UC-NRLF C E 773 355 BIOLOGY LIBRARY G THE OXYCEPHALIDS BY CARL BOVALLIUS. WITH EIGHTY-SEVEN ILLUSTRATIONS IN THE TEXT AND SEVEN PLATES. (PRESENTED TO THE ROYAL SOCIETY OF SCIENCES OF UPSALA THE 27*h OF APRIL, 1889.) UPSALA. PRINTED Bi" EDV. BERLING. 1890. BIOLOGY RA G CONTESTS. Introduction p. 1. I. Historical notes on the Oxycephalids. Bibliography 2. II. The systematical position of the Oxycephalids. Characteristics for the distinction of the subtribes of the Am- phipoda Hyperiidea 14. Characteristics for the distinction of the families of the Hype- riidea curvicornia 15. Characteristics for the distinction of the genera and species among the Oxycephalids 16. Synoptical table of the families of the Amphipoda Hyperiidea . 17. III. Morphological notes on the Oxycephalids. 1 . The shape of the body 20. 2. The integument 21. 3. The head 22. 4. The eyes 23. 5. The first pair of antennas 24. 6. The second pair of antennas 26. 7. The mouth-organs 28. 8. The perseon and its appendages 30. 9. The pleon and its appendages 36. 10. The urus and its appendages 37. 11. The nervous system 41. 12. The glands 41. 13. The young > 43. IT. The Oxycephalidean genera and species, The family Oxycephalidas _ 45. Synoptical table of the genera 46. 1. Simorhynchotus 47. S. antennarius 49. S. Stebbingi 50. S. Lilljeborgi 52. 2. Oxycephalus 54. O. piscator ... 56. O. Clausi .... 60. 43776G O. tuberculatus , t p. 62. O. pectinatus 64. O. latirostris 66. 3. Tullbergella 68. T. cuspidata 69. 4. Calamorhynchus 72. C. pellucidus . 73. C. rigidus 74. 5. Dorycephalus 75. D. Lindstroemi 76. D. ambobus 79. 6. Streetsia 80. S. Challenger! 82. S. porcellus 83. S. pronoi'des , 84. S. carinata t 87. S. Steenstrupi 89. S. longiceps 92. S. (?) mindanaonis 93. 7. Cranocephalus 94. C. Goesi 95. 8. Stebbingella > 97. S. sclerotica 93. S. typho'ides 100. S. ThSeli 101. 9. Glossoc ephalus 105. G. Milne-Edwardsi 106. G. spiniger jQS. 10. Leptocotis no. L. spinifera m L. teuuirostris 113 The family Xiphocephalidse 114. Xiphocephalus \\(j X. armatus 119. X."AVhitei 125. X. Lilljeborgi 131, X. brevicaudatus 133. Alphabetical Index 135 List of illustrations in the text 133 Plates. THE OXYCEPHALIDS. _L he Oxycephalids are perhaps the most interesting of all the Am- phipoda Hyperiidea on account of their strange external forms as well as the curious transformations which several of their internal and extern- al organs are subjected to. However, it is not only because they pre- sent such remarkable features as these that I have been induced to take them up now. The many intricate questions connected with the syno- nymy of the genera and species have to-day acquired an immediate inter- est through the publication of two great works both of the highest im- portance for our knowledge of the Amphipods, Die Platysceliden)) by C. GLAUS in 1887, and Report on the Amphipoda collected by H. M. S. Challenger during the years 1873 1876, by the Rev. THOMAS R. R. STEBBING in 1888. As my own opinions with regard to the synonymy and relationship of many of the Oxycephalidean species, briefly stated in a preliminary paper ^Systematical list of the Amphipoda Hyperiidea)), differ on several points from those of the eminent carci- nologists just mentioned, I have felt in duty bound to state more fully the reasons on which my systematical arrangement of this abnormal group of animals is founded. Besides, I wish to make use of this op- portunity for communicating some new results arrived at by the exami- nation of fresh material recently obtained. As an introduction I am going to give some historical and mor- phological notes on the Hyperids in question. Nova Acta Reg. Soc. Sc. Ups. Ser. 1 CARL BOVALLIUS, THE OXYCEPHALIDS. I. HISTORICAL 1SOTES 01N THE OXYCEPHALIDS, The Oxycephalids seem to have been comparatively late attended to by the zoologists, the first who described such an animal was H. MILNE EDWARDS, 1, the great reformer of the carcinological system, giving in 1830 the diagnosis of the genus Oxycephalus. He placed the new genus among the Hyperines anormales)), but said Ces amphipodes s'eloigneut de la plupart des Hyperines par la forme grele et allonge"e, par leur tete aplatie et lanceolee, etc.)) The species he named Oxy- cephalus plscatoris. Two years later ANASTASIO Cocco, 2 and 2 a, established the new ge- nus Orio, which partly is synonymous with Oxycephalus, his species Orio Ornithoramphus, however, is very difficult to identify, owing to the incom- plete diagnosis ; a facsimile of his drawing is given by STEBBING 1. c. p. 145 ; as far as I know it is the first drawing of an Oxycephalid, met with in the literature. NICOLO PRESTANDREA, 3, described in 1833 a new species of the just mentioned genus, Orio oxyahingus; it is very unclear where this species might be placed, in the systematical arrangement below it is omitted as being deficiently described; STEBBING 1. c. p. 152 assumes that Oxyahingus must be a misprint for Oxyrhingus, and this probably a malformation for Oxyrhynchus. In 1836 F. E. GUERIN, 4, instituted the BIBLIOGRAPHY. 1. H. MILNE EDWARDS Extrait de Recherches pour servir & 1'Historie naturelle des Crustace's amphipodes*. Annales des Sciences naturelles. Tome 20 me , p. 396. Paris, 1830. 2. ANASTASIO Cocco. Su di alcuni nuovi crustacei de' mari di Messina Lettera del dott. A. C. al celebre dott. WILLIAM ELFORD LEACH, etc.* Effemeridi scientifiche e letterarie per la Sicilia. Tomo 2 do , Pa- lermo, 1832. 2 a see p. 10. 3. NICOLO PRESTANDREA. Su di alcuni nuovi crustacei dei mari di Messina. Effemeridi scientifiche e letterarie per la Sicilia. Tomo 6 to , Pa- lermo, 1833. 4. F. E. GUERIN. Description de quelques genres nouveaux de Crustace's appar- tenant a la famille des Hyperines. Magasin de Zoologie; 6 me Anne"e. 1836. Classe VII, p. 10, pi. 18, fig. 2. Paris, 1836. HISTORICAL NOTES. 3 new species Oxycephalus oceanicus, which he himself later, 8, declared to be synonymous with the typical species of H. MILNE EDWARDS. In 1839, 6, he gave a new drawing of it. In 1838 H. MILNE EDWARDS, 5, quoted the two species Oxycephalus $>piscatoriw and 0. oceanicus; and in 1840, 7, he gave a drawing and a fuller description of the type species, changing the name into Oxycepha- lus pise a tor; at the same time he gave a diagnosis of a new species. 0. armatus, which is the first representative for the genus Rhabdosoma or Xiphocephalus In the same year, or in 1841, EYDOUX and SOULEYET, 8, gave a long description of Oxycephalus armatus, H. MILNE EDWARDS, accom- panied by excellent drawings ; it is, however, not the true 0. armatus, but identical with Rhabdosoma Whitei, established in 1862 by SPENCE BATE, as will be proved below under this species. After the description of the animal they say 1. c. p. 271 that F. E. GUERIN, who had aided them with the examination, regarded the species generically distinct from Oxy- cephalus, and established the genus Xyphocephalus for its reception. Thus Xyphocephalus or Xiphocephalus is the eldest name for the genus and must be accepted instead of the later names'. In 1848 ARTHUR ADAMS and ADAM WHITE, 9, instituted the genus Rhabdosoma for the same Oxycephalus armatus, not knowing the just men- tioned work; they gave a short generic diagnosis and said that Rhabdo- soma armatum forms a singularly interesting link between the Amphipoda and the Lasmodipoda, uniting, as it were, the two. They gave also a tolerably 5. J. B. P. A. DE LAMARCK. Historie uaturelle des Aniinaux sans vertebres. 2 me Edition par G. P. DESHAYES et H. MILNE EDWARDS. Tome 5 me , p. 308. Paris, 1838. 6. F. E. GUERIN. Typhis. Dictionnaire pittoresque d'Historie naturelle et des phe*nomenes de la nature. Sous la direction de F. E. GUERIN. Tome 9 me , p. 498, pi. 708, fig. 4 and 4 a, Paris, 1839. 7. H. MILNE EDWARDS. Historie naturelle des Crustace's. Tome 3 me , p. 99 102, pi. 30, fig. 1012. Paris, 1840. 8. EYDOUX et SOULEYET (and F. E. GUERIN). Voyage autour du monde execute' pendant les anne"es 1836 et 1837 sur la corvette la Bon it e. Zoo- logie. Tome l er , p. 267271, pi. 4 fig. 1332. Paris, (1841?). 9. ABTHUR ADAMS and ADAM WHITE. Crustacea. The Zoology of the voyage af H. M. S. Samarang, under the command of Captain Sir ED- WARD BELCHER, p. 63, pi. 13, fig. 7. London, 1848. 4 CARL BOVALLIUS, THE OXYCEPHALIDS. good drawing, which clearly shows that the specimen described by these naturalists is identical with the type species of H. MILNE EDWARDS. In 1849 HERCULE NICOLET, 10, mentioned Oxycephalus oceanicus from the seas of Chile, and H. LUCAS, 11, the genus Oxycephalus among the Hype'rines anormales)). In 1850 GUISEPPE NATALE, 12, promoted the specific name Ornitho- ramphus, given by Cooco, see above p. 2, to a generic name, and cal- led the species 0. Coccoi. He described this species at length but with- out good characteristics; if it is identical with Cocco's species or not, is very difficult to settle. STEBBING 1. c. p. 241 suggests, that it possib- ly may have some relation to Oxycephalus typhoides, GLAUS; according to my opinion it is rather near to Oxycephalus latiros tr is, GLAUS, ho wbeit, the genus is most likely a synonym for Oxycephalus, H. MILNE EDWARDS. NATALE was doubtful if Ornithoramphus might be rather an Isopod, allied to Sphceromidce, than an Amphipod. In the same year the same author, 13, pu- blished in an other paper one more new generic name, Erpetoramphus, also probably synonymous with Oxycephalus. Owing to the vague description it is impossible to identify the species which he named E. Costce. In 1851 H. LUCAS, 14, mentioned and figured Oxycephalus oceanicus, GUERIN, not adding any new characteristics. In the same year Fr. GUGL. HOPE, 15, issued a catalogue of the crustaceans of the Mediterranean, he enumerates without any charcteri- stics, under Amphipoda, the following, above mentioned Oxycephalids viz; Orio oxyrhynchus, PRESTANDREA, 0. ornithoramphus, Cocco, Ornithoramjjhus Coccoi, NATALE, and Erpetoramphus Costce, NATALE. 10. HERCTILE NICOLET. Historia fisica y politica de Chile . . . por CLAUDIO GAY. Zoologia. Tomo 3 r , p. 249. Paris, 1849. 11. H LUCAS. Hype"rines. Dictionuaire universel d'Historie naturelle .... dings' par CHARLES D'ORBIGNY. Tome 6 me p. 782. Paris, 1849. 12. GUISEPPE NATALE. Descrizione zoologica d'una nuova specie di plojaria e di alcuni crostacei del porto cli Messina, etc", p. Messina, 1850. 