LRAT} kK RNa et inl) ‘ SS aete reer payee aye haa oe Ts ONz) HAA A a ies Wi +o EOE tye ot mt Key rise Vi yd bet ot Wh a tg Taya JOURNAL OF THE LINNEAN SOCIETY. ZOOLOGY. VOL, XXXIV. 27/ $4.2 LONDON: SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON HOUSE, PICCADILLY, W. 1, AND BY LONGMANS, GREEN, AND CO., AND WILLIAMS AND NORGATE, 1918-1922. Dates of Publication of the several Numbers included in this Volume. No, 225, pp. 1-93, published May 22, 1918. » 226, ,, 95-188, % January 23, 1919. » 227, ,, 189-259, » March 30, 1920. » 228, - ,, 261-843, ‘ December 7, 1920. », 229, ,, 845-456, , April 28, 1921. » 230, ,, 457-509, (Index, Title, etc.), ‘i February 16, 1922. PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET. ES OP APRS: Page ALEXANDER, W. B., M.A. The Vertebrate Fauna of Houtman’s Abrolhos (Abrolhos Islands), Western Australia. (Communicated by Dr. W. J. Darn, Professor of Zoology in the University of Liverpool.)............ 4 Bipper, G. P., Se.D., F.L.S. The Fragrance of Calcinean Sponges and the Spermatozoa of Guancha coriacea and Sycon raphanus. (Plate 24.) ... 2 — Synerypta spongiarum, nova. (Text-figures.) .....0...6.0. ccc cece 3 — Notes on the Physiology of Sponges. (Text-figures.) 2.0.0... 00.00.00... 3 Cuapman, Freperick, A.L.S., F.R.M:S., Palwontologist to the National Museum, Melbourne; Hon. Pal. Geol. Surv. Vict. Sherbornina: A new Genus of the Foraminifera from Table Cape, Tasmania ............ 5 Cotuince, Watrer E., D.Sc., F.L.8. Description of Paracubaris spinosus, a new Genus and Species of ‘l'errestrial Isopoda from British Guiana. (later errs ceaeeat wettest canteen guano at mercer oer Nore ne ae ame Sitisat ae — On the Oral Appendages of certain species of Marine Isopoda. (HARRIS YU)! -bacébobaspasdoneighonadbacnasaconaoadedatanbondsGoatsectaocerocsecornenn Daxty, W. J., D.Sc. F.L.S. The Perey Sladen Trust Expeditions to the Abrolhos Islands (Indian Ocean). Report I.—Introduction. General Description of the Coral Islands forming the Houtman Abrolhos Group. The Formation of the Islands. (Plates 10-14, and 12 Text- HEUGESS) eestor mene sen seesics cae csiyealeceet cb ldiisteyatarotersielarelefaine s etleiaisiie sugiosieteisiolersers 06 1 —— Sce ALEXANDER, W. B.; FaAvvet, P. Drassie, Eric, D.Sc., F.L.S., see Hoesen, L. T. Evans, Auwen M., M.Se. (Manch.). On the Structure and Occurrence of Maxillule in the Orders of Insects. (Communicated by Dr. A. D. Ines; "Haes.)) (Plate: stand 17 Mext-Geures:).c......00 svsetsrseessueses 426 DT 99 Ol 61 Lo “I Fauve, Pierre. Annélides Polychétes de VArchipel Houtman Abrolhos (Australie Occidentale) recueillies par M. le Prof. W. J. Dakin, F.L.S. (Communiqué par le Professeur W. J. Dakin, D.Sc., F.L.S.) (Avec Ds Page 2ifigures dans le: textes) oni saceneceeracieeety asst litsare cnt eee eee 487 Garensy, J. Bronrt, B.A., B.Sc., D.Phil. The Germ-Cells, Fertilization, and Early Development of Grantia (Sycon) compressa. (Communicated by Prof. E.S. Goopricn, F.R.S., Sec.L.S8.) (Plates 19-23, and 4 Text- IIT 400 91) ARERR AR EE Mr Sn aAda cd nacre lawl andednuccmmrariaatacosumedosocodsac 261 Gitcurist, J. D. F., M.A., D.Se., Ph.D., F.L.S8. A Post-Puerulus Stage of Jasus lalandii (Milne Edw.), Ortmann. (Plates15 & 16, and 13 Text- fISUTESs) 2 ran seiadeae seeaclasmasine -eReE ARENT ce CRE ROMEL Aes erin tai caia eeeL tee eee eee 189 GoopricH, HE. 8., F.R.S., Sec.L.8. Restoration of the Head of Osteolepis. (GMextatiouress) ser acu. maetentenno ene bencsncne cee hi acnademeinuish scuisn cosentaneme 181 — On a new Type of Teleostean Cartilaginous Pectoral Girdle found in young’ Clupeids) ac. <-ci cms osasceseranacceet ssl fy coameneern reese TOokee Aone 505 —— See Gatensy, J. B. Haswett, W. A., M.A., D.Sc., F.R.S., F.L.8. The Hxogonee. (Plates 17 & 18, and 2) Rexbehouress) 4.2. ueaateunes saencdrtiacecers hie aeons aCe EEE 217 Herpmay, W. A., F.R.S., F.L.8. Spolia Runiana, IIT., 1V. (With 6 Text- HSULES) se scicisatyisaese aie oeeisactinrose te shi Cen Maree eee eee aera eee 95, 247 —— See Tarrersaty, W. M. Hoesen, Lancetor T., B.A., B.Sc. On certain Nuclear Phenomena in the Oocytes of the Gall-fly Meuroterus. (Communicated by Dr. E. Drassze, F.L.8.) (With 2 Text-figures.) ..........0..0 22.5 ceeceee vee 327 Inns, Dr. A. D., see Evans, ALWEN M. Kernnarp, A. §., F.G.S., and Woopwanrp, B. B., F.L.S. On the Linnean Species of Non-Marine Mollusca that are represented in the British Fauna, with Notes on the Specimens of these and other British Forms nn the imnneaniCollectronsstse <5 copra eee eee eee eee eee 203 Morrram, J. C., M.B. (Lond.). Some Observations on the Feeding-habits of Fish and Birds, with special reference to Warning Coloration and Miniecry. (Communicated by Prof. E. B. Pouuron, F.R.S., F.L.S.) Gre Teib eB 3) SSO BCLS OAS ag REA a LEE Nate Une Be ER 47 Poutton, E. B., F.R.S., F.L.S., see Morrram, J. C. Sreppine, Rev. T. R. R., F.R.S., F.L.S., see Unwin, E. E. Tarrersaty, Waurer M., D.Sc. (Vict.). Report on the Stomatopoda and Macrurous Decapoda collected by My. Cyril Crossland in the Sudanese Red Sea. (Communicated by Prof. W. A. Herpmay, F.R.S., F.L.S.) (Bla test2 75.282) mrcape mt sae theo Dn oss avaste aie assaeiee e Mumm eaten seer oi 345 Unwin, Ernest E., M.Sc. (Leeds). Notes upon the Reproduction of Asellus aquaticus. (Communicated by the Rey. T. R. R. Srespine, M.A., HERES ae Hel Saye (elatesta DweOa)eersicacccsc-enetesases oe peneete dar keee re caecee 335 Waters, ArrHuR W. Some Collections of the Littoral Marine Fauna of the Cape Verde Islands made by Cyril Crossland, in the Summer of 1904,— Bryozoa. (Plates 1-4, and 2 Text-figures.) ......00..00......0.000, 1 —— Observations upon the Relationships of the (Bryozoa) Selenariade, Conescharellinide, ete., Fossil and Recent. (Plates 29, 30.) ............ 399 Woopwarp, B. B., see Kennarp, A. 8. 13, 14. 15, 16. 17, 18. 19-23. 24. " 25, 26. 27, 28. 29, 30. 31. 32. vi EXP LAN AT RONG Ops clsEve Cape Verde Islands. (Bryozoa.) Artificial Insects (Halford’s Patterns). Paracubaris spinosus. Oral Appendages of Isopoda. Limestone Cliff, Turtle Bay. Lesser Noddy nesting on Mangroves, Noddy Terns. Abrolhos Islands. Jasus lalandit. Australian Exogonee. Grantia (Sycon) compressa. Spermatozoa in Sponges. Asellus aquaticus. PLATES. Stomatopoda and Macrurous Decapoda from the Red Sea. Bryozoa. Petrobius sp. Sherbornina atkinsoni, F. Chapm. ETE EG Ac Den Page 48, line 26, for M, danica read LE. danica. 155 ,, 18 from bottom, for oae read one. GY A oe at BI airs Ki for hewever read however. 212 ., 26, for rotundata, Poiret, read leucostoma, Millet. » 9 from bottom, for ratdina read navidina. BIO) a a Sustifeva read fustifera. 236 ., 13, for naudina read natidina. 954 ,, 3, for StyErLopsts read STYELA. »—o9y,—«15, 18, 22, for Styelopsis read Stycla. 293 ., 15 from bottom, for 63 read 64 pt. 3. 404 ny 17, for Haloporella read Holoporella, pay ‘ oo y boating ff Rear Sie Mb renthane tig ; ees ~ Sian aes hati! " yt KOE Siar ie, May 22, 1918. Price 20s- —— AV HOR J: URENSAMIEI eo) 067, aA OF NOV H 5) 19794 THE LINNEAN SOCIETY Vou. XXXIV. ZOOLOGY. No. 225. CONTENTS. Page I. Some Collections of the Littoral Marine Fauna of the Cape Verde Islands made by Cyril Crossland, M.A., B.Sc., F.Z.S., in the Summer of 1904.—Bryozoa. By ArrHur Wu. Warners, F.L.S., H.G.8. (Plates 1-4, and 2 Text-figures.)..............0..setec0000s 1 II. Some Observations on the Feeding-habits of Fish and Birds, with special reference to Warning Coloration and Mimicry. By J.C. Morrram, M.B. (Lond.). (Plate 5.) (Communicated byglgroteelie Ber POULTON HiFu. Ses) Hela: cs) amen rene. ceeeee «oe 47 III. Description of Paracubaris spinosus, a new Genus and Species of ‘Terrestrial Isopoda from British Guiana. By Watter E. Cotuinesr, D.Sc., F.L.S., etc., Carnegie Fellow, and Research Fellow of the University of St. Andrews. (Plate 6.)............ 61 TY. On the Oral Appendages of certain Species of Marine Isopoda. By Water E. Coutiner, D.Sc., F.L.S., ete. (Plates 7-9.).. 65 LONDON: SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON HOUSE, PICCADILLY, W., AND BY LONGMANS, GREEN, AND CO., AND WILLIAMS AND NORGATE, 1918. LINNEAN YSOCIERY (OF LON DION: LIST OF THE OFFICERS AND COUNCIL. Elected 24th May, 1917. PRESIDENT. Sir David Prain, C.M.G., O.LE, F.R.S. VICE-PRESIDENTS. James Groves, Esq. Dr. D. H. Scott, F.R.S. Horace W. Monckton, F.G.S. Dr. A. E. Shipley, F.R.S. TREASURER. Horace W. Monckton, F.G.S. SECRETARIES. E. 8. Goodrich, M.A., F.R.S. | Dr. A. B. Rendle, F.R.S. GENERAL SECRETARY. Dr, B. Daydon Jackson. COUNCIL. Mrs. Arber, D.Sc. Miss G. Lister. Edmund G. Baker, Esq. Gerald Loder, M.A. W. Bateson, F.R.S. Horace W. Monckton, F.G.S. E. T. Browne, M.A. Dr. G. E. Nicholls. R. H. Bume, Esq. Prot. H. G. Plimmer, F.R.S. A.D. Cotton, Esq. Sir David Prain, C.M.G., C.LE., F.R.S. Sir Frank Crisp, Bt. Dr. A. B. Rendle, F.R.S. EK. S. Goodrich, M.A., F.R.S. Dr. D. H. Scott, F.R.S. James Groves, Esq. Dr. A. E. Shipley, F.R.S. Dr. B. Daydon Jackson. Lt.-Col. J. H. Tull Walsh. LIBRARY COMMITTEE. The Officers ex officio, with the following in addition :— E. G. Baker, Esq. Dr. W. G. Ridewood. W. Bateson, F_R.S. Dr. E. J. Salisbury. i. T. Browne, M.A. C. E. Salmon, Esq. Gerald Loder, M.A. | B. B. Woodward, F.G.S. R. I. Pocock, F.R.S. | OF THE LINNEAN SOCIETY, (ZOOLOGY.) “Some Collections of the Littoral Marine Fauna of the Cape Verde Islands, made by Cyrin Crossianp, M.A., B.Sc., F.Z.S8., in the Summer of 1904. —Bryozoa. By Arraur Wu. Warers, F.L.S8., F.G.S. (PrLatrs 1-4, and 2 Text-figures.) [Read 16th November, 1916. ] THE collections, kindly submitted to me by Mr. Cyril Crossland, were made from Boa Vista, Cape Verde Islands (1); from underneath a coal-lgiter in St. Vincent Harbour, marked (2); dredged, in 10 fathoms, in St. Vincent Harbour from a bottom of foraminiferous sand and nullipore nodules—also clinkers, &e. (3). This communication completes the description of Crossland’s tropical collections of Bryozoa. Two papers have been published by the Linnean Society on the Red Sea material, and two by the Zoological Society on that from British East Africa and Zanzibar. These last are referred to as “ Bry. from Zanzibar.” Specialists have already described parts of the collections in the pages of this Journal and in those of the Zoological Society’s ‘Proceedings.’ This Cape Verde collection was made with the aid of a grant from the Carnegie Trustees. My plan, for a long time, has been practically to complete any paper by comparison of my own collections, and then to check my results in the LINN. JOURN.—ZOOLOGY, VOL. XXXIV, 1 B MR. A. W. WATERS ON THE MARINE Natural History Museum, especially regarding types ; and I have again to thank Mr. Kirkpatrick for giving me facilities on these periodical visits. The literature on the Bryozoa of the neighbouring seas includes :— Surrr.— Floridan Bryozoa,” Kongl. Svenska Vetensk.-Akad. Handl. vols. x. & xi. 1872-73. Busx.—‘‘On some Madeiran Polyzoa,” Quart. Journ. Micr. Se. vols. vi., vii. 1858-1859 ; “ Catalogue of Polyzoa collected by J. Y. Johnson at Madeira,” op. cit. vol. viii. 1860-1861. Hincxs.—‘ The Madeiran Polyzoa,” Ann. Mag. Nat. Hist. ser. 5, vol. vi. 1880, p. 69. JouNsoN, J. YaTE.—‘‘* New Cyclostomatous Bryozoa found at Madeira,” Ann. Mag. Nat. Hist. ser. 6, vol. xx. 1897, p. 60. Waters.—“ Bryozoa from Madeira, &e.,” Journ. R. Mier. Soc. 1899, pp. 6-16. Norman.— Polyzoa of Madeira,’ Journ. Linn. Soc., Zool. vol. xxx. 1909, p. 275. Ospurn, R. C.—‘ The Bryozoa of the Woods Hole Region,” Bull. Bur. of Fisheries, vol. xxx. 1912; ‘* Biological Survey of the Waters of Woods Holeand Vicinity,” op. cit. vol. xxxi. pt. 1, p. 102, pt. 2, p. 595. 1913 ; ‘The Bryozoa of the Tortugas Islands, Florida,’ Publ. 182, Carnegie Inst. of Washington. 1914. Points of Special Interest. (1) The discovery of Tubulipora (Proboscina) Lamourouaii, Audouin, which has never been understood, so that the generic name Proboscina has been incorrectly used by some authors for forms allied to Stomatopora: p. 31. (2) The way in which certain species of Schizoporella grow in layers, with the younger layer taking definite positions, frequently passing over the opercula. Together with this, the closures of the Cheilostomata are con- sidered: pp. 15 & 16. (3) The spines surrounding the zowcia of Cribrilina radiata, Aud., and the formation of the zocecia are considered : p. 11. (4) A very interesting Lichenopora with confluent zoaria is more fully described. Some subcolonies are multiserial, others uniserial, and it is difficult to decide whether they should be considered as one or two species : p. 33. (5) Several points relating to the classification of Crista are dealt with : p. 24. The collections made by Mr. Cyril Crossland consist of 45 species or varieties, of which 25 were already known from the Atlantic, 16 are British, 24 Mediterranean, and probably 17 are Australasian. Besides these, a list of 41 species already described from the Cape Verde Islands is given, BRYOZOA OF THE CAPE VERDE ISLANDS. A Table of Distribution from West to East. . a 3 3 oye Rid /O].|/8 SB /S/8 8) 8) ee te o i=} — pio 3 o am — wo] S |H|(Slcolalwlala ese Se | cote |e Oa coe leeae oul mame A }/4 Ala) a JOl|4/e PA CLEC MECLOsSITINCKS sere sess cceses sia 5+ 5] + {+} + + Japan, E. Africa, Amboina, Lifu, Bugula dentata, Lamx. ..........6.... 5 | + Tarn. Lea Scrupocellavia Bertholletii, Aud. ...) 5 | + +)+ Canary Is., 8. Afr. a Macandrei, Busk...... 6] + ]..-/4]..) + + Zanzibar. UGENLALA, Spo NOVeses|) vel) |Pscz \lneees| eee |\ecet|! eres Ile Beania hirtissima, Heller ............ Sales a le ale tonic ye Membranipora tenuirostris, Hincks.| 9 | + |+)}+]. +/+ i curvirostris, Hincks .| 9]... {+ ]+ quadricornuta,sp.nov.| 9 | ... |... |e | eee] oo ae Cribrilina radiata, Moll.............+. lO} +/+) +...) + ]...| +] +) Mauritius, Brazil. Thalamoporella Rozieri, Aud Af 1B] A P+] A] +] A yA] e it Schizoporella unicornis, Johnst. ...... 4/4 /+/4+/+ +]... Japan. . spongites, Pallas ...... UGS leee | foees|freeey | oee lessen | oe 4 oligopus, Robertson ...| 18 | ... |...|.2.|...| «+ Jee]... |...| California. trichotoma, sp. nov. ...| 19 | ... |... | «+ oo Bes fees | Osthimosia avicularis, Hincks ...... 19 |...) +] + + wef Ht Hippothoa distans, MacG. ...........- 20]... ;+]+ + |/+]+/+) S. Africa. x divaricata, Lamx. | 20) + JI + + +/+] S. Africa. " hyalina, Linn, 20; + )+}+ fee eters Arthropoma Cecilii, Aud, .......1.65- 20) + J} +) + /+?]4+ A +)+)+ Chorizopora Brongniarti, Aud. ...... PAI ets | be earl eed | aca +/+ Smittina trispinosa, var. protecta,Th.| 21 | ... |... [...) + IN. +) + | tropica, Waters PAL NF eee don {ate | Lepralia cucullata, Busk 21) + +/+ +| California. »» peristomata, Waters ..... ...| 21 | + | Schizotheca fissa, Busk............+000+8 227)... Ve Holoporella pusilla, Smitt .........6++ 22) + a aperta, Hinoks ............| 28 | + + Microporella ciliata, Pallas............ 23) + ]+)+]4+)/+yj+]+]+] Japan. Adeonella contorta, Mich. ............ 23) + secu tell ieteeelhece loess | eect me Braz le Crisia tubulosa, Busk .............000+ 26 | + (+ | , denticulata, var.verdensis,noy.| 27 | ... |... | + » sigmoidea, Waters ..........- PAN cee lane Par » vincentensis, 8p. NOV. Be (2 Neer eee Reel bene | lerapon kistioal Petind fier | Tubulipora pulchra, MacG. ASO lecce [tees |tesetl ieee econ [essa (=a en mus n@alitornias | . Tamourouxit, Aud. ..s2-.) ol | 2. |... | =-- eealines |e a incrassata, Smitt.......... O24 |eace | eene| keke Lichenopora irregularis, Johnson ...| 83 | + 7H yp yar. composita, M1 OWasee oo | esem [tae [i ss Flustrella hispida, Fabr. ........ Ba Bist laces a i | Amathia tortuosa, T. Woods .. OOu| eset [teas == |e cein| +: ey Vidovici, Heller ............| 41 | + sl eaalt verse Roscoff, B.E. Africa. Zoobotryon pellucidum, Eh. ......... 41) + +/+) + + | Bowerbankia pustulosa, var. alter- | nata, nov. Pea Keri cack Keel tes 4 P05 [oa | ; i. Barentsia discreta, Busk...........06.+ 2 |+ a sep teael lees [teen tees teeta tose] BE eas Hiatal | Pedicellina cernua, Hincks ............ 43} + J+] +]... ... Jee] + * Previously known from the Atlantic, 1* 4 MR. A. W. Busk WATERS ON THE MARINE described a few species from the Cape Verde Islands in the ‘Challenger’ Reports, also a few in his British Museum Catalogues, and Calvet described a considerable number in the “ Bryozoaires du ‘ Travailleur’ 299 et du ‘ Talisman’ ” ; Atlantic. British. Mediterranean. Red Sea. Indian Qcean. Ceylon. | Japan. Australian. | Fossil. Caberea ligata, Calv. Cellaria biseriata, Maplestone Beania magellanica, Busk Membranipora patellaria, Moll....... bellula, Hincks cervicornis, Busk...... granulifera, Hincks.. ie (?) Alam, Jull.......... Onychocella angulosa, Reuss Setosellina Roulet. Caly. .........40:.4: ? Schizotheca Talismani, Calv. 4 Smittina carvicornis, Pallas i jacobensis, Busk Porella levis, var. subecompressa, Busk Trupostega venusta, Norman Haswellia alternata, Calv. Porina borealis, Busk beer Lepratia brancoensis, Caly. ........+.+. s Poissonii, Hincks Ichthyaria avicularia, Calv. Mamillopora simplex, Kosch. (Mamit- ” ” ” i Edwardsi, Jull. ? Myriozoum strangulatum, Calv. ... Retepora Imperati, Busk a ramulosa, Calv.... Cellepora intricata, Calv. Cupularia Oweni, Gray st umbellata, Def. ........-..- Adeona Heckeli, Reuss. (Microporella’ violaced) ......+5 Adeonella insidiosa, Jull.............655 Crisia ramosa, Harmer (elongata, (Norm) yecee sees a conferta, Busk........... ¢ tenella, Caly. Entalophora raripora, VOrb. .. Os reticulata, Calv. Be subverticillata, Calv. .. Idmonea atiantica, Forbes ............ Filisparsa superba, Jul. ......... 2.242. * Hornera brancoensis, Calv. Stomatopora granulata, M.-Edw. ... lopora Smittiz, Caly.)} ... +++ 4+ P+4+ ! +: me South Africa, South America. Ss. Aricas .|4 Amboina, Antarctic. S. America, Marion Isl. Lifu, Mauritius. N. Europe. China, ‘Tahiti, Mauritius, Zanzibar. Kerguelen, Cape Horn. S. Pacific? Magellan, Tristan da Cunha. * Hornera eburnea, Calvet, is mentioned by Calvet from Cape Verde Islands. ‘Bryozoa from Zanzibar,” Proc. Zool. Soc. 1914, p. 836, I said it was doubtful why it was placed with Hornera, and since then Prof. 8. J. Hickson has shown that it is}(Stylasterina) Hydrocoralline. so that the following are also known from the locality. A few species are omitted where there was doubt about the determination. In BRYOZOA OF THE CAPE VERDE ISLANDS. 5 AETEA RECTA, Hincks. BuGua DENTATA (Lamourour), Busk. (Plate 1. fig. 1.) Achamarchis dentata, Lamx. Hist. des Polyp. corall. p. 185, pl. 3. fig. 3 (1816). For synonyms see Miss Jelly’s Catalogue, and add :— Waters, “ Bry. N.S. Wales &c.,” Ann. Mag. Nat. Hist. ser. 5, vol. xx. (1887) p. 91, pl. 4. fir. 14; Kirkpatrick, Proc. R. Soe. Dublin, n. s. vol. vi. (1890) p. 614; Ortmann, “ Die Japan. Bry.,” Arch. f. Naturgesch. vol. i. (1890) p. 25, pl. 1. fig. 20 ; Philipps, “ Rep. on the Polyzoa coll. by Dr. Willey, from the Loyalty Isl. &c.,” Willey’s Zool. Results, pt. 4, p. 443 (1899) ; Thornely, “ Mar. Poly. of the Indian Ocean,” Trans. Linn. Soc., Zool. vol. xv. (1912) p. 141; ‘Calvet, “Bry. d’Amboine,” Rev. Suisse de Zool. vol. xiv. (1906) p. 617, pl. 21. fig. 1; Norman, “ Polyzoa of Madeira,” Journ. Linn. Soe., Zool. vol. xxx. (1909) p. 285, pl. 36. fig. 3. Krauss speaks of the zoarium being “ bleifiirbig und dunkelgraue ”’ ; Macgillivray says greenish or leaden colour ; Busk says “ ovicell blue” (but did he not mean to refer to the zoarium ?) ; Kirkpatrick says “the zoarium is of a leaden-blue colour” ; Miss Philipps, in her Loyalty Island paper, says “‘zoarium leaden-blue” ; and Miss Thornely, in her Indian Ocean paper, says “ ovicells not blue” ; Calvet did not tind any colour, but says that the specimen had been many years in spirit. No one, however, seems to have noticed the curious fact that the polypides and growing tissue are blue, looking just as if they had been stained, and in these Cape Verde Islands specimens it is a fairly bright blue. The specimens left in spirit have lost their colour, and one set of spirit-specimens were always without colour. A spectroscopic examination of this species would surely be interesting The geographical range is very great, occurring in many places from Japan to the Cape Verde Islands. There are about 14 tentacles. Loc. Australia, New Zealand, Tasmania, 8S. Africa, Torres Str., 6-7 fath. ; Lita (Lhornely), Amirante, 29-34 fath. (Zh.), Amboina (Calvet), Sagamibai, Japan, 70 fath. (Ortmann), Madeira (Vorman); Boa Vista, Cape Verde Islands, 20 fath., collected by Crossland. ScRUPOCELLARIA BeRTHOLLETII (Audouwin), Hines. For synonyms see Waters, “ Mar, Biol. of the Sudanese Red Sea, Bryozoa,”’ Journ, Linn. Soe., Zool. vol. xxxi. (1909) p. 188, and add :— Scrupocellaria Bertholletii, Norman, ‘‘Polyzoa of Madeira,” Journ. Linn. Soc., Zool. vol. xxx. (1909) p. 283, pl. 36. figs. 1,2; Calvet, Expéd. Sc. du ‘Travailleur’ et du ‘Talisman,’ Bryozoaires, vol. viii. (1907) p. 376. There are some specimens ‘closely attached to Adeonella contorta, Mich., with the creeping growth and attachment by radicles, as known in S. reptans (Linn.), Gray. On account of the resemblance of the two species [ at one time spoke of S. reptans, var. Bertholleti. There are some very large srect median avicularia, much larger than I have seen in any other specimens of S. Bertholletii. Hincks bas mentioned MR. A. W. WATERS ON THE MARINE Cow the irregularity of the median avicularia, and this I can confirm from Mediterranean specimens, in which some parts are without them, and others have them fairly regularly. Loc. Naples, Capri, Rapallo, Trieste, Suez, Madeira; Cape Verde Islands, 110-180 met. (Calvet), and St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. ’ ScrupocELLaRIA Macanprel, Busk. (Plate 1. figs. 3-7, 11.) For synonyms see Waters, “ Bry. from Zanzibar,” Proc. Zool. Soc, 1913, p. 477, pl. 58. figs. 5, 6. This has a separable operculum (fig. 3), which seems to be general in Caberea, but not in Scrupocellaria. The oral parts are thrown back as in C. Darwinii, Busk, and the scutum comes up to the oral aperture, much as in that species. The vibracular chambers are larger than is usual in Serupocellaria and spread diagonally halfway across the dorsal surface, but they are not as large as is usual in Caberea. The seta of the vibraculum is smooth as is general in Scrupocellaria*. It will be seen that this species has some characters of Caberea, showing how closely that genus is allied to Scrupocellaria. The seta is serrate, or, more strictly, spinous at one side only, in C. grandis, Hincks, C. lata, Busk, C. rostrata, Busk, C. Ellisia, Flem., C. Boryii, Aud., C. Hookeri, Busk, C. glabra, MacG., C. bursifera, Ort., C. Darwinii, Busk, C. minima, Busk. The vibracular seta has several irregular projections at the base, and at each side there is a separated curved chitinous piece, against which the vibraculum evidently works. The central vibracular muscles draw down the seta, and the sac-like “peculiar body ” naturally varies in position with the movements of the seta; but, besides the large + muscles, there is another muscle placed further away from the beak (fig. 4, m.), attached to the flexible wall proximal to the seta, and thus the chamber-wall is drawn in, reminding us of the way in which the front membranous wall is drawn down in Micropora, Steganoporella, Cupularia, and other genera. The lateral muscles are attached to projections on the base of the vibracular seta (fig. 5). There is a small median avicularium placed near to the base of the scutum. Miss Thornely refers to tall median avicularia. There is a chitinous arch from the inner side of the area below the scutum. ‘The ovicell is imperforate, and there are 13 tentacles. Loc. Coast of Spain; Adriatic ; Lifu (Zhornely) ; Queensland (Haswell) ; Zanzibar, 8 fath. (Waters) ; in the Indian Ocean (Providence, 50-78 fath., Amirante, 29 fath., Farquhar reef, Cargados, 30 fath., Seychelles, 34 fath.) * In my paper on the Zauzibar Cheilostomata, Proc. Zool. Soc. 1913, p. 474, I unfor- tunately stated that the vibracula of Caberea were smooth, which presumably was a slip in transcribing. + With the circular movement of the seta it is questionable whether any of the muscles should be called retractor or occlusor BRYOZOA OF THE CAPE VERDE ISLANDS. 7 (Th.); St. Paul’s Rock, N. Atlantic (Chall.); St. Vincent, Cape Verde Islands, 1070-1150 fath. (Chall.), Cape Verde Islands, 110-180 met. (Calv.) ; Boa Vista, Cape Verde Islands, collected by Crossland. SCRUPOCELLARIA TRIDENTATA, sp. nov. (Plate 1. figs. 9, 10.) Zoarium spreading in all directions, forming a circular growth, in most respects agreeing with S. cervicorns, Busk, but below each bifurcation there is a very large raised avicularium, of which the mandible has three long narrow prongs, and the avicularian chamber folds inwards round three projections to receive these prongs as in a half-open tube (PI 1. fig. 10). In other parts there is an anterior avicularium on each zocecium, just below the area*. Hincks describes avicularia below the area on a tall columnar process, in S. cervicornis, Busk. Fig. 1. Scrupocellaria tridentata, sp. nov. x 50. There are about six oral spines, none of which are bifurcate ; down the middle of the zoarium there are very long stout erect spines, and these in one internode will all belong to the zocecia on the right side, whereas in the companion internode of the same age they will belong to the zocecia on the left side, in both cases growing from near the distal end. Hincks refers to tall spines on each side of the branch in S. cerwicornis, and when a colony of S. tridentata is seen from the side the spine appears to grow from the side. The lateral avicularia are very minute, being smaller than those figured by Smitt 7 in his S. cervicornis. Smitt calls attention to the variation in the size of the lateral avicularia in S. cervicornis, and also makes comparisons of S. pusilla, Smitt, S. cornigera, Smitt, and S. cervicornis, Busk. MacGillivray figures the Australian form with large avicularian mandibles. This species, S. cervicornis, Busk, S. cornigera, Smitt, and S. obtecta, Hasw., have very * Hincks, Ann. Mag. Nat. Hist. ser. 5, vol. xi. (1883) p. 103. + “Floridan Bryozoa,” Pt. I. p. 15 (1872). fo) MR. A. W. WATERS ON THE MARINE similar scuta. There are about seven zocecia in an internode. There is no separable operculum, the ovicell is perforate, the radicles are serrate, and the smooth vibracular sete are about three times the length of a zocecium. Loc. Boa Vista, Cape Verde Islands, 20 fath., collected by Crossland. Brania Hievissima, Heller. (Plate 1. fig. 2.) Diachoris hirtissima, Heller, “ Die Bryozoen des Adriat. Meeres,” Verhand. der k.k. zool.-bot. Gesellsch. Wien, vol. xvii. (1867) p. 94, pl. 1. figs. 6,7; Busk, Rep. of the Voyage of H.M.S. ‘ Challenger,’ vol. x. Polyzoa, p. 61 (1884). Chaunosia hirtissima, Busk, Q. Journ. Micr. Se. n. s. vol. vii. (1867) p. 241, pl. 36. figs, 12-14. Beania hirtissima, Waters, “On the Use of the Avicularian Mandibles in determination of the Cheil. Bry.,” Trans. Micr. Soc. ser. 2, vol. v. (1885) pl. 14. fig. 5; “Bry. from Rapallo, &e.,” Journ. Linn. Soc., Zool. vol. xxvi. (1896) p. 17; “ Bry. from Madeira,” Journ. R. Micr. Soc. 1899, p. 15; Calvet, “Bry. Mar. de la Rég. de Cette,” Trav. Inst. de Zool. de l’Univ. de Montpellier, ser. 2, Mém. 11, p. 24 (1902); Expéd. Se. du ‘ Travailleur’ et du ‘Talisman,’ Bryozoaires, vol. viii. (1907) p. 392; Norman, ‘ Polyzoa of Madeira,” Journ. Linn. Soc., Zool. vol. xxx. (1909) p. 286. In the specimens from the Cape Verde Islands there are no radicle-tubes, and none seem to occur in the typical form of this species from other localities, so far as examination has gone, but there are, in these Cape Verde specimens, delicate spines which sometimes are entire, but more often they bifurcate near the base, and usually there are two such spines on the dorsal surface of each zocecium. Busk, in describing his Chaunosia the same year as Heller with the same specific name, refers to the bi-trifurcate dorsal spines. The zoarium of Busk’s Chaunosia is like the form of cylindrica of Hincks. ‘In looking carefully through some specimens from Naples and Rapallo a few such spines, similarly bifurcating, were found, but only in isolated cases, whereas they are very numerous in the Cape Verde specimens, which in most respects correspond fairly closely with the Mediterranean specimens. On the anterior border there are about nine delicate spines on each side, while further away from the border, with no great regularity, there are a number of much stouter spines. Round the distal part of the aperture there are usually seven fairly stout spines with other spines nearer to the dorsal surface. The variety robusta, Hincks, which I think might well be separated as a species, has long thin dorsal tubes or radicles; 6. conferta, MacG., has numerous long dorsal spines: 6. spinigera, MacG., has no dorsal spine or tube; 8. elongata, Hincks, has a small dorsal tube near the distal end; B. magellanica, Busk, has a wide radicle-tube near the distal end, thus showing the value of the dorsal surface in determination. Loc. Adriatic, Naples, Capri, Rapallo, Villefranche-sur-Mer ( Waters), Ajaccio, (Calvet); Madeira ( Waters § Norman); New Zealand (var. robusta); St. Vincent, Cape Verde Islands (Chall.); St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. BRYOZOA OF THE CAPE VERDE ISLANDS. 9 MermpranNIPORA TENUIROSTRIS, inchs. Membranipora tenuirostris, Hincks, Anu, Mag. Nat. Hist. ser. 5, vol. vi. (1880) p. 70, pl. 9, fig. 3; op. cit. ser 5, vol. x. (1882) p. 7; op. cit. ser. 5, vol. xix. (1887) p. 314; Waters, Journ. R. Mier. Soc. ser. 2, vol. v. (1885) p. 775, pl. 14, fig. 41; Quart. Journ. Geol. Soc. vol. xlvii. (1891) p. 11; Journ. Linn. Soc., Zool. vol. xxi. (1898) pp. 668 & 685, pl. 47. fig. 7. Membranipora Flemingti, Waters, Aun. Mag. Nat. Hist. ser. 5, vol. iii. (1879) p. 122, pl. 18, fig. 2. In the specimens from the Cape Verde Islands some parts have the spines as figured by Hincks, whereas in others there are none. In my paper on Membraniporidee I referred to the spines not occurring in Naples and Rapallo specimens. Loc. Queen Charlotte Islands(Hincks); Madeira on Pinna (Hincks) ; Naples, low water to 40 fath.; Capri; Rapallo; Adriatic; Oran (Algiers), 54 fath. ; St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. Fossil. Bartonian and Pliocene of Italy. MEMBRANIPORA CURVIROSTRIS, [Hincks. Membranipora curvirostris, Hincks, Ann. Mag, Nat. Hist. ser. 38, vol. ix. (1861) p. 295 (89), pl. 7. fig. 4; Brit. Mar. Polyzoa, p. 153, pl. 20. figs. 5, 6 (1880) ; Ridley, Proc. Zool. Soc. 1881, p. 46; Waters, Journ. Linn. Soc., Zool. vol. xxvi. (1898) pp. 668 & 684, pl. 47. figs. 2, 12. Some of the zoccia have a median spine below the area, also one or two spines in various other positions. There is the unoccupied space above the avicularium as in the British and Naples specimens. Loc. British, Polperro, 40 fath : Singapore? ; Brazil, 35 fath. (Ridley) ; Naples ; St. Vincent Harbour, Cape Verde Islands, 10 fath., on clinker and from a diver, collected by Crossland. * [ Dwarka, Arabian Sea (Thornely). | MEMBRANIPORA QUADRICORNUTA, sp. noy. (Plate 1. fig. 8.) There are incrusting specimens from St. Vincent Harbour, Cape Verde Islands, 10 fathoms. The opesia is oval, and at the distal end on each side there is a short cervicorn spine, and lower down there is another pair. There are a considerable number of vicarious avicularia just like those of JZ. crassi- marginata, Hincks {—in fact, the species is like J/. erassimarginata, except that there are two pairs of cervicorn spines. The avicularia are about as long as the ordinary zocecia, but not quite so wide. ‘lhe border is crenulated. There is one distal rosette-plate about the middle of the distal wall, and two lateral rosette-plates about halfway between the anterior and basal wall. There is considerable irregularity in the spines, for the lateral ones at first have only two points, whereas later there may be several divisions, and those [* Thornely, Hyd. & Poly. coll. at Okhamandal in Kattiawar. 1916, + Ann. Mag. Nat. [Hist. ser. 5, vol. vi. (1880) p. 71, pl. 9. tig. 1. 