellates and can live on round algae 15-80u in diameter. Move- ments of the limbs produce a current which flows forward under the labrum, where the food organisms are secured. Triops does not always feed in this way, but the fact that it can do so suggests how filter-feeding may have originated. TASCH: Lochhead’s observations on feeding in the notostracan Triops can be 1963 supplemented by fossil evidence. Trusheim found indications that Triassic Triops (identical with living forms) indulged in cannibalism. +Lepidocaris from the De- vonian Rhynie Chert had its anterior ap- pendages modified for a special type of feeding—possibly detrital feeding. We should envisage a variety of possible modes of feeding in some branchiopods—surely for some known from the geologic past. DAHL: In the mysids the mouth is not directed backwards to the same extent as in the branchiopods; still there is a marked difference from those malacostracans which feed on large pieces of food directly below or in front of the mouth parts. And as Manton has shown, there is a clear cor- relation between head topography and mode of feeding in the Anaspidacea. Para- naspides is much more of a filter-feeder than Anaspides and has a larger labrum and the atvium oris directed more back- wards. On the other hand, the semiterres- trial Koonunga is almost prognathous. SANDERS: Concerning the primitive filter-feeding habit in the Crustacea, the evidence from the Cephalocarida indicates that filter-feeding may have been preceded by a detritus-feeding habit, and the type of detritus-feeding we are referring to is probably secondarily derived from a filter- feeding habit. Cephalocarids may have the primary detritus-feeding habit since all the components needed for branchiopod filter- feeding are present. The limbs are similar. There is a well developed endopodite which is not present in the more generalized branchiopods. There are endites; the proxi- mal one is completely unmodified which, as Cannon postulated, must have been the ancestral condition. There are anterior and posterior setae on the protopod and endo- pod which catch relatively large masses of detritus that are put in suspension by the DISCUSSION 19 naupliar-like or sweep-net movement of the second antenna and the endopods of the trunk limbs. The detritus is first caught by the posterior setae during the anterior or suctional phase and then pushed back into the median chamber by the interdigi- tating anterior setae. At this point the detritus is drawn dorsally and moved by the spines of the proximal endites to the head region. MAYR: What is believed about where the crustaceans join any other branch of the arthropods? WHITTINGTON: I prefer to keep the trilobites entirely separate from Crustacea. The lines of descent of crustacean groups may go back and join within some limit that we know in the fossil record, but be- fore that, the trilobites had separated. Feeding with the mouth facing back is an old and general habit. MOORE: I would like to ask about trilobitomorphs other than the trilobites. ROLFE: As Tiegs and Manton (1958, Biol. Rev., 33:292) pointed out, we don’t know how many of them have genu- inely trilobitan limbs. The “pseudocrusta- cean” ¥Canadaspis [= tHymenocaris| from the Burgess Shale was asserted by Raymond (1920, Mem. Conn. Acad. Arts Sci., 7:113) to have trilobitan limbs. Of 202 specimens in the Museum of Comparative Zoology, only a few have thoracopods which show anything of their segmentation. Up to eight segments can be counted in the limb, the eighth segment bears four terminal claws, and a large proximal flap is present which is presumably pre-epipodial. This basal lamella is not filamentous as it is in the trilobitan limb. Simply from the number of segments in the limb and the presence of a basal (branchial?) structure it could be a normal crustacean limb minus exopod. But until the United States National Museum specimens have been critically re- studied, it seems worthless to speculate on some of these. Tiegs and Manton’s idea of the “‘Pseudocrustacea” being mosaic forms is most intriguing, but unfortunately my preliminary study of the limbs of +Cana- daspis does not seem to verify it for this genus at least. (A. Simonetta has recently published a restudy of +Marrella, 1962, Monitore Zoologico Italiano, 69(3-4) :172- 185). MANTON: Before we build theories as to whether an animal was a trilobite or a crustacean relative we should, I agree, have good evidence of the nature of the outer ramus of the limb. SANDERS: Would the paleontologists care to speculate on the actual validity of the so-called trilobite limb, since appar- ently it is known from so few specimens. There may have been differential preser- vation resulting in only a small fraction of the entire spectrum of limb variation being available to us. WHITTINGTON: You may be right. What we really know about trilobite limbs is based upon Stdrmer’s sections, and they show one type of limb in the Ordovician. We do know a somewhat similar limb, not so well, in the Devonian. Then we have the Burgess Shale forms in the Cambrian. It is a reasonable assumption that they were like this, but the evidence is slim. PALMER: In olenellids and agnostids we know nothing of the limbs, and these are two major trilobite groups. BROOKS: Stgrmer brought out the most significant point here in recognizing the amandibulates and the mandibulates among the arthropods. Trilobites, lacking a mandible, are obviously distinct phylo- genetically. How are we going to explain the mandible as a biting structure in the 20 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, Crustacea by saying that they were primi- tively filter-feeders? HESSLER: The mystacocarid mandible is a trilobitan limb except that its basal endite is an elongate gnathobase and there are fewer segments on the endopod. These are not great differences. MANTON: Fine-food feeders require a mandible mainly for rubbing and squeez- ing, but not for biting. I do not think that biting is a primitive attribute of the Ar- thropoda (see p. 111). A crustacean man- dible performing rolling, squeezing and grinding movements suits small-food feed- ing whether this food is derived from the bottom or from suspension. SANDERS: Certainly the mandible, at least primitively, was more than merely a biting or masticating structure. In adult copepods and mystacocarids, and in the nauplii of a large number of groups, the mandible is used to collect food particles. It can also transport particles or, in the case of the larval cephalocarids, detritus to the region of the atrium oris. In other words, the mandible primitively is a gen- eralized limb doing a number of functions. The elaboration we usually see is further modification or simplification, primarily for mastication. DAHL: What does the formation of an exoskeleton carry with it in the form of change of ecology? If we assume for the sake of argument an annelid ancestor, it has a highly flexible head, it can turn any- where to pick up food, crawl on or through the bottom, feeding all the time. If you reach a crustacean stage with a firm exoskeleton, that flexibility tends to be lost, and that must inevitably affect the feeding methods. GLAESSNER: The late pre-Cambrian annelid 7Spriggina has a head which resem- bles that of the annelid Tomopteris. The 1963 latter has a head consisting of several fused segments. To what extent that would affect the position of the mouth and the feeding habits I do not know, because the living tomopterids are secondarily adapted to a pelagic mode of life. The fossil that resembles them is not so adapted and that does not necessarily rule out the relation- ship. The head could develop a strongly sclerotised integument. We have to decide whether that would go with an anterior position of the mouth, and at what stage it would lead to that downward curvature which is one of the key points when we compare annelids and arthropods. We should keep the background of possible annelid-arthropod relationships in mind when we talk about Crustacea. The latter have inherited their segmentation from somewhere, say from an annelid, and the question is what happens to that segmen- tation ‘subsequently in connection with changes in locomotion, respiration, etc. That inherited segmentation is one of the problems that is being overcome in a vari- ety of ways: either by concentration and formation of a cephalothorax, or by reten- tion and development of a uniform seg- mentation which leads to a different mode of locomotion. The Stomatopoda overcome the problem by shifting more of their or- ganization into the abdomen, whereas oth- ers concentrate it in the anterior portion of their body. What of the development of the crustacean limb from the ancestral locomotive limb, not the feeding limb? Is that a new development, or is it another shift in position of a pre-existing, pre- crustacean feature? DAHL: I find it hard, at the primitive crustacean level, to distinguish between a locomotory limb and a feeding limb. The one almost postulates the other. MANTON: At a primitive level many DISCUSSION 21 functions are done by every limb includ- ing the mandible: feeding, locomotion, res- piration, etc. Lochhead’s point, that Triops is capable of feeding on minute particles for unlimited periods of time, without spe- cial structures to facilitate this process, is of importance in all considerations of crus- tacean feeding and speculations as to how the trilobites fed. MOORE: Do the Onychophora fit into the picture? MANTON: A good argument can be advanced for supposing that present-day Onychophora are not degenerate or secon- darily simplified. Functional considerations can account for their soft body wall, un- striated muscle, connective tissue skele- ton, limited scute formation, undifferen- tiated gaits, etc. The Onychophora are so similar in superficial appearance to tAy- sheaia as to suggest a common derivation, but this must be very far removed from that of Crustacea. The Onychophora are associated with the Myriapoda and Hexa- poda on embryological and comparative anatomical grounds. One can see no com- mon basis in either the type of limb or the type of jaw present in the Crustacea and in the Onychophora-Myriapoda-Hexa- poda series. Thus a study of the Onychoph- ora does not help with problems of crustacean evolution. WHITTINGTON: What sort of adap- tation, mode of feeding or of locomotion makes a crustacean? You’ve been saying that possibly the early ones lived on the bottom. So did the trilobites, and the Onychophora. What is the essential dif- ference? How could it be in the environ- ment, if this was presumably somewhat similar? What was the feature that led to the evolution of Crustacea? MANTON: Work on myriapods has shown how evolution has taken place in association with divergent habits which have been established in the same type of environment. Mandibles, locomotory limbs, and differing trunk characters which are diagnostic of large groups (classes and orders), have evolved in association with these habits. This is a quite different type of evolutionary advance from adaptation to particular niches which occurs in the adaptive radiations of classes and orders, usually taking place at a later evolution- ary stage. WHITTINGTON: Annelids can undu- late the body in the horizontal plane, but trilobites cannot. Is this a different habit in the same environment? MANTON: The great difference be- tween annelid and arthropod surface loco- motion (not burrowing) is the predomi- nant use of trunk muscles as well as parapodial muscles in the former, and ex- trinsic and intrinsic limb muscles but not trunk muscles in the latter, for providing the locomotory force. Body undulations caused by trunk musculature are usually a disadvantage, and in the arthropod are controlled up to a point by various means. The biramous legs of trilobites and of Crustacea have probably evolved quite in- dependently from those of +Aysheaia. SANDERS: Regarding the trilobite limb and mode of feeding, one tends to think in terms of a filter-feeding current system. Manton feels that one of the dif- ficulties of such a scheme is that the en- ditic or gnathic spines did not meet along the midline, so that it is difficult to imag- ine how the food was passed forward. However, the trilobites may have fed in a manner similar to the Cephalocarida utilizing large masses of detritus, or may have been predaceous like Limulus. In such a case, there would be no necessity for the limbs to meet along the midline 22 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 since the food could be carried forward from limb to limb as long as the food mass was large enough. LOCHHEAD: Emphasis is often put on the need for particular setae or other structures to transport or push the food forwards to the mouth under the labrum. In a great many cases the current alone will do this. If you watch a captive filter- feeding anostracan or cladoceran and con- centrate on individual particles that are being sucked along in the feeding current, you'll see that a great many particles get all the way to the mouth and in under the labrum, without ever touching any setae at all. The setae are there as a sort of insurance for the ones that go astray. A large percentage of particles will reach the mouth without any help from special structures to push them there, simply be- cause there is a backwash current that will take them up under the labrum. WATERMAN: The visual organs of crustaceans are of particular interest to me and it is notable that among the apparently primitive crustacean groups a remarkable variety of different eye types occur, often quite distinct from the highly developed compound eye of the decapods (Waterman, 1961, ‘Physiology of Crustacea,” 2:1-64.) Considering the crustaceans alone, one might conclude from such evidence that evolutionary exploration of various kinds of visual organ ultimately led to standardiza- tion in one particular type in the most highly evolved forms of the class. Yet the compound eyes of pterygote insects show remarkable structural similarities with those of decapod crustaceans. This parallelism is marked both in gross and microscopic anatomy not only in the structure of the retina but also in that of the eye’s dioptric apparatus which is almost identical. It is true that corneagenous cells present in crustacean eyes are absent in the tracheate compound eye and that certain aspects of the screening pigment differ but otherwise the cell-for-cell detail is astonish- ingly close. Thus a one-to-one correspond- ence may be seen in the relations and struc- ture of the more peripheral retinula with its typical 7 +- 1 pattern of neurosensory cells, the rhabdom made up of radial microtubules perpendicular to the optic axis, the basilar membrane and strands of primary visual axons as well as the more central three suc- cessive optic ganglia (lamina ganglionaris, medulla externa and medulla interna) sep- arated by the external and internal chias- mata. One major neurological difference, which appears as a notable exception, is the medulla terminalis of stalk-eyed crustaceans absent in insects and in sessile- eyed crustaceans, too. This ganglionic mass is believed to have been derived from part of the protocerebrum and to have migrated peripherally in connection with the evolu- tion of movable eyestalks (Hanstrém, 1928, “Vergleichende Anatomie des Nervensys- tems der wirbellosen Tiere.” 628 pp. Springer, Berlin.) In view of these extensive similarities, Dr. Manton’s remark that the hexapod-myri- apod line of arthropod evolution, on the basis of a wide variety of evidence, does not seem to be close to the crustacean line raises some interesting points. If indeed the closely similar compound eyes of the higher insects and higher crustaceans arose inde- pendently in a strong evolutionary conver- gence, one is forced to draw rather drastic physiological conclusions, namely that each of the numerous parallel details in these two kinds of eyes must represent some func- tional component without which a highly efficient eye of this kind cannot operate. Thus one should then ask what it is operationally that requires the presence of DISCUSSION 23 7 retinular cells plus 1 basal or eccentric cell in each ommatidium. Perhaps color perception or sensitivity to the plane of polarized light is dependent on the cluster- ing of retinal cells in this specific way. Hy- potheses of these sorts have been proposed (Hanstrém, 1927, Z. vergleich. Physiol., 6:566-597; Autrum and Stumpf, 1950, Z. Naturforsch., 5b:116-122.) but no direct evidence of their validity is yet available. Similarly the physiologist should demand an explanation for the presence of three op- tic ganglia separated by two chiasmata in these optic tracts. We do know that exten- sive processing of optic information takes place in these regions in decapods (Wiersma, Waterman and Bush, 1961, Science, 134: 1435 (Abst.); Waterman and Wiersma, 1963, J. Cell. Comp. Physiol., 61:1-17) but cannot yet tie together structure and function effectively. These unanswered questions raise still an- other challenging problem. This relates to the existence, or not, of evolutionary alter- natives in the development of certain func- tions. Certain biological processes seem to be the unique solution available to animals for effecting a particular task. Thus all ani- mals known use the carotenoid retinene, (or the very closely related retinenes) as the chromophore of their visual pigments. No exceptions are known although cephalo- pods, insects, crustaceans and many verte- brates have been studied in this regard. Furthermore, these visual pigments, whose individual characteristics are endowed by the opsin moiety of the whole molecule, are invariably located either in submicroscopic lamellae (vertebrates) or in submicroscopic oriented fine tubules (arthropods and ceph- alopods) (Fernandez-Moran, 1959, Rev. Mod. Physics, 31:319-330.) Such monotonous regularity in animal groups which cannot conceivably be con- sidered closely related in terms of their eye evolution suggests some strikingly stringent functional requirements and _ limitations. Yet it is equally clear that at certain levels the evolution of the eye as a whole has been enormously affected by natural selection, as witness the many differences between cam- era eyes, compound eyes and ocelli as well as their extensive variations. The attempt to distinguish a priori between these two types of biological characteristics is a chal- lenging and important point which has received little attention. DAHL: This is one of the most chal- lenging problems which the student of ar- thropod anatomy meets. I am struck by the great structural differences in the eyes of various Crustacea. Still more so if you regard the arthropods as a whole, finding what you regard as the final product with almost identical eyes. In the branchiopods Hanstrém showed that the types of gan- glion cells and interpart connections there are much simpler than in decapods and insects. It seems to be the same in the eyes of Maxillopoda, which also have much simpler pathways. It is hard to conceive how one of these eye types could have been derived from the other ones at that high level. In the embryology of Noto- straca, there is, more or less in the middle portion of the head, a large proliferation area of undifferentiated cells which mi- grate out to form the optic apparatus of the compound eye. Also, there is one strip of cells wandering in to constitute the lamina ganglionaris, and another strip of cells going to form part of the ganglion layer of the medulla, while other parts of the ganglion cells forming the medulla come from the brain. At the same time elements of this proliferation zone wander into the vast nauplius eye complex. So whatever the final function of these 24 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 cells may be, they come from a quite dif- ferent area of the head from that in the Malacostraca, where the rudiment of the optic ganglia forms a proportionately much larger portion of the optic apparatus. This is a further challenge to the correlation of structure and function. TASCH: Is there a development which Waterman attributes to natural selection and another development outside of nat- ural selection that is the unique solution? WATERMAN: Not correctly so since I believe natural selection to be the molding force in both types of biological character- istics I was referring to. In the case of the carotenoid part of visual pigments, how- ever, natural selection has nearly zero scope for action since in all known cases the molecular pattern of this component is essentially the same. Therefore, if vision is going to be present, apparently either reti- nene, or retineney must participate. On the other hand the form, shape, size, color, dioptrics and other such details of eyes obviously can be varied enormously under natural selection and have in fact under- gone fascinating changes of all sorts as al- ready mentioned. In some cases there may be only one way of doing something bio- logically; in other cases there may be sev- eral or many alternatives. TASCH: I am always afraid of the type of formulation which speaks of unique so- lutions, out of random selection. WATERMAN: I agree with your dis- trust if you mean that lack of evidence is a dangerous substitute for negative evi- dence. However, it seems fair to say that the nature of the universe and of the physics and chemistry which underlie living systems surely put definite restrictions on what may or may not take place in evolu- tion whatever the mechanism of the latter. This thesis was elegantly developed many years ago by L. J. Henderson in “The Fit- ness of the Environment” (1913). TASCH: Or you can speak of natural selection amongst compounds. MAYR: I think there are three or four things we might distinguish. One is that certain basic enzymes are so important for life that they start almost simultaneously with life. So if they are now widespread both in animals and plants, it is simply that they were so essential that living or- ganisms could not have gotten very far without them. There have been enzyme studies in microorganisms which have re- vealed the enormously wide distribution of certain enzymes. However, sharing such an enzyme or metabolic pathway cannot be used as proof of common descent be- cause, in the case of the chemistry of cer- tain metabolic phenomena, it seems that only one solution is possible. To find the same macromolecule in different organisms does not necessarily prove their origin from a common ancestor who also pos- sessed such a molecule. The case of hemo- globin is a good illustration for the poly- phyletic origin of a complex molecule, as pointed out by Waterman. Hemoglobin, apparently, transfers oxygen more effi- ciently than any other respiratory pigment, and whenever there was strong selection pressure in favor of efficient oxygen trans- fer, hemoglobin was “invented.” This has happened at least three times, indepen- dently, in the animal kingdom. Cilia, and many components of cell structure, pose the question whether there is only one possible solution or whether the existing similarities are due to extreme phylogenetic age. Electron-microscopy pic- tures of the mitochondria of animals and plants, for instance, look very much the same to me. The peculiar lamellar struc- ture is presumably the key component of DISCUSSION 25 this organelle. Most likely, we have here a combination of a unique solution and great phylogenetic age. We now come to the fourth and most difficult example. All these things up to now were simple things, in one case macromolecules, in the other case, very simple structures. But when we get to something as complex and yet as similar as the compound eye of the deca- pods and the insects, it puts one’s faith in natural selection to a severe test, and that in two separate ways. The first is the as- sumption that the compound eye is the only truly superior eye that an arthropod can have; the second one is that natural selection could have put together all of these pieces in such a way that indepen- dently the same kind of eye emerged, and yet, this is what the phylogeny at first sight seems to suggest. A possible solution to this puzzle is as follows: Evolutionary changes do not happen as the early Mendelians thought by a gene turning up that creates a new character. This atomistic thinking has been com- pletely refuted. There is a total genotype, and it is highly “integrated,” co-adapted, “cohesive,” and it has certain potentiali- ties. One can assume that the potentialities that produced this type of eye go back to the common ancestors of the decapods and the insects. It wasn’t until a particular selection pressure set in that demanded an eye that was better than the eyes of some of the lower crustaceans or primitive rela- tives of the insects, that this similar com- pound eye emerged. It emerged because the two lines possessed the same genetic potential, a potential which responded in an analogous manner to a similar selection pressure. This is rank speculation, but we have a puzzling situation, and we have to propose an explanatory model. DAHL: The anostracans and phyllopods have an eye consisting of generally five plus one retinular cells, a lamina and a medulla without a chiasma between and few connections, and on the other hand, the decapod eye with seven plus one retinular cells, two chiasmata and various synapses in the pathway. Can one expect any functional differences, with respect to image formation, or color vision, or analy- sis of polarized light, etc? WATERMAN: This raises an interest- ing point about arthropod organization in general. As a whole arthropods seem to be most economically organized on a cellular basis. Thus the central nervous system, motor efferents and the sense organs them- selves contain far fewer neurons than do the comparable elements of vertebrates. In Daphnia for example there are only 22 ommatidia (and hence perhaps 150 neuro- sensory cells) in its fused median eye. Yet these creatures are very strongly sensitive to polarized light, differentiate colors and show good evasive action if you try to catch them. Even in decapods there are only a few tens-of-thousands of retinular cells involved in the eye, yet our studies on visual information transfer in these sys- tems (Waterman and Wiersma, 1963, J. Cell. Comp. Physiol., 61:1-17) indicate that the functional performance is closely similar to that found in the frog (Matur- ana, Lettvin, McCullock and Pitts, 1960, J. Gen. Physiol., 43(6), supp. 2:129-175). In crustaceans there are movement receptor units, intensity receptors and novelty recep- tors that respond only to sudden changes in the visual field and not to general illumina- tion or other sustained aspects of the po- tential stimulus. DAHL: I follow Tiegs and Manton’s view that the compound eyes have been formed independently in various arthro- pods. It is hard to explain the derivation 26 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 of the various arthropod groups without assuming that. If you don’t assume that you encounter other convergences, which are equally difficult. It is not due to pure chance that we have compound eyes at these rather different levels of cellular organization. MAYR: There is an amusing parallel with the phylogeny of the mammals. It is not many years ago that perhaps more than half of the people writing on the subject said they couldn’t possibly see how the mammalian middle ear and the jaw mechanism could have ever evolved from the reptilian one—it was an impossi- bility. Now, it is firmly established that on the basis of a basic potentiality for it, this shift has happened independently at least five times, and quite likely seven times. It is evident that the reptilian an- cestors of the mammals had the basic potentiality in their genotype that permit- ted parallel evolutionary changes in sev- eral independent lines. This is that basic potentiality I was talking about as having independently given rise to the compound eye both in the arthropod and in the insect line. WATERMAN: Dr. Mayr suggests that some basic genetic potentiality of a major group like the arthropods might remain un- expressed until some of its component taxa had evolved into quite distinct lines, like crustaceans and insects. Such an idea seems difficult to correlate with the observed emergence of typical compound eyes only in the most highly developed members of these two classes. In other words as their over-all organization drifts further and further apart, we find an important complex sensory system developing a high degree of convergence. PHYLOGENY AND EVOLUTION OF CRUSTACEA Museum oF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 III Adaptation, A Primary Mechanism of Evolution By Otto Kinne Department of Zoology, University of Toronto Toronto 5, Ontario, Canada! Adaptation is a universal phenomenon, characteristic of all living things. Like all great concepts, the concept of adaptation is basically a simple one, stating “that any living thing is somehow fitted to live where it does in fact live” (Simpson et al., 1957, p. 13). In its broadest sense adaptation refers to all alterations of living things which favor survival. Adaptation is there- fore essentially an ecological concept—a concept of compensation for changes that occur in the nonliving and living environ- ment of a given species. And there is no environment—nonliving or living—that does not change with time. Mentioning time emphasizes at once that adaptation is also an evolutionary concept, referring to gradual changes in organisms during the course of phylogeny, changes that in- crease the chances of survival, that make for increased fitness. Darwin himself has proposed that evo- lution could be accounted for completely on this basis. Stanier ef al. (1957) state that “in nature, genetic adaptation is the principal mechanism of evolution” (p. 1 Present address: Biologische Anstalt Helgo- land, Zentrale Hamburg-Altona, Germany. 410), and Simpson emphasizes that .. . “the origin of adaptation .. . is the prime problem of evolutionary biology” (1958, p. 521). As a primary mechanism of evo- lution, adaptation is based on genetic vari- ation and subsequent natural selection; it is—as Mayr (1960) has put it—‘‘a com- promise between conflicting selection pres- sures” (p. 497). Now that the statement contained in the title of the present paper has been documented and justified, let me restrict its ambitious sounding scope. I propose to set here before you a brief outline of our present knowledge of nongenetic and ge- netic adaptation in Crustacea to temper- ature, salinity and to life on land. I shall attempt to illustrate some general trends on the basis of a few subjectively selected examples from literature, with emphasis on the intact, whole organism and on “ecological” conditions. But first let us clarify some terms and concepts that are often used in different ways. The term “adaptation” is taken here to mean adjustments of living sys- tems to one or more factors of their nat- ural environment, which result ultimately in an increase in their capacity to com- 28 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 pete, i.e. to survive and reproduce. Such adjustments are ecologically ‘advantag- eous” in an objective sense, and on this basis, distinguishable from mere responses. Adaptations may be effective continuously, or during certain periods only; they may be different in different life cycle stages, e.g. in nauplius, cypris and adult of the barnacle Balanus, which exhibit a series of different adaptations, a phenomenon known as serial adaptation. Not all functions or structures of a phenotype are necessarily adaptive. An ef- fort was made therefore to exclude papers reporting mere responses. A clear-cut sep- aration between response and adaptation, however, proved to be difficult or even im- possible in many cases. Assessment of the adaptive value of an adjustment requires more knowledge of the ecology and phy- logeny of a species than is frequently available. For detailed analysis, distinction is nec- essary between nongenetic adaptation in- volving nongenetic changes in the response mechanism, also known as acclimation or acclimatization, and genetic adaptation involving genetic changes in the response mechanism. Adjustments known as “‘sen- sory adaptation,” for example, accommo- dation of eyes to different intensities of light, do not generally produce actual changes in the response mechanism per se; such rather simple and fast adjustments represent a special case and are excluded here. Nongenetic adaptations as such are not passed on to the next generation; however, the ability to adapt and the mechanisms involved are evolutionary products. In this sense then, nongenetic and genetic adap- tations represent different aspects of the same basic phenomenon, even though they operate at different levels. In practice, these two kinds of adaptation can be dis- tinguished on the basis of breeding ex- periments and performance tests under different environmental conditions. Both nongenetic and genetic adaptation may be expressed at a functional level, e.g. in changes of rates and efficiencies of metabolism or in changes of behavior, and at a structural level, e.g. in changes con- cerning the architecture of cells, organs or the whole organism. Most of our present knowledge on adaptation in Crustacea is based on results obtained on macroscopic forms, especially decapods. There is great need for information on smaller forms such as branchiopods, ostracods, copepods and cirripeds. Some nongenetic and ge- netic adaptations in Crustacea are univer- sal among animals—for example, func- tional adaptations to temperature such as shiftings of lethal limits, changes in activ- ity or in metabolic rates; others appear principally to be related to specific crus- tacean features, such as their primarily aquatic way of life, respiration through gills, urine formation in antennal or maxil- lary glands, exoskeleton and body shape. Species names have been used in accord- ance with The Physiology of Crustacea (T. H. Waterman, ed., Academic Press, New York, 1960/61) regardless of the names originally employed by authors of cited articles. NONGENETIC ADAPTATION Let us first consider the nongenetic type of adaptation. The ability to acclimate appears to be greatest in species that en- counter extensive alterations in their nat- ural nonliving or living environment. While it is usually necessary for a qualified anal- ysis to begin with studies of acclimation to single factors such as temperature, sa- linity, light or social and behavioral as- KINNE: ADAPTATION 29 pects, it should be kept in mind that an organism acclimates to its total environ- ment rather than to single factors. The capacity for acclimation depends on—besides genotype and environment— physiological condition and age of the in- dividual and may vary at different life cycle stages. In general, the capacity seems to reach its maximum during early ontogenetic development, that is, in eggs or early postnatal stages, and to decrease with increasing age of the individual. Non- genetic adaptations that have been ac- quired during the most sensitive phase of a life cycle or that are the results of repeated reinforcements, may be _trans- ferred to the next generation or even through several life cycles by nongenetic transmission (examples in Prosser, 1958). It is necessary in such cases to distinguish between individual nongenetic adaptation (without nongenetic transmission) and superindividual nongenetic adaptation (with nongenetic transmission). No criti- cally analysed cases of superindividual nongenetic adaptation among Crustacea have, however, come to my attention. In the time course of nongenetic adap- tation, three successive phases may be dis- tinguished: (i) immediate responses, be- ginning seconds or minutes after change of environment and involving increased fluctuation of performance, i.e., over- and undershoots and shock behavior; (ii) sta- bilization, beginning minutes or hours after the change and leading to progres- sively increasing constancy of perform- ance, thereby gradually approaching a steady level; (iii) new steady state, begin- ning hours, days or weeks after the change, i.e. after completion of the most effective adjustments. Immediate responses to sudden changes in temperature or salinity may involve fluctuations in the overall activity, changes in behavior and over- or undershoots of metabolic rates (examples in Kinne, 1963). Of particular interest here are the over- and undershoot responses ob- served in Cyclops strenuus by Scherbakoff (1935), in Neomysis integer, Hemimysis lamornae, Diaptomus gracilis, Artemia salina and developing eggs of Astacus pal- lipes by Grainger (1956), and in Daphnia magna and Simocephalus vetulus by Mei- jering (1960). A typical example is the oxygen consumption of the brine shrimp, Artemia salina (Grainger, 1958) (Figs. 3, 4). Oxygen consumption of Artemia over- shoots upon changing the water tempera- ture from 10°C to 30°C (Fig. 3). This +10? ><«————_-. 30.0°-—_—_—______—_> 0.2 0.1 #1 O2/ min. /0.01 gm. HOURS Fic. 3. Immediate response of oxygen con- sumption in the brine shrimp, Artemia salina, following a sudden increase in temperature. The four shrimps used in this experiment had pre- viously been kept at 10°C for several weeks (After Grainger, 1958). metabolic overshoot response ends largely after 30 minutes and completely after 1 hour. Oxygen consumption remains then fairly constant for the next 8 hours and thereafter decreases to what may be con- sidered the final new level. The period between 2 hours and 26 hours (Fig. 3) seems to indicate the phase of stabiliza- tion, and the period beginning after 26 hours, the commencement of the new 30 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 steady state. Metabolic over- and under- shoots occur even in Artemia salina anes- thetized with ether and lying motionless in the respiratory chamber (Fig. 4). The only motor activity in the anesthetized individuals was the beating of their hearts and occasional slight twitches of their in- testines. This interesting result indicates that the immediate responses observed are not simply due to changes in loco- motory activity but based on actual rate changes of basal metabolism. Sudden changes of rather constant tem- peratures may also occur under natural conditions, and the immediate responses reported above may therefore be consid- ered a normal first step in the acclimation process. Our present knowledge on this subject is meager and does not allow fur- ther generalizations. A more normal situ- ation is doubtless gradually changing tem- peratures. These may cause fluctuations of rates around the steady state curve (obtained after acclimation to various con- stant temperatures) with the tendency of overshooting that curve under conditions of slowly rising temperatures and of under- shooting it at slowly declining tempera- tures (Fig. 5). The term phase of stabilization consti- tutes at present not much more than a brief description of the period between the highly unstable immediate response and the rather stable new steady state. The duration of the phase of stabilization var- ies in different species and for different processes. Within a given species it de- pends on age, physiological condition, tem- perature and the severity of the stress em- ployed. In general, speed of acclimation seems to be proportional to metabolic rate. McLeese (1956) has transferred the ma- rine lobster Homarus americanus from 14.5° to 23.0°C and shown that thermal acclimation was practically completed in about 22 days (Fig. 6); substantial accli- mation to low salinity had occurred within one week. Schwabe (1933) exposed the <—300%>+-—_—_—_—_100*____>_—_300*____> 0.08 0.07 0.06 0.05 0.04 0.03 pl O2 /min./0.01 gm. 0.02 0.01 O i 2 3 4 5 HOURS Fic. 4. Under- and overshoot responses of oxygen consumption, following abrupt tempera- ture changes, in anesthetized, motionless Artemia salina. Two of the nine individuals died during this experiment (After Grainger, 1958). KINNE: ADAPTATION 31 ) © 1) w i 7) a Ww © w a iis w fe 10 Fic. 5. 20 2 Typical immediate responses of heart activity to slowly rising temperatures in two 16mm long Gammarus duebeni males. A, steady state curve. Male 7 had been kept at 6-7°C for 14 days and was then exposed to temperatures rising to 23.0°C within 3% hrs; male 2 had been kept at 14°C for 14 days and was then exposed to temperatures rising to 26.1°C within 2 hrs. All points represent averages of 10 counts. 10%. salinity. (Modified from Kinne, 1952.) freshwater living Astacus astacus into blood isosmotic water of 15% salinity? and found that blood osmoconcentration reached a new steady state after 12 days, and O» consumption after 20 to 30 days; the crab Potamon acclimated to new salin- ity levels in about 12 days (Drilhon- Courtois, 1934). Shorter periods were found in the freshwater isopod Asellus aquaticus and the Australian mangrove crab Heloecius cordiformis. According to 2 The symbol %- expresses salinity in parts per thousand. Edmonds (1935), Heloecius transferred into 25% sea water, reached a new steady state of its blood osmoconcentration after about 12 hours; this period, however, is longer after transfer into more extreme salinities. From these and other references, it may be concluded that the phase of stabilization may last anywhere from a few hours to several weeks. The new steady state relative to the original steady state has been dealt with in innumerable papers, particularly with reference to lethal limits of temperature w bo MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 AVERAGE SURVIVAL TIME IN HOURS 0 2 4 6 8 IO N2.-14216 ACCLIMATION TIME 18 20 22 24 26 28 30 32 34 IN| DAYS Fic. 6. Duration of phase of stabilization in the lobster Homarus americanus. Given is the gain in average survival time of groups of 6-10 lobsters which had been kept at 14.5°C; the groups were transferred to an acclimation temperature of 23.0°C for periods ranging from 1 to 31 days and then tested at 30°C. Thermal acclimation is practically complete after about 22 days (After McLeese, 1956). and salinity and metabolic rate. In most cases, acclimation to low intensities of temperature or salinity tends to shift the lower lethal limit downward, and acclima- tion to high intensities tends to shift the upper lethal limit upward; examples are: Armadillidium, Porcellio, Oniscus, Ligia (Edney, 195la,b; 1960), Gammarus (Kinne, 1953a), Streptocephalus (Moore, 1955), Homarus (McLeese, 1956) (Fig. 7), Artemia (Grainger, 1958). Matutani (1960), however, reports a case of inverse compensation for the marine copepod, 77- griopus japonicus: while 30°C-acclimated Tigriopus exhibit a higher heat resistance than 20°C-acclimated ones, acclimation temperatures below 20°C (10°; 5°C) cause progressively increasing heat resist- ance. Metabolic acclimations to temperature may result in characteristic quantitative differences between the original and the new steady state which have been de- scribed and classified by Precht (e.g. 1949, 1958) and later, in a modified version by Prosser (1958). Precht distinguishes be- tween compensations (types 1 to 3), no compensation (type 4), and “‘inverse”’ com- pensation (type 5); type 3 represents the normal case, namely, partial compensation; type 1 represents over-compensation, and type 2, perfect compensation resulting in constant rate functions. Metabolic accli- mations have been demonstrated in many KINNE: ADAPTATION 33 100 (o,) © Oo Oo PERCENT MORTALITY psy oO 20 25) 26 e7 TEST TEMPERATURE The effect of nongenetic thermal adaptation on the lethal levels of temperature in Homarus Fic. 7. 28 29 30 3) 32 ° e: americanus. Numerous lobsters were exposed to various test temperatures after acclimation to 5°, 15° or 25°C. The curved lines are drawn through points representing per cent mortality at the respec- tive test temperatures after 48 hours; the temperature that would cause 50% mortality is in- dicated by dotted lines and arrows. 30% salinity; 6.4 mg O./l (After McLeese, 1956). crustaceans, including Daphnia (Precht, 1949), Gammarus (Kinne, 1952, 1953a, b; Krog, 1954), Pachygrapsus (Roberts, 1957a,b), Balanus (Barnes and Barnes, 1958) and Uca (Vernberg, 1959a,b,c). Acclimation to different temperatures may also involve changes in behavior (ori- entation, migration, territorialism) and in the quantity of physiologically important substances as well as in body size and structure. In Eriochei sinensis kept in fresh water, for example, acclimation to 1° to 3°C results in a marked decrease of free proline concentration in the intracel- lular pool of amino acids in muscles, rela- tive to individuals kept at 15°C (Ducha- teau and Florkin, 1955) and in Daphnia kept at a constant Oz pressure, temperature increase augments the synthesis of hemo- globin (Florkin, 1960). In various species, different levels of temperature cause func- tional and structural color changes which may have adaptive value. Structural or morphological color changes involve actual changes in the amount of pigment per pigment cell or changes in number of chro- matophores per unit area (e.g. Brown, 1934). Changes in body shape are docu- mented by the well known phenomenon of cyclomorphosis in Cladocera (Wesenberg- Lund, e.g. 1900; Ostwald, 1904; Wolter- eck, e.g. 1913; Brooks, 1946, 1947, 1957; Lieder, 1951) and Copepoda (Margalef, 1955). Species from small bodies of water in high temperate latitudes, such as Daphnia cucullata and Daphnia retrocurva, 34 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 exhibit spectacular variations of helmet, head crest and spines, particularly if the parthenogenetic young develop at high temperatures of at least 18° to 20°C. These structural modifications per se, how- ever, do not seem to have adaptive value. They seem to be incidental expressions of a functional acclimation to the overall seasonal conditions, especially to high temperature and increased water turbulence. Water temperature and turbulence affect relative growth presumably through an in- crease in metabolic rate (Brooks, 1957; see also Hrbacek, 1959). Metabolic acclimation to different levels of salinity leads to new steady states which may be (1) unaffected, i.e. identical with the original state, (2) higher in subnormal salinities and/or lower in supranormal salinities, and (3) higher both in sub- and supranormal salinities. An unaffected, con- stant respiratory rate has been reported in Eriocheir sinensis after acclimation to fresh water, 15%o or 32%o (Schwabe, 1933; Krogh, 1939). Examples for the second case—higher metabolic rates in subnormal salinities and/or lower metabolic rates in supranormal salinities—include most spe- cies tested, e.g. Carcinus maenas, Eriphia spinifrons (Schlieper, 1929; Schwabe, 1933), Gammarus locusta (Schlieper, 1929), Pagurus longicarpus (Maloeuf, 1938), Uca spp. (Gross, 1957a), Hemi- grapsus oregonensis (Dehnel, 1960), As- tacus astacus (Schwabe, 1933), Potamon edulis (Raffy, 1934) and Gammarus due- beni (Kinne, 1952). Examples for the third case—higher metabolic rate in both sub- and supranormal salinities—include: Ocy- pode quadrata (Flemister and Flemister, 1951), Palaemonetes varians (Lofts, 1956) and Metapenaeus monoceros (Rao, 1958). As has already been pointed out, or- ganisms acclimate to their total environ- ment rather than to single factors. Studies concerned with acclimation to factor com- binations and with acclimation in the natu- ral habitat are therefore of great impor- tance. A critical, detailed analysis of such studies is, however, difficult for three reasons: (1) the situation is extremely complex; (2) sufficient information on en- vironmental and biological history of habi- tat and organisms is mostly not available and difficult to obtain; (3) nongenetic and genetic components of an adaptation can- not readily be distinguished and assessed. It is evident though, from literature, that quality and quantity of a given acclimation to one environmental factor depend not only on the physiological condition of the individual involved and its genetic back- ground, but also on other environmental factors. The capacity and rate of an accli- mation to salinity, for example, may be different at different levels of ambient tem- perature or oxygen tension. In other words, the efficiency and perfection of a given ac- climation may be different at different levels of other simultaneous acclimations. Har- mony or disharmony between concomitant acclimations may therefore well be a fun- damental way of increasing or decreasing the total capacity for nongenetic adaptation in complex environments. GENETIC ADAPTATION Evidence for genetic adaptation to tem- perature, salinity and life on land comes largely from crustaceans living in different latitudes, in sea or fresh water or in aqua- tic or terrestrial habitats. An attempt to attribute a given adaptation to a single ecological master factor like temperature or salinity is therefore difficult, certainly more so than in the case of nongenetic adaptations which are documented largely on an experimental basis. KINNE: ADAPTATION 35 In Crustacea, genetic adaptation to tem- perature is much less pronounced than to salinity or to life on land. Scholander and his collaborators (1953) and others have established that metabolism of arctic crus- taceans is much higher at a given tem- perature than that of their tropical counter- parts. These adjustments may be consid- ered a genetic adaptation of basal metab- olism to life at low temperatures, since locomotory activity appears to be similar in both groups and since metabolism in isolated tissues shows parallel ad- justments. Such adaptations result often in considerable differences in tempera- ture optima (Wingfield, 1939), and lower and upper limiting temperatures (Krog, 1954; Takeda, 1954; Spoor, 1955; South- ward, 1958) and may consist of a genetic as well as a nongenetic component. A variety of interesting genetic adap- tations to salinity have made it possible for numerous crustaceans to leave their oceanic home and to establish themselves in a wide range of salinity conditions such as brackish water, fresh water and brine, and even to conquer land. The adjustments known so far are largely of an osmotic nature involving changes in (1) quantity and quality of active absorption and ex- cretion of salt and water, (2) surface per- meability to salt and water, (3) osmo- concentration and ionic composition of body fluids, (4) tissue tolerance to fluc- tuations of osmoconcentration and _ ionic composition of blood, (5) salt and water storage in tissues, and (6) behavior. In an attempt to classify different de- grees of genetic adaptation to salinity, four major groups may be distinguished: (1) Polystenohaline inhabitants of the ocean with its rather constant salinity, (2) euryhaline inhabitants of coastal, estuarine or brine habitats characterized by reduced, fluctuating or extreme salinity, (3) oli- gohaline inhabitants of fresh water and (4) holeuryhaline inhabitants of sea water, brackish water and fresh water. (1) Polystenohaline species are osmo- conformers with ion and volume regulation but little or no osmoregulation; examples are: Maja verrucosa, Hyas araneus, Can- cer antennarius, Emerita, Callianassa, U po- gebia, Speocarcinus, Lophopanopeus, Pagu- vus and Palinurus (e.g. Duval, 1925; Schlieper, 1929; Schwabe, 1933; Robert- son, 1949, 1960; Gross, 1957a) and pre- sumably most other oceanic Crustacea. Osmoconformers swell rapidly in diluted sea water and gain salts in concentrated sea water. Their blood osmoconcentration is isosmotic to the surrounding medium (A = —1.9°C) and conforms readily to any salinity changes (Krogh, 1939). (2) Euryhaline species have developed a reduced permeability of gills and cara- pace to water and salt as well as improved mechanisms for differential absorption and excretion of ions; examples are: Carcinus maenas (e.g. Duval, 1925; Schlieper, 1929; Nagel, 1934), Rhithropanopeus harrisii (Kinne and Rotthauwe, 1952), various species of Gammarus (e.g. Widmann, 1935; Beadle and Cragg, 1940a; Kinne, 1952; Werntz im Prosser and Brown, 1961), Palaemonetes, Palaemon, Penaeus, Meta- penaeus (Panikkar, 1939, 1940a, b, 1941a, 1950), Crangon (Broekema, 1941), Ar- temia (Croghan, 1958b), Uca, Pachygrap- sus (Jones, 1941; Prosser et al., 1955; Gross, 1955, 1957a; Green et al., 1959), Ocypode (Flemister and Flemister, 1951), Hemigrapsus (Gross, 1955), Heloecius, Leptograpsus (Dakin and Edmonds, 1931; Edmonds, 1935) and Birgus (e.g. Gross, 1955, 1957a). On the basis of their osmo- regulative performance, these euryhaline species can be subdivided into two groups: 36 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 the hyperosmotic regulators which are hyperosmotic in diluted sea water but more or less isosmotic in higher salinities (e.g. Carcinus, Rhithropanopeus, Gammarus (Fig. 8) and the hyper-hypo-osmotic regu- lators which are hyperosmotic in diluted fresh water 10 20 sea water as well as hypoosmotic in higher salinities (e.g. most shore shrimps, brine shrimps and semi-terrestrial and terrestrial crabs) (Fig. 8). Hyporegulation is always correlated with hyperregulation suggesting that hyper-hypo-regulation is indicative of 30 40 50 eS 100% sea water (34.5 %ooS) Gammarus 4 vA duebeni / vA oe Rhithropanopeus -~ harrisii .