13. Su pochi Crostacei del porto di Messina. Lettera del Dottor G. de N. . . al Sg. ACHILLE COSTA, con una tavola in rame. Na- poli, 1850 (teste STEBBING, 1. c. p. 16211624). 14. H. LUCAS. Historic naturelle cles Crustaces, des Arachnides et des Myria- podes, p. 240, pi. 18, fig. 9. Paris, 1851. 15. Fr. GUGL. HOPE. Catalogo dei Crostacei Italian! e di molti altri del Medi- terraneo, p. 21 and 22. Napoli, 1851. HISTORICAL NOTES. 5 J. D. DANA, 1617, in 1852 placed the Oxycephalids as a sub- family Oxycephalince in the family Typhidce. In 1858 C. SPENCE BATE, 18, described the new genus Macro- cephalus, with M. longirostris, n. sp., later, 20, he himself dropped both names, finding them synonymous with Ehabdosoma, ADAMS and WHITE, and Eh. armatum, H. MILNE EDWARDS; here below it is quoted as synonym for Xiphocephalus armatus, H. MILNE EDWARDS. In 1861 the same author, 19, gave some interesting notes on the young of some Hyperids allied to Oxycephalus; the following year he, 20, established, in his catalogue of the Amphipoda in the- British Museum, the family Oxycephalidae, with two subfamilies, Synopince and Oxyce- phalmw, he described there Oxycephalus piscator, H. MILNE EDWARDS, 0. oceanicus, F. E. GUERIN, and 0. tuberculatus, n. sp. this latter was described from a mutilated specimen but with characteristics sufficient to show that it is a species of its own, not synonymous with any one of the previously known species; further Ehabdosoma armatum, H. MILNE EDWARDS, and Eh. Whitei, n. sp. of the former species he gave a new, enlarged description, adding some good characteristics to those mentio- ned by H. MILNE EDWARDS; of the latter species he gave a detailed de- scription showing beyond doubt that it was a separate species. Unfor- tunately he did not know both sexes of the two species, but only the female of the former and the male of the latter. C. GLAUS was proba- bly misled by this fact to interprete Eh. Whitei, SPENCE BATE, as the male form of Eh. armatum, H. MILNE EDWARDS. SPENCE BATE gave also drawings of all these species. 16. J. D. DANA. On the Classification of the Crustacea Choristopoda or Tetra- decapoda. The American Journal of Science and Arts. 2 d Se- ries. Vol. 14, p. 316. New Haven, 1852. 17. United States Exploring Expedition during the years 1838, 1839, 1840, 1841, 1842, under the command of CHARLES WILKES. U. S. N.; Crustacea. Vol. 2, p. 1009 and 1443. Philadelphia, 1852. 18. C. SPENCE BATE. On some new Genera and Species of Crustacea amphi- poda. Ann. and Mag. of Nat. Hist. 3 d Ser. Vol. 1, p. 361. London, 1858. 19. On the Morphology of some Amphipoda of the Division Hyperina. Ann. and Mag. of Nat. Hist. 3 d Ser. Vol. 8, p. 1 15. London, 1861. 20. Catalogue of the specimens of Amphipodous Crustacea in th e collection of the British Museum, p. 341 346. London, 1862. 6 CARL BOVALLIUS, THE OXYCEPHALIDS. In 1864 ACHILLE COSTA, 21, without referring to the genus Oxyce- phalus or to the family Oxycephalidae, instituted a kind of subfamily, Ornithoramphini, for the reception of the genus Ornithoramphus, NATALE, and the two new genera Natalius and Carcinornis. On the synonymy of the first genus see above p. 3, the genus Natalius is probably syno- nymous with Oxycephalus, and the new species N. candidissimus, pos- sibly is 0. piscator, H. MILNE EDWARDS; the genus Carcinornis is diffi- cult to identify, it comes perhaps near to Streetsia, STEERING, orLep- tocotis, STREETS, because the form of the first two pairs of pera3Opoda alluded to in the diagnosis is most similar to the form of those pairs in the two mentioned genera; of this latter genus Carcinornis he established two new species C. acutirostris, and C. inftaticeps; the suggestion of STEERING 1. c. p. 347 that the latter may prove to be synonymous to Oxycephalus typhoides, GLAUS, is, according to my opinion, not supported by the few characteristics given by COSTA; see below under Stre etsia. In 1867 COSTA, 22, mentioned Ornithoramphus Coccoi, NATALE, among a collection of Mediterranean crustaceans sent to the exhibition in Paris that year. In 1871 C. CLAUS, 23, gave some short notices on the anatomy of the Oxycephalids, mentioning a vesicular organ in the hind part of the head; this vesicle he regarded as an auditory organ; he described further the structure of the eyes, the nervous system, and the organs of circula- tion. At the same time he gave a new diagnosis of the genus Oxyce- phalus, and described as 0. piscator, H. MILNE EDWARDS, the animal which here below is named 0. Clausi, he argued further that 0. oceanicus, GUERIN, is synonymous with 0. piscator, H. MILNE EDWARDS, a fact which GUERIN-MENEVILLE already in 1841, himself had proved, 8, and that 21. ACHILLE COSTA. Di alcuni crostacei degli acalefi, e di un distomideo para- sitto. Rendiconto delPAccademia delle Scienze Fisiche e Mate- matiche. Anno 3 ro . Fas 4 to , p. 8689. Napoli, 1864. 22. ACHILLE COSTA. Saggio della collezione de 1 Crostacei del Mediterraneo Del Museo Zoologico della University di Napoli spedito alia Espo- sizione di Parigi del 1867. Annuario del Museo Zoologico della R. University di Napoli. Anno 4 to , p. 44. Napoli, 1867. 28. C. CLAUS. Untersuchungen fiber defl Ban mid die Verwandschaft der Hype- riden. Nachrichten von der K. Gesellschaft der Wissenschaften und der Georg-Augusts-Universitat aus dem Jahre 1871, p. 149 157. Gottingen, 1871. HISTORICAL NOTES. 7 Rhabsosoma Whitei, SPENCE, BATE, is not a species by itself but only the male form of Eh. armatum, H. MILNE EDWARDS, a mistake, which will be duly cleared up under ))Xiphocephalus below. He established also the new genus Simorhynchus, with the new species S. antennarius, and a new species of Occycephalus, 0. tenuirostris, mentioned below as Leptocotis tenuirostris. In the same paper he said, Die Gattung Synopia, DANA, die DANA and SPENCE BATE irrthiimlich mit den Oxycephaliden vereiuigt haben, gehort zu den Gammariden)). Among the Oxycephalidae he mentioned then Snehagenia, n. g. later transferred to the Lycseids. In 1877 THOMAS H. STREETS, 24, described under the name Oxy- cephalus tuberculatus, SPENCE BATE an animal which I really think may prove to be identical with SPENCE BATE'S species ; he further instituted the new genus Leptocotis with the new species L. spin if era. In 1878 GLAUS, 25, gave some interesting notices about the heart and vessels of Oxycephalidee and other Hyperids. In the same year THOMAS H. STREETS, 26, described from the Pa- cific the following Oxycephalids viz; Oxycephalus tuberculatus, SPENCE BATE, which, however, is not identical with SPENCE BATE'S species but with 0. Clausi, C. BOVALLIUS; 0. bulbosus, n. sp., 0. scleroticus, n. sp. Lepto- cotis spinifera, STREETS, Calamorhynchus, n. g. with C. pellucidus, n. sp., Rhabdosoma Whitei, SPENCE BATE, Rh. armatum, H. MILNE EDWARDS, and Rh. longirostris, SPENCE BATE, the latter synonymous withXipho- cephalus armatus, H. MILNE EDWARDS. It is noticeable that he did not himself find the specific difference between the Oxycephalus tuber- culatus, described by him in 1877 and the 0. tuberculatus, mentioned in this his latter paper. The descriptions are accompanied with drawings which are not very good, being executed in a small scale. 24. ThoMAS H. STREETS. Contributions to the Natural History of the Hawaiian and Fanning Islands and Lower California*. Bulletin of the Uni- ted States National Museum, 1877. N 7, p. 136138. Was- hington, 1877. 25. C. GLAUS. Uber Herz und Gefass-system der Hyperiden. Zoologischer An- zeiger, herausgegeben von J. V. CARUS. l ster Jahrgang, 1878, p. 269271. Leipzig, 1878. 26. THOMAS H. STREETS. Pelagic Amphipoda. Proceedings of the Academy of Natural Sciences of Philadelphia. 1878, p. 276290, pi. 2. Philadelphia, 1878. 8 CARL BOVALLIUS, THE OXYCEPHALIDS. In 1879 GLAUS, 27, made a revision of the known Platyscelids, and quoted among them the following genera and species as belonging to the family Oxycephalidge. Oxycephalus piscator^ H. MILNE EDWARDS, as observed above, p. 6, identical with 0. Clausi, C. BOVALLIUS, 0. similis, in my opinion probably a sjmonym for 0. piscator, H. MILNE EDWARDS, 0. latirostris, n. sp., 0. tenuirostris, GLAUS, quoted as Leptocotis tenuirostris below, 0. porcellus, n. sp., = Streetsia porcellus, 0. longiceps n. sp. = Streetsia longiceps, 0. typhoides n. sp. = Steb- bingella typhoides. He adhered to his former statement that Rhab- dosorna Whitei, SPENCE BATE, was synonymous with Rh. armatum, H. MILNE EDWARDS. He transferred the genus Simorhynckus to the family ^Lycosidce^. In 1880 GLAUS, 28, placed Oxycephalince as the fifth subfamily of the fourth Hyperiidean family Platyscelidce, quoting the same genera as in the just cited work. In the same year OTMAR NEBESKI, 29, added important facts to our knowledge of the anatomy of the Amphipods;he did not investigate any one of the Oxycephalids but many of his statements throw light upon analogue peculiarities in this family. Just the same ought to be said about YVES DELAGE, 30, and his -admirable treatise on the organs of circulation of the marine Edriophthalmata. In 1884 GEO. M. THOMSON, 31, described a new species of Oxy- cephalus, 0. Edwardsii, which, however, probably is synonymous with 0. piscator, H. MILNE EDWARDS, see below under this species. 27. C. GLAUS. Die Gattungen und Arten der Platysceliden in systematischer Uber- sicht. Arbeiten aus dem Zoologischen Institute der Universitat Wien und der Zoologischen Station in Triest. Herausgegeben von D:r C. CLAUS. Tom. 2, p. 188198. Wien, 1879. 28. Grundzuge der Zoologie. 4 te Auflage. Vol. 1, p. 588. Marburg, 1880. 29. OTMAR NEBSEKI. Beitrage zur Kentniss der Amphipoden von Adria. Arb. Zool. Inst. der Universitat Wien etc. Toni. 3, p. 111 162. With 4 plates. Wien, 1880. 30. YVES DELAGE. Contribution a 1'^tude de 1'appareil circulatoire des crustace's e"driophthalmes marins. Archives de la Zoologie Exp erimentale et Ge-nerale. Vol. 9, p. 1172, pi. 12. Paris, 1881. 31. GEO. M. THOMSON. Descriptions of new Crustaceans. Transactions and Proceedings of the New Zealand Institute, 1883. Vol. 16, p. 237240, pi. 12, fig. 1421, pi. 13, fig. 1. Wellington, 1884. HISTORICAL NOTES. 