L/^10 HARVARD UNIVERSITY Library of the Museum of Comparative Zoology BULLETIN OF THE MUSEUM OF COMPAEATIVE ZOOLOGY AT HARVARD COLLEGE, IN CAMBRIDGE VOL. 109 CAMBRIDGE, MASS., U. S. A. 1953 The Cosmos Press. Inc. Cambridge. Mass., U.S.A. ^ U' CONTENTS PAGE No. 1. — Notes on Siphonophores. 2. A Revision of the Abylinae. By Mary Sears. May, 1953. 1 No. 2. — Report on the McCabe Collection of British Columbian Birds. By J. C. Dickinson Jr. May, 1953. 121 No. 3. — New and Little Known Sharks from the Atlantic AND FROM the Gulf OF Mexico. By Henry B. Bia;elow, William C. Schroeder and Stewart Springer. July, 1953 211 No. 4. — The Birds OF Japan, THEIR Status AND Distribution. By Oliver L. Austin, Jr. and Nagahisa Kuroda. (plate.) October, 1953 .277 Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vol. 109, No. 1 NOTES ON SIPHONOPHORES 2. A REVISION OF THE ABYLINAE By Mary Sears Woods Hole Oceanographic Institution CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM May, 1953 Publications Issued by or in Connection WITH THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE Bulletin (octavo) 1863 — The current volume is \'o!. 109. Breviora (octavo) 1952 — No. 13 is current. Memoirs (quarto) 1864-1938 — Publication was terminated with \'ol. 55. Johnsonia (quarto) 1941 — A publication of the Department of Mollusks. Vol. 2, no. 31 is current. Occasional Papers of the Department of Mollusks (octavo) 1945 — Vol. 1, no. 17 is current. Proceedings of the New England Zoological Club (octavo) 1899- 1948 — Published in connection with the Museum. Publication terminated with Vol. 24. These publications issued at irregular intervals in numbers which may be purchased separately. Prices and lists may be obtained on application to the Director of the Museum of Comparative Zoology, Cambridge 38, Massachusetts. Bulletin of the Museum of Comparative Zoology AT HARVARD COLLEGE Vol. 109, No. 1 NOTES ON SIPHONOPHORES 2. A REVISION OF THE ABYLINAE By Mary Sears Woods Hole Oceanographic Institution CAMBRIDGE, MASS., U. S. A. PRINTED FOR THE MUSEUM May, 1953 No. 1 — Notes on Siphonophores. 2. A Revision of the Ahylinae By Mary Sears^ Woods Hole Oceanographic Institution CONTENTS Page Introduction 4 Subfamily Abylinae 6 Key to the genera of the Abylinae 17 Genus Ahyla Quoy and Gaimard 18 Key to the species of Ahyla 33 Ahyla trigona Quoy and Gaimard 35 Ahyla carina Haeckel 37 Ahyla schmidti n. sp 38 Ahyla haeckeli Lens and Van Riemsdijk 39 Ahyla ingehorgae n. sp 42 Ahyla peruana n. sp 44 Ahyla hicarinaki Moser 45 Ahyla hrownia n. sp 46 Ahyla toUoni n. sp 47 Genus Pseudahyla n. gen 49 Key to the species of Pseudahyla 52 Pseudahyla irregularis n. sp 52 Pseudahyla duhia n. sp 53 Genus Ceratocymba Chun 54 Key to the species of Ceratocymha 62 Ceratocymba sagittata Quoy and Gaimard 63 Ceratocymba leuckarPii Huxley 67 Ceratocymha dentata Bigelow 69 Ceratocymba intermedia n. sp 71 Genus Pseudocymba n. gen 72 Pseudocymba asymmetrica n. sp 75 Pseudocymba anomala n. sp 75 Genus Abylopsis Chun 76 Key to the species of Abylopsis 80 Abylopsis tetragona Otto 80 Abylopsis eschschoUzii Huxley 84 Genus Abylopsoides n. gen 87 Key to the species of Abylopsoides - 89 Abylopsoides dorsalis n. sp 89 1 Papers from the "Dana" Collection No. 31, and Contribution No. 602 of the Woods Hole Oceanographic Institution. 4 bulletin: museum of comparative zoology Abylopsoides ventralis n. sp 90 Abylopsoides basalis n. sp 91 "i Abylopsoides sp 92 Genus Pseudabylopsis n. gen 92 Pseudabylopsis anomala n. sp 93 Genus Bassia L. Agassiz 94 Bassia bassensis Quoy and Gaimard 94 Genus Enneagonum Quoy and Gaimard 98 Enneagonum hyalinum Quoy and Gaimard 98 Bibliography 102 Appendix. List of specimens 107 INTRODUCTION Probably no plankton collection other than that of the "Discovery" has yielded such an abundance of siphonophores for studies of geo- graphical distribution as that of the Carlsberg Foundation's Oceanog- raphical Expedition Round the World, 1928-30, on the "Dana" (Carlsberg Foundation, 1934; 1944). In addition, the "Dana" ma- terial has provided an extensive series of abylids which were repre- sented in earlier collections by only an occasional individual. Thus, it provides one of the first opportunities in recent years to re-examine this subfamily systematically. As a result, a number of misconcep- tions that have appeared in the literature during the last seventy-five years or more^ can now be amended. From our observations of well- preserved colonies, the polygastric generation of most of the known members of this group can now be recognized with little difficulty. In addition, the characteristics of the eudoxids for the long established genera are now well defined. The new information at hand is sufficient to make it worthwhile to record our observations at this time with the hope of establishing a more natural grouping of the Abylinae, similar to that already developed by Totton (1932, pp. 345-346) for the Diphyinae, and to call attention to the points requiring further study when better material becomes available. Several important contributions to our knowledge of the Abylinae are afforded by the "Dana" specimens listed in the appendix to this ' It should be remembered that earlier workers whose observations do not always appear correct, were handicapped in the days before formalin came into general use as a preservative (about 1900). Due to the shrinkage of specimens preserved in alcohol, they were forced to make most of their observations and sketches in the field and could not recheck them later. Dr. Bigelovv informs me, for example, that all of Dr. Mayer's figures were made from life, in the field, for this reason and that he did not have the specimens available when wTiting his papers. sears: notes on siphonophores o report. (1) Four new species of Abyla are reported: A. schmidti and A. tottoni (both previously confused with A. trigona), A. ingeborgae, (with an extra facet similar to A. haeckeli) and A. brownia. In addi- tion, a new species, A. peruana, found in the "Albatross" material in the U. S. National Museum is described. (2) Abyla carina, long con- sidered as a synonym for A. trigona, is reinstated as a good species. (3) The inferior nectophore of Abyla bicarinata Moser is now definitely known. (4) The truly prismatic superior nectophores of A. haeckeli were taken together with cormidia and small inferior nectophores attached within the hydroecium of several superior nectophores. (5) Two species, leuckartii and dentata formerly referred to the genus Abyla, are now transferred to the genus Ceratocymba. (6) The eudoxid of Ceratocymba dentata Bigelow previously identified by Totton (Moore, 1949) is figured and described for the first time. (7) The superior nectophore of a third new species, apparently a transitional one between dentata and sagittata is also described as C. intermedia. (8) It is possible to corroborate Bigelow's (1918) characters for dis- tinguishing the bract of Ceratocymba sagittata Quoy & Gaimard from those of all other abylids. (9) Several additional, though minor, characters have been found for distinguishing more readily the superior nectophores and gonophores of Abylopsis tetragona Otto and A. eschscholtzii Huxley — even despite poor preservation — than was possible with Bigelow's (1931) criteria. (10) Well preserved gono- phores can now be identified at least to genus in the absence of the bract, in all cases where the eudoxid is known. (11) Finally, four genera, Pseudabyla, Pseudocymba, Psevdabylopsis and Abylopsoides, are provisionally described. The structure of these suggests a ten- dency, not generally recognized among abylids, for the superior nectophores to develop asymmetrically as well as to increase or de- crease the number of facets. The revision is chiefly based on the "Dana" material, but I have also been able to compare this series with specimens taken on the "Albatross" and "Bache" in the collections of the Museum of Com- parative Zoology at Harvard College and of the U. S. National Museum. These specimens were used in studies of the group made by Bigelow (1911; 1913; 1918; 1919) and I have therefore had the oppor- tunity of checking my identifications with his. I have also had the privilege of receiving Dr. Bigelow's friendly advice and criticism, which encouraged me to attempt such a comprehensive review. In b bulletin: museum of comparative zoology addition, I wish to thank Captain A. K. Totton of the British Museum for his generosity in providing me with sketches, a discussion of his observations, and other material. Without his ready assistance I could not have clarified the status of Abyla trigona and A. carina. Similarly, Mr. A. Franc of the Museum National d'Histoire Naturelle, Paris, has kindly helped me in ascertaining the identity of Quoy and Gaimard's Abyla trigona. Dr. H. Engel of the Zoologisch Museum, Amsterdam, loaned me not only the "Siboga" specimens of A. trigona, but also the type of A. haeckeli. I am also extremely indebted