10 MR. A. W. WATERS ON THE MARINE from opposite sides may interdigitate, or in a few cases may be fused together, thus forming a good protection. In some cases there is a second spine on each side near the proximal end, but this is not common; also in one zoccium I have found a central proximal spine. The oral spines stand erect, and therefore it is impossible to give an idea of them as seen from above—consequently a few pressed down are drawn to show the shape. CRIBRILINA RADIATA (J/oll), Smitt. The specimens from the Cape Verde Islands are fairly typical C. radiata, a species subject to much variation. They have about 11 ribs, and there are 5-7 pores inaray. ‘There is no suboral pore visible, but in dried specimens, afterwards decalcified, a row of 6 pore * disks is seen below the aperture with a single pore in the middle, the same size as the others, but this semicircle of pores must probably be considered as the upper series of the radial pores. On each side of the central pore there is usually a small nodule. By the side of the more or less oval zocecia there are elongate avicularia, as fre- quently described; as is generally the case in Cribrilina, there is no avicularian cross-bar, though one occurs in C. monoceros, Busk, which, how- ever, most workers would now remove from C7ribrilina on other grounds. The large bordering spines to which reference will be made are about 11, the two upper ones, being much longer than the others, have been mistaken for sete or vibracula. In a stained specimen, from Rapallo, the small frontal pores are each in the centre of a much larger disk. The ovicells have nodules, and sometimes a keel dividing to each lower corner. The opercula are separable and are of quite the same type as in C. pune- tata, Hass., C. Gattye, Hincks, C. Balzaci, Waters, and I fail to see any reason for separating a genus Puellina, Jullien, as maintained by Levinsen. The primary zocecium has 11 spines, while that of C. punetata, Hass., has 12 (Harmer), that of C. projecta, Waters, has 12, that of C. monoceros, Busk, has 9 (Harmer), that of C. Gattye, Hincks, has 11, while the closely allied C. Balzaci, Waters, has a primary zocecium similar to the ordinary zocecia, and this, according to Harmer, is also the case in C, annulata, Faby. The primary of Membraniporella nitida, Smitt, has 10 spines fairly regularly surrounding the zocecium, but it may be seen that 6 are oral spines and 4 zocecial. Hincks evidently made a mistake in describ- ing 14 spines, for his figure in Brit. Mar. Pely. pl. 27. fig. 6 is clearly not a primary. In J. nitida a number of zoccia, after the primary, have a Membraniporidan character somewhat like M/. lineata, Hincks, and taking a fairly typical colony of IZ. nitida from Oban, Scotland, the second series have * In speaking of pores it is not meant that there is a direct opening to the interior. BRYOZOA OF THE CAPE VERDE ISLANDS. ll 6 oral spines and two or three delicate zocecial ones. The following series have 4 distal spines and 2 oral. None of the subsequent zocecia have lateral spines, while in the next series (circle) there are two zocecia with two distal spines and one oral on each side; whereas two other zoccia in the same series have the flat spines characteristic of nitida and meeting in the median line. Thus the early zocecia are distinctly Membraniporidan, in each row approximating gradually to the mature nitida, There are several cases in the Cellularidee, and other families, in which the zocecia immediately following the primary are intermediate between it and the following zoccia. Hincks, Norman, and Levinsen have placed M/embrantporella under Cribrilinidee, and Norman * has written a lengthy description of the bars or spines, comparing them with other Cribrilinide, but what he describes as the loop is only the base of the spine arising from the lateral wall. His ficures (pl. 8. figs. 8, 9) are difficult to understand without examination of specimens and are far fromsatisfactory. The bars are simply spines such as we see in many Membranipore, and they do not touch their neighbours at the side continuously for the whole length. These spines are an arch over the frontal membranous wall with which they are not in contact, and the operculum, which is of a Membraniporidan type, is not separable, and is in the membrane in no way connected with the spines. We may ask if there is any reason for generically separating it from Jembranipora, and must answer, No--as Smitt has already done. The ridges of Cribrilina have been supposed to show that it was very closely allied to Membraniporella nitida, but the difference is much greater than first appearances suggest. In 1879 ¢ I mentioned minute spines round the area of C. Gattyv, Hincks, and since then Harmer? has described spines surrounding C. radiata §. I have made decalcified and stained preparations, and even decalcified preparations of dried specimens of various species, for they often show im- portant structures, and these preparations led me to doubt the theory of the frontal wall being formed of confluent spines. My decalcified preparations of C. radiata from several places show a row of bluntly pointed, erect spines round the border of the zocecium, with the two distal ones much longer and narrower than the others, and acute. Of course, these free erect spines have * “Natural History of Hast Finmark,” Ann. Mag. Nat. Hist. ser. 7, vol. xii. (1903) pl. 8. fig. 18, pl. 9. figs. 4, 6. + “Bry. of the Bay of Naples,’ Ann. Mag. Nat. Hist, ser. 5, vol. iii. (1879) p. 36, pl. 9. fig. 6 a. ¢ “On the Morphology of the Cheilostomata,” Q. Journ, Micr, Sci. vol. xlvi. p. 826. § Probably these must be compared with the large stout spines which I figured round Hippothoa Brongniartii, @Orb. in Ann. Mag. Nat. Hist. ser. 7, vol. xy. (1905) p. 10, pl. 1. figs. 2-4; and also with the spines of Lepralia Poissonit, Hincks, 12 , MR. A. W. WATERS ON THE MARINE nothing to do with the ribs (ridges). Iam thus able to confirm part of what Harmer has written on this point, though it may be somewhat differently stated, and the rows of small radia] pores are, as mentioned by Harmer, in line with these spines, and in between are the ribs. We have been accus- tomed to speak as if the ribs were the most important frontal structure, whereas it seems that it is the pores which are the more important, with the ribs produced later and of secondary importance, In specimens of C. Balzaca, and in other species the young zoocia, with only a membranous front, show the pore disks before any calcification takes place, and we may see in many forms that calcification has formed round these pore disks ; and Norman in his figures of C. nitido-punctata, Smitt, and C. Gattye, Hincks, shows a thicker calcareous deposit round the disks, though he did not appreciate the importance of his observation. Where there is a distinct area, as in C. Balzact, &e., and in C. figularis, Hincks, there is a surrounding row of large pore disks, often dark and chitinous, with smaller transparent pores, radiating from the centre, from these large disks. These large pores are probably vestigial, seeming to give a record of ancestors having had spines. The frontal wall of C. radiata oceurs right up to the spinous circle, and may be attached to the spines, but this does not justify our assuming that the frontal wall of either C. radéata or other Cribriline is formed by overarching spines, and itis the supposed resemblance to Membraniporella nitida which has led to this assumption. There are a great many species in various genera in which there is a more or less radial arrangement of the pores, and a fuller study of Cribrilina should explain various structures, for of course the spines and pores show a close developmental connection, even though former explanations may not be correct. The opercula measured range in size from C. radiata (form called setosa, Waters) 0°07-0°08 mm.; C. radiata, Cape Verde Islands, 0°08 mm.; C. Gattye, 0:07 mm.; C. radiata, some specimens, 0°l mm.; C. Balzacr, O-lmm.; C. punctata, Hass., 0°12 mm.; C. patagonica, Waters, 0°15 mm. ; C. figularis, Hincks, 0-17 mm.; C. latimarginata, Busk, 0°2 mm.; C. mono- ceros, Busk, and C. acanthoceros, MacG., 0°22 mm. C. patagonica, Waters, has on the under surface of the operculum pro- jections of a Membranipora type. Norman* still considers that there are two species, C. radiata and C. innominata, although most workers have united them, but in my own collection I am unable to find characters by which they can be separated. In British, Naples, and Madeira specimens, which Norman would prebably call innominata, there are frequently many avicularia, though some specimens may have few or none. The number of ribs varies from 3-20 ; the elevation of the so-called bars is not a satisfactory distinction. * ¢ Polyzoa of Madeira,” &c., Journ, Linn. Soe., Zool. vol. xxx. (1909) p. 291. BRYOZOA OF THE CAPE VERDE ISLANDS. 13 It is advisable to eall attention to the similarity between the primary zowcia described by Smitt in his Discopora coccinea *, D. ventricosa +, Smitt, D. stenostoma t, Smitt, and the mature zocecia in certain Cribriline such as C. Gattye, C. Balzaci, and to a certain extent C. figularis, Hincks. Loe. Very widely distributed. St. Vincent Harbour, 10 fath., collected by Crossland. Fossil. Wocene, Miocene, and Pliocene. THALAMOPORELLA RoziErt (Audowin), Hincks. (Plate 4. figs. 9, a, b, ¢.) For synonyms and localities see Waters, ‘ Bryozoa of the Sudanese Red Sea,” Journ. Linn, Soc., Zool. vol. xxxi. (1909) p. 141, pl. 15. figs. 12-15, and add :—? Robertson, “ Inerust.° Chil. Bry.,’ Univ. of California, Pub. Zool. vol. iv. (1908) p. 277, pl. 17. figs. 27, 27 a, 28, 29, pl. 18. fig. 30; Levinsen, Morph. & Syst. Studies on the Cheil. Bry. p. 181, pl. 6. figs. 6 a-G k, pl. 64. figs. 1 a-8 6 (1999) ; ? Osburn, “ Bry. of the Tortugas Islands, Florida,” Publ. 182, Carnegie Inst. of Washington, p. 196 (1914). Vincularia nove-hollandie, Haswell, Proc. Linn. Soc. N.S. Wales, vol. y. (1881) p. 41, pl. 3. fig. 3. Thalamoporella nove-hollandie, Levinsen, Morph. & Syst. Studies on the Cheil. Bry. p. 185, pl. 6a. figs. 8 a-8 f (1909). The form from the Cape Verde Islands is the same as that from the Red Sea, described by me in the “Sudanese Red Sea”? Report, which must be taken to represent the type well figured by Savigny, who apparently found no avicularia, and the operecular knobs were very pronounced; but in the Cape Verde Islands specimens there is great variation in this respect, and sometimes in the younger parts of a colony none are seen, whereas in older parts they occur as figured by Savigny. This is a very variable character throughout the family. Of the varieties figured by Hincks § the one most nearly approaching the Red Sea form is Steganoporella Rozieri, var. indica, Hincks, whereas the variety jalciferuw seems entitled to specific rank as proposed by Levinsen. Probably var. labiata, Levinsen, should also be given generic position as the type of avicularia is different. Most of the zocecia are about 0°65 mm. long, but some are longer, others shorter, and this is about the size of co-type specimens of 7. novee-hollandie, Hasw., sent to me by Haswell. These I have considered were 7. Rozieri, but as my specimens were without membranes they were not suitable for studyiny the spicules; however, examination shows a ‘‘ curve” (spicule), which Levinsen says does not occur. [tis somewhat larger than the average size in 7. Roziert from the Red Sea and Cape Verde Islands. My specimens of nove-hollandie ave cylindrical and hollow, while Levinson speaks of the colonies being partly incrusting, partly free hollow branches, and I see no * © Kyit. Foérteck. ofv. Skand. Hafs-Bry.,” Oty. Kongl. Vet.-Akad. Férh. (1867), Bih, pl. 27. fig. 163. + Ib. (1871) pl. 21. fig. 31. t Tb. (1871) pl. 21. fie, 29. § Ann. Mag. Nat. Hist, ser. 5, vol. vi. (1880) p. 379, pl. 16. fig. 1. 14 MR. A. W. WATERS ON THE MARINE sufficient ground for separating 7. nove-hollandie from Roziert. Further cleaning of one of my specimens, sent by Haswell, revealed an avicularium similar to the one figured by Levinsen. With ovicells and avicularia unknown, it was difficult to be certain of the specific identity. Levinsen’s studies of the spicules in Thalamoporella are most important, but we do not know much as to the positions in which they occur, nor as to their object. In 7. Roziert from both the Red Sea and Cape Verde Islands there are two bundles of the long straight spicules starting from near the opercula and passing diagonally by the walls of the opesiules to the lateral walls, near the basal wall (Pl. 4. fig. 9). The long straight spicules are mostly free in the zocium, and at one time I wondered whether they were arranged in this way by the polypide passing in and out of the zocecium, but these bundles occur in very young zocecia before any calcareous frontal wall or operculum is formed. There are on, or in, the membranes spicules which Levinsen called curves, but in the sponges they would be called sigmas or arcuate spicules, and perhaps it would be well to speak of them as sigmas. These are remarkably abundant in Th. Jervoisi, H., and 7. mamillaris, Lamx., from various localities. A specimen of Jervois¢, from (Queensland, in my collection, has these spicules in great abundance in both the basal and frontal membranes, and at the growing end where the beginning of the lateral wall is only just indicated, on the supporting seaweed, these sigmas are massed in abundance. Dr. Alice Robertson considers that her specimens, from La Jolla, California, are var. indica, Hincks, which hardly seems to differ from the type, others from San Pedro are articulated, and I expect this will ultimately be generically separated. Osburn considered his specimens from Tortugas Island were var. labiata, Levinsen, but the difference of the avicularium as figured by Levinsen may suffice to make labiataa genus. 1 donot consider that the presence or absence of «vicularia is of first class importance, as often some colonies of Thalamoporella may have them while others are without ; on the contrary, the type of avicularium and mandible is of material importance. Localities, additional. California, few fath.: San Diego (Robertson), Florida, Jamaica, Tortugas Bay ; Cape Verde Islands, St. Vincent Harbour, 10 fath., collected by Crossland. ScHIZOPORELLA UNICORNIS (Johnston), Pergens. (Plate 2. figs. 14-17, 22.) For synonyms see Waters, “ Bryozoa of the Sudanese Red Sea,” Journ. Linn. Soc., Zool. vol. xxxi. (1909) p. 143, pl. 12. figs. 12, 18, and add :— Nordgaard, Die Bry. des West Norweg. p. 86 (1903); Hydrog. & Biol. Invert. in Norwegian Fiords, p. 165, pl. 5. figs. 28-25, 27 (1905); Calvet, Expéd. Sc. du ‘ Travailleur’ et du ‘Talisman,’ vol. viii. (1907) p. 417; Norman, “ Polyzoa of Madeira and neighbouring Islands,” Journ. Linn. Soe., Zool. vol. xxx. (1909) p. 808; Osburn, “ Bry. of Woods Hole of the Tortugas Islands, Florida,’”’ Publ. 182, Carnegie Inst. of Washington, p. 205 (1914) ; Waters, “ Bry. from Zanzibar,” Proc. Zool, Soc. 1913, p, 501, BRYOZOA OF THE CAPE VERDE ISLANDS. 15 Crossland sent some thick pieces from underneath a coal-lighter, in St. Vincent Harbour, Cape Verde Islands, and says, “ this is the first of the host of forms to make an attachment to the lighter.”. One block was about 170 mm. X 140 mm. and 7 mm. thick. Some colonies occur on a stone brought up by a diver. In my last two papers * it has been pointed out that the zocecia are frequently superimposed, and this I referred to in a species of Meliceritites. Reuss has shown the same thing in Cumulipora angulata, v. Maehr, and Lee f calls attention to Ulrich f, saving, “in the Trepostomata and many Cryptostomata the tubular zocecium really represents a series of superimposed cells.” This species shows the “closures” more clearly and distinctly than any other I have come across. The calcareous matter is evidently deposited by a cellular tissue over the operculum. When a zoarium consisting of several layers is examined from below, the chitinous opercula in the older layers may nearly all be seen remaining in the oral aperture in their natural position, and when the zocecia from these lower layers are examined from above the opercula are mostly without any calcareous growth, while others have a calcareous layer over part or all of the operculum. Near the centre of the operculum, or rather distal to the centre, a tubule grows through this super- opercular deposit, which may be quite short (figs. 14, a, b) or may form a noticeable tube, for in one case a fine tubule joins the inner wall of the next layer of zocecia (fig. 14,2). The mandibles in 8. spongites, Smitt, are also sometimes similarly covered with a calcareous deposit. Schizoporella viridis, Thornely, in large masses, shows in the upper Jayers no calcareous deposit on the opercula, nor any tubule or tubercle ; but when the upper layer is scraped away, then the opercula of the underlying zocecia are mostly seen to have a stout tubercle over the middle, but only in a few cases is there any calcareous deposit. The new zoccia in S. wiridis are placed irregularly, quite independently of the layer below, and only accidentally do the walls of the new zocecia pass over the old opercula, but where this has occurred there is in one or two cases a thickening of the operculum under the new wall. We thus see that the position of the layers of zocecia is quite different in S, viridis (fig. 18) to what occurs either in S. unicorns or in S. porelliformis §, nov. (Pl. 2. figs. 19-21, an African * “ Bry. Sudanese Red Sea,” Journ. Linn. Soc., Zool. xxxi. (1909) p. 144; “ Bry. from Zanzibar,” Proc. Zool. Soc. 19138, p. 502. + “Brit. Carb. Trepostomata,” Mem. Geol. Surv. of Great Brit., Paleont. vol. i. pt. 3, p. 145 (1912). } Ulrich, E. O., “Pal. Bry.,” Pal. of Minois, Geol. Survey Ilinois, vol. viii. pt. 2, p. 822 (1890). § Schizoporella porelliformis, sp. noy. (Pl. 2. figs. 19-21). It is much like S. nivea in most characters, but is larger with larger aperture, &e. The surface and the large round ovicells have large pitted pores. It is bi-multilaminate. The oral aperture is nearly round, as the sinus is nearly the width of the aperture. The operculum has the muscular attachments at the side, and not from bosses distant from the border. There are one or two small oval avicularia at each side of the aperture, 16 MR. A. W. WATERS ON THE MARINE species which I have long known in manuscript), there being important specific differences in each. The closures and tubules are very important, as the same thing occurs in various recent and fossil Cheilostomata *, though often called “ opercula,” as it has frequently been supposed that the aperture was closed by this calcareous film, independently of the chitinous cperculum, whereas from analogy we may feel sure that the calcareous deposit was formed over the operculum, there being therefore an important structural difference in the closures of the Cheilostomata and Cyclostomata, so that, in fact, it is doubtful whether the name closure should be used for the deposit in the Cheilostomata. In Meliceritites (fossils very abundant in the Jurassic and Cretaceous formations) the closures with tubules are very common, and I have shown that the family had avicularia with mandibles; it also had opercula, and we now know that such closures and tubules are not restricted to the Cyclostomata. The new layer in some species ef Meliceritites has the distal wall passing over the operculum of the under layer just as described in Schizoporella nivea t+, Busk. Further, the shape of the zocecia in Meliceritites is not regu- larly tubular as in Cyclostomata. As we have several truly Cheilostomatous characters I have protested against Meliceritites being called Cyclostomata ; on the other hand, Gregory and Levinsen have shown that the large ovicells have Cyclostomatous characters, which may be sufficient to prevent the family being placed with Cheilostomata ; certainly it should never be called Cyclostomata. Levinsen ¢ speaks of layers of growth being a Cyclostomatous character, but there are several Schizoporelle with multi-layered growth, and this is often the case in Adeonella, Cellepora, Micropora, Xe. In examining the opercula of S. wnicornis in decalcified material I came across a very abnormal double one, having two proximal ends, each with a wide are, like those of the normal form, and there are four muscular dots (fig. 22). Loc. Generally distributed in the northern hemisphere and the tropics. St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. ScHizoPoRELLA sponaitEs (Pallas), Smitt. (Plate 2. figs. 10-13.) Eschara spongites, Pallas (pas), Klenchus Zoophytorum, p. 47 (1766). Hippothou spongites, Smitt, Floridan Bryozoa, p. 42, pl. 8. figs. 161-168 (1873). Schizoporella spongites, Thornely, Report on Pearl Oyster Fisheries of the Gulf of Manaar, p. 114 (1905); Levinsen, Morph, & Syst. Studies on the Cheil. Bryozoa, p. 324, pl. 18. figs. 4a-d (1909); Osburn, ‘The Bryozoa of the Tortugas Islands, Florida,” Publ. 182, Carnegie Inst., p. 207 (1914). * “Bry. from Zanzibar,” Proc. Zool. Soc. 1918, p. 504. + “ Bry. from Zanzibar,’ Proc. Zool. Soc. 1913, pp. 503, 504. t “Studies of the Cyclostomata Operculata,” K, Danske Vid. Sky. vol. x. (1912) p, 19, BRYOZOA OF THE CAPE VERDE ISLANDS. 17 From the Cape Verde Islands there are two specimens, one of which is ll cm.x7 cm. They are solid, many-layered masses (up to 25 layers), rising in places into large mammillated mounds. These specimens are white and have been dead some time, as there are no opercula. Besides these there are a few colonies growing on stone forming one or two layers. Lower than the oral aperture, on one or both sides, there is a small avicularium with acute mandibles, and there is sometimes a small central umbo. The frontal surface is perforated, except just below the aperture, and there are large vicarious avicularia scattered about. Above the oral aperture there is a lined or crenulated ridge. The oral aperture has two denticles by the proximal edge,as can be seen from above (fig. 11), or better from the interior (fig. 12), and the operculum shows two hollows into which these denticles fit ; these are what Busk calls “foramina” in Cellaria, and I have shown * that they are sockets into which the teeth fit. The operculum is drawn from a specimen from Manaar, kindly given to me by Miss Thornely, and with regard to the minute details of the aperture and the surrounding ridge, &c., it corresponds in every detail with the specimens from Cape Verde, but does not have more than two layers of zocecia in any part. The ovicell is very large, being about twice the size of that of Schizoporella unicorns, Johnst., with a pitted surface, and at the opening two denticles directed distally, but I have not seen them nearly meeting, as figured by Levinsen. The ovicell usually quite conceals the oral aperture, and it is quite exceptional for any trace of the aperture to be seen, as already mentioned by Levinsen. The large vicarious avicularia occur frequently in a large number of species called Cellepora, and certainly most of the older writers would have placed this, with its solid irregular growth, under Cellepora; however, it has long been recognized that many species known as Cellepora must be united to Schizoporella. I have said + that the Cellepora spongites, Pall., was Schizoporella unicornis, the form which I called errata, and this I consider to be the case, but apparently, from the reference to previous figures, and from his description, Pallas was considering three distinct species, namely Schizoporella unicornis, and S. sanguinea, Norm., both from the Mediterranean, while the one he last refers to from America is what Smitt placed as S. spongites, though where only zoarial characters are referred to it is impossible to speak with absolute certainty, as two species may correspond zoarially. The first two species are now well known under other names and “ spongites” has been used by several authorities, so that even if rules should suggest a new name it would seem as if the plan adopted by Levinsen is the best—that is, to call it S. spongites (Pallas), Smitt. * Report of the Voyage of the ‘Challenger,’ vol. xxxi. pt. 79, p. 16. + Journ. Linn. Soc., Zool. vol. xxxi. (1907) p. 145. LINN. JOURN.—ZOOLOGY, VOL. XXXIV. 2 18 MR. A. W. WATERS ON THE MARINE In figure 10 the zocecia from various parts of the colony are placed together to show the characters. Loc. Florida, 13-35 fath. (Smitt) ; Bermuda (Verrill) ; Ceylon (Thornely) ; West Indies and Malacea (Levinsen) ; on a coral from 8. Africa in the Jelly Collection in the Victoria University Museum; Cape Verde Islands (St. Vincent Harbour, 10 fath.), collected by Crossland. ? SCHIZOPORELLA OLIGOPUS, Fobertson. (Plate 2. figs. 5, 6.) Schizoporella oligopus, Robertson, ‘The Incrusting Chilostomatous Bryozoa of the West Coast of North America,” Univ. of California, Pub. in Zool. vol. iv. (1908) p. 292, pl. 20. figs. 50, 51, 52. The radicles occur to nearly all the zocecia, and are seen at the growing edge, even when no polypides have been formed, the number is most frequently 6-8 but there may be 10, and a few zocecia have none. Similar chitinous radicle-tubes occur in many Cheilostomata—for instance, in M/em- branipora patellaria (Moll), Waters*, which is often thus attached to large Melobesia, and the specimen of S. oligopus from the C. Verde Is. is also on Melobesia. These radicles are at first a tubular projection at the end of which is a chitinous tube attached to the Melobesia. Ai the junction of the zocecia there is a space, as mentioned by Dr. Alice Robertson. My specimens of S. argentea, Hincks, from Tahiti, which are co-types, do not show the spaces at the junction of the zocecia, nor do those from the Red Sea, which have a calcareous opaque area in the middle of the dorsal surface, as described by Hincks, and there is considerable irregularity in the number of radicles. The anterior surface of the zocecia does not usually show pores round the border, but, when examined from the interior, a row at the border is seen between the granula and over the surface generally, but none are visible until Hau de Javelle has been used. The position of the avicularium close up to the oral aperture on one side and the thin operculum led to my calling it Rhamphostomella before recognizing that it had been described. It will be noticed that while the umbo is central, when there is no avicularium, yet when an avicularium eccurs the umbo is on the other side. The aperture of the ovicelligerous zoccia are much wider than those of the other zocecia and are differently shaped, as the sinus is less marked. This is not alluded to by Dr. Alice Robertson, but the figures indicate a difference. No difference is noticeable between the zoccial and ovicelligerous apertures of S. argentea. No spines have been seen. There are about 16 tentacles. It must remain an open question whether the difference (if any) in the frontal pores justifies a variety or new species. * Waters, “Bry. of the Bay of Naples,” Ann. Mag. Nat. Hist. ser. 5, vol, iii. (1879) p. 120, pl. 10. figs, 8, 9. BRYOZOA OF THE CAPE VERDE ISLANDS. 1g) Loe. San Pedro, California (Robertson). St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. SCHIZOPORELLA TRICHOTOMA, sp. nov. (Plate 2. figs. 1-4.) The zoarium is adnate, incrusting clinkers. The zoccia are ovate, with pores spread over the surface, which under a low power appear stellate ; the oral aperture is large, with a distinct sinus on the straight lower edge, and straight sides; at the distal end of the zocecium there are four spines. The operculum has a diagonal ridge at each side. The decalcified frontal zocecial wall shows that the upper membrane has, at the position of each pore, three oval disks, then below this in the lower membrane there is a circular disk with a spot in the centre (fig. 3). The raised ovicell has a median ridge and a mucro, with ridges from the mucro to the distal end of the aperture. There are pores round the edge of the ovicell. The primary has eleven spines. Only dried specimens have been met with. Schizoporella divisopora, Waters”, and S. pulehra, Nevianit, both have stellate pores, but the ovicells do not quite correspond, though the main characters are the same in S. divisopora, so that the recent species are closely related. Loc. St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. OsTHIMOSIA AVICULARIS (LTincks), Waters. For synonyms see Miss Jelly’s Catalogue under Cellepora avicularis, Hincks, and Calvet’s Expéd. Se. du ‘Travailleur” et du ‘Talisman,’ Bryozoaires, vol. viii. (1907) p. 441. Cellepora americana, Osburn, “ Bry. of Woods Hole,” Bull. of the Bureau of Fisheries, vol. xxx. (1910) p. 288, pl. 25. figs. 52 a, 6; pl. 31. fig. 99. There is only one specimen, which is about 5 mm. high, the zoarium is massive with three or four short lobes. The rostrum to the oral aperture is usually short, but there is generally in this species much variation in this respect. In specimens from Naples sometimes there is quite a long rostrum with avicularia on the inner side, while others in the same specimens have much shorter ones with the mandible near the end. It therefore seems that the Mediterranean form includes the differences mentioned by Osburn, who says his C. americana is related to avicularis. The large vicarious avicularia, with nearly parallel sides, are abundant in some parts of the specimen. Loc. Arctic: British; Mediterranean at Naples, Rapallo, Cette, Corsica, * “Bry. from N.S. Wales, &c.,” Ann. Mag. Nat. Hist. ser. 5, vol. xx. (1887) p. 198, pl. 5. fig. 4. + “Bri. foss. della Farnesiana,” Pal. Ital, vol. i. p. 110, pl. 5, figs. 386, 37 (1895). 9% 20 MR. A. W. WATERS ON THE MARINE Oran (Algiers) ; Florida; Vineyard Sound, Buzzard Bay, 1-19 fath., No Man’s Land, Nantucket, Sandwich (Osburn); from underneath a coal-lighter in St. Vincent Harbour, Cape Verde Islands, collected by Crossland. HipporHoa pIstans (/fincks). Cosmopolitan. HrpporHoa DIVARICATA, Lamourou.. For synonyms see Miss Jelly’s Catalogue, and add :— Waters, Résult. du Voyage du 8.Y. ‘ Belgica,’ Bryozoa, p. 53, pl. 7. fig. 8 (1904); “ Bry. from near Cape Horn,” Journ. Linn. Soc., Zool. vol. xxix. (1904) p. 238; ‘Bry. from Zanzibar,” Proc. Zool. Soc. 1913, p. 501; Nordgaard, Hyd. & Biol. Invest. in Norwegian Fiords, p. 165 (1905) ; Robertson, “ Incrust. Chil. Bry. of N.W. Coast of America,” Univ. of California, Pub. Zool. vol. iv. (1908) p. 296, pl. 21. figs. 59, 60 ; Norman, “ Polyzoa of Madeira,” Loc. Cosmopolitan, including Woods Hole, Madeira, and between Fayal and Pico. From slight depths to 1000 fathoms. HipporHoa HYALINA (Linneus), Waters. For synonyms see Schizoporella hyalina in Miss Jelly’s Catalogue, and add :— Mollia hyalina, Barrois, Kmb. des Bry. p. 163, pl. 9. figs. 4-17 (1877). Schizoporella hyalina, Levinsen, Zool. Dan. ‘ Danske Dyr’ (1894), p. 66, pl. 5. figs. 47-57; Robertson, “ Bryozoa,” Proc. Washington Acad. of Sc. vol. ii. (1900) p. 326; Calvet, Bry. Mar. de Cette, p.44 (1902); Hamburg. Magalh. Sam, Bry.p. 25 (1904); Expéd. Se. ‘Travailleur’ et ‘Talisman,’ vol. viil. (1907) p. 415; Robertson, Incr. Chil. Bry. p, 289, pl. 19. figs. 43-45 (1908). HHippothoa hyalina, Waters, “Bry. from Franz Josef Land,” Journ. Linn. Soe., Zool. vol. xxviii. (1900) p. 70, pl. 8. figs. 16-18; “ Bry. from Chatham Island,” Ann. Mag. Nat. Hist. ser. 7, vol. xvii. (1906) p. 19; Norman, “ Nat. Hist. Finmark,” Ann. Mag. Nat. Hist. ser. 7, vol. xii. (1903) p. 108; Nordgaard, Hydr. & Biol. of the Norwegian Fiords, p. 165 (1905); Kluge, “Erg. wb. die von der Olga gesamm. Bry.,” Biol. Anst. auf Helgoland, vol. viii. (1906) p.39; Levinsen, Morph. & Syst. Studies on Cheil. Bry. p. 276 (1909) ; Osburn, Bry. of Woods Hole, p. 235, pl. 24. fig. 47 (1912) ; Sumner, Osburn & Coles, “ Biol. Survey of the Waters of Woods Hole,” Bull. Bur. of Fisheries, vol. xxxi. (1913) p. 603. Celleporella hyalina, Norman, Ann. & Mag. Nat. Hist. ser. 6, vol. xiii. (1894) p. 129; Bidenkap, “ Bry. von Ost Spitzbergen,” Zool. Jahrb. vol. x. (1897-8) p. 621; *‘ Bry.,” Anstalt Helgoland, vol. iv. (1900) p. 252; “Bry. von Spitzbergen und Konig Karls Land,” Fauna Arct. vol. i. (1900) p. 512; “Forteg. ov. de Arkt. Bry.,” Bergens Mus. Aarbog, 1905, No. 9, p. 18. There is one small colony growing on the seaweed with Amathia. Loc. Fairly cosmopolitan from both hemispheres, occurring from tide- levels to over 2000 metres. Cape Verde Islands, colleeted by Crossland. ARTHROPOMA CEcILIt (Audouin), Levinsen. Waters, ‘ Bry. from Zanzibar,”’ Proc. Zool. Soc. 1913, p. 508. Schizoporella Cecilii, Robertson, ‘The Incrusting Bryozoa of the W. Coast of N. America,” Univ. of California Pub. in Zool. vol. iv. (1908) p. 288, pl. 19. tig. 42. Loc. General in the north temperate zones; Indian Ocean, Australia; Zanzibar shore ( Waters); Red Sea ; Japan ; Cape Verde Islands, collected by Crossland, BRYOZOA OF THE CAPE VERDE ISLANDS. 21 CuorizoporA Bronenrartit (Audouin), Hincks. For synonyms see Miss Jelly’s Catalogue, and Calvet, Expéd. Sc. du ‘Travailleur’ et du ‘Talisman,’ vol. viii., Bryozoaires, p. 413, and add :— Levinsen, Morph. & Syst. Studies on the Cheil Bry., pp. 275, 276 (1909); Angelis D. Joaquin, “ Los primeros Briozoos enc. en los Depdésitos Plioc. de Catalutia,” Mem. de la C. y Artes de Barcelona, p. 36 (1900). Loe. British and French coasts; Guernsey ; Mediterranean ; Australia ; Dunedin and Foveaux Straits, New Zealand; S. Africa ; Canaries; Azores ; Madeira; St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. SMITTINA TRISPINOSA, var. PROTECTA, Thornely. Waters, ‘‘ Bry. from Zanzibar,” Proc. Zool. Soc. 1918, p. 513, Loe. Gulf of Manaar ( Thornely) ; Red Sea (Waters) ; Wasin, Brit HE. Africa, 10 fath.; St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. SMITTINA TROPICA, Waters. Smittina tropica, Waters, Journ. Linn. Soc., Zool. vol. xxxi. (1918) p. 174, pl. 17. figs. 10-14. Loc. Red Sea; St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. LEPRALIA ? CUCULLATA, Bush. For synonyms see Waters, “ Mur. Biol. of the Sudanese Red Sea, Bryozoa,” Journ. Linn, Soc., Zool. vol. xxxi. (1909) p. 150. pl. 15. figs. 1-6, 10, and add: — Schizoporella cucullata, Jullien et Calvet, Bry.proy. des Camp. de lHirondelle, p. 141, pl. 16, figs. 7,7 a (1903); Calvet, Expéd. Se. du ‘Travailleuy’ et du ‘Talisman,’ Bryozoaires, vol. viii. (1907) p. 415. The opercula of specimens from the Cape Verde Islands are somewhat larger than those from the Mediterranean. Loc. Mediterranean ; Red Sea; Mazatlan; Ceylon; California ; 8. Africa; Azores; Iles Brancot, Cape Verde, 110-180 met. (Calvet); St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. LrePrALIA PERISTOMATA, Waters. Lepralia peristomata, Waters, “Bry. from Madeira,” Journ. Roy. Micr. Soc. 1899, p. 10, pl. 3. fig. 20; Norman, “ Polyzoa of Madeira, &c.,” Journ. Linn. Soc., Zool. vol. xxx. (1909) p. 305. Lepralia Magnevilla, Busk (non Sav. et Aud.), Q. Journ. Mier. Sci. vol. viii. (1860) p. 284, pl. 31. fig. 5. Two small tubercles can be seen on the front of the operculum towards the proximal part in the position figured by me (loc. cit. fig. 20). They can be examined in dried specimens in which the apertures are closed by the opercula, 22 MR. A. W. WATERS ON THE MARINE In Lepralia peristomata the peristome entirely surrounds the aperture, whereas in Lepralia crassimarginata, Hincks, afterwards described by Jullien as L. collaris, there is only an elevation a short distance below the proximal edge of the aperture. There is, however, in L. peristomata a great range in the development of the peristome, as some zocecia show hardly any, whereas in others it is much raised. Loc. Madeira (Busk, Waters, Norman), 15-70 fath.; St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. ScHizoTHeca FIssa (Bush), Hincks. Lepralia fissa, Busk, Q. Journ. Mier. Se. vol. iv. (1856) p. 308, pl. ¥. figs. 8-10; Waters, Ann. Mag. Nat. Hist. ser. 5, vol. iii. (1879) p. 43, pl. 11. fig. 6. Schizotheca fissa, Hiucks, Brit. Mar. Polyzoa, p. 284, pl. 41. figs. 1-3 (1880); Ann. Mag. Nat. Hist. ser. 5, vol. xix. (1887) p. 303; Neviani, “ Bri. post plioc. di Spilinga,” Atti d. Accad. Gioen. di Sc. Nat. in Catania, vol. ix. (1896) p. 34, fig. 17; ‘ Bri. delle Form Plioc. e Postpl.,” Bull. Soc. Geol. Ital. vol. xvii. (1898) p. 12; “ Bri. neog. della Calabria,” Pal. Ital. vol. vi. (1900) p. 202, pl. 18. figs. 4,5; Calvet, ‘Bry. du Corse,” Trav. de l’Inst. de Zool. de l’Univ. de Montpellier, ser. 2, Mém. 12, p. 25 (1902); Exp. Se. du‘ Travailleur’ et du ‘Talisman,’ Bryozoaires, vol. viii. (1907) p. 429; Levinsen, Morph. & Syst. Studies on the Cheil. Bry. p. 294 (1909). There are only one or two small pieces, and in these a number of the earliest or disk stages of the ovicell occurs, but very few completed-ones ; in these the “fissure” is not as large as in my British and Mediterranean specimens, but the median portion is thin. Levinsen does not deal, in evtenso, with the genus, merely putting it in the family Reteporide. ‘This has long seemed to be right, as the open ovicell is similar to those of the group of Retepore with the wide fissure as in R. Imperati, Busk*. The labial fissure or pore is also very well marked, and this, as I have said, is an important character of the genus. oc. Guernsey ; Sidmouth; Plymouth ; Cornwall, 30-40 fath. ; Ireland ; Naples; Capri; Adriatic; Corsica, St. Bonifacio, 55-77 met. (Calvet and M. Hdw.), Bastia, 40-60 met. (Calvet), Bay of Cadiz, 717 met. (Calvet); St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. HOLOPORELLA PUSILLA (Smitt), Waters. (Plate 2. figs. 7-9.) Discopora albirostris, forma pusilla, Smitt, Floridan Bry. p. 70, pl. 12. tig. 238 (1873). Cellepora seaspinosa, Waters, “ Bry. from Madeira,” Journ. R. Micr. Soc. 1899, pl. 3. fig. 12; Norman, “‘ Poly. of Madeira,” Journ. Linn. Soc., Zool. vol. xxx. (1909) p. 811. Lepralia Watersi, Calvet, Expéd. Sc. du ‘Travailleur’ et du ‘Talisman,’ Bryozoaires, vol. vill. (1907) p. 412, pl. 27. fig. 11. The younger zoccia, near the border, show six long spines, whereas the older ones have only one at each side. The ovicell is small, raised, and * Waters, “ Med. & N. Zeal. Reteporidw,” Journ. Linn. Soc., Zool. vol. xxv. (1894) p. 256. BRYOZOA OF THE CAPE VERDE ISLANDS. 