9 In the same year J. S. KINGLEY, 32, gave a drawing of Rhabdosoma Batei, n. sp., which name however is only a synonym for Xiphocepha- lus armatus, H. MILNE EDWARDS. In 1885 J. V. CARUS, 33, recorded the Oxycephalids hitherto men- tioned from the Mediterranean viz; Oxycephalus similis, GLAUS, doubtfully referring to Natalius candidissimus, COSTA, as synonymous with it; he further mentioned 0. typhoides, GLAUS; and enumerated also among the Oxy- cephalidae, but with a sign of interrogation, Carcinornis acutirostris, COSTA, C. inflaticeps, COSTA, Orio Zandeus, Cocco, Chiropristis messanensis, Cocco, and Ornithoramphus Coccoi, NATALE. Of these animals Orio Zancleus is an Eupronoe, GLAUS, and Chiropristis an Anchylomera, H. MILNE EDWARDS; for the probable synonymy of the remaining ones see, above p. 4 and 6. Iii 1886 A. GERSTAECKER, 34, gave Oxycephalidas as the fifth family of the Hyperina anormalia, including the two genera Oxyce- phalus and Rhabdosoma and supported the view of GLAUS, that Rh. ar- matum and Rh. Whitei were the same species. He gave also a copy of GUERIN'S drawing of Oxycephalus oceanicus. In 1887 the author of this treatise, 35, published a preliminary pa- per on the systematical arrangement of the Amphipoda Hyperiidea, enu- merating and describing briefly the following Oxycephalids, Gloss o- cephalus, n. g. with the species G. Milne Edwardsi, n. sp., and G. spiniger, n. sp.; Oxycephalus piscator, H. MILNE EDWARDS, 0. Clausi, n. sp. 0. tuberculatus, SPENCE BATE, 0. pectinatus, n. sp., 0. latirostris, GLAUS, 0- porcellus, GLAUS, O.pronoides, n. sp., 0. Steenstrupi, n. sp., and 0. longiceps, GLAUS, the last four species being recorded in the 32. JOHN STERLING KINGSLEY. The Standard Natural History. Vol. 2; Crusta- cea and Insects. Boston, 1884. 33. JULIUS VICTOR CARUS. Prodromus Faunse Mediterrauese sive Descriptio Ani- malium Mediterranei incolarum, quam comparata silva rerum qua- tenus innotuit, adiectis locis et nominibus vujgaribus eorumque auctoribus in comtnodum Zoologorum congessit J. V. C. Vol. 1, p. 427428. Stuttgart, 1885. 34. A. GERSTAECKER. D:r H. G. BRONN'S Klassen und Ordnungen des Thier-Reichs, wissenschaftlich dargestellt in Wort und Bild. 5 ter Band. 2 te Ab- theiluug. Gliederfiissler, Arthropoda, p. 486 487, pi. 25, fig. 6 and 6 a. Heidelberg, 18841886. 35. C. BOVALLIUS. Systematical list of the Amphipoda Hyperiidea. Bihang till Kongl. Svenska Vetenskaps-Akademiens Handliugar. Band 11. N"> 16, p. 3440. Stockholm, 1887. Nova Acta Reg. Soc. Sc. Ups. Ser. III. 2 10 CARL BOVALLIUS, THE OXYCEPHALIDS. present treatise as belonging to the genus Streetsia, STEBBING; Oxyce- phalus typhoides, GLAUS, and 0. scleroticus, STREETS, both here transferred to the genus Stebbingella, n. g. ; Leptocotis Lindstromi, n. sp., here named Dorycephalus Lindstromi; Leptocotis tenuirostris, GLAUS; Tullbergella, n. g., with the new species T. cuspidata; Calamorhyn- chus pellucidus, STREETS; Rhabdonectes, n. n. instead of Rhabdosoma, which name the" writer supposed to be preoccupied, but wrongly, as STEBBING (1. c. p. 1606) has clearly proved. Diagnoses were also given of Rh. armatus, H. MILNE EDWARDS, and Rh. Whitei, SPENCE BATE. In 1887 the important work of GLAUS, 36, Die Platysceliden)) was issued. He gave there good drawings of all the species described in his previous paper of 1879 (see p. 8). He maintained his opinions about Oxycephalus piscator, H. MILNE EDWARDS, and Rhabdosnma armatum, H. MILNE EDWARDS, spoken of above (p. 7 and 8). The descriptive part is a reprint of his paper of 1879, with some few additions. In the same year G. M. GILES, 37, described a new species of Rhab- dosoma, Rh. investigatoris, which is however undoubtedly synonymous with Xiphocephalus Whitei, SPENCE BATE; GILES expressed the opinion that GLAUS was quite right in considering Rkabdosoma Whitei, SPENCE BATE, as the male of Rh. armatum, H. MILNE EDWARDS. In December 1887 I gave, 38, a diagram of the systematical rela- tions of the Hyperiidean families, in which the Oxycephalidae are recorded as the thirteenth family and placed between the Tryphcenidce and the Pronoidce. 36. C. CLAUS. Die Platysceliden, p. 6775, pi. 2226. Wien, 1887. Fol. 37. G. M. GILES. On six new Amphipods from the Bay of Bengal*. Journal of the Asiatic Society of Bengal. Vol. 56, p. 219. Calcutta, 1887. 38. C. BOVALLIUS. Contributions to a Monograph of the Amphipoda Hyperi- idea. Part I. 1. Kongl. Svenska Vetenskaps-Akademiens Hand- lingar. Band 21. N 5, p. 3. Stockholm, 1887. 39. Th. R. R. STEBBING. On the Amphipoda collected by H. M. S. Challenger during the Years 1873 1876 Report on the scientific results of the Voyage of H. M. S. Challenger during the years 18731876 .... Zoology. Vol. 29. 2 a. ANASTASIO Cocco. Descrizione cli alcuui Crustacei di Messina. Giornale di Scienze, Lettere e Arte per la Sicilia. Tomo 44 to - Anno XI. Palermo 1833. HISTORICAL NOTES. 11 Lastly I have to mention the splendid work of the Rev. THOMAS R. R. STEBBING, 39, on the ))Challenger Amphipoda, published at the end of last year, in the bibliographical part of which the accomplished author gives us many valuable hints concerning the synonymy of the Oxyce- phalidean genera and species, and in the second half of the descrip- tive part of wich he most accurately sets forth the characteristics of the species obtained during the celebrated ))Challenger Expedition. The descriptions are accompanied by magnificent plates. He describes the following genera and species, Oxycephalus Clan si, C. BOVALLIUS, Oxycephalus porcellus, GLAUS, 0. lonyiceps, GLAUS, the last two below being named Streetsia porcellus, GLAUS, and S. longiceps, GLAUS; Leptocotis ambobus, n. sp., synonymous with Dorycephalus Lindstro mi, C. BOVALLIUS, as will be proved below; Leptocotis mindanaonis, n. sp.; Calamorhynchus rigidus, n. sp.; Streetsia n. g., with the new species S. challenge ri; Rhabdosoma armatum, H. MILNE EDWARDS, of which he with some doubt quotes Rh. Whitei as a synonym, and lastly the new species Rh. brevicaudatum. The name Simorhynchus, GLAUS, is changed into Simorhynchotus, n. n., and he places this genus in the family Lycceidce. In this treatise I propose to distinguish the Oxycephalidse, SPENCE BATE and the Xiphocephalidse as two separate families, and my reasons for doing so will be given below; I further think it necessary to transfer the genus Simorhynchotus, STEBBING, from the Tryphcenidce to the Oxycephalidae, in order to get uniformity in the generic cha- racters; in this latter family I remove Oxycephalus porcellus, GLAUS, 0. pronoides, C. BOVALLIUS, 0. longiceps, GLAUS, and 0. Steentrupi, C. BO- VALLIUS, from the genus Oxycephalus, sensu stricto, to STEBBING/S new genus Streetsia; for the reception of 0. typhoides, GLAUS, 0. sclero- ticus, STREETS, and a new species with the specific name The"eli, I pro- pose the new genus Stebbingella. For Leptocotis Lindstromi the new genus Dorycephalus is instituted. A hitherto undescribed Oxycephalid is the type for the new genus Crauocephalus , with the specific name C. GOESI. The names Rhabdosoma, ADAMS and WHITE, and Rhab- donectes, C. BOVALLIUS, are dropped in favour of the generic name Xi- phocephalus, proposed by GUERIN as early as in 1841 to this genus a new species X. Lilljeborgi is added. 12 CARL BOVALLIUS, THE OXYCEPHALIDS. II. THE SYSTEMATICAL POSITION OF THE OXCYCEPHALLDS, H. MILNE EDWARDS was the first who arranged the Hyperi- idean genera, into subdivisions, establishing in 1840 the following three tribus. 1. vHyperines gammaro'ides'j), 2. ^Hyperines ordinairesv and 3. ^Hy- perines atiormalesy>. The characteristics used by him for distinguishing these subdivisions were: 1. The size of the head, Tete petite)) and Tete tres grosse)). 2. The width of the body, Corps com prime" 5 and Corps renfle)). 3. Maxillipeds with palps, or wanting a palp. 4. The second pair of antennae straight, or zigzag-folded, ))Antennes de la seconde paire styliformes)), and ))Antennes de la seconde pair se reployant sur elles memes, etc.)) Of these characteristics the third one, depending on a wrong observa- tion, may be dropped without further discussion; the second characteristic varies, according to our present knowledge of the Hyperiidean species, so very much, not only among closely allied genera and species, but also in the case of the sexes of the same species, that it has no value what- ever. The first characteristic varies in the case of two so nearly rela- ted families as Vibilidce and Cyllopodidce and may thus be assigned the value of a family-characteristic, but not that of a subtribal one. The fourth may be considered as a good characteristic; with regard to the first subtribe, it may be observed however that the form of the second pair of antennae of the first five families in the following synoptical dia- gram show some tendency to the ))folding type)), so strongly developed in the Qxycephalidce and their allies. J. D. DANA, the next systematizer of the group, placed the Hyperids in three families, 1. Hyperidce, 2. Plironimidce, and 3. Typhidce. The principal characteristics were: 1. The second pair of antennae extant, or able to be hidden under the head or the body. 2- The hind part of the body able to be inflected or doubled up under the body, or not possessing this faculty. 3. The fifth and sixth pairs of legs mediocre, simple, or thick, elongated, often prehensile. SYSTEMATICAL NOTES. 13 The first is essentially the same as MILNE EDWARDS' fourth char- acteristic but is more vaguely expressed. The second has, in my opi- nion, no value, because almost all the Hyperids I have studied have the power of doubling up the urus under the body, though this power is not equally developed in the different genera. The third characteristic is utterly artificial, and the use of it would lead to the separation of closely related genera, and would place them in different families as for instance Euthemisto among the Phronimidce instead of close to Hyperia. SPENCE BATE in 1862 followed DANA as to the distinction of Hy- peridce and Phronimidce from the other Hyperids, but for the division of these latter into three families Platyscelidce, Phorcidce and Oxycephalidce, he used the following characteristics. 1. The head rounded. or long, anteriorly produced. 2. The eyes occupying the whole head or only a part of it. 3. The fifth pair of perseopoda with the femur largely developed, or imperfectly developed. The first and second characteristics are not valuable for distinguish- ing subtribes and scarcely constant as family-characteristics, while the third is a purely generic characteristic. C. GLAUS in 1879 included under the name DPlatyscelidem the fol- lowing five families 1. Typhidce, 2. Scelidce, 3. Pronoidce, 4. Lycceidce, and 5. Oxcyephalidce, and used the following /a?m7y-characteristics. 