~~ tridentatus -7-~ Uca crenulata O 1.0 Fic. 8. Crangon crangon Artemia salina AC MEDIUM 2.0 Blood osmoconcentration as function of salinity (fresh water to double strength sea water) in euryhaline Crustacea. Hyperosmotic regulators: amphipod Gammarus (from Kinne, 1952), and brachyure Rhithropanopeus (Kinne and Rotthauwe, 1952); both had been acclimated to the respective salinities for several weeks at temperatures between 20° and 21°C. Hyper-hypoosmotic regulators: land crab Uca (Jones, 1941), shore shrimp Crangon (Fliigel, 1959), and brine shrimp Artemia (Croghan, 1958b); all three had been acclimated to the respective salinities for at least 2 days, Uca at 17°-18°C, Crangon at 15°C and Artemia at 18-24°C. Blood osmoconcentration may change significantly with temperature. KINNE: ADAPTATION 37 an advanced stage of genetic adaptation to fluctuating or extreme salinities. Hyper- hypo-regulation is found also in sea water- inhabiting insects and represents presum- ably the most elaborate genetic adaptation to osmotic stress within the invertebrates. (3) Oligohaline species inhabit fresh water and are characterized by a well de- veloped hyperosmotic regulation meeting fresh water 10 2.0 4 the osmotic requirements for life in very dilute media. Usually their osmoregulatory mechanism collapses in salinities above 5 to 10%o; examples are: Potamon (Schlieper and Herrmann, 1930; Shaw, 1959), Palae- monetes antennarius (Parry, 1957), Gam- marus pulex, Gammarus lacustris (Lock- wood, 1961) and presumably most other freshwater living crustaceans (Fig. 9). 20 30 foo S JL, Eriocheir U1 £7 sinensis A Astacus = m7 astacus Ys Hoe / Gammarus pulex via A yy magna A °C MEDIUM 0 0 1.0 2.0 Fic. 9. Blood osmoconcentration as function of salinity in oligohaline Crustacea; Potamon (from Duval, 1925; Schlieper and Herrmann, 1930), Astacus (Herrmann, 1931; Beadle, 1943), Gammarus (Beadle and Cragg, 1940a), Daphnia (Fritzsche, 1917), and in the holeuryhaline crab Eriocheir (Scholles, 1933). Details on acclimation period and temperature are mostly not available. 38 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 Genetic adaptations to life in fresh water include: very low surface permeability to water and salt, active salt absorp- tion by gills, reduction of normal os- moconcentration of body fluids, salt re- absorption, water expulsion and accumula- tion of food and salt in the egg, making the species more independent of the me- dium during its most critical stage. Gaining the ability of perfect hyperregulation ap- parently goes hand in hand with losing the ability to adjust to increasing salinities. (4) Holeuryhaline species are rare and poorly investigated. They inhabit sea-, brackish- and fresh water, migrating as individuals from one medium to the other, or establishing populations in all three media. The best investigated example is the crab Eriocheir sinensis (Scholles, 1933; Schwabe, 1933; Schlieper, 1935; Krog, 1954) (Fig. 9); another example is its close Indian relative, Varuna litterata (Panikkar, 1950). Adaptation to life in the whole salinity range from fresh water to sea water includes genetic and nongenetic plasticity, exceedingly low surface perme- ability to water and salt, highly advanced absorption and excretion of salt against steep gradients, and high tissue tolerance to fluctuations in blood osmoconcentration. The mechanism of hyperosmotic regu- lation appears to be primarily based on an antagonism between loss of salt from ex- ternal surfaces and antennary glands, and active salt uptake from the medium. In Gammarus duebeni, G. locusta and G. ob- tusatus changes in blood osmoconcentration are due to salt movements rather than to water movements (Beadle and Cragg, 1940a). G. duebeni is capable of reducing salt loss under conditions of hypoosmotic stress (Shaw and Sutcliffe, 1961). In salin- ities below 50% sea water, such reduction is accomplished by producing urine which is hypoosmotic to the blood but hyper- osmotic to the external medium (Lock- wood, 1961) and possibly also by changes in surface permeability (Shaw and Sut- cliffe, 1961; Lockwood, 1961). As the salin- ity is decreased below 50% sea water, G. duebeni increases the rate of urine flow (until in fresh water it reaches the equiv- alent of 70% total body water/day) and decreases its urine concentration. Such adjustments may be achieved within 2 hours, a fact that appears to be important in view of rapid salinity fluctuations to which G. duebeni is often exposed in its natural habitat. Blood-hypoosmotic urine is presumably also produced by G. zad- dachi and G. salinus when they are in dilute media (Lockwood, 1961). In the shore crab Carcinus maenas, genetic adap- tation to life in reduced salinities includes reduced surface permeability to water and salt, high tissue tolerance to lowered osmo- concentration of body fluids and active salt absorption from the external medium (probably by gills). The most important osmoregulatory organ seems to be the gill. In sea water and brackish water the an- tennal glands produce urine which is ap- proximately blood-isosmotic; they play no significant part in osmoregulation. Urine output increases with decreasing salinity, and it is assumed that the gills replace such progressive salt loss by a reciprocal in- crease in active salt absorption (Nagel, 1934). The mechanism of hypoosmotic regu- lation is not yet sufficiently investigated to generalize. The most thoroughly ana- lyzed species is the brine shrimp, Artemia salina, an anostracan which exhibits pro- nounced genetic adaptation to life in high salinities (Croghan, 1958a-e). Artemia swallows its medium continuously and takes up water from the gut lumen. The osmotic KINNE: ADAPTATION 39 pressure of the gut fluid is appreciably higher than that of the blood but in more concentrated media is considerably below that of the medium. Regulation occurs in gills (salt balance) and gut (water bal- ance). The ionic ratios of the hemolymph are relatively constant, and very different from those of the medium. Changes in hemolymph osmoconcentration that may occur as salinity is varied, are due more to net movements of NaCl than to water movements. The low osmoconcentration of body fluids, the type of ionic regulation and the low internal Mg concentration re- semble conditions found in freshwater ani- mals and have been interpreted as evidence for the freshwater ancestry of brine living forms (e.g. Robertson, 1960). shows an appreciable degree of perme- ability, especially in the gut epithelium. It can actively excrete (first 10 pairs of gills) and absorb (probably first 10 pairs of gills and gut) NaCl. The gut has apparently become adapted as a mechanism for active uptake of water, controlling water balance and preventing dehydration in hyperos- motic media (Croghan, 1958b, c,d). These mechanisms are similar to those employed by marine teleosts. Palaemonetes varians and Palaemon longirostris produce rather large amounts of blood-isosmotic urine over a wide range of salinities (Panikkar, 1939; Parry, 1955, 1957); there must conse- quently be an intensive absorption of salt, particularly in rather diluted media. Pachy- grapsus crassipes produces slightly blood- hypoosmotic urine when in a diluted me- dium, and blood-isosmotic urine when in hyperosmotic sea water (Prosser et dal., 1955). By immersing Pachygrapsus cras- sipes in different salinities containing varying concentrations of Mg, it was shown that the urine Mg concentrations are not a direct function of Mg influx, but rather Artemia of water influx. Furthermore, it could be demonstrated that the muscle tissue of P. crassipes swells if the crab is immersed in dilute sea water and shrinks if it is im- mersed in concentrated sea water. The volume changes of muscles take place at the expense of the blood space; the crab does not change weight (Gross and Mar- shall, 1960). In Uca pugnax and Uca pugi- lator kept in 100% and 175% sea water, urine osmotic and electrolyte concentrations are significantly blood-hyperosmotic. The chief sites of entrance of water and salt are the stomach and the gills, and the chief sites of regulation are the antennal glands and the gills with some regulation by the stomach and possibly the midgut gland (Green et al., 1959). Ocypode quadrata reabsorbs water in its antennal glands when in air or in blood-hyperosmotic salinities. It excretes water in its antennal glands and may also excrete chloride when in hypoosmotic salinities. The gill membrane is assumed to function in the reverse way (Flemister and Flemister, 1951). Reab- sorption of water in antennal glands has also been demonstrated in the land crab Gecarcinus lateralis (Flemister, 1958). The antennal glands play little or no osmo- regulative role in Palaemonetes, Palaemon and Pachygrapsus but may assist in ionic regulation (Pachygrapsus). In the semi- terrestrial Uca, however, and in the quite terrestrial Ocypode and Gecarcinus, anten- nal glands have become progressively ca- pable of reabsorbing water and of excreting salt against the gradient—capabilities that are clearly effective in osmoregulation and that must be considered adaptations to life on land. Behavioral mechanisms of hypoosmotic regulation have been reported for the land crab Pachygrapsus crassipes which selects suitable salinities if given a choice (Gross, 40 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 1957b), for Coenobita perlatus, which varies the frequency with which it visits water of different salinities as well as the time spent in or outside of such waters (Gross and Holland, 1960) and for Birgus latro, which can moisten its respiratory membranes with the help of its appendages and drink water even from small sources (Gross, 1955). In semiterrestrial and terrestrial crabs it has been suggested that the primary im- portance of hypoosmotic regulation is an adaptation to evaporation from the bran- chial chamber, and that regulation against branchial fluids reaching extremely high concentrations by evaporation would pre- vent the blood from attaining critical os- motic pressure (Jones, 1941). This hy- pothesis was later rejected, mainly on the grounds that the salt contained in the small volume of branchial fluid (Pachy- grapsus) is insufficient to effect a harmful rise in osmoconcentration of body fluids even if it were completely absorbed (Gross, 1955). The fact remains, however, that semiterrestrial crabs may be subjected to extreme osmotic conditions, i.e. rain and extensive evaporation, and that they may have to enter water of deviating salinity even if only for short periods. Hyperregu- lation then seems to be of adaptive value in conditions of heavy rainfall or other contact with fresh water or water of re- duced salinity, and hyporegulation seems to be of adaptive value in meeting osmotic stress endured during submersion in water of high salinity (such submersion also pre- sents a respiratory problem) or during periods of prolonged desiccation, and they may aid in water conservation. In air, gills must be kept moist to facilitate respiration (Krogh, 1919); consequently, they lose water. Excessive water loss from evapo- ration leads to significantly increased blood osmoconcentration (Jones, 1941; Parry, 1953). Hyporegulation could conceivably be of advantage here in two ways: (1) in facilitating active salt excretion against the gradient and thus compensating to some extent for the increase in osmoconcentration of body fluids, (2) in maintaining a high salinity on the moist gill surface which would lead to a somewhat decreased vapour pressure and hence reduce the rate of evaporation. With respect to the osmoregulatory mechanism employed by oligohaline spe- cies, two groups may be distinguished: the first group produces a more or less blood-isosmotic urine, e.g. Potamon edulis (Schlieper and Herrmann, 1930), Potamon niloticus (Shaw, 1959), Palaemonetes an- tennarius (Parry, 1957); the second group produces a blood-hypoosmotic urine, e.g. members of the family Astacidae in which the urine has usually only about 10% of the blood osmoconcentration (Robertson, 1960), Gammarus pulex, G. lacustris (Lockwood, 1961) and presumably numer- ous other freshwater living crustaceans. The river crab Potamon appears to be rather poorly adapted to life in fresh water: it has a high blood osmoconcentration (A =—1.1° to —1.2°C in P. edulis); it is more permeable to water and salts than many other freshwater organisms, and it produces urine, which is practically blood- isosmotic. However, if compared to Car- cinus or Eriocheir, it shows an all-around reduction in surface permeability, both to water and salts. Potamon produces only small amounts of urine and actively ab- sorbs sodium and potassium from the ex- ternal medium (Schlieper and Herrmann, 1930; Shaw, 1959). The shrimp Palaemo- netes antennarius has a lower blood osmo- concentration (A = —0.75°C) than Pota- mon, but loses large amounts of salt via KINNE: ADAPTATION 41 an almost blood-isosmotic urine (A = —0.67°C), which is produced at the rate of about 2% body weight/hr (Parry, 1957). Palaemonetes antennarius is not particularly well adapted to its freshwater environment. Not only does it have to use energy in order to compensate for consider- able salt losses via diffusion and excretion, but it appears in addition to be limited by a “threshold” Na-concentration in the medium which effectively stops the uptake mechanism below a finite Na-concentration. This “threshold” lies between 0.125 uM Na/] and 0.183 uM Na/I. Similar thresh- olds may exist for other monovalent ions and may attribute to the prawn’s discon- tinuous geographic distribution. The po- tential difference across the body wall is negative with respect to the inside of the prawn, implying that Cl— is taken up actively, and Na* follows passively. Di- valent ions may be equally important, since they can seriously affect the permeability of the prawn’s surface. A low concentration of Ca++ in the medium would, for ex- ample, tend to increase cuticle permeability, which in turn would increase urine flow and salt loss, thus forcing the prawn to increase the rate of active salt uptake (Parry, 1961a). Within the second group, the Astacidae, blood osmoconcentration varies between A =—0.6° and —0.8°C_ (Robertson, 1960). In an external medium of A= —0.018°C, the crayfish Astacus astacus maintains a blood osmoconcentration A = —0.81°C and excretes a very dilute urine of A = —0.09°C. In fresh water its urine output amounts to 4% of its body weight per 24 hours (Herrmann, 1931) or to 0.175% body weight/hr (Scholles, 1933). Urine output decreases with increasing salinity, approaching zero in a blood-isos- motic medium (Scholles, 1933). Procam- barus clarkit compensates for osmotic water inflow in fresh water by excreting blood- hypoosmotic urine at the rate of 5.2% of its body weight per 24 hours (Lienemann, 1938). Under none of the conditions offered in experiments has a crayfish been observed to excrete a completely salt-free urine. A 50 gram crayfish loses 600mM of Cl— daily (Prosser and Brown, 1961). Very low blood osmoconcentrations are found in smaller forms such as Gammarus pulex and Asellus sp., which have A’s of —0.4° to —0.6°C (Beadle and Cragg, 1940a; Parry, 1953), and Daphnia magna with a A of —0.2° to —0.3°C (Fritzsche, 1917). Gammarus pulex lacks the capacity to vary its urine concentration (which is present in G. duebeni); consequently, in solutions more concentrated than 20 to 30mM/I, its urine becomes hypoosmotic not only to the blood but also to the me- dium (Lockwood, 1961). Its main regu- latory mechanisms seem to be active ion uptake and differential surface permeabil- ity. Asellus aquaticus is fairly permeable to salt and water, and maintenance of its internal concentration against a gradient of approximately 100:1 must result pri- marily from replacement of ions from the medium at the same rate as they are lost from the body by diffusion and in urine. Continued maintenance of blood osmocon- centration during 8 days of starvation shows that NaCl loss can, if necessary, be replaced solely by active uptake from the medium, i.e., independent of the food sup- ply (Lockwood, 1959). The freshwater liv- ing branchiopod Triops cancriformis main- tains its blood osmoconcentration by (1) relative surface impermeability, and (2) salt uptake from food. Its osmoregulation breaks down in slightly blood-hyperosmotic media (Parry, 1961b). Salt uptake from food has also been demonstrated or sug- 42 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 gested in other freshwater crustaceans, e.g. Branchipus (Krogh, 1939) and Chiro- cephalus (Panikkar, 1941b). Beadle and Cragg (1940a,b) have sug- gested that adaptation to life in fresh water has proceeded by two main stages: (a) Maintenance of a high blood osmoconcen- tration (as in Potamon) associated with a large blood/tissue Cl gradient; at this early stage sudden increase in salinity can still be tolerated. (b) Evolution of renal salt-reabsorption and lowering of both blood osmoconcentration and blood/tissue Cl gradient to levels more easily maintained (as in Gammarus pulex and most other freshwater species); at this advanced stage higher salinities are lethal. Pearse and Gunter (1957) consider the most essential requirement for permanent establishment in fresh water to be the accumulation of food and salt in the egg, making the spe- cies in its most critical stage relatively in- dependent of the medium (see also Pearse, 1950). Once the transition from the sea to fresh water has been achieved, it is seldom reversed. Gaining the ability of perfect hyperosmoregulation apparently goes hand in hand with losing the ability to adjust to changing or increased salinities (e.g. Adolph, 1925). The main osmotic problems facing an organism migrating into fresh water from the sea or from brackish water are: con- tinuous inflow of water which has to be expelled; paucity of ions which have to be actively absorbed from a very dilute medium and from food; increased vari- ability of the ionic composition of the medium, and of its temperature and chem- istry. Adaptations in crustaceans to life in fresh water that have been demonstrated or suggested, include very low differential surface permeability to water and salt, active salt absorption by the gills, reduction of normal osmoconcentration of body fluids, salt reabsorption and water expulsion. More information on extrarenal routes of water expulsion as well as on the regulatory capacities of gills, intestinal tract and antennal (maxillary) glands is urgently needed. Calculations concerning the thermody- namic work performed in osmotic regulation suggest that the reduction of normal osmo- concentration of body fluids is the princi- pal means of easing osmotic stress in crustaceans migrating into brackish and fresh water, and that production of a hypoosmotic urine affords little advantage to the organism until a salinity well below 17%o is reached; in freshwater animals, however, hypoosmotic urine may reduce osmotic work by 80 to 90% (Potts, 1954). These considerations have been criticized by Shaw (1959), pointing out that Potts’ calculations are based on the assumption that the outer surfaces are semipermeable, which is not true for several species in- cluding Potamon niloticus and Eriocheir sinensis (the latter was used by Potts as an example to illustrate his arguments). According to Shaw (1959) reduced differ- ential surface permeability to water and salts is of greater importance than reduced osmoconcentration of blood or urine. If, on the other hand, reduction in permeabil- ity is restricted largely to salts and the animal remains relatively permeable to water, as in Astacus, then the production of a dilute urine or the further reduction in normal blood osmoconcentration would have adaptive value (Shaw, 1959). Little is known about the mechanism employed by holeuryhaline species. In Eriocheir sinensis blood osmoconcentration was found to be high (A =—1.1° to —1.2°C) in individuals exposed to fresh water, and urine output low (3 to 5 ml/day KINNE: ADAPTATION 43 in a 60 gram individual). The urine is isosmotic or slightly hyperosmotic to the blood both in fresh water and sea water (Scholles, 1933; Schlieper, 1935), and chloride and ammonia losses are the same whether the excretory pores are open or closed. Metabolic rate remains practically constant in fresh water and sea water (Schwabe, 1933). Salt (NaCl) is actively absorbed from very dilute media by the gills (Schwabe, 1933; Koch, 1954). The existence of a K pump separate from the Na absorbing mechanism has been indi- cated, suggesting the presence of a similar mechanism as has been reported for larvae of insects, i.e. Chironomus and Aedes (Koch and Evans, 1956a), and various aspects of ionic exchange have been studied (Koch and Evans, 1956b,c). Eriocheir resembles the river crab Potamon in (a) maintaining a high blood osmoconcentra- tion in fresh water, (b) actively absorbing sodium and potassium, and (c) excreting small amounts of more or less blood-isos- motic urine. Structural genetic adaptations to salinity have been reported by several authors; however, only a few cases have been worked out in some detail. Most reports refer to size or structure of regulatory organs such as gills, gut and antennal glands. Pearse (1929a,b), for example, has shown that the gills of some estuarine or freshwater crabs are reduced in number or size, lead- ing to a decrease of total area through which exchange diffusion with the diluted medium occurs. And Schwabe (1933) has demonstrated that the nephridial canals are longer in the freshwater living Gam- marus pulex than in the brackish-marine Gammarus locusta; these canals have been shown by Peters (1935) to be more highly differentiated in the freshwater Astacus astacus than in the marine Homarus gam- marus. Genetic adaptations to life on land are well documented. The major route of land immigration appears to have been, and still is, from the sea via the littoral zone. Thus terrestrial Crustacea are mostly closer related to marine species than to estuarine or freshwater ones (Vandel, 1943). In spite of a variety of functional and structural adaptations to life on land, crustaceans have “never quite made it,’ and are— compared to insects—rather poorly equip- ped for life on dry land. Even the most successful terrestrial representatives, the isopods, cannot fully exploit the ecological opportunities offered by the terrestrial habitat and must avoid completely terres- trial conditions. In order to be able to oc- cupy their present niches on land, crus- taceans did not have to change very much; all successful immigrants were to some ex- tent pre-adapted. The most important pre- adaptations to land life are: (1) the hard exoskeleton, (2) jointed, strong extremities, (3) internal or quasi-internal fertilization, (4) the carapace-covered gill chambers of crabs, and (5) the egg protecting brood pouches of isopods and amphipods. Our present knowledge on terrestrial ad- aptations in Crustacea has recently been competently reviewed by Edney (1960). Edney comes to the conclusion that re- markably few profound changes have re- sulted from assuming the terrestrial way of life, and that even in land isopods, all devices for land life were present in the aquatic ancestor, or if not, are to some degree makeshift. Thus there is no effective protection against surface evaporation (no wax layer in epicuticle); respiration is still accomplished by gills, which have been only slightly modified, and the pseudo- tracheae are but short bunches of tubes; 44 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 the eggs are by no means cleidoic and must be carried by the parent; excretion is still predominantly ammonotelic; osmotic changes are tolerated rather than control- led; and high ambient temperatures are suffered only at the expense of increased transpiration. Vandel (1943; 1954) has arranged the families of land isopods on the basis of their fitness for survival in dry air in the following order: Ligiidae, Trichoniscidae, Oniscidae, Porcellionidae, Armadillidiidae. As a first approximation, this order seems also to be indicative of the increasing de- gree of morphological specialization to in- creasingly drier habitats. We may sum- marize the most obvious functional genetic adaptations to life on land as follows: (1) increased tolerance to extreme temperature, (2) reduced evaporation, (3) collection of water from small sources and absorption of water against a gradient, (4) active salt secretion, (5) reduction of the total nitrogen excreted per unit weight, and pos- sibly a general suppression of nitrogen metabolism, (6) reduction of total osmo- concentration of body fluids, (7) develop- ment or improvement of behavioral devices: for selection of suitable microhabitats, for feeding, for controlling extreme body tem- perature by exposure to conditions which cause different rates of evaporation (hu- midity, wind), and for orientation, sensory perception, walking and mating. Let us now finally consider a few struc- tural adaptations to terrestrial life. Bliss (1956) has shown that in Gecarcinus later- alis the pericardial sacs are adapted for water absorption, and premolt swelling can occur only when these organs are in contact with a moist substrate. Verhoeff (1917; 1920) demonstrated a series of water conducting channels in Oniscus and other advanced land isopods, which run along externally on both sides of the body from head to uropods, with cross channels on the pereion. This capillary system can pick up water from the substrate via the apposed uropods and conduct it to the respiratory surfaces on the pleopods. If no free water is available, regurgitated fluid may be conducted from the mouth to the pleopods to keep the respiratory surfaces moist (Edney, 1960). Some shore and land isopods are able to roll themselves into balls, e.g. Sphaeroma, Armadillidium and Tolypeutes. Such rolling protects the softer, appendage-carrying subsurface from poten- tial enemies and may also reduce water loss in dry habitats; it is often accompanied by a relocation of eggs from the external brood pouch into internal brood sacs (Han- sen, 1905; Kinne, 1954). The gill has undergone several modifi- cations in terrestrial species: (1) it has received additional support by sclerotiza- tion and ridges (van Raben, 1934); (2) its functional surfaces have been reduced (Ayers, 1938; Gray, 1953; Pearse, 1929a, b, 1950); and (3) its external surfaces are continuously moistened. Uca and Ocypode have special respiratory openings between their third and fourth legs (Edney, 1960). Land isopods use the same structures for respiration as their aquatic ancestors, the pleopods. These show, however, definite adaptations to air breathing in more terres- trial species (Modlinger, 1931): the semi- terrestrial Ligiidae and Trichoniscidae have still unmodified pleopods; Oniscidae have the exopodites of their pleopods hollowed out below; Porcellionidae and Armadilli- diidae possess hollow tuft-like invaginations known as pseudotracheae which facilitate respiration in air of reduced humidity. The most profound respiratory adap- KINNE: ADAPTATION 45 tations to land life are the vascular tufts and the vascularization of gill chamber walls in Ocypode, Coenobita and Birgus (Harms, 1932; van Raben, 1934). Both are developments de novo. With respect to the topic of this confer- ence—Evolution of Crustacea—the facts and brief considerations that I have pre- sented here demonstrate two things: (1) They illustrate important relation- ships between environmental factors and organismic functions and structures, which promise to provide useful tools in an at- tempt to analyse and understand the forces at work and the biological mechanisms in- volved in evolutionary processes. (2) They show that our present knowl- edge of nongenetic and genetic adaptations to temperature, salinity and to life on land cannot explain the substantial macroevo- lutionary differences of crustacean body plans which we witness today. Macroevolutionary changes are presum- ably based on adaptations to conditions that were effective before the first crusta- ceans began to leave their oceanic home. They may represent adaptations, for ex- ample, to the different feeding habits of hunters and lurers, substrate- and filter feeders and to differences in modes of loco- motion and habits in pelagic, benthic, ses- sile, vagile, epizoic and endopsammic forms (e.g. Kaestner, 1959). Some macroevolu- tionary differences in structure may turn out to be based on random variations with more or less neutral adaptive value, or to be incidental by-products of primarily functional adaptations. It will certainly be a long and difficult task to bring to light the complex and manifold relationships between environ- ment and organism in such a large and heterogeneous group as the Crustacea. And it is to be hoped that the modern studies of nongenetic and genetic adap- tation—both at the functional and at the structural level—will produce useful, com- plementary means for opening up addi- tional avenues for a new, comprehensive and dynamic approach to an old, funda- mental problem of biology, the problem of evolution. SUMMARY A brief review is presented of our pres- ent knowledge of nongenetic and genetic adaptation in Crustacea (mostly macro- crustaceans) to temperature, salinity and life on land. Some of the adaptations re- ported are universal among animals, for example, functional adaptations to tem- perature involving shiftings of lethal lim- its, changes in activity or in metabolic rates. Other adaptations appear to be re- lated to specific crustacean features, i.e. their primarily aquatic way of life, respi- ration through gills, urine formation in antennal or maxillary glands, exoskeleton and body shape—for example, the differ- ent mechanisms involved in osmoregula- tion, respiratory adaptations and adapta- tions to life on land. Important relations are illustrated, between environment and function or structure, which promise to provide useful tools in an attempt to ana- lyse and understand the forces at work and the biological mechanisms involved in evolutionary processes. Macroevolutionary differences in structure cannot be ex- plained on the basis of the outlined adap- tations to temperature, salinity and land life. They are assumed to be based on dif- ferences in feeding, respiration and loco- motion, to represent incidental by-products of primarily functional adaptations, or random variations with rather neutral adaptive value. 46 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 REFERENCES ApotpH, E. F. 1925. Some physiological distinc- tions between freshwater and marine orga- nisms. Biol. Bull. 48:327-335. Ayers, J. C. 1938. Relationship of habitat to oxygen consumption by certain estuarine crabs. Ecology 19:523-527. Barnes, H. anp M. Barnes. 1958. The rate of development of Balanus balanoides (L.) larvae. Limnol. and Oceanogr. 3:29-32. Beapte, L. C. 1943. Osmotic regulation and the faunas of inland waters. Biol. Rev. 18:172-183. Brave, L. C. anp J. B. Cracc. 1940a. Studies on adaptation to salinity in Gammarus spp. I. Regulation of blood and tissues and the prob- lem of adaptation to fresh water. J. Exp. Biol. 17:153-163. 1940b. Osmotic regulation in fresh- water animals. Nature, London 146:588. Butss, D. E. 1956. Neurosecretion and the con- trol of growth in a decapod crustacean. In: Wingstrand, K. G. (ed.), Bertil Hanstrom. Zoological papers in honour of his sixty-fifth birthday, November 20th, 1956. Lund, Zoo- logical Institute, Sweden, pp. 56-75. BROEKEMA, M. M. M. 1941. Seasonal movements and the osmotic behaviour of the shrimp, Crangon crangon L. Arch. Néerl. Zool. 6:1- 100. Brooks, J. L. 1946. Cyclomorphosis in Daphnia. I. An analysis of D. retrocurva and D. galeata. Ecol. Monographs 16:409-447. 1947. Turbulence as an_ environ- mental determinant of relative growth in Daphnia. Proc. Natl. Acad. Sci., U.S. 33:141- 148. 1957. The species problem in fresh- water animals. Jn: Mayr, E. (ed.), The Species Problem. Am. Assoc. Adv. Sci., Washington, D.C., pp. 81-123. Brown, F. A., Jr. 1934. The chemical nature of the pigments and the transformations respon- sible for color changes in Palaemonetes. Biol. Bull. 67:365-380. CrocHan, P. C. 1958a. The survival of Artemia salina (L.) in various media. J. Exp. Biol. 35: 213-218. 1958b. The osmotic and ionic regu- lation of Artemia salina (L.). J. Exp. Biol. 35: 219-233. 1958c. The mechanism of osmotic regulation in Artemia salina (L.): the physi- ology of the branchiae. J. Exp. Biol. 35:234- 242. 1958d. The mechanism of osmotic regulation in Artemia salina (L.): the physi- ology of the gut. J. Exp. Biol. 35:243-249. 1958e. Ionic fluxes in Artemia salina (L.). J. Exp. Biol. 35:425-436. Daxiy, W. J. ano E. Epmonps. 1931. The regu- lation of the salt contents of the blood of aquatic animals, and the problem of the per- meability of the bounding membranes of aquatic invertebrates. Australian J. Exp. Biol. and Med. Sci. 8:169-187. Deunet, P. A. 1960. Effect of temperature and salinity on the oxygen consumption of two intertidal crabs. Biol. Bull. 118:215-249. Drityon-Courtots, A. 1934. De la régulation de la composition minérale de l’hémolymphe des Crustacés. Ann. physiol. physicochim. biol. 10: 377-414. DucHATEAU, G. AND M. Frorxin. 1955. Influence de la température sur l’état stationnaire du pool des acides aminés non protéiques des muscles d’Eriocheir sinensis Milne-Edwards. Arch. intern. physiol. biochim. 63:213-221. Duvat, M. 1925. Récherches physico-chimiques et physiologiques sur le milieu intérieur des ani- maux aquatiques. Modifications sous |’influence du milieu extérieur. Ann. Inst. Océanogr., Monaco, ns. 2:233-403. Epmonps, E. 1935. The relations between the in- ternal fluid of marine invertebrates and the water of the environment, with special reference to Australian Crustacea. Proc. Linn. Soc. New South Wales, 60:233-247. Epney, E. B. 1951a. The evaporation of water from woodlice and the millipede Glomeris. J. Exp. Biol. 28:91-115. 1951b. The body temperature of woodlice. J. Exp. Biol. 28:271-280. 1960. Terrestrial adaptations. Jn: Waterman, T. H. (ed.), The Physiology of Crustacea. New York and London, Academic Press, 1:367-393. FiLemister, L. J. 1958. Salt and water anatomy, constancy and regulation in related crabs from marine and terrestrial habitats. Biol. Bull. 115: 180-200. FLeMister, L. J. AnD S. C. FLemister. 1951. Chloride ion regulation and oxygen consump- tion in the crab Ocypode albicans (Bosq.). Biol. Bull. 101:259-273. KINNE: ADAPTATION 47 Frorxkin, M. 1960. Ecology and metabolism. Jn: Waterman, T. H. (ed.), The Physiology of Crustacea. New York and London, Academic Press, 1:395-410. Frtcer, H. 1959. Zum Problem der Osmoregu- lation im tropischen Brackwasser. Naturwiss. 46(6):213. FritzscHe, H. 1917. Studien itiber Schwankungen des osmotischen Druckes der KOrperfliissigkeit bei Daphnia magna. Intern. Rev. Hydrobiol. 8: 22-80, 125-203. Grarncer, J. N. R. 1956. Effects of changes of temperature on the respiration of certain Crus- tacea. Nature, London 178:930-931. 1958. First stages in the adaptation of poikilotherms to temperature change. In: Prosser, C. L. (ed.), Physiological Adaptation. Am. Physiol. Soc., Washington, D.C., pp. 79-91. Gray, I. E. 1953. Comparative study of gill area in crabs. Anat. Record 117:567-568. Green, J. W., M. Harscu, L. Barr anv C. L. Prosser. 1959. The regulation of water and salt by the fiddler crabs, Uca pugnax and Uca pugi- lator. Biol. Bull. 116:76-87. Gross, W. J. 1955. Aspects of osmotic regulation in crabs showing the terrestrial habit. Am. Naturalist 89:205-222. 1957a. An analysis of response to osmotic stress in selected decapod Crustacea. Biol. Bull. 112:43-62. 1957b. A behavioral mechanism for osmotic regulation in a semiterrestrial crab. Biol. Bull. 113:268-274. Gross, W, J. AND P. V. Hortanp. 1960. Water and ionic regulation in a terrestrial hermit crab. Physiol. Zool. 33:21-28. Gross, W. J. anp L. A. MarsHatr. 1960. The in- fluence of salinity on the magnesium and water fluxes of a crab. Biol. Bull. 119:440-453. Hansen, H. J. 1905. On the propagation, struc- ture and classification of the family Sphaero- midae. Quart. J. Micr. Sci. 49:69-135. Harms, J. W. 1932. Die Realisation von Genen und die consecutive Adaptation. II. Birgus latro L. als Landkrebs und seine Beziehungen zu den Coenobiten. Z. wiss. Zool. 140:167-290. HERRMANN, F. 1931. Uber den Wasserhaushalt des Flusskrebses (Potamobius astacus Leach). Z. vergleich. Physiol. 14:479-524. Hrpacek, J. 1959. Circulation of water as a main factor influencing the development of helmets in Daphnia cucullata Sars. Hydrobiologia 13: 170-185. Jones, L. L. 1941. Osmotic regulation in several crabs of the Pacific Coast of North America. J. Cell. Comp. Physiol. 18:79-92. Karstner, A. 1959. Lehrbuch der Speziellen Zool- ogie, Teil I: Wirbellose (4. Lieferung). Jena, VEB Gustav Fischer, pp. 659-979. Kinng, O. 1952. Zur Biologie und Physiologie von Gammarus duebeni Lillj., V: Untersuchun- gen uber Blutkonzentration, Herzfrequenz und Atmung. Kieler Meeresforsch. 9:134-150. 1953a. Zur Biologie und Physiologie von Gammarus duebeni Lillj., I. Z. wiss. Zool. 157:427-491. 1953b. Zur Biologie und Physiologie von Gammarus duebeni Lillj., II. Uber die Hautungsfrequenz, ihre Abhangigkeit von Tem- peratur und Salzgehalt, sowie iiber ihr Ver- halten bei isoliert gehaltenen und amputierten Versuchstieren. Zool. Jb. (Physiol.) 64:183-206. 1954. Eidonomie, Anatomie und Le- benszyklus von Sphaeroma hookeri Leach (Iso- poda). Kieler Meeresforsch. 10:100-120. 1963. Crustacea. In: Handbook of Physiology; Adaptation to the Environment. Amer. Physiol. Soc. (in press). KiyynzE, O. ann H.-W. RorrHauwe. 1952. Bi- ologische Beobachtungen und Untersuchungen uber die Blutkonzentration an Heteropanope tridentatus Maitland (Decapoda). Kieler Meeres- forsch. 8:212-217. Koc, H. J. 1954. Cholinesterase and active trans- port of sodium chloride through the isolated gills of the crab Eriocheir sinensis (M. Edw.). In: Kitching, J. A. (ed.), Recent Developments in Cell Physiology. New York, Academic Press, pp. 15-27. Kocu, H. J. anp J. Evans. 1956a. On the influence of lithium on the uptake of sodium and potas- sium by the crab Eriocheir sinensis (M. Edw.) Mededel. Kon. Vlaamse Acad. KI. Wet. 18(6): 3-10. 1956b. On the absorption of sodium from dilute solutions by the crab Eriocheir sinensis (M. Edw.). Mededel. Kon. Vlaamse Acad. Kl. Wet. 18(7) :3-15. . 1956c. Influence of a basic dye, thio- nine, on the absorption of sodium by the crab Eriocheir sinensis (M. Edw.). Mededel. Kon. Vlaamse Acad. KI. Wet. 18(8):3-11. 48 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, Kroc, J. 1954. The influence of seasonal environ- mental changes upon the metabolism, lethal temperature and rate of heart beat of Gam- marus limnaeus (Smith) taken from an Alas- kan lake. Biol. Bull. 107:397-410. Krocu, A. 1919. The diffusion of gases through animal tissues. J. Physiol., London 52:391-408. 1939. Osmotic regulation in aquatic animals. London and New York, Cambridge Univ. Press, 242 pp. Lieper, U. 1951. Der Stand der Zyklomorpho- seforschung. Naturwissenschaften 38:39-44. LienemMaAnn, L. J. 1938. The green glands as a mechanism for osmotic and ionic regulation in the crayfish (Cambarus clarkii Girard). J. Cell. Comp. Physiol. 11:149-161. Locxwoop, A. P. M. 1959. The osmotic and ionic regulation of Asellus aquaticus (L.) J. Exp. Biol. 36:546-555. 1961. The urine of Gammarus due- beni and G. pulex. J. Exp. Biol. 38:647-658. Lorts, B. 1956. The effects of salinity changes on the respiratory rate of the prawn, Palaemonetes varians (Leach). J. Exp. Biol. 33:730-736. McLeersrt, D. W. 1956. Effects of temperature, salinity and oxygen on the survival of the American lobster. J. Fish. Res. Bd. Canada, 13:247-272. Matoeur, N. S. R. 1938. Echanges d’eau_ et d’électrolytes chez un pagure (Pagurus longi- carpus). Arch. int. Physiol. 47:1-23. (Quoted from v. Buddenbrock, 1956.) Marcater, R. 1955. Temperatura, dimensiones y evoluciodn. Publs. Inst. Biol. Apl. (Barcelona) 19: 13-94. Maturant, K. 1960. Studies on the temperature and salinity resistance of Tigriopus japonicus. I. Changes in heat resistance in relation to acclimation temperatures of Tigriopus japonicus reared at 20°C. Physiol. and Ecol. 9:35-38. Mayr, E. 1960. Chairman’s introduction to the symposium on adaptive evolution. Proc. XIIth Internat. Ornithological Congr., Helsinki, 1958, pp. 495-498. Meryertnc, M. P. D. 1960. Herzfrequenz und Herzschlagzahlen zwischen Hautung und Ei- ablage bei Cladoceren. Z. wiss. Zool. 164: 127-142. Moptirncer, G. 1931. Morphologie der Respira- tionsorgane der Landisopoden. Studia zool. 2: 25-79. Moore, W. G. 1955. Observations on heat death 1963 in the fairy shrimp, Streptocephalus seali, Proc. Louisiana Acad. Sci. 18:5-12. Nace, H. 1934. Die Aufgaben der Exkretionsor- gane und der Kiemen bei der Osmoregulation von Carcinus maenas. Z. vergleich. Physiol. 21: 468-491. Ostwa Lp, W. 1904. Experimentelle Untersuchungen iiber den Saisonpolymorphismus bei Daphnien. Arch. Entwicklungsmechanik 18:415-451. PanrkkKar, N. K. 1939. Osmotic behavior of Pala- emonetes varians (Leach). Nature, London 144: 866-867. 1940a. Osmotic properties of the common prawn. Nature, London 145:108. 1940b. Influence of temperature on osmotic behaviour of some Crustacea and its bearing on problems of animal distribution. Nature, London 146:366-367. 1941a. Osmoregulation in some pala- emonid prawns. J. Marine Biol. Assoc. U.K. 25:317-359. 1941b. Osmotic behavior of the fairy shrimp Chirocephalus diaphanus Prévost. J. Exp. Biol. 18:110-114. 1950. Physiological aspects of adap- tation to estuarine conditions. Proc. Indo- Pacific Fisheries Council (Australia), 2nd meet., pp. 168-175. Parry, G. 1953. Osmotic and ionic regulation in the isopod crustacean Ligia oceanica. J. Exp. Biol. 30:567-574. 1955. Urine production by the an- tennal glands of Palaemonetes varians Leach. J. Exp. Biol. 32:408-422. 1957. Osmoregulation in some fresh- water prawns. J. Exp. Biol. 34:417-423. 1961a. Osmoregulation of the fresh- water prawn Palaemonetes antennarius. Mem. Ist. Ital. Idrobiol. 13:139-149. 1961b. Chloride regulation in Triops. Nature, London 192:468-469. Pearse, A. S. 1929a. Observations on certain lit- toral and terrestrial animals at Tortugas, Florida, with special reference to migrations from marine to terrestrial habitats. Papers Tortugas Laboratory 26 (Carnegie Inst., Wash- ington, Publ. no. 391): 205-223. 1929b. The ecology of certain estu- arine crabs at Beaufort, N.C. Jour. Elisha Mitchell Sci. Soc. 44:230-237. 1950. The emigrations of animals from the sea. Dryden, N. Y., Sherwood Press, 210 pp. KINNE: ADAPTATION 49 Pearse, A. S. AND G. Gunter. 1957. Salinity. In: Hedgpeth, J. W. (ed.), Ecology vol. 1, Treatise on marine ecology and paleoecology. Geol. Soc. America, Mem. 67:129-157. Peters, H. 1935. Uber den Einfluss des Salzge- haltes im Aussenmedium auf den Bau und die Funktion der Exkretionsorgane dekapoder Crustaceen. Z. Morph. u. Okol. Tiere 30: 355-381. Ports, W. T. W. 1954. The energetics of osmotic regulation in brackish- and freshwater ani- mals. J. Exp. Biol. 31:618-630. Precut, H. 1949. Die Temperaturabhingigkeit von Lebensprozessen. Z. Naturforsch. 4(b) :26. 1958. Concepts of the temperature adaptation of unchanging reaction systems of cold-blooded animals. Jn: Prosser, C. L. (ed.), Physiological adaptation. Am. Physiol. Soc. Washington, D.C., pp. 50-78. Prosser, C. L. 1958. General summary: The nature of physiological adaptation. Im: Prosser, C. L. (ed.), Physiological adaptation. Am. Physiol. Soc., Washington, D.C., pp. 167-180. Prosser, C. L. anp F. A. Brown, jr. 1961. Com- parative Animal Physiology. 2nd ed. Philadel- phia and London, W. B. Saunders Co., 688 pp. Prosser, C. L., J. W. GREEN AND T. J. CHow. 1955. Ionic and osmotic concentrations in blood and urine of Pachygrapsus crassipes acclimated to different salinities. Biol. Bull. 109:99-107. Rarry, A. 1934. Influence des variations de salinité sur l’intensité respiratoire de la telphuse et de l’écrevisse. C. R. Acad. Sci., Paris 198: 680-681. Rao, K. P. 1958. Oxygen consumption as a func- tion of size and salinity in Metapenaeus mono- ceros Fab. from marine and_brackish-water environments. J. Exp. Biol. 35:307-313. Roperts, J. L. 1957a. Thermal acclimation of metabolism in the crab Pachygrapsus crassipes Randall. I. The influence of body size, star- vation, and molting. Physiol. Zool. 30:232-242. . 1957b. Thermal acclimation of metab- olism in the crab Pachygrapsus crassipes Ran- dall. II. Mechanisms and the influence of season and latitude. Physiol. Zoél. 30:242-255. Rosertson, J. D. 1949. Ionic regulation in some marine invertebrates. J. Exp. Biol. 26:182-200. . 1960. Osmotic and ionic regulation. In: Waterman, T. H. (ed.), The Physiology of Crustacea. New York and London, Academic Press, 1:317-340. ScHerBAKorr, A. P. 1935. Uber den Sauerstoffver- brauch von einigen Planktoncrustaceen. Arbei- ten der limnol. Station Kossino der Hydro- meteorolog. Administration der USSR 19: 67-89. SCHLIEPER, C. 1929. Uber die Einwirkung niederer Salzkonzentrationen auf marine Or- ganismen. Z. vergleich. Physiol. 9:478-514. 1935. Neuere Ergebnisse und Prob- leme aus dem Gebiet der Osmoregulation was- serlebender Tiere. Biol. Rev. 10:334-360. SCHLIEPER, C. AND F. HerrMANN. 1930. Bezie- hungen zwischen Bau und Funktion bei den Exkretionsorganen dekapoder Crustaceen. Zool. Jahrb. (Anat.) 52:624-630. SCHOLANDER, P. F., W. Fracc, V. WALTERS AND L. Irvinc. 1953. Climatic adaptation in arctic and tropical poikilotherms. Physiol. Zool. 26: 67-92. ScHoLtes, W. 1933. Uber die Mineralregulation wasserlebender Evertebraten. Z. vergl. Physiol. 19:522-554. ScHwase, E. 1933. Uber die Osmoregulation verschiedener Krebse (Malacostracen). Z. vergl. Physiol. 19:183-236. SHAW, J. 1959. Salt and water balance in the east African freshwater crab, Potamon_ nilo- ticus (M.Edw.). J. Exp. Biol. 36:157-176. SHAW, J. anp D. W. Surcrirre. 1961. Studies on sodium balance in Gammarus duebeni Lilljeborg and G. pulex pulex (L.). J. Exp. Biol. 38:1-15. Smrpson, G. G. 1958. Behavior and Evolution. In: Roe, A. and G. G. Simpson, Behavior and Evolution. New Haven, Yale University Press, pp. 507-535. Simpson, G. G., C. S. PitrenpricH Ano L. H. Tirrany. 1957. Life. An introduction to biology. New York, Harcourt, Brace and Co., 845 pp. SoutHwarp, A. J. 1958. Notes on the tempera- ture tolerances of some intertidal animals in relation to environmental geographical distribution. Assoc. U. K. 37:49-66. Spoor, W. A. 1955. Loss and gain of heat-toler- ance by the crayfish. Biol. Bull. 108:77-87. STanier, R. Y., M. Doupororr, anp E. A. ADEL- BERG. 1957. The Microbial World. Englewood Cliffs, N.J., Prentice-Hall, Inc., 682 pp. TaKepA, N. 1954. Thermal adaptation in the marine copepod, Tigriopus japonicus Mori. Physiol. Ecol. (Japan) 6:49-54. temperatures and J. Marine Biol. 50 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 VANDEL, A. 1943. Essai sur lorigine, l’évolution et la classification des Oniscoidea (Isopodes terrestres). Bull. biol. France et Belg., Suppl. 30:1-136. . 1954. La répartition géographique des Oniscoidea (Crustacés Isopodes terrestres). Bull. biol. France et Belg. 74:221-272. van Rasen, K. 1934. Verdnderungen in dem Kiemendeckel und in den Kiemen einiger Brachyuren (Decapoden) im Verlauf der An- passung an die Feuchtluftatmung. Zool. A145:425-461. VERHOEFF, K. W. 1917. Zur Kenntnis der Ent- wicklung der Trachealsysteme und der Unter- gattungen von Porcellio und Tracheoniscus. Sitzber. Ges. naturforsch. Freunde Berlin, 1917: 195-223. 1920. Uber die Atmung der Landas- seln, zugleich ein Beitrag zur Kenntnis der Entstehung der Landtiere. Z. wiss. Zool. 118: 365-447. VERNBERG, F. J. 1959a. Studies on the physio- logical variation between tropical and tem- perate zone fiddler crabs of the genus Uca. I. Thermal death limits. Ecology 40:589-593. 1959b. Studies on the physiological Z. wiss. variation between tropical and temperate zone fiddler crabs of the genus Uca. II. Oxygen consumption of whole organisms. Biol. Bull. 117:163-184. 