23 widely open in front, and at one side of the zocecia near the aperture there is a small acute avicularium ; in a few cases there is one on each side, with occasionally an ayicularium in other positions. Generally there is no mucro to the ovicells, though in one specimen many of the ovicells are mucronate, or even might be said to support a long spinous process, and the frontal projection, as figured by Calvet, has been seen. There was no ovicell to the specimen from Madeira. The surface of the zocecia and of the ovicell is finely granular, and there are a few large pores round the border of the zocecia, The aperture measures about 0-08 mm. Loc. Florida, 9 fath. (Smitt); Madeira; Ile Brancot, Cape Verde Is., 20-25 met. (Calvet) : St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. HoLoporELLa APERTA (LTincks), Waters. For localities see Waters, “Marine Fauna of Brit. E. Africa,” Proc. Zool. Soc. 1913, p. 522. There is only one small colony, growing on a piece of coal, so that no preparation could be made, but there seems no doubt as to the identity. Loe. Singapore or Philippines; Ceylon; Indian Ocean; Zanzibar and Brit. E. Africa ; Cuba; St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. MicroporeLia cinta (Pallas), Harmer. For localities see Waters, ‘‘ Marine Fauna of Brit. E. Africa,” Proc. Zool, Soc. 1918, p- 523, and for synonyms add: Lepralia ciliata, Barrois, Embryologie des Pryozoaires, p- 149, pl. 7. figs. 4, 6, 7, 15, 19 (1877). Microporella ciliata, Harmer, ‘ Morph. Cheil.,” Q. Journ. Micr. Se. vol. xlvi. (1902) p. 316; Waters, ‘ Mar. Biol. Sudanese Red Sea,” Journ. Linn. Soc., Zool. vol. xxxi. (1915) p. 443; Osburn, “Bry. of Woods Hole Region,” Bull. Bur. of Fisheries, vol. xxx. (1912) p- 233; Osburn, Sumner & Cole, “ Biol. Survey of the Waters of Woods Hole,” Bull. Bur. of Fisheries, vol. xxxi. (1913) p. 601. Loc. Cosmopolitan; Cape Verde Islands, collected by Crossland. ADEONELLA contorta (Michelin), Waters. (Plate 4. tigs. 10-15.) Laminopora contorta, Mich. Magasin de Zoolozie, 1842, pl. 3; Waters, “On some Recent Bryozoa in d’Orbigny’s Collection,” Ann. Mag. Nat. Hist. ser. 7, vol. xv. (1905) p. 16. Schizoporella contorta, Calyet, Expéd. Sc. du ‘Travailleur’ et du ‘Talisman,’ Bryozoaires, vol. viil. (1907) p. 420. ? Gemellipora arbuscula, Calvet, loc. cit. p. £26, pl. 27. figs. 16-19. Adeonella contorta, Waters, “ A Structure in Adeonella (Laminopora) contorta, Mich. . .. together with Remarks on Adeonide,” Ann. Mag. Nat. Hist. ser. 8, vol. ix. (1912) p. 489, pls. 10 & 11. A common character is frequent branching at right angles, often forming but very short branches, as figured (fig. 15) from a colony, collected by Cross- land, from St. Vincent, Cape Verde Islands. In some specimens these branches 24 MR, A. W. WATERS ON THE MARINE form a semispiral foliaceous growth round the main branch. The specimens under consideration, and those seen in various museums, are generally violet or red, but the colour is not mentioned by Calvet in Gemellipora arbuscula, nor are the frequent and short branches, but, as young and even older colouies of A. contorta are known without this branching, I consider that it is probably a synonym. It may be well to call attention to the fact that the Adeonide are nearly all darkly coloured, but specimens bleached quite white are often seen. There is an interesting specimen of Adeona in the British Museum quite white on the one side, though very dark on the other, it evidently having been exposed to the light. The zocecia are raised when quite young, but are depressed when older. There is an avicularium on one side, sometimes on both, at about the level of the oral aperture. ‘The oral aperture is elongate with a long sinus and denticles just above the sinus. The zocecia are small and may be entirely filled up by the embryo, which is about 0°25 mm: long, and, so far as I have seen, there is no external indication as to which are goncecia. The older opercula often have a raised tubercle on the front, about the middle, and sometimes there is a calcareous layer over the operculum, forming what has been incorrectly called a calcareous * operculuam—from this arises the calcareous tubercle. In many of the Cheilostomata a calcareous layer may be formed over the older opercula, and this has been dealt with on page 15. It does not seem right, in any case, to speak of a calcareous operculum. There are about 13-15 tentacles, and all the Adeonidee examined have about this number, namely, A. platalea, Busk, 13; A. polystomella, Reuss, 16 ; A, polymorpha, Busk, 15; A. lichenoides, Busk, 14; Adeonellopsis distoma, Busk, 14-16; A. Crosslandi, Waters, 13-14. In my paper on Adeonella contorta, on page 493, the characters are given of the group dealt with by Busk as Adeonella. Loc. Evidently abundant trom the Cape Verde Islands, and, as stated in my paper referred to, it appears to occur in the Red Sea; the John Adams Bank, Brazil, and ? Cape St. Vincent +, Spain; St. Vincent, Cape Verde Islands, 5-20 fath., collected by Crossland. CRISIA. When this paper was almost finished Harmer’s valuable ‘Siboga’ Report was published, and as I was not sure how far I agreed with him upon some * See page 15. + I wrote asking My. Kirkpatrick if there was any possibility of the British Museum specimen from “Cape St. Vincent, Spain,” having come from St. Vincent, Cape Verde Is., and he kindly replied that it was ‘ Rattlesnake’ material, purchased through a dealer, and that 5 fath. looks like offshore at Cape Verde, not St. Vincent, Spain, and that there was probably a sale catalogue mistake. BRYOZOA OF THE CAPE VERDE ISLANDS. 25 points, this genus again received further examination, confirming most of the conclusions previously arrived at. The ovicells are generally of three forms * :— Ist. Those that are free, with the oceciostome on the dorsal surface, as Crisia Edwardsiana, VOrb., C. biciliata, MacG., C. howensis, MacG., C. ker- guelensis, Busk, C. inflata, Waters, C. cuneata, Maplestone. 2nd. Those witha pomiform oviceil, as C. aculeata, Hass., C. eburnea, Hincks (pars), C. elongata, Harmer (M. Ed. ?), all of which have an opening with more or less of a slit and no pronounced tube; then C. conferta, Busk, and C. sertularoides, d’Orb., have a tubular oceciostome. A funnel oceurs in C. conferta, Busk, which in many other respects corresponds with C. elongata, Harmer. 3rd. There are a number with pyriform ovicells, most of which seem to have a tubular oceciostome, as C. tubulosa, Busk, and C. jistulosa, Heller, with large wide zocecia about 0°1 mm. or more in the aperture. C. opereculata t, Rob., C. pacijica, Rob., C. maxima, Rob., C. pugeti, Rob., have also elongate pyriform adnate ovicells, but the zocecia are smaller, being about 0:07 mm. ; then C. ramosa, Harmer, C. Holdsworthii, Busk, C. eburneo-denticulata, Smitt, C. tenuis, MacG., C. cribraria, Stimp., C. geniculata, M. Ed. We can only give typical shapes of the ovicells as there is a certain amount of variation—for instance, I have C. denticulata, from Roscotf, sent to me by Joliet, in which there are several long ovicells (Zeppelin form) almost tapering at the end (see Busk, Cat. Mar. Poly. pt. iii. pl. 4. figs. 2, 3), whereas from the same colonies there is one with a shorter ovicell and flattened distal end. A colony of Harmer’s elongata from the ‘Siboga’ Expedition shows considerable variation of the ovicells, so that they might be called pyriform, pomiform, or wider. Canu ¢ gives the photograph of the fossil Crisca Corbini, Canu, with a very large and wide ovicell, about double the width of the branch, with the end flat. Jullien and Calvet § say that.they consider the “formules algébroides” introduced by Smitt, and followed by Harmer, with one slight simplification, to be valueless, and these have never appealed much to me, for in description we must try and give the usual or typical characters, rather than take an individual in such a genus. However, there may be cases where such a formula may be useful, as, for instance, using it instead of an extra figure or in sending particulars to a correspondent. The formula has, however, been * “Bry. from Zanzibar,” Proc. Zool. Soc. 1914, p. 840. + Some of Robertson’s species are only judged from the figures, and this is the case with C. tenuis, MacG. } “ Bry. des Terrains Tert. des Environs de Paris,” Ann. de Paléont. vol. ii. (1907) p. 104, pl. xii. fig. 6. § Jullien et Calvet, Bry. prov. des Camp. de l’Hirondelle, p. 109 (1903). 26 MR. A. W. WATERS ON THE MARINE made unnecessarily puzzling by the use of the plus (+) sign, where plus is not really meant, for the number of zoccia in a branch is first given, then where a new branch occurs is indicated, but this is not a plus, it is not additional, unless the subbranching zocecia are first mentioned separately ; and by putting a comma instead of the + the disconcerting “algébroide” appearance is done away with. We thus give the number of zoccia, say 9, and show that the new branch occurs after the second zocecia, and whether the branch is on the right or the left by the position of 7. There is, of course, not the objection to brackets that there is to the plus, but it is more easily printed with a colon. A specimen from the Cape Verde Island collection reads :— Ry BS iP 8) ZR Gm SS o oy ay OES nies, Geel Sc t 12. 57, 10035, Xue This is a piece near the growing end, whereas at the base there are a number of internodes without zocecia, or very few, before the typical inter- nodes are formed. The example formula only deals with three branches, while many colonies will have fifty, and for some colonies a large sheet of paper is required, though the main features of a hundred colonies might sometimes be given in a few lines. It should be noticed that the growing branch frequently contains a larger number of zocecia than other branches. A further complication in the formula occurs in Harmer’s last paper, a branch on the right has + and a small figure below, whereas a branch on the left has a large figure and a small 7 above, but, although it occurs throughout, it must surely be a printer's error. In giving the measurements of the distance from zocecia to zocecia it, of course, must not be supposed that there is no variation, but if mature branches are taken, though not the lower zocecia on a branch, then it will usually be found that the variation from the average is but slight. Crista TUBULOSA, Busk. (Plate 3. fig. 1.) Crisva tubulosa, Busk, Cat. Mar. Polyzoa, pt. iii. p. 7, pl. vi. A. figs. 3, 4 (1875). In the Boa Vista specimens the zoccial tubes expand at the end, the aperture measures about 0°12 mm., which is a trifle smaller than the British Museum specimen, and the zocecial tube is free for a considerable extent. The ends of the zocecia are 0°5 mm. apart, and except in the lower internodes there are more zoccia than given by Busk, sometimes as many as 16. The branches arise from the second, third, or fourth pairs. The ovicells are elongate wo “I BRYOZOA OF THE CAPE VERDE ISLANDS. pyriform, to some of which there is a tube without any funnel, but other ovicells show no tube. In the British Museum specimens there is a slight funnel. The aperture of the ovicell tube is about 0°08 mm. There is usually a branch on each side near the base of the ovicell. The bases rami are long— that is, they reach to the next zocecium below. This is much like C. pacifica*, Robertson; but from Dr. Alice Robertson’s figure the distance from zocecium to zocecium is only about 0°33 mim., so that it isa more slender species than tubulosa, Busk. C. pacijica, Rob., C. maxima, Rob., C. puget, Rob., all seem closely allied. Loe. Cape Verde Islands (Busk); and Boa Vista, Cape Verde Islands, collected by Crossland. CRISIA DENTICULATA (Lamarck), MW. Edw.,var.VERDENSIS, nov. (Text-fig. 2.) There are several colonies from the Cape Verde Islands which are about 12 mm. loug, forming close tufts much less straggly than is usually the case Fig. 2. Crista denticulata (Lam.), var. verdensis, nov. in Crisia denticulata, as the branches very soon have the mature form, and this may be the case after one short internode, whereas frequently in * “ Cyclostomatous Bryozoa of the West Coast of N. America,” Uniy. of California, Pub. Zool. vol. vi. (1910) p. 242, pl. 20. figs. 16, 17. 28 MR. A. W. WATERS ON THE MARINE C. denticulata there are a series of short internodes before gradually becoming normal however, this may be only a character of local importance. There are generally 7-9 zocecia in an internode, the branches usually commence after the first or second zocecium on that side, being alternate where there is no ovicell, with one branch to each internode, followed in the next internode by one on the other side. In the internodes with an ovicell there is usually a branch after the first zocecitum or near it, and a few zocecia further up is the ovicell, as well as a new branch on one or both sides near to it, corre- sponding with the branching and ovicell described by Harmer in his C. elongata*. The ovicell, though not as short as in Harmer’s figured specimen, is short and pomiform, but does not spread out at the distal end asmuch. It has a slit or oval opening in the distal end of the ovicell as described by Harmer for both elongata and denticulata, The branches are fairly straight, with sometimes a slight sigmoid curve as may also be found in C.denticulata. The zocecial aperture is about 0°06 mm., the distance from zocecium to zocecium is about 0°26 mm., the joints are black. The denticle on the outer side of the zocecium is a most inconstant character in this and other species, for some internodes may have it strongly marked, while it is quite absent in others. It would seem that round denticulata f we have a group including probably C. acropora, Busk, C. sinclairensis, Busk, C. elongata, M. Edw., C. elongata, Harmer, C. serrata, Waters. Whether we consider them as species or varieties T cannot think that the form I figured as C. elongata t, believing it to be the form described by Milne Edwards, is less entitled to a distinct position than the others. Its internodes are remarkably straight, whereas Harmer describes his as usually with a sigmoid curve. I suppose that about 100 internodes of the form figured were examined, all straight and long with numerous zoecia, and only one branch growing near the distal end. The base of the branches, both main and lateral, of the Zanzibar elongata is very broad, about 0°13 mm., but as this is not much broader than in C. denticulata much importance cannot be attached to it. In my paper “On some Species of Crista,” § I have referred in more detail to Harmer’s C. elongata, and indicated that two * “ Polyzoa of the ‘Siboga’ Expedition,” Ent., Ctenost. & Cyclos. p. 96 (1915). + The specimen of C. denticulata in the ‘Challenger’ collection from St. Paul’s Rock, Chall. Rep. Polyzoa, Pt. II., p.4, pl. 2. fig. 3, has short ovicells without any tubular oceciostome, the lower joints are black, and it is somewhat like Harmer’s C. elongata. ‘The specimen from Macclesfield Island, Tristan Da Cunha, has the joints light, the basis rami is not wedged in, and it does not seem to be denticulata. The specimen from Cape York has fairly straight internodes with the branches placed high, and is not unlike my C. elongata. + “ Bry. from Zanzibar,” Proc. Zool. Soc. 1914, pl. 1. figs. 3, 4 (1915), and see Ann, Mag. Nat. Hist. ser. 8, vol. xviii, (1916) p. 474. § Ann. Mag. Nat. Hist. ser. 8, vol. xviii. (1916) p. 474. bo ite) BRYOZOA OF THE CAPE VERDE ISLANDS queried specimens may have influenced him to make a comparison with my elongata. The variety verdensis has comparatively short regular internodes, few branches in the internodes where there is no ovicell, black joints, base of branch about 0°08 mm., most of the bases rami are wedged in, while others are longer ; though, as some of these characters may not be very important, verdensis and elongata may be moderately closely allied. On the other hand, the Cape Verde Islands specimens of verdensis are much like C. sinclairensis, Busk. At any rate, verdensis has not the long straight internodes with one branch to each, as in what Buskand I have considered to be elongata, M. Edw. Loc. Boa Vista, Cape Verde Islands, 5-20 fath., collected by Crossland. CRISIA SIGMOIDEA, Waters. Crisia sigmoidea, Waters, Ann. Mag. Nat. Hist. ser. 8, vol. xviii. (1916) p, 476, pl. 16. figs. 9, 10. Crisia denticulata, Waters, “ Bry. of the Bay of Naples,” Ann. Mag. Nat. Hist. ser. 5, vol. iii. (1879) p. 269, pl. 23. fig. 2. There is one piece, from the Crossland Cape Verde collection, which I think is this species. The zoarium is broad, with the ends of the zomcia free projecting upwards, the ends of the zocecia are about 0°28 mm. apart, and the zocecial apertures are about 0:06-0:07 mm. wide. The joints are light, and the base of a branch is about 0-1 mm. wide, occurring after the second or third zocecium on that side, with another branch on the other side a few zocecia further up. Unfortunately there is no ovicell, and the C. sigmotdea and C. conferta, Busk, are very similar, so that without an ovicell determination is difficult, but the type-specimen of C. conferta is slightly wider than C. sig- moidea and at the same time the basis rami is wider, being about 0°13 mm. in conferta. The ovicells of the Mediterranean sigmoidea do not show any tubular oceciostome, whereas the funnel-shaped oceciostome is very marked in conferta. This differs from C. denticulata and the C. elongata, Harmer, in haying light joints, but the position of the branches is very similar to Harmer’s elongata. Loc. Naples, Rapallo, Villefranche-sur-Mer, Oran (Algiers). Cape Verde Islands, collected by Crossland. CRISIA VINCENTENSIS, sp. nov. (Plate 3. figs. 2, 3.) This is a form which has been difficult to place, for although there are several small specimens an ovicell was only found in one case, and of such an unusual form as to raise the question as to whether it is abnormal. It is broad throughout, as though two zocecia might have been forming ovicells 30 MR. A. W. WATERS ON THE MARINE which became agglomerated. At first it was thought to be C. eburnea, and probably the variety Java, Busk. The lower branches usually start from below the aperture of the first zoecium—that is, from the side of the first zocecitum,—while in older parts the new branch arises above the aperture of the first zocecium. In C. eburnea, Hincks, there is some irregularity as to the branches starting from the first or second zocecium, though it is most frequently from the first. In C. vincentensis there are usually 7 zocecia in an internode, the zocecial aperture is 0:05-0:06 mm., the oceciopore is 0°03 mm. without any funnel; the branches are about 0°12 mm. wide, the distance from zocecium to zocecium is about 0°26 mm., the base of a new branch is about 0°07 mm., and the joints are light. The basis rami differs from that of C. eburnea in being longer, sometimes reaching to the next zocecium, or it may stop a little short of this, whereas generally the basis rami in eburnea is short with what T should eall the graft character. A ‘Challenger’ specimen of C. eburnea, var. lara, in the British Museum from the Busk collection has a large, long, adnate ovicell much raised at the distal end with the oceciostome contracted, so that it is much wider than deep as in the C. eburneo-denticulata figured by me * from West Greenland, and on this account the present form is not placed under axa. Harmer f, with a query, places C. eburnea, var. lava, as a synonym of C. kerguelensis, Busk, but the specimen alluded to in the British Museum, with the attached ovicell, proves that it is not the C. kerguelensis, which Harmer figures with a free ovicell, having a tubular oceciostome on the dorsal — surface. A new name is given with considerable hesitation to the present Cape Verde specimens. C. vincentensis has about 12 pores in each square 0°1 mm., C. elurnea and C. fistulosa have fewer, C. ramosa about 5. TUBULIPORA PULCHRA, MacGillivray, var. nov. (Plate 3. figs. 8, 9.) Tubulipora pulchra, MacG., “ Descr. of new or little-known Polyzoa,” pt. vii. Trans. & Proc. R. Soc. Viet. vol. xxi. (1885) p. 95, pl. 2. fig. 1; Robertson, ‘Cyclost. Bry.,” Univ. of California, Pub. Zool. vol. vi. (1910) p. 250, pl. 23. figs, 82-35. Tubulipora fimbria, var. pulchra, Waters, Ann. Mag. Nat. Hist. ser. 5, vol. xx. (1887 p. 258, pl. 7. figs. 1-3. The Cape Verde Islands specimens have the zoccial tube much larger (0°14 mm.) than those from Port Jackson, Australia (0°7 mm.), and it may be necessary to consider them as a variety. Apparently the specimens examined by Dr. Alice Robertson are coarser than those from Australia, but not so large as the present form. * Ann. Mag. Nat. Hist. ser. 8, vol. xviii. (1916) pl. 16, figs. 4, 5. + “Polyzoa of the ‘Siboga’ Expedition,” Ent., Ctenost, & Cyclost. p. 105 (1915), BRYOZOA OF THE CAPE VERDE ISLANDS. 31 In the colonies from the Cape Verde Islands the young branches are narrow with only a few zovwcia and frequently anastomose, though ultimately forming a wide spreading growth. They have evidently grown on a calcareous seaweed. 7. organizans, d’Orb., sent to me by Jullien from Cape Horn, has similar attachments, and Harmer mentions them in 7”. plumosa, W.Thomp. It is possible that the species under consideration should be called 7. organizans, WV Orb. Loe, Victoria (Mae(rillivray), Port Jackson ( Waters) ; Southern California, San Diego shore, 11-32 fath. (ubertson) ; St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. Tusuttpora Lamovurouxm (Audouin § Savigny), Waters. (Plate 8. figs. 4, D565 LONE) Proboscina Lamourouwit, Aud., ‘‘ Descrip. de ’ Egypte,” Hist. Nat. p. 236, pl. 6. fig. 5 (1826). This occurs as a narrow band, with very long zowcia on each side, showing a certain amount of regularity as in /dmonea, or it may spread in opposite directions with two oval subcolonies (figs. 4a & 10a), or even spread out ina flabeiliform manner, and in one of these the ovicell has a distinct purple colour in the dry state, so that at first it was taken for 7. liliacea (Pall.), Harmer. The zowcia are very exceptionally small, having the aperture about 0°5 mm., and the projecting portion of the zocecial tubes show no pores. The ovicell is central, forming a raised inflation with a simple oceciostome tube, about 0°03 mm. The dorsal surface is very characteristic, having raised ridges curved from the median line, and alony these ridges there is a row of small pores for the attachment (fig. 5). No Yubulipora has been described with so small an oceciostome tube, and in all the cases referred to by Harmer, with the exception of 7. jlabellaris, Fabr., the oceciopore is larger than the zooecial orifice, whereas here it is smaller. Harmer * says the owciostome in 7. liliacea is larger than the orifice of a zowecium, mentioning 7. phalangea, Couch, as about the same size, in T. aperta, Harmer, as larger than the zocecial orifice, in 7. plumosa, W. Thomp.,as also larger, while in 7. fabellaris it is somewhat lessthan the diameter of the zocecial orifice. Dr. A. Robertson ft says that the oceciostome in T. occidentalis is smaller than the aperture of a zocecium. I met with a few cases of some very delicate tubes, three or four starting from near the same centre, and these were taken at first for an unknown form (PI. 3, fig. 7 juv.), but some young forms of 7. Lamourouwxit, with primary, convinced me that young forms with long erect tubes give much * “On the Development of Tubulipora,’ Quart. Journ. Mier. Se. vol. xi. n. s. p. 91. + “Cycl. Bry. of the West Coast of N. America,” Uniy, of California, Pub, Zool, vol. vi, (1910) p. 249, pl. 22. figs. 29-31, aya ec MR. A. W. WATERS ON THE MARINE this appearance. Looking down on the tubes they are so much foreshortened that it is impossible to give satisfactory figures, and this remark applies also to the oceciostomes of fig. 4. The discovery of this species, placed by Audouin in his genus Proboscina, a name adopted for a genus by d’Orbigny and others, is most important, for some authors, especially paleontologists, have considered that in the Stomatoporidx, Stomatopora should be used for uniserial adnate forms, while Proboscina would include multiserial adnate forms. Undoubtedly the name Proboscina will have to be entirely dropped, and, while I think that Stomatopora will have to include many forms which are more or less multi- serial, and that some things described as Proboscina will fall into Tubulipora, T am not prepared to state that all will find their places in these two genera ; however, if a generic division is required for many now known as Proboscina there is the genus Criserpia * of H. Milne Edwards in which these can find a home+. Milne Edwards proposed to place all the known Cyclostomata in the “family” Tubuliporide, but we now see that many families as now understood were included, though until we understand more about the most important genus Stomatopora both as to its primary attachment and ovicells we cannot speak with any certainty concerning the group in question. Miss Jelly and others have considered that Savigny’s figures of Proboscina Boryi and P. Lamourouaii represented the same species, but in this I cannot agree, for in the figures the difference in size of the zocecia is very marked. Loc. Savigny’s specimens were probably trom the Red Sea or Mediter- ranean; St. Vincent Harbour, Cape Verde Islands, 10 fath., collected by Crossland. TUBULIPORA INCRASSATA (Smitt), Waters. (Plate 3. fig. 7. to) Proboscina incrassata, Smitt, “ Krit. Fort. 6fver Skand. Hafs-Bry.” Ofv. Kongl. Vetensk.- Ak. Foérh. 1866, pp. 402, 458, pl. 5. Tubulipora incrassata, Waters, ‘Bry. of the Bay of Naples,” Ann. Mag. Nat. Hist. ser. 5, vol, iii. (1879) p. 272. ? Proboscina Boryi, Aud., Descrip. de ’Egypte, Hist. Nat. p. 236, pl. 6. fig. 4 (1826). * “Mémoire sur les Crisies,” &c., Ann. Sci. Nat. 2me sér., Zool. ix. (1838) pp. 193-238. The importance of this paper has not always been fully appreciated, for here the Cyclostomata were divided from the Cheilostomata, though under other names, and since then there has been much elaboration. Audouin and Milne Edwards,' in 1828, separated the Bryozoa as “family” 4, meaning by this what we should now call a class. Although several zoologists had, for many years, realized that a separation of zoophytes should be made, they were the first to make it. To have shown not only that there was this class, but also to have realized the large divisions of Cheilostomata and Cyclostomata shows what good pioneer work Milne Edwards gave us. {‘ Audouin et Milne Edwards, ‘‘ Résumé des recherches sur les animaux sans yvertébres faites aux iles Chausey,” Ann. Sci. Nat. vol. xv. (1828) p. 18.] + Ann. Sci, Nat. vol. xv. (1828) p. 41, BRYOZOA OF THE CAPE VERDE ISLANDS. 33 There are some specimens of Tubulipora growing irregularly in a band- like form, with large zocecia, the zocecial aperture being about 0:14 mm., whereas in 7. Lamourouxii the aperture is only about 0°05 mm. A young specimen of 7. Lamourous (Pl. 8. fig. 7, juv.), growing on the present species shows the great difference in size. There are in the interior of the zocecia a few pin-head spines, and in one zocecium, not very far down, there is a row of short teeth. In a specimen from Naples there is an anterior ovicell spreading among several zocecia, with a short tubular oceciostome, about the same diameter as a zocecium, without any funnel. I do not consider that the Stomatopora inerassata, Hincks, is the Tubulipora (Proboseina) inerassata, Smitt, nor am I sure that the form described by Smité in 1871 * is the same species as he originally described. Without a considerable amount of material it is difficult to decide whether to provisionally retain Smitt’s name or to give a new one, as it is evident that there has been some confusion. D’Orbigny first gave the name without figures, and the description is insufficient. There is a certain amount of resemblance to Fil/sparsa tubulosa, Busk, from Naples, which, however, grows free or only very slightly attached, and has an oceciostome with one axis, usually the transverse, much the longer, with an irregular funnel as in 7. aperta, Harmer. The zoarium spreads out slightly towards the distal end without being flabelliform, and although the zoarium does not expand as much as in Harmer’s specimens, it would seem to be the same as aperta. Harmer, in his ‘Siboga’ Report, p. 143, says that 7. aperta may possibly be the Hornera tubulosa, Busk. Loc. Naples. St. Vincent Harbour, Cape Verde Islands, on clinkers, collected by Crossland. LICHENOPORA IRREGULARIS (Johnson), Norman. (P1.4. figs. 2,3, 5, 6,7, 8.) Radiopora irregularis, J. Yate Johnson, ‘ Cyclost. Bry. from Madeira,” Ann, Mag. Nat. Hist. ser. 6, vol. xx. (1897) p. 63. Lichenopora irregularis, Norman, “ Poly. of Madeira and neigh. Islands,” Journ, Linn. Soc., Zool. vol. xxx. (1909) p. 282. There are several zoaria growing on a large stone, brought up by a diver from off St. Vincent, Cape Verde Islands, and a small piece on a clinker from St. Vincent Harbour. “ The very thin lamella is closely attached to the stone, and is more or less circular, quite independent of the outline of the subcolonies ; the central sub- colony is circular or oval, while the outer ones are lobulate or often form “subtriangular lobes” with uniserial zocecia, and in these uniserial forms ovicells have been found. There are also various similar colonies, which from their general appearance seem as if they should be united, but they have bi-multiserial rays throughout the colony (PI. 4. fig. 1) with subcolonies * Ofvers. Kongl. Vetensk.-Ak. Férh. 1871, p. 1119. LINN, JOURN.—ZOOLOGY, VOL. XXXIV, 3 34. MR. A. W. WATERS ON THE MARINE somewhat larger than those of the uniserial forms (figs. 2, 8), and this bi-multiserial form is provisionally called var. composita. Returning to the uniserial forms it isin these that the ovicells have been found, and the zoarial growth is similar to that of Radiopora Francquana, d’Orb.