1. The body broad and short, or more or less compressed, elongated. 2. The urus able to be doubled up, or without this power. 3. The femora of the fifth and sixth pairs of perseopoda largely de- veloped, forming covers for the underside of body, or the femora elongated and narrowed. 4. The urus very short, or elongated. 5. The mouth-organs broad and short, - - or elongated and narrow. The first characteristic I have already said is too vague and, in my opinion, only of a specific value, for if you compare for instance the female of Hyperia Latreillei with the female of Eutyphis ovoides, we find the breadth of the body to be almost the same, and the body often not more depressed in Eutyphis than in Hyperia. Also among the Lycceidce there are species almost fully as broad as a Dithyrus or an Amphithyrus. The second is, as I have already observed when discussing the family- 14 CARL BOVALLIUS, THE OXYCEPHALIDS. characteristics used by DANA, only a relative quantity though it is true that it is carried to the extreme point in the Eutyphidce. The third is a good characteristic for distinguishing the Eutyphidce and the Scelidce from the Oxycephalidce and the Phorcidce, but unhappily the Pronoidce, and some forms of the Lycceidce are intermediate. The fourth is, in my opinion, very uncertain and liable to sexual variation. The fifth is a valuable characteristic for distinguishing the family Eutyphidce from the family Sce- lidce, but it is difficult to use it in external examination. After a long study of the different forms of the Amphipoda Hy- periidea I have arrived at the conviction that there are a number of char- acteristics, little or not at all appreciated by previous authors, but useful for the distinction of subtribes, families arid genera within the tribe Amphipoda Hyperiidea, of them I quote. the most important here. A. Characteristics for the distinction of the Subtribes of the AMPHIPODA HYPERIIDEA. 1. The point of fixation of the first pair of antennce. 2. The shape of the same pair of antennce. 3. The number of joints of the flagellum of the same pair of antennae in the male. According to these characteristics I propose to divide the Hype- rids into three subtribes: I. Hyperiidea recticornia. Diagn. The first pair of antennas are fixed at the anterior part of the head, they are straight, the first joint of the flagellum is large, the following few in number and terminal. II. Hyperiidea filicornia. Diagn. The first pair of antenna are fixed at the' anterior part of the head, they are straight, the first joint of the flagellum is large, the following many in number (c?), filiform and terminal. III. Hyperiidea curvicornia. Diagn. The first pair of antennae are fixed at the inferior part of the head, they are curved (j 1 ), the first joint of the flagellum is very large, the following few in number and subterminal SYSTEMATICAL NOTES. 15 These three subtribes correspond exactly with the three subdivi- sions A, B, and C in the synoptical diagram given in my ^Contributions to a Monograph of the Amphipoda Hyperiidea)), (p. 3) issued in 1887. The first two subtribes have not the same limits as the first two subdivisions proposed by MILNE EDWARDS and DANA, but the third has almost the same limitation as the Hype"rines auormales)) of MILNE EDWARDS, the family Typhidce of DANA, or the families Platyscelidce, Phorcidce and Oxycephalidce together, of SPENCE BATE, and lastly the same limitation as the scelidenj) of GLAUS. B. Characteristics for the distinction of the Families in the subtribe HYPERIIDEA CURVICORNIA. 1. The form of the femur of the sixth pair of perceopoda. 2. The form of the seventh pair of perceopoda. 3. The telson coalesced with the last ural segment, or free. 4. The development of the mouth-organs. 5. The second pair of perceopoda transformed, or not transformed. The first characteristic distinguishes the Parascelidce and Euhyphidce from the remaining families. The second distinguishes the Phor cor aphides, the Lycceidce, and the Oxycephalidce from the Pronoidce, Eutyphidce, and Xi- phocephalidce, this last family being instituted here by me for the recep- tion of the genus Xiphocephalus. The third characteristic separates the families Phorcoraphidce, Pronoidce, Euthamneidce, and Xiphocephalidce from the families Lycceidce,- Oxycephalidce, Parascelidce, and Eutyphidce. The fourth characteristic distinguishes the Lycceidce^ from the Oxyce- phalidce and the Xiphocephalidce, and the Parascelidce from the Eutyphidce. The fifth distinguishes the Phorcoraphidce from all the other families. Within the subtribe I place the family Phorcorhaphidce first because it comes nearest to the Hyperiidea filicornia owing to the situa- tion of the first pair of antenna?; last in the whole tribe I place the family Eutyphidce, because it seems to show the most affinity of any of the Hyperiidean families with the Isopods, especially with the Sphseromids, owing to the form of the body and the very intimate coalition between 1) As this characteristic is of no use in an external examination another but less good characteristic is used in the synoptical table p. 18. 16 CARL BOVALLIUS, THE OXYCEPHALIDS. the telson and the last ural segment. The likeness between the Xiphoce- phalidce and the Caprellids, suggested by previous authors, is very super- ficial and is no reasn for placing the Xiphocephalidce last in the tribe 1 ). C. Characteristics for the distinction of the Genera among the OXYCEPHALIDS. 1. The general form of the head and the rostrum. 2. The first pair of perceopoda being simple, subcheliform, or cheliform t 3. The second pair of perceopoda being subcheliform, or cheliform. 4. The inner rarni of the second and third pairs of uropoda coalesced with the corresponding peduncles, - - or free. 5. The relation between the length of the ural segments. These five characteristics are sufficient for the distinction of the genera given here below as constituting the family Oxycephalidce. D. Characteristics for the distinction of the Species of the OXYCEPHALIDEAN GENERA. These characteristics vary of course much in the different genera, but a few which are very useful and of common occurence may be pointed out here. 1. The armature of the first two pairs of perosopoda. 2. The seventh pair of perceopoda. 3. The relation between the outer and inner rami of the uropoda. 4. The relation between the telson and the last ural segment. Within the genus Xiphocephalus particularly the following good specific characteristics may be indicated: 1. The form of the epimeral of the first pair of perceopoda. 2. The form of the lateral parts of the seventh perceonal segment. 3. The relation between the length of the first two pairs of uropoda. 4. The relation between the last pair of uropoda and the telson. 1) For further information on these points see my forthcoming paper Remarks on the system and synonymy of the Amphipoda Hyperiidea. SYSTEMATICAL NOTES. 17 To illustrate the systematical following schematic table is given : position of the Oxycephalids the A. HYPEEIIDEA RECT1CORNIA. The first pair of antennae are fixed to the an- terior part of the head, they are straight, the first joint of the flagellum is large, the following few in number and terminal. The second pair are not similar to the first, slender, often angulate. a 1. The head is small, not tumid. aa 1. The first joint of the flagellum of the first pair of antennae is styliform, not tumid. aa 2. The first joint of the flagellum of the first pair of antenna? is lanceolate, tumid. aaa 1. The seventh pair of perreopoda are not transformed. aaa 2. The seventh pair of perseopoda are transformed. a 2. The head is large, tumid. aa 3 The first joint of the flagellum of the first pair of antennae is tumid. aaa 3. The seventh pair of peraeopoda are transformed. aaa 4. The seventh pair of peraeopoda are not transformed. aa 4. The first joint of the flagellum of the first pair of antennas is styliform, not tumid, a 3. The head with the peraeon is transformed into a halloon-like bladder. 2. LANCEOLTD^E. 3. VIBILID.E. 4. CYLLOPODID^. 5. 6. THAUMATOPSID^E. 7. MlMONECTID-ffi. B. HYPEEIIDEA FILICORNIA. The first pair of antennae are fixed to the an- terior part of the head, they are straight, the first joint of the flagellum is large, the following many in number filiform (J), terminal. The second pair are subsimilar to the first, not augulate. fo 1. The uropoda with distinct rami. fob 1 The mandibles with palp. bb 2. The mandibles without palp. b 2. The uropoda transformed, without rami. Nova Acta Reg. Soc. Sc. Cps. Ser. III. 8. HYPERIID^E. 9. PHRONIMID.E. 10. ANCHYLOMERID^}. 3 18 CARL BOVALLIUS, THE OXYCEPHALIDS. C. HYPERIIDEA CURVICORNIA. The first pair of antennae are fixed to the inferior side of the head, they are curved (tf), the first joint of the flagellum is very large, the following few in number and subterminal (tf). The second pair are folded (<3% or wanting (9). 1. The femur of the sixth pair of perseopoda is not operculiform. CC 1. The second pair of perasopoda are transformed. The seventh pair of perseopoda are not transformed; the telson is free. CC 2. The second pair of perseopoda are not trans- formed. ccc 1. The head is scarcely produced, not forming a rostrum. The maxillse con- sist of two lamina?. CCCC 1. The seventh pair of perseo- poda are reduced to two or three joints. The seventh pair of peraao- poda are complete, not transformed. cccc 2. ccccc 1. ccccc 2. The telson is free, not co- alesced with the last ural seg- ment. The telson is coalesced with the last ural segment. 11. PHORCORAPHID.E. 12. PRONOID^E. 13. EuTHAMNEIDvE. 14. LYCLEID.E 1 ). 