1959c. Studies on the physiological variation between tropical and temperate zone fiddler crabs of the genus Uca. III. The in- fluence of temperature acclimation on oxygen consumption of whole organisms. Biol. Bull. 117: 582-593. WESENBERG-LuND, C. 1900. Von dem Abhingig- keitsverhaltnis zwischen dem Bau der Plank- tonorganismen und dem spezifischen Gewicht des Siisswassers. Biol. Zbl. 20:606-619, 644- 656. Wipmann, E. 1935. Osmoregulation bei ein- heimischen Wasser- und Feuchtluft-Crustaceen. Z. wiss. Zool. 147:132-169. WINGFIELD, C. A. 1939. The activity and metab- olism oi poikilothermal animals in different latitudes. IV. Proc. zool. Soc. London A109: 103-108. Wottereck, R. 1913. Uber Funktion, Herkunft und Entstehungsursachen der sogen. “Schwebe- Fortsatze” pelagischer Cladoceren. Zoologica, Stuttgart 26:475-550. PHYLOGENY AND EVOLUTION OF CRUSTACEA Museum or CoMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 IV On the Relationship of Dromiacea, Tymolinae and Raninidae to the Brachyura By Isabella Gordon British Museum (Natural History), London In 1950 I described in some detail the spermathecae of females belonging to species of the families Dromiidae and Thelxiopidae (— Homolidae), and showed how the male intromittent organs were adapted to suit these two kinds of sper- mathecae. I did not at that time discuss the relationship of the Dromiacea to the brachyuran crabs, because it was my in- tention to continue the study of other crab-like forms that are often referred to the Brachyura. This work, however, was delayed, partly because of other commit- ments, partly for lack of certain essential material. As mentioned by Balss (1957, p. 1616), I found in the Raninidae a special kind of unpaired female sperma- thecal opening or pit, quite different from the paired ones of Dromiidae or Thelxiopi- dae. Quite recently I obtained some material of the so-called Tymolinae which enabled me to examine the sternal furrows in the female, structures that are absent in females of the supposedly related sub- family Dorippinae. I hope ere long to publish a detailed account of these sper- mathecae; here I can only deal with some of them briefly. TYMOLINAE In Figure 10A the thoracic sternum of a female of Tymolus japonicus Stimpson is represented, tilted so that the almost vertical posterior sternites are seen. In true ventral aspect, most of the sternum behind the ridges (7) on sternites 5 is not visible. In the tilted position, however, the sternal furrows are obvious; they resemble rather closely those of the family Dromiidae. But the large spermatophores, or spermatophoral masses, visible through the thin sternal wall, are situated near the spermathecal openings (Fig. 10A, f, o and s). As in Dromia, these sternal fur- rows in the female are just the 7/8 sutures carried forwards and anteriorly modified; the papillae on which the spermathecal openings are placed are situated on a level with the sockets for peraeopods 2 and thus well in front of the genital openings on the coxae of peraeopods 3. In this tilted position, the coxae of peraeopods 4 and 5 are also visible (Fig. 10A, p4 and p5). I have been able to confirm the presence of very similar sternal furrows in the now fragmentary female syntype of Xeinostoma eucheir Stebbing. In the genus 52 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 Fic. 10. A. Tymolus japonicus Stimpson, 2 Cymonomus, on the other hand, the tho- racic sternum is bent much less abruptly than in Tymolus and the large oval areas that indicate the spermathecal openings are visible in ventral aspect. In Figure 10B, however, the specimen is tilted so that sternites 7 and 8 are seen more com- pletely. Here the 6/7 suture runs across the full width of the thorax; thus the 7/8 suture is shorter than in Tymolus, ending opposite the coxa of peraeopod 3, but still in front of the genital opening on that coxa. Only the distal half of the 7/8 suture forms the spermatheca, the posi- tion of which is indicated by a distinct bulge on sternite 8; the sutural margin of sternite 8 overlaps that of sternite 7, as shown in Figure 10B. The spindle shaped area (0) presumably belongs to sternite 7 and the entrance to the sper- matheca is along its posterior edge. The from Manazuru, Japan; thoracic sternites tilted to show the almost vertical posterior ones and the sternal furrow (f) ending distally in the sper- mathecal opening (0). B. Cymonomus granulatus Norman, left thoracic sternites of @, slightly tilted, to show oval area (0) at the spermathecal opening. Abbreviations: 1-8, thoracic sternites; g, genital opening on coxa of peraeopod 3; 7, ridge on sternite 5; p1-p5, coxae of peraeopods 1 to 5; s2, socket for peraeopod 2; s, indicates position of the spermatophore. position of the relatively small sper- matophoral mass is indicated by a broken line (s, Fig. 10B). This type of sperma- theca is reminiscent of that found in the Thelxiopidae. There seem to be, within the so-called Tymolinae, two different kinds of sper- mathecae. When the intromittent organs of the male are examined (Figs. 11A and 11B), they differ chiefly as regards the apex of pleopod 2. In Cymonomus and in Tymolus there is a large penial projection on the coxa of peraeopod 5; the terminal segment of pleopod 1 (= the endopod) is a hollow, folded, leaf-like structure; pleopod 2 is large in both. But the apex of the last segment (= the endopod) re- sembles a hypodermic needle in Tymolus (2a, Fig. 11A); in Cymonomus it is like the sole of a boot (apex of 2, Fig. 11B). In Tymolus only the needle-like tip of GORDON: CRAB RELATIONSHIPS 53 B mm. ——) Fic. 11. Intromittent organs of ¢ granulatus Norman. Abbreviations: c5, coxa of peraeopod 5; », penis. pleopod 2 probably enters the small open- ing of the female spermatheca and the seminal fluid is probably poured into the spermathecal sac. In Cymonomus, pleo- pod 2 probably acts as a sort of piston, the “sole” pushing aside the spindle-like flap to place a spermatophore in the spermathecal pocket. At least that is how I interpret these structures. This fits in nicely with Ihle’s subdivision of the “Tymolinae” into two tribes Cyclodorippae and Cymonomae (Ihle, 1916, p. 154). Since Cyclodorippe is a synonym of Ty- molus the tribe should be called Tymolae. RANINIDAE In the Raninidae the thoracic sternites are specially modified, presumably in con- nection with their burrowing habits. Figure 12A represents the thoracic sternites of a very immature female of Ranina ranina (L.), carapace length 63mm. Owing to the fact that the sternum bends abruptly of: A. Tymolus japonicus Stimpson and B. Cymonomus 1, pleopod 1; 2, pleopod 2; 2a, apical part of pleopod 2; upwards, at an angle of nearly 90°, in the region of the genital or sixth somite, the posterior part is foreshortened. Sternites 8 are not visible because the last pair of peraeopods are dorsal and somewhat anterior to peraeopods 4. The separate figure of sternites 7, at the same scale, gives a better idea of their length al- though the portion behind the articulation of peraeopods 4 is still foreshortened. Even at this immature stage sternites 7 can be distinguished from those of a male because of the median depression in the anterior half, which indicates an incipient spermathecal opening. In the adult female this single spermathecal pit is very con- spicuous (Fig. 12B). It is situated in the anterior half of sternites 7 behind the small genital openings on the coxae of peraeopods 3. The single spermathecal opening is even more conspicuous in the much smaller species Notopoides latus Henderson (Fig. 54 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 Fic. 12. Ranina ranina (L.), [3 Q. A. Thoracic sternum of a very immature specimen (carapace length = 63 mm.). B. Thoracic sternites 7 and 8 of an adult (carapace length = 138 mm.), to show the large spermathecal opening (so). Abbreviations: 1-8, thoracic sternites 1 to 8; c4, c5, coxae of peraeopods 4 and 5; p2-p4, sockets for peraeopods 2 to 4; g, genital opening on third coxa. 13A). On dissection, I found that the spermathecal pit leads obliquely back- wards and inwards to a_spermathecal pocket (s) enclosed in part of the endo- phragmal skeleton. The endophragmal sys- tem of a raninid differs strikingly from that of a typical brachyuran crab such as Maja squinado (Herbst)—compare Figures 13B and 14B. One of the pecu- liarities of Notopoides is the very high median apodeme arising from the median suture of sternites 6 and 7, respectively. In a male of Notopoides the apodeme arising from sternites 7 is similar to that arising from sternites 6 (see Bourne, 1922, pl. 4, fig. 9 of Ranina); in the female, however, part of apodeme 7m is modified to form the spermathecal pocket and the passage leading to it (s, so, Fig. 13B). The male intromittent organs are also modified in a special way, pleopods 1 be- ing fused basally and the free portions being relatively slender closely apposed so that both can enter the single spermathecal opening. and GORDON: CRAB RELATIONSHIPS 55 Fic. 13. 5 mm. Notopoides latus Henderson, @ syntype. A. Posterior almost vertical portion of thoracic sternum, to show the large spermathecal opening (so) on sternites 7. B. Posterior part of endo- phragmal system of same, in median aspect. Abbreviations: g, genital opening of coxa on peraeo- pod 3; 64, p5, sockets for peraeopods 4 and 5; 5m-7m, median apodeme from sternites 5 to 7; s, spermathecal pocket. BRACHYURA Maja squinado (Herbst): This com- mon spider-crab may be taken as an example of a typical brachyuran crab. The endophragmal system has been de- scribed and figured by Drach (1939, pp. 369-373). The thoracic sternites of the female are much broader than those of the male and all are visible in ventral aspect (Fig. 14A). The genital opening is sternal, situated on sternite 6 just behind the 5/6 suture line. The posterior part of the endophragmal system of the same female is represented in median aspect in Figure 14B. The sella turcica or turkish saddle (ts) with its wing-like extension (w), and the way in which the five posterior endo- pleurites above and the four posterior endosternites below (4/5-7/8) are all conjoined, give great strength to the whole. The vagina lies in the space between endo- sternites 5/6 and 6/7 but it and its sper- mathecal portion (s) are quite free, not incorporated in any part of the endo- skeleton. DISCUSSION The division of one family Dorippidae into “Dorippidae sternitremen” (Dorip- pinae) and “Dorippidae peditremen”’ (Tymolinae) has always seemed odd to me. Because, if the Decapoda are taken as a whole, all the Natantia and many of the Reptantia are “Hneditremen” (with genital opening coxal in the female). It is only in the Brachyura 1 The classification in Balss (1957) may be taken as the latest, though Balss was ultra con- servative in places. 56 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 A 2\em. Fic. 14. Maja squinado (Herbst), ing (g) on sixth sternite. Abbreviations: B @. A. Right thoracic sternites 1 to 8, with genital open- B. Posterior part of endophragmal system of same, in median aspect. 4/5-7/8, endosternites 4 to 7 arising from the sutures between the sternites in- dicated; s, spermathecal part of vagina; ts, sella turcica; w, wing of sella turcica. that a few families and subfamilies are “peditremen,” all the rest being ‘“sterni- tremen,”’ with the female genital opening sternal. The “peditremen” forms among the Brachyura are the Dromiacea, the Raninidae and the Tymolinae, and, as I have shown, it is these that possess various kinds of spermathecae in the formation of which the endophragmal system is in- volved. To me it seems logical to exclude all these from the true Brachyura, restrict- ing the term to the vast majority of crabs with the female genital openings sternal (the “Decapoda sternitremen”). In 1922 Bourne made a careful study of various raninids and separated them from the Oxystomata, placing them in the new superfamily Gymnopleura. Previous work- ers like Boas, and Milne Edwards and Bouvier also thought the raninids were not related to the Oxystomata. The dis- covery of a special, unpaired spermathecal pit leading to a spermathecal pocket in the endophragmal system, together with the specialisation of the male pleopods, fully supports Bourne’s conclusion and justifies the term Gymnopleura. I do not know why Bourne missed the conspicuous spermathecal pit, but the abdomen and the pleopods have to be turned right back in order to see it. He happened to dis- sect a male Ranina for the endophragmal system (Bourne, 1922, pl. 4, fig. 9). Certainly the so-called Tymolinae with sternal furrows and coxal genital pores should not be placed in the same family as the dorippids without sternal furrows and with the genital openings of the female sternal. The tymolids should at least be placed in a separate family, Tymolidae; GORDON: CRAB their true place in the classification is with or near the Dromiacea. All the females of Tymolidae that I have examined have been in poor condition and the endo- phragmal system is either poorly or not at all calcified, at least in the region of the spermathecae. The internal structure of the spermathecae has been studied as far as the imperfect condition of the material allowed, and will be described more fully elsewhere. If the “Tymolinae” are given family rank, then Ihle’s two tribes become the subfamilies Tymolinae and Cymonom- inae. If the term Brachyura is restricted to the eminently successful “Decapoda sternitremen” and the Dromiacea, Gymno- pleura and Tymolidae placed where they belong in the ‘Decapoda peditremen,” the term Anomura will probably have to be abandoned. At present it is in Monod’s words simply a “rag bag” into which is thrown an assortment of superfamilies and families that do not fit in anywhere else. Our knowledge of the endophragmal system of the Decapoda is still fragmen- tary and imperfect. Drach (1950) has pointed out that it is present in the Decapoda Natantia, contrary to what is generally supposed. He also thinks that RELATIONSHIPS 57 the Eryonidea should be excluded from the Palinura because their endophragmal sys- tem proves to be of a different kind, much more primitive. He also says that the homolids (Thelxiopidae), homolodromiids and ‘‘certain raninids” differ as regards endophragmal system from the rest of the Brachyura. Much work has still to be done before we can arrive at a satisfactory clas- sification of the Decapoda. REFERENCES Barss, Heryricu. 1957. Decapoda. Jn Bronn’s Klassen und Ordnungen des Tierreichs, 5, Abt. I, Buch 7, Lief. 12: 1505-1672, figs. 1131- 1199. Bourne, G. C. 1922. The Raninidae: a study in carcinology. J. Linn. Soc. Zool. 35:25-79, 6 pls. DracH, Prerre. 1939. Mue et cycle d’intermue chez les Crustacés Décapodes. Ann. Inst. Océanogr. Paris (n.s.) 19:103-391, pls. II-VII. . 1950. Les étapes évolutives de l’endos- quelette chez les Crustacés Décapodes. C. R. Acad. Sci. Paris, 281:1563-65. Gorpon, ISABELLA. 1950. Crustacea Dromiacea. John Murray Exped. Sci. Rep. 9(3): 201-253, 1 pl., 26 figs. Inte, J. E. W. 1916. Die Decapoda Brachyura der Siboga Expedition. II. Oxystomata, Dorip- pidae. Siboga Expeditie, Uitkomsten 39b (livr. 78): 97-158, figs. 39-77. § +, * wine ig Ohiie HA &. PHYLOGENY AND EVOLUTION OF CRUSTACEA Museum oF CoMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 V The Pericardial Sacs of Terrestrial Brachyura By Dorothy E. Bliss The American Museum of Natural History New York, N.Y. and The Albert Einstein College of Medicine New York, N.Y. INTRODUCTION Associated with the evolution of ter- restrial life in brachyuran crustaceans are striking modifications in their physi- ology and morphology, particularly of their respiratory organs. Due to greater curvature, the branchial chambers grow more capacious. They become either par- tially or completely filled with air. Vas- cular channels appear in the membrane that lines the branchial chambers, with branchial tufts arising from the membrane in some forms. The vascularized mem- brane and tufts may serve as respiratory structures. Gills become fewer in number and reduced in volume and surface area, these modifications being possible because of the greater amount of oxygen per unit volume in air compared with that in sea water. With less surface then available for respiratory exchange, the loss of water through the gills is reduced. No single species necessarily displays all the modifications just enumerated. The ghost crab, Ocypode quadrata (Ocypodi- 59 dae), possesses 12 gills and prominent branchial tufts. Compared with the gills of most marine crabs, its gills have less surface area per gram body weight (Gray, 1957; Bliss, in preparation). Gecarcinus lateralis and Cardisoma guanhumi (Gecar- cinidae) lack branchial tufts but have capacious branchial chambers and numer- ous vascular channels in their branchial membrane. Like many marine crabs they have 18 gills. However, the surface area of their gills per gram body weight is less than that of the gills of marine crabs (Bliss, in preparation). Furthermore, the blood of Cardisoma guanhumi has a high affinity for oxygen, the gas penetrating rapidly enough to compensate, at least partially, for the reduction in gill surface area (Redmond, 1962). Crabs better adapted to life on land show generally higher respiratory rates than do strictly marine forms (Ayers, 1938). Significantly, this relationship ex- tends to isolated gills but not to excised midgut gland (Vernberg, 1956). However, 60 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 the degree of activity characteristic of each animal affects such results. An active marine crab like the blue crab, Callinectes sapidus, maintains a higher basal meta- bolic rate than does a sluggish marine form like the spider crab Libinia, and may have a rate comparable to that of an intertidal form like Panopeus—but never- theless not as high as that of the active terrestrial crab Ocypode quadrata (Vern- berg, 1956). These and various other features of terrestrial crabs have been reviewed by Pearse (1929a, 1929b), Edney (1960), and Wolvekamp and Waterman (1960). Detailed studies on the osmotic concen- trations of the blood in terrestrial crabs have been made by Pearse (1934), Jones (1941), Gross (1955), Flemister (1958), Green et al. (1959), Gross and Holland (1960), and Dehnel (1962). One adaptive feature of land crabs that seems not to have been investigated, however, is the greater development of the pericardial sacs. A discussion of these important organs follows. GROSS ANATOMY The first published account of the peri- cardial sacs to come to the attention of the writer is that of Milne-Edwards (1834) who noted in decapod crustaceans the existence of a spongy, whitish organ ex- tending on both sides of the body from the posterior portion of the branchial cavity to the beginning of the abdomen. Milne-Edwards described the organs as enveloped in a fold of the tegumentary membrane and resting on the flancs (thoracic epimera). Cuénot (1891) appears to have been next to study the pericardial sacs. In Maja and Carcinus he remarked particularly on the relationship of these organs to the heart, which they partially surround, and on the adherence of their internal border to the overlying pericardial membrane and to the adjacent calcified endoskeletal wall. He also noted that they contract sharply if pricked with a needle, that they contain whitish, pulpy material, and that their external border is free. He emphasized that the pericardial sacs are external in the same way as are the gills, that is, both types of organs are bathed by the external medium. Subsequently, Cuénot (1893) reported pericardial sacs to be present in all brachyurans and, in reduced form, in palinurids, galatheids, and some pagurids. He noted their absence from _ other macrurans, including crayfish. In a classic monograph on Cancer, Pearson (1908) pictured and described the pericardial sacs, which he termed “‘peri- cardial pouches.” He found each sac to be covered with a cuticle that constitutes an extension of the chitinous wall of the branchial chamber and to have a cavity continuous with the pericardial sinus. The cavity, he said, is broken up by connec- tive tissue cells and muscle fibers. The pericardial sacs of three species of terrestrial crabs appear in Figures 15-17, those of two species of marine crabs in Figures 18, 19. In the marine form Callinectes sapidus Rathbun (Fig. 19), the pericardial sacs are narrow, elongate, and pouch-like. They lie on the flancs (thoracic epimera) just posterior to the tips of the gills, which barely overlap them. In Cancer borealis Stimpson, also a marine crab (Fig. 18), the gross struc- ture of the pericardial sacs and _ their relationship to other organs differ little from Callinectes sapidus. In Cancer borealis the pericardial sacs are likewise narrow, attenuated, and pouch-like; they 61 PERICARDIAL SACS OF LAND CRABS BLISS ‘apIs JYSII ay} uO S{[Is oy} Jo uoTAOd wv pue podi[ixew jsiy oy} Jo YOuRIGqOS seu 94} Os[e VARY SB “PpaAOWaI UVaq sey PU[S JNSprur ayy sased [[e uy “suatayap sea ‘A favjas JO 4YJN} [eUIa}Xa ‘CS foes [eIPIeotIed ‘gq {yoRUIO}s ‘QO sauvIG “Wout [RIPOIyye papfoy “PL + (juswisas [eurmopqe jsiy sy} Jo UoNeSuofoid sAota}ue) deg “T {ys ‘H foueg ‘gq {ovs [erprvottad ay} Jo uotsua}xa Jorsysod “q !yivay ‘Y isuUOT}VIADIqGY “UMOYS a1B sqvId a[eUT ATUQ ‘snpidDS saz2aUIIDJ “61 ‘SYDaL0g 4a9UDD “ST ‘sqeAD auTAeUI Jo so¥s [eIPIeoTIAd ‘6 ‘SI ‘SsIq “Dypponb apogkrQ "LT “wunyuvns vmosipapyg ‘ot ‘SYDAIJD] SNUIIADIAH “ST :sqeio purl Jo sovs [eIpreotied ‘/[-ST “SSIQ ‘“6I-ST ‘SOIA 6) ( ml 62 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 too lie on the flancs and are overlapped only slightly by the gills. In the terrestrial crab Ocypode quadrata (Fabricius) (Fig. 17) an outward exten- sion of the pericardial sacs has occurred, with the result that the lateral tip lies far out on each flanc. The tip, nevertheless, is pointed rather than broad and there is little overlap of the pericardial sacs by the gills. The same sort of lateral extension of the pericardial sacs occurs in Cardisoma guanhumi Latreille (Fig. 16). However, the lateral tip, although still attenuated, has penetrated farther into the branchial chambers. Furthermore, a new feature is apparent in Cardisoma guanhumi. There has occurred a posteroventral extension of each pericardial sac towards the ventral surface of the abdomen. The tip of each extension bears setae which lie close to the many external setae that border the abdomen. The most extensive modifications of the pericardial sacs have occurred in Gecarcinus lateralis (Fréminville) (Fig. 15). Laterally the pericardial sacs not only extend far out on the flancs and well into the branchial chambers but each lateral extension terminates in a broad expanded lobe. As a result the gills overlap the peri- cardial sacs extensively. Posteroventrally in Gecarcinus lateralis, as in Cardisoma guanhumi, an extension of the pericardial sacs has taken place. The tip of each posteroventral extension is fringed with setae which lie in close proximity to an external group of setae situated on the lateral edges of the first three abdominal segments (Figs. 15, 20). Perhaps the most striking modification in Gecarcinus lateralis is the extensive de- velopment of the arthrodial membrane that links thorax and abdomen. In this crab the membrane is both sturdy and voluminous. It stretches in folds across the medial region between the two peri- cardial sacs and over the sacs themselves (Fig. 15). It is firmly attached to the movable chitinous flap that constitutes an anterior prolongation of the first abdominal segment. In effect the membrane forms an extensible chitinous bag within which the pericardial sacs lie protected. As the peri- cardial sacs swell during premolt water uptake in Gecarcinus lateralis, the arthro- dial membrane unfolds and the “bag” expands enormously (Fig. 24). Reference to Figures 15-19 leaves little doubt that the pericardial sacs of the three terrestrial species of crabs are larger than those of the two marine species. Furthermore, the pericardial sacs of Gecarcinus lateralis seem to be more ex- tensively developed than are these organs in Cardisoma guanhumi and Ocypode quadrata. Yet it is rather unsatisfactory to base conclusions on gross observation alone when quantitative data can be ob- tained. Therefore, the surface area of the pericardial sacs in the five species of crabs was determined with a Compensating Polar Planimeter (Keuffel and Esser Co., number 4236). Results appear in Tables 2-4. Because Cancer borealis was not always available, the surface area of the pericardial sacs of Cancer irroratus Say also was determined. Morphologically the two species are similar. It was not always possible to obtain and dissect the crabs in the live state, hence some of the data of Table 2 are expressed in terms of fixed weight. In every case the crab was blotted and drained of preservative (70% ethyl alco- hol) before it was weighed. For Gecarcinus lateralis the ratio of fixed weight to live weight is approximately 0.9. BLISS: PERICARDIAL SACS OF LAND CRABS All data of Table 2 concern pericardial sacs that either have been taken from al- ready preserved crabs or have been re- moved from fresh crabs and then fixed for 24 hours in 70 per cent ethyl alcohol. The amount of shrinkage that a fresh 63 pericardial sac undergoes during fixation varies with species. For 12 specimens of Gecarcinus lateralis the fixed pericardial sacs have a mean surface area that is approximately 70 per cent that of the fresh sacs. Yet for two specimens of AhaAR AAs Fics. 20, 21. carapace cut away to show the posterior extension (E) of the left pericardial sac, the setae at its tip, and the setae-lined channel (C). One external tuft of setae (S), on the opposite side of 20, ventral aspect of Gecarcinus lateralis with portions of the abdomen and the the abdomen, also is shown. Abbreviations: N, abdomen; V, vas deferens; D, left first pleopod. 21, ventral aspect of a female specimen of Ocypode quadrata showing the external row of setae at the base of the second and third walking legs when the legs have been separated (left arrow), and when the legs are in normal resting position (right arrow). 64 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 TABLE 2 THE SURFACE AREA OF FIXED PERICARDIAL SACS Surface Area of Crab Pericardial Sacs mm?2/cm. Carapace Weight carapace mm?/gram Species Sex Width (cm) (g.)1 width body weight Gecarcinus i) 5.49 58.9 84.6 7.9 lateralis 3 5.29 47.9 70.7 7.8 ) S21 SYS) 81.7 8.1 Q 5.15 47.2 74.8 8.2 Q 5.09 45.2 60.8 6.8 Q 4.88 40.1 78.4 9.5 3 4.81 42.5 67.0 7.6 Q 4.69 36.0 82.3 10.7 ) 4.44 33.9 81.4 10.6 3 4.31 29.8 59.9 8.7 Q 4.16 29.4 62.0 8.8 Cardisoma 2 8.23 176.8 90.9 4.2 guanhumi 3 6.80 113.7 75.9 4.5 g 6.10 94.7 50.8 3.3 3 5.89 76.7 56.9 44 3 4.03 24.8 38.4 6.2 3 3.70 23.8 34.9 5.4 Ocy pode 3 3.24 17.9 35.8 6.5 quadrata 3 3.06 13.4 29.5 6.7 2 2.90 11.8 35.5 8.7 3 2.47 7.5 41.7 13.8 3 2.29 5e/; 33.8 13.5 3 2.23 4.9 23.1 10.6 3} 1.77 2.5 14.6 10.4 Callinectes é 13.29 191.7 29.1 2.0 sapidus 3 13.61 173.9 27.3 2.1 } 14.11 169.5 28.0 2.3 ) 14.26 160.5 26.6 2.4 3 13.71 155.5 271 2.4 3 12.72 139.6 34.5 3.1 Q 14.14 135.5 33.7 3.5 3 12.64 125.0 30.5 Sul re 12.17 119.6 23.7 2.4 Cancer 3 8.98 113.2 18.3 1.5 borealis 3 7.62 66.3 30.4 3.5 Cancer 3} 11.23 200.5 48.3 2.7 irroratus 3 10.84 190.0 30.9 1.8 3 11.17 185.0 31.8 1.9 1 For G. lateralis, Callinectes sapidus, Cancer borealis, and last 2 specimens of Cancer irroratus, live weight is given; for C. guanhumi, O. quadrata, and first specimen of Cancer irroratus, fixed weight is given. BLISS: PERICARDIAL SACS OF LAND CRABS 65 Cancer irroratus, one of Cancer borealis, and eight of Callinectes sapidus, the sur- face area of the fixed pericardial sacs averaged 83 per cent that of the fresh sacs. Body weight is a good basis on which to make comparisons of surface area of the pericardial sacs in various species, provided all the species in question have approximately the same mass of exo- skeleton. Body weight is not a good basis for making such comparisons if one species has a light, fragile exoskeleton and a second species has a heavy, sturdy one, for the body weight of the second species will include much more inert skeletal material than will the body weight of the first species. Thus in Table 3 the mean sur- face area per gram body weight as cal- culated for the pericardial sacs of Gecar- cinus lateralis, Cardisoma guanhumi, Cal- linectes sapidus, and the two species of Cancer decreases in the order given. Yet in the midst of this series lies Ocypode quadrata with what appears to be an aberrantly high value. O. quadrata pos- sesses a very light, fragile exoskeleton, whereas the other five species have firm, heavy shells. In O. quadrata abnormally high values for surface area of the pericardial sacs on a weight basis may be attributed, at least in part, to the fact that the body weight of this crab does not include the weight of considerable inert skeletal matter, whereas the body weights of the other five species do. For closely related species of com- parable body form (e.g. Gecarcinus later- alis and Cardisoma guanhumi) or for dif- ferent individuals within the same species, carapace width is a good basis for com- parison. Thus, in Table 2 the surface area of the pericardial sacs per gram body weight for Cardisoma guanhumi is approxi- mately the same regardless of size, and therefore age, of the individual crab. Yet on the basis of carapace width there is evident an increase in surface area of the pericardial sacs with increasing size of crab. For this species, because of the con- siderable increase in body weight with age, carapace width is the more valid basis on which to make _ intraspecific comparisons. In the case of Gecarcinus lateralis, on the contrary, aging leads to no marked in- crease in size of claws nor in weight of exoskeleton. Hence body weight and carapace width are equally valid criteria on which to base intraspecific compari- sons. With either unit of measurement, TABLE 31 MEANS AND STANDARD ERRORS FOR SURFACE AREA OF FIXED PERICARDIAL SACS Arithmetic Means and Standard Errors for Surface Area of Pericardial Sacs mm2/cm. carapace width mm?/gram body weight Number in Species of crab Sample Gecarcinus lateralis 11 Cardisoma guanhumi 6 Ocypode quadrata 7 Callinectes sapidus 9 Cancer borealis and Cancer irroratus 5 [Selgst 2.8 8.6 + 0.4 57.9 + 8.9 4.7404 30.6 + 3.6 10.0 + 1.1 28:9 ete 2.6 + 0.2 31.9 + 4.8 2.3404 1 Original data on which this table is based appear in Table 2. 66 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 G. lateralis shows no consistent variation in surface area of its pericardial sacs with age—or with sex (Table 2). In individuals of comparable size and with carapace width as a basis for com- parison, the surface area of the peri- cardial sacs in Gecarcinus lateralis is greater than that of the corresponding organs in Cardisoma guanhumi (Table 2). Yet because of the increase in size of pericardial sacs per centimeter carapace width that occurs with age in C. guanhumi, the difference in surface area per centi- meter carapace width between the samples of G. lateralis and C. guanhumi is not statistically significant (Table 4). On the basis of body weight there is a statistically significant difference; nevertheless, this must be discounted because of the dis- proportionate increase in size of the claws and in weight of exoskeleton that occurs in older C. guanhumi. Therefore, we are forced to base our conclusions regarding these two species on values for individuals of comparable size (Table 2). Then we find that the surface area of the pericardial sacs is greater in Gecarcinus lateralis than in Cardisoma guanhumi. On the basis of carapace width (which is the more valid for the following com- parisons), there is a_ statistically sig- nificant difference in mean surface area of the pericardial sacs between Gecarcinus lateralis and Ocypode quadrata, and also between Cardisoma guanhumi and Ocypode quadrata (Table 4). The existence of a significant difference in the surface area of these organs when Ocypode quadrata is compared with the three marine species TABLE 41 VALUES FOR t AND P ror MEAN SuRFACE AREAS? OF FIXED PERICARDIAL SACS Per centimeter Per gram carapace width body weight t P t P G. lateralis and C. guanhumi 2.0 0.06 6.7 < 0.0001 G. lateralis and O. quadrata 9.5 < 0.0001 1.5 0.2 G. lateralis and C. sapidus 13.3 < 0.0001 13.7 < 0.0001 G. lateralis and C. borealis + C. irroratus 7.8 < 0.0001 10.4 < 0.0001 C. guanhumi and O. quadrata 3.1 0.01 4.2 0.001 C. guanhumi and C. sapidus 4.0 0.002 5.2 0.0002 C. guanhumi and C. borealis +- C. irroratus 2.4 0.03 4.2 0.002 O. quadrata and C. sapidus 0.52 0.6 7.4 < 0.0001 O. quadrata and C. borealis + C. irroratus 0.23 0.8 SS 0.0002 C. sapidus and C. borealis + C. irroratus 0.89 0.4 0.94 0.3 1 Original data on which this table is based appear in Table 2, means and standard errors in Table 3. 2 A difference in mean surface area is considered significant if P < 0.05, highly significant if P < 0.01. BLISS: PERICARDIAL SACS OF LAND CRABS 67 of crabs is highly questionable. There appears to be a significant difference on the basis of body weight but not on the basis of carapace width. As a matter of fact, neither body weight nor carapace width is a good basis for comparison when Ocypode quadrata is being compared with Callinectes sapidus and with the two species of Cancer because of the great differences in amount of inert skeletal material and in body form. Thus there are no reliable data from which to draw con- clusions regarding the surface area of the pericardial sacs in Ocypode quadrata as it compares with that of the pericardial sacs in the three marine species of crabs. In summary we can say: (1) that on the basis of both body weight and cara- pace width the surface area of the peri- cardial sacs in the terrestrial species Gecarcinus lateralis and Cardisoma guan- humi is significantly greater than that of the pericardial sacs in the three marine species Callinectes sapidus, Cancer bore- alis, and Cancer irroratus; (2) that on the basis of individual values for specimens of comparable size (the best basis for this comparison) the surface area of the peri- cardial sacs of Gecarcinus lateralis is greater than that of the pericardial sacs of Cardisoma guanhumi; (3) that on the basis of carapace width (the best basis for this comparison), the surface area of the pericardial sacs in Gecarcinus lateralis and Cardisoma guanhumi is significantly greater than that of the pericardial sacs in Ocypode quadrata; (4) that there exists no valid quantitative basis for comparison of the surface area of the pericardial sacs in Ocypode quadrata with that of the pericardial sacs in the three marine species of crabs. Thus, with the exception of the data for Ocypode quadrata, the quantitative results confirm earlier conclusions based on gross morphology, namely: (1) that the pericardial sacs of terrestrial crabs are larger than those of marine crabs; and (2) that of the six species included in this study the one with the largest pericardial sacs is Gecarcinus lateralis. MICROSCOPIC ANATOMY Thus far histological study of the peri- cardial sacs has been confined to the single species Gecarcinus lateralis (Fig. 22). Externally these structures have a soft covering that probably consists of epi- cuticle and undifferentiated endocuticle (Leo Schatz, personal communication). Just within lies a single layer of epidermal cells. A considerable capacity for expan- sion is suggested by the many convolu- tions that are present in the outer layers of the pericardial sacs. The main mass of a pericardial sac seems to be composed of loose connective tissue similar to the sub-epidermal con- nective tissue of Panulirus argus (Travis, 1955, 1957). Within this loose connective tissue of the pericardial sacs of Gecarcinus lateralis are bands of striated muscle fibers oriented in many directions, al- though primarily transversely. Much of a pericardial sac appears vacu- olar in character. There probably exist both intracellular, vacuoles and_ blood lacunae continuous with the pericardium. In stained sections large amounts of blood, including hemocytes, can be seen scattered throughout the loose reticulum of con- nective tissue. Also visible, particularly within certain areas of a pericardial sac, are large inclusions of unknown composi- tion. According to Mrs. Mary Weitzman (personal communication), such inclusions are found generally within the connective tissue of Gecarcinus lateralis. 68 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 Fic. 22. epidermis (E), vacuolated connective tissue (T), bands of striated muscle fibers (M), hemocytes (H), coagulated blood (B). Many of the histological details given above agree with the early description of the pericardial sacs of Maja and Carcinus reported by Cuénot (1891). He described the “glandes lymphatiques pericardiques” as covered by a chitinous cuticle that is secreted by a palisade type of epithelium. According to Cuénot, the organ is filled with a loosely formed meshwork contain- Microscopic structure of a pericardial sac of Gecarcinus lateralis. Note the cuticle (C), ing numerous bands of striated muscle, most of which are oriented circularly. Cuénot observed the reticulum to contain blood cells and large, clear, rounded vesicles, as well as smaller vesicles bear- ing refringent material that Cuénot be- lieved to be of an albuminous nature. In a later paper Cuénot (1893) charac- terized the connective tissue as composed BLISS: PERICARDIAL SACS OF LAND CRABS 69 of two types of reserve cells, namely, cells of Leydig and proteic cells. These two types of cells, according to Travis (1955, 1957), correspond to the “reserve cells” that she has described in the loose con- nective tissue of Panulirus argus, as well as to the “‘lipo-protein cells” described by Sewell (1955) for Carcinus. Travis (1955) has noted that the reserve cells of integu- mentary tissues resemble those of the mid- gut gland, which are known to store large quantities of fat. In summary, we can conceive of the pericardial sacs as two diverticula of the pericardium that possess a marked ca- pacity for expansion and contraction and a considerable potential for storage. Let us now consider possible functions for these organs. FUNCTIONS The earliest investigators to study the pericardial sacs were not loathe to suggest a possible function. Thus Milne-Edwards (1834) thought that he could detect a canal running from each sac to the ex- terior and wondered if these organs could be the site of an excretory product analo- gous to urine. In 1891 Cuénot felt that the pericardial sacs might serve both for storage of reserve materials and as a site for developing amoebocytes. Two years later, however, he denied any storage or “lymphatic” (hemopoietic) function and suggested that the pericardial sacs may play a mechanical role of unknown sort. Generally speaking, more recent authors have refrained from assigning a function to the pericardial sacs. Thus Pearson (1908), Borradaile e¢ al. (1935), and Lochhead (1950) have stated that the function of these sacs is unknown. Pyle and Cronin (1950) have not mentioned the pericardial sacs at all. Borradaile et al. (1958) have included the pericardial sacs in figures illustrating the internal anatomy of Carcinus maenas but have not referred to them in the figure legends or text. Drach (1939), however, has studied the pericardial sacs in relation to molting. He has reported that in the marine crabs Cancer and Maja the “poches _ peri- cardiales” are extremely elastic and can undergo rapid changes in volume as the pressure within the hemocoel is altered. He has suggested that the sacs may regu- late and limit hydrostatic pressure during molting, by storing water while it is being absorbed. In Cancer and Maja Drach has found the site of water absorption during molting to be the new lining of the diges- tive tract, which undergoes rapid changes in permeability when the old lining is shed. Our own studies on the pericardial sacs of Gecarcinus lateralis (Bliss, 1956) have suggested that in this terrestrial crab these structures play an important role in the premolt uptake and retention of water, which after ecdysis is used to expand the new soft exoskeleton before it hardens at its larger dimensions. In Figure 23 appears the posterior portion of an unswollen intermolt specimen of Gecarcinus lateralis. In Figure 24 is another specimen of the same species as it appears one day before ecdysis. The pericardial sacs are so swollen that they bulge far out from under the carapace. During ecdysis, as the old carapace is forced forward and the crab gradually emerges, the pressure exerted by the old shell on the soft body of the crab diminishes and the pericardial sacs begin to decrease in size. By the end of ecdysis they appear once again as in Figure 23. Our studies (Bliss, 1956, 1962) have indicated that the neuroendocrine system regulating growth and molting also regu- 70 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 lates premolt water uptake and retention by the pericardial sacs. Thus light, which acts hypothetically through the medium of the neurosecretory system to inhibit limb regeneration and molting, also inhibits pre- molt water uptake and retention. Dark- ness, which acts hypothetically via the neuroendocrine system to favor limb regeneration and molting, also favors pre- molt water uptake and retention. Crabs from which eyestalks (and_ therefore sources of the molt-inhibiting hormone) have been removed may show extreme swelling of the pericardial sacs before ecdysis. It is likely that the pericardial sacs of 23. Posterior dorsal view of Gecarcinus lateralis during the intermolt period. There is no external evidence of the pericardial sacs, which are unswollen and lie entirely beneath the carapace. Fic. 24. have become so swollen that they protrude from under the carapace. They lie protected within the unfolded arthrodial membrane. Abbreviations: P, pericardial sac; G, gill. Posterior dorsal view of Gecarcinus lateralis just prior to ecdysis. The pericardial sacs BLISS: PERICARDIAL SACS OF LAND CRABS 71 Gecarcinus lateralis are the primary means of water uptake and retention during the intermolt period as well. The water may replace that lost by evaporation (e.g. from the gills). Perhaps the significant altera- tion that occurs just prior to ecdysis is an increased rate of water uptake and mmcreased retention of water. If so, then a change in membrane permeability, in os- motic pressure of body fluids, or in both may play a role at the time of molt. In order for the pericardial sacs to take up water, there must be some means whereby water can reach the sacs, which are situated under the carapace partly within and partly posterior to the branchial chambers (see Fig. 15). It has already been noted briefly (Bliss, 1956) that the pericardial sacs of Gecarcinus lateralis swell prior to ecdysis only if the crab is on a moist substratum. No swelling occurs if drinking water is the only type avail- able. The question arises, therefore, as to how a crab gains access to the water in a damp substratum.t A possible answer to this question has been suggested by the results of the following experiments. If a specimen of Gecarcinus lateralis is held ventral side up and distilled water containing the vital dye, neutral red, is placed carefully on the external tufts of setae located on each side of the abdomen (Fig. 20), the fluid disappears with extraor- dinary speed. This procedure may be repeated several times before the rate of disappearance of the fluid begins to de- cline. If a few moments are allowed to pass and then drops of fluid again are applied, they vanish rapidly once more. 1 Gross (personal communication) has con- firmed in detail our observations that Gecarcinus lateralis takes up fresh water from damp sand. Gross’ observations were made entirely on inter- molt crabs. When a small portion of the exoskeleton that covers the extension of a pericardial sac at the base of the fifth walking leg is cut away, a setae-lined channel, formed by adjoining segments, is revealed (Fig. 20). This channel leads directly from the external tuft of setae to the setae-fringed tip of the pericardial sac. Here a small lacuna, large enough to hold fluid, is located. Should dye-containing distilled water now be placed drop by drop on the ex- ternal tuft of setae on the same side as the cutaway, the drops can be seen to run swiftly along the setae-lined channel and to collect in the lacuna at the tip of the pericardial sac. Within the next moment or so, the fluid disappears from the lacuna as if being blotted by absorbent tissue. One may speculate that this is the normal pathway followed by tiny drops of moisture (e.g. dew, soil water) from the substratum to the pericardial sac. The presence of a lacuna at the tip of the sac may provide a temporary reservoir for fluid when it is present in excess. Re- call that distilled water applied to the external tufts of setae disappears at a decreasing rate, and that if a few moments without application of fluid are allowed to pass, the rate rises again. It seems likely that the drops of applied fluid run along the setae-lined channel to the lacuna, where they collect and then are slowly absorbed by the pericardial sac. Once the lacuna is full, no more fluid can be carried there by the setae until the excess is absorbed. The conduction of water from the ex- ternal tufts along the setae-lined channels to the lacunae may be due entirely to capil- lary action. Subsequent absorption of fluid from the lacunae by the pericardial sacs may involve either simple diffusion or ac- 72 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 tive uptake. Evidence regarding this proc- ess awaits physiological experimentation. In an effort to obtain some measure of the amount and rate of fluid uptake, application of drops of distilled water was made with a one-milliliter hypodermic syringe and needle. When fluid was thus applied to the external tufts of setae on a crab that had been without food or water for four days, 0.085 milliliters of distilled water were taken up within 32 minutes. Three days later the experiment was re- peated on the same crab which in the meantime had had no food or water. Now 0.290 milliliters of distilled water were taken up within 72 minutes. The mean rate of uptake in these two experiments was 0.0034 milliliters (or 3.4 microliters) per minute. Probably of considerable significance is the observation that drops of sea water, when applied to the external tufts of setae, do not disappear readily. The first few drops vanish, presumably into the lacunae, but subsequent drops remain on the ex- ternal tufts. It would appear that this mechanism of water uptake does not function if sea water is the only moisture available.” In Cardisoma guanhumi setae are found around the entire edge of the abdomen, with longer, thicker ones occurring in the area near the extensions of the pericardial sacs. It is likely that in Cardisoma guan- humi water is carried by setae to the peri- cardial sacs through capillary action, just as postulated above for Gecarcinus lateralis. 2 Gross (personal communication) has _ ob- served that when intermolt specimens of Gecar- cinus lateralis are maintained on sand moistened with sea water, few survive for even two weeks; furthermore, the gills of these crabs appear extremely dry. Setae occur in Ocypode quadrata on each side of the body at the base of the second and third walking legs. At each of these locations the setae mesh with one another to appear as one tuft (Fig. 21). If, however, the second and third legs on one side are separated, two distinct rows of setae become visible. These lead dorsally to other setae which in turn lead through an orifice in the exoskeleton directly into the branchial chamber, close to the gills and the lateral tip of the pericardial sac. Fluid applied to the external setae at the base of the legs moves with great speed up into the branchial chamber. Regardless of species, the external tufts of setae occur in areas of the body that make contact with the substratum when the crab is at rest. The ventral surface of Ocypode quadrata is considerably more rounded than is that of Gecarcinus later- alis and Cardisoma guanhumi. When Ocypode quadrata is at rest, not the posterior portion of the abdomen but the portion adjacent to the base of the second and third walking legs is in contact with the substratum. The presence of two large tufts of setae in this area ensures that moisture from the substratum will be con- ducted into the branchial chambers where it can moisten the gills and the pericardial sacs. The same mode of water uptake from damp sand has been described for Ocypode gaudichaudii by Koepcke and Koepcke (1953). A final point concerns the relationship of gills, pericardial sacs, and pericardium. Injection experiments by the writer on Gecarcinus lateralis have confirmed the early findings of Cuénot (1891) that the pericardial sacs are continuous with the pericardium. If several tenths of a milli- liter of neutral red in sea water are in- jected into the central portion of one peri- BLISS: PERICARDIAL SACS OF LAND CRABS 73 cardial sac, it spreads quickly throughout the sac both towards the heart and towards the gills. After injection into the posterior extension of the pericardial sac, the dye moves rapidly towards the heart. Follow- ing such injections, however, no dye is visible in the other sac. There can be no doubt that blood flows into the pericardial sacs from the peri- cardium and thence close to the external surfaces of the gills. It is conceivable that diffusion of water (transpiration) from the hemolymph to the cells of the peri- cardial sacs and thence across the cuticular membrane of the sacs into the branchial chambers can take place. The portion of the pericardial sac that is overlain by the gills has a thin cuticle and numerous blood spaces. Indeed, one may question parenthetically whether this part of the sac may not also serve as a respiratory apparatus. Edney (1957) has commented that to some extent water can probably pass across the entire cuticle of most arthropods, for no osmotic barrier exists and even a waxy cuticle, such as occurs in insects, spiders, and ticks, is somewhat permeable. No waxlike layer has been demonstrated in the exoskeleton of crustaceans. If transpiration across the cuticle of the pericardial sac does occur, then this structure may serve to maintain a high relative humidity in the branchial chamber and thus to reduce the rate of evaporation from the gills and the branchial membrane. This effect may be enhanced by the actual contact that exists between some of the gills and the pericardial sacs. The more terrestrial a crab may be, the more important the functions of the peri- cardial sacs may become. Therefore, in the next section, we shall consider briefly the distribution of the three species of land crabs on an island in the Bahamas and relate it to the degree of independence that each species shows with respect to an aqueous environment. THE DISTRIBUTION OF LAND CRABS, WITH REFERENCE TO THE AVAILABILITY OF WATER? The property of the Lerner Marine Laboratory in Bimini, Bahamas, B. W. I., covers an area that is approximately 600 feet by 200 feet and extends across the island of North Bimini from the waters of the Bahama Bank to those of the Florida Straits. A vertical profile along the south- southwesterly boundary of this property appears in Figure 25. On it is indicated the distribution of the three species of land crabs that occur there. A primary factor governing the distribu- 3 This section is a brief summary of material that will be presented in detail in a forthcoming paper. d GECARCINUS poor — OCYPODE CARDISOMA |\OCYPODE SEA LEVEL HORIZONTAL SCALE =a) 20 0 20 FEET Fic. 25. VERTICAL EXAGGERATION Two TIMES Vertical profile along the south-southwesterly boundary of the property of the Lerner Marine Laboratory, Bimini, Bahamas, B. W. I. The distribution of Gecarcinus lateralis, Cardisoma guanhumi, and Ocypode quadrata on the property is indicated. 