* The central depressed region of eavh subcolony has large pores (about 0:07 mm.) about the size of those of the zocecia, whereas the pores between the rays are usually smaller. The inner zocecia of a ray are the most raised, whereas in var. composita there is very little difference—in fact, most seem cut off nearly straight, without being apparently worn ; in this uniserial form the zowecia are well raised, and especially near the edge of the colony the spinous processes are more pronounced than in the variety. In the British Museum specimen from Madeira f, in the Norman collection, most of the zocecia show a blunt spinous process at the outer border and occasionally on both the inner and outer, also the internal pin-head spines in the cancelli are very small and rather rare. In one colony, from Cape Verde Islands, there are two ovicells, which are very narrow and interdigitate in narrow bands between the rays (figs. 2,6). The surface of the ovicell has numerous fine perforations. Another colony has a very deep pit in the centre of several of the subcolonies (fig. 5), but, on carefully focussing to the base of one pit, the narrow bands of a similar ovicell are seen, so that we are looking upon the upper part of an ovicell now at the bottom of this pit, and a zoarial growth must have taken place subsequent to the formation of the ovicell, but whether it indicates a regular second layer it is impossible to say. In one zoarium, in the centre of a subcolony, the base is found of an ovicell in formation, which is smooth and imperforate, proving that in the case previously mentioned we were not looking on the base of an ovicell at the bottom of the pit. In the section of another Lichenopora I have seen an ovicell in two stories—that is, one ovicell above another. Harmer ¢ describes and figures a similar ovicell in what he calls Lichenopora nove-zelandiv, Busk, but it is unfortunate that he should have taken the name nove-zelandiw apparently because it was published a page sooner than L. Holdsworth, Busk, for this latter is well described from characteristic specimens found in Ceylon. On the other hand, the description and figure of nove-zelandie were not very satisfactory, and among the specimens which Busk so named in his collection there seem to be more than one species. I have some specimens which have always seemed to me to be this species, and one from New Zealand was sent to me so determined, also the British Museum specimens have been looked through more than once, * Pal. Franc. pl. 782. figs. 8-8. + In one case the ovicell only throws out arms on one side of the subcolony. There are some deep pits as described in the Cape Verde Islands specimens. { Polyzoa of the ‘Siboga’ Expedition, pt. i. p. 155, pl. 12. fig. 11 (1915). BRYOZOA OF THE CAPE VERDE ISLANDS, sata) without finding reason for change of opinion. Apparently Busk did not figure the specimen he was describing—in fact, the figure is from a badly preserved specimen; whereas he had in his collection specimens, so named, of what I have identified as L. nove-celandiw. It is a small form in which the zocecial tubes of the much raised inner zocecia of the rays meet or nearly ineet in the centre, forming when there is no ovicell a very depressed inverted cone. These havea raised flat ovicell* over the centre with most minute perforations and a bordering ridge; besides the cancelli are frequently formed with bars across, asin L. radiata, Hincks, and, although there are some “ pin-head” spines in the cancelli and zocecia, they are not very frequent, whereas in £. Holdsworthii the round cancelli are very regular, with very numerous large pin-head spines. In my specimens of L. Holdsworthii, in two cases, there are hollow grooves in the interradial spaces, and in other places on the central wall, which seem to relate to the formation of an ovicell such as that figured by Harmer (loc. cit. fig. 11). T have a large number of Lichenopore with ovicells, and, when writing my short paper t on them, much time was spent in comparing the Natural History Museum types and specimens. Since then I have had some additions to my collection, and the Museum has received several valuable collections, such as Busk’s and Hincks’s, but on the whole it seems that what I then said is supported. My paper was written at a time when hardly anyone took any notice of the ovicells, in fact they were generally ignored ; but certain types of ovicells were indicated, and whether the identifications will ultimately stand was a matter of secondary importance, though I am unaware of any reasons for changing the names then used. I showed that there was the flat type of ovicell as in L. nove-zelandie, the raised, rounded, dome-shaped type with trabeculee over the wall as in L. echinata, McG., and LL. victoriensis, Waters, and, lastly, the ovicell spreading up to the rays asin L. californica (Busk) Waters. I stated that this should be considered as the species of Busk, as there was some doubt as to d’Orbigny’s description, and Harmer, in calling attention to the biserial rays of Busk not being indicated by d’Orbigny, does not seem to have remembered what I had already said. However, the question of uncertainty through Busk’s species being bi- to multiserial has no importance now, as I have a specimen from Western Port, Victoria, with uniserial rays and an ovicell as figured { by me, and another from the same locality with biserial rays and a similar ovicell. ‘The question of uni- and biserial rays has not only broken down * “Bry, New South Wales, etc.,” Ann. Mag. Nat. Hist. ser. 5, vol. xx. (1887) pl. 7. fig. 8. + “On the Ovicells of some Lichenopore,’ Journ. Linn. Soc., Zool. vol. xx. (1888) pp- 280-285. { “Bry. New South Wales, etc.,” /. c. p. 261, pl. 7. fig. 8. 3* 36 MR. A. W. WATERS ON THE MARINE as a generic character, but also in several cases as a specific one. All the same the Lichenopore have yet to be brought into order, and to do this we require more spirit-specimens ; but to place as synonyms what I con- sidered as L. nove-zelandiv, Busk, L. Holdsworthii, Busk, and L. victoriensis, Waters, does indeed seem retrograde. Harmer’s ‘Siboga’ specimens must be called L. Holdsworth, Busk *. Since I wrote on the ovicells, among others, two specimens from Ceylon, named by Busk L. Holdsworthii, have been added to the National Collection. The central part is somewhat depressed with an ovicellular wall covering it, and the rays extend into the ovicell; by the side of this there is an oceciostome tube, and on one of the specimens at a higher level there is an ovicell spreading in and out between the rays. These Cape Verde specimens of L. irregularis are not, as Norman supposed, what I called Radiopora pustulosa t, d’Orb., from Naples, each subcolony of which has some multiserial and some uniserial rays, and the subcolonies are much larger than those of the Cape Verde Island forms. The specimens alluded to, from Naples, are not closely attached, nor were Peach’s 1. mean- drina, Although unable to see any material difference between the recent and fossil R. pustulosa, there is some uncertainty in identifying a fossil about which we know little, and therefore it would have been a safer course to have named the recent one L. neapolitana, as it does not seem to be the same as L. hispida, Flem. The specimens of L. meandrina, Peach, in the Nat. Hist. Museum from Busk’s and Hincks’s collections, have much smaller suBcolonies than what I called pustulosa, d’Orb., and the zocecia are 0:07-0:08 mm., while the cancelli are about 0°06 mm.—that is, smaller than in pustulosa, where they are about 0°1 mm., with the zocecia about 0:09 mm. In the Museum specimen of ‘‘meandrina” the subcolonies are often very elongate, instead of circular, and there are more definite serial ridges than in the solitary L. hispida, Flem., and we may certainly doubt the advisability of considering the composite form a variety of hispida. Peach’s figure and description of ZL. meandrina correspond more with the Mediterranean pustulosa than do the Museum specimens, which are not from Peach’s collection. However, I think we have as separate species L. irregularis, Johns., L. pustulosa, 7 Orb. (Waters), and L. meandrina, Peach. The ovicells are unknown in the last two species. Other composite species are L. pristist, MacG., L. magnifica, MacG., L. bullata, MacG., and several fossil ones are knownas L. (Radiopora) formosa§, d’Orb., L. (Radiopora) Francquana||, d’Orb., * There is, however, a small specimen from between Nusa Besi and N.E. point of Timor (St. 282) with long erect biserial rays, which may almost be said to be small bundles of rays, and the central zoccia are very irregular. This I cannot consider to be the same species as the other so-called nove-zelandie. + Pal. Frane. pl. 649. figs. 1-4. } Trans. Roy. Soe. Vict. vol. xx. (1883) p. 126, Pal. Franc. pl. 782. figs. 1, 2. || Pal. Franc. pl. 782. figs. 3-8, BRYOZOA OF THE CAPE VERDE ISLANDS. 37 L. (Radiopora) multistellata*, W’Orb., L. conjuncta t, Mich., L. cumulata t, Mich., L. (Multitubigera) micropora §, Reuss, L. (Multitubigera) gregaria ||, @Orb., L. suecica J, Hennig. The early stage of Lichenopora is flabelliform and is like that of Tubulipora (say, fabellaris,Fabr.),but it soon completes the circle and then the flabelliform growth can only be seen by examining the under side. Now clearly in a colony of L. irregularis, L. meandrina,and L. pristis there is only one primary, and the new subcolonies may be said to bud out from the older ones (see figs. 1, 2, 3, 4, 5), and this requires further study with sections. It does not seem strictly correct to speak of these as confluent, as is usually done. There are also many cases of Lichenopora where the two colonial disks partially coalesce : for instance, I have specimens of L. verrucaria where the touching lamina of two disks form one growth, as well as some of the adjacent zoecia uniting. In Lichenopora boletiformis**, d’Orb., we find the same thing, and also one colony growing on the top of another, which also occurs in many other cases—as, for example, Discosparsa marginata, d’ Orb. Lichenopora with multiserial rays was separated from Discocavea with uniserial rays, but most recent workers have recognized that the difference of the rays cannot be retained as a generic character, though usually useful specifically, while some both of the confluent and simple species show that even specifically the character may not always be of value. The type is Lichenopora turbinata ff, Defrance, also photographed by Canu tt, which, although with but a small attachment, has been considered by Pergens and Canu to belong to Lichenopora as now generally understood. As described by d’Orbigny and by Canu L. turbinata has eancelli. Lichenopora is not the only genus which forms round disks, and there are a number with the zocecia radiating in bundles without pores between the rays. Many of these have been called Defrancia, others Actinopora, Apsenidesia, Pelagia, &c. These Gregory §§ would place partly in Lichenopora (having removed what has so long been considered Lichenopora to Discocavea) and partly under Apsendesia, considering that Pelayia clypeata, Lamx., was the same species as Apsendesia cristata, Lamx., * Pal. Frang. pl. 649. figs. 5-7. + Michelin, Icon. Zooph. p. 277, pl. 68. fig. 16 (1840-7). t Loe, cit. p. 319, pl. 77. fig. 1. § Reuss, “ Anth. & Bry, von Crosaro,’” Denk. Ak. Wien, vol. xxix. (1869) p. 259, pl. 36. fig. 16. || Loe. cit. pl. 752. figs. 9, 10. q “Bry. Sver. Kritsystem,” Lunds Univ. Arsskrift, vol, xxx. (1894) p. 35, pl. 2. figs. 88-36. ** Waters, “‘Cyclost. Bry. from Australia,” Q. Journ. Geol. Soc. vol. xl. (1884) p. 695, pl. 31. figs. 20, 21. tt Dict. Sc. Nat. vol. xxvi. p. 257, pl. 4. figs. 4, 6 (1823). {{ “Bry. des terr. Tertiaires,” p. 138, pl. 17. figs. 13-15, in Ann. de Paléont. (1907). §§ “ Cretaceous Bry.,” Cat. Fossil Bry. in Brit. Mus. p. 247 (1909). 38 MR, A. W. WATERS ON THE MARINE but from Lamouroux’s figure of cristata nothing can be made out ; however, Michelin, Haime, and Gregory have given figures of what they believed was the Apsendesia cristata. Haime thought that the two genera must be united, while Gregory * considers that they should be united as one species, but since Pelagia clypeata, Lamx., occurs, from various localities, as a complete and mature species I cannot at present accept this view. Perhaps it would be most reasonable to retain Defrancia, but time will show, and in this group, in all cases with which I am acquainted, the ovicell is near the border of the zoarium. In a specimen of “ Actinopora regularis,” d’Orb., which I collected from St. Croix, Switzerland, the locality from which d’Orbigny described it, there is fortunately an ovicell extending from one ray to the next and not very large. The ovicell was not previously known, and my specimen is as good as a co-type, and shows that it is not Lichenopora but Defrancia, if this name is retained. For a long time, while progress was being made with the classification of the Cheilostomata, that of the Cyclostomata seemed most hopeless, but now that more is being learnt about the ovicells and other characters, there is every reason to hope that the collection and study of more species will make the classification more natural. Harmer in his ‘ Siboga’ Report, received after the above descriptions were written, considers that Lichenopora should be used for what we have so long understood by this name, and Pergens, Canu, and others have expressed the same opinion. Loc. Madeira (Johnson, Norman); St. Vincent, Cape Verde Islands, collected by Crossland per diver. LICHENOPORA IRREGULARIS, var. COMPOSITA, var. nov. (PI. 4. figs. 1, 4.) This is the bi-multiserial form mentioned as occurring on the same stone as L. irregularis, Johnson, and about which I have been in doubt as to whether the two forms should be united as one species. The subcolonies are some- what larger than those of the uniserial form, and the zowcia, usually without any spines, are less raised, but otherwise they seem similar. FLUSTRELLA HISPIDA (Fabricius), Gray. For synonyms see Miss Jelly’s Catalogue, and add :— Flustrella hispida, Prouho, “ Recherches sur la larve de la Flustrella hispida,” Arch. de Zool. Expér. et Géner. 2nd ser. vol. viii. (1890); Robertson, “ Bryozoa,” Harriman Alaska Exped., Proc. Wash. Ac. Se. vol. ii. (1900) p. 331; Nordgaard, Hydr. and Biol. Invest. in Norwegian Fiords, p. 175 (1905). Only a very small piece was found, but it was enough for certain deter- mination. This is widely distributed in northern seas, and is common on the *« Jurassic Bry.,” Cat. Fossil Bry. in Brit. Mus, p. 171 (1896), BRYOZOA OF THE CAPE VERDE ISLANDS. 39 British and North French coasts; and Dr. Alice Robertson reports it from California, but if has not been mentioned from the Atlantic or from the Mediterranean. Hincks describes a form from Victoria as var. cylindrica, but judging from my specimens from Port Phillip I should certainly not call it a variety of F’. hispida, for it has spines regularly all round the zocecium, which is not the case in the northern form, which also is a larger and inore solid form. Flustrella hispida has 23-27 tentacles. Harmer * considers that the genus /lustrella should be retained. Loc. Northern Seas; California; Boa Vista, Cape Verde Islands, 5-20 fath., collected by Crossland. AMATHIA ToRTUOSA, Tenison Woods (non Bush). For synonyms see Waters, ‘‘ Mar. Biol. of tue Sudanese Red Sea, Bryozoa,” pt. ii. p .243, pl. 24. fig. 5 (1910). Dr. Harmer ft has doubted the determination of this species from the Sudan, perhaps partly through my figures being merely outlines. I gave three very small figures of forms believed to be three species, so that a comparison could be made of the size and position, but merely to show these characters in question. In Amathia the lower parts of the zocecia are usually connate, whereas the upper part may be more or less free, varying in appearance according to the condition of the specimen and of the mount. My figure 5 of A. tortuosa (also figured by MacGillivray) is depicted looking down on the top of the zocecia when the upper part is seen separated, giving naturally a short appearance, but it was advisable to show it in this position. The length of the zowcia from Cape Verde Island are about 0-4 mm., with the lower half of the zocecia connate and the upper free. With regard to my figure of A. distans, Busk, the scale was too small for it to be possible to show the ends in detail, buat it will be seen that in two or three cases projecting zocecia are shown, and this was more distinct in the original drawings, for though taken from actual specimens the size prevented their being more than diagrammatic in most respects ; the length of the zocecia is relatively the same as figured by T. Woods ft, and surely if he had been speaking of the length of the zocecia he would haye said the length of the cells and not of “the pair: of cells,” and he meant the length of the group of the pair of cells. This unfortu- nate sentence of T. Woods has misled several authors, but the figure does not show long cells. My determination of the Sudan and Zanzibar * Polyzoa of the ‘Siboga’ Expedition, p. 40 (1915). + Polyzoa of the ‘ Siboga’ Expedition, Hunt., Ctenos. & Cyclos. p, 68 (1915). =~ “On the Genus Amathia,” Trans. Proc. Roy. Soc. Victoria, vol. xvi. (1879) p. 90, fig. 6. 40) MR. A. W. WATERS ON THE MARINE Amathie was made after prolonged comparison of the British Museum and ‘Challenger’ specimens, and I cannot think there is anything material to alter in what I then wrote. In the Cape Verde Island specimens there are usually 14 pairs of zocecia. The zoarium grows from a seaweed, and the free stolon when passing over the narrow branch of any part of the seaweed becomes attached across it, again growing free beyond it. Where the stolon is thus attached the zocecia are arranged horizontally on each side of it, thus forming two lateral series without any sign of the spiral arrangement. From the front of the attached stolon, and in the middle of the group of zocecia, one or two fresh erect branches arise, a mode of growth and branching quite unknown in Amathia, raising rather large questions On the same seaweed were entangled pieces of Amathia Vidovici, Heller, but on careful separation the colonies were found to be quite distinct. The seaweed having been some time in spirit has entirely lost any colour it may have had, and resembles a figure I have before me of Polyides rotundus, Grev., which may by now be known by another name. I followed MacGillivray in considering that A. conneaa, Busk, was a synonym, although stouter; of this lam now doubtful, and in the determina- tion of some of these spiral forms we are on rather uncertain ground. A specimen in the British Museum determined as A. connewa by Busk I found had the stem only 0°2 mm. diameter, which is much below Busk’s measure- ment. A. conneva is the largest of these spiral forms with more or less free ends, then comes} A. tortuosa, then A. distans, and there is also A. Vidovier, with the zocecia much separated, but it may be the differences are not as important as now supposed. In A. obliqua, MacG., the ends are free, and even in A. lendigera, Lamx., a considerable part is often free; and while we usually find that the species with the zocecia almost entirely connate have the two series also united, yet in A. lendiyera the two series may sometimes be found almost unattached in the middle. | We must remember that we are not quite sure what characters are of most value in Amathia. The Zanzibar A. distans has the ends of the zocecia free as in A. tortuosa, and a small fragment from Sydney of what I have always considered to be A, distans has the zoarium and zoccia the same size as those from Zanzibar and the ends are free. Loc. New South Wales ; Victoria ; Red Sea; Cape Verde Islands, collected by Crossland. BRYOZOA OF THE CAPE VERDE ISLANDS. 41 Amatuia Vipoyict (feller), Waters. Amathia Vidovici, Waters, “ Marine Fauna of Brit. East Africa and Zanzibar,” Cyclos, Ctenos. & Endoprocta, Proce. Zool. Soc. London, 1914, p. 848, pl. 4. figs. 1, 2, which see for synonyms, This is growing on the same seaweed as A. tortuosa, T. Woods. There are usually about 8-9 pairs of zocecia (about 0° mm. long) in a short spiral near the joint, with the lower part connate, the upper part free, as in A. tortuosa ; and this is the rule in A. Vidovici from other places, but a specimen from Genoa has the zoccia more free—in fact, some of the zocecia near the ends of the branches are quite free. From the stem there are a few radicles with digitiform processes. The stem is about 0°14 mm. in diameter, and there is a certain amount of variation in the thickness of the stems, but the piece measured from Zanzibar must have been abnormally compressed, as most are approximately the same as the present. I find that the piece sent to me by Joliet as Serialaria semiconvoluta, from Roscoft, has biserial zooecia arranged in a spiral which is longer than that of Vidovici, being somewhat like A. tortuosa, and the zowcia are much separated, so that many are quite free, which led me to think it was a modified form of A. Vidovici, but there is a little uncertainty as to what it should be called. We have seen that A. tortuosa, Woods, A. distans, Busk, and A. Vidovici are very closely related. I have a sketch of an Amathia in the British Museum, trom the Aigean Sea, named Serialaria lendigera, 91.71.4878, which seems to be A. Vidovici. Loc. Mediterranean ; Bermuda ( Chall.) ; Brit. Hast Africa ; N.E. Coast, United States ; Cape Verde Islands, collected by Crossland. ZOOBOTRYON PELLUCIDUM, Lhrenberg. For synonyms and localities see Waters “ Bry. from Zanzibar,” Proc. Zool. Soc. 1914 p- 849; and Journ. Linn. Soc., Zool. vol. xxxi, (1910) p. 243, pl. 24. figs. 12, 15. The specimens from Boa Vista, Cape Verde Islands, correspond closely with those from Naples and the Red Sea. As previously said, I think that Zoobotryon and Bowerbankia will have to be united. Harmer does not agree, as he thinks the zoarial characters are sufficient to warrant the separation, but the only difference is whether the branching is usually with two new branches or three, whereas in a closely allied genus, Amathia, there are considerably greater differences, without new genera being made. For example, A. Wilsoni, Kirkpatrick, an un- doubted Amathia, has also three branches at a joint. BoWERBANKIA PuSTULOSA (Jillis § Sol.) Hincks, var. ALTERNATA, var. nov. Specimens from Boa Vista, growing on Zoobotryon pellucidum, Khr., as well as on the supporting seaweed, and others from St. Vincent Harbour, 42 MR. A. W. WATERS ON THE MARINE 10 fathoms, are smaller than B. pustulosa*, Hincks. The stolon (0°04 mm.) is smaller, the zocecia (0°35-0°45 mm.) are shorter, and the gizzard is about 0°07 mm. in diameter. The zoarium is usually creeping, but small pieces are free, some of which may have been attached. From the creeping stolon the zocecia grow on the opposite sides, though not as a rule in pairs, but alternate in groups of four to ten (usually about ten) : after a group there is a diaphragin, followed by a bare stolon for a moderate distance, and then the next group. A few branclies grow at right angles to the main stolons, and then there is a diaphragm near the beginning of the new branch. : Var. ALTERNATA. This is much like B. gracillima t, Hincks, but the stolon is smaller in the var. alternata, also the zocecia are smaller than those of typical pustulosa, whereas Hincks describes his gracillima with larger zocecia than those of pustulosa, though from his figure there does not seem to be much difference. Wherever this may ultimately be placed, it seems safest to now call it var. alternata. Loc. Boa Vista and St. Vincent Harbour, 10 fath., Cape Verde Islands, | collected by Crossland. BARENTSIA DISCRETA (Busk), Karkp. Ascopodaria discreta, Busk, Zool. Chall. Exp. vol. xvii. p. 44, pl. 10. figs. 6-12 (1886). Pedicellina australis, Jullien (non Ridley), Mission Scientifique du Cap Horn, p. 13 (1888). Barentsia discreta, Harmer, Polyzoa of the ‘Siboga’ Exp. p. 29, pl. 2. figs. 8, 9 (1915). As Harmer gives full references I have struck out mine and refer to his paper for synonyms. I have previously mentioned that the connection of the pedicel with the polypide is subject to considerable variation, sometimes the rings mentioned by Busk are seen, but in other specimens there are none. ‘This has been confirmed by Osburn. Asajiro Oka t has described Barentsia misakiensis as * Harmer, in his “ Polyzoa of ‘Siboga,’” p. 72, has, from my figure, doubted whether I had obtained Bowerbankia imbricata from the Sudan (Journ. Linn. Soc., Zool. vol. xxxi. (1909) pl. 25. fig. 8). But I made sketches of the three conditions of the zocecial chamber, and it was intended to be obvious that fig. 8 represented a zocecium in which histolysis had quite altered the polypide, but, nevertheless, in spite of these great changes having taken place, the shape of the zocecial chamber was not as yet at all altered. If the object had been to show a perfect zocecium a much more detailed figure would have been drawn, and certainly a little further explanation on my part would have avoided false conclusions. The figure merely represents any typical B. imbricata. + Hincks, Brit. Mar, Poly. p. 525, pl. 75. fig. 6 (1880), t “Sur la Barentsia misakiensis,’ Zoolog. Magazine, Tokyo, 1895, pp. 76-86, pl. 12. figs. 1-8. BRYOZOA OF THE CAPE VERDE ISLANDS. 43 closely allied to Barentsia discreta, which he, however, distinguishes by the number of tentacles, his species having 20-24, whereas Busk describes B. discreta as having 16-20 ; it, however, seems doubtful whether they should be separated, and Harmer also unites the two species. Loe. Tristan da Cunha, 100-150 fath. (Busk) ; China Seas, 27 fath. ( Kirkpatrick) ; Ceylon, India, Chile, Cape Horn, 40 met. (Jullien) ; Vineyard Sound and Buzzard Bay, Beaufort, N.C., and Tortugas Isl., Florida, 18 fath. (Osburn) ; 5° 28' 8.-184° 53’ E., 57 met.; 1° 42’ S.-130° 47’ E., 32 met. ; Makassar, 0-32 met. ; Saleyer, 0-36 met. (Harmer) ; St. Vincent Harbour, Cape Verde Islands, from under stones, collected by Crossland. PEDIGELLINA CERNUA (Pallas), Smitt. For synonyms see Miss Jelly’s Catalogue, and add :— Uljanin, “ Anat. & Entw. Gesch. der Pedicellina,’ Bull. Soc. Imp. de Natur. de Moscou, 1870, pls. 5, 6; Harmer, ‘ Life-Hist. of Pedicellina,” Q. Journ. Micr. Se. 1886, p. 239; Jullien, Mission du Cap Horn, Bryozoaires, p. 9 (1888); Seeliger, ‘“‘ Ungesch. Verm der Endoproct. Bry.,” Zeit. Wissen. Zool. vol. xlix. (1889) pls. 9, 10; Levinsen, Zool. Danica, ‘Mosdyr,’ vol. iv. p. 96, pl. 9. figs. 18-29 (1894); Calvet, “ Bry. Mar. de la Reg. de Cette,” Trav. Inst. Zool. de l’Univ. de Montpellier, ser. 2, mém. 11, p. 94 (1902); Jull. & Calvet, Bry. prov. des Camp. de l’Hirondelle, p. 25 (1903); Norman, “ Nat. Hist. East Finmark,” Ann. Mag. Nat. Hist. ser. 7, vol. xi. (1903) p. 574; Osburn, “ Bry. of Woods Hole Reg.,” Bull. Bur. of Fisheries, vol. xxx. (1912) p. 213, pl. 18. figs, 8a-d; Osburn, “ Bry. of Tortugas Isl., Florida,” Pub. Carnegie Inst. p. 185 (1914). The specimens on Scrupocellaria Bertholetti, Hincks, from St. Vincent, Cape Verde, are about 0°4 mm. long, whereas British forms in my collection are about 16-2 mm. long. Loc. Northern Seas; British and French coasts, Mediterranean, Australia (Kirkpatrick) ; Senegal, Gulf of Guinea, Liberia, Egypt, Smyrna (Jullien & Calvet) ; Canary Islands (Jullien); Woods Hole Region (Coburn) ; Cape Verde Islands (Jullien § Calvet); St. Vincent, Cape Verde Islands, collected by Crossland. : EXPLANATION OF THE PLATES. PUATES Ls Fie. 1. Bugula dentata, x 25. The tissues drawn dark are coloured a dark blue, and this coloration is specially marked in the growing tissue and tips of the tentacles. 2. Beania hirtissimu, X 25. Dorsal surface. Serupocellaria Macandrei. (a) Operculum, x 250; (b) seta, X 85; (¢) base of seta, X 250; (d) mandible of lateral avicularium, x 250: (e) mandible of anterior avicularium, x 250. 4. Do, do. Avicularian chamber decalcified, x 250. Do. do. Base of seta with muscles and lateral chitin pieces (cA.), x 250, ot Fig. MR. A. W. WATERS ON THE MARINE 6. Serupocellaria Macandrei. Avicularium, x 100. U6 Do. do. x 865. 8. Membranipora quadricornuta, sp. nov., X 25. The distal spines usually stand erect, but some are given bent down to show the shape. 9. Serupocellaria tridentata, sp. nov., X 85. 10. Do. do. Mandible fitting into the avicularium. 11. Serupocellaria Macandrei, x 85. Decalcified, showing the position of the ovary (ov.) and the testes (¢.). PLATE 2. 1. Schizoporella trichotoma, sp. nov., X 50. 2. Do. do. x 85. Operculum. 3. Do. do. x 250. Pores in decalcified membranes of the frontal surface of the zocecia, showing the larger and smaller rings on the inner and outer wall. 4. Do. do. x 12. Showing the primary zowcium. 5. Schizoporella oligopus, xX 25. 6. Do. do. x 25. Dorsal surface. 7. Holoporella pusilla, x 25. 8. Do. do. x 85. Operculum, 9. Do. do. x 85. Mandible. 10. Schizoporella spongites, X 25. Showing ovicell and vicarious avicularium. nil, Do. do. Xx 85. Showing the orai denticles from the front. 12. Do. do. x 85. Oral aperture showing the denticles from the interior. 13. Do. do. x 86. Operculuin. From Manaar. 14. Schizoporella unicornis, X 85, (w) Operculum in the oral aperture from lower layer with closure and small tubule; (4) do., with large tubule; (¢) do.; (d) do., showing connection from the tubule to the distal wall of the zocecium ; (e) do., with calcareous covering commencing over the distal part of the operculum. 15. Do. do. x 85. Operculum. 16. Do. do. x 25. Zocecium with part of the front wall removed, showing the zocecium of a lower layer in an exactly similar position. 17. Do. do. x 4. Side view of the zoarium, showing the superimposed layers of the zocecia, with some, ovicells. 18. Schizoporella viridis. Similar multilateral zoarium, showing the irregular position of the layers. From the Red Sea. 19. Schizoporella porellformis, X 25. Showing walls commencing for the next layer, and these walls crossing over the opercula. [From Port Elizabeth, S. Africa. 20. Do. do. x 60. 21. Do. do. x 85. Operculum. 22, Schizoporella unicornis, X 85. Abnormal double operculum. PLatTE 3. 1. Crista tubulosa, X 26. 2. Crisia vincentensis, sp. nov., X 25. Showing long ovicell. Do. do. x 85. Oceciostome. XXXIV. Pl. 1. Journ Linn. Soc. Zoon. Voreeee Pet Waters Gambridge University Preos A.W Waters, del. CAPE VERDE ISLANDS. (BRYOZOA) = , ‘ x n 5 i _ i ’ 3 ee r . 4 ne : . ; e / 4 . ; v a . i 5 i ; ' S 5 : j A 7 ‘ : ; j ‘ ' . é = a : : : z . 23 t XXXIV. Pl. 2. Waters Journ Linn. Soo. Zoor Vou See > y) h 8s Nyon) BIB x i A.W Waters, del. Cambridge University Press CAPE VERDE ISLANDS. (BRYOZOA) Ay XXXIV. Pl 3. Journ. Linn. Soc. Zoou Vora Pes. Waters Cambridge University Prees. A.W.Waters, del. CAPE VERDE ISLANDS. (BRYOZOA) ay Waters A.W Waters, del. XXXIV. Pl 4. Journ Linn.Soc. Zoo. Vo, *S4-ee+= A See see en) enna ane / yes Cambridge University Press CAPE VERDE ISLANDS. (BRYOZOA) Fig. BRYOZOA OF THE CAPE VERDE ISLANDS. 45 4, Tubulipora Lamourourti, X 25. (o@c.) oceciostome. (a) X 3. 5. Do. do. x 25. Dorsal surface of fig. 4. 6. Do. do. x 25. The lower part is obscured by Lithothamnium growth. 7. Tubulipora incrassata, X 25. The magnification of this is the same as of tigs. 4, 5, 6, 10. (Juv.) is the early growth of 7. Lamourouavit, Aud., and the difference in the size of the zocecia of the two species is very marked. 8. Tubuliporu pulchra, X 8. 9. Do. do. x 25. Dorsal surface. 10. Tubulipora Lamourouati, X 25. (a) xX 3. ibe Do. do. x 12. Primary and a few zocecia. PLATE 4. 1. Lichenopora irregularis, var. composita, var. noy. This form is bi-multiserial, x about 4. 2. Lichenopora irregularis, X about 4. This form is uniserial and there are ovicells. 3. Do. do. x 2:5. Small uniserial colony. 4, Lichenopora irregularis, var. composita, var. noy. Uniserial form showing the elongate extension of the subcolunies. 5. Lichenopora irregularis, x 3'5. Ovicells showing several pits. 6. Do. do. Ovicell of fig. 2, more magnified. is Do. do. Showing pit at the base of which the upper wall of an ovicell is seen, proving the existence of a subcolony at a lower level. 8. Do. do. x 25. Zocecia showing the spinous elevation surrounded by cancelli. 9. Thalamoporella Rozierti, X 25. Dorsal surface, showing through the walls bundles of spicules near the opesiules. (a) Spicule nearly straight, x 85; (6) spicule “curve,” x 250; (c) bent spicule, x 250. It should be noticed that the last two are more magnified than (a). 10. Adeonella contorta, x 85. Aperture. 11. Do. do. X 26. Zocecia near growing end. 12. Do. do. x 85. Mandible. 13. Do. do. X 25. Operculum. 14. Do. do. x 25. Lateral view, showing large vicarious avicularia. 15. Do. do. Natural size. MeN A potearklyet a. tg ON THE FEEDING-HABITS OF FISH AND BIRDS. AT Some Observations on the Feeding-habits of Fish and Birds, with special reference to Warning Coloration and Mimicry. By J. C. Morrraw, M.B. (Lond.). (Communicated by Prof. E. B. Pounron, F.R.S., DES ) (Puate 5.) [Read Ist March, 1917.] MarerraL tor this paper has been extracted from a journal, wherein are recorded observations of the riverside habits, and especially the feeding- habits, of fish and birds during the years 1909 to 1915 inclusive ; they were not collected with any special object in view, and were made in the British Isles and during one season in New Zealand. The paper is divided into two parts: the first deals with observations which have a bearing on the hypothesis of Warning Coloration ; the second with those which appear to throw fresh light on certain aspects of Mimicry. PART 1, Feeding-Habits of Salmo fario, Linn. (the Brook Trout ). The fish is strictly carnivorous. Its food consists of small fish, crustaceans, molluses, annelids, aquatic and floating insects. Vegetable matter was found on only very few occasions out of five to six hundred autopsies. In New Zealand, on one occasion, the stomach of a fish was filled with Spirogyra, Link ; subsequently it was proved that the fish took the weed in order at the same time to capture a small Trichopterous larva. The yellow bloom of the furze, Ulex europeus, Linn., was also taken on account of a small grub, probably Tineina. In this country pieces of Ribbon-weed, Potamogeton, Tourn., are taken in order to obtain Stmulium, Latr., colonies, either larva or pupa. Sometimes small pieces of wood, of grass stems, and grass and other seeds are found in the alimentary canal; these will be accounted for subsequently. When the fish are feeding on floating insects conditions are particularly favourable for detailed observation : the insect can be clearly seen and cannot be taken by the fish without a marked disturbance of the surface. The under-water feeding is less easily observed, but examination of stomach and intestinal contents gives reliable information. The feeding-habits depend to a large extent on the quantity of the food- supply. In waters where food of all kinds is abundant, as in many chalk- streams, the fish, although they have the choice of many kinds of food, 48 MR. J. C. MOTTRAM ON THE always take a mixed diet. It is rare on opening a fish to find the remains of only one kind of food : in one part of the intestine will be found, for instance, snail-shells ; in another fish-bones, and in the stomach insect-remains. The arrangement of the food in the alimentary tract indicates that the fish take their food in batches, the remains of the various foods are more often than not unmixed and occupy different parts of the gut. Observations in every respect confirm this: a fish picking up snails from the river-bed will never stop to take a floating insect; fish taking floating insects will entirely ignore a shoal of minnows in the neighbourhood ; on the other hand, when a fish is taking minnows other foods are neglected, This hunting of one prey at a time is most clearly seen when the fish are taking floating insects. It is the rule in chalk-streams to observe fish taking only one species of insect when there are two or more on the water. These observations indicate that fish are subject to what may be called a special appetite, as distinguished from a general one. When a fish is observed to be feeding on one insect and neglecting another, it cannot he concluded that there is more than a temporary difference in palatability between these two insects ; it is not unlikely that another fish may be seen feeding on the second insect and neglecting the first. There is some evidence that these individual and temporary preferences depend upon what food the fish has had in the immediate past ; if, for instance, the fish has recently fed upon insects, it will be more likely to take some other food when next hungry. In many rivers during early June Ephemera danica, Miill. (the May-fly), hatches out in immense quantities and the fish at once gorge themselves with it ; nevertheless, towards the close of the May-fly season, the fish may often be seen taking other food, whilst M. danica is hatching out in abundance the fish may even be seen taking smaller species of the Ephemeride or some species of Diptera. Obs.