1) In my preliminary paper sSystematical list of the Amphipoda Hyperiidea, I thought it convenient to place the genus Trypluena, A. BOECK, in the same family as Lyccea, DANA, and to apply to the family, containing these two genera and the allies of Lyccea, the name Trypluenidce, which preceded by nine years the family-name Lycceidce, proposed by CLAUS in 1879 in his systematical arrangement of the Platyscelids. Further study of these interesting Hyperids has however convinced me that Try- phcena is more closely allied to the genera Phorcoraphis, STEBBING, and Lycceopsis, CLADS, than to Lyccea and the other genera placed there by CLAUS. It is principally the peculiar transformation of the second pair of perseopoda that makes it desirable to place Tryphcena close to the other members of the old family Phorcidtf, beacuse it would be very strange to suppose such a homologous development of the same organ in genera not closely allied. To this characteristic come others which also, but not so thoroughly, point to a closer relation between the three genera, Phorcoraphis, Lycteopsis, and Tryphcena, as for instance the form of the sixth pair of peiseopoda, the form of the urus and its appendages, and in some way the shape of the first pair of antennae. From the point of view of priority the family- SYSTEMATICAL NOTES. 19 CCC 2. The head is more or less produc- ed forming a rostrum. The maxillae are rudimentary. CCCC 3. The telson is coalesced with the last ural segment. The seventh pair of perseopoda are complete, not transformed. 15. OXYCEPHALID^:. CCCC 4. The telson is free l ), not co- alesced with the last ural segment. The seventh pair of perseopoda are transformed. 16. XIPHOCEPHALHXE. C 2. The femur of the sixth pair of perasopoda is more or less operculiform. CC 3. The seventh pair of perseopoda are complete, not transformed. The mouth-organs are narrow, protruding. 17. PARASCEUD^E. CC 4. The seventh pair of perseopoda are reduced. The mouth-organs are short and broad. 18. EUTYPHID^E. name Trypfusnidee ought to be applied to the family thus composed, but as it has been used as a sy- nonym for LycfEidee by myself and by STEBBING it would make confusion in the synonymy. I therefore prefer to use a family-name derived from the new form given by STEBBING to the old generic name Phorcus, which was found to be preoccupied. To maintain, as does STEBBING. the family-name Phorcida, when the generic name Phorcus is rejected, is not convenient as the family-name must refer to a generic name in use within the family. I have also divided the old family Tryphcenidce into two, Euthamneidte comprising the two genera Thamyris, SPENCK BATE, (or Brachyscelus, according to STEBBING) and Euthamneus, (the former name Thamnem being thus modified to avoid confusion) and Lycceida, CLAUS, restricting it to the two genera Lyccea, DANA, and Pseudolyccea, CLADS. Of the old members of the family Try- plicenidai, as it was composed in my paper, which i have just referred to, I have thus placed Tryphcena in the family Phorcoraphidce; Paratyceea in the Pronoidce, (this transposition was really done in 1887 but owing to a typographical error it stands into Systematical listt> with its number from Pronoida among the Tryphcenidce) ; Thamyris and Euthamneus in the new family Euthamneidce ; Lycoea and Pseu- dolyccea in the Lycceidce; and lastly Simorhynchotus in the Oxycephalidce. 1) A very puzzling exception to this rule is Ehabdosoma brevicaudatum, described by STEBBING (1. c. p 1612, pi. 208). He says namely that the telson is coalesced with the last ural segment and much shorter than this segment, broadly rounded and pectinate at the apex, but all these statements are strikingly opposed to my experience which is founded upon the examination of many specimens of the two old species, Xiphocephalus armatus and X. Whitei. I for my part am much inclined to believe that the single specimen which was the type for the new species proposed by STEBBING, may have been abnormal, perhaps injured. III. MORPHOLOGICAL SOTES OIS THE OXYCEPHAL1DS. /. The shape of the body. The genus Xiphocephalus is certainly the most extreme of all the Hyperiidean forms with regard to its development in length and narrowness, but this extravagant, rod-like form of body is however very naturally connected with the other Hype- rids by intermediate links. Thus, if examining the genera of the Oxycepha- lids, we find first in the series the thick and stout Tullbergella, (fig. 1) in shape of peraeon very similar to a true Hyperia, and then Oxycephalus, with the body somewhat more compressed, and next Stebbingella, both of them with the head and the hind part of the body tolerably elongated, from them is an easy transition to the slender Street- si a. This slender form of body is more pronounced in Leptocotis and Dorycephalus, which, on the other hand, are readily found to be connected with Xiphocephalus (fig. 2) Fig. 1. ^Tullbergella cuspidata. Fig. 2. Xiphocephalus armtaus. J 1 . by the rod-like form of body of Calamorhynchus. The body of the females is usually somewhat broader than that of the males, except in Xiphocephalus, where such a sexual difference is not present. MORPHOLOGICAL NOTES. 21 2. The integument In Tullbergella (fig. 27) and Stebbingella (fig. 3) the integu- ment is very thick and hard, and calcareous; in the latter it shows many tubercles, fine notches, scrolls and grooves; in Oxycephalus (fig. 4) the integument is thinner, sometimes forming longer or shorter longitudinal and transversel keels or folds, but these keels are only angulated eleva- tions of the integument, not formed by calcareous deposits as in the two above mentioned genera. In Cranocephalus, (fig. 5) Dorycephalus, Fig. 3. Stebbingella Theeli. . Fig. 4. Oxycephalus Clausi. Fig. 5. Cranocepha- Im Goesi. cf- Leptocotis and Xiphocephalus, the integument is still thinner and not at all calcareous, showing in some species small epidermal grooves which will be spoken of under the head glandsy> below. The hypodermis seems to be either composed of hexagonal cells, or hexagonally reticulated but this reticulation is seldom to be seen distinctly; at the head, the epirnerals, and parts of the ural appendages it is most distinct. The colour is very variable, red, yellow, green, and white; many species are entirely hyaline. Only in Tullbergella cuspidata, and in Xiphocephalus White i, have I observed larger pigment-cells, or chromatophors of some kind, of more or less regular star-like form. There are also to be seen separate calcareous (?) concrements in the shape of small, irregularly ovate discs or buttons in the thin integument of some species of Oxycephalus; in the thick, calcareous integument of Tullbergella they are very numerous. 22 CARL BOVALLIUS, THE OXYCEPHALIDS. 3. The head. The shape of the head of the Oxycephalids shows forms gradually developing almost from the normal type in Lycceidce to the strange-look- ing, extremely elongated form in Xiphocephalidae. The genus Glosso- cephalus has a distinct, but thick and obtuse, tongue-shaped rostrum (fig. 6 and 6 a), Oxycephalus (fig. 7 and 8), Stebbingella (fig. 3), and Streetsia (fiig. 9) offer different steps of development of a more and more elongated and sharp rostrum, open at the under-side; in Dory- cephalus and Leptocotis (fig. 10) it has reached the form of a needle, which form is so enormously developed in Xiphocephalus, that it may be properly called filiform (fig. 2). An aberrant development of the rostral or anterior part of the head occurs in Calamorhynchus where the dilation of the basal portion of the rostrum, which is indicated already in Tullbergella (fig. 12), is exaggerated into the large wing-like projections shown in fig. 11. Fig. 6. Fig. 6 a. Glossocephalus Milne Edivardsi. Fi(j 7. Oxycephalus Claim. <$. Fig. 8. Oxycephalus Cluusi. But also the hind part of the head shows a gradual development from being in Streetsia riot constricted, in Oxycephalus, Crano- cephalus, Tullbergella, and Stebbingella, feebly constricted and scarcely forming a distinct neck, to Do rycephalus, Leptocotis and Cala- morhynchus, where the hind part of the head is strongly constricted, forming a distinct, but short neck, and lastly to Xiphocephalus, which has a long and narrow neck, longer than the central part of the head (fig. 2). The highest part of the head, which may be situated close to the anterior margin of the first peraeonal segment, as in Simorhynchotu s MORPHOLOGICAL NOTES. 23 and Streetsia, or at a considerable distance therefrom as in Xipho- cephalus, is occupied by the eyes. The under side of the head is excavated, forming a groove or channel from the base of the head to the anterior end of the eyes, where the channel is enlarged into a vaulted chamber, at the upper anterior wall of which the first pair of antennae are fixed. In this channel the second pair of antennas are received when folded, extending into the lower part of the chamber below the Fig. 9. Streetsia pronoides. . Fig. 26. Glossoccphulus spiniger and more slender than the second. In the family Xiphocephalidae the flagellum consists of only one joint, much longer than the peduncle, bulbous or tumid at the lower end, and carrying some long olfactory hairs. 6. The second pair of antennae. They consist in the male of five joints in all the genera except Glossocephalus, where they are composed of six. The first four joints are elongated, articulating with each others in such a way as to allow of being folded up like a carpenter's rule. The last or the last two joints in Glossocephalus are minute, and the articulation allows of but little move- ment. The first joint of the second pair of antennas varies in length, ac- cording to the length of the head, being shorter than half the second joint in the genera Simorhynchotus, Tullbergella and Cranocephalus (fig. 27 and 28). With this shortness of the first joint is connected a very important difference in the manner of hiding the antennae when folded. In the three just mentioned genera the folded antennae are concealed for more than half their length under Fig. 27. Tullbergella cuspidata MORPHOLOGICAL NOTES. 27 the perseon between the perseopoda, as in the immediately preceding families of the Amphipoda curvicornia, but the anterior part of the folded antennae, concealed under the head, is comparatively longer than in those families. This feature is very interesting as showing the close relation between the Oxycephalidas and the Lycoeidce. In the other genera the first joint is only a little shorter than the second, and the second pair of antennas are, when folded, concealed for their whole length under the head within the above mentioned hollow channel, and within the lower part of the spacious chamber, in which the first pair of antennae are fixed. The length of the antennae in these latter genera depends naturally Fig. 28. Simorhynchotus Lilljtborgi. Fig. 29. Oxycephalus pectinatus upon the length of the head, so that they are comparatively short and stout in the genera Glossocep halus and Stebbingella (fig. 30), tole- rably long in Oxycephalus (fig, 29) and Streetsia, and very long and slender in Dorycephalus (fig. 31) Leptocotis 1 ), and Xiphocepha- Fig. 30. Stebbingella typhoides. Facsi- mile from Glaus Platysceliden, pi. Fig. 31. Dorycephalus Lindstroemi. lus (fig. 2). In the three genera, Simorhynchotus, Tullbergella, and Cranocep halus the second joint is more than twice as long as the first, while in all the other the second it is equal to or only a little longer than the first. In all the genera the third and fourth joints are about 1) The male of Calamorhynchus is hitherto not known. 