74 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 tion of these crabs is the availability of water. Ocypode quadrata remains on the beaches at either end of the property, where it can run into the ocean frequently, thus moistening its gills and replenishing its body fluids. Cardisoma guanhumi is found only in areas of very low elevation. It burrows down to ground water, that is, to the fresh water lens that the natives tap for their drinking water. When this crab is dug out of its burrow, it is taken from wet mud or even from mucky water. Gecarcinus lateralis ranges widely throughout the laboratory property, over- lapping Cardisoma guanhumi in open areas of sun or light shade. Where there is deep shade and a high population density of Cardisoma guanhumi, as in the area marked Cardisoma on Figure 25, no Gecarcinus lateralis occur. The burrows of Gecarcinus lateralis never descend to ground water and the crab never enters the ocean (except fe- males with spawn—to drop their eggs). The soil out of which Gecarcinus lateralis is dug feels powdery dry to the touch and contains only a_ small percentage of moisture. Of the three species of land crabs, there- fore, Gecarcinus lateralis is the only one to which copious supplies of water are not always available. Only during heavy rain showers does this crab become thoroughly wet. Yet particularly during the winter dry season, rains may be very infrequent. Unpublished experiments by the writer have indicated that in air Gecarcinus lateralis is far more resistant to desicca- tion than is Ocypode quadrata and some- what more resistant than is Cardisoma guanhumi. Gecarcinus lateralis loses water at a significantly lower rate than does Ocypode quadrata and is able to with- stand a greater total amount of water loss than can either of the other two species. Accordingly its period of survival in air is longest. Clearly Gecarcinus lateralis has acquired the greatest independence of an aqueous environment. How has this been done? The answer to this question involves in part a recapitulation of much that has already been discussed in this paper. The gills of Gecarcinus lateralis, although num- bering 18, have the smallest surface area per gram body weight of any of the crabs considered here and smaller than any species listed by Gray (1957) in his ex- tensive study of marine, intertidal, and terrestrial crabs. The gill lamellae of Gecarcinus lateralis are strongly scle- rotized and are held apart by ridges that ensure free circulation of air between the lamellae. The branchial chambers are capacious and their epidermal lining is highly vascularized. The pericardial sacs have the largest surface area of any crab so far studied. They penetrate far into the branchial chambers where they may help significantly to keep certain gills moist by contact and where they may also reduce drying of the gills and the branchial membrane by maintaining a high relative humidity. Accordingly, Gecarcinus lateralis is assured adequate respiratory exchange without marked danger of desiccation. Not only have various morphological adaptations served to emancipate Gecar- cinus lateralis from dependence upon an aqueous environment, but climatic con- ditions have done so as well. This crab is restricted to maritime locations. For example, North Bimini is an oceanic island where the relative humidity of the atmosphere is very high, averaging 85 per cent for the whole year and varying only slightly from summer to winter. Maritime BLISS: PERICARDIAL SACS OF LAND CRABS 75 air is moist, regardless of season. The air that filters into the burrows of Gecarcinus lateralis on this island is laden with moisture. At its highest point the island of North Bimini is not far above sea level, maxi- mum elevation on the laboratory property being about 20 feet. Ground water lies close by and capillarity serves to raise some of this ground water to upper levels. The relative humidity of the soil atmos- phere at the level to which many burrows of Gecarcinus lateralis descend, therefore, is high. At night the atmosphere often reaches saturation, with light to heavy dews common in both summer and winter. Earlier in this paper it was postulated that moisture coming in contact with external tufts of setae on the abdomen of Gecarcinus lateralis follows setae-lined channels to lacunae situated at the tip of the pericardial sacs. It was also postulated that either by simple diffusion or by active uptake this water enters the pericardial sacs. Now it is suggested that the prin- cipal form of natural water to come in contact with the external tufts and sub- sequently to be absorbed by the peri- cardial sacs is dew. While Gecarcinus lateralis rests within its burrow or while it runs outside of its burrow during nocturnal forays for food, dew may form on the body of the crab or on objects with which the external tufts make contact. This dew then may be absorbed by the pericardial sacs. The fre- quency with which dews occur would make it unnecessary for the crab to depend upon sporadic showers during which to re- plenish its body fluids. Thus, through a combination of mor- phological and physiological adaptations and favorable climatological factors, Ge- carcinus lateralis has become almost totally independent of an aqueous medium. Cardisoma guanhumi and Ocypode quad- rata possess many of the same adaptive features as does Gecarcinus lateralis, but they depend upon them for survival far less. For Cardisoma guanhumi and Ocy- pode quadrata, water in quantity must always be available for immersion, and due to the limits of their distribution, it always is. These species regularly immerse themselves in ground water or in the ocean, as the case may be. Gecarcinus lateralis returns to the ocean only to provide its young with the en- vironment that they must have for de- velopment. After completing larval de- velopment this species remains entirely dependent for its well-being upon the adaptive changes that evolution has wrought to equip it for life in a relatively dry terrestrial environment. Among these changes, not the least important appears to be the greater development of the peri- cardial sacs. SUMMARY 1. The comparative gross morphology of the pericardial sacs is discussed in three species of terrestrial crabs, Gecar- cinus lateralis, Cardisoma guanhumi, and Ocypode quadrata, and in three species of marine crabs, Callinectes sapidus, Cancer borealis, and Cancer irroratus. 2. Determination of the surface area of the pericardial sacs indicates that, in general, terrestrial crabs have larger peri- cardial sacs than have marine crabs and that Gecarcinus lateralis has the largest pericardial sacs of any species studied. 3. Histologically, the pericardial sacs of Gecarcinus lateralis are characterized by the presence of a thin covering (epi- cuticle and probably undifferentiated endo- 76 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 cuticle), a single layer of epidermis, con- siderable vacuolated connective _ tissue, and numerous bands of striated muscle fibers oriented in many directions. They also contain blood spaces in which hemo- cytes are numerous. The presence of many convolutions in the outer layers and numerous muscle bands within the sacs suggest a considerable potential for ex- pansion and contraction. 4. For many years the functions of the pericardial sacs were virtually unrecog- nized. Drach (1939) suggested a probable function for these structures in marine forms, that of regulating hydrostatic pres- sure within the hemocoel during water uptake at ecdysis. Our own studies have indicated that in the terrestrial crab Gecarcinus lateralis the pericardial sacs take up water from a damp substratum. Just prior to ecdysis water uptake and retention are intensified so that the peri- cardial sacs may become extremely swol- len. At all times these structures may help to maintain a high relative humidity in the branchial chambers and thus to reduce the rate of evaporation of water from the gills. 5. In Gecarcinus lateralis an extension of each pericardial sac posteriorly towards the base of the fifth walking leg where there is a small lacuna may serve as the actual site of water uptake. External setae and a setae-lined channel are believed to conduct minute drops of water (e.g. dew) from the substratum to the lacuna where the water is then absorbed by the peri- cardial sacs. A somewhat comparable arrangement may exist in Cardisoma guanhumi, but not in Ocypode quadrata, where the external setae are located in a more anterior position. Nevertheless, in the latter they may serve the same func- tion, that of conducting water to the peri- cardial sacs—and in this case directly to the gills as well. 6. The distribution of the three species of terrestrial crabs is considered in rela- tion to the availability of water. Gecar- cinus lateralis is the only one to which plentiful supplies of water are not regu- larly available. It is also the species most resistant to desiccation. A combination of morphological, physiological, and ecologi- cal factors have combined to provide this species with a high degree of independence of an aqueous medium. ACKNOWLEDGMENTS Some work reported in this paper was supported in part by Research Grants G-4006 and G-11254 from the National Science Foundation. Field observations were made by the author in 1955 when she was a Research Fellow in Biology at Harvard University and a guest investigator at the Lerner Marine Laboratory, Bimini, Bahamas, B.W.I. Financial support for these field studies was provided in part by an In- stitutional Grant to Harvard University from the American Cancer Society. Assistance in obtaining several species of crabs was provided by Mr. Harold Ber- man, Dr. Phyllis Cahn, Mr. Adrian Gage- steyn, Dr. Evelyn Shaw, and Mr. Mal- colm Taylor, and also by the staff of the Lerner Marine Laboratory and that of the Bermuda Biological Station. Histological preparations that served as a basis for Figure 22 were made by Miss Kate Gruen. Figures 20-24 were drawn by Miss Maria Wimmer when she was a member of the Department of Graphic Arts at the American Museum of Natural History. Figures 15-19 were modified by the writer from sketches by Miss Wim- mer. Topographical data upon which BLISS: PERICARDIAL SACS OF LAND CRABS 77 Figure 25 was based were obtained from the staff of the Lerner Marine Laboratory. Stimulating discussions regarding the histology and gross anatomy of the peri- cardial sacs were held with Mrs. Mary Weitzman and Mr. Leo Schatz. Some biological material used in determining the surface area of pericardial sacs was contributed by Mrs. Weitzman. Dr. War- ren J. Gross kindly made available certain unpublished observations regarding water uptake in Gecarcinus lateralis. The contributions of all named above are gratefully acknowledged. AYERS, JoHN C. 1938. Relationship of habitat to oxygen consumption by certain estuarine crabs. Ecology 19:523-527, 1 fig. Buiss, Dorotuy E. 1956. Neurosecretion and the control of growth in a decapod crustacean. In: Wingstrand, Karl Georg (ed.), Bertil Hanstrom. Zoological papers in honour of his sixty-fifth birthday, November 20th, 1956. Lund, Zoological Institute, pp. 56-75, 7 figs. . 1962. Neuroendocrine control of loco- motor activity in the land crab Gecarcinus lateralis. Mem. Soc. Endocrinol., no. 12, Neuro- secretion: 391-410, 12 figs. Borraparte, L. A., F. A. Potts, L. E. S. East- HAM, AND J. T. SAUNDERS. 1935. The Inverte- brata. Second edition. New York, The Mac- millan Co., Cambridge, University Press, pp. i-xv, 1-725, 483 figs. . 1958. The Invertebrata. Third edition revised by G. A. Kerkut. Cambridge, Univer- sity Press, pp. i-xvii, 1-795, 523 figs. Cutnot, L. 1891. Etudes sur le sang et les glandes lymphatiques dans la série animale. 2¢ partie: Invertébrés. Arch. Zool. Exp. Gen. (ser. 2) 9:13-90, pls. 1-4, 15-18. 1893. Etudes physiologiques sur les crustacés décapodes. Arch. Biol. 13:245-303, pls. 11-13. DEHNEL, PAut A. 1962. Aspects of osmoregula- tion in two species of intertidal crabs. Biol. Bull. 122:208-227, 7 figs. Dracu, Prerre. 1939. Mue et cycle d’intermue chez les crustacés décapodes. Ann. _ Inst. Océanogr. 19:103-391, 13 figs., pls. 2-7. Epney, E. B. 1957. The water relations of ter- restrial arthropods. Cambridge Monogr. Exp. Biol., no. 5, pp. i-vi, 1-109, 32 figs. —. 1960. Terrestrial adaptations. Jn Water- man, Talbot H. (ed.), The physiology of Crustacea. New York, Academic Press, 1:367- 393, 7 figs. FLEMISTER, LAuNCE J. 1958. Salt and water anatomy, constancy and regulation in related crabs from marine and terrestrial habitats. Biol. Bull. 115: 180-200, 2 figs. Gray, I. E. 1957. A comparative study of the gill area of crabs. Biol. Bull. 112: 34-42, 3 figs. GREEN, JAMES W., Mary Harscu, Lioyp Barr, AnD C. Lapp Prosser. 1959. The regulation of water and salt by the fiddler crabs, Uca pug- nax and Uca pugilator. Biol. Bull. 116:76-87, 1 fig. Gross, WARREN J. 1955. Aspects of osmotic regu- lation in crabs showing the terrestrial habit. Amer. Nat. 89:205-222, 2 figs. Gross, WARREN J. AND PAut V. HoLtaAnp. 1960. Water and ionic regulation in a terrestrial hermit crab. Physiol. Zool. 33:21-28. Jones, Lowett L. 1941. Osmotic regulation in several crabs of the Pacific coast of North America. Jour. Cellular Comp. Physiol. 18: 79-92, 2 figs. Koercke, H.-W. anp M. Koercke. 1953. Con- tribucion al conocimiento de la forma de vida de Ocypode gaudichaudii Milne Edwards et Lucas (Decapoda, Crust.). Publ. Mus. Hist. Nat. “Javier Prado,” ser. A. Zoologia 13: 1-46, 14 figs. Locuueap, J. H. 1950. Callinectes sapidus. In Brown, Frank A., Jr. (ed.), Selected inverte- brate types. New York, John Wiley and Sons, Inc., London, Chapman and Hall, Ltd., pp. 447-462. Mitne-Epwarps, H. 1834. Histoire naturelle des crustacés. Paris, Librairie Encyclopédique de Roret, 1:i-xxxv, 1-468. Pearse, A. S. 1929a. The ecology of certain estuarine crabs at Beaufort, N. C. Jour. Elisha Mitchell Scien. Soc. 44:230-237, pl. 8. . 1929b. Observations on certain littoral and terrestrial animals at Tortugas, Florida, with special reference to migrations from marine to terrestrial habitats. Papers Tortugas Laboratory 26 (Carnegie Inst. Washington, Pub. No. 391):205-223, 3 figs. 1934. Freezing points of bloods of certain littoral and estuarine animals. Papers 78 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 Tortugas Laboratory 28 (Carnegie Inst. Washington, Pub. No. 435) :93-102. PEARSON, JOSEPH. 1908. Cancer (the edible crab). Proc. Trans. Liverpool Biol. Soc. 22:291-499, 13 figs., 13 pls. PyLe, ROBERT AND EUGENE CronIN, 1950. The general anatomy of the blue crab, Callinectes sapidus Rathbun. Maryland Board of Natural Resources, Dept. Research and Education, Pub. No. 87:1-40, 19 figs. RepMonpD, JAMES R. 1962. Oxygen-hemocyanin relationships in the land crab, Cardisoma guanhumi. Biol. Bull. 122:252-262, 3 figs. Sewet1t, M. T. 1955. Lipo-protein cells in the blood of Carcinus maenas, and their cycle of activity correlated with the moult. Quart. Jour. Micros. Sci. 96:73-83, 5 figs. Travis, Dorotuy F. 1955. The molting cycle of the spiny lobster, Panulirus argus Latreille. II. Pre-ecdysial histological and histochemical changes in the hepatopancreas and integumen- tal tissues. Biol. Bull. 108:88-112, 34 figs. . 1957. The molting cycle of the spiny lobster, Panulirus argus Latreille. IV. Post- ecdysial histological and histochemical changes in the hepatopancreas and integumental tissues. Biol. Bull. 113:451-479, 23 figs. VERNBERG, F. JoHN. 1956. Study of the oxygen consumption of excised tissues of certain marine decapod Crustacea in relation to habi- tat. Physiol. Zool. 29:227-234, 1 fig. WoLtveKAMP, H. P. anp Tatpot H. WATERMAN. 1960. Respiration. Jn Waterman, Talbot H. (ed.), The physiology of Crustacea. New York Academic Press, 1:35-100, 10 figs. PHYLOGENY AND EVOLUTION OF CRUSTACEA MusEuM oF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 VI Discussion Following Papers by Kinne, Gordon Bliss and WATERMAN: Dr. Kinne has restricted the coverage of adaptation to its ecological aspects in which there are direct inter- actions between various environmental pa- rameters, such as temperature or salinity, and the organism. Considering the organism as a self-perpetuating steady-state system, there are four other important aspects of adaptation which also deserve to be men- tioned: physiological, genetic, develop- mental and evolutionary (Waterman, 1961, “Physiology of Crustacea’’). Also we must keep in mind the adaptive implications in the “fitness of the environment.” These different facets of adaptation show considerable overlap; yet they can be rather easily distinguished on the basis of the time scale over which they are signifi- cant. Thus physiological adaptation is mainly concerned with the immediate short- term mechanisms maintaining the steady state whether these deal with sensory func- tions, nutrition, or the molecular basis of hormone action. The adaptations with which Dr. Kinne was mainly concerned involve time scales of hours, days or weeks which are significant for physiological ecology. Developmental adaptations extend through- out most of a generation; genetic adapta- tions, at least at the population level, ex- tend through several or many generations while the time scale for evolutionary adap- tation is the longest of all. 79 Although it is a common opinion, perhaps a nearly universal one in the present com- pany, that evolution is the explanation in biology, I would like to propose a different point of view which seems to place the study of evolution in a more reasonable perspective. Since whatever happens in bio- logical systems must be adaptive in order to persist (otherwise it would throw the organism out of its essential steady state), the basic explanation which we are looking for in all kinds of biological phenomena is their adaptiveness. This would seem to be a valid generalization for the whole field regardless of whether we are more specifi- cally interested in evolutionary, embryonic, genetic, ecological or physiological events. Obviously this view is consistent with the central role of natural selection in evolu- tionary studies. KINNE: I have tried to define and restrict the often loosely and widely ap- plied term “adaptation” to provide a use- ful tool for the evolutionist as well as for the physiologist. Waterman’s enlarged concept of adaptation is practically iden- tical with biology as a whole, comprising genetics, ecology, embryology (develop- ment), evolution, and physiology and ap- pears to be less specific and less useful. How do you distinguish, for example, between genetic and evolutionary adap- tation, and between internal and chrono- 80 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 logical adaptation? It is certainly true that the so-called internal adaptations (i.e. the functional and structural mechanisms in- volved on the enzyme-, hormone-, cellular-, and organ-levels) deserve more attention. However, our knowledge concerning the physiological mechanisms of adaptation is rather limited and does not provide an adequate basis for a review that could be integrated with the primarily ecological approach that was chosen here. GLAESSNER: Adaptation, unless more strictly defined, is not a very useful con- cept in paleontology. When we discuss things from the point of view of adapta- tion, we end up relegating evolution to a minor subdivision, or to a general philo- sophical context, and bringing everything under an enlarged concept of adaptation. We should not consider everything as either an adaptation or not an adaptation. What is zot an adaptation is what the Crustacea have inherited from their pred- ecessors. What follows when they come into being and differentiate is their adap- tation. In the higher Malacostraca, the division of the body into a cephalothorax and an abdomen, the division of the vari- ous organs, and the distribution of loco- motory, respiratory, and sensory organs, is inherited, but not acquired by that group as an adaptation. The modifications of that inheritance are the adaptations that we are discussing here. Let us talk more about evolution and restrict the concept of adaptation so that it remains useful. KINNE: As long as we talk about adaptation in Crustacea, it is correct to say that what the Crustacea have inher- ited from their phylogenetic predecessors (the characteristics which make the Crus- tacea a distinct group of animals) is not to be considered an adaptation. I would like to add a word concerning the over- emphasis of the structural aspect in evo- lution. It appears desirable to relate de- fined structures to functions, e.g. different modes of life, locomotion, and respiration, or to different habits of feeding. Adapta- tion provides a useful, synthetic concept, emphasizing the relationships between structure and function of a given organism and its environment. MANTON: A hermit crab, adapted in innumerable ways to the gastropod shell in which it lives, is an example of adapta- tion related undoubtedly to a particular environmental niche. Another type of evo- lutionary change, which is not directly im- posed by any property of the environment, is much the more important concerning the major steps in evolution, as pointed out by D. M.S. Watson (1949, Proc. Linn. Soc. London, 160 (2): 75-84) for +Plesi- osauria. I can only give you comparable examples outside the Crustacea, but the principles probably apply to the Crustacea also. In the same decaying log today you can find Peripatus, centipedes, millipedes, Symphyla, scorpions and other arachnids, apterygotes, etc. Their taxonomic morpho- logical characters are related to divergent habits, which were probably established at an early terrestrial stage. A hundred or more conspicuous body features recently have been shown to be correlated with these habits: the form of head, trunk and legs; the shape, size and number of scutes, their fusion together or doubling in num- ber; the musculature and endoskeleton, etc. These taxonomic characters suit: the bulldozer burrowing habits of the diplo- pods; the fast moving predatory habits of the chilopods, with extreme adaptation towards manipulating food in shallow crevices; the gaining of shelter by twisting and turning without pushing of the Sym- DISCUSSION 81 phyla; the amazing powers of body dis- tortion in the Onychophora, by which they gain shelter (again without pushing) through narrow crevices far too small to permit pursuit by predators large enough to harm them, etc. Adaptation to a par- ticular niche leads to the edge of a “preci- pice” when conditions change. But adap- tations which facilitate particular habits of life, which lead to better or easier living in a variety of habitats, result in per- sistent and far-reaching evolutionary ad- vance. It is not easy to recognise which, among the many things an animal does, represents the habit or habits of real evo- lutionary importance. Great advances have been made with the Onychophora and Myriapoda along these lines. Com- parable work on the aquatic Crustacea may be much more difficult, but it is an approach which should be made. KINNE: This is certainly an important aspect of adaptation. But was it not men- tioned in my paper when I talked about behavioral adaptation? WATERMAN: I do not quite agree with that interpretation of what Dr. Man- ton said. My understanding of her com- ments would be as follows. If a variety of different arthropods were to approach a log of wood, some of these would walk around it, others would push through it, some would wriggle under it, and so on. These forms differ in their behavior when confronted with the same environmental situation. Such differences are clearly of adaptive value since those which conform to the structural and functional needs of the particular organism will confer a selec- tive advantage on it. I think Dr. Manton has made an important point which had previously been neglected, namely that a crucial issue in evolution and adaptation is the initial presence of some particular habit or behavior pattern which emerges from the structural and functional details of the creature’s organization. MOORE: Does what Manton calls habit really mean the behavior that is con- trolled in terms of what Waterman said? WATERMAN: For the present discus- sion we may perhaps consider this be- havior as spontaneous. To go into some ex- planation of its origin would take us too far afield. However, I again emphasize the importance of time scale in speaking of adaptiveness. For example, it seems im- possible for me to accept Glaessner’s con- tention that hereditary information which each individual organism receives from its parents is not adaptive. Clearly the in- dividual could not go on living unless it received this information. However, I agree that in the immediate sense to which he was referring it is not adaptive, which il- lustrates the key position of time scale in such considerations. Also I object somewhat to Kinne’s state- ment that what we, or other organisms, are now is just the result of evolution. From a strictly operational approach, such as most physiologists, biochemists and _bio- physicists take, an organism is what it is now because of its immediate relations to itself and to its environment. What it may have been in the remote past or what it may become in the distant future are quite irrelevant from this point of view. SANDERS: There is a different way of viewing the same problem. Physiologists are most interested in those environments where the physical conditions tend to fluc- tuate rather widely demanding physiolog- ical adaptations. However, when one thinks of the number of species on our globe, only a small fraction lives in such environ- ments. One might think of two pure types of 82 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, i963 communities: (1) the physically con- trolled, and (2) the non-physically con- trolled where physical factors such as tem- perature, salinity, light, etc. vary very little in the course of the year, and have remained essentially unchanged for ex- tensive periods of time. In the predom- inantly physically controlled community, the number of species is small, and there is marked numerical dominance by few species; in the predominantly non-physi- cally controlled community there are many species with no species markedly abun- dant. In the physically controlled com- munity the niches are broad, large, and tend to be diffuse. In other words, biolog- ical interactions are not well developed— the adaptations are to physical conditions. In the other type of environment the niches are smaller, more circumscribed as a product of biological interactions result- ing in biological accommodation. Probably most of the functional adaptations that we have been talking about evolved in the non-physically controlled environment as a response to biological accommodation. Thus the physiologists find the physically controlled environment inherently more in- teresting while the functional morpholo- gists find that the reverse is true. WATERMAN: Can you give an ex- treme example of each of those? SANDERS: An extreme example of a physically controlled environment would be a hypersaline bay where only three species may be present. The salinity may vary from zero to sixty or even seventy parts per thousand, and there may also be appreciable temperature change. The high Arctic is another example. A tropical rain forest or the deep sea represent rela- tively stable environments, and in such situations one finds non-physically con- trolled communities. WATERMAN: The organism that is living in the deeps at a thousand atmos- pheres, in the dark and in a place where there is no primary food production is under a very difficult biochemical and physiological stress. SANDERS: A limitation of food does not necessarily imply a limitation in the variety of animals, but rather a limitation in the number of animals. WATERMAN: As I understand it you distinguish between the degree of stress and the rate of change of such stress. SANDERS: If the environment remains constant in time with a stress factor such as high pressure in the deep sea, the ani- mals have made the necessary physiolog- ical adaptations a long time ago. From that period on, biological interactions came into play resulting in biological ac- commodation, and the numerous niches of the non-physically controlled community. WATERMAN: Does this tie in with the rate of speciation? SANDERS: I believe so. Most of the centers of evolutionary activity have been in non-physically controlled environments —for example, southern Asia for fresh- water fish. TASCH: I should like to stress the con- cept of the antiquity of a given adapta- tion. In the region of the Great Salt Lake, recent cores show that Artemia salina was an inhabitant at least 600,000 years ago. Its adaptation to a hypersaline environ- ment was established by then, and un- doubtedly long before that time. Take the matter of ephippial eggs in cladocerans. The oldest known cladoceran fossil is a daphniid ephippium from the Oligocene, 40 million years ago. Ephippial eggs occur in times of stress in the environment. Adaptations of this type are very ancient. WATERMAN: One of the most im- DISCUSSION 83 portant things that paleontologists can do is to tell us in which direction and how fast evolution has in fact moved in its historical course. On the basis of comparative mor- phology, ethology, physiology, biochemis- try, development, and so on, biologists can construct a plausible ladder or tree of evolu- tionary change. However, none of this com- parative evidence can tell us anything certain about the direction or actual rate of change in the proposed phylogeny. This is largely true even for the population geneticists who are directly studying evo- lutionary mechanisms. However, if the geo- logical record yields relevant data from more than one stratum both direction and rate of evolution can be defined in terms of the history of the sediments in question. Such information seems to me crucial in controlling the speculative nature of phy- logenies otherwise obtained. BOWMAN: To verify what Sanders said, in terms of planktonic animals: in the study of populations of calanoid cope- pods of the southern coast of the United States we get definite zonation of the num- ber of species into inshore and offshore. Inshore, the number of species is quite limited, but the number of individuals is much greater than offshore where you get more species. MANTON: Details could be given of habit reversals and their morphological consequences or accompaniments, in which the direction of change is not a matter of speculation, but is quite clear. These habits are not adaptations to environmen- tal niches. For example, there are a few millipedes which have given up their basic habit of burrowing by pushing and show covergent resemblances to centipedes in accomplishments and structure, superim- posed upon a basic diplopodan morphology. There is also a centipede (Craterostigmus ) which has given up the ability of fast running and its structure is much modified permitting great flexibility of the trunk. In this case I do not know the full nature of the habits associated with these fea- tures. They clearly are not adaptations to a particular niche. We now know that there are many arthropodan groups whose basic evolution has not been adaptive to the immediate environment or to Sanders’ narrow niches, but to habits which can be employed in various circumstances. There is much too great a readiness at the present time to believe that structure, and the physiological capabilities of that struc- ture, are largely explicable as direct ad- aptations to the environment, although general adaptations to a terrestrial or an aquatic environment are also important. WATERMAN: It is also familiar that independent parallel adaptations appar- ently appear over and over again in dif- ferent families of the same group. GLAESSNER: If we work on the basis of the traditional classification, for ex- ample, then we are discussing whether the differences between the Raninidae and the Oxystomata should be given a certain taxonomic rank, a problem in which nature does not give us much guidance. If we assess the differences in the structure of the genital organs, we may be in danger of arriving at a single character classifica- tion on the basis of important differences. We have to take into consideration the fossils, no matter how incomplete the record, because we know something of what happened in the past, in this transi- tion from Macrura to Brachyura, and the origin of crabs. We know from the Creta- ceous a number of undescribed crabs of this group, including Dorippidae. We have to study the evolution of these systems of functioning organs before we decide what 84 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, importance to give them in classification. We have to study evolution, not only on the basis of comparative morphology or of embryology, but also on the basis of paleontology. I do not agree that they should be given different values and that paleontological evidence should be given more value than the others. Certain things Gordon has said appeal to me. For ex- ample, that there is a distinction between Macrura, Anomura, and Brachyura was a rather crude approach in the past when little was known about the decapods. “Natantia,’ ‘“Reptantia” was a_ purely functional approach to classification, and not the evolutionary one which our system should represent. I diverge somewhat from Gordon’s opinion in believing that the peculiar habit of the Dromiacea and Dorippidae (to have used their reduced last pairs of legs or one pair of legs for carrying objects on their backs) is a dis- tinctive pre-adaptation. It is connected with the reorganization of the body in the internal skeleton, from a macruran type with its cylindrical internal skeleton to the conical internal skeleton of the crabs, with a different habit of locomotion and different arrangement of muscles of the branchial chamber and so on. There was 1963 not enough space, until this reorganization was completed, for the muscles of the last pairs of legs (which are affected by the loss of the ancestral lobster tail) to be fully developed and remain fully functional. Only when the stage of the Dromiacea and Dorippidae was passed did the organization of this internal skeleton allow the full development of the attach- ment of the last pairs of legs and of their muscles to come into play again. Nothing was lost completely. In the transitional stages the legs were utilized for other functions, for carrying objects. This was a reasonably successful group, but when the crab habit was fully developed in the later, more advanced forms, then the last pairs of legs could grow to their full size, could be adapted for swimming as in the portunids, and in all sorts of ways, and the balance of the organism was restored in a new way, leading to adaptive radiation. GARTH: As J listen to various people, I realize that what we are talking about would be called adaptation by Kinne, physiology by Waterman, and_ probably evolution or phylogeny by Glaessner. I think Gordon and I would include most of it under systematics. PHYLOGENY AND EVOLUTION OF CRUSTACEA Museum oF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 vil Studies in Malacostracan Morphology: Results and Problems By Rolf Siewing Zoologisches Institut und Museum der Universitat, Kiel, Germany For more than a hundred years there has been discussion of phylogenetic rela- tionships within the Crustacea. But, up to the present we have not succeeded in re- constructing a system reflecting in all respects the natural relationships. There are seyeral reasons for this, including the insufficient paleontological data, the in- terpretation of which is partially disputed. Nearly always we need to work with Recent material. The first step in reconstructing the phylogeny is the establishment of natural systematic categories (Fig. 26). This is done by the determination of homologies, using exact criteria, following the method recently elaborated and interpreted by Remane (1956). We can carry out this step in a number of crustacean groups, so that we get a horizontal or a two-dimen- sional arrangement of the different groups. The second step consists of a transition into a three-dimensional arrangement. The homologous similarities of the natural groups depend on their common descent. We can thus draw a picture of phylo- 85 genetic connections between the systematic groups, and depict them in a schematic phylogenetic tree (Fig. 27). This picture represents the best demonstrable and best explained way of understanding the phylo- genetic relations of one group. New inves- tigations and discoveries may lead to the alteration of details. The third step in morphological science is the reconstruction of the phylogenetic ancestor of the different natural groups (Fig. 28). It may be suitable first to define several words. “Primitive” means that a structure is simple. But simple and primitive do not necessarily imply a phylogenetic state. Primitiveness or simplicity can also origi- nate secondarily. An early phylogenetic state, “ursprtinglich” in German, I have translated as “original.” The opposite term is “derived.” With the above of I have investigated different malacostracan groups, and have obtained the following results. The Malacostraca is the only group of point view, 1963 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 86 “‘poyyoul UONeASdAUT-AZojouroy ay} Aq paedtiap wovsysOoR[ePL 9Y} UTYIIM sdnois [einyeu Jo JUoWaSuLAIe [LUOISUIUITP-OM} JY} Surmoys weiseiq 97 “O17 e390 =eudisboaejads 0 epodos| ‘0 eaoepidseuy O Bae . eaoeyjauAyjeg 0 eaoewn PBapiyo eeoeisneydng 0 PadeU epodiydwuiy ‘0 eplesayen + ‘0 Ppodo}JEW0j}S O =Auosdwe94 0 epodesaq 0 =deqsowsey| O eacepishj| ‘0 esoel]}eqen O DpDD0\doH 9] | DpuDIUAG'G] | DpiDoNy 7} |OpuD.UDY €} | DPD DIAZ} |D2D.}S0}d97 | NOISIAIC NOISIAIC NOISIAIG NOISIAIG POBIISODETEW S:SVe7OeNS 87 SIEWING: MALACOSTRACAN MORPHOLOGY ‘suoT}IoUUOD IoUasoTAYd YIM “vovIZsOOR[ePL 94} UIYIIM sdno1Z [emnyeu jo JUsWaZURIIe [eUOISUDUIIp-daIYy} dy} SuIMOoYS WeISeIGT ‘LZ ‘OI epodo}ew0}js ‘O DpluD90)doH 9 NOISIAIG eaoepidseuy oO eaceyjauujeg ‘OQ Baplyo =Auosdwe9+ ‘O DplDIUAS G NOISIAIG eaoeisneydng O eaoeuU epodesagq 0 =3eqSOWJay| ‘O ppuDan3 | |opuDsuDY ¢ NOISIAIG NOISIAIG BICIJSONDSIEW SSVIDGNS ea9 =eudiiBoaejads O epodos| O Baoepieue| O easeWND O epodiydwiy ‘QO eaoepish O BpHEByeN + 0 BadeeGeN O DPIDIDIAd Z NOISIAIG poDI}s0}d97 | NOISIAIG 88 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 Ovd Thorax K Abdomen Fic. 28. Hypothetical tennula; 2.A = antenna; eye; D-=intestine; Ed—endites; Enp = endopodite; furca; Gh = brain; H = heart; MN = maxillary-nephridium; NAu=nauplius eye; Ov = ovary; Ovd = oviduct; Plp = pleopods; T = telson; = maxillula; 3 = maxilla. subclass rank of which we can say with certainty that it represents a natural group; this means that it has a mono- phyletic descent. Marked by the constant situation of genital openings, the Mala- costraca possess moreover a constant num- ber of segments, consisting of six segments in the head, eight in the thorax, and originally seven in the abdomen. Since the synthesis of Calman (1904), the Malacostraca have been classified into divisions or superorders (Fig. 26), and these have the character of natural groups. There has long been agreement that the Leptostraca must be regarded as the most original Malacostraca. This basic position is proved by the great carapace, which is free of all the thoracic segments; more- over they are original in having a true furca and seven abdominal segments. In a certain sense also the extension of the heart through almost the whole body is original and particularly its invasion into the hind part of the head region (Fig. 29). Nearly all the paired lateral arteries con- reconstruction of the phylogenetic ancestor of Malacostraca. 1.A = an- AN = antennal-nephridium; Bm = ventral nerve cord; CoAu = compound Ex = exopodite; Fu = Ot = ostium; 1= mandible; 2 Ep = epipodite; Thp = thoracic legs; sist of two components, one of which pro- vides for the extremities, the other for the internal organs (Fig. 30). I have found a similar composition in the Anaspi- dacea, and it is present, more or less clearly, in other Malacostraca. It is in- teresting to note that similar conditions are also found in some errant polychaetes. Perhaps direct phylogenetic connections exist here. Possibly therefore the bipartite arteries are original. The possession of two pairs of nephridia is also original. Diagnostic characters of the Leptostraca are the change of legs into phyllopod-like extremities, the structure of the stomach with its gastric mill, the structure of the first antenna and the lack of exopods on the second antenna. Further, the carapace has developed into a bivalved shell, the halves of which can be moved by an ad- ductor muscle. One finds parallels for these characters in other crustaceans. The differences do not justify a separa- tion of the Leptostraca from the rest of the Malacostraca. Moreover these dif- 89 SIEWING: MALACOSTRACAN MORPHOLOGY “AdoyIe [eyUsWIZas Jo qouviq [eI19IsIA = YA :WNAjsor Jo AraqIV = OTY !umajsor = y { AroAw yesdje] [eyuswBes pared v jo youesq [elpod = yg ‘AivA0 = AQ ‘eso = JO suniqe, Jo AlgIV = OHIO ‘unipruydou Areypixew = NW “yoruloys = WY Seay = A ‘uleiq= yy ‘vony = y ‘atpodoxa = xq :aytpodida = gy ‘ayipodopua = (NA ‘Are [esiop= 9yq ‘{aoedeies — UVO ‘e49= YY ‘10Le}sod ej10e = gy ‘unipuydeu jeuuajue = NY :Iojonppe = qqVv ‘1OHaIUe Row = WY ‘evuuaUe— WZ ‘eNUUZJUR=— WT (‘9S6T ‘SuTMaTS) ‘uvoRIyso}da] & JO UOT}EZIURSIO dy} JO WRISLIP IeUIDYIS “67 ‘OLY dy 70NF 90 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 Fic. 30. Nebalia bipes. Cross-section through the heart and its lateral arteries with visceral (VK) and podial (PK) components. (Siewing, 1956.) H = heart; KV =lips of the blood vessel valve; PK = paired podial blood vessel; PS = pericardial septum; TA = outer layer (connective tissue) of the heart; TM = inner layer (muscular tissue) of the heart; VK = paired visceral blood vessel. ferences from the other Malacostraca are markedly moderated by the statement that +Nahecaris and other +Archaeostraca have a typical malacostracan organization (Fig. 31). The first antenna has a normal structure and has two flagella. The second antenna possibly possesses an exopod. The carapace seems not to be bivalved, and finally, the extremities are not phyl- lopod-like (Broili, 1928). Thus +Nahecaris and other +Archae- ostraca are important connecting links be- tween Leptostraca and the lower Malacos- traca. They are simultaneously original. Recently other representatives of the former ‘‘“;Archaeostraca” have been totally separated from the crustaceans by St@rmer (1944). A highly specific homologous organ, that connects the Leptostraca and the Stomatopoda, is the so-called pro- SIEWING: MALACOSTRACAN MORPHOLOGY 91 Fic. 31. +Nahecaris, reconstruction. (Broili, 1928.) cephalon (the segment of the head forward Leptostraca, so that a near relationship of the second antenna which is movable between Leptostraca and Stomatopoda is by a number of specific muscles). very probable. Although the Stomatopoda In the Stomatopoda there appears a have a considerably more exceptional or- further articulation of the head, which ganization than the Leptostraca, a posi- makes the segment bearing the eyes mov-_ tion outside the Malacostraca for the last able (Fig. 32). The muscles in the head’ group has never been discussed. region are almost identical to those of There are a few original characters in YY) DS Wencenn : 7 MAS \ Me : AGC yj Wf Ct tre by h IFT holy 4 yey ‘fo, Pr hips ATTY Hoon T OMe te | bess wily / i H i yi i 7 fs “ ' TL WES HN Ces Ma Fic. 32. The procephalon of a stomatopod: (A) Dorsal view, (B) Ventral view. (Snodgrass, 1951.) 1 Ant=antennula; 2 Ant=antenna; Cp —=carapace; cvMb=cervical membrane; d= fastigial plate; e = ocular plate; Epst = epistome; f= postocular plate; g—=posterior dorsal plate of head; Lm=Labrum; Md =mandible; R= rostrum. 92 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 08.3 NX Y,; 59.5 me m4 3 py N wale >) ‘ < ) He} ta 5 y \c tur enw emer’ L\ Kl sf 2 Sie SS LA) x= & y .~} YH = S ™~ MY) oy Y NS =e 4 a) 8, Gs %s D ax ees De Seas, La wt W se : ws) wy I | = . p4-s WA re eo AS Re PSA TS lpé a Wats . AK LS? “ps6 péfe- APS dp | Ac Fic. 33. Blood system of Stomatopoda: (A) Larva of the last stage of Squwilla. (Balss, 1938 after Claus, 1883), (B) Adult of Squzlla oratoria (Balss, 1938 after Komai and Tung, 1931). al = arteria antennularis; an=—arteria antennalis; ao = aorta anterior (—cephalica), br = thoracic gill; ca. ao = aorta posterior; c. an—=common origin of both arteries of first and second antenna; c.ao = aorta anterior (=cephalica); ce =arteries of the brain; dp = aorta posterior; en = endopodites of uropods; ex = exopodites of uropods; H = heart; lc = arteria lateralis cephalica; lp = paired lateral arteries of abdomen; 0, os = ostia; oc =artery to the compound eye; oph= ophthalmic artery; SIEWING: MALACOSTRACAN MORPHOLOGY 93 stomatopods (Fig. 33): (1) the heart with its paired segmental vessels and the segmental ostia is so original that we can take it as a model for the most primitive vessel system in the Crustacea. Paired vessels and ostia are present for nearly all segments in the thorax and abdomen, and moreover there is a pair for a lost seventh segment in the abdomen. Obviously this original organization of the central vessel system is connected with the presence of gills on the pleopods. But I would prefer to refute the argument that the gills are new organs and therefore the heart should extend secondarily into the abdomen. The heart and all marks connected with it are so clearly original that one can assume that the gills on the pleopods are also primitive relicts. In the Cephalocarida and Notos- traca also we have a shift of the epipods onto the exopods. Transference of epipods we know also in amphipods and decapods. Thus perhaps it can be argued that the presence of gills on the pleopods is an original character. This last conclusion is still hypothetical, but I am sure that the heart in the stomatopods is original. By means of the abdominal arteries one can demonstrate that in the Stomatopoda —in contrast to the other Malacostraca— the number of six abdominal segments is the result of the reduction of the first abdominal segment. This conclusion seems to conflict with the presence of a petasma. This copulatory organ in the Anaspidacea and in the whole Eucarida is formed from the appendages of the first and second abdominal segments. Since the first seg- ment is reduced in Stomatopoda, here the petasma must be constructed by the second and third pleopods. Therefore, a homology with the petasmae of other crus- taceans is to be excluded. In fact the structure of the petasma in stomatopods shows very little similarity with that in the other groups (Fig. 34). It seems to be an euanalogy. A convergent formation of a petasma is also demonstrable in many isopods. After this digression we have to com- plete the list of original characters in stomatopods (Fig. 35). There is still to mention (2) the extension of genital organs through the whole body. As special charac- ters we have the subneural artery with a great number of rami communicantes. Furthermore, there are five pairs of claws with the simple structure of subchelae, the surface structure of the whole body is modified in connection with the current of respiratory water, and in addition are the pereiopods and the eyes. Judged by their original characters the Stomatopoda stand near the basis of Malacostraca and therefore near the phyllocarids. Moreover a specific homology connects these two groups: the movable procephalon with its complicated muscu- lature. Up to the present, in Malacostraca, nowhere is found a procephalon of such a perfect structure. Nevertheless from the Phyllocarida to the Hoplocarida there is an important phylogenetic step: the furca is replaced by the transformed last pair of abdominal extremities, the uropods. They form to- p4-5 = limit between the fifth and sixth abdominal segment; p6/te = limit between the sixth abdom- inal segment and the telson; pl = branches of arteries to the pleopods; ro = branch of artery to the rostral plate; rc = blood vessel branches between lateral arteries and the subneural artery (= ramus communicans) ; sc = arteries to the frontal part of larval carapace; sg = paired segmental arteries; sn = arteria subneuralis; th = arteries of the thoracopods; thp = limit between thorax and abdomen; up = arteries to the uropods; Z=artery to the hind median spine of larval carapace. 04 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 A Ir. an. di. ir. an. pr. Fic. 34. The so-called petasma of Squwilla mantis: (A) first pleopod, (B) second pleopod, show- ing how it differs from the petasma of Decapoda and Anaspidacea. (Giesbrecht, 1921.) Cr=keel; Fol = foliated blade; Ir.an.di.= frontal distal piece of armour; Ir.an.me.3 = third frontal median piece of armour; Ir.an.pr.= frontal proximal piece of armour; Pr.a. = hook-shaped projection; Pr.tu.= tubular projection; Ret = retinacula. «— Abdomen Ma Aoa Thorax ———| . & AS Ki ==! ===. \RRQAY c\ ] \ \ N Pp, Bm Art. subn. \N XA Fe ; Fic. 35. Schematic diagram of the organization of a stomatopod. (Siewing, 1960, fig. 4.) Aoa = aorta anterior; Aop = aorta posterior; Art. subn = subneural artery; Bm = ventral nerve cord; Da = intestine; Ho = testes; Ki= gills; KoAu = complex eye; Ma = gizzard; Md = intestinal diverticula; P,.. = thoracopods; Pp = pleopod; Up = uropods; 1 = mandible. SIEWING: MALACOSTRACAN MORPHOLOGY 95 gether with the telson a broad terminal paddle. The furca is seen embryologically in a greater number of Malacostraca. An advance is also to be seen in the fusion of the four frontal thoracic segments and a part of the fifth in the stomatopods. This fusion is here, as in many other crusta- ceans, achieved independently. The Stomatopoda have much in com- mon with the Anaspidacea within the Syn- carida. A specific homology is the so-called pars ampullaris on the entrance of the caeca into the pyloric chamber. The superorder Syncarida is a group poor in species, with an accumulation of negative characters and having many re- lationships to other Malacostraca. In all species a carapace is wanting, and no breeding mechanism is evolved. Within the group, the first thoracic segment tends to fuse with the head. In this respect the Bathynellacea and the +Gampsonychidea are original—the latter because of their seven free abdominal segments, provided partly with extremities, the former be- cause of their furca in addition to the uropods. A reduction abdominal segments is very common in all the other representatives of Syncarida and furthermore of Malacostraca. In the Anaspidacea, which can be ana- tomically studied, we see further original characters (Fig. 36): the long extended heart with vessels consisting of a podial and visceral component; the structure of the thoracic legs that carry endites and in which a well-developed praeischium is present. The gonads extend through the whole body in the male. The statement of closer relationships is based on the following specific homologies: a real petasma and a statocyst in the first antenna connect the Anaspidacea with the Eucarida, especially with the Decapoda. On the other hand, the statocysts in the uropods of +Gampsonychidea point to the Mysidacea. Which character is the more important? That is the problem. The stem and fusion of 4 Ant Fic. 36. Schematic diagram of the organization of an anaspidacean 1.Ant = anten- (Anaspides) . nula; 2.Ant—=antenna; Aoa=aorta anterior; Ao desc. = aorta descendens; Aop = aorta posterior; Art. supran = arteria subneuralis; Da = intestine; He = heart; KoAu = complex eye; Md = abdom- inal caeca; MxN = maxillary nephridium; Ost = ostium; Ov = ovary; P = thoracopod; Pp = pleo- pod; Te=telson; Up =uropods; 1 = mandible; 2 = maxillula; 3 = maxilla. 