— 15. 6.113, R. Kennet, “ Found two trout taking ‘ Olive Dun’ (sub-imago of Baétis vernus, Curtis) and neglecting the May flies which were hatching out at the same time. Also saw several dace (Leuciscus leuciseus, White) feeding upon ‘Reed Smuts’ (Stmuliwm, Latr.) and neglecting the May fly.” It is well known to fly-fishermen that for some weeks after the May-fly season the fish more or less neglect insect diet: ‘‘ The -May-fly carnival means a long interval, during which the rod may as well be put by, for the glutted fish take a ‘cure’ of at least three weeks, during which they abstain from insect-food of all descriptions” (from ‘ Happy Hunting Grounds,’ by A. EH. Gathorne-Hardy). In some streams Brachycentrus subnubilus, Curtis, (Grannom) hatches out in immense quantities during April, and, in a similar way, the trout quickly satiate themselves with this insect, FEEDING-HABITS OF FISH AND BIRDS. 49 A general survey of the feeding-habits of trout, where food is plentiful, brings out three distinet facts: (1) the fish prefer mixed diet ; (2) fish feed upon one food at a time; (3) fish have a special appetite as apart from a general appetite. In waters where food is scarce the fish only grow to a small size. Under these conditions their feeding-habits are entirely different: they take all food that comes within their reach, they are forced by general hunger to eat that which is supplied, they have no opportunity of exhibiting preference. Their method of taking food is shown on examination of their stomach contents ; the following are two typical examples :— “Aug. 21, 1905, R. Lambourne, Great Shelford, part of river where food is scarce, stomach contents were 2 house flies, 2 bluebottles, 1 wasp, numerous reed smuts and other small Diptera, 2 shrimps, 1 snail. * Sept. 10, 1910, Pennel burn, North Wales, stomach contents were 1 grasshopper, 2 wasps, 1 black beetle, several small species of Coleoptera, 1 large Ichneumon fly, 1 daddy-longlegs, 2 hoppers, 6 house flies, 2 caddises, numerous small Diptera, numerous red and black ants male and female, several duns and spinners (sub-imago and imago of Ephemeridze).” Had observations been confined to such fish, conclusions would have been drawn that trout appear to have no preference for particular foods. Evidence that Trout recognize a Difference in Palatability between different Fuod. In order to prove that fish do discriminate, a long series of observations must show that, although the fish take both foods, nevertheless, when they have the choice of both, they much more often take one than the other. During May the trout have often choice of two floating insects, the sub- imago of Baétis vernus, Curtis (the “Olive Dun”), and the imago of Bibio johannis, Meig. (the “Black Gnat”). It will be seen that some fish are taking both insects, others the Bibio only, and others the dun. ach fish thus falls into one of the three classes: by observing a number of fish, distribution of feeding-habit is obtained, as in the following observation :— May 16, 1910, “Black gnats and olive duns on the water from 10,0 to 11.30 a.m. : of 20 feeding fish, 11 were taking duns only, 7 black gnats only, and 2 were taking both flies.” Similar observations made on many days and on different waters show that the dun, Baétis vernus, is taken more frequently than the gnat, Biblio johannis. The number of fish taking both inseets was found to depend on the stream on which the observations were made ; where the food-supply is LINN. JOURN.—ZOOLOGY, VOL. XXXIV, 4 50 MR. J. C. MOTTRAM ON THE scanty the fish most often take both insects, no doubt because their general appetite is very good. Neglecting this class of fish, the following figures were obtained :—Number of fish observed 201, fish taking only duns 149, fish taking only black gnat 52. Similar observations indicate that the sub-imago of Baétis pumilus, Burm. (“Tron Blue Dun’’), is more palatable than B. binoculatus, Leach (‘“ Pale Watery Dun”), and that the imago of the smaller Ephemeridz is more so than Simulium (“the Reed Smut”). . Even these observations are not quite free from doubt, the relative prevalence of the insects may play a part, or one insect may be more easily captured than another, or more easily seen. Nevertheless, there is stronger evidence that fish do appreciate a difference in palatability. There are two species of the Ephemeride, Heptagenia sulfuria, Mill. (‘‘ Yellow May Dun”), and Leptophlebia marginata, Haton (“Turkey Brown Dun”), which are almost never taken by trout. A ‘“‘ Yellow May Dun” (sulfuria) has never been seen to be taken by a trout, though a great number have been observed floating down where fish were feeding. On two occasions considerable hatches of this insect were seen, but the fish left them alone. Obs.—‘“‘ Sept. 25, 1909, R. Teme, Eardistone, saw a good hatch of the ‘ yellow May dun’ (H. sulfuria), many on the water at the same time; this is rarely seen ; trout and grayling rising at other flies, but no fish feeding on the ‘ May dun.’ “« Sept. 28, another hatch of the ‘ yellow May dun’ (su/furia), fish not taking the insect.” As regards the “turkey brown” (LZ. marginata), I have seen this insect taken on one occasion by two fish, one of which was captured and two recovered from the stomach. Obs.—‘“ 10. 5.1914, R. Lambourne, Newbury, a hatch of ‘turkey browns’ (L marginata), saw two fish take these insects: killed one and recovered two from the fish’s stomach ; this is the first time I have seen this insect taken by fish.” It is the experience of anglers that these insects are very rarely taken. These two insects are among the less common species of the Ephemeride, and it is possible that their scarcity may be a factor in causing them to be neglected by the fish; on the other hand, fish are often seen feeding on other species even when the hatch is very sparse. There are two other insects which are almost immune from the attacks of trout, Gerris thoracica, Fabr., and Velia currens, Latr. (‘“ water-skaters”). In the vast majority of trout-streams these insects abound, but I have never seen one taken by a tish; on two occasions I have recovered single specimens from the stomach. ‘The taking of these insects has been gbsenyad by anglers on a few occasions. It is possiblen the fish neglect them because thee are difficult to capture ; they are very active, buoyant, and difficult to drown, FEEDING-HABITS OF FISH AND BIRDS. Sal The only other animal which trout appear to avoid is the tadpole. In lakes and in still pools of most trout-streams tadpoles are found, but I have never seen trout feeding upon them, nor have they been recovered from the alimentary canal. On the river Exe, during 1911, a small pool was observed in which were four trout and tadpoles estimated to the number of two hundred: no diminution of the number was detected during four days. Similar observations were made in a hatch-hole on the river Lambourne at Shelford in 1910, during ten days, with similar result ; the possibility of tadpoles dropping down from above was excluded. Nevertheless, on a few occasions anglers have observed trout to be feeding on tadpoles in waters where there was a searcity of food. The colour of the tadpole deep black, its gregarious habits, its making no attempt at concealment, indicate that the apimal is not much open to attack. It is also noteworthy that the vivid yellow of the “yellow May dun” (HZ. sulfwria) makes this insect a con- spicuons object on the water ; it is certainly more easily seen than any other of the smaller British Ephemeride. These observations indicate that trout do recognise a difference in palata- bility between different foods; also that the relatively unpalatable foods are oceasionally devoured. No small animal which lives either in, or upon the surface of, fresh water has been observed to be entirely immune from the attacks of trout. The feeding-habits of other fish, in respect of floating insects, is of con- siderable interest: Thymallus vulgaris, Nilss. (the Grayling), takes floating insects as freely as trout: Leuciscus leuciscus (the Dace) also feeds freely on floating insects ; Leuciscus cephalus, lem. (the Chub), in many streams, only feeds on floating insects during the “ grannom” (Brachycentrus sub- nubilus) and ‘* May-fly ” seasons, or when an especially large hatch of some other insect occurs, in other streams it rises almost as freely as the dace ; Leuciscus rutilus, Flem. (the Roach), is very seldom observed to be feeding on floating insects, but during the ‘“ May-fly” (Zphemera danica) season it not infrequently takes the sub-imago. As will be seen, a similar sequence is found among the birds: some feed only upon the “ May-fly” (2. danica), others take other water-insects as well, whilst others again feed freely upon almost every riverside species. Before considering the significance of these conclusions with regard to warning coloration, some observations on the feeding-habits of birds will be described. Cypselus apus, Il. (Swift), Hirundo rustica, Linn. (Swallow), Chelidon urbica, Boie (Martin), and Cotile riparia, Boie (Sand-Martin), feed freely on the sub-imago of the Ephemeride, either taking the insects whilst resting on the water or after they have risen into the air. Itisrare to see a hatch of these insects without these birds hunting them, so much so that by watching 52 MR. J. C. MOTTRAM ON THE these birds the angler knows when to begin fishing; nevertheless, they occasionally become satiated with these insects. During the close of the “ May-fly ”? (Ephemera danica) season it is by no means uncommon to see the swallows hunting over the fields whilst May-flies are hatching-out. Obs.— June 5, 19138, R. Kennet, Thatcham, whilst a good hatch of E. danica was on, swallows were hunting over the meadows to the windward of the river and were therefore not feeding on the ‘ May fly,’ which for several days they have been eagerly devouring.” In contrast to their liking for the Ephemeride, they have only on a very few occasions been observed to take the “ Black Gnat” (B. johannis) : there evidently appears to be a distinct difference in palatability between these insects. A difference in the ease with which they may be captured cannot account for the selection, because the birds take other insects whose flight is quite as erratic as that of the “ Black Gnat,” and also the “ Black Gnat” is often over the water in such dense swarms that the birds would have but to fly through them in order to obtain a mouthful, a method of feeding which they often employ for the capture of many small insects which swarm in the air. Emberiza scheniclus, Linn. (the Reed-Bunting), is very adept at picking the sub-imago of the Ephemeridee off the water; it in the same way feeds upon Brachycentrus subnubilus (the Grannom), and has never been seen taking B. johannis (Black Gnat). Motacilla lugubris, Linn., M. boarula, Linn., and M/. ray, Boaap. (Wagtuails), are commonly seen taking Ephemeride : on two occasions, during ,a double hatch of “Grannom” and “Olive Dun,” it was noticed that the birds took the Ephemeride by preference, allowing ‘“ Grannom,” untouched whilst flying several yards for a “ Dun.” close at hand, to pass Obs.—‘* April 16, 1910, R. Lambourne, a mixed hatch of ‘ Grannom’ and ‘ Olive Dun,’ 11.0 a.m. to 1.0 P.M., a pair of WM. lugubris feeding on both flies; when a dun was available it was always taken though ‘Grannom’ were often closer on the water.” A second observation with J/. rayi, on the same river, was precisely similar. A large number of different species of birds collect at the river-side during the “ May-fly”’ season: Fringillidee (Sparrow, Chaffinch, Greenfinch), Sturnus vulgaris, Naum. (Starling), Laride (Black-headed Gull and Common Tern), Anatide (Ducks), Muscicapa grisola, Linn. (Flycatcher), Anthus pra- tensis, Bechst. (Meadow-Pipit), Sylviide (Warblers), and other birds which are not, as a rule, observed feeding upon the smaller species of the Ephemeride. These observations indicate that the several British species of FEEDING-HABITS OF FISH AND BIRDS. 08 Ephemeride could be arranged in an order of palatability, both for birds and fish, headed by Lphemera danica, as probably the most palatable, and Heptagenia suljuria, as the least. The habits of the Swift are of particular interest because of its great agility, and the great pace at which it hunts would undoubtedly enable it to take any flying insect—for instance, when feeding on the ‘‘ May-fly,” the impact of bird and fly can be distinctly heard at a distance of ten to fifteen yards. I have seen them take white butterflies on two occasions. Obs.—“ 7.5. 1913, R. Kennet, swifts feeding on * May fly,’ many Pierines cross- ing the river: the bird caught one with the greatest ease and at once rejected if, the insect fell on the water and was not recovered. 15.6.1910, ‘swift’ took a small garden white at a single swoop, then passed behind a willow tree, out of view.” These observations are recorded because they show that the birds could capture these butterflies did they care to. As a proof that much less agile birds are capable of capturing rapidly flying insects, it may be mentioned that on three oceasions Sparrows have been observed amongst new-mown hay successfully chasing Tryphana pronuba, Hiibn. (Yellow Under-wing Moth) ; on two occasions small flocks of these birds were observed feeding in this manner for several hours. Wagtails on several occasions have been observed systematically hunting Musca domestica, Linn. (the House-fly ), and Scatophaga stercoraria, Latr. (the Dung-fly). The Significance of these Observations with respect to Warning Coloration. It appears that whether or no an animal will prey upon another depends upon a number of factors, of which the following are the most important:— (1) The animal’s general hunger: a hungry trout will take all that comes within its reach, sometimes even tadpoles; a less hungry fish will only take the more palatable foods, whilst an almost satisfied fish will only take the most palatable. (2) The animal’s special hunger: trout, and less certainly birds, prefer a mixed diet, if an animal has lately become satiated with a particular food, then this food is temporarily lowered in the seale of palatability. (3) The readiness with which a food can be gathered : animals difficult to capture will be neglected at the expense of those easy to take. (4) The prevalence of a food-supply : trout, and less certainly birds, appear to concentrate their energies upon hunting one prey at a time; for this reason they will attack a prevalent prey in preference to a scarce one (it has been often observed that birds will feed upon insects when they are unusually abundant and which they have never before been seen to attack). 54. MR. J. C. MOTTRAM ON THE (5) The animals appreciate a difference in palatability. Possibly there are other factors which control the selection. It is obvious that in order to estimate the palatability of food all these determining factors must be taken into account. It may here be mentioned that relative palatability may itself depend upon several factors, of which taste and digestibility are not necessarily the most important. The readiness with which the prey can be disposed of is an important one: for instance, when butterflies are attacked by small birds the large expanse of wing protects the small body, the insect is likely to be seized by one or both wings, when the bird shifts its hold in order to dispose of it the wing is momentarily released and the scarcely injured insect escapes, requiring to be recaptured; further, when the insect has been killed, the stiff dry scale-covered wings are difficult to swallow, and are, as a rule, removed before the body is eaten. It would appear that a butterfly is unpalatable in this respect, apart from taste. If warning coloration advertises unpleasant taste there seems no reason why it should not also advertise difficult disposal. It is also conceivable that warning coloration may similarly advertise difficult capture ; on the other hand, conspicuous coloration would give the hunter a good mark to follow. Inconspicuous coloration is of use during motion: a covey of grouse flying over heather can only be followed with the eye fora short distance; should there be a partially white bird among the covey it can be followed over the moors for a mile or so. Entomologists find some inconspicuous butterflies very hard to capture on the wing, not because of rapid flight, but because they are being constantly lost in the background. The feeding-habits of birds upon butterflies is of special interest because selection by birds has been considered to bea factor in the production of mimicry in these insects. On this account, particular attention has been paid to the feeding of birds upon butterflies. Marshall (1) has collected together the records of the attack of birds upon butterflies up to the year 1909 and given reasons for the paucity of the evidence, which has been considerably increased since then, The evidence at present collected appears to show that— (1) Birds seldom feed upon butterflies ; (2) They are less often attacked, and are less often found in the alimentary canal of birds, than are many other insects. (Further evidence may, however; negative both these statements. When- ever accurate field-obseryations are made these instances of birds feeding upon butterflies are by no means infrequently recorded. Swynnerton has shewn that these insects are so broken up in the alimentary tract of birds that a careful microscopical examination is required for their recognition.). Or On FEEDING-HABITS OF FISH AND BIRDS. It has therefore been concluded that on broad lines buttérflies are less palatable than many other insects, and that their diurnal habits, relatively slow flight, and conspicuousness may be looked upon as warning characters. The wings of butterflies, apart from pattern and coloration, are conspicuous on account of their large size (size is a very powerful factor as regards visibility), and also on account of their opacity (transparent wings are a great aid to concealment). Although butterflies as a whole may be less palatable than many other insects, it does not follow that among them there may not be difference in palatability, just as there is in the Ephemeroptera towards fish and birds; and just as butterflies may as a whole exhibit con- spicuous (warning) characters so, among them, according to their palatability, some may exhibit more conspicuous characters than others. If animals can be arrranged in an order of palatability, and if warning coloration and protective resemblance are the result of selection by enemies, then animals liable to attack should also form a series presenting all grades, from a most pronounced conspicuous coloration to a great protective resemblance, the intermediate forms showing something of both. Some evidence has been brought forward that this occurs in the Ephemeride. In a previous paper (5), the factors for conspicuous pattern were ascertained by experiment with artificial pattern, and it was shown that among Indian Lepidoptera some of the insects present all the factors for conspicuousness ; others, some only; whilst others, only a few. Swynnerton (2) has shown that animals, which form the food-supply of any species, can be arranged in order from the most palatable to the least palatable ; and opinion is expressed that animals hear some distinctive mark whereby a preyer can recognise and distinguish the relatively unpalatable from the relatively palatable. The “distinctive” coloration need not necessarily be very conspicuous. Unpalat- able animals can afford to carry a conspicuous mark, but the relatively palatable can only carry one, which is to a small extent conspicuous, or one which is only displayed when concealment has failed. This conception is similar to the author’s, but is better in that, by “warning coloration,” it brings into line those slightly replacing the term conspicuous characters which cannot be called “warning” characters. The conception presumes that preying animals have a memorising power of con- siderable magnitude, but this presumption does not appear to be always necessary : for instance, a very strong family likeness runs through long series of protected species of butterflies. Swynnerton says of this, ‘ This can be accounted for by the advantage of maintained notoriety.” If, however, a number of different butterflies present very conspicuous characters, they must be very similar in appearance, because the majority of the factors for conspicuonsness must be present in each. Poulton (3), so long ago as 1887, observed that certain colours and patterns associated with unpalatability, 56 MR. J. C. MOTTRAM ON THE do constantly recur among insects, both in the larva and adult states, It follows that in this case unpalatability might be associated in the preyer’s mind, not with a particular form, colour, pattern, or other character, but with conspicuousness. If a bird be given a conspicuous insect which it has never before seen it will be shy in attack ; whereas when the insect is inconspicuous the bird’s attack will be bolder. Variations in edibility may be associated in the preyer’s mind with degrees of conspicuousness rather than with distinctive markings: many distinctive marks appear to be related to con- cealment rather than the reverse ; nevertheless, as a working hypothesis, Swynnerton’s conception may prove to be of considerable value. (3) Birds have been observed sometimes to exercise no choice with respect to butterflies on which they are feeding. As has been already seen, the conclusion that they do not recognise a difference in palatability cannot be drawn. (4) Birds have been sometimes observed feeding upon presumably pro- tected butterflies, Hupiceines and Danaines. Conclusions cannot be drawn that they are not relatively unpalatable; trout will sometimes feed upon tadpoles. (5) Birds have on a few oceasions (4) been observed to feed upon some butterflies, whilst neglecting others. This might be the result of a temporary special appetite of the bird, the difference in the ease of capture of the insects, or by a difference in the relative prevalence of the insects. Only after exclusion of these and other factors can it be concluded to be due to a difference in palatability. PART 2. Observations which show that Salmo fario mistakes other things for Floating Insects, and fails to distingwish Artificial Flies from Natural Ones. As already mentioned, examination of the alimentary contents often reveals the presence of small pieces of wood, stick, grass-stem, and seeds, more especially grass-seeds. If a trout, which is feeding on floating insects, be observed over « long period, it will be seen often closely to examine many such small floating things ; the majority of these the fish will neglect, a few will be taken into the mouth to be at once rejected, a small proportion is swallowed. ‘There can be no doubt that the fish mistakes these things for the insect on which the fish is feeding at the time, and it may be concluded that its vision is not very acute. If the fish is feeding on one particular insect, as is most often the case in chalk-streams, mistakes are much more seldom made than when the fish is feeding on any floating insect which the stream may bring down, FEEDING-HABITS OF FISH AND BIRDS. ay The following are samples of the observations :— Obs.—“ March 30, 1911, lake Okeraka, New Zealand, stomach contents of trout were four grasshoppers, two cicadas, and three short pieces of stick of about the same length and thickness as the grasshoppers. “July 21,1911, R. Kennet, Newbury, stomach contents were many ‘ Pale Watery Duns’ (Baétis binoculatus) and three grass seeds similar in size to the bodies of the insects. ; * July 5, 1909, R. Lambourne, Shefford, watched a fish for half an hour taking ‘blue-winged olives’ (Zphemerella ignita, Poda) and ‘Reed Smuts’ (Stmulium), fish always examined any small floating object which at all resembled the insects ; on more than a dozen occasions such things were mouthed, and twice, evidently swallowed.” More certain evidence that trout are easily deluded is provided by the fact that the dry-fly fisherman is able to kill fish with crude floating imitations of the natural insects made of fur and feather. Dry-fly fishing consists of floating over the fish and making to rest on the surface of the water, a dry artificial insect in imitation of the natural one on which the fish has been observed to be feeding. In streams which are little fished the fish are easily deluded ; but, in heavily fished waters, they become educated and are able, on close inspection of the insect, to detect the counterfeit. This education is the result of the fish having been either caught and returned, or hooked and lost on many occasions. The Plate (Pl. 5) shows a series of artificial insects photographed against a high light, as the fish see them. It illustrates the roughness of their resemblance to the natural insects. The artificial insect may be looked upon as an unpalatable insect mimicking a palatable one, and the facts show that a crude mimicry is of considerable power for delusion. Observations which indicate that Birds mistake inanimate things jor Flying Insects, and the Anglers Fly floating on the water for Insects similarly situated. If Muscicapa grisola (spotted flycatcher) or any of the species of swallows or martins be watched, over a long period of time, whilst they are feeding on flying insects, it will be noticed that not infrequently the birds start out from where they are resting, or alter their line of flight, in order to capture small pieces of feather, seed-plumes, leaf-scales, chaff, or other light bodies floating in the air, which they obviously mistake for flying insects; as a rule, closer inspection shows them their mistake. Sometimes the object is taken in the bill and subsequently dropped; or very occasionally it appears to be LINN, JOURN,—ZOOLOGY, VOL, XXXIV, 5 58 MR. J. C. MOTTRAM ON THE swallowed. The following are details of some of the observations, given to show that the above statement is not open to doubt. Obs.—‘‘ 3. 6.1913, watched a spotted flycatcher feeding upon small flying insects, chiefly gnats, observations extending over half an hour, bird seen to set out after small pieces of feather, chaff, d&c. on three occasions, and a fourth time the object was taken in the bill. “© 13. 5.1914, swallows feeding upon ‘olive duns,’ bird took small piece of feather which was not seen to be discarded. “18, 4.1910, wagtails feeding upon ‘ Grannom,’ on several occasions bird made for small floating objects resembling the insect, and twice picked these off the surface of the water, then dropped them.” Whilst making such observations, it was noted that the birds often flew long distances after passing insects, as much as twenty yards, and that many of the mistakes were then made. The dry-fly fisherman’s evidence that birds are easily deluded, is very clear. Swallows, martins, and swifts very commonly pick the artificial insect off the water, when an imitation of one of the Ephemera is being used. Two typical experiences are given :— Obs.—“ 5. 7.1909, R. Lambourne, Great Shelford, 3.0 to 4.0 P.m., hatch of ‘ Blue-winged Olives’ (Ephemerella ignita), swallows picking insects off the water as well as taking them in the air, my artificial ‘ Blue-winged Olive’ was repeatedly picked off the water and carried for a yard or two in the air, by both swallows and martins : so persistent were the birds that they interfered with the fishing. “ 90. 4.19138, hatch of ‘Grannom,’ 11.0 to 12.30, black-headed bunting picking insects off the water, on two occasions bird mistook my artificial insect.” Similar mistakes have been personally witnessed in the case of the following birds: Motacilla rayi, M. boarula, Sterna (a tern in New Zealand), Anthus pratensis, Acrocephalus schenobenus, Newton, Caprimulgus europeus, Linn., Anser (domestic duck), Podiceps fluviatilis, Deg]. & Gerbe, and species of Bat (undetermined). It is clear therefore that birds, like trout, frequently mistake crude imitations for living insects. The Importance of these Observations with respect to certain Aspects of Mimicry. It has been shown that a crude resemblance suffices to delude both birds and trout. It would appear therefore highly probable that if one insect only slightly resembled another, which the bird was avoiding, then the mimicking insect would, to some extent, be also avoided. The fact that birds have been observed to begin their attack from considerable distances would help to give value to a crude resemblance ; because at such distance all details of pattern and coloration (especially during motion) cannot be seen. FEEDING-HABITS OF FISH AND BIRDS. og These considerations are of some importance, because it is difficult to conceive how mimicry can have had a beginning, except by a variation of considerable magnitude ; further, it has been thought that birds would be able to see through a disguise, unless it were good. However, the obser- vations which have been recorded indicate that both birds and fish are easily deceived, and that a crude resemblance would be likely to give a mimicking insect some protection, and especially because birds often begin their attack from such a distance that many details of pattern and coloration are invisible. On the one hand, a bird may allow to pass at twenty feet a mimic which is only a poor imitation of a protected species; on the other hand, at two feet, a good imitation may be necessary for immunity from attack. It would follow that, although a poor imitation may gain some protection, a good mimic would gain more ;, so that from a rough resemblance, a good resemblance could conceivably be built up by the agency of selection by birds. There is ample evidence that butterflies, as a whole, are less palatable to birds than many other groups of insects. An examination of their form, pattern, coloration, and habits shows that they exhibit several characters which must render them, relatively to other insects, conspicuous in Nature. Nevertheless, if butterflies, as a whole, be considered to present some warning characters, it obviously does not follow that, within the group, there may not be degrees of palatability associated with variations in the amount of warning coloration, and that therefore there may not be also within the group mimicry. SUMMARY. Attention is especially directed to two aspects of the relations between preyer and preyed upon. First, that there are many factors which determine whether or not, at any particular time, one animal will prey upon another, of which relative palatability is only one. In order, therefore, to demonstrate a difference in palatability between various foods, all these determining factors must be taken into account. Second, that observations indicate that both birds and fish are deluded by rough resemblances to the insects upon which they are at the time feeding. BIBLIOGRAPHIC REFERENCES.°* (1) Marsuaty, G. A. K.: Trans. Ent. Soc. Lond., Sept. 1909, p. 329 (2) Swynnerton, C. J. M.: Proc. Ent. Soc. Lond. 1915, p. xxxii; Ibis, 1916, vol. iv. p. 264, and 1917, vol. v. p. 529. (3) Poutton, E. B.: Proc. Zool. Soc. 1887, p. 191. (4) Carpenter, —: Proc. Ent. Soc. Lond. 1915, pp. lxix—lxxii, Ixxv. (5) Morrram, J. C.: Proc. Zool. Soc. 1916, p. 383. 60 ON 'THE FEBDING-HABITS OF FISH AND BIRDS. PLATE 5, ArtiriciaL Lysects. (Halford’s patterns.) The Explanation is printed at the foot of the Plate. Mo??RAM. JOURN. LINN. SOC., ZOOL. VOL. XXXIV. PL. 5. 3 » 4. Female sub-imago of B. vernus, Curtis (* Olive Dun”), » 5. Male Bibio johannis, Meig. (“ Black Gnat”’). = Female Brachycentrus subnubilus, Curtis, carrying green egg-sac (“ Grannom ”), iY ’ ’ Alaa ~Sat =) DD ' : 0 7 , ' + + ‘ . { : OA, . je i We : Wey ‘ , i ’ et! u ne rn py moe Caer . s! ‘ i e \ * ea j : 7 ’ ‘ 7 ¥ 1 ; \ 7 . ; ey ‘| 1 ‘ va M < ‘ er . ’ 7 ~ at - ' ‘ * ® ON A NEW GENUS OF TERRESTRIAL ISOPODA. 61 Description of Paracubaris spinosus, a new Genus and Species of Terrestrial Isopoda from British Guiana. By Watrer E.