to Dr. A. Vedel Tuning, Director of the Marinbiologisk Laboratorimn, Charlottenlund, Denmark, for entrusting me with the examination of the "Dana" siphonophores and for his hospitality at the laboratory on two occasions. Finally, I am beholden to the Milton Fund for the grant to Dr. Bigelow which enabled me to make the first visit to Den- mark in 1934 and to the Rask-Orsted Foundation for a most generous grant which permitted me to make a second trip in 1946. ABYLINAE L. Agassiz, 1862 Within recent years, the group of species under discussion here have been generally accepted, with two exceptions (Totton, 1932, p. 328; Leloup, 1934, p. 4), as forming the Subfamilies, Abylinae L. Agassiz and Ceratocymbinae Moser, of the Family Diphyidae (Bige- low, 1911; 1931; Moser, 1925). Totton (1932), however, elevated the Abylinae to the status of a family because observations of living speci- mens led him to believe that the functions of the nectophores were so distinctive that separation from the Diphyidae was warranted. What- ever the merits for such action may be, function cannot be used as a means for distinguishing siphonophore families in most plankton collections. It therefore seems to me that structure, rather than function, is preferable for differentiating the families in this group, in a report such as this which is based entirely on preserved specimens. As a result, the closely related genera under consideration are placed in the Abylinae, a subfamily of the Diphyidae,^ following other recent students of the group (Bigelow, 1911; Moser, 1925; Browne, 1926) rather than Totton (1932) and Leloup (1934). I The Diphyidae, broadly speaking, include all siphonophores with two nectophores of dis- similar structure (Bigelow, 1911). For exceptions, see remarks in next paragraph on certain monophyid species. sears: notes on siphonophores I do subscribe, however, to certain innovations made by Totton as they appear to lead to a more natural grouping. In discussing his Aby- lidae, Totton (1932, p. 331) compares Ceratocymha (under the name Diphyabyla^) with Ahyla, in such a way that it is to be presumed that he tacitly, at least, abandoned the Subfamily Ceratocymbinae of Moser (1925, p. 267), as had Browne (1926, p. 58) before him, in ac- cordance with Bigelow's (1911, pp. 215-216) earlier views. This course has also been taken more recently by Leloup (1934, p. 4). Most recent authors have agreed that the peculiar prolongation of the apex of the superior nectophore of Ccratocymba sagittata hardly seems sufficient to warrant a special subfamily for this one species. At that time the genus was monotypic but, as already mentioned, two species pre- viously referred to the genus Ahyla, Icuckartii and dentata, together with a new species, intermedia, are at present included in it. These form a well-defined transition from a species with no apical pro- longation {leuckartii) to a marked one {sagittata) (Fig. 1)*. In short, *e fHiur fia Fig. 1. Outline sketches of superior nectophores of Ceratocymha. A. Cera- tocymha leuckartii. B. C. dentata. C. C. intermedia. D. C. sagittata. ' This name was superseded by Cerntocymha sagittata Quoy and Gaimard, when it was defi- nitely proven that the eudoxid of that name belonged to the superior nectophore of Lens and Van Riemsdijk's Diphynbyla hubrecl.ti (Moser, 1913; 1925; Bigelow, 1918, p. 412). 2 The drawings throughout this paper have been prepared by Mr. Gale G. Pasley, to whom I am indebted for his painstaking efforts to devise a technique for illustrating transparent, geometric objects such as siphonophores, for his accurate observation of obscure details, and for his cooperation in making the drawings conform to the rather unusual requirements occasioned by the peculiarities of the group. 8 bulletin: museum of comparative zoology there now appears to be less justification than ever for maintaining a separate subfamily for Ccratocymba sngittata. Likewise, it is now agreed (Leloup, 1934; Bigelow and Sears, 1937, pp. 4-5) that Totton's (1932, p. 327) treatment of monophyid species results in a more natural classification of the heterogeneous genera previously referred to the Monophyidae. Within the limits of the present paper, Enncagonum hyaUnum Quoy and Gaimard is the only such species definitely proven to be monophyid. In the Abylinae, as here defined, then, are included E. hyalinuvi and Ceratocymba sagittata, together with three other species now assigned to that genus. The Abylinae differ from the other subfamilies of the Diphyidae in that the superior nectophores are rather generally prismatic and are distinctly smaller than the inferiors (with the exception of the modification found in Ccratocymba sagittata). The inferior nectophores are typically diphyid in character, with a basic plan of five ridges. There is a definite tendency, however, for one or another of the ridges to be suppressed {Ahyla, Ccratocymba and Bassia) while others may be greatly expanded to form wing-like structures {Abyla). Super- numerary ridges may also be present (Ccratocymba). The bracts, like the superior nectophore, are prismatic and are larger and more conspicuous in the known species of Ccratocymba than those of other diphyids. Despite the diverse shapes of the bracts in this subfamily, they all have a basic plan which has been homologized and well il- lustrated by Totton (1932, p. 337, text fig. 17). A modified copy of his figure (Fig. 2) is included here to dift'erentiate the bracts more clearly than can be done by words alone. It will be noted that in addition to diff"erences in external shape, various portions of the somatocyst may be absent. Thus, in some the anterior median horn (or oleocyst) is missing, in one the median dorsal descending branch, and in another, the two ventro-lateral branches. Abylid gonophores are usually larger and more robust {Abyla, Ccratocymba, Enncagonum) than those of other diphyids and in many ways resemble the inferior nectophores of the group especially in the arrangement and size of the oral teeth. In one genus (Abylopsis), they appear both in size and in the arrange- ment of the ridges somewhat like those of the Diphyinae but the prominent apophysis, especially in the young stages, readily distin- guishes them. The latter is markedly developed in all known genera of this group. Before proceeding further, it appears desirable not only to present G s PA seer, P£-i. Fig. 2. Bracts of Abylinae. (After Tottoii, lSC Fig. S. Lateral view of superior nectophores. A. Abyla carina with a dorsal facet of 6.9 mm. in length. B. A. trigona, with a dorsal facet of 4.9 mm. in length. C. A. srhmidti, with a dorsal facet of 7.4 mm. in length. D. .4. pprunna. with a dorsal facet of about '^^X^ mm. in length. sears: notes on siphonophores 23 6.G. PASUr, £>£i. D. Fig. 9. Ventral view of same specimens as in Figure 8. A. Abyla carina- R. A. trigona. C. A. schmidti. D. .4. peruana. 24 BCLLETIX: MUSETM OV ( OMPAKATIVF: ZOOLOGY S G. PASLfy^, />£C D. Hg. 10. Dorsal view of same specimens as in Figure 8. A. Ab'jla carina. H. A. t,ri(io)uj. C. ^4. schmidii. D. .4. peruana. n Fig. lo. Ventral view of same specimens as in Figure 12. A. Abi/la hicarinata. B. A. tottoni. C. A. brownia. D. A. haeckeli. E. A. ingeborgae. ee P'st^ey "n Fig. 14. Dorsal view of same specimens as in Figure 12. A. Abyla bicarinata. B. A. toltoni. C. A. brownia. D. A. haeckeli. E. A. ingeborgae. «*• PXSier cei Hg. 15. Left lateral view of inferior nectophores. A. Abyla hicarinala 19 mm. n length. B. A. tottoni of about 22 mm. in length. C. A. schmidti, of about .2.3 mm. in length. 80 bulletin: museum of comparative zoology (1888a), and others is still open to question now that we know the eudoxids are so similar that small cormidia cannot be distinguished. Until better material is available, than that found unattached in the samples of the "Dana", it is impossible to ascertain the characters of value in determining the individual species. The bracts are prismatic with six facets. The dorsal surface is in general smaller than the ventral. As a result, the basal and lateral facets are not in one plane, but veer outwards toward the ventral border. The dorsal facet is rectangular; the others, the apical, ventral, basal, and two laterals, are all trapezoidal. However, the laterals are imperfectly so, since the ventrobasal corners are cut away for the opening of the hydroecium. The latter occupies the basal half of the ventral surface and extends up into the interior of the bract almost to the apical facet. Above it lies the much swollen descending branch of the somatocyst. From its anterior end two thin ventrolateral branches extend down toward the ventrolateral corners of the apical facet. Gonophorcs (Fig. 26). The gonophores of Abyla, like those of Cera- tocymba and Enneagonum have five stout oral teeth surrounding the opening to the nectosac as well as a stout hook folded in toward the hydroecium from one of the ventral ridges.