96 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 form of syncarids surely had only one pair of statocysts, either in the uropods or in the antenna. Only one of these two can be a homologous organ, the other must be a convergence. It seems to be uncertain whether the impressions in the uropods of +Gampsony- chidea are different organs or even arti- facts. If the latter, the relationship of the Syncarida would be more directed to the Eucarida on the basis of the statocyst in the basal joint of the first antenna. But perhaps, also the small depression in the antenna of Hansenomysis is a vestige of a static organ. If this is correct a closer re- lationship of the Anaspidacea to the Mysi- dacea, which represent the basic group of the Peracarida, would be indicated. It is not yet certain whether the lacinia mobilis-like structures, recently found in some South American Anaspidacea, are ho- mologous organs or only convergences to the similar organs in Peracarida and Pancarida. But certainly the Syncarida branched off from the main malacostracan stem earlier than the Peracarida and Eucarida. This follows not only from their geological age but also from their usually original organi- zation in comparison with Eucarida and Peracarida. Direct phylogenetic relationships to Isopoda and Amphipoda do not exist. One of the most diverse superorders is the Peracarida. The Peracarida are charac- terised by a specific breeding mechanism that is unique among arthropods. The eggs, and the embryos after hatching, are kept in a brood pouch of the female that is periodically formed by oostegites and more or less reduced afterwards. The relationships within the Peracarida have been clarified by the dissolution of the so-called ‘‘Ar- an artificial group, throstraca,” the main categories of which were the Isopoda and Amphipoda. In fact, there are several similarities between the two groups (Fig. 37), such as the reduced carapace and the fusion of the first one or two segments with the head. Further, the exopods of the thoracic legs and the eyestalks are reduced. The coxal joints are transformed to coxal plates and some of the thoracic legs have turned forwards, others backwards. Are these common characters specific homolo- gies and are they sufficient to prove a natural relationship? The greater part of these characters de- pend on reduction. An independent evolu- tion of such characters is possible, and they cannot prove a close relationship be- tween Isopoda and Amphipoda. We are facing euanalogies. Concerning the turning of thoracic legs in the two groups, von Haffner demon- strated that it is achieved independently. Thus the coxal plates remain as the only specific character common to both groups. But it is demonstrable that such broaden- ing of the coxal joints is present also in other crustacean groups. Thus we have grounds for the suspicion that there are no nearer relationships be- tween isopods and amphipods. Their union as “Arthrostraca,” in several cases to- gether with the Syncarida, is artificial. We have to investigate characters that prove a separation of the Isopoda from the Am- phipoda. In fact, there are a great num- ber of such characters (Fig. 37): (1) the Isopoda have a maxillary gland, the Amphipoda an antennal gland; (2) gills are formed by epipods of the thoracic legs in Amphipoda but by the pleopods in the Isopoda; (3) in connection with the localisation of the respiratory organs we find the heart in the thorax in Amphipoda; SIEWING: MALACOSTRACAN MORPHOLOGY 97 in the Isopoda it lies mainly in the abdo- men. (It is highly probable that it has secondarily transferred to this situation.) Many Isopoda possess a more or less well- developed arteria subneuralis, which is i = —— absent in all Amphipoda; (4) the differ- ences in the structure of abdominal ex- tremities are important: while the Isopoda have five pairs of pleopods and only one pair of uropods, in the Amphipoda there eeeeee 2. Ant Thorax ——>«— Abdomen z le Le von SS ae gh A FOG coco y (ey Ge A2 Gene CS’ yt =, Ups Fic. 37. Schematic diagrams of the organization of an isopod (A) and an amphipod (B): for comparison. (Siewing, 1960, fig. 7.) ANphr = antennal nephridium; Ao = aorta; He =heart: Ki— gills (= epipodites) ; Mars = marsupium; Mda = anterior dorsal caecum; Mdp = abdominal caeca of intestine; S Art = segmental artery; Te = telson; breviations see Figs. 29, 35, 36. Up = uropods; I= maxillipede; for other ab- 98 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 are three pairs of so-called uropods; (5) the Isopoda have a manca stage, but this is wanting in Amphipoda; (6) the egg-embryos in the Isopoda are curved dorsally, in Amphipoda ventrally (Fig. 38); (7) in Isopoda the oostegites are al- most completely reduced after every brood, in the Amphipoda only their marginal bristles are reduced; (8) further dif- ferences are connected with the microscopic structure of the gut, especially of the stomach. This great number of differences sug- gest that within the Peracarida there exist other relationships than hitherto assumed, relationships that are based on homologous correspondence. In fact it is demonstrable that the Iso- poda are much more nearly related to the Tanaidacea and Cumacea. They are connected with these groups partly by just those characters in which they differ from the Amphipoda (Fig. 39). The Amphipoda, on the other hand, are closer to the mysids. The reconstruction of the natural system of the Peracarida by means of homologies shows that isopods and amphipods are end points of two divergent lines of evolution which arise near the mysids. One of these lines com- prises the Cumacea, Tanaidacea, and Isopoda; the other line comprises only the Amphipoda. In both lines we have a reduction of the carapace, eyestalks, and thoracic exopods. Along with the reduc- tion of the carapace as the main respiratory organ, and the restriction of the epipods to one pair, we see a functional substitu- tion by the pleopods, which has already begun in the tanaidaceans. The transfer of the heart of isopods to the hind part of the body (Fig. 37) accompanies this secondary transfer of respiratory organs from the thorax to the abdomen. In this connection we can understand the re- markable structure of the heart with its blind caudal end, and the different inser- tion of arteries from that in amphipods. Fic. 38. Embryos with rudiments of extremities of an amphipod (A) and an isopod (B), showing the different curvature of the egg-embryo. Embryo just before hatching of an amphipod (C) and an isopod (D). (Siewing, 1960, Fig. 8, after Weygoldt and Nair.) SIEWING: MALACOSTRACAN MORPHOLOGY 99 Fic. 39. ogous organ. (A) Amphipoda; (C) Cumacea; (I) Isopoda; Hatching with all extremities; (2) pyloric funnel of the gastric mill; (3) Caeca anteriora dorsalia; (4) antennal nephridium; Diagram showing relationships within the Peracarida. Each line corresponds to a homol- (M) Mysidacea; (T) Tanaidacea. (1) (5) pyloric bristle-chamber; (6). epipodial-gills on the thoracic legs; (7) oesophageal valve of the gastric mill; (8) tripartite cardiacal-ventral-piece of the gastric mill; (9) in- ner surface of carapace functioning as a gill; maxillary nephridium in the adult animal; “hood-plate” in the gastric mill; (14) manca oostegites after each brood. (Siewing, 1951.) In the isopod line we have a greater number of radical secondary transforma- tions, whereas the Amphipoda preserve a few original characters such as the heart and the epipods on the thoracic legs. The degree of certainty with which we can reach conclusions can be considered a test for the quality of the morphological method. In the current investigations of the relationships within the Peracarida, I have expressed the opinion that the post- embryonic development in Tanaidacea must take a similar course to that in Cumacea and Isopoda. These two orders are, according to our investigations, more closely related to the Tanaidacea than to any other peracaridan group. A few years (10) embryos in the egg curved dorsally; (12) stage of the embryo; (11) syncytial-like midgut-epithelium; (13) dorsal (15) stage with reduced ago Lang (1953) was able to verify this assumption. Here we see clear evidence for the exactness of the morphological method. The Mysidacea, including the Lopho- gastrida which are in many respects orig- inal, represent the connecting link to other Malacostracan groups, particularly to the Eucarida and Pancarida. One _ specific homology between the Peracarida and Eucarida is the arteria subneuralis. In both divisions it is very similar in structure, in the arrangement of the vessels towards the extremities, and in the position of the aorta descendens (Fig. 40). However, we find an artery leading to 100 aa a B SD SP A SL VEE > i) asn MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 VAS A x Fic. 40. Comparison of the main vessels of the blood showing the identical position of the aorta descendens Decapoda (B), subneuralis (asn) and its branches. the ventral nerve cord also in Stomatopoda, Anaspidacea and many Isopoda. A com- parative investigation shows that these arteries are not altogether homologous structures, but that they have originated independently in several lines. Only for Eucarida and Mysidacea is the homology highly probable. Besides this important connecting link, we find further charac- ters in the structure of the stomach that agree even in microscopical detail. A common root for the two superorders of Eucarida and Peracarida can be accepted. In the neighborhood of this root stands a group of small crustaceans with very few species, the Pancarida, with the single order Thermosbaenacea (Fig. 41). They are very closely related to the Peracarida by the lacinia mobilis of the mandible. This structure is indeed a most charac- teristic mark of the Peracarida. But this alone does not suffice to incorporate the Thermosbaenacea in the Peracarida, par- in Lophogastridea (A) and (ad) and of the arteria system ticularly if we find such structures in other arthropods; it occurs in some newly dis- covered anaspidacean species (Noodt, 1963). The uncommon breeding habit of the Pancarida argues in favor of making this group a separate division. The primitive structure of the maxillipeds indicates a branching off below the Peracarida but above the Eucarida. Thus, we can envision the phylogenetic connections as shown in Figure 42. Finally I should try to answer the ques- tion of the stemform of the Malacostraca (Fig. 28). The body consists of the head, eight thoracic and seven abdominal seg- ments; the telson carries a furca. Each of the homologously formed thoracic legs has a three-jointed protopod, an endopod with six joints, endites on several of these joints beside those of the protopod and furthermore two epipods and a flagelli- form exopod. The abdominal extremities SIEWING: MALACOSTRACAN MORPHOLOGY possess a similar homologous form, but their structure is far simpler and certainly secondary. The mandible has a palp, but no exopod; the second antenna bears an articulated exopod or squama. The cara- pace covers all thoracic segments, and these are not fused. Stalked eyes are present. The gut has a stomach, and ventral and dorsal caeca are present in the most anterior part of the midgut. The circulatory system consists of a heart, extending through the whole body 101 and into the hind part of the head. It is furnished with segmental, paired ostia and arteries, each of which is composed of a podial and a visceral component. There are two pairs of nephridia. In addi- tion to the stalked compound eyes a rudi- mentary nauplius eye may have been present. The gonads extend through the whole body. The openings of the testes lay in the eighth, that of the ovaries in the sixth thoracic segment. The development is indirect, there is a Fic. 41. nula; 2.A=antenna; Aa =aorta anterior; 1. Schematic diagram of the organization of a pancaridan. (Siewing, 1957.) 1.A = anten- Aag = artery to the antennula; 2 Aag = artery to the antenna; Aba = abdomen; Ac = brain artery; Amp = ampulla of the brain artery; Ap = aorta posterior; Cp —=carapace; Cz —further brain arteries; Dc —deutocerebrum; Ed=rectum; En, End = endopodite; Ex = exopodite; Fz = frontal blood vessel of the artery of labrum; H = heart; Hd = testes; Km = stomach; Md = mandible; Mdd = intestinal gland; Mx,, Mx, = maxillula and maxilla; Mxp = maxillipede; Mdm = intestine; Ol = labrum; Olg = artery of labrum; Ot = ostium; Pc = protocerebrum; Pe=copulatory organ; Plp=pleopods; Pp=pereiopods; Ps = pericardial septum; Ra=circumoesophageal artery; Tc = tritocerebrum; Up =uropods; Vd=vas deferens; X= point of curving back of the vas deferens. 102 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 | SOPODA TANAIDACEA SPELAEOGRIPHACEA ? a CUMACEA AMPHIPODA MYSIDACEA OSTRACODA COPEPODA DECAPODA: MYSTACOCARIDA -- Peracarida EUPHAU- SIACEA THERMOSBAENACEA ? EUTHYCARC INUS STOMATOPODA BRANCHIURA Eucarida-* Se ae BATHYNELLACEA x _ Hoplocarida : CIRRIPEDIA ANASPIDACEA LEPTOSTRACA l +GAMPSONYCHIDEA | : TNAHECARIS ASCOTHORACIDA Syncarida Phyllocarida Vag Maxillopoda EUANOSTRACA Malacostraca 9 +LIPOSTRACA CEPHALOCARIDA Anostraca PHYLLOPODA Gnathostraca tPSEUDOCRUSTACEA CRUSTACEA Fic. 42. Phylogenetic tree of the Malacostraca showing the possible relations of Malacostraca to other Crustacea (Siewing, 1960, fig. 20). SIEWING: MALACOSTRACAN MORPHOLOGY nauplius larva, and formation of segments follows serially. This ancestor, compared with other crustaceans outside the Malacostraca, is much more original in many respects. It seems to be unjustified to maintain the opinion that the Malacostraca are the “higher” crustaceans. Certainly some groups of Malacostraca have reached a highly evolved level of organisation, but it is most probable that they branched off very early from the common crustacean stem (Fig. 42). REFERENCES 1928. Sitzungsber. Abt. Broivi, FERDINAND. Nahecaris. math.-nat. 2 figs. Carman, W. T. 1904. On the classification of the Crustacea Malacostraca. Ann. Mag. Nat. Hist. (7) 18:144-158, 1 fig. HAFFNER, KoONSTANTIN von. 1937. Untersuchungen uber die urspriingliche und abgeleitete Stellung der Beine bei den Isopoden. Z. wiss. Zool. 149:513-536, 15 figs. Lanc, Kary. 1953. The postmarsupial develop- ment of the Tanaidacea. Ark. f. Zool. (2) 4:409-422, 4 pls., 6 figs. Beobachtungen an Bayer. Akad. Wiss., 1927/1928, pp. 1-18, 1 pl. 103 Noopt, Wo.LrramM. 1963. Anaspidacea (Crustacea, Syncarida) in der siidlichen Neotropis. Zool. Anz. 26 supp.: 568-578, 6 figs. ReMANE, ADOLF. 1956. Die Grundlagen des natiirlichen Systems, der vergleichenden Ana- tomie und der Phylogenetik. 2nd ed. Leipzig, Akademische Verlags, Geest und Portig, vi + 364 pp., 82 figs. SrEWING, Ror. 1951. Besteht eine engere Ver- wandtschaft zwischen Isopoden und Am- phipoden? Zool. Anz. 147:166-180, 1 fig. . 1953. Morphologische Untersuchungen an Tanaidaceen und Lophogastriden. Z. wiss. Zool. 157:333-426, 44 figs. Untersuchungen zur Morphologie der Malacostraca (Crustacea). Zool. Jahrb. (Anat.) 75:39-176, 75 figs. 1957. Anatomie und Histologie von Thermosbaena mirabilis, ein Beitrag zur Phylogenie der Reihe Pancarida (Thermos- baenacea). Abhandl. Akad. Wiss. Literatur, Mainz, Math. Nat. Kl. 1957 (7):197-270, 43 figs. 1960. Neuere Ergebnisse der Ver- wandtschaftsforschung bei den _ Crustaceen. Wiss. Zeitschr. Univers. Rostock, Math.-Nat. Reihe 9:343-358. SrgRM_ErR, Lerr. 1944. On the relationships and phylogeny of fossil and Recent Arachno- morpha. Skr. Norsk. Vid.-Akad. Oslo, I. Mat.- Nat. KI1., No. 5, 158 pp., 30 figs. aay PHYLOGENY AND EVOLUTION OF CRUSTACEA MuseuM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 Vill Discussion Following Siewing’s Paper SANDERS: In regard to the so-called primitive second antenna of the Malacos- traca, one thinks of a scale and a flagel- liform endopod. In +Nahecaris, however, the endopod of the second antenna is reminiscent of the condition in many entomostracans. How is the thoracic limb to be interpreted? I presume what we see in Figure 31 are trunk limb exopods. The supposedly primitive exopod in the Mala- costraca is a flagelliform structure and we do not have such an exopod here. Proba- bly the flagelliform exopod in the Mala- costraca is a secondary acquisition. It seems that we are trying to impose a caridoid pattern of limb morphology on +Nahecaris that is not necessarily indi- cated in the reconstruction. GLAESSNER: +Nahecaris is the one animal of this group in which the append- ages are preserved. ROLFE: One significant feature of +Nahecaris pointed out by Giirich and Hennig, but denied by Broili (1930), con- cerns the structure of the carapace. +Nahe- caris is preserved in pyrite and is normally cleaned with wire brushes, so that any traces of segmentation in the limb or in the carapace, would be obliterated simply by this method of preparation (although setae are beautifully preserved). As the reconstruction (Fig. 43) shows, the cara- pace is typically rhinocaridid in having a median dorsal plate between the carapace valves, and a movably articulated rostral plate. Thus +Nahecaris does not justify the maintenance of a discrete order +Nahe- carida. The body structure shows that it median dorsal plate rostral plate mesolateral carina Fic. 43. Reconstruction of +Nahecaris stuertzi Jaekel, after Broili pleomere 7 telson (1929, Sitzungsb. d. Bayer. Akad. d. Wiss. math.-naturw. Abt.), carapace structure revised by W. D. I. Rolfe (cf. Fig. 31). 105 106 belongs in one of the well known sub- orders of the Phyllocarida +Archaeostraca: the +Rhinocarina. This structure of the carapace with the median dorsal plate is specialized and would seem to exclude ;~Nahecaris from any truly intermediate position between the +Archaeostraca and the mainly Recent Leptostraca. This is only one character, however, and I would agree with Siewing that the biramous an- tennules (if Broili’s 1929 reconstruction of these is genuine—cf. his 1928 reconstruc- tion) do suggest a link between the yArchaeostraca and the Eumalacostraca. GLAESSNER: Do you still consider the Palaeozoic phyllocarids as the ances- tors of Recent Nebaliacea? ROLFE: Yes, although we have no one genus which we can point to as the ancestor for the Leptostraca. Malzahn (1958, Z. deutsch. geol. Gesell., 110:352- 359) has found the abdomen of a nebaliid in the Upper Permian of Germany, which shows that the Leptostraca are at least that old. DAHL: The biramous antennule of the Malacostraca is peculiar; have Siewing or the paleontologists anything to tell us on that? SIEWING: There is no known reason for this unique feature turning up in the Malacostraca. One could speculate that this is a multiplication of organs. Thus the three flagella in the first antenna of the Stomatopoda indicate perhaps an in- crease in the sensory surfaces of this organ. Such multiplication of organs is sometimes found in the animal kingdom. BROOKS: Paleozoic Eumalacostraca have one pattern: a peduncle of three joints and two flagella. MANNING: When did_ the branched first antenna show up? BROOKS: We know little about the antennae of the Paleozoic pre-stomatopods. three- MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 Peach (1908, Geol. Surv. Gt. Britain, Palaeontology, Mem., p. 39) figured 7Perimecturus from Scotland with a bira- mous first antenna. MANNING: _ Siewing’s_ morphological work finds support in Shiino’s work (1942, Mem. Coll. Sci. Kyoto Imp. Univ., Ser. B, 27 (1): 77-174) on the embryology of Stomatopoda, where the seventh abdominal segment shows up in the blood vessels and in the nerve cord. But the sixth and seventh segments fuse, embryologically, and the first is not reduced. SIEWING: The conclusion that the first abdominal segment is reduced in Stomatopoda is based only on the com- parative anatomy of the blood-vessel sys- tem. There is an artery on the limit be- tween thorax and abdomen that leads to no extremity but ends blindly in the tis- sue. A pair of ostia also corresponds to this lateral artery. Therefore, here the first abdominal segment may be reduced as in Scorpionidea. But there is no embryolog- ical evidence; in the embryo, the seventh abdominal segment is reduced as in other Malacostraca. MANTON: The embryology of a mysid and of Nebalia shows the formation of seven pairs of abdominal mesodermal som- ites and a seventh ganglion, but no cor- responding limb develops on this segment. The uropods arise on the sixth abdominal segment, not the seventh, and the caudal furca is borne by the telson and is not a segmental appendage. SIEWING: We can see in this a dis- crepancy in that the uropods are present on the seventh segment in the }Gampsony- chidae. BROOKS: This is one of the errors that has been perpetuated [see Rolfe, 1962, Palaeontology, 4:548; and Brooks, 1962a, p. 168; 1962c, p. 236 for reviews of this error. | DISCUSSION WATERMAN: Siewing mentioned cer- tain similarities between Leptostraca and annelids. Is it unorthodox to assume that the Malacostraca are the central stem of Crustacea and that other so-called primi- tive ones are simplified or aberrant forms which come from the same basic roots or perhaps from other groups? Are the Crus- tacea monophyletic? SIEWING: There are more original characters preserved in the Malacostraca than in other Crustacea. Probably these other Crustacea lost these original char- acters secondarily. It is hard to say how we should derive the non-malacostracan Crustacea. It is in- consequential to assume they were derived phylogenetically independently from other Arthropoda. Furthermore, there is no evi- dence that other groups of the “lower Crustacea” are derived from other aber- rant Malacostraca. A possibility of the relationships of Malacostraca with the other Crustacea is given in Figure 42. I also suggest that the Malacostraca Leptostraca Tf Pseudoctustacea 107 yArchaeostraca probably have an exten- sive reservoir of characters, from which may be derived the so-called lower Crus- tacea (Fig. 44). I have expressed the opinion that the Malacostraca have preserved more original characters than other Crustacea; this is valid for the original groups in this sub- class. There is no doubt that they have reached a high evolutionary level in the decapods. This means that other Crus- tacea lost their original characters more quickly than the Malacostraca (Fig. 42). WATERMAN: Earlier, Dahl implied that the Entomostraca are more primitive. DAHL: In the Malacostraca we find a definite number of segments and a divi- sion of the body into tagmata. It seems plausible that forms with a large and varying number of segments and a vary- ing number of limbs are more primitive. In all Crustacea most segments are added one by one from a proliferation zone in front of the telson. It is not surprising that in a long series of segments, as e.g. in Maxillopoda Gnathostraca TTrilobita Fic. 44. Possible relations of {Archaeostraca to other Crustacea and to {Trilobita; by Rolf Siewing. 108 the notostracan phyllopods, the posterior segments should have poorly developed limbs or no limbs at all. It seems that there exists a general tendency in a series like this to get more and more imperfect segments further back. Numerous parallels can be found among polychaetes. Such an arrangement appears to me more primi- tive than the arrangement in malacos- tracans. The mandibular palp was specially mentioned; it is true that there is no mandibular palp in adult branchiopods and cephalocarids, but it is present in the larvae, and it is well developed in many maxillopods and ostracods. A_ primitive heart is found in the phyllopods. The com- pound eyes are more highly differentiated in the malacostracans than in any other Crustacea. I think that e.g. cephalocarids, many branchiopods and, among the maxil- lopods, at least the mystacocarids are at a more primitive level of organi- zation than the Malacostraca as we know them today. But the Malacostraca may have separated from other crustacean evolutionary lines at an early stage of differentiation and have certainly retained various primitive traits. SANDERS: Claus (1876, Untersuch- ungen z. Erforschung d. genealogischen Grundlage des Crustaceen-Systems. Wien) long ago pointed out that the malacostra- can palp was secondarily derived; from ontogeny we know that there is a biramous palp in the nauplius. The palp then dis- appears at the end of the naupliar series resulting in a palpless mandible. At a later developmental stage the characteris- tic uniramous palp appears. SIEWING: Examples of the temporary absence of an organ in the embryogeny or morphogeny are distributed throughout the animal kingdom (e.g. lophophores in phoronid larvae, and in Bryozoa the loss MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 of the gut during metamorphosis). A temporary disappearance of an organ does not necessarily mean that if it re- appears it is to be considered secondary. SANDERS: In this case, however, the palp in the nauplius is terminal. The secondarily acquired uniramous palp ap- pears much more proximally on the limb, fairly close to the molar process. I inter- pret this to mean that these are not homologous. SIEWING: We have more extensive translocations in the ontogeny of animals. For example, one would have to conclude on the same basis that the carapace of cirripedes is not homologous with that of other Crustacea. Another example may be derived from the Bryozoa (Membrani- pora), where the gut disappears in ontogeny and reappears after metamor- phosis from another germ layer and in another position. Nobody doubts that the gut of larva and adult is homologous. MANTON: The embryonic mandible of Nebalia and of Hemimysis is long and finger-like. The gnathobasic part grows out in late stages and the distal part re- mains as the palp. There is no degenera- tion of a palp and a regrowth. SANDERS: I’m thinking of the pe- naeid decapod developmental series—the naupliar, protozoeal, mysis and postmysis series. GLAESSNER: Are the stomatopods primitive or could they have arisen from a modification, not of any of the exist- ing Malacostraca, but of some primitive Malacostraca which we know only as fossils? They have a peculiar habit; their locomotion is different from that of other Malacostraca, and they developed their abdomen perhaps more strongly in con- nection with these habits and locomotion. The division of the head may have some- DISCUSSION thing to do with their habits, rather than being altogether a primitive feature. I cannot see, however, where the amphipods would come from other than somewhere near the origin of the isopods. Siewing’s diagram does not place them far from the origin of the isopods, tanaids, and cumaceans. Again, there is a peculiar habit of locomotion in amphipods. This group did not differentiate into anything, as did the decapods or isopods, but never- theless it was successful. On a different point, Heldt (1954, Bull. Soc. Sci. Nat. Tunis, 6:177-180) has called attention to similarity between a penaeid larva and Walcott’s +Waptia. This might have been mentioned when we discussed the question of whether the Malacostraca were more primitive than we think. ROLFE: The latter idea did not origi- nate with Heldt; both Fedotov and Hen- riksen (see references in Rolfe, 1962, Breviora M.C.Z., 160:5-6) discussed the similarity of +Waptia to a penaeid pro- tozoea. As they and others have pointed out, the adult must have been large if +Waptia was a larva, and indeed large carapaces are found in the Burgess Shale which could be regarded as the adults. But again, as Manton has pointed out, we need to look at it and see how close is the similarity. The general facies of it is the same but, if we are to believe Stgrmer, +Waptia has a trilobite limb again, which is far from any protozoeal limb. HESSLER: If the Malacostraca is the most specialized and advanced subclass of the Crustacea, why is it alone in having the supposedly primitive abdominal ap- pendages? Either this means that the primitive crustacean from which all sub- classes were derived had a larger series of appendages than we find in the so- 109 called entomostracan groups, or it means that a segment which has lost its append- ages through evolution is capable of getting them back again. GLAESSNER: I don’t think that it would get them back. Siewing’s Figure 28 shows two groups of appendages, and they don’t work together. They serve different purposes at different times: pleopods for swimming in the swimming forms are adapted to different functions where the abdomen is reduced. The pereiopods have acquired specialized functions. Preserva- tion of abdominal appendages in the Malacostraca does not exclude the pos- sibility that they are the more advanced Crustacea. There is much scope for reduc- tions in the lower Crustacea, and the pres- ervation of certain primitive features in the Malacostraca. LOCHHEAD: On the question of whether a limb once lost could ever be regained, some interesting findings were reported by Etienne Wolff in a lecture at Woods Hole in 1961. Wolff stated that in the embryonic limbs of birds and other vertebrates it is possible to cut out some of the tissue which would develop into a dominant digit. When this is done, a sup- pressing effect is removed, and digits which have long since disappeared in evo- lution may grow in a quite normal fashion. Thus the genes for the vanished digits are still there, despite the seeming total loss of the phenotypic structures. GLAESSNER: Is it really agreed that at any stage in crustacean evolution the telson could have been lost? In the decapods the abdomen is reduced in a most extreme way and yet the telson cannot be lost. DAHL: I never intended to imply that the telson was lost, just that segment for- mation stopped at an earlier stage. 110 SIEWING: Have you any model or perhaps an embryological clue in the Mala- costraca for this assumption? DAHL: Only that crustaceans appear with large variations in the number of segments. We have short crustaceans, such as the ostracods where limbs go all the way back and with few segments, the MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 copepods which have few segments but no limbs on those posterior segments which should correspond numerically to segments not so far back in Malacostraca, and some of the phyllopods or anostracans, that have a far greater number of limbs than the Malacostraca. PHYLOGENY AND EVOLUTION OF CRUSTACEA Museum or CoMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 IX Jaw Mechanisms of Arthropoda with Particular Reference to the Evolution of Crustacea’ By British Museum (Natural History) and Queen Mary College, London The elucidation of phylogenetic rela- tionships within the Arthropoda is essen- tially speculative since the fossil record is non-committal. The facts concerning jaw mechanisms of arthropods which have recently been worked out bring to light considerable evidence concerning inter- relationships and phylogeny within the Arthropoda which is not of a speculative nature. Arthropods can be classified on their head structure alone and the form and use of the mandible together with the anterior sensory equipment is all important in determining the various types of head structure. Many widespread assumptions concerning arthropod jaws have _ been found to be untenable. The most im- portant of these are: (1) that the man- dibles in all classes correspond, (2) that biting in the transverse plane is a primi- tive arthropod attribute, and (3) that many mandibles are worked by simple adductor muscles without any abductors, the abductor force being, supposedly, elasticity of the cuticle and/or hydrostatic pressure. 1 A full account of this work, which covers the myriapods and insects, will appear in Phil. Trans. Roy. Soc., London, Ser. B. 111 It is important to use living animals in order to determine the exact movements which are employed, and full use of tech- nical methods is essential in elucidating anatomy. ‘“‘Generalised” simplified anatomy as a basis for consideration of jaw mech- anisms is most misleading and unsuitable. The head endoskeleton of most Arthropoda has been inadequately described. The form and function of head apodemes and segmental tendons, their movement and/ or rigidity, play an integral part in the feeding mechanisms of many animals. The study of jaws must of necessity embrace a comparative survey of head endoskeleton and its associated musculature as well as the skeleto-musculature of the jaws them- selves. A comparative detailed study of the anatomy and mechanism of move- ments of jaws shows clear evolutionary trends and in many cases points out plainly which are the more primitive and which the more specialised types. It is generally agreed that an ability to tackle large and hard food is an advance on the feeding on soft food. The strongest biting takes place in the transverse plane but this does not appear to be a primitive attribute in either Crustacea or Hexapoda. 112 Only in Limulus (king crab) and the Myriapoda does such biting appear to be primitive and it has certainly been in- dependently evolved in both these cases. It is not usually appreciated that ap- proximation and parting of paired coxae or coxal endites does not necessarily need any specific adductor or abductor muscles or an axis of movement supported by more than one close union of the jaw on the head. When a walking or a swimming leg swings about an axis situated exactly in the transverse plane the promotor and remotor movements do not alter the dis- tance between the coxae. It is immaterial whether the axis of swing lies horizontal, i.e. on the ventral side of the body, or whether the axis lies obliquely up the side. But if the lateral end of this axis lies a little posterior to the mesial end, as in the walking legs of an iuliform dip- lopod, the remotor swing brings mesially directed coxal spines or lobes together and the promotor swing parts them. Such movements can also be seen in some branchiopod thoracic limbs, and it is clear that an exploitation of this effect has led to apparent adductor-abductor move- ments of incisor processes which are effected by promotor and remotor muscles, very little direct coxal adduction or abduc- tion taking place (Fig. 50). Two movements seen in typical ambu- latory limbs have been utilised in the evolution of jaw mechanisms. (1) The common promotor-remotor swing of the coxae on the body and (2) the direct pre- hensile movement of a telopodite or telo- podite and coxa causing gripping or ad- duction in the transverse plane. Type (1): the promotor-remotor swing appears to have been used by the Crustacea and Hexapoda in mandibular evolution giving first a mandible capable of scratching, MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 squeezing or grinding small particles with- out biting and being modified secondarily to give strong holding and cutting in the transverse plane. Type (2): leg move- ment has been exploited by the Myriapoda and Chelicerata in giving direct transverse biting. Mechanical difficulties have confronted all arthropods during their attainment of biting jaws capable of dealing with hard and large food. Adduction is easy but ab- duction of mandibles occupying the whole head width presents important mechanical difficulties which have not previously been recognised. These microengineering prob- lems have been resolved in different ways by the various classes of arthropods. Some- times the resolutions are entirely different one from another and must indicate in- dependent phylogeny. Sometimes the reso- lutions show convergent resemblances superimposed upon different basic mor- phologies, and, as with locomotory and burrowing mechanisms, the anatomical changes which become associated with each type of resolution of the biting prob- lem become evolutionary one-way streets. The ontogenetic development of jaws in arthropods shows that they arise in two different ways. The present-day series of mandibles of larval and of adult Crus- tacea shows the unquestionable derivation of the mandible in this class from the proximal endite or gnathobase and proxi- mal part of the coxa, the distal telopodite in the adult being reduced to a bi- or a uniramous palp or it may be absent. Bit- ing is similarly done by the leg base in Limulus and in the Arachnida and the embryonic derivation is the same. In the Onychophora-Myriapoda-Hexapoda assem- blage, the mandible, on the contrary, is formed from the whole of an embryonic limb. In the Onychophora the two terminal MANTON: JAW MECHANISMS jaw blades resemble the two claws on each walking leg. The unsegmented mandibles of insects are developed from a _ whole limb rudiment, the most distal part form- ing the gnathal armature or mandibular tip. Most of the myriapod classes possess jointed mandibles, a feature correlated with the manner of use of the mandibles and providing flexures serving adduction in the transverse plane, as in a telopodite. There is no reason to suppose that myria- pod and insect mandibles are not of the same whole limb type although the dif- ferences in the mechanisms of movement and in details of morphology are so great as to preclude the origin of the one from the other. That a basic promotor-remotor swing has been employed by the gnathobasic mandibles of Crustacea and the whole limb mandibles of hexapods is seen to be a parallel evolution, and the basic dif- ferences in mandible derivation of these two groups must indicate the independent evolution of mandibles within them. Bit- ing in the transverse plane has been in- dependently evolved several times within the Crustacea and also within the hexa- pods. There are great differences in the use of the gnathobases in Limulus and in the Crustacea, differences which suggest a deep-seated lack of affinity between the Merostomata and Crustacea indicating that jaw evolution has proceeded inde- pendently in these two groups also. THE MORE PRIMITIVE TYPES OF CRUSTACEAN JAW MECHANISMS A survey of the mandibular mechanisms of the more primitive Crustacea shows a surprising uniformity in basic form, those of Chirocephalus and Daphnia among the Branchiopoda and Hemimysis among the Malacostraca are extraordinarily alike 113 and primarily serve the grinding or squeez- ing of soft fine food. Greater biting ability is present in Paranaspides and Anaspides.” The molar areas of these mandibles do not function by a basic adduction of the jaws from their single dorsal point of closest union with the head. The primary movement is a derivative of the promotor- remotor roll of a walking leg. Figure 45 shows a side view of the mandible of Chirocephalus (A) at the end of the promotor forward swing and (B) at the end of the remotor backward swing, the axis of movement being dorsoventral. The latter movement rolls the molar areas forwards and across each other as indi- cated by Figure 46. The musculature caus- ing the promotor-remotor swing is shown in Figures 46 and 47. The head endo- skeleton consists of a wide transverse mandibular tendon, anchored by struts to the cervical groove and linked with other units developed from basement membrane. The remotor muscles are the longer, larger and stronger; they comprise muscles 4, Sa and the direct muscle 5c uniting the posterior margins of the mandibles. The promotor muscles 3 and 5b roll the molar processes across each other in the opposite direction and can cause slight abduction owing to the exact shape of the mandible. Food arriving at the point marked x on Figure 46A, by a route situated close to the ventral surface of the body (see arrows on Figure 49E) emerges at y, Figure 46C, and is sucked into the mouth. The fre- quency of the mandibular grinding move- ments is much less than that of the fil- tratory trunk limbs. At times slight direct abductor move- ments occur which part the mandibles. * Paranaspides and Anaspides are essentially similar in the aspects of mandibular mechanisms here considered. 114 The dorsal point of close union of the mandible on the head is not a firm articu- lation; the mandible can slide up and down against a small sclerotised rib situ- ated laterally in the cervical groove (Fig. MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 48C). Muscles 5d and 5e can strongly adduct the mandible during the remotor swing. Contraction of muscles 6 and 5f and relaxation of muscles 5d and 5e result in sliding the dorsal part of the mandible CHIROCEPHALUS cervical groove \ ~. s maxillary __ SUNG og land@acaiiy! \ Noes en y a: NS oe i . eye antenna | a \\ dorsal articulation of mandible ~ antenna 27 \: thoracic leg Fic. 45. (A) Lateral view of the head of Chirocephalus diaphanus Prevost showing the mandible in its position of extreme forward roll (promotor swing) about a dorsoventral axis. The only point of close union with the head lies dorsally at the black cross (see Fig. 48C). (B) The eye and first trunk limb are in the same position as in A, but the mandible is at the end of the backward roll (remotor swing), displaying the straight anterior mandibular margin and the inward and forwardly directed molar process. (C) Transverse view of the mandible on the body. 115 JAW MECHANISMS MANTON “BUIMS AIOAOIAI OY} B}JOWLOId Gg sapsnuI {19430 yord JSUIeSe PUL SPAILMIOJ SLIIe IL[OW dy} SBUIMSs YOTYM ‘D-VY UT Uses se YONs ‘a[qIpuelM oy} Jo Javed [esI}e[ BY} JO [[OI IOJOWIIT preMyIeq oy} asnvo Bs SapSNT “aUIT payjop ev Aq po}eoIpUr Buteq ‘vare Ie[OUI oy} SurTIvsq ‘a[qipueW oy} Jo UOTJIOd [eAZUA daI¥ BY} JO UOT}ISOd dy} ‘aAoqe WoT, PaMmaIA snyoydar0uy J JO IqIpuRUI yYBI ay} JO UOTJDes [eJUOI v SyUaSaIdoy (CZ) “MOIA [eJUOIY UT o[qIpUeUT UROvIJsOUe Ue SuT[quiaseI sadeys prevoqpies jo aed & jo sjied [e19}e[ ay} JO SIXe [eOT}AVA & JNOGe OS YSno1Yy} Sums 10}JOWeI plwMyeq B JO Jaya oy} MOYS 0} SUIRIBeIq (D-VY) 9b “OIY JuswsAoW jo SIxe U 4] , qd Joliaisod (s022nppe ) Jojowas 4ejnqipuew SE ah ee ese Jolsaysod JOlajue JOljajue JUsWIAOW JOJOWSI 116 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 CHIROCEPHALUS cervical groove mandibular promotor ( abductor ) mandibular remotor ( adductor) 4 = Jremotor (adductor) 4a Qe QQ transverse mandibular tendon mandibular adductor § ¢ longitudinal endoskeleton mandibular adductor 5 ¢ nerve cord A mandibular promotor ( abductor ) B 73 B endoskeletal suspension maxillary gland = longitudinal endoskelecon Bend of transverse mandibular tendon } _ paragnath \ \ mandible transverse mandibular remotor( adductor) 5 c \ nerve cord Fic. 47. Transverse views of the head of Chirocephalus diaphanus passing in front of and through the mandible at progressively more posterior levels. A-B lies just in front of the mandible. Level A shows the oesophagus (stippled) uniting the mouth with the foregut and the cut longitudinal bar of head endoskeleton formed by an elaboration of the basement membrane. Level B passes through MANTON: JAW MECHANISMS slightly downwards and inwards, so ab- ducting the molar areas. Biting in the transverse plane by more specialised Crus- tacea is not developed from these small adductor-abductor movements mediated by muscles 5d, 5e and 5f and 6, but by a modification of the much stronger muscles causing the promotor-remotor roll. In Hemimysis the axis of promotor-re- motor swing is no longer vertical; the dorsal end is tipped backwards a little, a position further accentuated in Para- naspides and Anaspides (Figs. 49F and 51). The axis also lies near to the anterior margin of the mandible. An incisor process used for biting projects from the mandible far from the axis. The promotor-remotor swing of the mandible now not only rolls the paired molar processes across each other as in Chirocephalus, but the shape of the mandible is such that this rotation results in approximately transverse biting by the incisor processes, as shown in Figure 50 for Anaspides. The musculature is little changed, promotor muscles 3 and 5b and remotor muscles 4, 5a and 5c correspond with those of Chirocephalus, an additional remotor muscle 2 pulls out- wards from the anterior margin of the 117 mandible, and remotor muscle 4 is much enlarged arising from a strong posterior mandibular apodeme. Muscle 6 arises in Hemimysis, as in Chirocephalus, near the dorsal apex of the mandible (Figs. 47D, 53C). This muscle arises from the anterior border of the mandible, lower down in Paranaspides (Fig. 51), where it forms an. effective promotor inserting on the tendinous endoskeleton as in Hemimysis (Fig. 53C). The head endoskeleton of Paranaspides and Anaspides is mainly tendinous as in Chirocephalus but more elaborate, consisting of the segmental tendons of mandible, maxillule and maxilla linked by paired longitudinal bars, the whole anchored to the body wall by many struts (Fig. 52). The shift in the position of the axis of promotor-remotor swing in these Syncarida enhances the apparent adductor-abductor movement of the incisor processes while still maintaining adequate grinding be- tween the molar areas which lie close to the axis of movement. It is clear that (1) the promotor-remotor swing of the mandibles of Chirocephalus and Hemimysis is caused by antagonistic muscles; (2) that minor direct adductor the origin of this endoskeleton from the lateral basement membrane, the transverse connection be- tween the longitudinal endoskeletal bars, and their fibrillar suspension from the dorsal body wall. The mandibular promotor muscle 3 inserts dorsally on to a cone of basement membrane. Two lobes of digestive gland are cut. Level C shows the anterior face of the mandible. Level D passes through the middle of the mandible and its dorsal articulation (see Fig. 48C), the posterior margin of the mandible being visible behind. The median connection between the mandibular tendon and the endoskeletal plate uniting the paired longitudinal endoskeletal bars lies on the posterior side of the mandibular tendon, and is shown also in level E. Level E passes through the posterior margin of the mandible just in front of its union by ample arthrodial membrane with the head. The endo- skeletal link between the inter-paragnath groove and the transverse plate uniting the paired longi- tudinal endoskeletal bars lies well behind the mandible. Muscles 3, 4, 5, 5b, 5c and 6 are represented in Paranaspides and Anaspides (Figs. 51-53). The prefixes to muscle numbers “adductor” and “ab- ductor” signify muscles causing direct movements in the transverse plane; “remotor (adductor)”’ and “promotor (abductor)” signify muscles primarily causing the remotor-promotor roll which facil- itate the rolling together of the molar areas in Chirocephalus or secondary transverse biting in derived crustacean types (Figs. 51-53 and 55-57). 118 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 \ mandibular suspension ; \ aN -> A a basement membrane of. oN oN ca, @ cuticle mandibular suspension PETROBIUS mandible cuticle a E BSS H —— ; 4 B pleural fold == = rt I Ce x HEMIMYSIS ANN MW, mandibular muscle 4 6 Vag < eaves basement membrane + ~“uU! Le wage] djed sejngipuew S3IdSWNV x 121 JAW MECHANISMS MANTON ‘gpiynd Ie[nqipueur 24} JO APAVIUOD dy} 0} ‘ATAATIOadSaI ‘spxeMIOJ PUB SpIeMjNO pu spIeMYPeq pue spreMyno ssed q¢ pu BS sopsNU JI WoT, {yoR[q ur uUMOYS sT uOpud} Iv[NqIpuvU IsIOASUPI] BY} Jo JAed [eII}Y[ OY, ‘adpa AL[NGIpuvU oY} Iv9U JULIGWIDU [eIPOAY}Ae WOT] aslIe g apIsnuUI JO Jed pue ¢ apsn]T ‘9S QPISNUL ISIOASUBI] SASIIB DIF Av[nqipueW ay} Jo JAed AOTIajsod pue aspa JOTIo}sod ay} WoA] pur ‘p spsNU [e19}v[OSIOp Surresq swaepode Moyjoy ay} 0} UISIO SaAIS a[qIpURU dy} JO UTSIeW JOTIA}sod ayy, ‘[[nd 9 pue ¢ sapPsNU YIYM 0} UO JIqIPUeUT 9Y} JO UISIvUI JOTIO}Ue 9Y} SMOYS DUT] po}}0p AYSY IL ‘JUIWIAOW BuT[OI oy} Jo sixe oy} syreu Jods yorlq pue ssold oY} UdaMJoq OUTT poop oYJ, ‘auIT AAvaY vB Aq UMOYS ST UIsIeW ddIF esoyM ploy [einajd MoyeYys ay} Japun pay pue s[qIpuvuU dy} JO UOTUN [RSIOP VSO] 94} SdJVIIPUT ssoId OY ‘syUOWIBAOW AR[NGIpuew [edourid 94} JO UOTPIIIP 94} PUB SapsNUL Iv[NqIpueUT JO suOT}Isod 9Yy} MOYS 0} UOSMIOY]L saD2UDUSD] SapidspUp JO peoy IY} JO MOIA [eIO}eYT “TS “OLY ssar0ud 40s!9uU! ‘ajqjpuew uopual ueipaw wor (4039NpPpe) soOWA4/ sejnqipuew ix Lil (4o32Nppe) 1020W94 yejngipuew assadsuesa Ba) 21221043 S| “ ajqipuew jo ugsew soiajue PJO} jesnajd euuoque awapode sejnqipuew Jolia3sod 2a0043 jesare] Yo ZL. - a]Nuuaaue Bz YY Than ‘s 9 f : aiqipuew JO ulZsew solsaaue Bs: wo. (4oinpge) soi0wosd sejnqipuew z ujZsew soliajue Wo.y( soinppe) soiowas sejngqipuew 5 ( ppe) Ingip € + 2A0018 JBD1A19> aueIquaW jeIposyrzse wos (4oIINpge ) soyowWoJd Jejnqipuew (s022Nppe ) soyowas 4ejnq!puew SIGIdSVNV 122 and abductor movements occur in Chiro- cephalus and Hemimysis, and the tilt of the axis of swing, more marked in the Syncarida than in Hemimysis, results in the promotor and remotor muscles in- directly causing adductor and abductor movements of the incisor processes. These crustacean mandibles are not worked by only three muscles comprising no abduc- tors, as claimed by Snodgrass (1950), and no elastic forces reside in arthrodial mem- branes sufficient to cause either abductor or promotor movements. Only sclerotised cuticle can provide such a force and arthrodial membranes linking the mandible to the head are typically lacking in sclerotisation which, if present, would be a hindrance to the effective movements. In certain cases sclerotised cuticle does provide a mechanism for recovery move- ments, e.g. abduction of the maxillule by the elasticity of the maxillulary sclerite (Jackson, 1926). TRANSVERSE BITING IN CRUSTACEA Strong transverse biting in Crustacea has been evolved many times, but by dif- ferent means, and two examples will be considered below, that of the Isopoda, of particular interest because of certain super- ficial resemblances to hexapods, and the resolution adopted by the crayfish Astacus fluviatilis (Potamobius astacus L.) and the crab Carcinus maenas (L.) among the decapods. Ligia is a_ large-food feeder capable of biting into food masses and swallowing sizeable particles via a wide oesophagus, the route for the food being from below upwards (see lower arrow in Figures 49F, G; cf. the narrow oesophagus of fine-food feeders, Fig. 49E). The axis of promotor-remotor swing of the mandible, already sloping backwards in the Syn- MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 carida, is approximately horizontal in Ligia and the pre-axial part of the man- dible is short, forming a very robust hinge extending along the dotted line between the cross and the black spot on Figure 54. The promotor-remotor swing now becomes a wide adductor-abductor movement in the transverse plane, the gape being also wide (cf. small gape in fine-food feeders). The musculature moving the mandible is a derivative of the primitive malacostracan type, and the head endoskeleton is more robust. The small pair of maxilla 1-2 apodemes of a mysid or Paranaspides is very large in Ligia, forming paired bars (the sternal alae of Jackson) extending through the head, and anchored on to the surface cuticle, entirely replacing the tendinous endoskeleton, which consists of transverse mandibular, maxilla 1 and maxilla 2 tendons and their longitudinal connections in Hemimysis and Anas pides. Reference to Figures 47 and 53 shows that the presence of a large mandibular tendon linking transverse mandibular muscles must inhibit wide abduction since great changes of length in the transverse muscles would be necessary if the gape should be wide. The elimination of the transverse mandibular tendon and _ the reduction of the transverse muscles aris- ing from the mandibular concavity to a remnant (5a on Figure 56A), free the mandible of this restriction and permit the employment of a wide gape. Remotor muscle 4 in Ligia is very elaborate, aris- ing in many sectors from a large and com- plex apodeme corresponding with that of the lower Malacostraca. These muscles fill a major part of the head cavity (Figure 55). The abductor mechanism is very ingeniously contrived. A long thin apo- deme leaves the pre-axial margin of the mandible and slopes upwards and out- MANTON: JAW MECHANISMS wards into the “cheek” (Fig. 54 and left side of 560A), and from the tip of this apodeme four small sectors of muscle 3 slope towards the middle line (Figs. 55 and 56A). Contraction of muscle 3 pulls the tip of this apodeme slightly towards the middle line and this movement results in wide abduction of the incisor processes and the extension of muscle 4. The remains of the transverse muscle 5a inserts now on the longitudinal head apodeme and serves as a weak adductor pulling at a rather poor mechanical advantage. The lateral endoskeletal node mandibular horacic Zndichocac remotor {adductor | 4 tergite Vv iv endoskeletal connective nerve cord maxillary gland Ist thoracic leg maxillulary muscles endoskeletal union with exoskeleton mandibular apodeme depressor to antennule Paragnath 123 molar areas are still present on the mandible but their triturating ridges are restricted to the proximal edges which alone come in contact as the mandibles swing together. Thus transverse biting and a wide gape have been achieved in the Isopoda by the adoption of a horizontal position of the axis of swing of the mandible, a dissolu- tion of the segmental tendon system of the head, together with most of the transverse muscles, and by the great elaboration of promotor and remotor muscles 3 and 4, ocular muscle depressor to antenna ii levator to antenna levator to antennule antennule < A Gm (eaereaina caconeN bar \ Repo *\ fs \ iy antenna mandibular remotor (adductor) 3\ from anterior margin ot mandible mandibular promotor (abductor) from anterior margin of mandible 6 mandible, molar process mandibular promotor (abductor) from anterior margin transverse mandibular tendon mandible , incisor process Fic. 52. Sagittal half of the head of Anaspides tasmaniae Thomson with the alimentary canal and digestive gland removed to display the cephalic endoskeleton and mandibular musculature. The posterior mandibular apodeme and the transverse segmental tendons and their connectives are shown in white, the median endoskeletal bridges cut in the sagittal plane are hatched. The positions of the mandibular margins and of the axis of movement between the dorsal union with the head (a cross) and the black spot are indicated. The web of fibrous connections between the transverse mandibular tendon, the lateral longitudinal tendinous bars and the three dorsolateral struts are more complex than shown. The levels A-D mark the planes of sections shown on Figure 53. 