28 CARL BOVALLIUS, THE OXYCEPHALIDS. the same length as the second. The extremities of the first four joints are dilated, showing powerful muscles plumosely arranged. In Stebbing- ella the antennal joints are much bent (fig. 30) but in the other genera they are more or less straight. The fifth joint, or in Glossocephalus the fifth and sixth joints are very short (fig. 31). The margins of all the joints are fringed with minute hairs. 7. The mouth-organs. These organs are less developed in the Oxycephalids than in most of , the other Hyperids. In the genus Xiphocephalus (fig. 32) the epistoma forms a rounded, shield-like plate, with a strongly protruding, Fig 32. Xiphocephalus Whitei. Fig. 33. Oxycephalus piscator. Fig. 34. Simorhyncliotus Lilljeborgi. central, spine-like process. In the true Oxyce phalidae it consists of a tu- berculous prominence in front of the mouth-organs. The lab rum (fig. 33) is more or less distinctly bilobed, the hind margin feebly curved, and almost entirely smooth. The mandibles (fig. 34 39) are best developed in the genera Tu 11 berg el la and Glossocephalus, and most simple in the genus Fig. 35. Oxycephalus piscator. Fig. 36. Oxycephalus piscator. Young c?- Xiphocephalus. In the first genus the right mandible has a large, straight, cutting or incisive process with a broad neck springing out from the somewhat elongated stem of the mandible; on the left mandible there is also an accessory process not half as wide as the principal. In Xiphocephalus the mandibles are less elongated than in the Oxyce- MORPHOLOGICAL NOTES. 29 phalidae and the cutting edge is not at all separated from the stem by a distinct neck. A molar tubercle is wanting in all the representatives of the two families. The palp is strongly developed in the males; among the Oxycephalidae the first joint is more or less elongated according Fig 37. Oxycephalus piscator. Fig. 38. Streetsia Steenstrupi. to the varying length of the head, and always nearly as long as the first joint of the second pair of antennae; the last two joints are short, subequal in length, and slender. I the Xiphocephalidae the first joint is enormously elongated, being many times longer than the two termi- nal joints together. Fig. 39. Dorycephalus Lindstroemi. Fig. 40. Tullbergella cuspidata. The labium (fig. 40) is soft, the lateral parts being rounded. Both the pairs of maxillae are very rudimentary being possibly represented by two minute soft laminae at the base of the mandibles. Fig. 41. Fig. 42. Oxycephalus piscator. Fig. 43. Glossoce- phalus spiniger. Fig. 44. Doryceplia- Fig. 45. Xiphoce- lits Lindstroemi. phalus Whitei. The maxillipeds (fig. 41 45) consist of a two-jointed, much bent stem; at the distal or lower margin of the second joint rise laterally two more or less spoon-shaped laminae, the homolpgues of the palps in the Gam- marids. In the middle between these laminae there extends a small, some- times tongue-like process, sometimes more or less cylindrical or tube- 30 CARL BOVALLIUS, THE OXYCEPHALIDS. like, bent inwards against the tip of the mandibles. The margins of the lateral laminae are often tuberculated and serrated or sparingly set with short bristles or fine hairs. 8. The perceon and its appendages. According to the varying form of body in the different genera the perseon is short and stout, as in Tnllbergella (fig. 1), or long and almost rod-like, as in Xiphocephalus (fig. 2), but it may be obser- ved that in the enormous elongation which occurs for instance in Xiphocephalus, has the perseon comparatively less share than the head, the urus, and the pleon. Fig. 46. The first and second peraeonal seg- ments of Xiphocephalus Whitei. c? Fig. 47. Th.e sixth and seventh personal seg- ments of Xiphocephalus Whitei. <5- The epimerals are distinct in all the genera of the Oxycepha- lidae (fig. 1, 3, 4, 5 and 27}, but in the Xiphocephalidee only the epimeral of the first pair of perseopoda is distinct, the others being coa- lesced with the corresponding segments (fig. 46 and 47). The branchial vesicles are attached to the second and four following pairs of pereeopoda in all species of both families. Those in the Oxycephalidee are comparatively small, while those in the Xipho- cephalidas are very large (fig. 46 and 47). The ovitectrices vary in form in the different genera of the Oxycephalidae; in Tullbergella, Simorhynchotus, and Stebbing- ella they are broadly ovate, in Dorycephalus, Streetsia, Oxyce- phalus, and Leptocotis they are more elongated, the margins being smooth without hairs, and in Glossocephalus they are more irregular in form, the lower end being produced into a sharp angle. MORPHOLOGICAL NOTES. 31 In the Xiphocephalidas on the other hand there exist no ovi- tectrices, but the eggs on leaving the oviduct are fixed on the under side of the body, probably by means of some kind of cement, which is pos- sibly secreted from some of the strongly developed glands in the joints of the perasopoda. In any case the manner of carrying the eggs arid embryos is very remarkable, and, as far as I know, has not been pre- viously observed (fig. 47). A different number of eggs are fixed under the different segments: under the first I have never observed a single Fig. 48. The fourth peraeonal segment Fig. 49. The sixth perseonal segment , of Xiphocephalus Whitei. Q. of Xiphocephalus Whiiei. egg, under the second three or four pairs are fixed at equal intervals, under each of the third and fourth four or five, under the fifth seven or eight, under the sixth five or six pairs and none under the seventh, one female thus carrying in all from forty to fifty eggs or embryos. I have observed very young eggs thus placed, they being in fact in the third or fourth stage of division, and quite globular (fig. 48), as well as embryos just on the point of breaking the egg membrane (fig. 49), the form of the egg in this case being elongate-ovate, and the young being provided with six pairs of peraeonal legs, distinctly developed and ar- med with strong claws. The first and second pairs of perceopoda are developed into strong grasping or prehensile organs, and show in the different genera almost all the modifications of such organs which are to be observed among the other Hyperids. We meet with the less complicated form in the first pair of pe- rseopoda of Simorhynchotus (fig. 50), that is to say: 1. The folding hand (fig. 50), in which the carpus is only a little di- lated, and has a more or less rounded hind margin, so that the metacarpus can be doubled up along it so as to impinge against its hind margin, the dactylus thus coming in contact with the tibia. 32 CARL BOVALLIUS, THE OXYCEPHALIDS. This is the same " form of prehensile organ which exists in the third and fourth and also in the fifth pair of peraeopoda in Euthemisto and Para- tkemisto. Moreover the folding hand is common in the Amplnpoda Gammaridea, in the first three pairs of peraeopoda in some families of Isopoda, and is probably most strongly developed in the first pair of Squ'dla and other Stomatopods. The next step in the development of these organs is represented in many genera, as for in- 8ta " ce in the e t C nd P1)ir hl SimorhynchotuB, the first pair in Streetsia (fig. 52), the first pair in Dorycephalus, the first and second pairs in Tullbergella (fig. 51), and in others, that is to say: 2. The subcheliform hand, (fig. 51 53) in which the carpus is broadly dilated backwards, but not produced downwards, or only very little produced, so that the meta- carpus impinges against the under margin of the carpus, the dactylus here coming in contact with the hind margin of the carpus. When the. \ J hind margin of the metacarpus is in contact with the under margin of the carpus, the meta- Fig. 51. The first pair of carpus is at a right angle with the axis of the leg, la cuspidata. the leg being supposed to hang straight down. Fig. 52. The first pair of Streetsia Steentrupi. Fig. 53. The first pair of Cranocephalus Goesi. The next form, the most developed prehensile organ, is best re- presented in the second pair of peraeopoda in the genus Oxycephalus (fig. 54), but is also met with in the first pair in the same genus, in the second pair in Streetsia (fig. 55), and in others; that is to say: MORPHOLOGICAL NOTES. 