- Couiiner, D.Sc., F.L.S., ete., Carnegie Fellow, and Research Fellow of the University of St. Andrews. (PLATE 6.) [Read 8rd May, 1917.) In June last I received from Mr. G. . Bodkin, the Government Economie Biologist at Georgetown, Demerara, British Guiana, two specimens of Terrestrial Isopods collected from decaying wood, Mazakuvi River, British Guiana, a very slight examination of which was sufficient to indicate that they differed in a number of important structural characters from the genus Cubaris, Brandt, to which they bore a superficial resemblance. A closer examination shows that whilst belonging to the Cubaridee, they are quite distinct from the genus Cularis or any of its allies, and necessitate the erection of a new genus, which I am here describing under the name of Paracubaris. PARACUBARIS, gen. nov. Body oblong-oval, segments strongly convex, whole of dorsal surface covered with short, blunt spines. Cephalon short and wide, cephalic lobes absent, epistome dorsally sloping and keeled in the median line. Hyes compound, situated dorso-laterally. Antennee with 2-jointed flagellum and elongated terminal style. Inner lobe of maxillipede without spines, termi- nating in a setaceous pad. Pleural plates of mesosomatic segments slightly excavate and terminally the 2nd to 5th bluntly pointed. Only the first seoment exhibits a slight thickening on the inner lateral margin. Uropoda extending beyond the telson; basal plate thick and flattened, exopodite articulating with the posterior inner border, short and broad and extending beyond the basal plate, endopodite long, both exo- and endopodite have a short terminal style. ‘Telson triangular, terminally bluntly rounded, proximally wider than the length. Paracubaris differs from the geuus Cubaris, Brandt, in the form of the cephalon, antenna, maxillipedes, and the first mesosomatic segment. Unlike any member of the latter genus the uropoda extend beyond the telson and in the form of both of these structures there are well-marked differences. In the position of insertion of the exopodite Paracubaris shows a relation- ship with Pseudarmadillo, Saussure*, Spheroniscus, Gerstaecker +, and * Rev. et Mug. Zool. (s. 2), vol. ix. (1857) p. 308, and Mém. Soc. Phys. et d'Hist. Nat. Genévye, vol. xiv. (1858) p. 483, figs. 45, 43 a, + Ent. Zeit. 1854, p. 314. 62 DR. W. E. COLLINGE ON A NEW GENUS AND SPECIES Haplarmadillo, Dollfus *, but differs from the first-mentioned genus in the form of the cephalon and first mesosomatic segment, from the second in the two-jointed flagellum of the antennze and in only the first two metasomatic appendages having pseudo-trachez, whereas in Sphwroniscus they are present in all the five appendages. In Haplarmadillo the flagellum of the antennze has but a single joint and the eyes are simple. The new genus is more closely related to the genus Minca, Pearse f, recently described from Colombia, which possesses a two-jointed flagellum, a triangular-shaped telson, and broad flat uropoda, which extend beyond the metasome. PARACUBARIS SPINOSUS, sp. nov. (Plate 6.) Body oblong-oval, segments strongly convex, dorsal surface covered with short blunt spines, finely granulose. Cephalon (figs. 1 & 2) short and wide, with well-marked anterior and posterior margins; cephalic lobes absent ; epistome dorsally sloping from the anterior margin of the cephalon, keeled in the median line, concave laterally. Eyes compound, situated dorso- laterally. Antennule (tig. 3) 3-jointed, distal joint with few stout spines laterally and fine terminal style. Antenne (figs. 4 & 5) of medium length, Ist joint short, 2nd grooved on the outer side, 5th joint elongated ; flagellum 2-jointed, distal joint the smaller, with elongated terminal style. First maxilla (fig. 6), outer lobe terminating in four outer, stout, curved spines and eight smaller inner ones; inner lobe terminally rounded, with two spines covered with fine short setee. Maxillipedes (fig. 7), the terminal joint of the outer palp is multispinous, two multispinous processes arise from the mniddle joint, and there is a single large spine on the inner border of the first joint ;.the inner lobe is without teeth or spines, but has a well-marked setaceous pad terminally. The segments of the mesosome are strongly convex, pleural plates all slightly excavate, those of 2nd to 5th segments terminally bluntly pointed, remainder truncate, posterior angles undeVeloped. The first segment has a slight fold on the outer margin, which ventrally appears as a thickening (figs. 8 & 9). Thoracic appendages normal, enlarging slightly posteriorly. Uropoda (fig. 10) extending beyond the telson ; basal plate thick and flattened, with slightly raised diagonal crest, exopodite short and broad, articulating with posterior inner border and extending beyond the basal plate a little, endopodite long, both spinous and with small terminal styles. ‘Telson (fig. 11) triangular, terminally bluntly pointed, proximally wider than the length. Length 20°35 mm. Colour (in alcohol) greenish- brown with lateral areas on the mesosome of brown and white mottling. Habitat. In decaying wood, Mazakuvi River, British Guiana, May 1916 (G. E. Bodkin). * Proc. Zool. Soc. Lond. 1896, p. 399. + Proc. U.S. Nat. Mus. vol. xlix. (1915) p. 544, fig. 6. Collinge. S.R.K. del.ad nat XXXIV. PL 6. Journ. Linn. Soc. Zoon. VoL. SOP e- C. Hodges &Son. Lith. & imp. PARACUBARIS SPINOSUS. OF TERRESTRIAL ISOPODA FROM BRITISH GUIANA. 63 Type. In collection of W. E. ©. Paracubaris spinosus is of special interest in that it represents a type of Cubaridee only known as yet from the Western Hemisphere. The short wide cephalon is very different from the form seen in Cubaris, Brandt, and the sloping epistome, present in some species of this last-mentioned genus, is also present here, a fact which would seem to considerably weaken the importance attached to this particular character by Budde-Lund, who regarded it as of generic importance. The antenne are distinctly Cubarid, and differ from those in Cubaris only by the absence of the grooves on the outer side, and in that the distal joint of the flagellum is shorter than the proximal one. The maxillee offer no characters of importance, but the inner lobe of the maxillipede is curious in possessing a setaceous pad terminally in place of the usual tooth-like spines. This feature occurs in many genera of Terrestrial Oniscoida, but not, so far as [ am aware, in any belonging to the Cubaridee. The pleural plates of only the second mesosomatic segment exhibit any thickening of the coxopodite, and these are of a very simple nature (figs. 8 & 9). The uropoda are very distinct from those of any other genus of the family, being thick and flittened with the exopodite inserted on the posterior inner border of the basal plate. They extend beyond the telson. On the outer lateral and posterior border of the basal plate are a number of strong short spines, and a few similar ones are present on the exo- and endopodites. As in Minea, Pearse, the telson is triangular and does not extend to the ° end of the metasomatic segments. EXPLANATION OF PLATE 6. Paracubaris spinosus, gen. et sp. nov. Fig. 1. Dorsal view of the cephalon. x 8. . Anterior view of the cephalon. x 8. . Dorsal view of the right antennule. x 70. Dorsal view of the right antenna. x 14. . Terminal style of the antenna. x 110. . Terminal portions of the inner and outer lobes of the right 1st maxilla. x 70. . Terminal portion of the left maxillipede. x 48. . Ventral view of the outer margin of the 2nd mesosomatic sezment. x 6. . Terminal margin of the same. ™ 8. . Dorsal view of the right uropod. x 14, . Telson and last metasomatic segment. x 6, SCO ONAanrwWDH Se H The author desires to thank the Carnegie Trust for the Universities of Scotland for a grant to defray artist’s charges. a BEA 0 WiBac the Hye 7 re 1 Wig ay i) iat r ON THE ORAS APPENDAGES OF MARINE ISOPODA. 65 On the Oral Appendages of certain Species of Marine Isopoda. By WALTER E. Continer, D.Se., F.L.8., Carnegie Fellow, and Research Fellow of the University of St. Andrews, the Gatty Marine Laboratory, St. Andrews. (PLATES 7-9.) [Read 3rd May, 1917.] CONTENTS. Syaivs, Richardson. . Symmius caudatus, Richardson. CurripoTEa, Harger. . Chiridotea cecq (Say). . Chiridotea tufts’ (Stimpson). Mesrporea, Richardson. . Mesidotea sabini (Kroyer). PrntipotEa, Richardson. . Pentidotea resecata (Stimpson). . Pentidotea wosnesenskit (Brandt). . Pentidotea whitet (Stimpson). GLYPTIDOTEA, Stebbing. 8. Glyptidotea lichtensteini (Krauss), Iporra, Fabricius. 9. Idotea rectilinea, Lockington, 10. Idotea phosphorea, Harger. ParrpoTeTa, Stebbing. 11. Paridotea ungulata (Pallas). , Yar, noy. atrovirens, 12. Paridotea reticulata, Barnard. 13. Paridotea rubra, Barnard. 14. Paridotea fucicola, Barnard. [vImoTeEa, gen. noy. 15. Euidotea peronti (Milne-Edwards). [eRICHSONELLA, Benedict. 16. Lrichsonella attenuata (Harger). SynipoTea, Harger. 17. Synidotea hirtipes, Milne-Edwards. 18. Synidotea palhda, Benedict. 19. Synidotea nebulosa, Benedict, 20. Synidotea angulata, Benedict. 21. Synidotea marmorata (Packard). . Synidotea bicuspida (Owen). 23, Synidotea nodulosa (Xvoyer). 24. Synidotea levis, Benedict. WATS TE HLO GARY eteetctenctarctssetcteka, sHe!-: slo) efaueletepenatatnativeyetrrehestaterere ertekeiebas oieie 90 co bh rl i NI Or LINN. JOURN.—ZOOLOGY, VOL. XXXIV. 6 G6 DR. W. E. COLLINGE ON THE T. INTRODUCTION. Tnere are few groups of animals that present greater difficulties in connection with their classification than the Order Isopoda. Much of the work of the earlier carcinologists has, in the light of recent research, proved to be very imperfect, and many of the genera and species classified as related to one another have been shown to be widely separated. Further, structural characters that hold good for certain divisions of the Tsopoda are quite unsatisfactory for others. his is particularly well illus- trated in the oral appendages. I have elsewhere (7) expressed the view, and shown, that in the suborder Oniscoidea, Sars, these appendages “are liable to a large amount of variation in individual species, and are therefore characters of only minor importance as compared with the form of the head, antenne, telson, uropoda, mesosomatic segments, and thoracic appendages,” and I believe that further investigations will serve to corroborate and strengthen this view so far as the strictly terrestrial forms are concerned, where considerable modification and degeneration have taken place. On the other hand, in certain suborders of marine Isopoda these same appendages would seem to be fairly constant in form, and to offer excellent data for the characterization of both genera and species (8-12). Recent work on the suborder known as Valvifera shows that the form of at least two of the four oral appendages affords most valuable aid in the discrimination of both genera and species, viz., the first maxillee and the maxillipedes. In this connection it is interesting to note, that in the members of the family Idoteide they differ very little in the immature stages from that obtaining in the adult ; they are thus frequently of great service for purposes of identification. A further interesting point relative to the two pairs of maxillee, and in a lesser degree the maxillipedes also, is the frequency with which the “casts” of these appendages remain attached to the newly formed ones, and they often serve a useful purpose in aiding in the elucidation of minute structural details, as, being free from all muscular attachments and almost void of pigment, they can easily and quickly be made transparent. It is therefore much to be regretted that many authors have given no description or figures of the form and structure of these appendages in quite a large number of species. In the present communication I have endeavoured to remedy this to some extent by describing and figuring the first maxilla (occasionally the second) and the maxillipede in 24 species referable to 9 genera of the family Tdoteidee, in none of which, so far as I am aware, have they both been described, and also to correct some errors in a few of the figures and descriptions where the maxillipede only has been partly described and incorrectly figured. ORAL APPENDAGES OF MARINE ISOPODA, 67 I have much pleasure in here expressing my thanks to the authorities of the U.S. National Museum, Washington, for specimens of many of the species examined ; to Professor D’Arey W. Thompson for examples of Mesidotea sabini (Kroyer); to Professor Chas. Chilton for various New Zealand species; to Mr. Keppel H. Barnard for many South African species ; and to Mr. Walter H, Baker for South Australian species. Under the different species the more important references have been given, but no attempt has been made to give the complete list of synonyms or references. II. Merruop or Preparation. Most of the errors in previously published figures and descriptions of the oral appendages of the Idoteidze are due to the fact that the different segments, Joints, spines, etc. are not always shown, and this has no doubt partly arisen owing to the appendages having been imperfectly prepared for microscopical examination. The old method of boiling or even soaking in a solution of caustic soda is unsatisfactory, as frequently the different parts become separated from one another and also become altered in form to a greater or lesser degree. After considerable experimentation, I think I have at last arrived at a method that is thoroughly satisfactory. So far as I am aware, the details have not previously been published ; it therefore seems desirable to describe the method in detail. The original idea of treating chitin as described below is Professor J. C. Trvine’s, who adopted it in a research on the chemical nature of chitin derived from different sources *. He has very kindly given me the essential details of the process, and with these I have made numerous experiments and modifications, ultimately evolving the following process for clearing and staining small chitinous objects, such as the appendages and other parts of the exoskeleton of crustacea. All the material dealt with had previously been preserved in alcohol. Upon removing the appendage or particular part to be studied, it was placed in a small quantity of a 15 °/, solution of HCl, and there allowed to remain for a period varying from twenty minutes to three hours, according to the size and thickness of the object. The receptacles used for containing the specimens and fluids were small flat glass dishes 39x39x9 mm. with a circular concavity 5 mm. in depth, and covered with a square of glass. Upon removing the specimen from the HCl it was well washed in distilled water and then transferred to a 5 °/, solution of caustic soda, in which it was allowed to remain for one to three hours, and then placed in a 4 °/, solution of sodium permanganate for a period varying from thirty minutes to two * Trans. Chem. Soc. Lond. vol. xey. (1909) pp. 564-570. 6" 68 DR. W. E. COLLINGE ON THE hours. Finally, it was again placed in a 15 °/, solution of HC] for ten to sixty minutes. After further careful washing the specimen was dehydrated and then stained with Meyer’s alcoholic carmine. Most of the material was examined in xylol, some was mounted in Canada balsam (xylol) and some in Farrant’s solution, but none of the specimens were improved by the use of either of these media—the former proved the better of the two. T should strongly recommend anyone trying this method to first experiment with some material of no value, as it is only by experience that one comes to appreciate just the right time to remove the specimens from one solution to another. If removed too soon little change has taken place ; on the other hand, a very short excess period is sufficient to ruin the object. If the above-mentioned small flat glass dishes are used the specimens can be exumined from time to time under a low-power of the microscope and the progress and changes noted. The value of this method of treating small, jointed chitinous objects lies in the fact that, if carefully handled, little or no displacement of the different parts takes place ; further, the objects are rendered perfectly transparent, all cellular tissue and pigment being removed, and a uniform stain is taken up by the margins of all joints, segments, spines, ete. Setee which appeared to arise as stiff spines from the surface of a body can, by this method of preparation, be seen to pass beneath the snrface at their proximal ends and to be lodged in tiny pits or articular cups, whilst others that appeared to be perfectly smooth are found to carry smaller spines, teeth, ete. It is with considerable pleasure that I here acknowledge my indebtedness to Professor Irvine for the basis of this method, and for the, at all times, kindly criticism and advice he has given. IL. Toe Orat APPENDAGES OF THE Faminy IDOTEIDA. The two principal appendages that have been employed in the classification of the members of the suborder Valvifera are the first maxille and the maxillipedes, and I propose to offer some few remarks upon these as presented in the family Idoteidee. The First Maxille.—Kach maxilla consists of a pair of elongated chitinous bodies—the inner and outer lobes—connected by a small basal segment. The outer lobe is slightly convex on the ventral side and almost flat dorsally, usually the anterior half, or even two-thirds, is wider. than the posterior portion. Setules, small plain spines, set in cup-shaped articular cavities may or may not be present on the anterior part of the ventral surface. Distally there are a variable number of curved spines. In many species these are divided into two distinct sets, viz., an outer series of stout, curved, bluntly ORAL APPENDAGES OF MARINE ISOPODA. 69 pointed ones, and an inner series of thinner, more slender, usually sharply pointed ones, which have one or more rows of teeth on their sides, which vary considerably in number and size. The two series are frequently separated by a long, slender, sharply pointed spine. There are often a series of setee on the anterior outer (and sometimes inner) border. The inner lobe shows considerable variation in the different genera. It consists of an elongated, narrow, posterior portion, which widens anteriorly into an oval or more or less triangular, flattened plate, with two to four setaceous spines at the distal extremity, in addition to which, one or more setules may be present. In a few cases the number of setaceous spines may be greater on the appendage of one side than on that of the other. The Mawillipedes—Each maxillipede consists of a divided coxopodite, a basipodite with a short inner lobe on the inner side, and an outer lobe or palp (endopodite) articulating with the basal segment on the outer side. On the outer side posteriorly there is a wide, more or less oval, plate, the epipodite. The divisions of the coxopodite lie respectively posterior to the epipodite and basipodite. The inner lobe of the basipodite is generally fringed distally with a number of setose and plain spines, and on its inner side and about its middle is a curved process which serves to hook together the appendages of the two sides. Occasionally there are two or more of these processes. The outer lobe or palp is composed of a varying number of joints, but the first or “collar ”’-joint, a short one immediately anterior to the basipodite, is, I believe, present in most genera. It would seem to be absent according to Miss Richardson (21) in the genus Chiriscus, for no such joint is shown in the figure of the maxillipede of Chzriscus australis, Richardson. In Edotia, Guérin-Mén., Syndotea, Harger, Chiridotea, Harger, and Macrochiridothea, Ohlin, there are three joints in the palp of the maxillipede ; in /dotea, Fabricius, Erichsonella, Benedict, Colidotea, Richardson, Eusym- merus, Richardson, Synsoma, Collinge, and Huidotea, Collinge, four joinis are present ; whilst in Crabyzos, Spence Bate, Mesidotea, Richardson, Pentidotea, Richardson, Pentias, Richardson, Zenobiana, Stebbing, Glyptidotea, Stebbing, Paridotea, Stebbing, Cleantiella, Richardson, and Lngidotea, Barnard, there are five divisions in the palp. It is very doubtful whether the structure of these parts indicates in any way the phylogeny of these crustacea. In any attempt therefore to estimate the value of these appendages for purposes of classification due consideration should be paid to the segmentation of the metasome, and also to any evidences of degeneration or special modification, It may be interesting to roughly classify the known genera according to the segmentation of the metasome and the number of joints of the palp of the maxillipede. The primitive Idoteidze, in all probability, possessed a metasome composed 70 DR. W. E. COLLINGE ON THE of five distinct segments and a palp on the maxillipedes having five joints. Such a type is probably represented by Protdotea, Racovitza and Sev., although the maxillipedes have not been seen of this extinct genus. destdotea, Richardson, would follow, and then Chiridotea, Harger, where a segment of the metasome has become fused, but a suture remains indicating it. Although undoubtedly related to the former genus, the palp of the maxillipede is composed of only three joints. The position of Macrochiridothea, Ohlin, and Chiriscus, Richardson, is somewhat doubtful. Ina second group we have Pentidotea, Zenobiana, Glyptidotea, Paridotea, Pentias, Crabyzos, Cleantiella, and Fngidotea, in all of which genera the palp of the maxillipede possesses five joints. In Zenoliana there is often a metasome composed of five segments, and in all the remaining genera there are indications of four segments. In a third group must be placed Idotea, Colidotea, Husymmerus, Mrich- sonella, Synisoma, and Huidotea, all with a four-jointed palp of the maxilli- pede and one to four metasomatic segments. Finally, in a fourth group we have dotia and Synidotea, each with a three-jointed palp of the maxillipede and a single metasomatic segment and one suture. The Table on p. 71 summarizes these and other structural features. Whilst there is fairly conclusive evidence to show that the genera of the first group are closely related to one another (cf. 19 & 10), I do not think that it is possible, in the present state of our knowledge, to state that a similar relationship is evident between the genera of the remaining groups. As I have elsewhere pointed out (12a), the classification of this interesting family of Isopoda is as yet very uncertain. ‘The known genera and species in all probability constitute only a very small part of those actually existing. The distribution of the members of this family, which contains some two dozen genera, including about one hundred and twenty-five species, as at present known, is curious. Miers (16) and the earlier writers all regarded them as having a wide distribution in all parts of the world. but being more abundant in temperate and colder seas than in the tropics. At that time the majority of species had been obtained in Northern latitudes, but of recent years many new species have been obtained from the Mid and Southern Pacific, Indian Ocean, and Antarctic. ‘Thus, Ohlin (18) has described two new species of Macrochiridothea and two of Edotia, and a new species of Evichsonella from South America; Chilton (5), the curious Jdotea festiva from New Zealand; Miss Richardson *, a new species of Glyptonotus from the Antarctic; Barnard (1, 2), three new species of Paridotea and one of Synidotea from South Africa. More recently t I have described a new * Bull. Mus. Paris, 1906, p. 187. + Ree. Indian Mus. vol. xiii. (1917) pt.1, p. 1, pl. i. ORAL APPENDAGES OF MARINE ISOPODA. 71 species of Synidotea from the Indian Ocean, and I have two new species of Crabyzos from South Australia, in addition to a number of other forms from tropical localities not yet worked out. There is no reason to suppose that the South Pacific and Antarctic Regions are any poorer in genera and species than the North Pacifie and Arctic Regions, although but few have yet been obtained from the former regions. A careful study, now extending over some years, of the existing forms leads me to the conclusion that future investigations will undoubtedly tend to link together more closely the various genera of this family and the different families of the Valvifera. As at present known the members of this suborder may conveniently be placed in two groups, viz. :— i, ASTACILLINEA, nov. Containing the families Astacillidee, Stebbing, Chectiliidee, Dana, and Amesopodidee, Stebbing. il. IDOTEINEA, noy. Containing the families Idoteide, Fabricius, Pseudidotheide, Ohlin, and Holognathide, G. M. Thomson. 1st Maxilla. | Maxillipede. Metasome. Number of | Number of Number of Spines on | Joints in | Inner Lobe. | Palp. Segments. | Sutures. Glyptonotus, Wights ............ 6 5 5 Symmius, Richardson ............ ‘ 3 (?) 3 Chipadoted, lax ceri se cies ceia 8 | 4 1 Macrochiridothea, Ohlin.......... pls 3 | | | Mesidotea, Richardson............ 3 5 3) Proidotea, Rac. & Sev. .......5.. ? ? 5 | | Chiriscus, Richardson ............ | 3 (?) 3 | | Zenobiana, Stebbing ............ 5 5 5-5 | Pentidotea, Richardson .......... 3 5) | 3 1 Engidotea, Barnard .............. 2 5 2 2 | Cleantiella, Richardson .......... 5 2 Paridotea, Stebbing ............ 3 or 4 5 | a 3 Glyptidotea, Stebbing ............ 3 | 5 | 1 3 Pentias+ Richardsoni sic: sent ee 3 | 5) 1 3 | Crabyzos, Spence Bate .......... 3 | 5 1 D# Uidote dey sbviclusiss erasiietels sel: 3 4 3 1 Burdoted, Collinge: os... oi ces 3 | 4 1 3 Colidotea, Richardson ............ 4 1 l Eusymmerus, Richardson ........ | Raia | 4 | iL 1 Erichsonella, Benedict ............ 31, 47, | 4 1 Synisoma, Collinge .............. Brent 3 | 4 J Edotia, Guérin-Mén. ............ wal 3 l 1 Srathiae, WINGS. Saono sono ongue 2 3 1 * The second suture is very small in some species, scarcely visible on the dorsal side. Ue DR. W. E. OOLLINGE ON THE IV. Synopric TABLE OF THE GENERA DEALT WITH. A. Coxal plates distinct on the 3 posterior segments of the mesosome. a. Lateral margins of the cephalon entire. Hyes dorsally situated. 6. Flagellum of antenne single-jointed. c. Metasome composed of 3 segments. d. Maxillipedes with a 3-jointed palp .............. B. Coxal plates distinctly separated on 2nd to 7th segments of the mesosome, a. Lateral margins of the cephalon cleft. Eyes dorsally situated. b. Flagellum of antenne multiarticulate. c. Metasome composed of 4 segments and 1 pair of sutures. d. Maxillipedes with a 3-jointed palp ...........-.. c’, Metasome composed of 5 segments. d', Maxillipedes with a 5-jointed palp .............. a’, Lateral margins of the cephalon entire. yes situated dorso-laterally. c''. Metasome composed of 3 segments and 1 pair of sutures .......... efoseuevedeeieleyeisyelayeeneievetel elastance ayer: c''. Metasome composed of a single segment and 3 pairs OUGMONE, gocesaccoaucosd HODES AGOOGoopoDIogad oxo c’. Metasome composed of 3 segments and 1 pair of sutures, d', Maxillipedes with a 4-jointed palp.............. c', Metasome composed of a single segment and 38 pairs ORSULUTEST c/a v- oseichertelepatre beteneletercn ee Rees en Merees 6’, Flagellum of antennz single-jointed. ce". Metasome composed of a single segment .......... C. Coxal plates not distinct on any mesosomatic segments. a. Lateral margins of the cephalon entire. Eyes situated dorso-laterally. 6, Flagellum of antennie rudimentary. c. Metasome composed of a single segment and 1 pair of sutures, d. Maxillipedes with a 3-jointed palp .............. bo’, Flagellum of antenne multiarticulate. c', Metasome composed of a single segmeut and J pair of SUCUTOS! Totes iora/ejoiis'tn/ oie. sso fetejel fet RRM MY Mamet r aaa Symmius, Richardson. Chiridotea, Uarger. Mesidotea, Richardson. Pentidotea, Richardson. Glyptidotea, Stebbing. Paridotea, Stebbing. Idotea, Vabricius. DLuidotea, Collinge. Evichsonella, Benedict. Edotia, Guérin-Mén. Synidotea, Harger. ORAL APPENDAGES OF MARINE ISOPODA. 73 V. List or GrenERA AND Srectes EXAMINED Symutius, Richardson. 1. Symmuius caupatrus, Richardson. Symmius caudatus, Richardson, Proc. U.S. Nat. Mus. vol. xxvii. (1904) p. 39, fies. 11-15. This interesting species was described by Richardson in 1904, from speci- mens obtained from Japan. The only reference to the oral appendages is in the generic diagnosis, where it is stated “* Maxillipedes with a three-jointed palp.” T have previously stated (10) that I question the accuracy of the figures given of this appendage (op. cit. p. 41, figs. 13a & b). In these two figures Miss Richardson shows a 3-jointed palp ; the basipodite, epipodite, and inner distal lobe united ; anda very curious form of coxopodite, in all of which characters it is totally unlike the condition obtaining in any other genus of the family. In the single specimen of this species in my collection, the oral appendages had been removed before it came into my possession. I am unable, therefore, to give a description or figures of either the first maxilla or the maxillipede. Carriporna, Larger. 2, CHIRIDOTEA caca (Say). (PI. 7%. fig. 1.) Idotea ceca, Say, Journ. Acad. Nat. Sci. Philad. vol. i. (1818) p. 424; Milne-Edwards, Hist. Nat. des Crust. vol. iii. (1840) p. 131; Harger & Vervill, Rept. U.S. Comms. Fish & Fisheries, 1878, pt. 1. p. 569, pl. 5. fig. 22. Chiridotea ceca, Harger, Amer. Journ. Sci. vol. xy. (1878) p. 874; Rept. U.S. Comms. I. & F. 1880, pt. vr. p. 388, pl. 4. figs. 16-19. Glyptonotus cecus, Miers, Journ. Linn, Soe., Zool. vol. xvi. (1881) p. 17. Chiridotea ceca, Richardson, Bull. No. 54, U.S. Nat. Mus. 1905, p. 553, figs. 380, 381. The Mawillipede (Pl. 7. fig. 1) —Both Harger and Richardson have figured the maxillipede of this species, but neither describes it. In both instances the figures are incorrect. This appendage in the genus Chiridotea is characterized by the short wide epipodite, small basipodite, and a three-jointed palp, the second joint of which is unusually long. In this species the whole appendage is extremely short. The coxopodite has the usual two joints of which the outermost is the larger. The basipodite is very small, having a straight inner margin and the outer one slightly expanded. Both the anterior and the posterior margin slope outwardly, but not to the extent shown in Harger’s figure. The three-jointed palp is rather 74 DR. W. E. COLLINGE ON THE more than two and a haif times the length of the basipodite: the first joint is small, the second very long, and the third smaller and more or less oval. Sete spinous. The inner distal lobe extends forward as far as the middle of the second joint of the palp and has the usual setose and plain spines terminally. ‘The epipodite is somewhat cone-shaped, very short, wider than the basipodite, and has its posterior margin excavate. 3. CHIRIDOTEA TUFTS (Stimpson). (PI. 7. fig. 2.) Idotea tuftsii, Stimpson, Smithscnian Contrib. to Knowledge, vol. vi. (1853) p. 39; Harger & Verrill, Rept. U.S. Comms. I’. & F. 1875, pt. 1. pp. 340 & 569. Chiridoteu tuftsti, Harger, Amer. Journ. Sci. vol. xv. (1878) p. 374; Rept. U.S. Comms, F. & F. 1880, pt. vi. p. 340, pl. 4. figs. 20-23. Glyptonotus tufts’, Miers, Journ. Linn. Soc., Zool. vol. xvi. (1881) p. 18. Chiridotea tuftsi’, Richardson, Bull. No. 54, U.S. Nat. Mus, 1905, p. 354, figs. 382, 383. The Mawillipede (Pl.'%. fig. 2).—The single specimen J have examined does not at all agree with either Harger’s or Richardson’s figures. I have no reason to doubt the identification of the species which was obtained at Cape Cod Bay, U.S.A., in 1879. A reference to the figure shows that in this specimen, whilst preserving most of the characters typical of Chiridotea, there are four joints in the palp. Mesiporna, Richardson. 4. Mestporea saBini (Avdyer). (PL. 7. figs. 3-5.) Idotea sabint, Wriyer, Nat. Tidssky. vol. ii. (1846-49) (s. 2) p. 401; G. O. Sars, Arch. f. Math. og Naturvidensk. vol. ii. (1877) p. 350. Glyptonotus subint, Miers, Jown. Linn. Soc., Zool. vol. xvi. (1881) p. 15, pl. 1. figs. 8-5; Richardson, Proc. U.S. Nat. Mus. vol. xxi. (1899) p. 844. Chiridotea sabini, Stebbing, Ann, & Mag. Nat. Ilist. (s. 7) vol. iv. (1899) p: 263. Mesidotea sabini, Richardson, Bull. No. 54, U.S. Nat. Mus, 1905, p. 350, figs. 877-379. Kroyer figures both of the maxillee and the maxillipedes, the former, how- ever, bear little or no resemblance to the actual form. In the first maxilla he shows an outer lobe with eight spines and an inner one with four, whilst in the second maxilla the two outer lobes are shown with four spines on each and a series of bluntly ending ones on the inner lobe. The maxillipede in his lower figure is fairly correct, but wrong in the upper one. The First Mawilla (V1.7. fig. 3)—The outer lobe terminates in eleven spines, most of which are stout, and there are numerous sete on the outer margin of the lobe, which extend downwards for some distance. The inner lobe terminates in two long setose spines and a small curved setule on the outer ventral border. The Second Mavilla (P1.'%. fig. 4}.—The two outer lobes terminate in six = On ORAL APPENDAGES OF MARINE ISOPODA. and seven long spines respectively, most of these are serrated ; the inner- most lobe has sixteen to twenty setose spines, varying in size, the last one on the inner margin being the longest. The Mawillipede (Pi.'¢. fig. 5)—In this species this appendage is com- paratively small, but of a very robust type and having its margins densely fringed with setee and spines. The inner division of the coxopodite is small, almost cuboid. The basipodite is also very small, measuring only about half the length of the palp. Its anterior and posterior margins are obliquely cut away, sloping towards the outer margin; the inner margin is fringed with fine, closely-set setee. he five-jointed palp has the first joint rather larger than in most species, the second joint is comparatively small, the third the largest and with a lobe-like extension on the inner side, the fourth joint is longer than the third, but much narrower, whilst terminally there is a small fifth joint. All the joints are fringed with sete, which are extra long on the fifth joint. The inner distal lobe extends to the anterior border of the second joint of the palp and is fringed with setose spines terminally. ‘The epipodite is almost circular and extends as far as the middle of the second joint of the palp. It is fringed with short, closely-set setee. Penriporea, Richardson. 