^ Similarly there are four ridges, two laterals and two ventrals, with a partial ridge extending upward from the dorsal tooth (as is also the case in Ceratocyinba and Enneagonum) . Hence, it is often difficult to determine the genus unless the specimens are well preserved. The outstanding differences among them occur on the apical half of the gonophore. In Abyla, the apico- dorsal and apicolateral ridges are arched and lie well below the apex of the nectosac. Beneath each arch is an indentation which is most prominent on the dorsal surface. Because of the position of these ridges, the apophysis is relatively more conspicuous in this genus than in Ceratocymba, but less so than in Enneagonum. Remarks In the closely related genus Abylopsis, the superior nectophores of the two species are so similar that it is only in recent years that it has become possible to distinguish them with any certainty in the absence of the inferior nectophores. Much the same situation apparently 1 Since the gonophores are mirror images of one another, the hook may occur on one or other of the ventral ridges, but not on both. SEAKS: NOTES ON SIPHONOPHORES 31 exists today especially among poorly preserved specimens of several species within the genus Abyla. It appears, however, that we are justified in distinguishing species of Abyla on the basis of seemingly minor differences in the superior nectophores but with greater dif- ferences in the inferior nectophores. We have found, for example, three superior nectophores {carina, trigona, schmidti) which unless well preserved, prove quite difficult to distinguish. We believe, how- ever, that in each instance, we have found sufficiently distinctive characters to separate them. In the older reports these three species and at least one other, were confused under one name, trigona} partly because so few specimens had ever been examined, partly because complete colonies were taken so infrequently that slight differences in the superior nectophores ap- peared to be merely individual variations, and partly because so many of the earlier figures and descriptions were lacking in detail. Thus, most of the earlier reports were diagnostic of the genus as a whole as we define it here rather than of any particular species. As a result, most subsequent authors referred their specimens to trigona rather than establish a new species for individuals with seemingly minor modifications. Until recently, therefore, when a large series of speci- mens could be examined only four species in the genus were recognized as distinct. Two of these, Icuckartii and dentata, are now referred to Cerafoci/tnba: the other two, hacckcli and bicnrinata, still remain in the genus Abyla. The two latter are quite distinctive and have probably seldom been confused with trigona.^ Thus, the superior nectophore of 1 Bigelow (1911; 1918; 1919), for example, lists as Ahyla trigona, three other species, carina, schmidti and a new species, peruajia, as listed below: M. C. Z. No. 1602, "Albatross" Sta. 4715, ^4. carina — eastern tropical Pacific. M. C. Z. No. 1603, "Albatross" Sta. 4684, A trigona — eastern tropical Pacific. M. C. Z. No. 16.58, "Albatross" Sta. 5601, .1. sc/im)rf«i — Celebes. M. C. Z. No. 3074, "Albatross" Sta. 5672, .4. schmidti — Philippines. U. S. Nat. Mus. No. 28349, "Albatross" Sta. 4673, Abyla peruana n. sp. — off Peru. U. S. Nat. Mus. No. 41.556, "Bache" Sta. 10163, .4. trigona — off North Carolina. U. S. Nat. Mus. No. 41557, "Bache" Sta. 10166, .4. carina — off South Carolina. V. S. Nat. Mus. No. 41558, "Bache" Sta. 10171, .4. trigona — off Bermuda. U. S. Nat. Mus. No. 41562, "Bache" Sta. 10194, .4. trigona — south of Bermuda. U. S. Nat. Mus. No. 41563, "Bache" Sta. 10207, -4. carina — Straits of Florida. U. S. Nat. Mus. No. 41564, "Bache" Sta. 10211, .4. trigona, A. carina? — north of Bahamas Banlc. In most instances, the original identification was based on detached superior nectophores. The three species concerned are, as stated above, difficidt to distinguish unless well preserved and unless one has been fortunate to study a series of cqmplete colonies. 2 Hu.xley (18.59, p. 48) states that, "It is with some little hesitation that I identify this species with the Abyla trigona of Quoy and Gaimard .... but there are so many points of similarity that I prefer to run the risk of making a species too few rather than one too many." Huxley's specimen has usually been referred to the synonymy of .4. hatckeli, but is possibly A. ingeborgae. Totton (1925) reports that among Haeckid's specimens of .4. carina, there was one of A. haecktli. 32 bulletin: museum of comparative zoology Abyla haeckcli (Lens and Van Riemsdijk, 1908) is characterized by an extra transverse ridge subdividing the apicoventral facet and that of A. bicarinata (Moser, 1925) by a marked expansion of the lateral ridges. However, a new species, A. ingeborgae, also has a transverse ridge separating the ventral and apicoventral facets and a second new species, A. broivnia, has expanded lateral ridges. It is possible, there- fore, that these species have been confused with haeckeli and bicarinata respectively. Fortunately, it has been possible to examine Quoy and Gaimard's (1827) specimens of irigona and to compare them with specimens de- scribed by Haeckel (1888a) as carina. Much of the confusion would appear to stem from the fact that both Haeckel (1888a) and Quoy and Gaimard (1827) apparently had two species in their samples. Today, Quoy and Gaimard's two superior nectophores are detached from the inferior nectophores. I believe that they never were attached, because the apophyses of the two inferior nectophores appear too big to fit into the hydroecium of either superior nectophore. Secondly, they are very similar to specimens in Haeckel's sample' and in the "Dana" collection, several of which also have small inferior nectophores at- tached within the hydroecium. These inferior nectophores are quite different from those in Quoy and Gaimard's sample. Finally, a colony in the Haeckel samples with a badly damaged superior necto- phore has an inferior nectophore which appears to be the same as Quoy and Gaimard's. The superior nectophore is, however, distinct from theirs, but resembles three others in Haeckel's sample. Since it thus seems certain that there are the same two species in both Quoy and Gaimard's (1827) and Haeckel's (1888a) samples, a decision must be made as to which should be trigona and which should be Haeckel's carina. It seems to me that Lens and Van Riemsdijk (1908, pp. 30-31) have already indicated the answer. They compared their specimens of Abyla trigona with those of Quoy and Gaimard and found them to be identical "in all respects". Re-examination of their specimens has shown their statement to be correct for the superior nectophores.^ It would therefore seem that the superior nectophores 1 One of these has a small inferior nectophore attached within the hydroecium. There is also a loose inferior nectophore quite unlike the inferior nectophores in Quoy and Gaimard's sample. ' Lens and Van Riemsdijk (1908, p. 28) show their hesitancy in identifying the inferior nectophore by recording it as ? Abyla trigona (Cat. 78D). This is in reality the new species A. schmidti. A second poorly preserved specimen (Cat. 126D) may possibly belong to this species. It certainly is not the same as those in Haeckel's (188Sa) and Quoy and Gaimard's (1827) samples. sears: notes on siphonophores 33 can be considered to have been designated by them as Abyla trigona. Since the inferior nectophores in Quoy and Gaimard's sample ap- parently belong to the same species as the colony and three superior nectophores in Haeckel's, it would seem appropriate to designate these as carina. Keys to the Nine Known Species of Abyla (Figs. 6-15) A. Based on characters of the superior nectophores. 1. Apicoventral facet subdivided by a transverse ridge 2 Apicoventral facet not subdivided 3 2. Ventral facet approaches a regular pentagon in shape; protrusion at juncture of horizontal and lateral ridges markedly overhangs basal half of ventrolateral surface; ridges elevated like rim of a pie plate . . Abyla haeckeli Lens and Van Riemsdijk Ventral facet elongate with basal sides of pentagon roughly three times as long as apical ones; protrusion at juncture of lateral and horizontal ridges not excessive; ridges well defined but not markedly elevated above facets Abyla ingeborgae n. sp. 3. Nectophores nearly circular in dorsal or ventral view due to pronounced expansion of lateral ridges 4 Nectophores elongate in dorsal or ventral view, but with a more or less pronounced knob at the juncture of the lateral and horizontal ridges . 