33 3. The cheliform hand, in which the carpus is not only dilated back- wards, but also produced downwards into a longer or shorter carpal process, so that the metacarpus impinges against the front margin of the carpal process and the dac- tylus comes into no contact whatever with the carpus itself. When the hind margin of the metacarpus is in contact with the front margin of the carpal process, the metacarpus is at an angle of from 45 to almost with the axis of the leg. Naturally there are many transitions Fig. 54. The second pair of Oxyce- between the three kinds of prehensile phalus Clattsi. Fig. 55. The second pair of Streetsia Steenstrupi. Fi(j. 56. The first pair of Dorycephalus Lindstroemi. hands, but for practical reasons I think it best to keep these three terms, at least as far as regards the tribe Amphipoda Hyperiidea. When the lower end of the metacarpus is dilated so that this joint shows an under margin, straight or excavated, and when the hinder corner of this under margin is produced so that the dactylus, or a por- tion of the dactylus, impinges against the under margin of the meta- carpus, I call the double prehensile organ thus formed: 4. A complexely subcheliform hand, (fig. 56), or 5. A complexely cheliform hand, (fig. 57) as the carpus is not or is produced, thus disregarding the fact whether the lo- wer hinder corner of the metacarpus is or is not produced. Nova Acta Reg. Soc. Sc. Ups. Ser. III. 5 34 CARL BOVALLIUS, THE OXYCEPHALIDS. None of the following pairs of legs are prehensile in any of the adult forms of the genera dealt with in this treatise. The seventh pair are complete in all the genera of the Oxycepbalidge (fig. 58) with the excep- tion of Tullbergella (fig. 59), in which the carpus and metacarpus are fused together, a feature closely resembling that which occurs in the families Vibilidce and Cyllopodidce among the Hyperiidea recticornia, and a feature also suggesting the relation of the Oxycephalidae with the Pronoidce. In Xiphocephalidas the seventh pair are transformed (fig. 47). Fig. 57. The first pair of Xiphocephahts arm fi tits. Fig. 58. Streetsia cari- nata. Fig. 59. Tullbergella cuspidata. The femur. In none of the Oxycephalids is the femur of the first four pairs of peraeopoda much dilated ; it is broadest in the first and second pairs in Oxycephalus but scarcely half as broad as long. The femur is narrowest, almost linear, in all the legs in the Xiphocephalidas. In the Oxycephalidre the femur of the third and fourth pairs is usu- ally narrow, and linear, while that of the last three pairs is on the other hand more or less dilated; in Glos socep halus it is comparatively nar- rowest; then come, with the femur increasing in breadth, Doryce- phalus, Leptocotis, C alamorhynchus, Streetsia, Tullbergella, Cranocephalus, Oxycephalus and Stebbingella, in which last the fe- mur is much dilated, somewhat tending to the form adopted by the Para- scelidce. In Stebbingella we have also the peculiarity, that the external surface of the joint shows some pits or holes which are probably for cutaneous glands (fig. 60). In all the genera except Stebbingella the femur of the sixth pair is broader than, or at least as broad as, the femur of the fifth pair. In Stebbingella the femur of the fifth pair is broadest. In Tullbergella the femur of the sixth pair (fig. 61) is produced at the lower hind corner into a long sharp process. In Streetsia the lower hind apex of the femur of the sixth pair (fig. 62) is produced downwards, so that the germ articulates sub-apically with the femur, a feature which shows some connection with that which oc- curs in the Pronoidce. The form of the femur of the seventh pair is MORPHOLOGICAL NOTES. 35 more or less ovate except in Stebbi n gel la (fig. 60), where it has a trapezoid form. The (jenu is always short and not dilated. The tibia is tolerably long, except in the first two pairs of peneo- poda; in the third and fourth pairs in Glossocephalus, and in the third and three following in Xiphocephalus it is much elongated. In the female of this last genus it is sometimes, and I think periodically, enor- mously inflated owing to the strong development of the glands within it. In the former genus it is much dilated, resembling that joint in Phor- corhaplns and Lycceopsis. The tibia is often provided with hairs or Fit,. 60. StclUngella Theeli. Fig. 61. TuUbergclln cuspidata. Fig. 62. Streetsia carinata. spines along the hind margin - - in the third and and fourth pairs - or on the front margin - - in the fifth and sixth. The carpus. The development of the carpus in the first and second pairs has been spoken of above ( p< 31 33). I n the third and fourth pairs the carpus is never dilated, as it is for instance in the genera Parathemisto and Euthemisto; it is often spiniferous or serrated, and in Xiphocephalus sometimes inflated, but not in the same degree as the tibia. In Glosso- Jv- 63 - . ,, ,.-, mi Xiphocephalus Whitei. cephalus it is dilated as well as the tibia. carpus of the fifth and sixth pairs is usually serrated or set with spines; in Xiphocephalus it carries a series of ovate discs or plates armed 36 CARL BOVALLIUS, THE OXYCEPHALIDS. with a row of four or five stout spines, which are probably sensitive or- gans of some kind (fig. 63). The metacarpus is usually narrow and linear; the hind margin of the third and fourth pairs, and the front margin of the fifth, sixth, and sometimes of the seventh is serrated or pectinated. In Xipho- cephalus the metacarpus of the fifth and sixth pairs shows sensitive organs like those of the car- pus mentioned just above (fig. 64). The dactylus of the first and second pairs as well as that of some or of all the other pairs has often a circular or ovate hole at the base of the hind margin, in the case of the first four pairs, and of the front margin in the case of the others. This hole is an outlet for the secretion of the glands Fi ' 65 c ^ ep}ialm which are present in many of the joints of the peneopoda. Sometimes the hind margin is chanel- led, and in some few cases the apex of the dactylus is open (fig. 65). The dactylus has often a tooth-like prominence in the middle of the hind margin. 9. The pleon and its appendages. The pleon differs less in form from that in the other Hyperids than any of the other 'portions of the body. In Xiphocephalus it is much elongated, longer than the perason (fig. 2); in Tullbergella it is very broad and stout, the lower hind corners of the last segment projecting into long sharp processes (fig. 74). In Stebbingella it is hard and tuberculous with grooves and ridges on the external surface, in Dorycephalus _ the integument is thin and marked with Fig. 66. Oxycephu- Fig. 67. Simorhymh- peculiar pits and excavations. The pleo- lus piscator. otus JMijeborgi. p O d a are nor mal; the peduncles are glo- bular, egg-shaped or almost cylindrical. On the inner side of the apex of the peduncle are to be seen those peculiar organs first mentioned by G. 0. SARS in his admirable work on the fresh-water Crustacea of Norway 1 ) under the name epines pa rticuliers)) and termed coup- 1) G. O. BARS. Crustaces d'eau douce cle la Norvge. I p. 5 and 133 pi. 5, fig. 8' and 8". Christiania 1869. MORPHOLOGICAL NOTES. 37 ling spines)) b}' STEERING in 1888, (1. c. p. XIV). They have usually the form of a club or a mushroom, the head armed with short, strong, retroverted teeth, their purpose being to hold the two pleopods to- gether (fig. 66 69). The rami o the pleopoda are also very similar Fig. 68. Fig. 69. Oxyceplialus piscator. Fig. 70. Simorhynch- otus Lilljeborgi. to those in the other Hyperids as well as in the Amphipoda Gammaridea. The first large joint of the inner ramus carries a Deleft bristle)) (fig. 68 70), also first noticed by G. 0. SARS and minutely described by Stebbing from many different forms of the ))Challenger)) Amphipoda. The following joints are provided each with one pair of feathered bristles. 10. The urus and its appendages. The first segment of the urus is usually short, much shorter than the last pleonal segment, except in Xiphocephalu s in which it is much elongated, rod-like, and longer than the last pleonal segment. The second and third segments are coalesced into one, and usually longer than the first segment, the only exception being Xipho cephalus ar- matus, in which the coalesced last ural segment is shorter than the first. The last coalesced segment is comparatively broad in Simorhynchotus (fig. 76), Tullbergella (fig. 74), Glossocephalus (fig. 71), and Steb- bingella, more elongated in C ranocephalus (fig. 72), Oxycephalas (fig. 75), Streetsia, Dorycephalus (fig. 73) Leptocotis, and Cala- morhynchus, and lastly very narrow and rod-like in Xiphocephalus. The telson is invariably coalesced with the last ural segment in all the genera of the Oxycephalidse, and this character not only suggests the relation of the Oxycephaiidae with the Lycceidce, the Para- 38 CARL BOVALLIUS, THE OXYCEPHALIDS. scelidce, and the Eutyphidce, but is also a feature of the Isopoda. In the Oxy cephalidge the telson is more or less broad, varying from the broadest form in Simorhynchotus (fig. 76) and Tullbergella (fig. 74) to the narrowly elongated form in Leptocotis and Dorycephalus (fig. Fig. 71. Glossocephalus Milne Edicardsi. Fig. 72. Cranocephalus Goesi. 73). It is usually sharp-pointed at the apex, but in Cranocephalus and Glossocephalus the apex, is broadly rounded (fig 71 and 72). In the Xiphocephalidag on the other hand the telson is free from, Fig. 73. Dorycephalus Lindstroemi. Fig. 74. Tullbergella cuspidata. and articulating with, the last ural segment; but the new species X. bre- vicaudatus, described in 1888 by STEBBING (1. c. p. 1612) shows a very remarkable exception to this rule. The case he reports of the telson being coalesced with the last ural segment may possibly be an indi- MORPHOLOGICAL NOTES. 39 vidnal abnormity in the only specimen examined owing to .some lesion, or if not, I am quite unable to see how such a contradiction is to be explained. In Xiphocephalus the telson is usually much longer than the rest of the urus. The uropoda show naturally the same gradation as the urus from stout, broad, laminar peduncles with broadly lanceolate rami, to extremely elongated, rod-like peduncles with more or less needle-shaped rami. In Gloss oc ephalus (fig. 71) they are very broad and short Fig. 75. Oxycephalus piscator. Fig. 76. Simorhynchotus Lilljeboryi. like those in the genus Hyperia; in Simorhynchotus (fig. 76), Oxy- cephalus, C ranocephalus, Tullbergella and Stebbingella, they are comparatively short; in Dorycephalus (fig. 73), Leptocotis and C alamorhy nchus they grow gradationally longer and narrower, the peduncle of the first pair even in Leptocotis being 8 or 10 times longer than broad. In Xiphocephalus the elonga- tion is carried to its extreme limit, the length of the rod-like peduncles being 30 50 times greater than their breadth. In some of the genera, see the diagram p. 46 all the rami are free, articulating with the pedunc- Fig. 77. Dorycepha- les (fig. 71 and 72), in others viz: Dorycephalus ^ ^dstroemi. and Tullbergella (fig. 73 and 74) the inner ramus of the last pair is coalesced with the peduncle while ail the other rami are free; and lastly in Oxycephalus (fig. 75 and 76) and others the inner rami of the last two pairs are coalesced with the corresponding peduncles, the other rami being free; and in Xiphocephalus also the inner rami of the last two pairs are coalesced with the corresponding peduncles (fig. 7880). 