5. PENYIDOTEA RESECATA (Stimpson). (PI. 7%. figs. 6, 7.) : ] y t 5 3 they Idotea resecata, Stimpson, Bost. Journ, Nat. Hist. vol. vi. (1857) p. 504, pl. 22. fi Miers, Journ. Linn. Soc., Zool. vol. xvi. (1881) p. 45. Pentidotea resecata, Richardson, Bull. No, 54, U.S, Nat, Mus. 1905, p. 369, figs. 400, 401. las The First Maailla (P\. 7. fig. 6)—This appendage in this species has not previously been figured or described. ‘lhe outer lobe terminates in twelve spines, the innermost five of which are denticulate; there is a long fine spine about the middle and then six stout curved spines on the outer side. The inner lobe terminates in three long setose spines and a setule on the outer anterior margin. The Mazillipede (Pl. 7%. fig. 7)—This appendage is greatly elongated in this genus and characterized by the large epipodite. Miss Richardson has given a figure of the mavxillipede, but it is inaccurate in detail. The basi- podite is narrow and greatly elongated and las straight inner, outer, and posterior margins. It is longer than the palp, which is five-jointed, the terminal joint being quite small. The inner distal lobe is narrow and the spines on its terminal margin setose. The epipodite is a large, elon- gated, wide plate extending forward as far as the middle of the third joint of the palp. It is slightly narrower anteriorly than posteriorly, but wider throughout than the basipodite. 76 DR. W. E. COLLINGE ON THE 6. PENTIDOTEA WOSNESENSKII (Brandt). (PI. 7. figs. 8, 9.) Idotea wosnesenskit, Brandt, in Middendorff’s Sibirische Reise, vol. ii. (1851) Crust. p- 146. Idotea hirtipes, Dana, U.S. Expl. Exped. vol. xiv. (11.) (1853) p. 704, pl. 46. fig. 6. Idotea oregonensis, Dana, Proc. Acad. Nat. Sci. Philad. vol. vil. (1854) p. 175. Idotea media, Dana, ibid. (1857) p. 175. Idotea wosnesenskit, Miers, Journ. Linn. Soc., Zool. vol. xvi. (1881) p. 40; Richardson, Proc. U.S. Nat. Mus. vol. xxi. (1899) p. 846. Pentidotea wosnesenskiit, Richardson, Bull. No, 54, U.S. Nat. Mus, 1905, p. 370, figs. 402-404. The First Mazilla (Pl. 7. fig. 8).—The outer lobe terminates in twelve stout curved spines, the outermost five being rather stronger than the remaining ones; none of them are denticulate. The inner lobe narrows considerably at its distal end and bears three slender setose spines. The Mawillipede (Pl. 7. fig. 9)—Miss Richardson has given a figure of the maxillipede, but I find many points of difference. The basipodite is elongated and somewhat shield-shaped, being produced anteriorly at each side of the first joint of the palp and narrowed posteriorly on both the inner and outer imargins. It is rather shorter than the length of the palp. All the five joints of the palp are large, the first and fifth being the smallest, the second and third are produced at their anterior margins, the latter more so than the former. The inner distal lobe is wide and surmounted by numerous setose and plain spines. The epipodite is a large wide plate extending forward as far as the commencement of the third joint of the palp. It narrows a little toward the anterior or apical end and is wider than the basipodite. 7. PENTIDOTEA WHITEI (Stimpson). (PI. 7. figs. 10, 11.) Idotea white’, Stimpson, Proc. Acad. Nat. Sci. Philad. (1864) p. 155; Miers, Journ. Linn. Soc., Zool. vol. xvi. (1881) p. 42. Pentidotea whitei, Richardson, Bull. No. 54, U.S. Nat. Mus. 1905, p. 373, figs. 405, 406. The First Mawilla (PI. 7. fig. 10).—The outer lobe has twelve terminal spines, the outermost five being stout and curved, the sixth is straighter and more slender, and all the remaining ones, excepting the innermost, have each four or five blunt tooth-like processes. On the ventral surface there is a large setule set in a cup-shaped articular cavity. There are a few sete on the outer margin. ‘The inner lobe more closely resembles that of P. resecata than P. wosnesenskti, being wide at its distal end. It has three setose spines, and a short strong setule on its inner anterior margin, The Mawillipede (P1. 7. fig. 11).—This appendage resembles somewhat that described in P. wosnesenskii, though larger and generally more robust. Miss Richardson’s figure does not show the divisions of the coxopodites or ORAL APPENDAGES OF MARINE ISOPODA. We those between the basipodite and the inner distal lobe, and the shape of the former is scarcely correct. The basipodite is rather shorter and wider than in P. wosnesenskii and as wide as the greatest transverse dimension of the epipodite. The five-jointed palp is considerably longer than the basipodite and ail the joints wider ani longer than in the preceding species, the fifth joint is comparatively much larger than in either of the two species of this genus described above. The inner distal lobe is slightly narrower than in P. wosnesenshii, but otherwise very similar. The epipodite is a large flat, elongated plate, narrowing some- what towards its apical margin, and extending forward to almost the middle of the third joint of the palp. GuiypripotEa, Stebbing. 8. GLYPTIDOTEA LICHTENSTEINI (Arauss). (PI. 7. figs. 12, 13.) Idotea lichtenstetn’, Krauss, Siidafrik. Crust. 1843, p, 62, pL iv. fig. 4; Miers, Journ. Linn. Soe., Zool. vol. xvi. (1881) p. 64. Glyptidotea lichtensteint, Stebhing, Sth. Afr. Crust. pt. m. 1902, p. 57, pl. 10. Stebbing was the first to describe and fivure the oral appendages, he states : « First maxillsee.—The outer plate is surmounted hy eleven curved spines, the outermost but one being the strongest, the innermost six slender, forming two sets, each consisting of three graduated spines. The inner plate has three plumose setz on the narrow apex.... Maxillipedes.—The inner margin of the first joint forms a rounded process beset with plumose sete ; its external part forms a broad base for the large distally narrowed epipod. The second joint is elongate, its apical process, distally fringed with sete, reaches beyond the second joint of the palp and somewhat above its base has a strong spine-hook, nearly at the level reached by the apex of the epipod. The first joint of the palp is small, the second widened cup-like, with the inner margin much longer than the outer, the third joint similar but larger, and with less difference between the two margins ; the fourth joint is much the longest, oval, but with truncate apex, on which is placed the small, but very distinet, oval fifth joint, this like the three preceding joints having sete on the inner margin.” With many other writers Mr. Stebbing does not distinguish between the two divisions of the coxopodite, and the inner distal lobe, which is distinct from the basipodite. The First Mawilla (PI. 7. fig. 12)—In all the specimens T have examined I find twelve terminal spines on the outer lobe and a large setule, set in a cup-shaped cavity, on the ventral side ; further, on the inner lobe, in addition to the three setose spines, there are two setules, DR. W. Ek. COLLINGE ON THE The Mavillipede (Pl. 7. fig. 13).—This appendage in this species shows a strong resemblance to the condition obtaining in the genus Pentidotea, only it is less robust and considerably narrower. The divisions of the coxopodite are large and slightly overlap one another. The basipodite is narrow and elongated, with a straight inner margin and narrowest about its middle. The posterior margin slopes upwards and outwards. The five-jointed palp presents all the characteristic features seen in Pentidotea, and is half again as long as the basipodite. The inner distal lobe reaches almost to the end of the extended inner margin of the third lobe of the palp. There are a number of setose and plain spines distally of a rather more robust type than in Penti- dotea. The epipodite proximally is wider than the basipodite, but its distal third narrows considerably, becoming bluntly pointed at its apical margin. It extends forward as far as the anterior outer margin of the second joint of the palp. It is somewhat doubtful whether or not this species is entitled to separate generic rank or whether it is not synonymous with Paridotea, Stebbing. In both genera the palp of the maxillipedes consists of five joints, and the meta- some is composed of a single segment and three pairs of sutures. The most striking difference between the two genera is seen, perhaps, in the form of the cephalon. Iporna, Fabricius. 9, IDOTEA RECTILINEA, Lockington. (Pl. 7. figs. 14, 15.) Tdotea vectilinea, Lockington, Proc. Cal. Acad. Sci. vol. vii. (1877) p. 36; Richardson, Proc, U.S. Nat. Mus. vol. xxii. (1900) p. 181, fig. 5¢; Bull. No, 54, U.S. Nat. Mus. 1905, p. 360, figs. 3889-391. The First Mawilla (PI. 4. fig. 14).—This appendage has not previously been described or figured in this species. The outer lobe terminates in twelve strong curved spines, of which three or four at the inner side are denticulate. The inner lobe has three setose spines and a small curved setule on the ventral outer margin ; there are numerous long fine sete on the inner margin. The Mazillipede (Pl. 7%. fig. 15).—Miss Richardson gives a figure of this appendage in which the coxopodite is shown as a single piece and the generai shape is scarcely correct. The coxopodite consists of two large stout nodules, one lying at the base of the epipodite and the other at the base of the basipodite. In all of the specimens examined these had the large cuboid form shown in fig. M5. The anterior margin of the basipodite is produced slightly on each side of the first joint of the palp. Tts inner, outer, and posterior margins are almost straight. The palp is composed of four joints, of which the third is rather smaller than usual. The inner distal lobe has an almost straight anterior ORAL APPENDAGES OF MARINE ISOPODA. 79 margin, it slopes outwardly, becoming wide at the base. There are a number of setose spines on the anterior margin. The epipodite is slightly narrower than the basipodite. It extends forward slightly beyond the anterior outer margin of the second joint of the palp and becomes a little narrower towards its apical margin. 10. Iporea PHospHoREA, Harger. (PI. 7. figs. 16-18; Pl. 8. figs. 19-23.) Idotea phosphorea, Harger with Verrill, Rept. U.S. Comms. F. & F. 1873, pt. 1. p. 569; Harger, ibid. 1880, pt. vr. p. 347, pl. 5. figs. 27-29, Idotea marina, yar. phosphorea, Miers, Journ, Linn. Soe., Zool. vol. xvi. (1881) p. 31. Idothea phosphorea, Richardson, Bull. U.S. Nat. Mus. 1904, p. 367, figs. 398, 399, Considerable difference of opinion exists amongst carcinologists as to the specific identity of this species. It was described by Harger in 1873, and in 1880 he gave figures of the animal, the antenna, the maxillipede, the first two walking limbs, the second metasomatic appendage, and the uropod. The original description is as follows :— “The head is narrowed behind. The eyes are of moderate size. The flagellum of the antennze is shorter than the peduncle, and consists of about ten to fourteen segments. The maxillipeds have the external lamella broader than in the preceding species [/. baltic], with its inner margin straight and its outer margin curving pretty regularly to a slightly attenuated tip. “The epimera of the second, third, and fourth pairs are rounded bebind, and those of the last three pairs are less acute than in J. baltica*. * Pleon ovate, a little constricted near the middle and pointed, its three proximal segments rather less acute than in the preceding species. The basal plate of the operculum tapers towards the end, and the terminal plate is triangular, a little longer than broad. The stylet on the second pair of pleopods in the male is slender, nearly straight, surpasses the lamella to which it is attached, and is obliquely truncate. “ Tenoth 25 mm. ; breadth 7 mm.” Miers regarded this species as a variety of J. balt/ca (Pallas), and remarks : “There appears to be no sufficient reason to distinguish /dotea granulosa of Rathke .... from the American J. phosphorea.” Whilst /. phosphorea is no doubt closely related to granulosa, Rathke, which is now regarded asa valid species, [ am of opinion that there are sufficient well-defined structural characters which clearly separate it from this latter species. Of these, the most important are, the form of the cephalon, the coxal plates of the mesosome, and the form of the metasome. In addition, the antennules, the antenne, the first maxillee, and the maxilli- pedes also exhibit differences. Finally, in all the specimens of /, granulosa that I have examined the dorsal surface of the body never shows tubercles ; * Termed J, irrorata by Harger. 80 DR. W. E. COLLINGE ON THE it is either finely granulated or almost smooth, whereas in all the specimens of I. phosphorea that I have seen there are large lateral and median tubercles. Harger states, “the body, especially of the young, is rough and tubercular along the median line and often, also, laterally. Older specimens are much smoother, losing their large median tubercles, but never becoming as smooth asin” J, baltica. I have seen no examples measuring more than 22 mm. in length, and in all of these both the lateral and median tubercles are quite prominent. A comparison of Surs’s figures (22, pl. 34. fig. 1) or those given by myself (12a, pl. 5. figs. 48-58) of J. granulosa, with those given by Harger (13, pl. 5. figs. 27-29) of I. phosphorea, at once show the difference in the form of the terminal segment of the metasome, the coxal plates of the mesosome, Kc. In the form of the cephalon (PI. 8. fig. 16) the differences from J. granu- losa are very marked, and are at once apparent in the width and the deep transverse groove, anterior to the posterior margin. The eyes are larger and situated more anteriorly, whilst the lateral margins posterior to the eyes curve inwards. The antennule (PI. 8. fig. 17) is shorter and more robust in this species than in 7. granulosa, as also the antenna (fig. 18). In J. granulosa I have described (12 a) the coxal plates of the mesosome as occupying the anterior two-thirds of the lateral margin of the second seg- ment, the third rather more, and the remainder the whole of the lateral margins, increasing in breadth from the fourth to the seventh segments. Sars (22) speaks of them as being “ comparatively small.” In J. phosphorea, whilst occupying approximately the same proportion of sach segment, they are slightly wider, the external margin being more expanded, so that those of the second, third, and fourth segments are roughly triangular, the apex of the triangle being rounded. Those on the fifth, sixth, and seventh segments have a sloping margin from the anterior to the posterior angle, whereas in /. granulosa the margins are almost truncate. Harger (13) remarks that young specimens cetemmle the young of 7. irro- rata (= baltica), but that they can be distinguished by the coxal plates of the second and third mesosomatic seements, eich do not occupy the whole of the posterior border of the segment. I have compared examples of J. phos- phorea of from 10 to 15 mm, in length with similar sized specimens of I. baltica, but IT must confess that I fail to note any resemblance between the two, the general shape and coloration at once serving to separate them. Although. Harger examined large numbers of specimens from twenty different localities, ranging from the New England coast northwards to Halifax, Nova Scotia, and the Gulf of St. Lawrence, and southwards as far as Cape Cod, he never observed a striped pattern of coloration, so common in J. baltica, with which species it was found associated, the colour being ORAL APPENDAGES OF MARINE ISOPODA. $1 usually dark green or brownish, with patches of yellow or whitish, trans- versely or obliquely arranged. The First Mawilla (Pl. 8. fig. 19) has the outer lobe strongly curved in- wards. There are eleven spines and a fine setule on the ventral surface. Four or five of the innermost spines are denticulate. The inner lobe is fairly large and has three stout setose spines terminally and a setule on the anterior oater margin. The Mazillipede (P1. 8. fig. 20).—Of the two figures given by Harger and Richardson of this appendage, that of the former is the better, although both differ considerably from the condition noted in the specimens I have ex- amined. The coxopodite has the two usual divisions. The basipodite is elongated and narrower than the epipodite. The four-jointed palp is half again as long as the basipodite, the first and third joints are small, whilst the second is longer and greatly expanded at the anterior margin on the outer side. The inner distal lobe extends forward almost to the end of the third joint of the palp, terminally it is surmounted with a number of setose and plain spines. The epipodite is shown by Harger with an almost straight inner margin and a narrow, bluntly pointed apical portion lying beneath the third joint of the palp. In all the specimens I have examined, both the inner and outer margins are curved outwards, and with the apical portion slightly narrower and not extending beyond the second joint of the palp. The metasome (PI. 8. fig. 22) is composed of two short joints, a single suture, and the large terminal segment. This latter is more triangular than in /. granulosa, rather shorter, and shows a faint constriction about its middle, then widening slightly, it terminates somewhat abruptly in a sharp point. The uropod (PI. 8. fig. 23) is characterized by the triangular-shaped endo- podite, which is rather longer than wide. ParipoTEa, Stebbing. 11. PartporEa uneuLATA (Pallas). (Pl. 8. figs. 24, 25.) Oniscus ungulatus, Pallas, Spicil. Zool. vol. ix. (1772) p. 62, pl. 4. fig. 11. Idotea ungulata, Lamarck, Hist. Anim. sans Vert. vol. v. (1818) p. 160. Idotea lalandii, Mine-Edwards, Hist. Nat. Crust. vol. iii. (1840) p. 182, pl. 31. fig. 7 ; Krauss, Siidafrik. Crust. 1848, p. 61. Idotea affinis, Milne-Edwards, Hist. Nat. Crust. vol, iii. (1840) p. 133; Krauss, Siidafrik. Crust. 1843, p. 61. Tdotea edwardsti, Guérin-Méneville, Icon. Réene Anim, 1829-44, Crust. p. 338, Idotea nitida, Heller, Verhandl. zool.-bot. Ges. Wien, 1861, p. 497. ? Idotea excavata, Haswell, Proc. Linn. Soc. N.S.W. vol. vi. (1882) p. 2. LINN. JOURN.—ZOOLOGY, VOL. XXXIV, 82 DR. W. E. COLLINGE ON THE Idotea ungulata, Miers, Journ. Linn. Soe., Zool. vol. xvi. (1881) p. 52; Chilton, Trans. New Zealand Inst. vol. xxii. (1890) p. 196. Paridotea ungulata, Stebbing, Sth. Afr. Crust. 1900, pt. 1. p. 58. The only two references to the oral appendages of this species are a figure of the mandible and the maxillipedes given by Milne-Edwards, and a short description of all the appendages by Stebbing. In this latter account the author remarks: “The first maxille have six strongly plumose set on the narrow inner plate, and ten short apical spines on the outer.’ In the description of the maxillipede the palp is spoken of as consisting of four joints. As will be seen from the following description and accompanying figures, I find many differences from Mr. Stebbing’s account. The First Mazilla (P1. 8. fig. 24).—This appendage is very large in this species. The outer lobe has five stout curved spines on the outer side, then a thin fine curved one and a further six, not so strong as those on the outer side; there is also a long setule, set in a cup-shaped cavity, on the ventral surface. The inner lobe is wide and has four stout setose spines terminally, and a small setule on the middle of the ventral surface close to the anterior margin. In none of the nine specimens examined was any variation in the number of spines observed. The Mawillipede (P1. 8. fig. 25).—This appendage is considerably elongated in this genus. The two divisions of the coxopodite are small, as also the basi- podite, which posteriorly terminates in a blunt point, with a straight outer margin ; anteriorly it flanks the first joint of the palp on both sides; the outer margin gradually curves inwards towards the posterior end. The palp is nearly twice the length of the basipodite. The first joint is small, the second more elongated and slightly expanded laterally, the third joint is still longer and more expanded laterally particularly on the inner side, the fourth joint is the longest, and this and the smaller terminal joint are both densely setose on their inner margins. There are also a few sete on the inner margin of the third and second joints. ‘The inner distal lobe extends forward as far as the middle of the third joint of the palp. It has an outwardly sloping anterior margin from which arise a number of setose spines ; there are three coupling-hooks towards the base of the inner margin. The epipodite pos- teriorly is wider than the basipodite, narrowing anteriorly, and then sloping inwardly its apical portion terminates in a blunt point. It extends forward to about the middle of the third joint of the palp. PARIDOTEA UNGULATA (Pallas) var. nov. ATROVIRENS. Whole of the body a very dark olive-green, almost black. Length 40 mm. Hab. Shore-pools, Brighton, Victoria, Australia: Dec. 18, 1893 (J. J. Lister). Type. In the University Museum of Zoology, Cambridge. ORAL APPENDAGES OF MARINE ISOPODA. 83 12. Partporea ReTicuLATA, Barnard. (PI. 8. figs. 26, 27.) Paridotea reticulata, Barnard, Ann. Sth, Afr. Mus, vol. x. (1914) p. 424, pl. 36 b. Barnard, in describing this beautiful species, states: ‘‘The mouth-parts resemble those of P. ungulata, except that the lateral margins of the epistome are angular, not evenly convex, and the inner lobe of the first maxilla has only three plumose sete.” The First Maailla (P). 8. fig. 26)—The outer lobe terminates in twelve spines, three or four of the innermost being denticulate. The inner lobe is narrower than in P. wngulata and has only three terminal setose spines and a single setule on the anterior outer margin. The Maxillipede (Pl. 8. fig. 27)— Whilst bearing a strong resemblance to the maxillipede of P. ungulata, this appendage in P. reticulata is rather more robust and setose. The basipodite is wider on its posterior margin and more obliquely cut away anteriorly. The second joint of the palp is smaller than in P. ungulata and the third longer and wider. The epipodite is about the same width as the basipodite, but narrows considerably towards its apical end. The whole appendage is densely setose, and there are a series of strong spines on the outer expanded margin of the third joint of the palp. 13. Partporgea ruBRA, Barnard. (PI. 8. figs. 28, 29.) Paridotea rubra, Barnard, Ann. Sth. Afr. Mus. vol. x. (1914) p. 426, pl. 37 a. The mouth-parts of this species have not hitherto been figured. Barnard’s description is brief—he states: * First maxilla, outer plate with ten spines, the two innermost denticulate, inner plate with three plumose sete... . Maxillipede seven-jointed, epipod reaching to the end of the fourth joint, apex incurved.” The First Mawilla (P1. 8. fig. 28).—The outer lobe terminates in ten spines, the outermost one is smaller than the second one, and the two innermost are faintly denticulate ; the inner lobe is narrow and has three setose spines and a prominent setule. The Mawillipede (P1. 8. fig. 29).—The two divisions of the coxopodite are small. The basipodite is narrow and elongated, with both inner and cuter margins almost straight, the posterior margin is obliquely cut away towards the epipodite, anteriorly the segment flanks the first joint of the palp on both sides. The five-jointed palp is typical of the genus. The inner distal lobe is rather longer than in the preceding species. The epipodite is slightly narrower than the basipodite, excepting at its base; it is curved inwards anteriorly and extends as far as the middle of the third joint of the palp, not to the end of the fourth joint as stated by Barnard. ahs 34 DR. W. E. COLLINGE ON THE 14. Parrporna FucicoLta, Barnard. (PI. 8. figs. 30, 31.) Paridotea fucicola, Barnard, Ann. Sth. Afr. Mus. vol. x. (1914) p. 427, pl. 86 5. Barnard has figured the maxillipede, but not altogether satisfactorily. The First Maailla (P1. 8. fig. 30).—This appendage differs from that in P. rubra in having more denticulate spines on the outer lobe and in the presence of a long setule on the inner ventral surface. It is also densely setose on both the inner and the outer margins, the sete on the inner margin being strong and spine-like. The inner lobe differs only in the position of the setule, which in this species is smaller and situated on the outer anterior marein. The Mawillipede (P1. 8. fig. 31).—Excepting that it is much smaller and not so elongated, the maxillipede is very like that of P. rubra. EuroreEa, gen. nov. Body narrow-oblong, not keeled, and nearly smooth. Cephalon anteriorly emarginate, lateral lobes somewhat prominent. Maxillipedes with a palp composed of four joints. Coxal plates narrow, in the second, third, and fourth mesosomatic segments they occupy a little more than half of the lateral margins, in the fifth they occupy almost the whole of the lateral margins, and in the sixth and seventh segments extend from the anterior to the posterior angle. Metasome composed of a single segment and three lateral sutures. 15. Eurporea Peroni (Milne-Edwards}. (PI. 8. figs. 32, 33.) Idotea peronii, Milne-ldwards, Hist. Nat. Crust. vol. iii. (1840) p. 133; Miers, Journ. Linn. Soc., Zool. vol. xvi. (1881) p. 55; Chilton, Trans. New Zealand Inst. vol. xxii. (1890) p. 199. Idotea distincta, Guérin- Méneville, Icon. Régne Anim, 1829-44, Crust. p. 33. Idotea stricta, Dana, U.S. Expl. Exped. 1853, Crust. ii. p. 704, pl. 46. fig. 7; Miers, Journ. Linn. Soe., Zool. vol. xvi. (1881) p. 62. Idotea caudacuta, Haswell, Proc. Linn. Soc. N.S. W. vol. vi. (1882) p. 1, pl. 4. fig. 4. Paridotea peronit, Stebbing, Ann, Sth. Afr. Mus. vol. vi. (1910) p. 483. This species was placed by Stebbing (24, p. 433) in the genus Paridotea, but an examination of the oral appendages at once shows that it cannot remain there, neither can it be placed in the genus Idotea, Fabr.; I have, therefore, proposed for its reception the new genus Hwidotea. The First Mawilla (P1. 8. fig. 32)—The outer lobe terminates in twelve stout spines, free of any denticulation, and a single fine spine ; the inner lobe has three setose spines, and three setules on its ventral surface. The Mawillipede (Pl. 8. fig. 33).—There are two somewhat large divisions of the coxopodite. The basipodite is rather short, but on its inner side ex- tends for some little distance beyond the first joint of the palp. It is about ORAL APPENDAGES OF MARINE ISOPODA. 85 the same width as the epipodite and shorter than the palp. The four-jointed palp has the first joint small, the second joint longer than the third though not so expanded laterally, and the fourth joint eonsiderably larger and oval in shape. The inner distal lobe is elongated, extending almost to the end of the third joint of the palp; terminally it has some strongly developed plain spines and a number of setose ones. The epipodite extends just beyond the anterior margin of the second joint of the palp. ERICHSONELLA, Benedict. 16. EricHsoNELLA ATTENUATA (Harger). (Pl. 8. fig. 34; PI. 9. fig. 35.) Erichsonia attenuata, Harger with Verrill, Rpt. U.S. Comms. F. & F. 1873, pt. 1. p. 570, pl. 6. fig. 27; Harger, ibid. 1880, pt. v1. p. 356, pls. 6, 7. figs. 36, 37. Erichsonella attenuata, Richardson, Bull. No. 54, U.S. Nat. Mus. 1905, p. 400, figs. 447, 448. The First Maailla (P1. 8. fig. 34).—The outer lobe terminates in ten spines, the middle ones being the largest, two or three of the innermost are faintly denticulate. The inner lobe has the anterior margin sloping inwardly and has sometimes four and sometimes three setose spines and a setule on the outer margin ; the inner margin is setose. The Mazillipede (Pl. 9. fig. 35)—Miss Richardson’s figure shows three divisions of the coxopodite, but I have failed to find more than two. The basipodite is rather short and a little narrower than the epipodite. The palp has four wide joints, the first being small and the second and third cup-shaped and gradually enlarging, their inner margins are widely expanded, the fourth joint (the largest) is oval in shape. ‘The setze are small and few in number. The inner distal lobe extends almost as far as the anterior margin of the third joint of the palp. ‘The epipodite is almost truncate at its posterior margin and narrows a little anteriorly and extends just beyond the middle of the third joint of the palp. Synoptic Key to the Species of the Genus Synidotea. A. Terminal seoment of metasome emarginate or notched. a. Lateral cephalic lobes well developed. b. Frontal spines overhanging frontal lobe. c. Outline of metasome triangular ................00-- S. hartepes (Milne-lidw.). S. pallida, Benedict. b', Frontal spines absent. c’. Outline of metasome subparallel.................. S. nebulosa, Benedict. c'. Outline of metasome arcuate ......0..0.2 eee ses . SS. angulata, Benedict. ce’. Outline of mesosome subtriangular. d. Outline of metasome subparallel ..............4. S. marmorata (Packard). d', Outline of mesosome strongly arcuate .......... S. bieuspida (Owen). 86 DR. W. E. COLLINGE ON THE B. Terminal seoment of metasome produced as a blunt spine. a. Lateral cephalic lobes small. 6. Frontal spines not on the margin ........ aera easttaorsl srs S. nodulosa (Kroyer). 3’. Frontal spines on the-margin . 00... eee eee eee S. levis, Benedict. Syniporea, Harger. 17. SyNIDOTEA HIRTIPES (Milne-Edwards). (PI. 9. figs. 36, 37.) Idotea hirtipes, Milne-Edwards, Hist. Nat. Crust. vol. iii. (1840) p. 184; Krauss, Siidafrik. Crust. 1843, p. 61. Edotia hirtipes, Miers, Journ. Linn. Soc., Zool. vol. xvi. (1881) p. 68. Synidotea hirtipes, Benedict, Proc. Acad. Nat. Sci. Philad. 1897, p. 403; Stebbing, Sth. Afr. Crust. 1902, pt. 11. p. 60. Stebbing has described the oral appendages of this species in great detail. He states: “The first maxillee have the outer plate surmounted by ten or sometimes eleven spines, some of which are denticulate, none very powerful. The inner plate is narrow at both ends, and has at the apex only two sete, which are rather long, and, as usual, setose. “The maxillipedes have the first joint short, the epipod nearly parallel- sided, not reaching the apex of the process of the second joint, though extending considerably beyond the first joint of the palp; its upper margin slopes inward. The process of the second joint is shaped as commonly in the Amphipoda Gammaridea, and similarly fringed with sete on the inner and apical margins, but here it is tied to its fellow, each member of the pair carrying a strong spine-hook for grappling the other. The first joint of the palp is smail and rather obscure, the second is very large, widening distally, its distal margin flatly rounded on the inner part and externally forming a little free projection. The third joint is also very large, its inner margin almost continuous with that of the preceding joint, feebly convex, fringed with short spines, its outer margin strongly convex, fringed with setsx-like spines, some of which also stand out from the surface.” The First Maailla (Pl. 9. fig. 36)—In none of the:specimens that I have examined have there been more than ten spines on the outer lobe, in addition to which there is a long setule set in a cup-shaped cavity on the ventral sur- face of the lobe. Ina like manner there are three setules on the ventral surface of tlie narrow inner lobe, the two setose spines of which are rather longer than in most species. The Mawillipede (Pl. 9. fig. 37).—This appendage in the genus Synidotea is characterized by the large size of the second and third lobes of the palp. In S. hirtipes the two divisions of the coxopodite are small. The basipodite has both its inner and outer margins almost straight, the posterior and anterior ones slope outwards. The three-jointed palp is almost twice the length of the inner margin of the basipodite. The first joint is small, the ORAL APPENDAGES OF MARINE ISOPODA. 87 second and third joints are very large. The inner distal lobe extends for- ward slightly beyond the middle of the second joint of the palp, and distally it has a number of setose spines. The epipodite is a wide flattened plate, with almost parallel sides ; it is wider than the basipodite. Its apical portion curves outwardly slightly and extends forward as far as a point a little beyond the middle of the second joint of the palp. 18. SynipoTEa PALLIDA, Benedict. (PI. 9. figs. 38, 39.) Synidotea pallida, Benedict, Proc. Acad, Nat. Sci. Philad. 1897, p. 396; Richardson, Bull. No. 54, U.S. Nat. Mus. 1905, p. 378, figs. 412, 413. The Mirst Mawilla (Pl. 9. fig. 38)—The outer lobe exhibits a strong inward curve. It has eiglit terminal spines, all of which are denticulate, and there are a few setae on the inner anterior margin. The inner lobe is compara- tively small ; terminally there are two long setose spines. The Mawillipede (P|. 9. fig 39).—Miss Richardson’s figure of the maxilli- pede of this species is incomplete and very unlike the actual appendage. The divisions of the coxopodite are small. The basipodite is small, but does not slope outward quite so much as in the preceding species, and the second joint of the palp is smaller, the third joint is twice the length of the second one; the inner distal lobe is also narrower than in S. hirtipes. The epipodite is a wide fattened plate with its apical portion narrowed and sloping outwards ; it extends as far forward as the anterior margin of the second joint of the palp. 19. SyNrpoTEA NEBULOSA, Benedict. (PI. 9. figs. 40, 41.) Synidotea nebulosa, Benedict, Proc. Acad. Nat. Sci. Philad. 1897, p. 397 ; Richardson, Proc. U.S. Nat. Mus. vol. xxi. (1899) p. 848; Bull. No. 54, U.S. Nat. Mus. 1905, p. 381, figs. 416, 417. The First Maxilla (Pl. 9. fig. 40).—The outer lobe has a strong inward curve. There are ten terminal spines, most of which are denticulate ; there are a few setze on both the outer and inner margins. The anterior portion of the inner lobe is roughly triangular with the distal end obliquely cut away and terminating in two stout setose spines. The Mavillipede (Pl. 9. fig. 41).