5 4. All ridges well defined; a definite angle at juncture of horizontal and lateral ridges, greatest width of ventral facet about one half length from insertion of horizontal ridges to basal tip . .Abyla broxcnia n. sp. Ridges delineating the apicodorsal facet as well as horizontal ridge rounded and often indistinct, lateral ridge circular throughout, not angular, greatest width of ventral facet greater than distance from insertion of horizontal ridges to basal tip . . . .Abyla bicarinata Moser 5. Greatest width of ventral facet is about the same (0.87 to 1) as its length from the insertion of the horizontal ridges to its basal tip 6 Greatest width of ventral facet is only about one half to two thirds (0.45 to 0.6) the length from the insertion of the horizontal ridges to its basal tip 8 t). In side view, apex of hydroecium considerably higher, apex of nectosac lower than that of somatocyst; horizontal ridge crosses somatocyst only slightly above its middle; no obvious depression ventrad to trans- verse apical ridge Abyla -peruana n. sp. Apices of nectosac and somatocj^st at about same level, apex of hydroe- cium only slightly higher; horizontal ridge crosses somatocyst well 34 bulletin: museum of comparative zoology above middle; obvious depression ventral to transverse apical ridge. . Abyla toltoni n. sp. 7. Transverse ridge in true side view lies above somatocyst, resulting in elongate apicodorsal facet Abyla schnddti n. sp. Transverse ridge in true side view lies above hydroecium 8 8. In true side view aijicodorsal facet almost vertical from insertion of lateral ridge to apical transverse ridge; lateral border of basal facet curved and tends to parallel horizontal plane; heavy and irregular serrations on lateral ridges Abyla trigona Quoy and Gaimard In true side view apicodorsal facet essentially flat; lateral border of basal facet diagonal and only slightlj' curved; serrations fine Abyla carina Haeckel B. Based on characteristics of the inferior nectophores.^ 1. About two to five teeth on comb of left ventral wing 2 About six to ten teeth on comb of left ventral wing 4 2. Two or three teeth on comb of left ventral wing; base of right ventral wing with thickening outlined by rather small teeth, the two ventral ones on both iimer and outer row being the heaviest Abyla haeckeli Lens and Van Riemsdijk Four or five teeth on comb of left ventral wing, right ventral wing triangular 3 3. Base of right ventral wing with inner row of stout teeth continuous with its ventral margin and the outer row of finer teeth projecting below on a more or less well defined triangular pad; left ventral wing not continuous with left ventral tooth Abyla schynidti n. sp. Base of right ventral wing with three or four stout teeth on inner margin, with a few weak teeth almost scallops on outer; left ventral wing continuous with left ventral tooth Abyla ingeborgae n. sp. 4. Outer row of teeth on basal margin of right ventral wing continuous with its ventral margin; with inner row ending on inner surface. . . .5 Inner and outer rows of teeth on basal margin of right ventral wing merge to become its ventral margin 6 5. Nectophore (exclusive of apophysis) about as wide as it is long, nearly circular in general appearance; usuall}^ about 7 teeth on comb of left ventral wing Abyla bicarinata Moser Nectophores (exclusive of apophysis); somewhat longer than wide; ovoid in general appearance; S-9 teeth on comb of left ventral wing. Abyla tottoni n. sp. 6. Ventral teeth elongate, heavily serrated; about six teeth on comb of left ventral wing; teeth on basal margin of right ventral wing heavy and prominent Abyla trigona Quoy and Gaimard ' Inferior nectophores have not yet been found for Abyla brownia n. sp., or A. peruana n. sp. sears: notes on siphonophores 35 Ventral teeth stubby, 9-10 teeth on comb of left ventral wing; teeth on basal margin of right ventral wing scarcely more than strong serrations. Abyla carina Haeckel Abyla trigona Quoy and Gaimard 1827 Abyla trigona, Quoy and Gaimard, 1827, pp. 14-15, pi. 2B, figs. 1-8 (partim); Gegenbaur, 1859, pp. 1-10, pis. 1-2, figs. 1-12; 1860, pp. 337-349, pis. 26- 27, figs. 1-12; Lens and Van Riemsdijk, 1908, pp. 29-34, text figs. 24-31, pi. 4, figs. 34-36; Bigelovv, 1911, pp. 221-222 (partim); 1918, pp. 408-409 (partim); Moser, 1 925, pp. 301-310, pi. 18, fig. 7 (partim); Moore, 1949, p. 13 (partim). ? Abyla krigorM, Eschscholtz, 1829, pp. 131-132; Blainville, 1830, p. 123; 1834, p. 135, pi. 4, fig. 4 (not seen); Chun, 1888, pp. 1160-1161; 1897, pp. 31-32; Haeckel, 1888, p. 35; Fewkes, 1889, p. 519; Schneider, 1898, pp. 90-91, 197; Bedot, 1904, p. 27; Moser, 1912a, fig. 20; Kawamura, 1915, pp. 578- 580, pi. 15, figs. 27-28; Browne, 1926, p. 62; Leloup, 1932, pp. 20-22, fig. 3; 1933, p. 21; 193.5a, p. 5; Totton, 1932, p. 3.32, fig. 17B. ■!A/nphiroa alata, Blainville, 1830, p. 121; 1834, p. 133, pi. 4, fig. 1 (not seen); Huxley, 1859, p. 64, pi. 5, fig. 1; Chun, 1888, p. 1160; 1897, pp. 31-32; Lens and Van Riemsdijk, 1908, p. 28, pi. 4, figs. 37-38. Diphyes abyla, Quoy and Gaimard, 1834, pp. 87-88, pi. 4, figs. 12-17 (partim). Non Abyla trigona, Vogt, 18.54, pp. 121-127, pi. 15, fig. 4, pi. 20, figs. 4-7, pi. 21, figs. .3-6, 10-13; Huxley, 1859, pp. 47-48, pi. 3, fig. 1; Bigelow, 1911, pp. 221-222; pi. 13, figs. 3-4 (partim); 1918, pp. 408-409 (partim); 1919. p. 334. Abyla carina, Haeckel, 18S8a, pp. 156-157, pi. 35 (partim). lAmphiroa carina, Haeckel, ]S88a, pp. 114-115, pi. 36. Superior ncdophorc (Figs. 8B, 9B, lOB). The superior nectophore of Ahyla trigona, even when poorly preserved, is perhaps the most readily distinguished of the three species, trigona, carina, and schmidti. Most of the ridges are heavily and irregularly serrated, the laterals and at times the basal ridges of the ventral facet and the laterals of the dorsal facet are especially so. In side view, the profile of the dorsal half is characteristic: the lateral ridges of the basal facet almost describe a semicircle close to the dorsal wall of the hydroecium, but the curvature widens gradually dorsad, to end parallel to the horizon- tal plane or to continue down onto the prominent dorsal teeth. Almost without exception, whether well preserved or not, the apico- ' At times the .sharp curvature is absent and the curve described may approach the diagonal of A. ciirina. 36 bulletin: museum of comparative zoology dorsal facet is sharply bent upward from the insertion of the lateral ridges to the transverse apical ridge. The latter lies above the center of the hydroecium. Just ventral to it the apicoventral facet drops away gradually so that there is only a slight depression in the furrow. The lateral protrusions are somewhat sharper and more prominent than those of A. carina. In addition, the facets of the nectophore as a whole are depressed below the ridges surrounding them. Thus, the nectophore often appears thinner and more fragile than carina. Inferior nectophore (Fig. 11 A). The inferior nectophore of A. trigona is only about one-half as wide as it is long. None of the ridges are markedly expanded. It is most readily distinguished from carina, however, by the decrease in number of teeth on the comb (6-8). In addition, in mature specimens, there are two rows of large sharp teeth on the basal margin of the right ventral wing, much as Haeckel (1888a, pi. 35, figs. 8-9) has shown them. These seem to vary in appearance with growth. In small specimens found attached within the hydroecium of the superior nectophore, the teeth are little more than a series of punctae and those on the outer row diverge somewhat from the inner and produce a small angular flap rather like that of A. schmidti. In somewhat older nectophores, the latter tends to dis- appear or to become very small. On such nectophores the two rows of teeth which are essentially parallel and separated only by a thicken- ing of the basal part of the wing, continue for a distance along the ventral margin. In the oldest specimens examined, however, the teeth were limited to the basal margin. The ventral teeth, likewise, show some variation with age. In younger specimens, they are straight and elongate rather like those on the inferior nectophore of Ceratocymba leuckartii. They become relatively shorter with age but they always remain straight, rather sharper than in most species, and are always coarsely serrated. The other oral teeth are also more heavily serrated than in any other known species of this genus. Eudoxid. Although the eudoxid of this species may have been described by Gegenbaur (1860) and Haeckel (1888a) we cannot be sure whether their specimens were A. carina or A. trigona. As no cormidia were found in the "Dana" collection sufficiently far advanced to determine the specific characters, it is not possible to describe the eudoxid of this species for certain. sears: notes on siphonophores 37 Remarks On examination of Lens and Van Riemsdijk's (1908) specimens, most were in every way similar to others of trigona we have seen. How- ever, two of their superior nectophores (Cat. 77B) had much more pronounced lateral protrusions. Nevertheless, they too appear to belong to this species. Abyla carina Haeckel 1888 Abyla trigona, Quoy and Gaimard, 1827, pp. 14-15, pi. 2B, figs. 1-8 (partim); Bigelow, 1911, pp. 221-222 (partim); 1918, pp. 408-409 (partim); Moore, 1949, p. 13 (partim). Abyla carina, Haeckel, 1888, p. 35; 1888a, pp. 156-159, pi. 35 (partim). Superior nedophore (Figs. 8 A, 9A, 10 A). The superior nectophore of A. carina has a number of distinguishing features, although they are not always apparent in poorly preserved specimens. In lateral view the sides of the basal facet are diagonal and only slightly curved because the dorsal teeth do not protrude below it as they do in trigona. The apicodorsal facet is usually a flat surface and in side view its lateral ridges likewise form a diagonal to the dorsal facet. There may, however, be a slight break at the insertion of the lateral ridge especially in somewhat shrunken and poorly preserved specimens. If so, the transverse apical ridge which lies above the center of the hydroecium is elevated slightly above the surface of the facet as a whole. This is never as exaggerated as is characteristic for trigo7ia. Ventral to it, there is almost always a slight depression in the furrow of the apico- ventral facet. The lateral protrusions as seen in ventral or dorsal view are not prominent and may be quite unobtrusive. The ridges of the nectophore as a whole are not raised above the facets, but most of these on the basal half of the nectophore are slightly serrated. Finally, the nectophore as a whole appears more massive than that of trigona. Inferior nectophore (F\g. 7). Quoy and Gaimard's specimens of the inferior nectophore, now referred to A. carina (p. 82), which are pre- served in alcohol, have apparently shrunken considerably because they are only about one-half as wide as they are long. A well preserved specimen in the "Dana" collection, on the other hand, is about two thirds as wide as it is long. This is due chiefly to the expansion of the right ventral wing, as the dorsal ridge is only moderately expanded. 38 bulletin: MXTSEnM of comparative zoology The general shape is not sufficient to differentiate this nectophore from that of other species. One feature which appears reliable is the number of teeth (9-10) on the comb of the left ventral wing. These were all the same size in both Quoy and Gaimard's and Haeckel's specimens, but they formed a graduated series on some of the "Dana" specimens which seemingly belong to this species. An equally distinctive char- acter is the dentition at the base of the right ventral wing. The teeth on the latter are not as robust and jagged as are those of A. trigona. The inner row of nine or ten small spines is parallel to the basal margin close to the two ventral teeth but ventrad it gradually swerves inward a short distance from the margin crossing the inner surface to merge with the outer row at the ventral margin much as is shown in one of Haeckel's drawings (1888a, pi. 35, fig. 1). The outer row has about the same number of teeth as the inner, which are likewise weak and relatively inconspicuous. The ventral teeth are often quite blunt except for a sharp point at the tip. Eudorids. The eudoxids belonging to this species have been de- scribed by earlier authors, but it has not been possible to check their observations with specimens of known parentage. It therefore seems better to omit a description as it would only be misleading. Abyla schmidti' n. sp. Abyla trigona, Lens and Van Riomsrlijk, 1908, pp. 29-.34 (partim); Bigolow, 1919, p. 334. Abyla bicarinata Moser, 192.5, pp. 29S-.301 (inferior nectophore) pi. 19, figs. 7-9 (partim). Abyla sp. Totton, 1950. The type specimen of Ahi/la schmidtl consists of a well preserved superior and inferior nectophore taken at "Dana" Station 3922, at 3°45'S, 56°33'E with 1000 meters of wire out, on 12 December 1929 at 1850 hours with an open ring trawl, 300 cm. in diameter. These will be deposited in Universitets Zoologiske Museum, K0benhavn, Denmark. Superior nrctophorr (F'igs. 8C, 9C, 10("). The superior nectophore of Abyla schmidii, although quite similar to the two previous species, is best differentiated by its proportionately larger apicodorsal facet. In a true lateral view, this is seemingly effected by a protrusion dorsad ' Named in honor of the late Professor .lohannes Schmidt, the le;Kler of the Carlsberg Fo>iii- dation's Oceanographical Expedition Round the World, 192S~.30. sears: notes on siphonophores 39 over the dorsal facet and by the more ventral position of the transverse apical ridge over the dorsal portion of the somatocyst. Ventral to this ridge the apical portion of the apicoventral facet gradually slopes ventrad without any marked depression, although it is slightly furrowed. A second feature which is helpful in identifying many specimens is a dorsal facet tapering toward the base rather than one which bulges in the middle as in trignna and carina. Inferior necfopliore (Fig. 15C"). On well preserved specimens of the inferior nectophore of A. schinidti, the right ventral wing is expanded so that the width of the nectophore nearly equals its length. In most specimens, however, it is damaged so that the relative measurements are not helpful in identification. Its triangular shape, however, pro- vides the best means for separating this species from others in the genus. The inner row of teeth along the basal margin of the right ventral wing is rather strongly developed and joins the outer row to continue apically as the ventral margin. The triangular protrusion hanging below the inner row is in reality the thickening between the two rows of teeth since the outer row of weak teeth delimits it before it merges with the inner. This species is also readily recognized by the teeth on the comb on the left ventral wing which are fewer in number (4-5), larger, and more robust than in any other known species of Abyla. Eudoxid. Cormidia have been found on stems attached to superior nectophores of .1. schmidti but they are not distinctive enough to assist in identifying loose bracts and eudoxids. About all one can determine is that the bract is a typical "Amphiroa". Small gonophores known to be schmidti likewise appear very similar to those found on cormidia attached to the superior nectophores of other species. On these the large tooth on one of the ventral wings is thin, elongate, and almost fingerlike. Insofar as we have been able to ascertain, teeth of this sort are characteristic of small specimens belonging to a number of other species. Other characters which might be diagnostic in difi'erentiating the species are not obvious. It is possi- ble that as the gonophore matures, differences not obvious on small specimens may appear. Abyla haeckeli Lens and Van Riemsdijk 1908 "^ Abyla trigona Huxley, 1859, p. 47, pi. 3, fig. 1.' ' Huxley's specimen actually may have been the species described below as inge.horyne. 40 bulletin: museum of comparative zoology lAmphiroa angulata Huxley, 1859, pp. 64-65, pi. 5, fig. 2. lAm'phiroa carina Haeckel, 1888a, pp. 114-116, pi. 36. ?Abyla alata Haeckel, 1888a, pp. 113 and 156. Non Amphiroa dispar Bedot, 1896, p. 373, pi. 12, figs. 5-6. Abyla haeckeli Lens and Van Riemsdijk, 1908, pp. 32-34, text figs. 32-40, pi. 5, figs. 39-41; Bigelow 1911, pp. 222-224, pi. 13, figs. 1-2; Moser, 1925, pp. 310-318, pi. 18, fig. 6. ?Abyla haeckeli Browne, 1926, p. 63^; Leloup, 1932, pp. 19-20; Totton, 1925, pp. 446-447; 1932, pp. 331-333, figs. 12-13, 17B2. Non Abyla haeckeli Totton, 1932, text fig. 12. Superior ncdophore (Figs. 12D, 13D, 14D). In outline, the superior nectophore of haeckeli is as wide as it is high. It is also more truly prismatic than that of any other abylid. Mentioned, but not stressed, by Lens and Van Riemsdijk (1908, p. 33) in their original description are the flat facets. These are accentuated by sharp raised ridges so that most facets have rims the shape of polygonal pie plates. Two other characters at once separate this species from those already described and contribute to its unique appearance. The apicoventral facet is separated by a second transverse ridge to form a quadrangular facet on the ventral half of the apical surface and a nearlj' regular pentagonal one on the ventral surface. The latter is nearly twice as wide (1.8) at the insertion of the horizontal ridges as thence to its basal tip. The apical ventrolateral facet is large and overhangs the basal one, due to the shelf-like protrusion formed by the horizontal ridge and the basal half of the lateral ridge. The surface of the basal ventrolateral facet, however, appears to be only slightly smaller than the apical one. Nevertheless, in side view the horizontal ridge appears to lie just above the basal tip of the somatocyst. In no other Abi/la have we found the horizontal ridge lower than the midpoint of the somatocyst. Inferior nectopliore (Fig. 11 B). Several inferior nectophores so small that they have been shielded within the hydroecial cavity, have enabled us to learn the peculiarities of this species. Study of these permits identification of larger poorly-preserved detached inferior ' The fact that Browne (1926, p. 63) states that "the species is likely to escape notice unless every anterior nectophore of Abyla trigona is carefully examined on the ventral side" suggests that he may have had ingeborgae rather than haeckeli. 