124 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 PARANASPIDES mandibular promotor ( abductor ) ocular muscle levator to antennule pleural fold paired longitudinal depressor to antennule endoskeletal bar mandibular Esau promotor (abductor transverse mandibular tendon mandible , molar process mandibular’ Temotor (adductor ) A labrum mandible , incisor process B i tor ( adductor ocular muscle mandibular rem ( )4 antennal levator mandibular labrum remotor adductor ) mandibular mandibular palp promotor (abductor ) ? + mandibular promotor “> (abductor) 56 IES depressor to antennule mandibular apodeme paired longitudinal endoskeletal bar a mandibular remotor (adductor ) transverse mandibular tendon transverse mandibular remotor ( adductor ) Se mandible transverse mandibular 5 ¢ remotor (adductor ) paragnath Fic. 53. Paranaspides lacustris Smith. (A-D) Thick diagrammatic transverse sections of the head at the levels indicated on Figure 52, at A-D, to show the mandibular muscles and cephalic endo- skeleton. The position of the axis of movement passing from the dorsal union of the mandible and head, marked by a cross on D, to the black spot on A is projected on to the other figures which show only part of the obliquely situated mandibie. The designations (adductor and abductor) to the muscles shown are made in the sense described in the text; these do not signify a basic adduction and abduction in the transverse plane, but a remotor-promotor roll. Level A shows the cuticular unions ii and iii of the paired longitudinal tendinous bars, iii being close to the oral angle braced by the termination of the transverse mandibular tendon shown in level B. Level B shows the lateral longitudinal tendinous bar, the anterior dorsolateral mandibular remotor (ad- ductor) 2, and the anterior extension of the transverse mandibular tendon. Level C shows the trans- verse mandibular tendon bearing remotor (adductor) and promotor (abductor) fibres (see labelling on level B), the transverse union of the paired longitudinal endoskeletal bars at the level where they bear promotor (abductor) 6 (see the labelled dotted outline of the origin of the muscle in Fig. 52), and promotor (abductor) 3 arising mainly from the arthrodial membrane be- low the carapace fold. Level D shows the longitudinal endoskeletal bar and its tendinous suspen- sions iv and y, the hollow posterior mandibular apodeme with its dorsolateral remotor (adductor) 4, and the transverse mandibular remotor (adductor) Sc lying behind and independent from the mandibular tendon. (E) Oblique horizontal section just below the mouth to show the mandibles and their molar processes at the end of the remotor roll, the transverse mandibular tendon, which sup- ports the lower end of the axis of movement (black spot), and the transverse muscles. MANTON: JAW which now become, in effect, abductor and adductor in function. The triturative ability of the mandibles, favoured by a vertical position of the mandibular axis of swing (as in Chirocephalus), is decreased by a slope of the axis shifting towards the horizontal, a position giving maximum efficiency to incisor process biting by muscles pulling from the same side of the hinge as the cutting blades. LIGIA occipital groove tergal ala Ist thoracic tergite MME a = y aa sunion of endoskeleton and sclerite s dt alar bar maxillipedal a, Ist thoracic leg maxilla maxillule Fic. 54. Ligia oceanica Roux. MECHANISMS 125 The decapod transversely biting man- dibles are derived from the same _ basic malacostracan pattern seen in Hemimysis and Paranaspides, but lever principles are employed and different muscles from those of Ligia become the effective abductors and adductors. A strong hinge is present but the cutting blades and the muscles supplying the adductor force are situated on opposite sides of the hinge in contrast to Ligza. marginal line frontal line supra-antennal line antenna mandibular abductor apodeme “clypeus* labrum mandible paragnath Lateral view of the head with the pleural plate and leg of the second thoracic segment cut short. The heavy dotted line marks the axis of movement of the man- dible, passing through the same morphological points marked by a cross and black spot as shown for Chirocephalus, Paranaspides, and Anaspides on Figures 45, 47, 49, 51, 53. Arthrodial membrane is indicated by white stipple on black. The “tergal ala’? apodeme of the head is visible through the arthrodial membrane passing to the second thoracic segment, here stretched a little. The notch, marked by a small circle, carries a cuticular thickening, suspending: the maxillipedal sclerite, the alar bar of the superficial sclerite system of the head wall, and the inferior lateral pterygoid process which forms one arm of the base of the “sternal ala’—a principal head apodeme (see Fig. 55). 126 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 LIGIA i remotor (adductor mandibular promotor ( abductor ) to cranium mandibular ( ) 4 mandibular promotor ( abductor) to marginal line 3 tergal ala __s frontal line exoskeletal muscle ——_ é antennule mandibular. abductor apodeme — antenn: mandibular remotor ( adductor ) to tergal ala b mandibular adductor apodeme antennal apodeme paired longitudinal - endoskeletal. bar sternal ala : union of endoskeleton and sclerites maxillipedal sclerite Inferior lateral pterygold process alar bar a . maxillulary basal sclerite clypeus oesophagus mandibular molar area edge of mandible maxillary pterygoid process median bar mandibular remotor (adductor) 5a ~ Ist thoracic leg labrum Jacinta mobills median lobe paragnath mandible, incisor process maxillule Fic. 55. Ligia oceanica Roux. Sagittal view of left half of the head to show the longitudinal apodemal endoskeleton, the mandible and mandibular muscles. The mandibular cuticle is drawn alone in Figure 56, B from the same aspect. The cut through the ventral cuticle of the head passes along the “median bar” (a superficial scute), and one arm is shown of its anterior bifurcation which ends on the paragnath, the cut end of the other being hatched. The antennal muscles passing to the genal fossa region of the cranium (between the circle and cross on Fig. 54) are removed; they pass between the mandibular adductor and abductor muscles. The extrinsic muscles of the maxillule and maxilla from the tergal ala and sternal ala (see dotted arrow) and the maxillipedal muscles from the maxillipedal sclerite and alar bar are also omitted. The mandibular hinge on the head lies between the cross and black spot. The broad mandibular adductor apodeme leaves the para- sagittal edge of the mandible and passes directly upwards (see Fig. 56, A). The narrow mandibular abductor apodeme leaves the anterodorsal edge of the mandible close above the hinge and passes outwards and upwards (see left side of Fig. 56, A). The mandibular adductor muscle forms three main sectors; a posterior sheet passes from the proximal posterior edge of the apo- deme upwards and backwards to the tergal ala. The most bulky sector fans out from the apex of the apodeme to the whole dorsal and anterodorsal face of the cranium above the “frontal line” (half of it is cut away to expose the abductor muscles); and an anterior sector arises from a tuft of fibres half way along the anterior edge of the apodeme and inserts on and around the “supra-antennal line” (see Fig. 54) of the cranium. The mandibular abductor comprises 4 sectors: two pass to the dorsal cranial wall just median to the eye; one passes to the “marginal line” (see Fig. 54) behind the eye; and another to the anterior face of the tergal ala. The principal head apodeme the “sternal ala” arises from a maxilla 1-2 intucking (white stipple) bearing the inferior lateral pterygoid process and the maxillary pterygoid process in its outer and inner angles, respec- MANTON: JAW MECHANISMS Figures 57A, B, show side views of the mandibles of the crayfish and crab with the lateral parts of the head removed as far as the union of the mandible with the head. The axis of swing of the man- dible, marked by a dotted line, is oblique and in both animals allows biting in the transverse plane. In the crayfish the main hinge is formed by the elongation of the dorsal articulation along the anterior margin and there is no freedom at the point marked by the black spot as in the weaker rolling type of mandible (Chiro- cephalus and Paranaspides). In the crab the dorsal articulation is very weak and a strong anterior articulation near the black spot lies between the epistome and man- dible at the base of the mandibular palp. The muscular system of Hemimysis and Paranaspides is clearly recognisable in decapods but additional apodemes are present. A small one projects from the anterior margin of the mandible in the cray- fish and a much longer one in the crab (white on Figure 57A to D). In the crab the anterior ventral position of the hinge enables the morphologically dorsal part of the mandible to be in line with the anterior apodeme so that the two form a very long lever (as noted by Snodgrass, 1950) which is responsible for the strong bite of the crab. In the crayfish, muscle 4 remains the principal adductor, but becomes very small in the crab. Remotor (adductor) 2 (adductor lateralis mandibulae of Schmidt, 1915) from the anterior apodeme passes outwards to the lateral head wall in both decapods, pulling at a much more ad- 127 vantageous angle than in Anaspides and Paranaspides (Figs. 51, 53B and 57D) owing to the presence of the anterior apodeme. This muscle in the crab is very large and pulls from a tendinous flap set at a sharp angle to the tip of the apodeme, so providing plenty of surface far from the hinge for the origin of muscles 1 and 2. Muscle 2 of Hemimysis and Paranaspides corresponds with the separate adductor 1 and 2 of the crab, passing forwards to the front wall of the head (Figs. 57B, C, D). Muscles 1 and 2 in the crab are now the principal adductors superseding muscle 4. Part of transverse muscle 5a is present in the crayfish but inserts on the head endophragmal skeleton, no transverse ten- don being present, and muscle 5c is absent; these features facilitate the wide gape of the mandibles. In the crab muscle 5c is absent and 5a further reduced. Promotor (abductor) muscles 3 and 6 are present in crayfish and crab as in Paranaspides, but with the dissolution of the transverse tendon muscle 5b is absent. Muscle 6 leaves the anterior apodeme in crab and crayfish and passes inwards and_back- wards to the endophragmal skeleton (Fig. 57A, B, C, the “abductor major” of Schmidt, 1915, and the “internal abductor” of Pearson, 1908). In the crab muscle 6 is much more bulky than muscle 3 and forms the principal antagonist of muscles 1 and 2. The lever formed from the combined upper mandible and anterior apodeme in the crab is several times the distance be- tween the anterior mandibular articulation and the cutting edge. An extensive dis- tively. The maxillulary basal sclerite is attached laterally to the maxillary pterygoid process and can be seen through the membranous part of the apodemal intucking. The circle marks the union of the superficial scutes with the inferior lateral pterygoid process, a point also supported by the end of the tergal ala. Setae of maxilla 2 are drawn as if the median lobe were transparent. 128 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 LIGIA Ist thoracic tergite mandibular promotor (abductor ) to cranium 3 occipital groove mandibular promotor( abductor ) to marginal line tergal ala promotor ( abductor ) to tergal ala mandibular abductor apodeme exoskeletal muscle antennal muscles mandibular adductor 4 longitudinal endoskeletal bar union of endoskeleton and sclerites mandibular adductor 5 a A mandibular adductor apodeme paragnath bar mandible paragnath mandibular abductor apodeme mandibular adductor apodeme hinge cavity of mandible molar process _ spine row lacinia mobilis incisor process Fic. 56. (A) Posterior view of the head of Ligia oceanica, intact dorsally and laterally as far as the notch surrounded by the circle, marked as on Figures 54, 55. Ventrally the mouth parts are removed leaving the left mandible and the right paragnath and mandible. The prin- cipal head apodeme, the sternal ala, is cut, and its process bears the very small remains of man- dibular remotor (adductor) 5a (cf. Figs. 51 and 53.). (B) Sagittal view of the mandibular cuticle devoid of muscles seen from the same aspect as in Figure 55. 129 JAW MECHANISMS MANTON ‘sdjed aepnqipueur oy} wsemjoq wINIqe] 24} Sulsopsip uado Ay[ny soqipueu oy} YIIM qvid dy} JO UOIZaI yyNoW Jo MoIA [eIJUDA (J) “Gq pue @@ YIOq UL pauazoYysaIOF ov soposnuI s}I pue sqIpUeUT IY], “pray 94} YUM YqIpurUI oy} JO UWOTe[NIAe IOMII}Ue dy} YIM J9Y4}050} sapsnut S}I pue stuepode Av[nqipurul JoLajUe IY} MOYS 0} YOR ynd peasy dy} YM qeID dy} JO MOIA JOLIOJUY (C) “SeUT] peyop 9TYM Aq uMOYs SuIeq aqoy [eqyeus oy} Jo uorpsod oy} ‘sapsnut tay} pue sowapode Ay[nqipueul JoL1a}sod pue IOLIOJUe MOYS 0} qvID JY} JO MaIA [esIOG, (DQ) “qeAD oY} UI Z pue T SapsNUI 07 sWepode AL[NGIpULUT JOLI9}Uv JY} WOI SpIeMIO} sassed uopud} B {yOeq yn peay oy} YIM qed pue YSyAvIO Jo o[qIpueUI YSII dy} JO SMOTA etoyeT (gq pue YW) ‘429uDD IO} ‘ZO6T ‘UOsIvag PUR SNIDISP OJ ‘ST6T “PIULYDS das sapsnuI oy} Jo s[lejap Jay.INy Jog “gq pue YW uo sMosIe Aq UMOYS ST 9 0} T Saposnur jo [nd Jo uoTpaIIp sy, “2577 pur sapidspup ‘snyoygar0«yD YI UosteduI0d Oy (J) spUuapUM snu2Z4DD qeId dI0YS ay) Ul (A-q) pue ("T SnavIsp sniqowpjog) syviavy Snovjsp YsYARII dy} UT (Y) SJWIWIAOCU IIaY}] PUR saTqIpueUT oY} SuIMoys sweIseIq “LS ‘OI J eyjpcew d yanow awapode sejnqipuew id uo}zejN2}wWe JOjsaIUe JO wojajsod 401413350) djed sejnqjpuew euuajue uopuaa awapode sejngipuew JOlJaque ave Z (4022Nppe ) dojowWas sejngipuew 9 (40x NPge ) sooWosd sejnqipuew awapode sejnqipueW Jojaque | avd ( 4022Nppe) soyowad Jejngipuew 9 € sSioiDNpge eS ¥ TZ | sso3nppe sajsnw sejngipuew ‘ uonejndiie jessop TATA euuaiue if ( 4032nppe) 4oI0Was 4ejngipuew 7(soaonppe) 4o30wWas 9 (s012npge ) dojowoid uejngipuew } (4022nppe )yorowas sejngipuew awapode jejngipueW Jojjaque HSISAVYD avud 130 MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 retractor dorsalis lateral groove dorsal flange of coxa pleuro- coxa] articulation a Dx pleurite 5 pleurite 6 coxal endite coxa leg 5 sternite coxal endite muscle cavity of coxa A sternite muscle anterior posterior , Fic. 58. (A) The prosoma of Tachypleus tridentatus Leach cut transversely in front of legs 5 and viewed from in front to show the extrinsic limb muscles. The digestive gland is omitted; portions of the heart, pericardial floor and intestine are indicated but unlabelled; pedal nerves are stippled and arthrodial membranes are shown by white dots. Ventrally the anterior end of the sternite is passing upwards towards the oesophagus (see Fig. 59). The endosternite is cut trans- versely; its ventral surface and that of the ventral (pleural) surface of the carapace are fore- shortened. Muscle numbers are those of Benham (iz Lankester, Benham and Beck, 1885). Mus- cles 52 attach the dorsal cornua of the endosternite to the carapace and muscles 18 comprise the branchio-thoracic muscles which pass forwards from the branchial limbs to the prosomal cara- pace. The coxa of leg 5 is entire on the left, and on the right it is cut away as far as its articulation with the pleurite, disclosing the “dorsal flange” (marked) which is covered externally by MANTON: JAW MECHANISMS placement of the tip of the apodeme re- sults in a lesser but stronger displacement of the biting cusps. Large movements of the tip of the anterior apodeme can only be effected by long muscles (note the great length of muscles 1 and 2 in the crab compared with the size of the gnathal lobe, made possible by the wide carapace; Fig. 57C, D). Long muscles 1, 2 and 6 are easily housed in the wide cephalo- thorax. Their bulk is presumably quite sufficient for the needs because the head space could easily house more bulky adductor muscles. This contrasts with the locust and Ligia where the head capsule is filled to capacity by the mandibular adductor muscles. There is no reason to suppose that biting and holding by the mandible of a crab is in any way less efficient than that of isopods and pterygote insects. These two examples show how the basic musculature associated with the promotor- remotor mandibular roll of the primitive Malacostraca has independently given rise to strong transverse biting, the principal antagonistic muscles being 3 and 4 in Ligia and muscles 1-2 and 6 in the crab. Transverse biting has been independently evolved in other specialised Crustacea which will not be considered further here. 131 THE BITING MECHANISM OF THE CHELICERATA The Chelicerata share with the Crus- tacea the utilisation of pairs of biting and chewing mesially-directed coxal lobes or gnathobases, but the Chelicerata differ from the Crustacea in that their biting is done primarily by direct adductor-abduc- tor movements in the transverse plane and not, as in Crustacea, by a derivative of the anteroposterior swing of the coxa on the body used in locomotion. In Limulus both locomotory and feed- ing movements are carried out by prosomal limbs 2 to 5, and fairly soft but large food, such as worms and molluscs, can be shredded and eaten entirely with rapidity. The arachnids, on the contrary, are fluid feeders, chewing portions of their prey with the coxae of the pedipalps for long periods; the distal parts of these limbs are usually not locomotory. However, the coxal-feeding movements of the Xiphosura and Arachnida are essentially similar. When walking, each pair of prosomal legs 2 to 6 is moved in similar phase by a promotor-remotor swing on the body effected by muscles pulling on the anterior and posterior proximal coxal margins (Fig. 58, promotors 27 and 41a, remotors a fold of arthrodial membrane (see Fig. 60, A). On the right, one limb of the Y-shaped pleurite, with which leg 5 articulates, and the entire pleurite articulating with leg 6, are shown. The extrinsic muscles attached to the anterior thickened rim of the coxa are indicated in white on the left, some of the more posterior muscles being indicated in black. The extrinsic muscles attached to the pos- terior thickened rim of the coxa are shown on the right where the anterior rim and its muscles are cut away. The coxal endite is cut short on the right to show the sternite. (B) Anterior view of the mesial part of the gnathobase of leg 5 to show the heavy spines (with black tips) and the endite set in ample arthrodial membrane which allows the proximal rim of the endite to tip into the rest of the gnathobase on forward flexure of the endite (see Fig. 59, B). (C) Posterior view of the same to show the straight hinge between the endite and the rest of the gnathobase and the muscle extending from the endite to the posterior rim of the coxa. Contraction of this muscle re- turns the endite into line with the rest of the gnathobase after the endite has been pushed forwards by adductor movements of the coxa which press the endite against food or against the roof of the food basin. 132 29 and 41p). The phase difference between successive legs is small. Food found in or on the substratum by the terminal pincers of the prosomal legs during burrowing or walking is placed deep in a food basin extending between the coxal gnathobases, the labrum ante- riorly and the chilaria posteriorly (Fig. 59). The walking coxal movements are re- placed by adductor and abductor move- ments in the transverse plane. Projecting dorsally above the ball and socket pleuro- coxal articulation (Figs. 58, 60) are two flanges carrying abductor muscles 25 and 26. A small inward displacement of these flanges results in considerable abduction of the coxal bases. The adductor muscles 42 and 43 (Fig. 58) insert ventrally on the endosternite. The forward slope of the mesial proximal corners of the coxae, brings the movable coxal endites of legs 3 to 5 close to the mouth (Fig. 59). Ad- duction of the paired coxae results in leva- tion of these endites because each coxa moves from a laterally situated articula- tion (Fig. 58). The posterior face of the movable endites slides against the ster- nite, thus pushing food into the mouth. Lamellibranchs are subjected to a pre- liminary ‘“nutcracking” treatment. The shell is gripped by the chilaria, bitten and cracked by the very strong cusped sixth pair of coxae, and the chilaria swing for- wards passing the broken prey to the more anterior shredding gnathobases (a fuller account of the feeding movements will be given elsewhere). The skeleto-musculature of the prosomal limbs of Limulus is suited to produce all the observable movements. Abductor mus- cles are not absent as stated by Snodgrass (1952). The direct transverse biting by means of movements taking place at right angles to the basic promotor-remotor am- MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 bulatory swing contrasts with the trans- verse biting in Crustacea which appears to be a secondary end-point acquired from a promotor-remotor roll more than once in their evolution. The direct trans- verse biting of the Limulus gnathobase appears to be primitive and to indicate a wide phylogenetic gap between the Cheli- cerata and Crustacea. This conclusion is in keeping with the differences in the nature of the biramous limb which have been stressed by Stgrmer, 1939-1951. The anatomy and movements of the coxae of Limulus cast doubt upon the validity of the evidence for a pre-coxal segment. The ridges and grooves at the base of the coxa (Fig. 60) are functionally necessary to brace the coxal cuticle for carrying out both the promotor-remotor and abductor-adductor movements. Many other sutures in arthropods, for example the cervical groove in Anostraca and the more primitive Malacostraca (Figs. 45, 48, 51 and 52), the many “lines” on the head of Ligia (Fig. 54), and most sutures on the heads of insects (Strenger, 1952), clearly have a mechanical significance and do not represent lines of fusion of origi- nally separate scutes. Whether or not a free pre-coxal seg- ment ever existed, there is no denying the similarity between the legs of Trilobita and Merostomata which suggests that the former may be the prototypes of the appendages of the Chelicerata. Trilobite reconstructions, however, do not show a movable proximal endite, and _ trilobite coxae near the mouth slope forwards ventrally (Stgérmer, 1951), while those of Limulus slope in the opposite direction, the proximal parts being farthest forwards (Fig. 59). Thus it seems probable that the free endites of Limulus may have been evolved within the Merostomata alone and MANTON: JAW MECHANISMS 133 endo sternite LIMULUS retractor dorsalis gizzard entapophysis | oblique muscles ventral longitudinal muscles nerve cord transverse tendon chilarium transverse endoskeletal connective 3 2 Y sternite | L posterior anterior Fic. 59. (A) Sagittal half of the prosoma of Tachypleus tridentatus Leach to show the form and positions of the gnathobases, the mouth, and the endoskeleton with its attachments. The lateral prosomal ridge and the posterolateral margin of the carapace are shown by a dotted line. Muscle numbers are those used by Benham (in Lankester, Benham and Beck, 1885). The heart, oesophagus, proventriculus and intestine are not labelled. Numbers 2-6 mark the gnathobases of legs 2-6; the moveable coxal endites of legs 3-5, which are directed towards the mouth, are not labelled. The positions of the coxal margins of legs 2-6, which are attached by arthrodial membrane to the flanks of the animal, are indicated by dotted lines and marked 2-6. The anterior face of coxa 2 is visible lateral to leg 1. The supra- and suboesophageal ganglia are cut. The cut endosternite and its foreshortened dorsal face are shown, the anterior cornu and two lateral cornua are attached by muscles (not shown) to the anterolateral part of the carapace; the dorsal cornu is attached by muscle 52 to the carapace and by muscle 53 from the base of the dorsal cornu, and muscles 54 and 55 from the dorsal face of the endosternite pass to entapophysis 1. A small “transverse endoskeletal connective” lies below the nerve cord above the gnathobase of leg 6; this skeletal bridge is united laterally with the endo- sternite and carries a muscle to the anterior part of the sternite as shown (the occipital ring of Patten and Redenbaugh). The oblique and ventral longitudinal muscles are shown diagrammatically (for details see Benham, im Lankester, Benham and Beck, 1885, pls. 74-5, figs. 1-3; Patten and Redenbaugh, 1900, pl. 8, fig. 4). Three extrinsic cheliceral muscles insert on the carapace; the largest and most posterior is the flexor and the most anterior is the extensor. (B) View of the gnathobase of leg 5 from the sagittal plane to show the free anterior and tight posterior union of the endite with the stiff cuticle of the rest of the gnathobase, and the range of movement of the endite in the parasagittal plane (see also Figs. 58, B and C). 134 only in the proximity of the mouth, since a series of such endites all along an un- differentiated body could not have pos- functions to those of sessed similar Limulus. MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 RELATIONSHIPS OF CRUSTACEA, MEROSTOMATA AND TRILOBITA The basic movements of the locomotory limbs of Limulus and of Crustacea are the same, a promotor-remotor swing. The dorsal flange of coxa pleurite pleural surface pleuro- coxal articulation dorsal flange of coxa pleurite fold anterior edge of Be cox arthrodial A 2S membrane X cavity of coxa coxa leg 5 coxa leg 6 dorsal flange of coxa coxal endite chilarium coxa leg 6 coxa leg 5 anterior edge of. coxa anterior —_____» Posterior edge of coxa lateral groove (a) Fic. 60. (A), External lateral view of the dorsal ends of the coxae of legs 5 and 6 of Tachypleus tridentatus Leach to show their articulations with the Y-shaped pleurites lying at the junction of the under surface of the carapace and the pleural arthrodial membrane. A fold of this mem- brane covers the intucked dorsal flange of each coxa. (B) Internal view from the sagittal plane of the coxae of legs 5 and 6 after removal of the viscera and endosternite. The cuticular thicken- ings forming the coxal rims and internal ridges are shown in white. (C) Dorsal view of the upper end of left coxa 5 to show the size of muscles 25-29 and the position of their carapace in- sertions. Muscle 27 is attached to a tendinous sheet projecting below muscle 25 from the lateral groove of the carapace (see Fig. 58). The dorsal flange of the coxa is foreshortened, and the an- terior and posterior coxal margins slope downwards and to the left because the proximal rim of the coxa does not lie in the exact transverse plane (see Fig. 59, A, dotted outline of the base of the coxa of leg 5). MANTON: JAW MECHANISMS fundamental difference between the feed- ing movements of mandibular gnathobases of Crustacea and of the prosomal gnatho- bases of Limulus is of great importance. This, and the associated differences in limb morphology, including coxal articu- lation with the body, must surely mean that the merostome and crustacean gnatho- bases have been independently acquired as a parallel evolution not indicative of affinity, although both groups bite with a gnathobase and not with the tip of the limb. If a common basis for the Limulus and crustacean gnathobasic mechanism has ever existed it could only be sought for in a minute arthropod, possessing no basal articulations to its limbs, in which an un- differentiated link by arthrodial membrane permitted a variety of slow movements by promotor, remotor, adductor and abductor muscles. Increase in size or in strength of movement must have been accompanied by the evolution of closer articulations at the limb bases. Thereafter, the Crustacea have used direct adductor-abductor man- dibular movements only to a minor extent (as in Chirocephalus and Hemimysis), the direct abductor muscles 6 of these two animals corresponding in general principle, but not in detail, with abductors 25 and 26 of Limulus prosomal limbs 2 to 6, while the Chelicerata have exploited this biting movement to a maximum. It is significant that the second feature which the Merostomata and Crustacea have in common, the biramous leg, is one of quite different construction. The outer ramus in the Merostomata and in the Trilobita is a proximal exite (pre-epipodite of Stérmer, 1939), while the outer ramus in Crustacea is a more distal structure borne on the end of the protopodite, and one or two proximal exites may be pre- sent as in Anaspides. 135 A third contrasting feature shown by Limulus and the Crustacea lies in the ventral flexibility of the limb-bearing body wall in the former and the rigidity of the trunk insertions of the locomotory limbs in the latter. The pedigerous ventral sur- face of the prosoma of Limulus can shorten when feeding to 60 per cent of the length shown when walking, a flexibility serving opposite needs. Coxae when closely packed one behind the next can chew without lateral escape of food from the food basin, while some spreading out of successive coxae is essential for the locomotory promotor-remotor movements. Thus, the study of jaw mechanisms emphasises the depth of the cleft between the Merostomata-Chelicerata and the Crus- tacea. The Trilobita have clear general resemblances in leg form to the limbs of Limulus, but the Trilobita lack as good a gnathobase and possess pre-epipodites all along the series. There is also the resemblance in trilobation of the carapace which may have a functional significance such as it has in Limulus (bracing the cuticle against the deformatory pull and the extrinsic limb muscles). Thus the evi- dence, as far as it goes, supports Stgrmer, 1944, in suggesting closer affinity between Chelicerata and Trilobita than between either of them and the Crustacea. In attempting to assess the taxonomic position of early arthropodan fossil ani- mals, many showing some resemblances to Crustacea, one would like to see a much more careful study of the basal regions of the head and trunk limbs where the state of preservation may allow it. There appears to be much too great a readiness to state that limbs are of the biramous trilobite type when in fact this is not proved. Details of the coxal articulations and outer rami of the appendages might 136 indicate the type of biramous limbs and type of gnathobasic or jaw-like movements which existed and would be of service in correctly interpreting the fossil record. THE RELATIONSHIPS OF CRUS- TACEA, MYRIAPODA AND HEXAPODA There is reason to suppose that the present-day Onychophora, Myriapoda and Hexapoda represent a related series of animals whose origin may have been far removed from that of either the Crustacea or the Trilobita-Chelicerata assemblage. The unsegmented whole-limb jaws of Peripatus place the Onychophora squarely with the Myriapoda-Hexapoda assemblage, a conclusion in keeping with the many other considerations brought forward by Tiegs, 1947, Manton, 1949, and Tiegs and Manton, 1958. The limited cephaliza- tion indicates perhaps a very early adop- tion of a jaw technique in feeding. The unique alternate anteroposterior slicing by entognathous jaws (Manton, 1937) is related to the onychophoran accomplish- ment of outstanding importance, that of being able to deform the body extremely so that access is gained, without pushing, to damp cavities where predators cannot follow (Manton, 1959, 196la). Large sclerotised mandibles working on a trans- verse mandibular tendon would be an im- possible mechanism for such a habit of life. The four types of jaws indicated by the vertical columns on Figure 61 appear to have been independently evolved, since the details of their mechanisms and structure preclude any one type giving rise to an- other. Since the crustacean mandible is primarily a gnathobase and the hexapod mandible a whole limb, it would be sur- prising to find more than convergent re- semblances between them, as indeed is MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 the case. No support has been found for the assumption (Snodgrass, 1951) that these mandibles are homologous coxal derivatives. It is the coxa-body joint only which is constant in all types of man- dibles. Since only two movements of am- bulatory limbs appear to have been used in jaw evolution, it is not surprising to find each of these movements to have been exploited more than once. The whole-limb jaws of the Hexapoda are unsegmented and the mandibular mechanism, which most nearly approaches a central primitive type among hexapods, is that of Petrobius. As in Crustacea, Petrobius uses a promotor-remotor rolling movement of the mandible (but with very many differences in detail) about an ap- proximately dorsoventral axis. The dorsal point of closest union with the head forms a ball-and-socket joint in Petrobius, per- mitting very free rotation (Fig. 48A). The corresponding point of closest union of the mandible and head of Hemimysis is shown in Figure 48B; here there is no articulation at all, just a link of thick flexible arthrodial membrane braced by a fibrillar suspension from the basement membrane above. Petrobius possesses a well developed transverse mandibular tendon functioning as in the more primi- tive Crustacea, but this is a potential attribute of all body segments (Manton, 1928, 1934 and 1956) and is no evidence of close affinity. The mandible of Petrobius essentially rotates in a slot, and food is scratched, sucked up, and ground by molar processes. The combined action of the mouth parts, the hypopharynx and super- linguae gives hydraulic efficiency around an oral cone, copious salivary secretions being produced by the labial segmental organs. The mandibular mechanisms of (i) MANTON: JAW MECHANISMS Ctenolepisma and the more _ primitive Pterygota and of (ii) the entognathous Apterygota (Collembola and _ Diplura) present two divergent types of mandibular evolution which could have originated from an archi-Petrobius type. Trend (i) leads towards an absence of the Petrobius type of hydraulic efficiency and to the acquisition of strongly hinged transversely biting mandibles, by changes in the posi- tion of the axis of movement much as in the Isopoda, together with a parallel utili- sation of the same type of adductor-ab- transversely biting jaws CHELICERATA CRUSTACEA transversely biting jaws rolling jaws GNATHOBASIC JAWS GNATHOBASIC JAWS t T Fic. 61. types of mandibles or jaws (below) and the derivation of the jaw mechanisms transversely biting ja 137 ductor musculature. The superficial re- semblances between the mandibles of Ctenolepisma and the Pterygota and of the Isopoda are no more than convergent. Trend (ii) leads towards a proximal free- dom of the mandibles permitting protrac- tor-retractor movements, as well as free rotator and counter-rotator movements, made possible by the growth of a pleural fold, such as is present in Petrobius, and leads to entognathy. The differences in the protractor-retractor mechanisms of Collembola and Diplura suggest that their == entognathous-= === protrusible jaws Van HEXAPODA MYRIAPODA rolling jaws transversely biting jaws unjointed jaws jointed jaws WHOLE-LIMB JAWS WHOLE-LIMB JAWS \ Diagram showing the conclusions reached concerning the distribution of the principal (above). The heavy vertical lines indicate an entire absence of common ancestry between the jaws referred to on either side; an interrupted vertical line indicates separate evolutions of the jaw mechanisms of Hexapoda and Myriapoda which probably had a common origin; and the shaded areas indicate mandibular mechanisms showing convergent similarities derived from unlike origins. 138 entognathy has been independently ac- quired. All these changes indicate a closer relationship between some archi-Petrobius type of thysanuran and a pterygote than between either Collembola or Diplura and the Pterygota. No hexapod which has em- barked upon the evolution of one or an- other type of entognathous mandible is at all likely to have been able to reverse its evolutionary trend back to a generalised state and then progress towards the ptery- gote condition. The closer affinity of the pterygote and apterygote groups to each other than to the Myriapoda is shown by their mandibu- lar mechanisms, by the details of the structure, musculature and functions of the anterior and posterior tentorial apo- demes and segmental tendons (described in the full account of this work for the first time) and by their hexapodous state. The mandibles of the Myriapoda, com- prising two segments in the Symphyla, three in the Diplopoda, three to five move- able scutes in the Chilopoda and un- jointed in the Pauropoda, undoubtedly represent whole limbs comparable with the unsegmented mandibles of the Hexapoda. The basic mandibular movement in the Myriapoda appears to be direct biting in the transverse plane, such as is employed by the gnathobase of Limulus, but using the tip and not the base of the limb. The usual difficulty concerning the abductor mechanism has had to be resolved. Seg- mentation of these mandibles facilitates adduction and all utilise the mobile an- terior tentorial apodemes, to different ex- tents, to promote mandibular abduction. The advantages of entognathy, namely protrusibility of mandibles and_ great freedom of movement, have been inde- pendently acquired in the Chilopoda and Pauropoda. The basic form and modes of MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 action of myriapodan mandibles point to a unity among these classes and a cleft between them and the Hexapoda. The Symphyla, in some ways so sugges- tive of originating from the ancestors of insects, differ basically from the hexapods in mandibular mechanism, in structure and mode of action of maxilla 1, in the presence of maxilla 1 segmental organs, and in the presence of the myriapodan type of head endoskeleton; these differences are so great as to make it clear that a suggested archi-Symphylan ancestry of insects must be abandoned. The very great apparent differences in external trunk characters of the myriapodan classes are correlated in detail with habits of life and could readily have been derived from a common type (Manton, 1954, 1956, 1958, 1961la and b, Parts 8 and 9 in preparation). The Diplo- poda and Chilopoda show the result of early habit divergence between fast mov- ing predatory animals and_ burrowing feeders on decaying vegetation. The stimu- lus for the evolution of the entognathy of the Chilopoda has been the predatory crevice-living habit and the advantage of extreme head flattening. CONCLUSIONS Thus the evidence derived from man- dibular mechanisms supports the view con- cerning a deep separation between (i) the Onychophora-Myriapoda-Hexapoda stem, (ii) the Crustacea, and (iii) the Mero- stomata-Chelicerata-Trilobita assemblage. A biting gnathobase has been _ inde- pendently evolved in the Merostomata and in Crustacea, and a whole-limb man- dible has evolved independently in the present day land forms, usually in seg- mented form in the Myriapoda, and un- segmented, using a different basic move- ment, in the Hexapoda. MANTON: JAW MECHANISMS A parallel evolution of jaws in arthro- pods must date from the earliest differen- tiation of the major classes, but there is no indication of the stage of advancement from which they came. There appears to be no justification for the use of the term “Mandibulata” in a taxonomic sense uniting the Crustacea, Myriapoda and Hexapoda. Similarly the ““Entognatha” and “Labiata” indicate grades of advancement and not taxonomic groups. To some persons the parallel evolution of a mandible is a most improbable sup- position. Yet Hinton (1957-1962) has demonstrated the independent evolution of spiracular gills of pupae and plastron respiratory structures of eggs of insects in a very large number of cases, and the similarity holds down to the electron microscope level. Clearly, deviations from the resolution of the respiratory problems are impracticable and consequently the same type of respiratory structure has been evolved many times. Mandibles are not quite like this. There are a limited number of ways of acquiring a mandible, and following them out throws much light on the past history of the arthropods. Increase in knowledge of the fossil record shows that reptilian and mammalian grades of organization have been reached independently many times. “The mam- mals are a polyphyletic group—by which is meant that mammals have no common ancestor which was itself a mammal. The common ancestor must be found among the reptiles’ (Kermack and Mussett, 1959). In view of these conclusions on the evolution of major vertebrate classes it would indeed be surprising to find no trace of polyphyletic evolution among so large a group as the Arthropoda. Some polyphyletic conception of arthropod evo- lution indeed seems inescapable, but it is 139 one thing to demonstrate the existence of clefts between modern groups of animals and quite another to speculate upon the depths of these clefts in geological time and to suggest what common type of ani- mal preceded the divergent lineages. REFERENCES Hinton, H. E. 1957. The structure and function of the spiracular gill of the fly Taphrophila vitripennis. Proc. Roy. Soc. London, B, 147: 90-120, 12 figs. 1958. The spiracular gills of insects. Proc. 10th Internat. Congr. Entom. 1956, 1: 543-548, 1 fig. . 1960. The structure and function of the respiratory horns of the eggs of some flies. Phil. Trans. Roy. Soc. London, B, 243:54-73, 14 text-figs., 1 pl. . 1961. How some insects, especially the egg stages, avoid drowning when it rains. Proc. S. London Ent. Nat. Hist. Soc., 1960, 138-154, 4 figs., 4 pls. . 1962a. Respiratory systems of insect egg-shells. Sci. Progr. London, 50:96-113, 23 figs. 1962b. The structure and function of the spiracular gills of Deuterophlebia (Deutero- phlebiidae) in relation to those of other Dip- tera. Proc. Zool. Soc. London, 138:111-122, 4 figs., 2 pls. Jackson, H. D. J. 1926. The morphology of the isopod head—Part 1. The head of Ligia oceanica. Proc. Zool. Soc. London, 1926, 885- 911, 11 figs., 4 pls. Kermack, K. A. Ano F. Mussett. 1959. The first mammals. Discovery, 20:144-151, 11 figs. LankesteEr, E. R., W. B. S. BENHAM, AND E. J. Beck. 1885. On the muscular and endoskeletal systems of Limulus and Scorpio; with some notes on the anatomy and generic characters of scorpions. Trans. Zool. Soc. London, 11: 311-384, 12 pls. Manton, S. M. 1928. On the embryology of the mysid crustacean, Hemimysis lamornae. Phil. Trans. Roy. Soc. London, B, 216:363-463, 32 figs., pls. 21-25. . 1934. On the embryology of the crusta- cean, Nebalia bipes. Phil. Trans. Roy. Soc. London, B, 223:168-238, 17 figs., pls. 20-26. . 1937. The feeding, digestion, excretion 140 and food storage of Peripatopsis. Phil. Trans. Roy. Soc. London, B, 227:411-464, 14 figs., 3 pls. 1949. Studies on the Onychophora. VII. The early embryonic stages of Peri- patopsis, and some general considerations con- cerning the morphology and phylogeny of the Arthropoda. Phil. Trans. Roy. Soc. London, B, 233:483-580, 7 figs., pls. 31-41. 1954. The evolution of arthropodan locomotory mechanisms. Part 4. The structure, habits and evolution of the Diplopoda. J. Linn. Soc. London (Zool.), 42:299-368, 8 figs., 4 pls. 1956. Idem. Part 5. The structure, habits and evolution of the Pselaphognatha (Diplopoda). J. Linn. Soc. London (Zool.), 43:153-187, 8 figs., 1 pl. . 1958. Idem. Part 6. Habits and evolu- tion of the Lysiopetaloidea (Diplopoda), some principles of the leg design in Diplopoda and Chilopoda, and limb structure in Diplopoda. J. Linn. Soc. London (Zool.), 48:487-556, 21 figs., 1 pl. 1958b. Hydrostatic pressure and leg extension in arthropods, with special reference to arachnids. Ann. Mag. nat. Hist., Ser. 13, 1:161-182, 5 figs., 1 pl. 1959. Functional morphology and taxonomic problems of Arthropoda. In Syste- matics Association Publication, No. 3, 23-32. —. 1961a. Experimental zoology and prob- lems of arthropod evolution. In The cell and the organism. Ramsay and Wigglesworth, Ed. Cambridge. Pp. 234-255, 1 fig., 2 pls. . 1961b. The evolution of arthropodan locomotory mechanisms. Part 7. Functional re- quirements and body design in Colobognatha (Diplopoda), together with a comparative account of diplopod burrowing techniques, MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 trunk musculature and segmentation. J. Linn. Soc. London (Zool.), 44:383-461, 35 figs., 3 pls. Patten, W. ano W. A. REDENBAUGH. 1900. Stud- ies on Limulus. Il. The nervous system of Limulus polyphemus, with observations on the general anatomy. J. Morph., 16:91-200, 18 figs., 5 pls. Pearson, J. 1908. Cancer. Mem. Liverpool Mar. biol. Comm. XVI, viii + 209 pp., 13 pls. Scumipt, W. 1915. Die Muskulatur von Astacus fluviatilis (Potamobius astacus L.) Zeit. wiss Zool. Leipzig. 113:165-251, 26 figs. Snoperass, R. E. 1950. Comparative studies ot the jaws of mandibulate arthropods. Smithson. misc. Coll., 116:1-85, 24 figs. . 1951. Comparative studies of the head of mandibulate arthropods. New York. viii + 118 pp., 37 figs. . 1952, A textbook of arthropod anatomy. New York (Comstock). viii + 363 pp., 88 figs. STORMER, LeEIF. 1939, 1941 and 1951. Studies on trilobite morphology I, II, III. Norsk. geol. Tidsskr., 19:143-273, 35 figs., 12 pls.; 21:49- 164, 19 figs., 2 pls.; 29:108-158, 14 figs., 4 pls. . 1944. On the relationship and phylogeny of fossil and Recent Arachnomorpha. Skr. Norske Videns. Akad. Oslo, 1 Math.-Nat. KI., 5:1-158. STRENGER, A. 1952. Die Funktionelle und mor- phologische Bedeutung der N&ahte am _ In- sektenkopf. Zool. Jahrb. (Anat.), 72:469-521, 18 figs. Trecs, O. W. 1947. The development affinities of the Pauropoda based on a study of Pauropus silvaticus. Quart. J. micr. Sci., 88:165-336, 29 figs., 10 pls. Tiecs, O. W. Ano S. M. Manton. 