40 CARL BOVALLIUS, THE OXYCEPHALIDS. The outer and the inner ramus are usually both well developed, and more or less equal in length. Exceptions from this are Dory- cephalus Lindstroemi (fig. 77), Xiphocephalus armatus (fig. 80) and X. brevicaudatus. In the first of which the outer rami of the first and third pairs and in the two latter those of all the three pairs are very reduced in size, diminutive. Usually we have well developed glands in the peduncles and rami; their outlets are usually situated at the upper end of the rami. where the rami are in contact with one another. In Xiphocephalus Fig. 78. The second Fig. 79. The third Fig. 80. Xiphoce- Fiy. 81. The inner Fig. 82. Telson pair of Xiphocepha- pair of Xiphoce- phalus armatus. ratmis of the third of Xiphocepha- lus Whitei. phalus Whitei. pair of Xiphoce- lm Whitei. phalus Whitei. the glands are highly developed, filling not only the uropoda but also the telson; at the apex of the telson a small circular outlet is to be ob- served; at the end of some of the rami there are also similar outlets to be seen (fig. 81 and 82). *) 1) In the following parts of the Morphological Notes I give a few preliminary notices on the nervous system, glands, etc., reserving a fuller account of the different organs of the Oxycephalids for the third volume of my Contributions to a Monograph of the Amphipoda Hyperiidea. MORPHOLOGICAL NOTES. 41 //. The nervous system. The ganglionic chain shows a strongly developed supra-oeso- phagial, or cephalic ganglion; from its anterior corners extend the very stout and thick optical nerves, dilating apically into an ovate or bean-shaped disc, which is more or less closely pressed against the layer formed by the retina. Between the optical nerves rise two stout nerves from the front margin of the cephalic ganglion, and run to the first pair of antennae; at the base of these antennal nerves are situated the two vesicles which have been interpreted by GLAUS *) as auditory organs. These vesicles are small in the Oxycephalidse but very largely develop- ed in the Xiphocephalidae. The commissural cords connecting the cephalic ganglion with the infra-oesophagial ganglion are thick and stout, much thicker than for instance in Lanceolidce. Close upon the infra- oesophagial ganglion follow the first two pairs of personal ganglia; they are not coalesced with the infra-oesophagial ganglion but in the Oxy- cephalidae they are so closely pressed against it that the commissural cords between them are scarcely perceptible. The following four pairs of personal ganglia are equidistant, situated each in the middle of the corresponding segments, the third, fourth, fifth, and sixth. The pair of ganglia supplying the seventh personal segment and its appendages with nerves are situated just at the limit between the sixth and seventh segments, close to the preceding pair of ganglia but not coalesced with it. In the hind portion of the body the ganglionic chain shows three pairs of ganglia, one pair situated in the middle of each of the pleonal segments, and the last consisting of two or three pairs, fused together and so supplying the whole urus with nerves. The ganglionic chain is thus composed of twelve pairs of ganglia, some pairs being placed clo- sely together. The commissural cords run close to one another, but are not coalesced. In the Xiphocephalidaa the first two pairs of pe- reeonal ganglia are more separated from one another and from the infra- oesophagial ganglion, so that the commissural cords are distinct, but not longer than the ganglia themselves. 12. The glands. Glands are present in most of the joints of the perseopoda and uropoda of all the species of both the families. They are either simple, 1) C. GLAUS, 23, p. 150; 36, p. 15. Nova Acta Reg. Soc. Sc. Ups. Set. III. 42 CARL BOVALLIUS, THE OXYCEPHALIDS. Fig. 83. Cutaneous gland in the margin of the first epimeral of Xi- phocephalus Whitei. consisting of only one cell, with a short, narrow duct opening into a small circular hole in the surface of the integument (fig. 83), which hole is usually guarded by one or more small simple hairs ; or they are composed of many cells, and packed into thick masses round a central lumen from which a com- mon duct runs to the outlet. The outlet is often placed, as mentioned above, at the base of the dactylus (fig. 84), more seldom at its apex (fig. 65), often in semicircu lar incisions at the upper end of the rami of the uropoda, or as in Xiphocephalus -- at the apex of the telson and of the rami of the uro- poda (fig. 81 and 82). Usually the glands of the legs are concen- trated in the femur, and only a few and scattered glandular cells are to be seen within the follo- wing joints, often tying round or close to the duct which runs from the chief mass of glands in the femur to the base of the dac- tylus (fig. 84); but in the fema- les of the three species of Xi- phocephalus which I have exa- mined, the tibia and carpus of the third and fourth pairs of perseopoda, as well as the same joints of the fifth and sixth pairs Fiq. 85. The tibia of the ' , n n4 ~J ^J are sometimes inflated, and fourth pair of peraeo- f poda of Xiphocepha- almost egg-shaped, owing to a lus Whitei. . strongly developed glandular mass surrounding the axis of the joint for the whole of its length (fig. 85). The development of these powerful glands in the tibia and carpus of those pairs of perseopoda is, however, in my opinion periodic- al, and may be supposed to have some connection with the fixation of the eggs on the under-side of the body, so I venture the supposition that they are cement-glands. I believe this enormous development of the tibial and carpal glands to be periodical because I have seen fullgrown females of X. armatus and X. Whitei without the joints in question inflated at all; but I have never seen females of the same species with eggs or young ones, which had not at the same time those joints more or less inflated. Thus it seems probable that the development of these glands may be connected with the maternal functions Fig. 84. Oxycephalus laiirostris. MORPHOLOGICAL NOTES. of the animal, and that we thus may find them increasing or decreasing in volume according to the stage of the development of the ovaria. The other glands present within the joints of the legs have pro- bably a different function, according to the place they occupy, and so I am inclined to think that the highly developed glands in the uropoda and telson of Xiphocephalus are poison-glands. 13. The young. The eggs are in the family Oxycephalidae packed in, and pro- tected by a pouch formed of five pairs of ovitectrices, attached to the second and four following pairs of peraeopoda. There the eggs rest until the time of hatching, when the young have a form very similar to that of a true Hyperia. In the Xiphocephalidse also the young remind one of the form occurring in Hyperia, Vibilia, and other genera of the first two subtribes. But the manner of bearing the eggs is a very peculiar feature. No ovitec- trices exist but the eggs are placed under the perseonal segments, between the lower parts of the sides where they are fixed in two regular rows, as mentioned above (p. 31). When the young are ready to be hatched we find that each egg is cemented to the under surface of the peraeon with the head-end, and that the young hang tail downwards (fig. 48 and 49). The first pair of antenna in the just hatched young one of Xi- phocephalus consist of two small tubercles fixed at the anterior side of the head, and no trace of a rostrum is to be seen. The first and five following pairs of peraeopoda are developed, all simple and sub- similar in form, and armed with strong claws, the first and second pairs subequal in length, and more than half as long as the following pairs. No trace of the seventh pair is to seen. The pleon and urus are com- paratively short, much shorter together than the perseon. The pleopoda are represented by egg-shaped vesicles. The second and third ural seg- ments are free, not coalesced, each carrying a pair of small vesicles, the uropods (fig. 86). fig. 86. A. young, just hatched, of Xiphocephalus Whitei. IV. THE OXYCEPHALIDEA^ GE1SERA A1SD SPECIES, As I have already said I think it convenient to divide the Oxyce- phalids into two families, instituting a new family Xiphocephalidse for the genus Xiphocephalus. I separate this genus from the other Oxy- cephalids, because it is distinguished from the other genera by charac- teristics which in my opinion have a more than generic value. So for instance Xiphocephalus has the seventh pair of pereeopoda rudimen- tary, arid consisting of only the femur, while the other genera have the same pair complete 1 ); Xiphocephalus normally has the telson free from, and articulating with, the last ural segment, while all the other genera have the telson fused with the last ural segment; and lastly the female of Xiphocephalus wants ovitectrices, and carries the eggs and young fixed under the peraeonal segments in two regular rows, while the other genera have well developed ovitectrices. On the other hand the two families are connected by some common characteristics, as for instance the reduction of the mouth-organs, the maxillae being rudimentary in both families, and the rostral elongation of the forepart of the head, and the more pronounced tendency to form an elongated body than is seen in any other family of the tribe. This tendency to elongation is besides expressed in the form of the femora of the fifth and sixth pairs of perseopoda in Xiphocephalidae, but not met with in these parts in Oxycephalidae, not even in the genera Leptocotis and Calamo- rhynchus, which show much elongated uropods. DANA was the first (17, p. 1009, see above, p. 5) who took the Oxycephalids as a whole placing them as a subfamily, Oxycephalince, of the family Typhidce, at the side of the subfamilies Typliinoe and Pro- noince; he then gave the following diagnosis of the Oxycephalince: Abdomen in ventrem se non flectens. Caput oblongum, antennis Imis superficiem capitis inferiorem insitis)). 1) Except Tullbergella which has the femur and three following joints. THE FAMILY OXYCEPHALID^E. 45 He distinguished the two genera Oxycephalus and Rhabdosoma from one another by the length of the rostrum and of the uropods. In 1862 SPENCE BATE (20, p. 341) made the subfamily a family of its own, and in 1885 CARUS (33, p. 427) considered it as a subfamily of the family Platyscelidce. The family OXYCEPHALIDJE, SPENCE BATE, 1861. Diagn. The head is more or less elongated and rostrate. The eyes are large, but do not occupy the whole head. The first pair of antennae are curved (