2 "One anterior nectophore, the identification of which rests upon the presence of a trans- verse ridge dividing the ventral facet into two parts" (Totton, 1932, p. 331) might equally well refer to ingeborgae. sears: notes on siphonophores 41 nectophores. The dorsal tooth, the largest of the five surrounding the oral cavity, juts forward and slightly downward. The two lateral oral teeth are relatively small, the right being distinctly the smaller. Rather than being midway in position between the dorsal and ventral teeth as is customary, both the laterals are displaced so that they lie close to the dorsal. The left ventral wing is continuous with the ridge of its corresponding tooth, the coarse serrations extending part way down the ridge of the latter. There are two or three teeth on the comb. The thickened basal margin of the right ventral wing is concave and is delimited by about four teeth on the inner and outer rows. Those on the inner row are somewhat heavier than the outer. The teeth at the ventral end of each row are stubby and connected to each other. The ventral margin above these teeth is likewise thickened. The ventral wings may be sufficiently expanded so that the nectophore as a whole may be nearly as wide as it is long. Bract (Fig. 26A). Cormidia attached to the stem of the superior nectophore appear to have a peculiar flat dorsal facet with raised rims of the sort found only on the superior nectophore of harckeli. The basal facet is at right angles to the dorsal and lacks all indications of teeth. Loose specimens have been found with these characters, but their gonophores have been too badly preserved to check their identity further. Gonophore. Insofar as we can judge, the only distinctive feature of the gonophore of A. hacckcU is the stubby engorged curved tooth on the midportion of one of the ventral wings. We have not found older specimens in sufficiently good condition to study further. Remarks The peculiar shape of the horizontal and lateral ridges of the superior nectophore of haeckrli is figured but not mentioned by Lens and Van Riemsdijk (1908, pi. 5, fig. 41). Re-examination of their specimens proves that our specimens are more exaggerated in this respect than theirs. In Bigelow's (1911, pi. 13, fig. 1) figure, the horizontal ridge appears to be rather higher and it together with the basal half of the lateral ridge apparently is not protruded to over- shadow the basal portion of the ventrolateral ridge. We have re- examined his specimens which are relatively small and perhaps not fully developed and his drawing is correct. Moser (1925, pi. 18, fig. 6) follows Bigelow in general outline but she depicted the facets as flat 42 Bia.LETlN: MUSEUM OF COMPARATIVE ZOOLOGY and surrounded by raised rims of the sort observed on our specimens. We have seen one rather poorly preserved specimen which is very Hke the one Moser (1925) figured. The question of the identity of all the Ahyla eudoxids is perplexing because insofar as we can see, the differences in each species are so slight that they cannot be determined for either the bracts or gono- phores when small specimens are attached to the stem within the hydroecium of the superior nectophore. We have found specimens of A. hacckeli with cormidia which are not too unlike the one described by Huxley (1859, pi. 5, fig. 2) and which we believe develop into the detached specimens of the sort shown in Figure 26. Other bracts previously have been referred to haeckeli (Bedot, 1896, p. 373, pi. 12, figs. 5-6 [Amphiroa disjxir]; Totton, 1932, text fig. 12) because they seemingly differed from the one described earlier as that of trigona. These, however, are not hacckeli, I believe, because the basal facet in haeckeli is at right angles to the dorsal and they do not have the angularity of Bedot's (1896, pi. 12, figs. 5-6) or the cleft of Totton's (1932, text fig. 12) specimens. Abyla ingeborgae^ n. sp. 1 Ahyla trigona, Huxley, 1859, pp. 47-48, pi. 3, fig. 1. Ahijla haeckeli Kawamura, 1915, pp. 577-578, pi. 15, figs. 24-26. 1 Ahyla haeckeli, Totton, 19.S2. pp. 881-33.3. The type specimens of Abyla ingehorgae are 23 superior nectophores from (/>. K 1.— Sta. No. 4762 at 8°13'S, 2°54'E taken on 11 February 1933 at 1930 hours with a stramin net 200 cm. in diameter. There are also five loose inferior nectophores which, we believe, belong to this species. These will be deposited in Universitets Zoologiske Museum, K0benhavn, Denmark. Superior nectophore (Figs. 12E, 13E, 14E). The superior nectophore of Ahyla ingehorgae is very similar to that of schmidti in general appear- ance, when well-preserved, but it is readily distinguished from it by a transverse ridge dividing the apicoventral facet into two. Because of this character it may have been referred to haeckeli in the past. Never- theless, the two species are quite distinct. The ventral surface is only about one half as wide at the insertion of the horizontal ridge as it is high from this point to its basal tip. This is seemingly due to the two • Named in honor of Mrs. .Tobannes Sphmidt, thr widow of the expedition's leader. sears: notes on siphonophores 43 elongate basal ridges of this facet. In addition, the width of the nectophore is only about two-thirds the height (0.69). The lateral protrusions are relatively inconspicuous and, in addition, are dis- tinctly higher than those of harrkrii. The horizontal ridge in side view appears to lie well above the middle of the somatocyst not well below it, as in hoeclctii. Hence, the apical ventrolateral facet is definitely smaller than the basal ventrolateral one below it. The dorsal facet is almost a perfect rectangle. In other words, it does not taper from the top as is usual in schmidii or bow on the sides as in a number of other species. The only serrations visible even under quite high magnifica- tion are at the base of the dorsal wall of the hydroecium. Inferior nectophore (Fig. IIC). The inferior nectophore which we believe to be that of A. ingeborgae has some characters rather like that of .1. haeckeli, some like those of A. schmidii. Thus, the comb has five teeth like schmidii, the right ventral wing is more nearly triangular rather than concave on its basal border and the dorsal tooth is smaller and less prominent. On the other hand, the position of the lateral oral teeth is close to the dorsal one as in haeckeli, but the right lateral tooth is almost as large as the left and has a well defined ridge. Likewise, the left ventral wing is continuous with the ridge of the left ventral tooth. .\lso, the dentition of the right ventral wing is more like that of haeckeli. There may be three or four stout teeth on the inner basal margin, but the outer one has only a weak scallop or two. There are, however, no strong teeth delimiting the ventral extent of the basal margin and the ventral margin is not as thickened as it is in haeckeli. The general appearance of the nectophore is very like that of haeckeli in the specimens we have seen. Thus, while it may be almost as wide as it is long due to the expansion of the ventral wings, these are fragile and easily damaged. Eudoxid. A cormidium was found attached within the hydroecium of one specimen. The bract is characteristic of that for the genus and the gonophore is seemingly like that of haeckeli with perhaps a con- spicuous dorsal tooth and ridge and variations in the serrations on the ventral ridges and teeth. '^rn^ Remarks One small inferior nectophore was found within the hydroecium of a superior nectophore of A. ingeborgae, but it was not possible to ob- serve characters to distinguish it from A. haeckeli. The above de- 44 bulletin: museum of comparative zoology scription is therefore based on two fairly well preserved specimens found in the same sample as the type (superior nectophores) and at one other station, where the superior nectophores of A. ingchorgae were found. It seems probable, therefore, that these inferior necto- phores are actually those of A. ingchorgae, because we have proof from the small attached inferior nectophores that they are very like those of A. haeckeli and because the larger detached specimens have always been found associated with the superior nectophores of A. ingchorgae. Abyla peruana n. sp. Abyla trigona, Bigelow, 1911, pp. 221-222 (partim); pi. 13, figs. 3-4. The type specimen of Ahyla peruana is a superior nectophore col- lected at "Albatross" Sta. 4673 off Peru by Dr. H. B. Bigelow on 21 November 1904 and was deposited in the U. S. National Museum (No. 28349). It has previously been referred to A. trigona. Superior nectophore (Figs. 8D, 9D, lOD). The superior nectophore of Abyla peruana superficially looks more like A. haeckeli than A. trigona. This is chiefly due to the fact that the horizontal ridge when seen in side view lies just above the