1958. The evolution of the Arthropoda. Biol. Rey. 33: 255-337, 48 figs. PHYLOGENY AND EVOLUTION OF CRUSTACEA Museum oF CoMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 x Discussion Following Manton’s Paper MANTON: It is of great importance to recognize that there are two quite dif- ferent ways in which animals react to a definite need. If there is only one or a very limited range in resolutions of the problem, the same morphological (and physiological) features turn up independ- ently in unrelated animals, as in the case of compound eyes and plastron respiratory structures within the Arthropoda. Some resolutions may persist in spite of all other changes; for example, the fine structure of a cilium is constant throughout the plant and animal kingdoms. If, on the other hand, there are various but a limited num- ber of ways of meeting a need, the same resolution may be independently adopted by unrelated animals, giving well known convergent similarities. Phylogenetically related animals inheriting the same resolu- tion of a problem show resemblances due to affinity. “Unique” mechanisms may account for very detailed convergent similarities. Mandibular mechanisms are not “unique.” There are a limited num- ber of resolutions of the need for jaws, and an understanding of them shows which characteristics are indicative of phylo- genetic affinity or the reverse, and which are convergent. HESSLER: In AHutchinsoniella the mandibular muscle pattern is clearly de- rived from that of the thoracic limbs. The extrinsic muscles for each thoracic limb 141 come from four main areas of origin (Fig. 62): two are dorsal, just lateral to the dorsal longitudinal trunk muscles; two are ventral (VM), either the ventral inter- segmental tendon anterior or posterior to the segment. Of the dorsal extrinsic mus- cles, the anterior (DAM) and posterior (DPM) groups have corresponding an- terior and posterior insertions on the limbs. The dorsal anterior muscles fall into groups which abduct (ABD) or adduct (ADD) as they exercise their main func- tion of promotion. Observation shows that there is a mediolateral component to the primarily fore-aft swing of a thoracic limb, so that the path of the limb is elliptical. The extrinsic muscles of the maxillule fall into the same basic groups found in thoracic limbs. The abductor-adductor groups of the dorsal muscles are more strongly differentiated than those of more posterior limbs. This change is related to the greater importance of abduction-ad- duction in a limb primarily concerned with moving food into the atrium oris. With this change in function, the origin of the largest ventral extrinsic muscle has shifted along the posterior intersegmental tendon to the midline, thus increasing its power as an adductor. In the mandible, the changes are of the same sort found in the maxillule, but more extreme. The origins of nearly all the powerful ventral extrinsic muscles MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 142 ‘IQIssoPT “WY AG ‘sapsnur quay ssuryxa yeusaossjsod pue -o1ajue ‘PA ‘sapsnu YuNs} [eUIpNz suo] [erjUA “TATA + seposnur quiyy oisutjxo s101ejsodosiop ‘Wd ‘apsnur yuns [eurpnzsuol [esiop ‘Td ‘aures yo dnois s0jnppe “Gd V-WV ‘sepsnur quit] QIsuIs]X9 Iolo}UROsIop jo dnoss s0jONpqe ‘GAV-WVA ‘29} JoLezUe oy} UO Ie 9}TYM UT asoy} !quUIT] ey} Jo dI"F IOTI0}sod ay} UO are YOR[G Ul sapsnyy ‘aqipuew ‘Op pue ‘epIxeur ysiy ‘gq ‘qUIT] 9DeIOYY ‘YW “DYIUDIDAIDIM DYaquosuzyIINFY JO IINje[NIsNU GUT “79 “OTT DISCUSSION have shifted to the midline. The origins of the abductor and adductor subgroups of dorsal anterior muscles have shifted even further apart. The abductor group is the only set of muscles with an origin lateral to the single hinge point of the mandible and therefore is the only muscle group capable of abducting the jaws, the same function it had in the more posterior limbs. I agree with Manton that the primary motion of the mandible is a promotor-remotor swing, but just as in the posterior limbs there is still a muscu- larly controlled abductor-adductor com- ponent. LOCHHEAD: On the question of turgor and the role of the exoskeleton in the earliest arthropods: usually emphasis is placed on some sort of rigid exoskeleton, providing advantages such as hinge joints and muscle attachments. But may not the turgor type of organization have been more primitive? MANTON: A turgor type of organi- zation is basic for most phyla. It is dif- ficult to visualize an animal in which the internal pressures are not of vital im- portance. Hydrostatic pressure is essential in coelenterates, annelids and mollusks as well as arthropods, and a pressure system is essential even to a sponge. I suggest that the absence of basal articulations to the legs of Cephalocarida and of Anostraca and many other Bran- chiopoda is correlated with the large an- terior and posterior area of the legs in proportion to the transverse sectional area of the body. This of necessity results in very wide basal unions with the body. It is mechanically simple to devise basal articulations between leg and body per- mitting much movement when the leg base is small, but much less easy when the base is large. The incurled posterior mar- 143 gin of the mandible of Chirocephalus (Fig. 46D) is a means of easing the swing of so large a limb base, and the absence of any firm proximal rim to the trunk limbs of Anostraca and Cephalocarida is an- other resolution (necessitating also the origin of extrinsic leg muscles on the anterior and posterior faces of the legs and not on the proximal margins). BROOKS: ‘Thus far in this conference, we have not defined a crustacean. Crus- tacea are of great diversity but one thing that characterizes them is the structure of the head. There are two pairs of antennae, a pair of mandibles and two pairs of maxillae. These are not comparable to the cephalic appendages of a trilobite. Tri- lobites and Crustacea represent two phylo- genetic lines, and the structures mentioned are not homologous. Yesterday it was stated that trilobites were filter-feeders. I believe Stérmer was correct in drawing an analogy between the feeding habits of trilobites and Recent king crabs. Thus, the trilobites were probably lowly scavengers and predators. Whether we believe the crustacean jaw primitively was for scraping, for a molar purpose, or for biting is not important. The significant fact is that all primitive Crustacea have heavily sclerotized, strong jaws. Are we to believe that this structure originated through natural selection in a filter feeder? The crustacean jaw struc- ture must not be disregarded in phylo- genetic considerations. MANTON: Chirocephalus is a typical filter-feeder with a very large mandible, but this mandible is not heavy. The whole cuticle of Chirocephalus is thin and deli- cate and that of the mandible is a trifle more fully sclerotized than that of the trunk. To describe the mandibles of filter- feeding Crustacea as “heavy” is mislead- 144 ing when their specific gravity approxi- mates that of water. I have shown how the simplest crustacean mandibles suit the squeezing or grinding of small particles of soft food, however the food is collected. A heavy mandible is only found in large- food feeders capable of strong biting, and this is a repeated end term in crustacean evolution. BROOKS: But we have to be careful about secondary modification. The biting triturating mandible is a diagnostic charac- teristic of Crustacea. LOCHHEAD: Chirocephalus has a large, heavy mandible, which it uses when filter-feeding. However, like most filter- feeders, Chirocephalus does not always filter-feed. Sometimes it goes down to the bottom and scrapes up material with the tips of its trunk limbs. Perhaps the heavi- ness of the mandible is associated with this habit. GORDON: Filter-feeders are so highly specialized that I think the whole mech- anism has arisen secondarily. Euphausia has a marvelous filter-feeding method for a pelagic crustacean. Even in the Neba- liacea Cannon has shown that the filter- feeding limbs are secondarily acquired. MANTON: I do not believe that an elaborate filter-feeding mechanism was ever primitive. With Lochhead I would emphasize that a crustacean can feed on minute food without any filter mechanism, as in Tviops. Trilobites may have done MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 the same, all limbs cooperating, but these animals could not have cut up their food as effectively as does Limulus since the paired limbs apparently could not meet ventrally. A case can be made for suppos- ing that the primitive malacostracan was a bottom-liver. Such a habit could account for the typical differentiation of thorax and abdomen, and such an animal proba- bly was a bottom-feeder on soft or small food. Filter-feeding may have turned up with the adoption of the pelagic habit, probably convergently in Syncarida, Eu- carida, and Peracarida. GLAESSNER: When we discuss the feeding of the trilobites, we should not forget the paleobiological evidence. Seila- cher has somewhat confused the evidence by referring to +Cruziana as a trilobite resting-track. However, at least from Ordovician time, trilobites scraped up material from the bottom incidentally creating large and conspicuous sedimen- tary structures, which show that they were searching for coarse material. We have here the possibility of looking not only at the organization of the limb but also at actual traces of its activity left in the rock. BROOKS: 7Cruziana is also found in Lower Cambrian strata. There are many specimens on which it is possible to see that the trilobites were plowing in search of food just as does the Recent king crab. PHYLOGENY AND EVOLUTION OF CRUSTACEA MuseuM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 XI Evolution of the Branchiopoda INTRODUCTION The subclass Branchiopoda, as now understood, consists of seven orders, three of which are extinct. There is agreement at present on the classification shown in Table 5. Three of the orders made their first recorded appearance in Lower Devonian time. These are the Conchostraca, Anos- traca, and fAcercostraca. The _ order yLipostraca apparently (Tasch, 1957) originated and became extinct during Middle Devonian time. The first recorded By Paul Tasch Department of Geology, University of Wichita, Wichita, Kansas appearance of the Notostraca was during the Upper Carboniferous. A recently de- scribed order, +Kazacharthra, first ap- peared in the Upper Jurassic. The Clado- cera apparently arose sometime between late Mesozoic and the older Tertiary (Fig. 63). DUBIOUS PRE-DEVONIAN CONCHOSTRACANS The family jLepidittidae (Kobayashi, 1954) consists: of calcareous-phosphatic and calcareous forms that Tasch considers TABLE 5 CLASSIFICATION OF THE BRANCHIOPOD ORDERS BY VWARIOUS WORKERS LINDER BROOKS, TASCH DAHL Superorder 1 Notostraca SERIES ~Kazacharthra PHYLLOPODA A yAcercostraca Superorder 2 Conchostraca Cladocera SERIES Superorder 3 B Anostraca ANOSTRACA {Lipostraca The three superorders will be named in the “Treatise on Invertebrate Paleontology, part R.” 145 146 unacceptable as conchostracans. Other workers are of the same opinion (Novoji- lov, 1960). Reference is often made to this family to prove the marine origin of conchostracans. Adamezak (1961), for example, accepts at least one of the genera of the above-named family, the genus ;Fordilla, and considers it to be the Cam- brian ancestor of both conchostracans and yEridostraca-type ostracodes. He refers further to a conchostracan found in Pro- Recent > a O la eS = —_ S le WwW WwW oO - Upper S Jurassic N (o) op) W = Middle Devonian [o) mi CONCHOSTRACA ie Lower oO Devonian WwW J S LIMNADOPSIS KERATESTHERIA Jurassic \ Triassic ECHINESTHERIA \ \ \& \ VERTEXIA Permian Mt —— \ \\ C€C am ©) yy Pek anor Ns me ma GABONESTHERIA CURVACORNUTUS PALEOLIMNADOPSIS LIMNADIOPSILEAIA Carboniferous a ee ea yy (Penn) a . (CG) itv Miss.) Ui aes Z i) ia eet ee aaa ae Vie Pre- , if Carboniferous A CORNIA 2 Wy. (¢Pemphicyclus) / Common Ancestral Type Fic. 65. nadopsis read Palaeolimnadiopsis.) Possible lines of descent for spined conchostracans and related types. > < J) NS, _-— — {PSILONIA (Dev.) (For Paleolim- 150 on the initial valve, appears to be ancestral to the genus ;Vertexia and also to have given rise to the genus, +Curvacornutus. The latter bears a curved or looped initial valve spine. Another genus apparently derived from +Cornia, is +Protomonocarina. This un- usual genus possesses a series of bead- like components that form a partial rib. If, as postulated by Tasch, this genus did arise from a fCornia population (Fig. 66), then a mutation of the gene or genes governing larval node or spine could have caused a shift of such a node or spine from the center to the periphery of the valve. As the first five growth bands were laid down during successive molts, a repetition of the mutant condition could account for the beaded, partial rib. Mississippian and Pennsylvanian leaiids studied by _ the writer often show bead-like components of valve ribs. Novojilov raised the condi- tion of leaiid ribs with bead-like com- PEMPHICYCLIID\ GENE POOL ~ MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 ponents to generic status in his genus, y/gorvarentsovia. Such recognition seems hardly warranted since the condition of bead-like components of leaiid ribs, as already noted, is visible in eroded speci- mens of many species of the genus 7Leaia. Nevertheless, the widespread occurrence of bead-like components in leaiid ribs of Carboniferous specimens appears to strengthen the view that the leaiid rib originated by the steps indicated earlier. This would not exclude other modes of development of ribbedness in conchostra- cans. Another genus from the Oklahoma Wellington Formation, +Limnadiopsileaia, possesses a single leaiid rib as well as posterior recurvature of growth lines. It apparently derived from a contempora- neous unribbed palaeolimnadiopsiid popula- tion. One can explain this occurrence if the palaeolimnadiopsiid line is traced back to tEstheria ortoni from the Conemaugh of Ohio. As noted earlier, this species bears an NORMAL DEVELOPMENT —S-O-@ Nec POSSIBLE GENE MODIFICATION LEADING TO PROTOMONOCARINA IGORVARENTSOVIA INGENS NOV. Gene modification of initial valve node and origin of the leaiid rib. = (LEAIA) Fic. 66. TASCH: BRANCHIOPODA initial valve spine and also has posterior recurvature of growth lines. Another pos- sibility is the derivation from an Upper Mississippian unribbed palaeolimnadiopsiid that did not bear an initial valve spine but may have carried the gene(s) for it as recessives. At any rate, +Protomonocarina at one end, and +Limnadiopsileaia at the other (see Fig. 65), appear to relate back to a pre-Carboniferous +Cornia ancestral type. This type is envisioned as having had an initial valve spine but lacking posterior recurvature of growth lines. Since five, four, and three-ribbed leaiids are known from the Upper Devonian of the Canadian Arctic, it is apparent that the leaiid rib has a long prehistory in the oldest Devonian or pre-Devonian. Until such specimens are found, the question of “rib origin” cannot be resolved. The Permian +Palaeolimnadiopsis line could have given rise to }Keratestheria in the Jurassic and also to Limnadopsis. The last-named genus is characterized by a serrated dorsal margin. The Triassic jEchinestheria, and the Permian 7+Ver- texia, appear to be derivatives of the Per- mian genus *Gabonestheria. TAXONOMY AND ITS BEARING ON EVOLUTIONARY TREND Novojilov (1960) expanded the num- ber of fossil conchostracan families to twenty. Tasch, in the “Treatise on In- vertebrate Paleontology,’ has found that these can be reduced to eleven. The reason for this reduction stems from com- pletely divergent interpretations of what constitutes specific, generic, and familial characters. Having rejected fine sculpture and ordinary shape variants of valves as being within the variability to be expected in a given conchostracan population, it follows that new categories erected on IEG such premises should be placed in syn- onymy. When that is done, different evolu- tionary trends may be discerned. The eleven conchostracan families rec- ognized by ‘Tasch embrace forty-eight genera. During the Devonian, four new families appeared, while five new families first occurred in the Carboniferous. Thus, nine new families apparently made their first appearance during the late-Middle to Upper Paleozoic, a span of some 130 mil- lion years. By the Mesozoic, five of these Paleozoic families became extinct and only two new families arose during Juras- sic to Cretaceous time, a span of some 110 million years. Two new conchostracan families arose in post-Cretaceous time, a span of seventy million years. The Cyclestheriidae cer- tainly is a post-Cretaceous family. Per- mian through Cretaceous genera, assigned by Novojilov (1958) to the Leptes- theriidae, have been shifted elsewhere by Tasch in the “Treatise” for reasons al- ready mentioned. That leaves only Re- cent forms under this family. Thus, Lep- testheriidae appears to have arisen during post-Cretaceous time. It is apparent that Middle to Upper Paleozoic time represents the heyday of conchostracan evolution. That, in turn, can account for the variability in certain extinct valve features discussed above. Since the end of the Paleozoic to the Re- cent, only four new families have made their appearance. By contrast, an equiva- lent number arose in the Devonian alone. The entire geological history suggests that, in an evolutionary sense, the conchostra- cans represent a stagnant group. Only three of the five living families, which embrace fourteen genera, have a good fossil record. The Cyzicidae range from the Devonian to the Recent; the 152 Limnadiidae from the Carboniferous to the Recent; and the Lynceidae from the Upper Cretaceous to the Recent. Soft-part anatomy.—Evidence on soft- part anatomy preserved in the fossil record is limited to a very few well-documented examples. }Limnestheria ardra from the Kilkenny Coal Measures of Ireland had antennae, mandibles, trunk, telson, and appendages preserved. Specialization of the first and/or second appendage in males for clasping the female during copulation was apparently established in its modern aspect by Carboniferous time. +Cornia cebennsis described by Des- chaseaux (1951) had mandibles, frag- ments of a biramous antenna, impression of an ocellus, and of the interior of the digestive tube, as well as caudal furca, preserved. This Carboniferous conchostra- can possessed a head with a modern pro- file, somewhat larger appendages than liv- ving forms, shorter antennae, and a longer caudal furca. In addition, eggs fossilized with the valves were both fewer in num- ber and larger in size than in modern forms. Deschaseaux noted that by Triassic time, conchostracan eggs became smaller and more numerous in a given individual. It is apparent that, except in varying dimensions which might fit a normal curve of distribution in a given population if more fossil examples had been found, Carboniferous conchostracans were modern in most aspects of soft-part anatomy. However, Deschaseaux’s observation sug- gests that the reproductive process in fe- males differed from modern forms at least in egg production. It is plausible to assume that larger eggs in Carboniferous conchostracans may reflect the transition from a marine to a freshwater environment via an estuary. Needham (1930) pointed out that the MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 transition from salt water to fresh water via the estuarine environment would in- volve eggs with larger yolks to accom- modate to the brackish conditions. A further bit of evidence on this theme is Kummerow’s (1939) conchostracan from the Lower Carboniferous of Germany, +Ouadriasmussia hercynica. This form definitely lived in a marine environment, having been fossilized with trilobites and brachiopods. Gross (1934) described a species of Cyzicus from the Lower Devo- nian of Germany that had been fossilized with marine ostracodes. The fossil record of conchostracans from Devonian through Permian time thus contains examples of marine, estuarine, and freshwater environ- ments. +Limnestheria ardra, for example, clearly lived in an estuarine environment. From these considerations, one may postulate that the Carboniferous was a time of transition from marine to fresh water for some conchostracans. Further- more, there probably were several pulses, or times of transition, since freshwater Devonian forms are known. CLADOCERA Fossil _record—The__ ephippia__ of {+Daphnia fossilis from the Oligocene brown coal of Germany, is the oldest presently known fossil cladoceran. Cal- cified ephippia of Miocene daphniids have also been reported from the late Tertiary Humboldt formation of northeastern Ne- vada. Interglacial and postglacial clado- cerans are quite numerous. As many as forty-three species and subspecies of liv- ing central European chydorids occur in such deposits along with species of other families (Frey, 1958). Evolution of Cladocera.—Frey assumed that cladoceran species from late glacial lake sediments could be compared mor- TASCH: BRANCHIOPODA phologically to living equivalents because no extensive change had occurred in the past 11,000 years. J. L. Brooks concluded that certain Daphnia species from deep glacial lakes arose during the Pleistocene, while southern species in shallow waters arose prior to that time. Scourfield indi- cated that Bosmina species occurring in his glacial lake cores, when compared to living equivalents, suggested a less- evolved condition. These observations, which are essentially confined to the Pleistocene and _post- Pleistocene, can now be extended to the older Tertiary. The mode of shedding ephippia containing eggs in order to with- stand the increasing severity of the en- vironment was operative in its modern aspects some 40 million years ago. In turn, this suggests that Oligocene to younger Tertiary cladocerans possessed a soft-part anatomy closely similar to that of modern forms, with perhaps minor modifications. CLADOCERAN-CONCHOSTRACAN RELATIONSHIP The cladoceran genus Leptodora is con- sidered an “aberrant conchostracan.” It is the only living cladoceran with a larval form that goes through a naupliar stage as do conchostracans. Its shell is reduced to an egg case. The conchostracan species Cyclestheria hislopi, by contrast, is very cladoceran-like. Unlike conchostracans in general, it bypasses the naupliar stage, and the young have the full complement of appendages and a well-developed valve at birth. The eyes are united into a single cyclopean organ as in cladocerans (Sars, 1887). As observed in the discussion on Con- chostraca, the family Cyclestheriidae arose sometime during the post-Cretaceous to older Tertiary. Since Oligocene cladocerans 153 are known, the origin of cladocerans can be further restricted to post-Cretaceous pre-Oligocene time. Cyclestheria hislopi, a living species, seems to be the most likely candidate for the transitional type leading from conchostracan to cladoceran. Living Leptodora, while an aberrant form, is a cladoceran. It is also known from post- glacial deposits. It probably arose during the older Pleistocene. If so, the fact that it has a naupliar stage merely relates back to the ancestral condition, i.e. the original derivation of cladocerans from conchos- tracans. ANOSTRACA-7LIPOSTRACA The fossil record of anostracans is very sparse. The oldest reported fossil anos- tracan is }Gilsonicaris rhenana from the Upper Devonian of Germany (Van Strae- len, 1943). This species possessed eighteen trunk segments, eleven of which bore appendages. It had closest affinities to the Eocene +Branchipodites vectensis but dif- fered in the greater number of segments. +Branchipusites anthracinus from the Up- per Carboniferous possessed eight trunk segments with lateral appendages resem- bling those of the lamellar branchial feet of living Branchipus. +Rochdalia parkeri from the British Coal Measures had eleven segments. Palmer’s Miocene anostracans also possessed eleven trunk segments. Exclusive of the family Polyartemiidae, extant families of anostracans all have eleven thoracic segments. Thus, by Car- boniferous time, in this respect, anostra- cans had a modern aspect. Branchipus- type appendages, noted above, also are known from the Carboniferous. In turn, this suggests that appendages also had a modern aspect at that time. The chief trend discernible in the fossil record of anostracan evolution is a de- 154 crease in number of segments between De- vonian and Carboniferous time. The order {Lipostraca known only from the Rhynie Chert of the Devonian Old Red of Scotland bears closest affinities to the Anostraca. Scourfield (1926) was im- pressed by the combination of primitive characters (such as the biramous II an- tenna) with greater specialization than is found in living anostracans. He cited fif- teen points of difference between lipostra- cans and anostracans. Nevertheless, other workers have considered lipostracans to be primitive anostracans. Scourfield else- where (see Tasch, 1957) remarked that in the earlier stages, the mandibular palp in lipostracans was ‘“‘remarkably identical” with that found in the anostracan Chiro- cephalus diaphanus. He thought that this pointed to a pre-Devonian ancestry for both. If we follow that line of reasoning, a pre-Devonian ancestral type gave rise to two branches, anostracans in Lower De- vonian (perhaps a }Gilsonicaris-type) and lipostracans in Middle Devonian time. Since Lipostraca are unknown after Rhynie Chert time, it appears that this order with two distinct series of append- ages (branchiopod and copepod) was an unsuccessful evolutionary experiment. By contrast, once reduction of segments was achieved by Carboniferous time, the Anos- traca continued to the present day rela- tively unchanged in basic plan while dif- ferentiating into several families. ,ACERCOSTRACA, NOTOSTRACA, AND 7KAZACHARTHRA The 7Acercostraca known only from the Hunsrtick shale, Lower Devonian of Germany, closely resemble the notostra- cans in general, and Triops (=Apus) in particular. The common characteristics in- clude: a dorsal carapace, a pair of sessile MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 eyes, small antennae, and about fifty pairs of appendages. However, the acercostra- cans lack the characteristic notostracan telson and furca (Lehmann, 1955). Notostracans known as fossils range from the Upper Carboniferous to Recent. Fossil forms are closely similar to living forms. This evaluation applies to species of both Triops and Lepidurus. The last- mentioned genus is known from the Trias- sic of South Africa and the Lower Creta- ceous of Turkestan. There is little doubt that by Carboniferous time notostracans had a modern aspect. It appears that Triops, possessing a telson and rudimen- tary supra-anal plate, gave rise to Lepi- durus, which possesses both a telson and a supra-anal plate, somewhere between Per- mian and the dawn of Triassic time. A recently named new order, +Kaza- charthra, known only from the Lower Jurassic of Kazakhstan, possessed a. tel- son without a terminal segment. The mouth parts were of the notostracan type. In general plan, it closely resembles notos- tracans. However, notostracan resemblance loses its taxonomic importance in Novojilov’s view because the number of short append- ages in the anterior portion of the ventral face is fewer in *Kazacharthra than the number found in notostracans (Novojilov, 1957). The three orders mentioned above show close affinities as well as marked differ- ences (Fig. 67). There are alternative in- terpretations of the evolutionary relation- ships between them. Either +Acercostraca was directly ancestral to Notostraca, or both orders arose from a common pre- Devonian ancestral type. If the last alter- native is accepted, then the +Acercostraca represent an unsuccessful Devonian off- shoot while the Notostraca are late Car- TASCH: BRANCHIOPODA 155 —_— eC, eee Recent ; Cretaceous e M . E S| Jurassic O Z O aus | | Triassic * LEPIDURUS C , Telson & supra- *. anal plate. Permian Carboniferous je KAZACHARTHRA : : Telson without eord terminal segment. *NOTOSTRACA (Triops) Telson; rudimentary supra- : anal_plate. a ———— Lower Devonian Pre- Devonian Fic. 67. boniferous descendants that persist to the present time. {Kazacharthra, in turn, may either be a synonym of Notostraca, or an unsuccessful branch of the notostracan stock in Lower Jurassic time. In this way, we are afforded a long-range view of notostracan evolution from pre-Devonian time on. Thus, notostracans, as was the case with the conchostracans, are clearly a stagnant group in an evolutionary sense (Longhurst, 1955). This might not have been the case if the unsuccessful orders, e ACERCOSTRACA No telson. — COMMON ANCESTOR Inferred relationships between three branchiopod orders. yAcercostraca and +Kazacharthra, had survived. SIGNIFICANCE OF THE PRE- DEVONIAN FOR BRANCHIOPOD EVOLUTION It is apparent from the discussion above that pre-Devonian marine and non-marine deposits (specifically Ordovician and Silurian beds) should contain crucial evi- dence bearing on the origin of several branchiopod orders: Conchostraca, Anos- traca, and 7Acercostraca. The problem 156 narrows down to the derivation of these three orders that made their known first appearances in Devonian time. All the other orders arose directly from one or an- other of these or from collateral lines: Cladocera arose from Conchostraca; Lipos- traca and Anostraca probably had a com- mon ancestor; +Acercostraca and Notos- traca may have had a similar origin; and ;+Kazacharthra appear to be a derivative of the Notostraca. H. L. Sanders (this volume, p. 172) presents the interesting thesis that the ancestral crustacean was not too distant from living Cephalocarida. He indicated five steps leading to the branchiopod mode of development by modification of the common ancestral type. These steps in- cluded, among others, reduction, modi- fication, and loss of appendages in the ancestral crustacean that ultimately led in the branchiopod direction. These steps must have involved a_ con- siderable number of genetic mutations and countless generations of cepha- locarid-like populations. Sanders’ fourth stage in the transition from the ancestral type to true Branchiopoda is the ‘‘evolv- ing of the branchiopod nauplius.” As a tentative theory attempting to place in time the various parts of this sequence, the present writer suggests the following time scale: (1) Late Pre-Cambrian to Lower Cambrian . . . origin of the cephalocarid- like crustacean ancestor. (2) The rest of the Cambrian to older Ordovician . . . head and trunk append- age reduction, modification, and/or loss. (3) Middle to Upper Ordovician time .. evolution of the branchiopod nauplius. (4) Close of the Ordovician .. . estab- lishment of branchiopod mode of de- velopment. MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 (5) Silurian . adaptive radiation, ultimately giving rise to three distinct branchiopod lines: ancestral stock of De- vonian conchostracans, common ancestral stock of anostracans and lipostracans, and common ancestral stock of acercostracans and notostracans. If Sanders’ outline of the steps leading to the branchiopod mode of development by modification of a cephalocarid-like an- cestral crustacean is plausible (and the consensus of crustacean specialists appears to be in favor of it), then the theory of small-step evolution leading from species to higher categories requires large time spans in which such evolution could have occurred. Without having any claim to accuracy, the proposed geologic time schedule outlined above for the pre-De- vonian evolutionary events provides a useful frame of reference to be improved, modified, or corrected as new fossil data become available. It is possible, for ex- ample, that steps 2 to 4 of this time schedule may have to be pushed back to the Cambrian. However, at the present writing the available evidence supports the time schedule as outlined. REFERENCES Apamczak, F. 1961. Eridostraca—a new suborder of ostracods and its phylogenetic significance. Acta Paleont. Polonica 6(1):29-104, 4 pls. Barrp, W. 1849. Monograph of the Family Lim- nadiadae, a family of entomostracous Crus- tacea. Proc. Zool. Soc., London 17:84-90. Brooks, J. L. 1957. The systematics of North American Daphnia. Mem. Conn. Acad. Arts Sci. 13:1-176, 61 text-plates. DeESCHASEAUX, C. 1951. Contribution a la con- naissance des esthéries fossiles. Paléontologie 37:3-10. DEFRETIN, S. 1958. Remarques a propos de la note de N. I. Novojilov sur quelques Con- chostracés chinois et africains. Ann. Soc. Geol. du Nord 67:244-260, 1 pl. Frey, D. G. 1958. The late glacial cladoceran TASCH: fauna of a small lake. Arch. f. Hydrobiol. 54:14-270, pls. 35-41, 113 figs., 6 tables. Gross, W. 1934. Eine Estheria aus dem rhein- ischen Unterdevon. Senckenbergiana 16:309. KosayAsut, T. 1954. Fossil estherians and allied fossils. Jour. Fac. Sci. Univ. of Tokyo 9(1): 1-192, 30 figs. Kumnerow, E. H. E. 1939. Die Ostracoden und Phyllopoden des deutschen Unterkarbons. Abh. Preuss. Geol. Landesanst. (N.F.) 194:78-79. Leumann, W. M. 1955. Vachonia rogeri, n. gen., n. sp., ein Branchiopod aus dem_ unter- devonischen Hunsriickschiefer. Palaont. Z. 29: 126-130, 2 pls., 2 figs. Loncuurst, A. R. 1955. Evolution in the Notos- traca. Evolution 9:84-86. Mattox, N. T. 1957. A new estheriid conchostra- can with a review of other North American forms. Am. Midland Naturalist 58(2) :367-377. NEEDHAM, J. 1930. On the penetration of marine organisms into fresh water. Biol. Zentralbl. 50(8) :504-509. Novoyitov, N. J. 1957. Un nouvel ordre d’Arth- ropodes particuliers: Kazacharthra du _ Lias des monts Ketmen (Kazakhstan, S. E., U.R. S.S.). Bull. Soc. Géol. France (6) WII:171- 185, 2 pls. . 1958. Recueil d’articles sur les phyl- lopodes conchostracés. Ann. Service Informa- BRANCHIOPODA 157 tion Géol. du B.R.G.G.M., 1958, no. 26, 135 pp. 1960. Sous-ordre Conchostraca Sars, 1846, Phyllopodes bivalves. In Osnovy Pale- ontologii, Chlenistonogie Chernysheva. Moscou, Gosgeoltekhizhat, pp. 220-252, figs. 455-586, pls. 13, 14, 15 (Trad. Mme. Streto- vitch). Sars, G. O. 1887. On Cyclestheria hislopi (Baird). Forhandl. Videnskabs-Selskabet. 1:3-60, pls. I- VIII (Christiania). ScourFIELD, D. J. 1926. On a new type of crus- tacean from the Old Red Sandstone (Rhynie Chert Bed, Aberdeenshire)—Lepidocaris rhyni- ensis gen. et. sp. nov. Phil. Trans. Royal Soc., London, Ser. B. 214:153-187, 2 pls., 51 figs. Tascu, P. 1957. Flora and fauna of the Rhynie chert: a paleoecological reevaluation of pub- lished evidence. Bull. Univ. of Wichita, 32, Univ. Studies, no. 36:1-24, figs. 1-7, app. 1, 2. . 1958. Novojilov’s classification of fos- sil conchostracans—a critical evaluation. Jour. Paleont. 32:1094-1106, 33 figs. 1961. Pemphilimnadiopseidae, a new family of fossil conchostracans. Jour. Paleont. 35:1117-1120, pl. 132. Van STRAELEN, V. 1943. Gilsonicaris rhenanus nov. gen., nov. sp., Branchiopode Anostracé de lEodévonien du Hunsriick. Bull. Mus. Roy. Hist. Nat. Belg. 19, no. 56:1-10, 1 pil. PHYLOGENY AND EVOLUTION OF CRUSTACEA Museum oF CoMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 XII Discussion Following Paper by Tasch, and Sum- maries of Recently Published Papers by Brooks’ and Rolfe’ SANDERS: In the conchostracan Lyn- ceus, the valves of adult and nauplius are different. A paleontologist finding the flat larval valve would not put it into the Conchostraca; he would probably make it a notostracan. Perhaps some difficulties in identifying fossils may be due to such differences in developmental stages? TASCH: That’s true for the Lynceidae which are known as fossils only from the Lower Cretaceous of the Transbaikal. For the other families, we have the addition of growth bands. Various studies have shown that individuals molt approximately every three days. You can count how long a given individual lived. Some of these individuals get to be quite a size. I ac- 1 Brooks, H. K. 1962a. The Paleozoic Eumala- costraca of North America. Bull. Amer. Paleont. 44:163-338, 66 pls. . 1962b. Devonian Eumalacostraca. Ark. f. Zool. (2) 15(19):307-315. 1962c. On the fossil Anaspidacea, with a revision of the classification of the Syn- carida. Crustaceana 4(3):229-242. 2 Rolfe, W. D. I. 1962a. Grosser morphology of the Scottish Silurian phyllocarid crustacean Ceratiocaris papilio Salter in Murchison. Jour. Paleont. 36:912-932. . 1962b. Two new arthropod carapaces from the Burgess Shale (Middle Cambrian) of Canada. Breviora, Mus. Comp. Zool. 160:1-8. 159 knowledge what you say as a caution, but I think workers on fossil conchostracans have been aware of this fact. GLAESSNER: In a recent paper (Jux, J. Paleont., 1960, 34:1129-1152) a large number of +Archaeostraca were con- sidered as Malacostraca. Has this been definitely ruled out? ROLFE: It is interesting historically that the argument over the Leptostraca being Branchiopoda or Malacostraca has always been present to some extent with the fossils. We have just had a renaissance of this problem by Ulrich Jux, who has suggested that all the +Archaeostraca are Branchiopoda. He bases this on one genus which he has studied: +Montecaris. And it is exactly this problem that I outlined previously: a maximum of three segments is preserved, projecting from the cara- pace of one specimen available to Jux; this is different from any malacostracan number and hence he concluded that ;+Montecaris must be a branchiopod. Now, however, we know that +Montecaris has seven pleomeres and at least seven thora- comeres. Impressions of the mandibles were interpreted by Jux as the labrum. To my mind there is no basis for even this one genus being a branchiopod. Some 160 archaeostracan genera are certainly mala- costracan (see Rolfe, 1962a, p. 930); others are incertae sedis. BROOKS: Tasch has eliminated pre- Devonian forms from the Branchiopoda. What would he do _ with +Fordilla, *+Bradoria, and the various Middle Cam- brian fossils from Nova Scotia? TASCH: I cannot place them in the Branchiopoda, but other workers consider that some of these belong to a transitional group. BROOKS: What do you think of the Ordovician +Douglasocaris, which was de- scribed as a phyllocarid rather than a branchiopod? It has cercopods rather than a furca, and a single ramus in the antenna. It is not a malacostracan. ROLFE: It does not have the produced archaeostracan telson. BROOKS: It has nine abdominal som- ites: it is a branchiopod. So the branchi- opod record does go back further than the Devonian. TASCH: I agree. You can infer this from the high development in Devonian forms. We have to distinguish between the acceptable (i.e. the undisputed) fossil record of branchiopods and their pre- Devonian origins inferred from biological data of living branchiopods and related forms. The undisputed fossil record of branchiopods that is presently available begins in the Devonian. BROOKS: About the feeding habits of Paleozoic Crustacea. All the older syn- carids had raptorial appendages, so that to some extent they were predaceous. In- testinal fillings are characteristic of Para- naspides and Anaspides which scrape ma- terial and in feeding take in a lot of detritus. Paleozoic forms and +Anas pidites in the Triassic, show that they probably had the same feeding habits as Anaspides MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 or Paranaspides: they have intestinal tracts filled with enough detritus to reveal the course of the intestine through the body. These intestinal fillings are so characteristic of the different genera (none of the forms that I would say was preda- ceous ever has them) that when we get to 7Palaeocaris, all we have to do is look at the intestinal filling, because it is a generic character. MANTON (subsequent written com- ment): Brooks states that Paranaspides and Anaspides scrape material and take in a lot of detritus, with no further quali- fication. I think that I am the only person who has studied these animals alive (1930, Proc. Zool. Soc. London, pp. 791-800, 1079) and I would not endorse this state- ment. Paranaspides is largely a_ pelagic feeder in a lake rich in plankton. I stated that “the stomach contents consisted of an unrecognizable mass of detritus, finely divided and containing numerous di- atoms,” a common type of content of ali- mentary canals of filter-feeders. I did not use the term “detritus” in the sense in which Brooks uses it to denote a mass of mineral particles. We need a careful defi- nition of detritus, and much more work on the gut contents of extant Crustacea be- fore we can interpret fossil alimentary canals with the certainty suggested by Brooks. Anaspides is partly predatory and it scrapes algae off stones. Particles are collected from an enclosed space by the filter mechanism which prevents food be- ing swept away by the flowing streams in which the animals live. I should also like to question the basis of the statement that all the older syn- carids had raptorial appendages. I should like to see much more detail of all the limbs before accepting this statement. GLAESSNER: What about the origin DISCUSSION and history of the carapace in these early stages? Is it the archaeostracan carapace that undergoes these changes? Why do these groups get rid of it and how? Par- ticularly the syncarids.* BROOKS: I have no proof that the Malacostraca are monophyletic, but rather that they are polyphyletic. Possibly +An- thracocaris was a stage in the reduction of the carapace from the primitive caridoid Malacostraca. There is an alternative in- terpretation of this: that, as in the branchiopods, there are those with and those without carapaces. It is possible that some of these forms, for example the Tanaidacea, are different, and that amphi- pods, isopods, and syncarids represent a different phyletic line that never had a carapace to start with. I think we have 3 See also discussion of Peracarida problem, p. 181. 161 to admit that all lines that we can draw on from the study of Recent Crustacea indicate that the caridoid facies in the Malacostraca is primitive. I think that the Eumalacostraca were derived from within the complex of the Phyllocarida. If this is so, the primitive eumalacostracan must have had a carapace. We can reject this alternative interpretation. The strati- graphic record now supports Calman and other theorists that the caridoid facies is primitive. MOORE: From all we have heard in this conference it makes eminent sense to group together, as a reservoir of ancient Crustacea, those forms that have these divergent relationships to known modern groups. DAHL: Do you think that your line leading up to the Stomatopoda has to be derived from the Phyllocarida too? BROOKS: I suspect so because of con- SUPERORDERS OF RECENT EUCARIDA PAN- PERACARIDA HOPLO-| syNCARIDA EUMALACOSTRACA CARIDA CARIDA 0 A .-¢ ORDERS OF RECENT EUMALACOSTRACA EUPHAUS DECAPODA THERMOS- BAENACEA MYSIDACEA SPELAEO GRYPHACEA CUMACEA ISOPODA -| AMPHIPODA — | STOMATOPOD, BATHY- NELLACEA ANASPIDACEA X Schim- \ \’ perelia XN \ Palaeo- | \pemphix] \Isopod-/ \ites / | oO. PAL AeO-, 4 “XN \ fo. ANTHRACO-/7% \ CARIDACEA BS Oo =z a = 12) fo} Oo 4 4 =< = > W (=) lJ E S oO = 4 o x N WW (o) z a WW oO < a W Dor oS (7) e a 4 W WJ a n 3 o a uJ a x= E = ier |e Yon Fic. 68. (1962, Bull. Amer. Paleont., 44, text-plate 16). ~~ 0. EOCARIDACEA <% _\ | CARIDACEA fh : \ nor} \ pale / "0, Phylogeny and classification of fossil and Recent Eumalacostraca, by H. K. Brooks 162 spicuous differences in the articulated rostrum. There is other morphological evi- dence to suggest the same thing. I could not see them grouped with other Eucarida. DAHL: You still think it probable that they are not a separate line, but join the others at the phyllocarid level? BROOKS: Yes. The problem is that we don’t know anything within the Phyl- locarida that would be a suitable ancestor. Our unknown would be a good link there. Interestingly, fHocaris does not have a rostrum. A rostrum in the regular caridoid is a secondary development. ROLFE: Would Manton care to com- ment on the “furcal rudiments” on the telson that Brooks has emphasized? Are they genuine furcal rudiments, or could they be interpreted as anything else? MANTON: It seems to me a good interpretation, although I don’t know much about furcal rudiments. I only know them from following them out in Nebalia and a mysid. In the mysid they are formed directly from the telson cuticle and at the first ecdysis they are shed with the old cuticle and are not re-formed. BROOKS: Furcae are still found in the euphausiids and in the Bathynellacea. There is no question but that the furca is a primitive crustacean characteristic. We even have it in the Eumalacostraca; the furca is more widespread in Malacos- traca than we realize. MANTON: The basic evolution of Crustacea probably took place in a marine environment. We know a large number of fossils from freshwater deposits. Have we in fact a reasonable representation of the marine Crustacea? We should not build MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 evolutionary castles on a preponderance of freshwater forms. BROOKS: I am confident that the oldest forms are marine. I believe that the evolution and adaptive radiation as we know it was in the marine environment. Much of the fossil record, is, however, in freshwater deposits, and is sporadic. GLAESSNER: Could you name _ the genera that are associated with marine fossils? BROOKS: 7Palaeopalaemon, from the Ohio Shale, and +Devonocaris from the Moscow Shale of New York and also from Belgium are associated with marine fossils. ;Eocaris is found in the uppermost Middle Devonian of Western Germany and _ is associated with the phyllocarid +Monte- caris and a dominantly marine fauna. The records of the others are sporadic; even when many are found, it is only in ab- normal environments. They were not pre- served in a marine environment, just as there is no record of the king crab in marine deposits, even though the modern form comes into estuarine waters and fresh water only to breed. Otherwise it is truly marine, and are we to believe that the king crab existed throughout the Ter- tiary in fresh water and then became marine in the Recent? No, it is selective preservation. GLAESSNER: May I add two or three unpublished facts to the record. Birshtein in Moscow has a Devonian crus- tacean of this general character; I think it was obtained in marine strata from a borehole. Malzahn in Hanover has found fossils from the Upper Permian, which are primitive tanaids and Cumacea with eye- stalks. PHYLOGENY AND EVOLUTION OF CRUSTACEA Museum or CoMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 XIII Significance of the Cephalocarida' INTRODUCTION This paper is concerned with external anatomy, and internal anatomy is con- sidered only insofar as it impinges on problems relating to external morphology. The objectives are to: 1, justify the sub- class ranking of the Cephalocarida; 2, in- dicate possible malacostracan-cephalocarid relationships; 3, demonstrate branchiopod- cephalocarid relationships; 4, indicate the basic phylogenetic position of the Cephalo- carida among the Crustacea. The intent is not to show that the present-day Cephalocarida represents a close replication of the original ‘“Urcrusta- cean,” but rather an early grade of de- velopment in the crustacean series. A more extensive work (Sanders, 1963) should be consulted for more detailed con- sideration of the Cephalocarida and docu- mentation of topics discussed here. SERIAL HOMOLOGY One of the most characteristic features of the Cephalocarida is the pronounced 1 Contribution No. 1230 from the Woods Hole Oceanographic Institution. This research was supported by National Science Foundation Grants G-4812 and G-15638. By Howard L. Sanders Woods Hole Oceanographic Institution Woods Hole, Massachusetts 163 serial homology of the limbs (Figure 69). The trunk limbs and second maxilla are essentially identical appendages consisting of a foliaceous exopod, a lateral foliaceous pseudepipod, an ambulatory endopod, and a flattened, although slightly posteriorly Fic. 69. Ventral view of an adult specimen of Hutchinsoniella macracantha. 164 concave, protopod with a series of endites on the medial edge. The entire limb is concave posteriorly. This limb series rep- resents the most primitive state of tag- mosis known in the Crustacea. The identity of the second maxilla has been questioned, since in the branchiopods this limb is frequently reduced to a minute protuberance and therefore easily over- looked. However, there can be no doubt that this limb is the second maxilla, for the large maxillary gland opens on its pro- topod. The first maxilla with its foliaceous exopod, ambulatory-like endopod, and pro- topodal endites is obviously related to the more posterior limbs, particularly in the larvae. The second antenna and larval mandible are structurally alike. In both there is a many-segmented exopod with similar seta- tion, and the endopod is two-segmented. The enditic spines are found on the pro- topod of these limbs and on both pairs of maxillae and all the trunk limbs as well. The external morphological evidence for pronounced serial homology in Hutchin- soniella finds strong support in Dr. Robert R. Hessler’s studies (unpublished) on the skeletal musculature of this crustacean. The muscle pattern is the same in all the thoracic limbs and the second maxilla. In the first maxilla and mandible the muscu- lature departs progressively from the pat- tern of the more posterior limbs but, nevertheless, is clearly derived from it. The limbs reflect their high degree of serial homology in their multiplicity of function. All limbs participate in the trophic function. Current production is generated by the movements of the first maxilla, second maxilla, and trunk append- ages. The first antenna, second antenna, larval mandible, second maxilla and trunk MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 limbs are locomotory. The second maxilla, trunk limbs, adult first maxilla, and larval second antenna, as well as the setal combs on the terminal abdominal segment and telson and the long terminal spines on the telson, contribute to grooming activities. The sensory function seems to be re- stricted to the first antenna. MODE OF NAUPLIAR DEVELOP- MENT To clarify the relationship of the Cepha- locarida to other Crustacea, particularly the Branchiopoda, a comparative study was made of the one type of larva com- mon to all the crustacean subclasses: the nauplius. The word “nauplius” is used here to mean both nauplius and metanauplius. The primitive mode of development in the Crustacea was most probably a continuous and sequential addition of seg- ments and limbs through a large number of moults (Calman, 1909). In Figure 70 the mode of naupliar-metanaupliar devel- opment in the Cephalocarida, Branchio- poda, Cirripedia, Copepoda, and Mystaco- carida is compared. The pattern in the Cirripedia departs appreciably from the proposed primitive pattern. Neither segments nor limbs are added during the first five stages. At the sixth or terminal stage a segment and the first maxilla are added. At the moult to the first cyprid stage a large number of segments and limbs are added. The Copepoda conform more closely to the suggested pattern. There is a gradual addition of segments up to stage 6 when three segments are added. The addition of limbs is not continuous, a single limb being added at stage 3, and two limbs each at stages 5 and 6. Mystacocarid naupliar development shows a fairly regular addition of segments nau nai SANDERS: CEPHALOCARIDA 165 through a long series of 10 stages. The stage 7 and not in sequential order, and addition of limbs is highly irregular with three at stage 8. a single limb appearing at stages 2 and 3, The mode of development in the Bran- none during stages 4, 5, and 6, two at chiopoda is characterized by the addition _ = J =) , Fe : — : 8 BRIBIEL 7 >) Silas) 6 *}6 a ] yy J Jf ds 15 5 4 Toon) acs 5 0} o 24 °l4 el, °}3 ] o}jojjolje 2} 3 e}3 0} |e}, ) }] J} Jejeojjeo} se}, ej2 o||e I dededed eds 5 5 x] 4 + Ole ©} |¢) mxeo 0}/0]/0}]0] |e] wxep ©) mx2 o}} el mx2 O}}oflo}foljo}jo}]o}{o} |e} mx2 2} Mx1 [e} 1¢ | mx1 [OH OH Ojo ieye ie rey fey mx elle elfejie| lel up ellelle e|lelmp elleljo|jellelfel/e|je| le) mp e\\e e@e|;e|;e @ AN2 @\j\ell\e e e AN2 Coie Sie ee ee ee ee ee) @ AN2 ejlejtejlestestey Ley ani ei} eile 2) lean Le}lejlejlejlejie jie ste) Le} ani Wa SOI" Ge 7 cies Gree, Se i : ; : ' di s Ue ct Hus IRRIPEDIA | or nauplius’ ¢opEPODA woomerodite.) | nauplius — ‘avsTacocaripa * Juvenile [2 5 419 ] J| Jus }] Ja | }} Jie 1 J15 he 14 5 at — ] ] 43 | ] 12 4 4 } AIG 11 [a ] ] PRBRIEA iP PPPEEPREIE SU SAIS 4] 4 4 4 4 olle e 8 ] AN 4} J 5 e) e e) 7 } ie 5 > y| ) °] elle e 6 J J o] °} e)|e} e) 5 } ) J fo} e| e)\e} ejjejje n Jlo}jo}le °} e t 5 > BY + °) elle ° 3 ) °) ejjejje e elle) 5 AIDE ollo o} ellellelle e| e] ellelle e 1 o}}e)}/e) e e] elle e ellelle e) |e) mx2 ellelle elle e||e} el\elle e MX1 el\ei\e e\\e e |e | elle e MD ellelle e [el elle elle e AN2 ellelle elle elle elle ANI ejlejjejjejlejle elle IDE2 ERS hrs es5 i 6h o7, iear2. 4567 8 9 11 12 uplius ' juvenil nauplius BRANCHIOPODA’ /YY<""'® CEPHALOCARIDA Fic. 70. Mode of naupliar development in the Cephalocarida, Branchiopoda, Cirripedia, Cope- poda, and Mystacocarida. The bar denotes the number of segments; circles denote limbs present. Black circles represent limbs with the definitive form, open circles represent limbs that are rudi- mentary. The graphs are derived from the descriptions of Heath (1924), Oberg (1906), Delamare Deboutteville (1954), and Sandison (1954). 