middle of the somatocyst. There is no danger that this is a young specimen of haeckeli because it lacks the extra ridge separating the ventral and apicoventral surfaces. Indeed, in the one specimen seen, the two surfaces are only at an angle of about 45° rather than at a right angle. This appears to be due to the fact that the apex of the hydroecium is distinctly higher than either the somatocyst or nectosac. The transverse apical ridge is thus corres- pondingly higher. The apex of the nectosac is also lower than that of the somatocyst. In other species of Ahyla these have been at the same level, although the apex of the hydroecium is usually slightly above the other two cavities. The ridges outlining the facets are in general more pronounced and wider than they are in haeckeli and presumably stood well above the facet when living. As the lower part of the dorsal facet and the basal facet are damaged, the only other peculiarity that one can see is an unusually long and heavy keel beneath the hydroecium. Inferior nectophore. Unknown. Eudoxid. Unknown. sears: notes on siphonophores 45 Remarks Although the specimens from two stations of the "Albatross" Ex- pedition to the Eastern Tropical Pacific have not been found, it seems very likely that it was this specimen which Bigelow (1911, pi. 13, figs. 3-4) drew. The specimen is now damaged as described above, but it is hard to reconcile Bigelow's (1911, pi. 13, fig. 4) ventral view either with his (1911, pi. 13, fig. 3) lateral view or its present condition. While it is possible that the ventral ridge of the basal facet is as wide as he shows it, it seems unlikely that the nectophore ever narrowed along the lateral ridge (1911, pi. 13, fig. 4). One other feature not shown in Bigelow's drawings is the very marked ridges. Abyla bicarinata Moser 1925 Abyla bicarinata Moser, 1925, pp. 298-299, pi. 19, figs. 3-6 (superior necto- phores). Non Abyla bicarinata Moser, 1925, pp. 299-301, pi. 19, figs. 7-9 (inferior nectophores). Superior nectophore (Figs. 12A, 13A, 14A). The superior nectophore of Abyla bicarinata is unique in that it is wider than it is high. This is due to the extraordinary wing-like expansion of the lateral ridges. In some specimens these appear to end above the basal margin of the hydroecium but it is usually possible to trace them to the base, at least as a series of punctae. In addition, the edges of the facets are all rounded and tumid, but the exact location of the ridges may be traced as fine hair-like lines. There is no depression or furrow ventral to the transverse apical ridge. The apical ventrolateral facets appear more as a slight depression between the apical surface and upper half of the lateral and horizontal ridges. Furthermore, they do not actually lie in the horizontal plane, but are part of the apical surface. The re- sult is that in a true lateral view the somatocyst may be viewed in its entirety through the basal ventrolateral facet. Several other characters make this species quite distinct from the next species {brownia) to be described. The width of the ventral facet at the insertion of the horizontal ridge is about three-quarters (0.76) the height, thence to its basal tip. In ventral view the vertical dis- tance between the insertion of the horizontal ridges to the ventral facet and the transverse apical ridge is proportionately greater in this 46 bi^lletin: museum of comparative zoology species than in brotvnia. In side view, the lateral ridges of the basal segment are diagonal almost as in A. carina. Inferior nrdophore (Fig. 15A). The inferior nectophore of Ahyla hicarinaia is the only one known so far which is as wide as it is long. In addition, the left lateral ridge is as pronounced and as conspicuous as the dorsal one. Both are more expanded than in any other species and the two together, in preserved specimens at least, bound a some- what circular rather than rectangular surface covering most of the left lateral half of the nectophore. In addition, the right ventral wing' forms an almost perfect semicircle. In fact, the inferior nectophore does not have the conventional diphyid outline but rather gives the impression of being two, flat, superimposed discs. While the teeth surrounding the opening to the nectosac are very like those of the other species in the genus, they are stronger and more prominent. The dorsal tooth lacks serrations. One further character helpful in dis- tinguishing the inferior nectophores is that there are 4-7 teeth on the comb. Finally, there are six teeth on the inner row of the basal por- tion of the right ventral wing. This row at first parallels the outer but diverges from it to end on the inner surface well in from the ventral margin, a characteristic also found in A. tottoni. Eudoxid. Unknown. Remarks The most obvious differences which Moser (1925) apparently failed to stress between this and closely related species is that in most speci- mens the nectophores are opaque as in hacckcU. Also, all of them are rather more turgid than others in the genus. In this respect, they re- semble Ceratocymba sagittata and C. dentata. In one colony from the "Dana" collection, however, only the outer part of the ridges is opaque and a few of the inferior nectophores are quite transparent. Abyla BROWNiA^ n. sp. ?Abyla sp., Totton, 1950. ' In some specimens viewed from tlie right side, there iippears to be a partial ridge. Careful exf:minatiQn indicates that this is in reality the line of attachment of the right ventral wing, which may be more or less pronounced depending on preservation. - Named in honor of its discoverer. Miss Joan A. Brown sears: notes on siphonophores 47 The type specimen of Ahyla hrownia is a superior nectophore, which came from "Dana" Sta. 3964vm at 25°19'S, 36°13'E on 15 January 1930 at 2355 hours in a stramin net 150 cm. in diameter, with 2000 meters of wire out. It will be deposited in Universitets Zoologiske Museum, K0benhavn, Denmark. Superior nectophore (Figs. 12C, 13C, 14C). This nectophore will not be confused with any other species except possibly bicarinata, for it too has expanded lateral ridges which end just above the basal margin of the hydroecium. Relatively, these are not as wide as in bicarinata; the greatest width is somewhat less than the height (0.85). The ven- tral facet is likewise narrower, being somewhat less than half as wide (0.45) at the insertion of the horizontal ridges as it is thence to its basal tip. Furthermore, the ridges are well defined and most of them are finely serrated. The basal margins of the lateral walls of the hydroecium are heavily serrated, as is the basal portion of the laterals. Also, in contrast to bicarinata the arc formed by the lateral margins of the basal facet, as seen in a lateral view, has a marked curvature rather like that in A. trigona. The apicodorsal facet appears to be distinctly shorter than in bicarinata. This may possibly be due to the fact that in ventral view the vertical distance between the transverse apical ridge and the insertion of the horizontal ridges and the ventral facet is proportionately shorter than in bicarinata. Finally, the apical portion of the apicoventral facet is slightly furrowed, but lacks any indication of a depression just ventral to the transverse apical ridge. Inferior nectopliore. Unknown. Eudoxid. Unknown. Abyla tottoni^ n. sp. Ahyla trUjona Moser, 1925, pp. 301-310, text figs. 42-47, pi. 16, figs. 6-7, pi. 18, fig. 7 (partim). The type specimen is a colony from "Dana" Sta. 3994^ at 15°45'S, 5°45'\V taken on 24 February 1930 at 1930 hours with a stramin net 200 cm. in diameter. It will be deposited in the Universitets Zool- giske Museum, K0benhavn, Denmark. 1 Named in honor of Captain A. K. Totton of the British Museum (N. H.) in appreciation for his many kindnesses in furthering my understanding of this genus. 48 bulletin: museum of comparative zoology Superior nectophore (Figs. 12B, 13B, 14B). The superior necto- phore of Ahyla tottoni is tumid, especially the apical portion. The transverse apical ridge is not distinct, but looks rather like a rounded chamfer or rolled joint. In large specimens this may also be true of the apicolateral and the apicodorsal ridges. The exact location of the ridges is, however, delimited by fine hair-like lines which may be seen with proper lighting. The horizontal ridge may be indistinct but it can always be traced by a series of punctate elevations. Likewise, in some of the larger specimens the projection formed at the junction of the lateral and horizontal ridges is a definite knob rather than angular, as it usually is in this and other species. These knobs might be caused by poor preservation, although it is seldom possible to detect any damage on such specimens. In any event, the lateral protrusions are more marked than in any species other than those with expanded lateral ridges. Thus, the nectophore is nearly (0.86) as wide as it is high.' Two other characters together with those just mentioned serve to distinguish it fi-om all other species even when detached from the inferior nectophore. At the insertion of the horizontal ridge, the ventral facet is as wide (1.09) as from that point to its basal tip. Finally, one of the most consistent features is the extremely deep de- pression just ventral to the transverse apical ridge. Inferior nectophore (Fig. 15B). The inferior nectophore of A. tottoni is usually about three-quarters as wide as it is long (exclusive of the apophysis). The dorsal ridge is so expanded toward the apex that it curves sharply before merging with the dorsal tooth. Likewise, the right ventral wing is a somewhat lopsided semicircle. The outer row of teeth (8-12) on its basal margin is continuous with serrations of its ventral margin. The inner row parallels the outer close to the ventral teeth, but gradually diverges to end on the inner surface of the wing well in from the ventral margin very much as it does in bicarinata. These teeth are coarser and may be slightly more numerous than those on the outer row. The characters enumerated above, together with the row of eight or nine teeth on the comb of the left ventral ridge make it possible to distinguish the inferior nectophore of tottoni without any difficulty. Eudoxid. Unknown. 1 Poorly preserved specimens, however, have been seen in whioh these protrusions were so damaged that thi« character cannot always be relied upon in differentiating this species. sears: notes on siphonophores 49 Pseud ABYLA n. gen. Genotype: Pseudabyla irregularis n. sp. Generic Characters Superior nedophore (Figs. 16 and 17). The genus Pseudabyla is established for three damaged superior nectophores. These resemble specimens of Abyla and might be referred to that genus if not examined carefully. They have a number of characters in common, a transverse apical ridge and horizontal ridges as well as the same internal arrange- ment and configuration of the basal region. Closer examination, how- ever, reveals that there are fewer facets and that the nectophore is asymmetrical due to the absence of a facet or ridge and perhaps to a shift in the position of one ridge or another. Inferior nectophore. Unknown. Eudoxid. Unknown. Remarks The description given for Pseudabyla is general to include speci- mens of what appear to be two species closely related to Abyla, but until more is known of these it appears premature to go into greater detail. At present, the known variants of Abyla are included here to emphasize their existence. In making this genus, as well as three to be described later {Pseudocymba, Abylopsoides, Pseudabylopsis), so inclusive, the usual practise of differentiating the genera of the Abyli- nae by the number and arrangement of the facets and ridges of the superior nectophore has admittedly been disregarded. Such a course seems warranted, in view of the fact that only two or three specimens are known in each genus, at least until such time as a sufficient number of specimens become available either to substantiate or to disprove such action. One more or less parallel case among the Diphyinae also makes it somewhat justifiable. In that gi'oup, most genera have five ridges on the superior nectophore, but lyi one genus, Lcnsia, the num- ber varies (Totton, 1941). This variation is one of the characters which is helpful in distinguishing the individual species. Totton (1941) had a suflRcient number of specimens to prove the existence of good species in Lensia. It is possible then that much the same situation 6 6 PJISirr Ml. --x^w llill/ -■*— \ 1 -■ r J 4 1 ]| t- :^V k 'Wv|// Fig. 16. Superior nectophore of Pseudabyla irregularis dorsal facet of about 3 mm. in length. A. Cross section. B. Apical view. C. Left lateral view. D. Ventral view. E. Right lateral view. F. Dorsal view. «.<£. ^Asier ab; Fig. 17. Superior nectophore of Pseudabyla dubia with a dorsal facet of nearly 2 mm. A. Cross section. B. Apical view. C. Right lateral view. D. Ventral view. E. Left lateral view. F. Dorsal view. 52 bulletin: museum of comparative zoology exists among the Abylinae, but as yet we have an insufficient number of specimens to feel certain. Another problem to be solved in the Abylinae, which does not arise in the Diphyinae, is the reason for the asymmetry in some of these variants. The chief difference between Ahyla and Ccratocyviba is the absence of the transverse apical ridge and the two horizontal ridges with a consequent reduction in the number of facets, but otherwise little change in their arrangement. The actual loss of a facet, however, seems to result in an asymmetrical arrangement. This has been found to be the case not only in this genus, but also in Pseudocymha and Ahylopsoides. So far, we have only two bits of evidence to suggest that the asymmetry is a specific character. Two specimens of Pseud- abyla irregularis have been found at widely separated localities and yet, despite the poorly preserved condition they were found in, it appears that the two are asymmetrical in exactly the same way. A second suggestion that this may, indeed, be true is that both species of Pseudocymha have triangular basal facets. Key to the Species of Pseudahyla A. Superior nectophore 1. Ventral ridge present Pseudahyla irregularis n. sp. Ventral facet present Pseudahyla duhia n. ep. Pseudabyla irregularis n. sp. The type specimen of Pseudabyla irregularis is a superior nectophore taken at "Dana" Station 3919iv, at 0°07'S, 63°56'E on 8 December 1929 at 1910 hours in a stramin net, 200 cm. in diameter, towing with 100 meters of wire out. A second superior nectophore was taken at "Dana" Station 3964^ at 25°19'S, 36°13'E on 15 January 1930 at 2030 hours in a stramin net 150 cm. in diameter, towing with 50 meters of wire out. The specimens will be deposited in Universitets Zoologiske Museum, K0benhavn. Superior nectophore (Fig. 16). The two specimens appear to be identical, insofar as can be determined in their damaged condition. The asymmetry which might be presumed to have resulted from poor preservation is actually caused by structural peculiarities as men- tioned below. The most obvious characteristic of this species of Pseud- abyla is the presence of a ventral ridge rather than a facet, but a sears: notes on siphonophores 53 more critical examination reveals that the left horizontal ridge crosses the ventrolateral surface diagonally and joins the ventral ridge at its basal end. The right horizontal ridge is in the position usual in Abyla. However, the apicolateral ridge on this side which in Abyla separates the apicoventral and apical ventrolateral facet is apparently missing. Consequently, the transverse apical ridge is diagonal and one of the right lateral ridges bordering the apicodorsal facet has seemingly dis- appeared. Thus, this facet is pentagonal rather than hexagonal as in Abyla. Inferior nectophore. Unknown. Eudoxid. Unknown. Remarks Since the two specimens taken at widely separated localities appear identical, even in their asymmetry, irregidaris would seem to be a valid species which may previously have been overlooked because it appears so similar to Abyla unless carefully examined. PSEUDABYLA DUBIA H. sp. The type specimen of Pseudabyla dubia is a superior nectophore taken at"^"Dana" Sta. 3921iii at 3°36'S, 58°19'E on 11 December 1929 at 1900 hom-s in a stramin net 200 cm. in diameter, towing with 300 meters of wire out. The specimen when found was damaged and, most regrettably, was dried up after the sketches (Fig. 17) were made. The specimen has, however, been saved and will be placed in Universitets Zoologiske Museum, K0benhavn. Superior nectopliore (Fig. 17). This species differs from the previous one in that the dorsal facet is replaced by a ridge. Also, a ventral facet is present. On the other hand, this is not symmetrical as in Abyla: the apical part being twisted to the left and the point of junction with the horizontal ridges is slightly higher on the right than on the left. In addition, the upright vertical ridge separating the dorsal facet from the right dorsolateral is seemingly missing. Strangely enough, the tooth at the end of the left upright (or dorsal ridge) is smaller and more irregular than the right. The apicodorsal facet is pentagonal, ap- parently due to the disappearance of the ridge which is present in Abyla and which separates the dorsal and apicodorsal facets in that genus. Finally, as in irregularis the transverse apical ridge is skewed. 54 bulletin: museum of comparative zoology Remarks Despite the poor condition the specimen is now in, because it was accidentally dried, these characters can still be observed and hence, it appears desirable to call attention to this peculiar variant of Abyla. Ceratocymba Chun 1888 Genotype: Ceratocymba sagittata Quoy & Gaimard 1827 Generic Characters Superior nedophore (Figs. 18A, 19, 22). In general appearance the superior nectophores of the several species of Ceratocymba form a graduated series (Fig. 1), varying from an almost rectangular (/«