166 of numerous segments and limbs at a single moult during the early naupliar stages. The full complement of segments and most of the limbs are already present midway through the naupliar series. Such a developmental pattern is applicable to both the phyllopod and anostracan_ bran- chiopods. A ETON Fic. 71. MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 The cephalocarid naupliar development approximates the suggested primitive pat- tern more closely than does any other group. There is a continuous addition of segments through a long series of 13 nau- pliar stages. The addition of limbs is much more regular than it is in other Crustacea. Differentiation of the limbs proceeds at a Representative branchiopod nauplii: A, reconstruction of the lipostracan +Lepidocaris rhyniensis (from figures of Scourfield, 1940); B, the conchostracan Limnadia lenticularis (adapted from Sars, 1896); C, the notostracan Triops cancriformis (adapted from Claus, 1873); and D, the anostracan Artemia salina. SANDERS: CEPHALOCARIDA nearly constant rate, and usually only the terminal appendage at any given stage is rudimentary. NAUPLIAR AFFINITIES If we disregard the Cephalocarida for the moment, a study of naupliar mor- phology reveals that crustacean nauplii 167 can be clearly separated into two groups. The Branchiopoda make up one group (Fig. 71), which can be characterized by: (1) a small and unsegmented first antenna with setation restricted to the distal edge; (2) the length of the protopod on the second antenna, which is more than half the total length of the limb; (3) the Fic. 72. Representative from Oberg, 1906); B, the mystacocarid Derocheilocaris remanei (adapted from Delamare Deboutte- ville, 1954); C, the cirripede Balanus balanoides; and D, the decapod malacostracan Penaeus duo- rarum (adapted from Dobkin, 1961). nonbranchiopod nauplii: A, the copepod Temora longicornis (adapted 168 presence of a single large spine on the medial surface of the distal protopodal segment of the second antenna; (4) the absence of setae from the medial surface oi endopod on the second antenna; and (5) a uniramous mandible. In contrast, the nonbranchiopod group (Fig. 72) has (1) a first antenna that is long and segmented, with setation not limited to the distal edge; (2) a protopod on the second antenna less than half the total length of the limb; (3) two or more small spines on the medial surface of the protopod of the second antenna (absent in penaeids since the nauplii do not feed) ; (4) setation on the medial surface of the endopod of the second antenna; and (5) a biramous mandible. * . a \ Fic. 73. Stage 4 nauplius of Huichinsoniella macracantha. The cephalocarid affinities (Fig. 73) are wholly with the nonbranchiopod naupliar group: (1) the first antenna is large and MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 segmented rather than small and unseg- mented, and setation is not limited to the distal edge but is present on several proxi- mal segments as well; (2) the protopod of the second antenna is appreciably less than half the total length of the limb; (3) three spines rather than one large spine arise from the medial surface of the distal protopod segment of the second antenna; (4) two setae rather than none occur on the medial surface of the endopod of the second antenna; (5) the mandible is biramous rather than uniramous. This comparison of naupliar mor- phology demonstrates the difficulty of at- tempting to include the Cephalocarida within the Branchiopoda. MALACOSTRACAN RELATIONSHIPS Except for the Leptostraca, the more primitive representatives of the major groups within the Malacostraca have a common type of morphology from which the more specialized elements diverge widely. The anatomical features compris- ing this general type of morphology are referred to as the caridoid facies, and it is assumed that these features were present in the hypothetical ancestor of the Eu- malacostraca. The major factor obscuring the past history of the Malacostraca is the almost universal tendency for development to continue within the egg to a very late stage; the juvenile that finally emerges has already acquired the adult features. The stomatopods, euphausiids, and de- capods do have free-living larval stages, and it is in the penaeid decapods that development retains many of what we may suppose to be its primitive features. Analysis of the larval stages of Penaeus setiferus (Fig. 74) demonstrates that, by and large, the caridoid facies appear late SANDERS: CEPHALOCARIDA 169 jst 20d 30 1s 2nd 1s 1 Nauplius Protozoea Protozoea Protozoea Mysis Mysis Pastmysis Pastaysis Adolescent a | , | : ZA / Le | d oo x |5Z Je ae | Schell | : | S| P ia | NZ Wee za eee | | = | Vie | E~ i a VEY T i T = = | a 73 | 7 4 = re] oan ra ee eo Nes (| 4 3 - see |} WE WA BT Vela = BO ila e a eee eel oe | | ES AY lh FSS : es = Sea | paca sco ae lo. iW : a | xz | ye eA EN PEA Ve |. /2 ond MY > Pos es as # Zz AS 43 So | oe eae a es ee a | Os. IL 2 pees — ! | ] 2m IPD ie | ae | ee | ow | | es | $ : 2 hese | 5 aeons = 7 ar | =a | | j 4 aM) C= | Ab | 2A | po) | Pe | | 42 eo rary 4 STV ) se Hi | i. 4 Fic. 74. Developmental stages of the limbs in the penaeid decapod Penaeus setiferus (recon- structed from Heegaard, 1953). The limbs are arranged serially along the ordinate, the develop- mental stages along the abscissa. in larval development. The incipient biramous first antenna, the scale-like exopod and flagelliform endopod of the second antenna, the palp on the mandible, and the functional uropods are first evi- dent during the mysis stages. The func- tional pleopods make their appearance during the post-mysis stages, and the flagelliform exopods of the maxillipeds are not present until the juvenile stages. From this developmental series it seems apparent that a number of the caridoid features—the uniramous mandible, the flagelliform exopod on the maxillipeds, the scale-like exopod and flagelliform endopod of the second antenna—are, at least mor- phologically, secondarily derived. The morphology of the penaeid larva at the protozoeal stages, before the ap- pearance of most of the caridoid features, is much closer to certain entomostracan groups, particularly the Cephalocarida, than to its own later developmental stages. This morphological similarity can be demonstrated for each limb. Only the penaeid and cephalocarid second antenna will be compared here. (For a detailed comparison of other limbs see Sanders, 1963.) In both there is a many-segmented exopod with each segment, except the most proximal two or three, carrying a large stiff seta on the ventral surface and with the terminal segment bearing three setae. A two-segmented endopod carries 4-5 setae or spines distally and bears two or more setae on its medial margin, and a two- or three-segmented protopod. It is concluded that the caridoid mor- 170 phology is acquired late in development. In those few malacostracan groups having a fairly complete series of larval stages, the morphology before and after the ac- quisition of the caridoid facies is decidedly different. The pre-caridoid morphology is remarkably similar to the morphology found in the Cephalocarida and suggests that the Malacostraca arose from a crus- tacean stock that had many features in common with the present-day Cephalo- carida. ESTHERIA TYPE DAPHNIA TYPE Fic. 75. 1933). For interpretation see text. MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 BRANCHIOPOD RELATIONSHIPS Cannon (1933), on the basis of his detailed and elegant functional study of the feeding mechanisms of the Branchi- opoda, has presented a very convincing scheme (see Fig. 75) explaining the rela- tionships of the various branchiopod limb patterns. He postulated that the proto- branchiopod trunk limb most probably had six unmodified, medially directed endites with spines along the enditic ridge ENDOPODITE EXOPODITE ANCESTRAL TYPE LEPIDOCARIS TYPE DIAPHANOSOMA TYPE CHIROCEPHALUS TYPE Interrelationships of the filtering trunk limbs within the Branchiopoda (from Cannon, SANDERS: and with a series of widely spaced anterior and a series of widely spaced posterior setae. From such a pattern two types of limbs evolved. One line led to the modern Notostraca and was brought about by the enlargement of the basal endite which turned forward and so overlaid the proxi- mal endite of the limb in front. The other line led to all the remaining branchiopods and was brought about by the backward projection of the proximal endite. The latter condition was realized in the an- terior limbs of the mid-Devonian +Lepido- caris rhyniensis. The backward projection of the other endites and the conversion of the widely spaced row of posterior setae into a closely meshed row of filter setae gave the condition present in the Con- chostraca, Cladocera, and Anostraca. Both the Cephalocarida and Cannon’s protobranchiopod have, in contrast to all known branchiopods, six unmodified en- dites, i.e. the proximal endite has not been transformed into a gnathobase and then further altered. The Cephalocarida share the following features with the Notostraca and }Lepidocaris among the _ Branchi- opoda: endites (distal endites only in the Notostraca and }Lepidocaris) medially or slightly posteromedially directed; a row of spines present on crests of endites; a row of widely spaced posterior and a row of widely spaced anterior setae (distal endites only in the Notostraca and +Lepi- docaris). Because of the large degree of agree- ment between the cephalocarid and the proposed protobranchiopod limb, and since the Cephalocarida share morphologi- cal features with only the most primitive branchiopod components (Notostraca and +Lepidocaris), it seems probable that the limb structure in the Cephalocarida rep- CEPHALOCARIDA resents a condition precursory to that found in the Branchiopoda. The limbs in the cephalocarid move with a marked metachronal beat which is initiated posteriorly and passes forward. The metachronal limb movements create currents which are used by the animal in conjunction with feeding (Fig. 76). In the adult, a strong flow of water is drawn posteriorly, medially, and dorsally into the median chamber between the paired thoracic limbs and thence into the inter- limb spaces on the forestroke or suctional phase (Fig. 76, positions 1-6). On the backstroke, with the reduction of the inter- limb space, water is forced out postero- laterally as a series of jets between the exopods and pseudepipods (Fig. 76, posi- tions 7—9). Detritus is put into suspension by the broom-like action of the second antennal exopod and the distal endopodal spines of the second maxilla and trunk limbs. This secondarily suspended detritus is drawn into the median chamber on the forestroke and is caught on the posterior setal row of the endites and endopod. Because of the interdigitation of the an- terior with the posterior setal row (Fig. 76, positions 9 and 1), these relatively large detrital masses are scraped back into the median chamber where they are carried dorsally on the next forestroke. The enditic spines of the protopod then move the detrital mass forward from limb to limb to the enditic process of the first maxilla, which passes it into the atrium oris. It requires no stretch of the imagination to derive the most primitive known bran- chiopod trunk limb feeding mechanism from the cephalocarid mechanism. All that is required is the loss of the endopod (pos- 172 ‘ y Vy as MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 Fic. 76. Sagittal view of the trunk limbs in Hutchinsoniella showing the various positions during a metachronal cycle together with the associated water currents. The arrow above each trunk limb indicates the relative movement. sibly as an adaptation to pelagic life) and the modification of the proximal endite. On the other hand, the cephalocarid trunk appendage uniquely has all the components present in the nebaliid mala- costracan trunk limb feeding mechanism. In fact, it is much easier to derive the leptostracan feeding mechanism from one similar to the cephalocarid type (Sanders, 1963) than from a_ eumalacostracan scheme, since such a derivation does not entail loss of parts, acquisition of new parts, or a significant number of hypothet- ical stages. The ladder-like arrangement of the ven- tral nerve tracts found in many branchi- opods (Anostraca, Conchostraca, and cer- tain Cladocera) has been considered to be a retention of the pattern found in some Annelida (Calman, 1909). However, those branchiopods that lack a food groove, the Notostraca and certain Cladocera, also lack the ladder-like arrangement of the ventral nerve tracts. Hutchinsoniella, like the Notostraca, has the paired ventral nerve tracts close together. They are just dorsal to the food groove; in fact, the nerve ganglia protrude ventrally into the ventral gully. The more deeply invaginated food groove of the Anostraca and Conchostraca is directly opposed to the midgut, and the ventral nerve tracts are situated lateral to the food groove. Probably the wide, ladder-like nerve tracts are merely the result of lateral displacement by the deepening food groove. THE CEPHALOCARID-LIKE CRUS- TACEAN PROTOTYPE If there is any justification for the belief that the Cephalocarida represent a retention of an early stage of morpholog- ical development in the Crustacea, then it should be possible to relate the diverse patterns in each limb of the various sub- classes to that found in the Cephalo- carida. This has been done for each of the appendages. The limb patterns present in SANDERS: CEPHALOCARIDA 173 OSTRACODA COPEPODA MYSTACOCARIDA MALACOSTRACA BRANCHIOPODA Fic. 77. Relationship of the first maxilla of various crustacean subclasses. The length of the heavier portion of the radiating lines gives a measure of the similarity of those limbs to their cephalocarid counterpart. Upper cephalocarid limb is naupliar and lower is adult. Copepodan limb A is naupliar and B is adult. Malacostracan limb A is decapod protozoeal and B is the adult decapod. 174 the various crustacean groups were found to be either directly comparable to their cephalocarid counterparts or could be derived from them by reduction and sim- plification (see Sanders, 1963). As an illustration, the first maxilla will be considered here (Fig 77). The larval cephalocarid limb consists of a flattened, leaf-like exopod, bearing setae laterally and distally, a multi-segmented endopod with each joint bearing two setae on its medial surface and carrying three or more spines or setae distally, and a protopod with two or more endites bearing gnathic spines or setae. In the protozoea of the penaeid Malacostraca these features are all present, except for the fewer endopodal segments and a reduction in the size of the exopod. Among both the Copepoda and Ostracoda there has been a fusion ANCESTRAL STAGE MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 of the endopodal segmentation, but other- wise all the components persist. The first maxilla of the Mystacocarida is precisely the cephalocarid larval pattern lacking, however, the exopod. In the adult Ce- phalocarida the naupliar endites disappear and are replaced by a larger, more proxi- mal enditic process bearing a number of anteriorly directed spines. The highly reduced branchiopod first maxilla is mor- phologically and functionally similar to the enditic process of the adult cephalo- carid first maxilla. This analysis of comparative limb mor- phology demonstrates that: (1) only the Subclass Cephalocarida, among the known subclasses, possesses a limb series suffi- ciently generalized to indicate a central position within the Class Crustacea; (2) most crustacean groups can be derived Pronounced serial homology. 2. Complex, generalized limbs with foliaceous exopods, ambulatory endopods and endites. BRANCHI OPODA 1. Reduction of head appendages. 2. Modification of proximal endite on trunk limbs. 3. Loss of endopod on trunk limbs re- sulting in phyllopodium. 4. Evolving of branchiopod nauplius. Branchiopod mode of development. MALACOSTRACA Malacostracan first antenna, second antenna, and mandible (ontogenetically preceded by limbs similar to the Cephalo- carida), LEPTOSTRACA EUMALACOSTRACA 1. Enveloping carapace; ambulatory nature of thoracic endopod lost; oar-like movement of first antenna created current which was drawn backwards by paddle-like move- ments of thoracic endopods. 2. Particles filtered off by setal series on endopod and protopod of thoracic limbs (as in Cephalo- carida),. 3. Loss of exopodal scale on second antenna, 4, Fixation of abdominal segmentation at?. Maxillary pump (anteriorly con- cave), 2. Flagelliform thoracic exopods. 3. Uropods. 4, Fixation of abdominal segmenta- tion at 6 in postembryonic stages, Fic. 78. A scheme for the derivation of the Branchiopoda and Malacostraca. SANDERS: CEPHALOCARIDA from an ancestor having most of the mor- phological features of the Cephalocarida; (3) crustacean evolution probably con- sisted of simplification, reduction, and specialization. A suggested scheme for the derivation of the Branchiopoda and Malacostraca is shown in Figure 78. REFERENCES Carman, W. T. 1909. Crustacea. In Ray Lan- kester (ed.), A Treatise on Zoology, Part VII, fasc. 3. London, 346 pp. Cannon, H. G. 1933. On the feeding mechanism of the Branchiopoda. Phil. Trans. Roy. Soc. London, (B) 222:267-352. Craus, C. 1873. Zur Kenntniss des Baues und der Entwicklung von Branchipus stagnalis und Apus cancriformis. Abhandl. K. Gesell. Wiss. Gottingen, 18:1-48. DELAMERE DEBOUTTEVILLE, Ca. 1954. Recherches sur les crustacés souterrains. III. Les développe- ments postembryonic des Mystacocarida. Arch. Zool. Exp. Gén., 91 (fasc. 1) :25-34. 175 Dopxtn, S. 1961. Early developmental stages of the pink shrimp, Penaeus duorarum from Flor- ida waters. U.S. Fish and Wildlife service. Fishery Bulletin, 61:321-349. HeatuH, H. 1924. The external development of certain phyllopods. Jour. Morph., 38:453-483. HEEGAARD, P. E. 1953. Observations on spawning and larval history of the shrimp Penaeus seti- ferus (L.). Pub. Instit. Mar. Sci. Univ. Texas, 3:74-105. OperG, M. 1906. Die Metamorphose der Plankton- Copepoden der Kieler Bucht. Wiss. Meere- suntersuch. Abt. Kiel, 9:37-103. Sanpers, H. L. 1963. The Cephalocarida. Func- tional morphology, larval development, com- parative external anatomy. Mem. Connecticut Acad. Arts Sci., 15:1-80. Sanpison, E. E. 1954. The identification of the nauplii of some South African barnacles with notes on their life histories. Trans. Roy. Soc. South Africa, 34(1):69-101. Sars, G. O. 1896. Phyllocarida og Phyllopoda. Fauna Norvegiae, 1:1-140. ScourFiELp, D. J. 1940. Two new and _ nearly complete specimens of the Devonian fossil crustacean Lepidocaris rhyniensis. Proc. Linn. Soc. London, 152:290-298. PHYLOGENY AND EVOLUTION OF CRUSTACEA Museum oF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 XIV Discussion Following Sanders’ Paper MANTON: The primary difference be- tween the second maxilla and the limbs behind is that it has one less endite. HESSLER: The _ skeleto-musculature systems of the Leptostraca and the Ce- phalocarida have much in common. Both have a segmentally repeated, ventral inter- segmental tendinous bar in the thorax, and in the cephalon a ventral endoskeleton formed from the fusion of such bars. In both groups the longitudinal trunk muscles run straight from segment to segment, a pattern quite unlike that in the caridoid facies. The Leptostraca, Cephalocarida, Branchiopoda, Mystacocarida, and Cir- ripedia, have dorsoventral thoracic trunk muscles. The only other Malacostraca clearly possessing such muscles are the Stomatopoda. These muscles are so similar in leptostracans and cephalocarids as to suggest specific homologies. Because of data such as these, the position of the Malacostraca does not seem to be so isolated as has been thought. DAHL: The Malacostraca and _ the Maxillopoda resemble each other in vari- ous respects. Brooks mentioned that he could think of deriving the Malacostraca from some copepod-like crustacean. I don’t agree; I think that such an ancestor must lie well below the copepod level of organization. But I agree that the mala- costracan and maxillopodan groups of evolutionary lines appear to be closer to 177 each other than either of them is to the branchiopod lines. From what we have heard it appears probable that the Ce- phalocarida play a central part in the discussion concerning a possible common ancestor. SANDERS: There is a danger of rep- resenting the present-day cephalocarid as the primitive condition. Dahl has pointed out one feature which certainly is not primitive. This is the nature of the brain, which he has related to the fact that the animal is blind. DAHL: In preliminary observations on the brain of Hutchinsoniella we have found what appears to be a complete change- over from visual to chemical orientation. Some structures are unique. There is a pair of, comparatively speaking, enormous lobes lying in front of the labrum which we can only interpret as the lobi olfactorit. In any case, they are connected with the antennules. There are also some very large and dense masses of neuropileum in the anterior part of the protocerebrum located in an area more or less corresponding to the one where you find the olfactory asso- ciation centers of the corpora pedunculata in the Malacostraca. These two observa- tions tend to give independent support to the suggestion that the olfactory sense is the principal one in Hutchinsoniella. This in its turn will certainly have to be inter- 178 preted as an adaptation to the peculiar habitat. GORDON: With regard to the table of larval and adult characters, I would agree with you if you cut out caridoid and merely talk about larva and adult. TASCH: On Figure 75 of the proto- branchiopod, what is the significance of the arrows, because there is one from +Lepidocaris through Sida, a cladoceran, to Estheria? SANDERS: I don’t believe Cannon was attempting to present a rigid phy- logeny. He was merely arranging the limb series from a functional point of view. SIEWING: In 1960 I expressed an opinion on the interrelationships in the Crustacea (Fig. 79). The Cephalocarida are shown near the base of the Superorder Anostraca (comprising Cephalocarida, yLipostraca and Euanostraca), but also Subclass MAXILLOPODA Subclass MALACOSTRACA Superorder ? TPSEU DOCRUSTACEA Fic. 79. ifLiposttace ANOSTRACA MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 not far from the root of Malacostraca and Maxillopoda. What evidence is there against this view? SANDERS: First, the head appendages of the Cephalocarida and Branchiopoda are almost diametrically different. Second, the purpose of showing the branchiopod and non-branchiopod nauplius was to demonstrate that the cephalocarids cannot be included within the Branchiopoda, since their larval morphology is entirely non-branchiopod. Third, limbs and _seg- ments are added in large blocks early in the naupliar series in the branchiopods, while in the cephalocarids there is a con- tinuous, gradual addition of limbs and segments throughout the long series of naupliar stages. SIEWING: Yes, but you have proved that the feeding mechanism is so similar to that of Euanostraca, and Hessler has Euanostraca Subclass GNATHOSTRACA Cephalocarida ~———— Phyllopoda Relationships within Crustacea, by Rolf Siewing. DISCUSSION shown that they have turgor extremities too, like Euanostraca. Furthermore, there are other similarities between Cephalo- carida, Euanostraca and _ j7Lipostraca. There is the eggsac connected with the position of gonads in the abdomen, and the external openings of the gonads. And there are, as you have demonstrated in former papers, similarities in the body shape between the forms in question. These similarities seem to speak for a nearer relationship. SANDERS: A malacostracan, mystaco- carid or copepodan adherent could simi- larly present good morphological reasons for including the Cephalocarida in his group. But I tried to point out that there are good reasons for keeping the cephalo- carids separate. The fact that the Cephalo- carida share morphological features with most of the other subclasses and, at the same time, have features that are distinc- tively their own is the most exciting part of the story. HESSLER: With reference to the feed- ing type, there is just as good reason for seeing close similarity to the Leptostraca as to the Branchiopoda. SANDERS: Hessler’s studies of the endoskeleton indicate a non-branchiopod condition in the Cephalocarida. Branchi- opoda have a laminar type of endoskeleton that is quite different from any other group. MANTON (subsequent written com- ment): Since I have described the endo- skeleton of a leptostracan and have looked at the endoskeleton of several Branchiop- oda, may I say that I do not think that 179 there is an important fundamental differ- ence between the two. GLAESSNER: Are the Notostraca so far from the Cephalocarida? SANDERS: That is a matter of degree. I’d rather not consider that the cephalo- carids are any closer to the branchiopods than to any other group. SIEWING: Near the phylogenetic an- cestor of the Anostraca perhaps? SANDERS: Not particularly; it is gen- eralized enough, but it is not an im- mediate branchiopod ancestor. GLAESSNER: These are not the ances- tors but they are the present-day rep- resentatives of the ancestors, and some of our difficulties in assigning them a specific place arise from the fact that something has happened to all of them, particularly the Malacostraca. SANDERS: We _haven’t dimension of time. DAHL: Yes, we have to project them all into Recent time. When scanty mate- rial of the cephalocarid was first known, I took up exactly the same position as Siewing is taking now, that the Cephalo- carida would have to be connected to the branchiopod line. That was why I created a new group, the Gnathostraca, to receive both. But the material presented has convinced me that there is no such close connection and that the Gnathostraca can- not be regarded as a natural unit. It may be that the present-day Cephalocarida in certain respects resemble the branchiopods, perhaps even more than they resemble other things, but we have to accept the fact that there are also fundamental dif- ferences, and to place them accordingly. the third PHYLOGENY AND EVOLUTION OF CRUSTACEA Museum or CoMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, 1963 XV Discussion of the Peracarida Problem In a discussion arising from earlier ses- sions, Manton drew attention to the unity of the Peracarida, as drawn up by Cal- man, and used by Gordon, Siewing, etc. and to the parallel loss of the carapace in the Isopoda and Amphipoda, resulting in convergent similarity with the Syncarida. Basic peracaridan features—the Jacinia mobilis, and brood pouch—are features not easily seen in fossil animals. BROOKS: The seminal receptacle has a polyphyletic origin in the Eumalacos- traca. Thus, the presence of a seminal receptacle in the Syncarida cannot be used to exclude them from consideration as possible ancestors of the Peracarida. The Paleozoic Eumalacostraca show divergence toward the different Recent taxa. GLAESSNER: I have one or two ques- tions here. Now, if we take the Peracarida as they are, how far back should those characters go? Do we always have to have Peracarida if we go back in the geological record, or do we lose all those groups, the Amphipoda, Isopoda, Tanaidacea, Cuma- cea and Mysidacea when these characters, that you mentioned as characters of the Peracarida, were not yet in existence? At what stage do these characters necessarily belong to all these groups? Or are there other characters which may have been present in ancestral forms of these various groups to enable us to trace them back? I think the difficulty of the problem that 181 is placed before the paleontologists here is that these characters are so specific and so unusual that without them we cannot even talk of ancestors of those groups. How would you trace back those groups which are now peracaridan? MANTON: Back to primitive mysid- like animals, which had a carapace and progressively lost it. Where you have mysid records of an early nature one would put back the origin of the rest of the peracaridan groups to an earlier stage than that, but from the mysid line, and not from the syncarid line. GLAESSNER: Now the difficulty here is that we can trace the mysids back and we end up with lophogastrids with a very strongly developed character. We arrive at an ancestral group which we originally called +Pygocephalomorpha, and now they are called +Eocarida, and which contain the ancestors of the decapods. You have to assume that the true peracaridan char- acters evolved at a certain time which is presumably also the Palaeozoic. I did not connect them all that closely with the Syncarida. BROOKS: I do not believe in a close relationship between Syncarida and the Amphipoda and Isopoda. An alternative interpretation is that the syncarids origi- nated independently of the Eumalacos- traca with a carapace. Malacostraca with a carapace tend to have the thoracic 182 somites reduced in size. Another interest- ing point is the forward imbrication of the anterior thoracic somites. How did the carapace originate if its point of origin is covered by an overlap from the first thoracic tergite? GLAESSNER: The second pleura of the abdominal tergite in the Caridea has a very strong forward and lateral overlap and I don’t think anyone has ever been able to find out why. MANTON (subsequent written com- ment): The forward overlap of the first free thoracic tergite in the Syncarida or of the second abdominal tergite of the Caridea have a clear functional explana- tion: the facilitation of flexure. The former example is necessary because of the taper- ing form of the body, the larger tergite more easily overlapping the smaller, and the caridean condition permits maximum Superorder PERACARIDA Amphipoda loss of 4 Isopoda ‘ex carapace / N Tanaidacea Cumacea ? acquisition of marsupium MUSEUM OF COMPARATIVE ZOOLOGY, SPECIAL PUBLICATION, Mysidacea 1963 freedom of flexure at the point where it is most needed. Within the Chilopoda there are many examples of forward over- lap by tergites, but the same tergite does not always imbricate in the same direc- tion, depending always on body shape. When the body does not taper, the tergites overlap posteriorly. GORDON: Calman thought that the syncarids never had a carapace, and that the isopods and amphipods had lost it secondarily. SIEWING: The different reconstruc- tions of malacostracan phylogeny made by zoological and paleontological morpholog- ists derive from their different methods. The paleontologist proceeds in the reverse direction from that of the zoologist. He presumes that one can read the phylogeny directly from the geological succession. Thus, the zoologist reconstructs the phy- Superorder EUCARIDA Superorder PANCARIDA S uperorder SYNCARIDA tT a -, loss of ca _7 Carapace Fic. 80. Relations of Peracarida, by Rolf Siewing. PERACARIDA PROBLEM logeny on the basis of the natural system, the paleontologist the natural system from the phylogeny. The effect is that the paleontologist gets more parallel lines of evolution than the zoologist. Another cause for this may be the lack of material and the overestimation of superficial char- acters. Paleozoologists and neozoologists must compromise in their methods to get a mutually consistent system. As demonstrated earlier, the Isopoda and Amphipoda are derived from a com- mon ancestor near the Mysidacea. The Isopoda are connected with the Mysidacea by transitional groups, such as the Cuma- cea, Tanaidacea, and perhaps the Spelaeo- griphacea. All these groups are peracari- dan crustaceans and therefore have in common the physiologically complex mar- supium, and a Jacinia mobilis. But the latter is also found in the Thermosbaena- cea and some recently discovered Anaspi- dacea (Stygocaris, Parastygocaris) from South America. I can see no reason for 183 an independent derivation of peracaridan groups from one another. The Peracarida are connected via the Mysidacea with the main stem of the Malacostraca, that leads to the Pancarida and Eucarida (Fig. 80). At the fork the Syncarida and Peracarida are not far apart, but we have no proof of a direct derivation of Peracarida from Syncarida. GLAESSNER: What is the biological explanation for the independent loss of carapace in the two lines shown on Figure 80? SIEWING: The reduction of organs occurs much more frequently than the new building of an organ. Carapace reduction occurs in the Crustacea in many lines (Copepoda, Mystacocarida, Anostraca sensu lato, Syncarida, Amphipoda, Tanai- dacea, Isopoda, Spelaeogriphacea). Ab- sence of carapace alone, therefore, by no means proves a nearer relationship be- tween Syncarida and the Isopoda-Tanai- dacea-Amphipoda. cy hl amar ie Oi a ee ira ste wet 7 a oe af sat pina ; soneeaish Index This is a subject and systematic index—authors are not indexed. Throughout the volume the sign {~ precedes the name of a fossil taxon. Abdomen, 5-6, 8, 109 displacement of organs from, 3-4 and locomotion, 3, 17-18 muscles of, 17-18 number of segments in, 88, 93, 95, 107-108 structure of, 3 Abduction-adduction of muscles, 111-138, 141- 143 Acclimation [Acclimatization] (see also Tem- perature, Salinity, etc.), 28-34 yAcercostraca, 145, 154-156 Adaptation, 3, 21, 25, 27-45, 79-84, 178 antiquity of, 82 freshwater, 37-38, 40-42 functional, 5, 28, 33-34, 45 genetic, 28, 34-45 non-genetic, 28-34, 45 to salinity, 34, 35-43 serial, 28 structural, 28, 33-34, 43, 44-45 terrestrial, 34, 35, 40, 43-45, 59-76 Adaptive characters, viii Adaptive radiation, 21, 84, 156, 162 Aedes, 43 yAgnostida, 19 Amandibulates, 19 Amphipoda, 3, 8, 9, 10, 96-99, 109, 161, 181, 182, 183, Table 1, Figs. 37-39 Anaspidacea, 88, 93, 95-96, 100, 159, 183 Anas pides, 18, 113, 117, 122, 127, 135, 160, Figs. 36, 50, 51 {Anas pidites, 160 Ancestors (see also Common ancestor), vii, 39, 177, 179 crustacean, 17, 20 of Leptostraca, 106 of Peracarida, 181 reconstruction of, 6, 85 Ancestral branchiopod, crustacean, etc. (see Primitive branchiopod, crustacean, etc.) Annelida, 20, 21, 107, 143, 172 Anomura, 51-57, 84 Anostraca, 1, 2, 3, 5, 6, 7, 8, 10, 11, 12, 24, 110, 132, 143, 145, 153-154, 155-156, 166, 171, 172, 179, 183, Table 1, Fig. 1 Antenna, 19, 88, 90, 91, 105, 106, 152, 154, 160, 164, 167, 168, 169 185 Antennal gland, 38, 39, 42, 43, 96 Antennule, 88, 90, 91, 105, 106, 167, 168, 169, 177 +Anthracocaris, 161 Apodemes, 111-138 Appendages (see also Antennules, etc., Trilobitan limb), 4, 5, 19-20, 88-90, 111-139, 141-144, 178 biramous, 21, 135-136, 154, 168, 169 of Branchiopoda, 152-156, 178 of Cephalocarida, 163-174, 178 differentiation of head, 5 evolution of, 21 and feeding, 17-21 of fossil Conchostraca, 152 of Malacostraca, 88-101, 105, 106 and primitiveness, 107, 109 Apterygota, 80, 137-138 Apus (see Triops) Arachnida, 80, 112, 131 +Archaeostraca, 90, 106, 107, 159-160, Fig. 44 Archetype, archetypal evolution, viii Argulus, 11 Armadillidiidae, 44 Armadillidium, 32, 44 Artemia, 29, 30, 32, 35, 38-39, 82, Figs. 3, 4, 8, 71 Arteria subneuralis, 93, 97, 99 Arteries, paired, 88, 93, 101 Arthrodial membrane, 62, 135, 136, Fig. 60 Arthropoda (see also various groups), 2, 23, 83, 141, 143 Arthrostraca, 96 Ascothoracica, 1, 2 Asellus, 31, 41 Astacidae, 40, 41 Astacus, 29, 31, 34, 41, 42, 43, 122, 127, Figs. 9, 57 Atomism, 24 +Aysheaia, 21 Bahamas, distribution of terrestrial crabs, 73-75, Fig. 25 Balanus, 10, 28, 33, Fig. 72 Bathynellacea, 95, 162 Behavior, 33, 39-40, 44, 81 Benthic habit, 3-4, 6, 45, 144 186 Biramous appendages, 21, 135-136, 154, 168, 169 Birgus, 35, 40, 45 Bosmina, 153 Brachyura, 8, 51-57, 59-76, 83 +Bradoria, 160 Brain, 90, 177 Branchial chambers, 40, 43, 59, 62, 74 Branchiopoda, 1, 11, 12, 13, 18, 23, 28, 108, 113, 143, 145-156, 159, 160, 161, 163-166, 167, 168, 170-172, 175, 177, 178, 179, Table 5, Figs. 1, 2, 63, 67, 70, 78 +Branchipodites, 153 Branchipus, 42, 153 +Branchipusites, 153 Branchiura, 1, 2, 8, 11, 12, Table 1, Fig. 1 Brood pouch, 43, 44, 96, 98, 181 Bryozoa, 108 Caeca, caecum, 7-8, 95, 101 Calanus, 9-10 Calanoida, 83 Callianassa, 35 Callinectes, 60, 65, 67, 75, Tables 2-4, Fig. 19 +Canadaspis [;Hymenocaris], 19 Cancer 35, 59, 62, 65, 67, 69, 75, Tables 2-4, Fig. 18 Carapace, 2, 88, 90, 108, 109, 154, 159, 161 absence of, 2, 94, 161, 183 loss of, 181-183 in jNahecaris, 105-106 parallel evolution of, 2, 181 reduced, 96, 98, 161 as size indicator, 65-67 Carcinus, 34, 35, 36, 38, 40, 60, 68, 69, 122, 127, Fig. 57 Cardisoma, 59, 62, 65, 66, 67, 72, 74, 75, 76, Tables 2-4, Figs. 16, 25 Caridea, 182 Caridoid facies, 3-4, 17, 105, 161, 162, 168-170, 177, 178 Cephalization, 5, 136 Cephalocarida,, 1;- 2;°33>5;26,5'7,--83 10-1125513; 18, 20, 93, 108, 141, 143, 156, 163-175, 177- 179, Figs. 1, 2, 70 7~Ceratiocaris, 159 Cercopods, 160 Characters, vii, viii, 88-102, 107-108, 138, 147- 154, 178, 181-183 adaptive, viii external morphological, 2-6 internal morphological, 7-12 and habits, 21, 80-81 INDEX loss of, 107, 108, 109, 156, 181-183 of Malacostraca, 85-102 phylogenetic, viii primitive, 6-7 Chelicerata (see also various groups), 112, 131- 134, 138 Chilopoda, 6, 80, 138 Chirocephalus, 42, 113, 117, 122, 125, 135, 143- 144, 154, Figs. 45-49, 75 Chironomus, 43 Chydoridae, 152 Cilia, 24, 141 Circulatory system (see also Arteries, Heart, etc.), 8, 11, 101, Figs. 33, 40 Cirripedia, 1, 2, 8, 11, 12, 13, 28, 108, 164, 177, Figs. 1, 70 Cladocera, 33, 82, 145, 152-153, 156, 171, 172, Table 1 Classification, 1-2, 12-15, 85, 88, 138, 139 of Anomura, 51-57, 84 of Branchiopoda, 145, 147, Table 5 of Eumalacostraca, Fig. 68 single character, 83 Coelenterata, 143 Coenobita, 40, 45 Cohorts, 13, Fig. 2 Collembola, 137-138 Common ancestor, vii, 24, 177 Communities, physically and controlled, 81-83 Compensation, 32 Compound eyes, 10-11, 12, 22-25, 93, 108, 141, Table 1 Concavity of limbs, 164 Conchostraca, 7, 145-152, 153, 155, 171, 172, Table 1, Figs. 64-66 Connective tissue, 67 Convergence, convergent evolution, 2, 7, 25, 141, 144, 181 of brain, 9 of habits, 144 of lacinia mobilis, 96 of mandibles, 112, 136, Fig. 61 of petasma, 93 of statocysts, 95-96 Copepoda, 1, 2, 3, 4, 5, 8, 9, 12, 13, 20, 28, 33, 83, 110, 164, 174, 179, 183, Figs. 1, 70 Copepodoidea, 2 +Cornia [+Pemphicyclus], Fig. 65 Coxal plates, 96 Crangon, 35, Fig. 8 non-physically 156, 159, 149M 150.151 152; INDEX Craterostigmus, 83 Crustacea (see also various groups), 143 Crustacea/Insecta relationship (see Insecta) +Cruziana, 144 Ctenolepisma, 137 Cumacea, 98, 99, 109, 162, 181, 183, Fig. 39 +Curvacornutus, 150, Fig. 65 Cuticle, 41, 43, 60, 67, 68, 73, 111, 135, 143, 162 Cyclestheria, 153 Cyclestheriidae, 151, 153 Cyclodorippae (see Tymolae) Cyclodorippe (see Tymolus) Cyclomorphosis, 33 Cyclops, 29 Cymonomae, 53 Cymonominae, 57 Cymonomus, 52-53, Figs. 10, 11 Cypris-stage, 28, 164 Cyzicidae, 151 Cyzicus, 147, 152 Daphnia, 25, 29, 33-34, 41, 113, 152, 153, Figs. 9, 75 Decapoda, 9, 22, 24, 28, 51-57, 83-84, 95, 108, 109, 125, 181, Table 1, Fig. 40 Decapoda peditremen, 55-57 Decapoda sternitremen, 55-57 Derocheilocaris, Fig. 72 Detritus (see also Feeding), 160, 171 Development, 75, 108, 151, 164-167, 168-170, Figs. 70, 74 +Devonocaris, 162 Diaphanosoma, Fig. 75 Diaptomus, 29 Digestive tract (see also Intestine, Mouth, etc.), Ps Diplopoda, 80, 83, 138 Diplura, 137, 138 Displacement of organs, 3-4, 9 Distribution, 41, 73-75, 76, Fig. 25 Divergence, divergent evolution, vii, 21, 80, 98, 139, 161, 181 Dorippidae, 83, 84 Dorippidae peditremen, 55 Dorippidae sternitremen, 55 Dorippinae, 51, 55 +Douglasocaris, 160 Dromia, 51 Dromiacea, 51-57, 84 Dromiidae, 51 137 Ecdysis, 69-70, 159, 162, 164, 166 of Conchostraca, 159 and pericardial sacs, 69-70, 71, Fig. 24 tEchinestheria, 151, Fig. 65 Ecology, 27, 28, 34, 152 Eggs, 38, 42, 44, 152, 153, 168 Eggsac, 179 Elasticity of cuticle, 11, 122 Embryology, 21, 108, 109-110, 112-113 of Amphipoda-Isopoda, 98, Fig. 38 of Notostraca, 23 of Stomatopoda, 106 Emerita, 35 Endites, 18, 21, 95, 100, 112, 132, 134, 164, 170, 171, 174, 177 Endophragmal skeleton, 54, 55, 56, 57, 84, Figs. 13, 14 Endoskeleton of head, 111, 113, 117, 122, 127, 138, 177, 179, Figs. 52, 53, 55 Entognatha, 139 Entognathy, 136, 137, 138 Entomostraca, 107, 109 Environment (see also Oxygen, Temperature, etc.), 45, 80, 81-82, 83 acclimation and, 28, 34 of basic crustacean evolution, 162 and divergence, 21 fitness of, 24 Enzymes, 24 {Eocaris, 162 Ephippia, 82, 152, 153 {Eridoconcha, 146 {Eridostraca, 146 Eriocheir, 33, 34, 38, 40, 42, 43, Fig. 9 Eriphia, 34 Eryonidea, 57 +“Estheria”, 149, 150, 178, Fig. 75 Euanalogy, 93, 96 Euanostraca, 178-179 Eucarida, 93, 96, 99, 100, 144, 162, 183, Fig. 80 Eumalacostraca, 106, 159, 161, 168, 181, Figs. 68, 78 Euphausia, 144 Euphausiacea, 162, Table 1 Euryhaline (see Salinity) Evolution (see Adaptation, Phylogeny, etc.) Evolutionary lines, 13-15, Fig. 1 Excretion, 39-41 Exoskeleton, 20, 143 Eyes (see Compound eyes) Eyestalk, 9, 70, 96 188 Feeding, 5, 6, 17, 20-21, 136, 143-144, 160, 164, 170, 171, 172, 178, 179 -current, 171 detritus, 18-19, 21 effect on head topography, 2, 9 filter-, 18, 20, 21-22, 143, 144, 160 fine-food, 20, 122 fluid, 131 habits, 45 large-food, 122, 144 particle-, 4, 8, 22, 112 of Trilobita, 143, 144 +Fordilla, 146, 160 Freshwater (see Adaptation) Frontal organs, 9-10 Function and structure, 80, 177-178, 182 Functional adaptation (see Adaptation) Furca, 93, 95, 100, 152, 160, 162 Furcal rudiments, 162 Fusion, of scutes, 132 of segments, 20, 95, 96, 106, 174 +Gabonestheria, 149, 151, Fig. 65 Gammarus, 32, 33, 34, 35, 36, 37, 38, 40, 41, 42, 43, Figs. 5, 8,9 +Gampsonychidae, 106 +Gampsonychidea, 95, 96 Gecarcinus, 39, 44, 59, 62, 63, 65, 66, 67, 69, 71, 72, 74, 75, 76, 77, Tables 2-4, Figs. 15, 20-25 Gene modification, 148, Fig. 66 Genetic adaptation (see Adaptation) Genetic potentiality, 25, 109 Genitalia, 51-57, 83, Figs. 10-14 Genotype, 24, 29 Gills, 59, 60, 62, 71, 73, 74, 76 and osmoregulation, 35-40, 42 on pleopods, 44, 93, 96, 98 structural adaptations, 43, 44, 45 on thoracopods, 96 ~Gilsonicaris, 153, 154 Glands (see Antennal gland, Maxillary gland, etc.) Gnathobasic mandible, 112, 113-135, 136, 138 Gnathostraca, 1-2, 12, 179, Fig. 42 Gonads, gonoduct, gonopore (see Reproductive system ) Grades, 139, 163 Grooves, 132 Gymnopleura, 56, 57 Habits, 83, 108-109, 138 INDEX and divergence, 21, 80-81 and macroevolution, 45 Hansenomysis, 96 Har pacticus, 9 Head, 2-3, 5, 12, 44, 109, 111, 143 Heart, 3, 4, 12, 88, 93, 95, 96-97, 98, 101, 108, Fig. 5 “Heavy” mandibles, 143-144 Heloecius, 31, 35 Hemigrapsus, 34, 35 Hemimysis, 29, 108, 113, 117, 122-127, 135, 136, Fig. 48 Hemoglobin, 24 Hermit crab, 80 Hexapoda (see Insecta) Histology, of pericardial sacs, 67-69, 75-76 Holeuryhaline (see Salinity) Homarus, 30, 43, Figs. 6, 7 Homolidae (see Thelxiopidae) Homolodromiidae, 57 Homology, vii, 5, 9, 10, 93, 96, 98, 99, 100, 101, 108, 111, 136, 143, 163, 164, 177 Homology investigation-method, 85, Fig. 39 Hoplocarida, 93 Hutchinsoniella, vii, 141-142, 164, 172, 177, Figs. 69, 73, 76 Hyas, 35 +Hymenocaris (see {Canadas pis) Hyperia, 8 jIgorvarentsovia, 150, Fig. 66 Insecta, 3, 21, 22, 37, 111, 112, 113, 132, 136-139 Insecta/Crustacea relationship, 3, 22 Intestinal filling, 160 Intestine (see also Midgut, Proctodaeum, etc.), 3, 7-8, 42, 43, 108, 152, 160 {I psilonia, Fig. 65 Isopoda, 3, 8, 9, 10-11, 43, 44, 96-100, 109, 123, 137, 161, 181, 182, 183, Table 1, Figs. 37-36 ;+Japanoleaia, 148, Fig. 64 Jaws (see Mandible) Juvenile stage, 11, 168-169 *+Kazacharthra, 145, 154-155 +Keratestheria, 151, Fig. 65 Koonunga, 18 Labiata, 139 Labrum, 17, 18, 22, 159 Lacinia mobilis, 95, 100, 181 INDEX Larval stages (see also Juvenile, Nauplius, etc.), vii, 75, 108, 109, 159, 168-169, 178 +Leaia, 147, 150, Figs. 64, 66 +Leaiidae, 147, Figs. 64, 66 Leander, 9 +Lepidittidae, 145-146 {Lepidocaris, 18, 171, 178, Figs. 71, 75 Lepidurus, 154 Leptestheriidae, 151 Leptodora, 153 Leptograpsus, 35 Leptostraca, 90-91, 106, 107, 159, 168, 172, 177, 179, Table 1, Figs. 29, 78 Libinia, 60 Ligia, 32, 122, 125, 131, 132, Figs. 49, 54-56 Ligiidae, 44 Limbs (see Appendages) Limnadia, Fig. 71 Limnadiidae, 152 {+ Limnadio psileaia, 150, 151, Fig. 65 +Limnadopsis, 151, Fig. 65 {Limnestheria, 152 Limulus, 21, 112, 113, 131-135, 138, 144 yLipostraca, 145, 154, 156 Locomotion, 4, 6, 35 and abdominal structure, 3, 108-109 and macroevolution, 45 Locusta, 131 Lophogastrida, 99, 181, Fig. 40 Lophopanopeus, 35 Loss of characters, 107, 108, 109, 156, 181-183 Lynceidae, 152, 159 Lynceus, 159 Macroevolution, 45 Macrura, 83 Maja | Maia] 35, 54, 55, 60, 68, 69, Fig. 14 Malacostraca (see also various groups), 1, 3, 4, 6, 7, 8, 9, 10, 11, 12, 13, 80, 85-102, 106, 107, 108, 109, 110, 122, 132, 159, 160, 161, 162, 177, 178, 179, 181, 182, Figs. 1, 2, 26, 27, 28, 42, 78 Manca-stage, 98 Mandible, 19-20, 21, 23, 111-139, 141-144, 152, 159, 164, 168, 169, Figs. 45-57, 61, 62 gnathobasic, 112, 113-135, 136, 138 “heavy”, 143-144 whole-limb, 112-113, 136, 138 Mandibular palp, 101, 108, 154, 169 Mandibulata, 19, 139 Marine fossil Eumalacostraca, 162 +Marrella, 19 i89 Marsupium (see Brood pouch) +Massagetes, 148, Fig. 64 Maxilla, 163, 164, 171, 174, 177, Fig. 77 Maxillary gland, 42, 96, 164 Maxilliped, 100 Maxillopoda, 1, 2, 3, 4, 8, 10, 11, 12, 13, 23, 108, 177, 178, Figs. 1, 2, 42 Meganyctiphanes, 18 Membranipora, 108 Merostomata, 113, 132, 134-135, 138 Metabolic rate, 28, 29, 30, 32, 34, 35, 60 Metachronal limb movements, 171, Fig. 76 Metamerism, 5 Metanauplius (see also Nauplius), 12 Metapenaeus, 34, 35 Midgut, 7, 8, 101, 172 gland, 39, 59 Mitochondria, 24 Mollusca, 23, 143 Molting (see Ecdysis) {Monoleaia, 147, Fig. 64 Monophyly, 88, 161 {Montecaris, 159, 162 Morphological series, vii Morphology, 85-102 exactness of method, 99 importance of, 7 internal and external, 2-12 Mosaic evolution, viii, 19 Mouth, 6, 17, 18, 19, 20, 22 Mouthparts (see also Mandible), 3, 12 displacement of, 9 evolution of, 4-5 of +Kazacharthra, 154 Muscles, 3, 4, 17-18, 84, 91, 111-139, 141, 143, 164, 177, Figs. 46-48, 51-53, 55-60, 62 Myriapoda, 21, 22, 81, 112, 113, 136, 138, 139 Mysidacea, 8, 95, 96, 99, 100, 106, 162, 181, 183, Table 1, Fig. 39 Mysis-stage, 169 Mystacocarida, 1, 2, 3, 4, 5, 8, 12, 13, 20, 108, 164-165, 174, 177, 179, 183, Figs. 1, 70 {Nahecaris, 90, 105-106, Figs. 31, 43 Natantia, 9, 55, 57, 83 Natural selection, 23-24, 27, 143 Natural system, 2, 13, 183 Nauplius, 28, 102, 108, 153, 159, 164-168, 178, Figs. 70-74 Nauplius eye, 9, 11, 12, 23, 101 Nebalia, 8, 106, 108, 162, Fig. 30 Nebaliacea, 106, 144 190 Neomysis, 29 Nephridia, 88, 101 Nephridial canals, 43 Nervous system (see also Brain), 3, 8-11, 172 Neuroendocrine system, 69-70 New steady state, 29, 30, 31-32, 34 Niches, 21, 81-82, 83 Non-genetic adaptation (see Adaptation) Non-physically controlled community, 82 Notopoides, 53-54, Fig. 13 Notostraca, 5, 7, 23, 93, 108, 145, 154, 155, 156, 159, 171, 172, 179, Table 1 Ocy pode, 34, 35, 39, 44, 45, 59, 60, 62, 65, 66, 67, 72, 74, 75, 76, Tables 2-4, Figs. 17, 21, 25 Oesophagus, 122 Olenellid trilobites, 19 Olfactory sense, 177 Oligohaline (see Salinity) Oniscidae, 44 Oniscus, 32, 44 Ontogeny, 7, 108, 112, 147 Onychophora, 21, 80-81, 112, 136 Organs (see also Frontal organs, etc.) , 9-12 displacement of, 3-4, 9 “Original” structure, 85, 107 Ornamentation, of Conchostraca, 147 Osmoconcentration, Figs. 8, 9 Osmosis, 35-42, 60 Ostia, 93, 101, 106 Ostracoda, 1, 2, 7, 8, 11, 12, 13, 28, 108, 110, 174, Table 1, Figs. 1, 2 Overlap of pleura, 182 Overshoot response, 29-30 Oxygen, consumption and acclimation, 29, Figs. 3, 4 Oxystomata, 56, 83 Pachygrapsus, 33, 35, 39, 40 Pagurus, 34, 35 Palaemon, 35, 39 Palaemonetes, 34, 35, 37, 39, 40, 41 {~Palaeocaris, 160 +Palaeolimnadiopsis, 151, Fig. 65 Paleontology, 82, 84, 85 adaptation and, 80 method of, cf. zoology, 182-183 need for, 4 ;+Palaeopalaemon, 162 Palinura, 57 Palinurus, 35 INDEX Pancarida, 96, 99, 100, 183, Figs. 41, 80 Pandalus, 9 Panopeus, 60 Panulirus, 67, 69 +Paraleaia, 147, Fig. 64 Parallel evolution, vii, 25, 83, 183 of carapace, 2, 181 of compound eyes, 22-23 of mandibles, 135, 139 Paranaspides, 18, 113, 117, 122, 125, 127, 160, Figs. 49, 52, 53 Parastygocaris, 183 Particle-feeding (see Feeding) Pauropoda, 138 Pelagic habit, 3-4, 6, 45, 144, 160, 172 +Pemphicyclus (see }{Cornia) Penaeidea, 109, 168, 169, 174 Penaeus, 35, 168, Figs. 72, 74 Peracarida, 3, 4, 96-100, Figs. 39, 80 Pericardial sacs, 44, 59-77, Figs. 15-24 function, 69-73, 76 gross anatomy, 60-67 histology, 67-69, 75-76 surface area, 62, 63-67, 75, Tables 2-4 +Perimecturus, 106 Peripatus, 80, 136 Petasma, 93, 95, Fig. 34 Petrobius, 18, 136, 137, 138, Fig. 48 Phoronid larvae, 108 Phyllocarida, 93, 106, 159, 160, 161-162 Phyllopoda, 1, 5, 6, 7, 8, 10, 11, 12, 24, 108, 110, Fig. 1 Phylogenetic series, vii, 143 Phylogeny, vii-viii, 24-25, 27, 85, 107, 112, 141, 182-183 of Arthropoda, 11, 134-135, Fig. 44 of Branchiopoda, Fig. 67 of Crustacea, 6, 7, 9, 13-15, 177-178, Fig. 79 of Eumalacostraca, 161, Fig. 68, 80 of Malacostraca, 4, 85-102, 161, 177, 178, Figs. 68, 80 reconstruction of, 85 Physically controlled community, 82 Pleopods, 18, 52-53, 54, 56 respiratory, 44, 93, 96, 98 +Plesiosauria, 80 Pleura, overlap of, 182 Polyartemiidae, 153 Polychaeta, 88, 108 Polyphyly, 139, 161, 181 Polystenohaline (see Salinity) Porcellio, 32 INDEX Porcellionidae, 44 Porifera, 143 Post-Mysis stage, 169 Potamobius (see Astacus) Potamon, 31, 34, 37, 40, 42, 43, Fig. 9 Potentiality, basic genetic, 24-25 +Praeleaia, 147 Preadaptation, 83 Precoxal segment, 132 Pre-epipodites, 135 Primitive (see also Ancestors), 2, 17, 85, 107-108 anomomerism, 107 branchiopod, 156, 171, 179, Fig. 75 characters, 6-7 crustacean, 102, 107-108, 143, 163, Fig. 78 adaptation and, 45 ?benthic or pelagic, 3 and Cephalocarida, 6, 12, 13-14, 156, 177 characters of, 6, 162 development of, 164 limb(s) of, 4-5, 20-21, 109 organization, 3, 6 and tagmosis, 6 eumalacostracan, 161 feeding mechanism, 17, 144 habit, 144 jaw mechanism, 111 Malacostraca, 144, 168 malacostracan, 3, 6, 85, 100-102, 105, 144, 168-170, 177, Fig. 28 tagmosis, 164 Primitiveness, 6-7, 17 and appendages, 107, 108, 109 of Cephalocarida, 177 of Malacostraca, 107-109 secondary, 85 and segmentation, 107 of Stomatopoda, 108-109 of turgor organization, 143 Procambarus, 41 Procephalon, 90-91, 93, 108 Proctodaeum, 8 Promotor-remotor movements of coxae, 112-136, 143, Figs. 45, 46, 50 Protobranchiopod (see Primitive branchiopod) +Protomonocarina, 150, 151, Figs. 65, 66 Protozoea, 108, 169, 174 +Pseudocrustacea, 6, 19, Fig. 42 Pseudotracheae, 43, 44 +Pteroleaia, 147, Fig. 64 Pterygota, 137-138 +Pygocephalomorpha, 181 107, 191 +Quadriasmussia, 152 Ranina, 53, 54, 56, Fig. 12 Raninidae, 51, 53-54, 56, 57, 83 Raptorial appendages, of Syncarida, 160 Recapitulation, vii Reduction, of organs and structures, 17, 93, 95, 98, 106, 109, 154, 156, 161, 174, 175, 183 Relative growth, 34 Reproductive system (see also Genitalia), 17, 179 of ancestral malacostracan, 101 of Copepoda, 4 displacement of, 3-4 of fossil Conchostraca, 152 position of gonopore, 3-4, 5 Reptantia, 55 Reserve cells, 69 Respiration, 4, 20, 34, 43, 44-45, 59, 73, 93, 139 Respiratory organs (see also Gills, etc.), 96, 98 Responses, 28, 29-30, Figs. 3, 4 undershoot and overshoot, 29-30 Reversal, evolutionary, 138 +Rhinocarididae, 105 *Rhinocarina, 106 Rhithropanopeus, 35, 36, Fig. 8 Rhodopsin, 23 Ribs, of Conchostraca, 150-151, Fig. 64 {Rochdalia, 153 Rostral plate, rostrum, 105, 162 Salinity, 30-42, 82, Figs. 8, 9 Scorpionidea, 80, 106 Scutes, 21, 132, 138 Secondary derivation, 85, 108, 144, 169 Secondary modification, 144 Segmentation, 5, 6, 20, 88, 164, 166, 174, Fig. 70 of abdomen (see Abdomen) of +Archaeostraca, 159 of Branchiopoda, 153-154 inherited from annelid, 20 of Malacostraca, 88 of mandibles, 138 and primitiveness, 107 variation in, 110 Selection pressure, vii, 24, 27 Selective preservation of fossils, 162 Sella turcica, 55 Seminal receptacle, 181 Sense organs (see also Compound eyes, etc.), 9-12 Serial adaptation, 28 Serial homology, 163-164 192 Setae, 17, 18-19, 22, 62, 71-72, 75, 76, 164, 167, 168, 169, 171, 174, Figs. 20, 21 Shape, changes in Conchostraca, 147 Sida, 178, Fig. 75 Simocephalus, 29 Single character classification, 83 Size, 65-67, 147 Speciation, 82 Species, abundance, 81, 83 Spelaeogriphacea, 183 Speocarcinus, 35 Spermathecae, 51-57 Spermatophores, 51 Sphaeroma, 44 +Spriggina, 20 Squilla, Figs. 33, 34 Stabilization phase, 30-31 Statocyst, 9, 95-96 Steady state, 29-33, 34 Sternal furrows, 51 Stomach, 88, 95, 98, 100, 101 Stomatopoda, 3, 4, 20, 90-95, 100, 106, 108-109, 161, 177, Table 1, Figs. 32, 33-35 Stomodaeum, 7, 8 Streptocephalus, 32 Stress, 82 Structural adaptation (see Adaptation) Structure, and function, 80 Stygocaris, 183 Sutures, 132 Symphyla, 80, 138 Syncarida, 95-96, 117, 118, 122, 144, 159, Table 1 Tachypleus, Figs. 58-60 Tagmosis, 5-6, 107, 164 Tanaidacea, 98, 109, 161, 162, 181, 183, Fig. 39 Telopodite, 112, 113 Telson, 95, 100, 106, 107, 109, 154 Temora, Fig. 72 Temperature, acclimation, Figs. 5-7 Tendons, 111-138 Terrestrial adaptations (see Adaptation) Terrestrial crabs, 59-76 Thelxiopidae [Homolidae] 51 29, 30, 32-34, 35, INDEX Thermosbaenacea (see Pancarida) Thoracica, 11, 13, Table 1 Thorax, 5, 6, 51-52 Thysanura, 3, 138 Tigriopus, 32 Time scale, for branchiopod evolution, 156 Tolypeutes, 44 Tomopteris, 20 Trends, vii, 27, 111, 137, 153-154 Trichoniscidae, 44 Trilobation, 135 7Trilobita, 19, 21, 132-135, 136, 138, 143, Fig. 44 Trilobitan limb, 19-20, 21, 109, 132, 135, 143, 144 +Trilobitomorpha, 19 Triops [Apus], 10, 18, 21, 41, 144, 154, Figs. 71, 75 Turgor, 143, 179 Tymolae [Cyclodorippae], 53 Tymolidae, 56-57 Tymolinae, 51-53, 55-57 Tymolus |Cyclodorippe], 51-52, Figs. 10, 11 Uca, 33, 34, 35, 39, 44, Fig. 8 Undershoot response, 29-30 Unique solutions, 22-24, 141 Upogebia, 35 Urcrustacean (see also Primitive crustacean), 163 Uropods, 44, 93, 95, 96, 97, 98, 106, 169 Urspriinglich, 85 Variation, 27, 147 Varuna, 38 Vertebrates, 23, 109 +Vertexia, 150, 151, Fig. 65 Vision (see Compound eyes) +Waptia, 109 Water uptake, 70-76 Weight of Brachyura, 62, 65, 66, 67 Whole-limb mandible, 112-113, 136, 138 Xeinostoma, 51 Xiphosura, 131, 143, 144, 162 X-organ, 9-10, 12 Harvard MCZ Library Date Due Pr E-a 6) S68 slice. nyo : Ais res mae a whi me SS ee =