—The shape of the basipodite at once serves to separate this species from any other member of the genus. The second joint of the palp is very large, but the anterior lateral margins are not so expanded as in S. hirtipes. 20. SYNIDOTEA ANGULATA, Benedict. (PI. 9. figs. 42, 43.) Synidotea angulata, Benedict, Proc. Acad. Nat. Sci. Philad. 1897, p. 395; Richardson, Bull. No. 54, U.S, Nat. Mus. 1905, p. 382, figs. 418, 419. The First Mawilla (P1. 9. fig. 42) —The strong inner curve of the outer 88 DR. W. E. COLLINGE ON THE lobe is not so pronounced in this species. There are ten terminal spines, most of which are faintly denticulate. The anterior portion of the inner lobe is oval in shape with two setose spines, longer and more slender than those of S. nebulosa. The Mazillipede (P1. 9. fig. 43) agrees very closely with that of 8. nebulosa, excepting that it is more robust and larger. The third lobe of the palp is more stunted and the basipodite much longer. 21. Synrporea MARMORATA (Packard). (PI. 9. figs. 44, 45.) Idotea. marmorata, Packard, Mem. Bost. Soc. Nat. Hist. vol. i. (1867) p. 296, pl. 8. fig. 6. Idotea bicuspida, Streets & Kingsley, Bull. Mssex Inst. vol. ix. (1877) p. 108. Synidotea bicuspida, Warger, Proc. U.S. Nat. Mus. vol. ii. (1879) p. 160; Rept. U.S. Comms. F. & F. 1880, pt. v1. p. 352. Edotia bicuspida, Miers, Journ. Linn. Soc., Zool. vol. xvi. (1881) p. 66. Synidotea marmorata, Benedict, Proc. Acad. Nat. Sci. Philad. 1897, p. 392; Richardson, Bull. No, 54, U.S. Nat. Mus. 1905, p. 384, figs. 422, 423. The First Mawilla (Pl. 9. fig. 44).—Both of the lobes in this species are comparatively small. The outer one is abruptly narrowed posteriorly ; anteriorly there are eight spines, all of which are denticulate. Sete are present on both the inner and outer margins for a short distance. The Mawillipede (P|. 9. fig. 45)—This appendage bears a strong resem- blance to that of S. nebulosa; there are, however, slight differences in the form of the basipodite and the second lobe of the palp. 22. SYNIDOTEA BICUSPIDA (Owen). (PI. 9. figs. 46, 47.) Idotea bicuspida, Owen, Crust. of the ‘ Blossom,’ 1889, p. 92, pl. 27. fig. 6. Idotea pulchra, Lockington, Proc, Cal. Acad, Sci. vol. vil. (1877) p. 44. Synidotea incisa, G. O. Sars, Crust. et Pycnog. nova etc., 1880, no. 8. Edotia bicuspida, Miers, Journ. Linn. Soc., Zool. vol. xvi. (1881) p. 66. Synidotea bicuspida, G. O. Sars, Crust. Norweg. Nth. Atlantic Exp. 1885, vol. i. p. 116, pl. 10, figs. 24-26; Benedict, Proc. Acad. Nat. Sci. Philad. 1897, p. 391; Richardson, Bull. No. 54, U.S. Nat. Mus. 1905, p. 385, fig. 424, : The oral appendages of this species have not previously been described or figured. They exhibit a relationship to 8S. marmorata, but differ from those in that species in being larger and of a more robust type. The First Maailla (Pl. 9. fig. 46).—The outer lobe is somewhat thickened and has ten spines, most of which are denticulate ; there are a few sete on the inner anterior margin. As in S. marmorata, the anterior end of the inner lobe is somewhat triangular in shape, but the inner margin is almost straight. There are two long setose spines, and a setule on the anterior inner margin, ORAL APPENDAGES OF MARINE [SOPODA. 89 The Muawillipede (PI. 9. fig. 47).—The form of this appendage in S. bicuspida approaches very closely to that found in S. pallida, only it is much larger and the third lobe of the palp is comparatively smaller, and also shorter than the second joint. 23. SYNIDOTEA NODULOSA (Ardyer). (PI. 9. figs. 48, 49.) Idotea nodulosa, Kroyer, Naturhist. Tidsskrift, vol. ii, (1846) p. 100; Voy. en Scand. rust. 1849, pl. 26. fig. 2. Synidotea nodulosa, Harger, Rept. U.S. Comms. F. & F. 1886, pt. v1. p. 35], pl. 6. figs. 35-35. Edotia nodulosa, Miers, Journ. Linn. Soc., Zool. vol. xvi. (1881) p. 67. Synidotea nodulosa, Benedict, Proc. Acad. Nat. Sci. Philad. 1897, p. 898; Richardson, Bull. No. 54, U.S. Nat. Mus. 1905, p. 388, figs. 429, 430. The First Maailla (Pl. 9. fig. 48).—There are ten sharply pointed spines on the outer lobe, the three most ventral of which are denticulate ; the inner lobe has two long setose spines and a single setule on the anterior outer margin. The Mawillipede (P1. 9. fig. 49) agrees very closely with that of S. beeus- pida; the epipodite, however, is rather longer and has its apical portion more pointed. The inner margin of the basipodite anteriorly is produced into a number of serrations. 24. SynipoTEa LZVIS, Benedict. (PI. 9. figs. 50, 51.) Synidotea levis, Benedict, Proc. Acad. Nat. Sci. Philad. 1897, p. 399; Richardson, Bull. No, 54, U.S. Nat. Mus. 1905, p. 389, figs. 481, 432. The First Mazilla (Pl. 9. fig. 50).—The outer lobe has a well-marked slope on its outer side and has eight terminal spines, two or three of which are denticulate ; the inner lobe is somewhat triangular anteriorly, as in S. nebulosa, and has two long setose spines. The Maaillipede (P1. 9. fig. 51).—This appendage in this species is charac- terized by the great development of the second joint of the palp. It is the longest of the three, and anteriorly considerably wider than long, owing to the great expansion of the anterior lateral margins. The epipodite is similar in shape to that in S. marmorata, only somewhat shorter. The inner margin of the basipodite is anteriorly continued forward for some little distance beyond the first joint of the palp. 90 DR. W. E. COLLINGE ON THE VI. BipiioGRAPHy. 1. Barnarp, K. H.—Contributions to the Crustacean Fauna of South Africa. I. Additions to the Marine Isopoda. Ann. Sth. Afr. Mus. vol. x. (1914) pp. 197-230, pls. 17-22. ILI. Additions to the Marine Isopoda, with notes on some pre- viously incompletely known species. bid. pp. 325 a-358 a & 359- 442, pls. 27-38. 3. Benepict, J. H.--A Revision of the Genus Synidotea. Proc. Acad. Nat. Sci. Philad. (1897) pp. 389-404. ‘Two new Isopods of the Genus /dotea from the Coast of California. Proc. Biol. Soc. Washington, vol. xii. (1898) pp. 53-55, figs. 12, 13. 5. Carron, CHas.—On a new Species of Jdotea. Ann. & Mag. Nat. Hist. (s. 5) vol. xv. (1885) pp. 123, 124, pl. 5. figs. 1-3. Revision of the New Zealand Idoteide. Trans. New Zealand Inst. vol. xxii. (1890) pp. 189-204. 7. Cottince, Waurer E.—On the Range of Variation of the Oral Ap- pendages in some Terrestrial Isopods. Journ. Linn. Soc., Zool. vol. xxxii. (1914) pp. 287-293, pls. 20, 21. Description of a new Species of Marine Isopod of the Genus Pentias, Richardson. Journ. Zool. Research, vol. i. (1916) pp. 33- 35, pl. 3. 9. —— On the Marine Isopod Idotea ochotensis, Brandt. Ibid. pp. 82- 85, pl. 4. 2. 4. 6. 8. 10. On the Structure of the Marine Isopod Mesidotea sibirica (Birula), with sone Remarks upon Allied Genera. bid. pp. 112-118, pl. 5. 11. A Note on the Marine Isopod Idotea elongata, Miers. bid. pp. 119, 120. 12. —— Some Remarks upon the Structure and Generic Position of Idotea lacustris, Thomson. Ibid. pp. 153-157, pl. 6. A Revision of the British Idoteide, a Family of Marine Isopoda. Trans. Roy. Soc. Edinb. vol. li. (1917) pp. 721-760, pls. 1-11. 13. Harcer, Oscar.—Report on the Marine Isopoda from New England and adjacent Waters. Rept. U.S. Comms. Fish & Fisheries for 1878, pt. vi. 1880, pp. 297-462, pls. 1-13. 14. HasweLtt, W.—On some new Australian Marine Isopoda. Pt. II. Proc. Linn. Soc. N.S.W. vol. vi. (1881) pp. 181-196, pls. 3, 4. 15. Kroyer, H.—Karcinologiste Bidrag. Naturh. Tidsskr. vol. xi. (1846- 49) I. pp. 1-123, 366-446. 16. Mizrs, EK, J.—Revision of the Idoteide, a family of Sessile-eyed Crustacea. Journ. Linn. Soc., Zool. vol. xvi. (1881) pp. 1-88, pls. 1-3. 12a. 17. 18. 19) 20. 21. 22. 23. 24. Fig. ORAL APPENDAGES OF MARINE ISOPODA. 91 Mityne-Epwarps, H.—Histoire naturelle des Crustacés. Paris, t. iii. (1840) pp. 115-284, pls. 81-33. Ouin, AxEL.—Isopoda from Tierra del Fuego and Patagonia. I. Val- vifera. Svenska Exped. till Magellan. Bd. ii. (1907) pp. 261-306, pls. 20-25. Racovirza, H. G., and Sevasros, R.—Proidotea haugi, n. g.n. sp., Isopode oligocene de Roumanie et les Mesidoteini nouvelle sous- famille des Idotheidee. Arch, Zool. exp. et gén. s. 5, t. vi. (1910) pp. 175-200, pls. 9, 10. Ricwarpson, Harrizr.—A Monograph of the Isopods of North America. Bull. No. 54, U.S. Nat. Mus. 1905, pp. lii+727, & 740 figs. —— Description of a new Genus and Species of Isopod Crustacean of the Family Idotheide, from the Mouth of the Rio de la Plata, Argentina, South America. Proc. U.S. Nat. Mus. vol. xl. (1911) pp. 169-171, figs. 1-4. Sars, G. O.—Account of the Crustacea of Norway. Vol. II. Isopoda. 1897. Srepprne, T. R. R.— South African Crustacea. Pt. I., 1902, pp. 14-66, pls. 1-4 (publ. July 20th, 1900). Pt. I1., 1904, pp. 1-92, pls. 5-16 (publ. Oct. 7th, 1902). Marine Investigations in South Africa. — General Catalogue of South African Crustacea. Ann. Sth, Afr. Mus. vol. vi. (1910) pp. 281-593, pls. 15-22. EXPLANATION OF THE PLATES. No attempt has been made to introduce the exact number of sete into the different figures. PuLatE 7. 1. Chiridotea ceca (Say). Ventral view of the right maxillipede. x 38, 2: tuftsti (Stimpson). Ventral view of the left maxillipede. x 38. 3. Mesidotea sabini (Kroyer). Ventral view of the terminal portions of the inner and outer lobes of the right first maxilla. x 19. 4. — Ventral view of the terminal portion of the second maxilla. x 19. 5, Ventral view of the right maxillipede. x 9. 6. Pentidotea resecata (Stimpson). Ventral view of the terminal portions of the inner and outer lobes of the right first maxilla. x 19. —<— Ventral view or the left maxillipede. x 7:5. 8. ——- wosnesenskti (Brandt). Ventral view of the terminal portions of the inner and outer lobes of the right first maxilla. x 19. 9, -— Ventral view of the right maxillipede. x 19. 10. whitei (Stimpson). Ventral view of the terminal portions of the inner and outer lobes of the left first maxilla. x 38. lil, Ventral view of the left maxillipede. x 19, Fig. Fig. 41. 42. DR. W. E. COLLINGE ON THE 2. Glyptidotea lichtensteini (Krauss). Ventral view of the terminal portions of the inner and outer lobes of the right first maxilla. x 38. Ventral view of the left maxillipede. x 12. . Idotea rectilinea, Lockington. Ventral view of the terminal portions of the inner and outer lobes of the left first maxilla. x 88. Ventral view of the right maxillipede. x 19, — phosphorea, Harger. Dorsal view of the cephalon. Dorsal view of the right antennule. x 38. — —— Dorsal view of the left antenna. x 21°65. PLATE 8, . Idotea phosphorea, Harger. Ventral view of the terminal portions of the inner and outer lobes of the left first maxilla. x 78. — Ventral view of the left maxillipede. x 38. Dorsal view of the lateral portions of the mesosomatic segments, showing the coxal plates. Dorsal view of the metasome. xX 3'5. Left uropod. x 10. . Paridotea ungulata (Pallas). Ventral view of the terminal portions of the inner and outer lobes of the right first maxilla, x 38. Ventral view of the right maxillipede. x 9°5. reticulata, Barnard. Ventral view of the terminal portions of the inner and outer lobes of the left first maxilla. x 38. Ventral view of the left maxillipede. x 9°5. —— rubra, Barnard. Ventral view of the terminal portions of the inner and outer lobes of the left first maxilla. x 88. —— — Ventral view of the left maxillipede. x 12, fucicola, Barnard. Ventral view of the terminal portions of the inner and outer lobes of the first maxilla. x 24. —— Ventral view of the left maxillipede. x 24. 32. Euidoteu peronii (Milne-Edwards). Ventral view of the terminal portions of- the inner and outer lobes of the right first maxilla. x 38. Ventral view of the right maxillipede. x 14. . Evichsonella attenuata (Harger). Ventral view of the terminal portions of the inner and outer lobes of the right first maxilla. x 75. PuateE 9. . Lrichsonella attenuata (Harger), Ventral view of the right maxillipede. x 88. . Synidotea lirtiyes (Milne-Edwards). Ventral view of the terminal portions of the inner and outer lobes of the right first maxilla. x 38. Ventral view of the left maxillipede. x 16. . — pallida, Benedict. Ventral view of the terminal portions of the inner and outer lobes of the right first maxilla. x 38. Ventral view of the right maxillipede. x 18. nebulosa, Benedict. Ventral view of the terminal portions of the inner and outer lobes of the left first maxilla. x 38. Ventral view of the left maxillipede. x 19. angulata, Benedict. Ventral view of the terminal portions of the inner and outer lobes of the right first maxilla. x 38. XXXIV, PL 7. Journ. Lunn. Soc. Zoor. Vor. SP Collinge. Az .J.J.del. ad. nat. C.Hodges & Son. ]ith.& imp. ORAL APPENDAGES OF ISOPODA. XXXIV. PL 8. Collinge. . Journ. Linn Soc. Zoon. VoL. SEP re— We : srw . oe . SS: i : : fe. SJ. del.ad. nat. C Hodges & Son.lith &imp. ORAL APPENDAGES OF ISOPODA., XXXIV. PL 9. Collinge Journ. Linn. Soc. Zoor Vor. ot Pr-o: Hi AR. J.J. del.ad nat . C. Hodges & Son. th.& amp. ORAL APPENDAGES OF ISOPODA. ORAL APPENDAGES OF MARINE ISOPODA. 93 Fig.43, Synidotea angulata, Benedict. Ventral view of the left maxillipede. x 12. 44. marmorata (Packard). Ventral view of the terminal portions of the inner and outer lobes of the right first maxilla. x 38. 45, —— Ventral view of the right maxillipede. x 19. 46, bieuspida (Owen). Ventral view of the terminal portions of the inner and outer lobes of the left first maxilla. x 26. 47. Veutral view of the left maxillipede. x 10. 48. -— nodulosa (Kroyer). Ventral view of the terminal portions of the inner and outer lobes of the left first maxilla. > 100. 49, —— Ventral view of the right maxillipede. x 38. 50. —— levis, Benedict. Ventral view of the terminal portions of the inner and outer lobes of the left first maxilla. x 38. 51. —— Ventral view of the left maxillipede. x 19. The author desires to thank the Carnegie Trust for the Universities of Scotland for a grant to defray artists’ charges. RULES FOR BORROWING BOOKS FROM THE LIBRARY. 1. No more than Six volumes shall be lent to one person at the same time without the special leave of the Council or one of the Secretaries. 2. All books shall be returned before the expiration of Six weeks from the time of their being taken out, but if not required by any other Fellow, they may, on application, be kept for a further period of Six weeks. 3. All books lent shall be regularly entered by the Librarian in a book appropriated for that purpose. 4. No work forming part of Linneus’s own Library shall be lent out of the Library under any circumstances. Norn.— Certain other works are included in this prohibition, such as costly illustrated works, and volumes belonging to sets which could not be replaced if lost. A GENERAL INDEX to the first twenty Volumes of the Journal _ (Zoology) may be had on application, either in cloth or in sheets foz binding. Price to Fellows, 15s.; to the Public, 20s. IN -OWRECIES Tue attention of the Fellows, and of Librarians of other Societies, is requested to the fact that TWO volumes of the Journal (Zoology) are in course of simultaneous issue, as follows :— Vou. 33. Nos. 222 and 223; No. 224 is in the Press, and will complete the volume. Vou. 34. No. 225, the present number. Authors are entitled to 25 copies of their communications gratuitously, and may obtain another 75 by payment, as shown on the printed slip which accompanies the proof. If more than 100 copies are wanted, application must be made to the Council. Abstracts of the proceedings at each General Meeting and Agenda for the uext are supplied to all Fellows. B. DAYDON JACKSON, General Secretary. JANUARY 23, 1919. Price 12s. EH JOU RN WL es 3 Bay OF Py ON oh ke ery aks THE LINNBAN SOCIETY, Vou. XXXIV. ZOOLOGY. No. 226. CONTENTS. Page I. Sponra Runtana.—III. The Distribution of certain Diatoms and Copepoda, throughout the year, in the Irish Sea. By W. A. Herpman, F.R.S., F.L.S., Professor of Zoology in the University of Liverpool. (With 21 Text-figures.) ............... 95 [Reprinted from Journ. Linn. Soc., Bot. vol. xliv. pp. 173-204. ] IJ. THe Percy Stapen Trust EXpaepirions To THE ABROLHOS Istanps (Indian Ocean).—Reporr I. Introduction, General description of the Coral Islands forming the Houtman Abrolhos Group, the Formation of the Islands. By W.J.Daxtn, D.Sc., F.LS., F.Z.8., Professor of Biology, University of West Australia. (Plates 10-14, and 12 Text-figures.).................. 127 III. Restorations of the Head of Osteolepis. By Epwin 8S. Goopricu, F.R.S., Zool.Sec.L.S., Fellow of Merton College, Oxford. (QW itheGelext-founes 3) ie cot ceneameremeec see tetecestesesiecsiesene 181 LONDON: SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON HOUSE, j PICCADILLY, W., AND BY LONGMANS, GREEN, AND CO., AND WILLIAMS AND NORGATE, 1919. LINNEAN SOCIETY OF LONDON. LIST OF THE OFFICERS AND COUNCIL. Elected 24th May, 1918. PRESIDENT. Sir David Prain, C.M.G., C.I.E., F.R.S, VICE-PRESIDENTS. Dr. W. Bateson, F.R.S. R. I. Pocock, F.R.S. Horace W. Monckton, F.G.S. Dr. D. H. Scott, F.R.S. TREASURER. Horace W. Monckton, F.G.S. SECRETARIES. E. 8. Goodrich, M.A., F.R.S. | Dr. A. B. Rendle, F.R.S. GENERAL SECRETARY. Dr. B. Daydon Jackson. COUNCIL. Mrs. Arber, D.Sc. Miss G. Lister. Edmund G. Baker, Esq. Horace W. Monckton, F.G.S. Dr. W. Bateson, F.R.S. Dr. G. E. Nicholls. R. H. Burne, Esq. R. I. Pocock, F.R.S. Sir Frank Crisp, Bt. Sir David Prain, C.M.G., C.L.E., F.R.S. Stanley Edwards, F.Z.S. Dr. A. B. Rendle, F.R.S. Prof. J. B. Farmer, F.R.S. Dr. D. H. Scott, F.R.S. E.S. Goodrich, M.A., F.R.S. Miss A. Lorrain Smith. Dr. B. Daydon Jackson. Arthur W. Sutton, Esq. C. C. Lacaita, Esq. Lt.-Col. J. H. Tull Walsh. LIBRARY COMMITTEE. The Officers ex officio, with the following in addition :— E. G. Baker, Esq. Prof. E. B. Poulton, F.R.S. Dr. W. Bateson, F.R.S. Dr. E. J. Salisbury. E. T. Browne, M.A. F. N. Williams, Esq. C. C. Lacaita, Esq. B. B. Woodward, F.G.S. R. I. Pocock, F.R.S. DISTRIBUTION OF DIATOMS AND COPEPODA IN THE IRISH SEA. 95 SpouiA Runiana.—lll. The Distribution of certain Diatoms and Copepoda, ~ throughout the year, in the Irish Sea. By W. A. Herpway; F.R.S., F.L.S., Professor of Zoology in the University of Liverpooke | __ (With 21 Text-figures.) [Read lst November, 1917.] aK Reprinted from Journ. Linn, Soc., Bot. vol. xliv. (1918), pp. 178-204. Jy Parr of the work * of the yacht ‘Runa’ for some years previous to 1914 consisted in taking periodic samples of the marine plankton at various localities around Port Hrin, at the south end of the Isle of Man, during the two most interesting times in the annual cycle—viz., spring (March-April) and autumn (July-September). During the remaining months, when the yacht was not in commission, plankton gatherings in Port Erin bay were taken with great regularity at the rate of six in the week, three at a time on two occasions per week, two of the three hauls being horizontal and the third vertical. This systematic plankton survey has been continued for fully 10 years (1907-1917 inclusive), and over 5000 + samples have been collected and examined. The general results of this intensive study of the plankton of a central area of the Irish Sea have been given in a series of reports t drawn up in collaboration with Mr. Andrew Scott, A.L.S., and others, and published by the Lancashire and Western Sea-Fisheries Committee; but the material and statistics collected still contain much information which has not yet been made use of. It is proposed in the present communication § to deal with the records of the occurrence throughout the year in our district of a few of the most abundant of the Diatoms and the Copepoda which make up the bulk of the phytoplankton and of the zooplankton respectively at those periods of the year when they are most abundant. At the time of the spring maximum (usually in April or May) a small silk tow-net hauled for about 15 minutes through about half-a-mile of the surface water of the Irish Sea will usually catch some millions of individual Diatoms (up to a couple of hundred millions || on occasions), constituting probably, on the average, some 999,999 out of each million of organisms in the gathering ™. This is almost * For Parts I. and I. of “Spolia Runiana”’ see Journ. Linn. Soce., Zool. xxxii. p. 163 (1913), and p. 269 (1914). + More precisely 5116, to the end of 1916. t Trans. Biol. Soc. Liverpool, xxii. (1908) to xxxi. (1917), § I wish to acknowledge, with thanks, the help I have received in the preparation of these plankton records from Mr, Andrew Scott, A.L.S., and from my secretary, Miss H. M. Lewis, B.A. Mr. Scott took for me the excellent photo-micrographs of the plankton from which most of the illustrations have been reproduced. || Estimated by counting measured samples. q The average of a number of cases where smaller, but still very large, hauls of Diatoms were taken is—Diatoms=about 99 per cent. of the total organisms present. LINN. JOURN.—ZOOLOGY, VOL. XXXIV. 8 96 PROF. W. A. HERDMAN ON THE DISTRIBUTION a pure sample of Diatoms—a “ monotonic phytoplankton.” Similarly, when the zooplankton is at its height in late summer (usually September) the same net may contain almost a pure gathering of Copepoda numbering some tens to hundreds of thousands of individuals (up to 214,000), and making up perhaps 999 ont of every thousand organisms present--a ‘“ monotonic zooplankton.” But we may still regard the gathering as a zooplankton if over 30 per cent. of the organisms are animals—on account of their greater bulk. } Moreover, these very abundant Diatoms and Copepoda belong in each case to very few species, so that one can select about half-a-dozen species of Copepoda which constitute by far the greater part of the summer zooplankton, and about the same number of Diatoms which similarly make up the bulk of the spring phytoplankton. These few species, belonging to these two very widely separated groups, thus come to be the most significant organisms Fre. 1.—‘‘ Hensen,” “ Nansen,” “ Funnel” and other plankton nets drying after use on the yacht. in relation to the annual metabolic cycle of our seas and the food-supply from our coastal fisheries. Consequently it is of both scientific and economic importance to obtain such data as seem possible from our long series of observations, extending over a decade, as to the occurrence of these dominant factors in the plankton. No doubt there are exceptional years with unusual occurrences which will have a disturbing effect, but the ten or eleven years’ results ought to give us an average of some value. We have endeavoured, in our work from the yacht, as the result of many experiments, to make use of a standard net in a constant manner so that the hauls should be approximately comparable. Our two horizontal tow-nets of fine-meshed miller’s silk (No, 20 and No. 9) measure 35 cm. (about 14 inches) OF DIATOMS AND COPEPODA IN THE IRISH SEA. 7 in diameter at the mouth, and are in our weekly gatherings throughout the year hauled through a definite course in the open water of Port Erin bay. The two together (“coarse”’ and “ fine” nets) constitute a ‘standard haul.” The gatherings taken outside the bay during the periods of the vernal and autumnal plankton maxima were made with the same nets as in the bay, at fixed “stations’’ respectively three and five (and ona few occasions ten) miles from land *. In addition to these standard hauls many others were taken with special nets, such as the closing “ Petersen-Hensen”’ and ‘“ Nansen,” the ‘“ shear- net” and “funnel-nets,” usually at depths of 5 to 30 fathoms, and on occasions down to 60 fathoms out in the deep central valley between the Isle of Man JAN Fea Mar Aru May June = Juv Aucr Ser Ocr Nov. Dec Fie, 2.—Typical Irish Sea plankton curves for the years 1913 (dotted line) and 1914 (whole line). and Ireland. Figure 1 gives a general view of the various nets used on one of our plankton cruises. Temperatures, surface and deeper, weather obserya- tions and water-samples were also taken systematically f. The plankton gatherings when taken are at once treated with formol to kill and fix the organisms and prevent further changes. They are then placed (at the Port Erin Biological Station) in graduated cylinders, and after some hours, when the material has fallen to the bottom and the super- jacent fluid is clear, the quantity of plankton in cubic centimetres is recorded. The fixed plankton is then concentrated and preserved in * Determined by cross-bearings, and by distance run calculated from the engities. + For further particulars as to methods, see the earlier reports published by the Lanca- shire and Western Sea-Fisheries Committee. 8* 98 PROF. W. A. HERDMAN ON THE DISTRIBUTION 5 per cent. formol, and is re-measured it may be weeks later when submitted to detailed examination. It is this last more accurate measure- ment that has been made use of for calculations and curves. The first estimation of the volume is only taken in case some accident happens later ; but we have been very fortunate in that respect: we have only lost about 10 gatherings in the 10 years out of over 5000 samples. Figure 2 shows typical plankton curves for the years 1913 (dotted line) and 1914, taken from the recorded monthly averages of the plankton hauls. As an example of the height to which individual hauls may rise in spring, I may quote from our records of this year that on April 18th, 1917, the standard haul measured 165 c.c., and on April 19th 150 c.c., and consisted mainly of Chetoceras. Our confidence that these samples are adequate and representative receives support from the fact that the same organismis are recorded in much the same quantities year after year, and that practically no new forms turn up. Mr. Andrew Scott, A.L.S., who has made a detailed microscopic examination of all the material for the purpose of determining the species, has met with none new to science. Some rare species previously unknown in British seas, such as the Norwegian Copepod Microcalanus pusillus, Sars, which appeared first in our deeper nets in the summer of 1907, and the Indo-Pacific Diatom Biddulphia sinensis, Grev., have occurred ; but throughout the series of over 5000 gatherings, extending over nearly every week of 10 years, no species actually new to science has been determined from the macro- plankton. The various new Copepoda which have been described from time to time from our work at Port Erin have all been bottom-living forms obtained by dredging. This is, so far, a satisfactory result of our work, as it seems to indicate that probably all the pelagic species of Copepoda in our sea are now known. It is not the rare species that are of most interest. They may have an interest of their own—morphoiogical or but for my present purpose it is the common species that are of most import- ance, those species which by their abundance in nature play their part in providing fish-food for man or in affecting the public health either by keeping the sea clean or by causing plagues. For the purpose then of arriving at some conclusion as to the distribution throughout the year of these really significant organisms, | have picked out from our records the following six species of Copepoda as being undoubtedly the most abundant and economically the most important representatives of that section of the plankton :—Oithona helgolandica (= similis), Pseudo- calanus elongatus, Acartia clausi, Temora longicornis, Paracalanus parvus, and Calanus finmarchicus. These are all cases of genera where there is only one species in our seas (e. g., Calanus) or only one common species (e. g., Oithona), so that we are foie om half-a-dozen very distinct forms, and there can be no doubt as to what is in question even if the genus only is referred to. OF DIATOMS AND COPEPODA IN THE IRISH SEA. 99 When we turn to the Diatoms the case is rather different. There several of the more prominent genera are represented by a number of common species, and moreover some of the species are closely related, and variable, so that doubts may arise as to the exact identifications, and authorities may differ as to the relative proportions in which certain species or varieties of, say, Chetoceras or Biddulphia are present in the plankton. Under these circumstances I have considered it will best serve my purpose, which is a general and economic rather than a detailed speciographic one, if I deal with generic titles only, grouping together for example all the species of Chetoceras that may occur under that one name. I have chosen the following genera as being the most important representatives of the Diatoms in our plankton :—Biddulphia, Chetoceras, Coscinodiscus, Rhizosolenia, Thalassio- sira, Guinardia, and Lauderia. In some of these genera (e.g., Guinardia and Lauderia) as in the case of the Copepoda there is only one possible species in question, in biddulphia it is in most eases only the species B. mobi- liensis, but in others (e. g., Chetoceras, Coscinodiscus, and Rhizosolenia) there are usually several allied species occurring together in profusion in any large gathering of the genus. I may add that our commonest species in the Irish Sea off Port Erin are not necessarily those that are most abundant in other seas of North-West Europe. For example, in the Baltic near Kiel, according to Lohmann, the most abundant Diatom is Skeletonema costatum, a comparatively rare form in our plankton, and George Murray similarly found that to be the commonest form he had met with in a plankton survey of some of the more sheltered lochs of the West of Scotland. It is, however, one of the minuter forms which readily escapes notice, and may to a considerable extent pass through the meshes of the net. Then again, in July 1911, in Upper Loch Torridon, on the West Coast of Scotland, I got a haul of 334,000,000 Nitzschta delicatissima, which is rare with us in the Irish Sea, but is apparently more abundant at Plymouth. I think it probable, however, that our Port Erin results will be found to hold good for the more open sea-water of high salinity * around the British Isles. A valuable paper which appeared recently on the Plankton of Plymouth Sound, by Dr. Marie V. Lebour +, while dealing mainly with the more minute Protozoa and Protophyta which escape the tow-net and can ovly be obtained by centrifuging samples of water, gave also some records of the occurrence of some of the larger forms which enables a comparison to be made between the plankton conditions in the English Channel and in the northern part of the Irish Sea. * The salinity off Port Hrin averages about 34-2 per mille. Its range for April, May, and June in the year when we took the most complete series of observations is from 34:02 to 34-4 per mille. + Journ. Mar. Biol, Assoc. vol. xi. 2 (1917), p. 188. 100 PROF. W. A. HERDMAN ON THE DISTRIBUTION There are certain differences in detail. For example, the total Diatom curve at Plymouth has three maxima or crests, in April, August, and October. At Port Erin the curve has only two crests, a much greater maximum in spring and a variable and smaller one in autumn, while Diatoms are usually wholly absent in August. On the other hand, there is a general agreement in regard to the distribution throughout the year of many of the more abundant organisms. For example, amongst Diatoms Coscinodiscus is a winter and early spring form, Biddulphia flourishes throughout the winter from November to April or May, Rhizosolenia is a summer form having its maximum in June, while Chetoceras and Lauderia have two maxima, the one in spring and the other in autumn, in the English Channel and the Irish Sea alike. Amongst Copepoda there seems to be a general agreement along with a certain amount of difference in detail which will be referred to below when discussing the species. I may recall that in November 1910 I read a paper before this Society * in which I made a comparison between the summer (July) plankton on the West Coast of Scotland and that of the Irish Sea, showing that in some of the deep fjord-like highland sea-lochs green-coloured phytoplankton can be obtained even in the height of summer, while a zooplankton may be found living in abundance a few miles away. This, of course, would be impossible in the Irish Sea, where a zooplankton and a phytoplankton do not occur simultaneously. DIATOMS. The seven generic forms I have selected for consideration taken together make up nearly the whole of the Diatom plankton of the year. No other genus occurs in anything like such profusion as these. In April, for example, when the Diatoms are usually at their climax, all the remaining genera (at most 10 or 12) taken together make up only about one-thousandth, or less, of the whole. Moreover, these common Diatoms often attain their greatest profusion successively, not simultaneously, so that single genera, or it may be single species of a genus, make up on occasions the bulk of the phyto- plankton. For example, in May 1916 the month’s average haul of Diatoms was 7,171,789, while the average for the genus Chetoceras taken alone was 6,947,333, leaving only 224,456 as the average of all the rest of the Diatoms. On the last two individual hauls, taken on May 25th and 29th, the actual numbers were as follows :— Chetoceras alone... ian 24,094,500 2a 19,461,600 C. sociale alone Sa oy 23,936,000 ae 19,396,000 All other Diatoms together... 166,300 ioe 228,900 So that on these occasions, and they are examples of many, one species makes up nearly the whole of the plankton. * Journ. Linn. Soc., Zool. xxxii, (1918) p. 28. OF DIATOMS AND COPEPODA IN THE IRISH SEA. 101 The maximum on the Diatom curve ranges from March to May. In 1907 it was in March, in 1908 in May, and in 1909 in April. In some years the Diatom maximum may be divided into two parts, an earlier due mainly to Chetoceras and Thalassiosira, and a later in June due to Fhizosolenia and Guinardia. A common order of succession for the species which contribute most largely to the Diatom maxima is-—biddulphia mobiliensis and Coscinodiscus radiatus in early April, Chetoceras debile in late April, Cluetoceras sociale in May, Chetoceras teres and Rhizosolenia Shrubsolei in early June, and Rhizosolenia (several species) and Guinardia in later June. The autumn Diatom maximum is constituted mainly in the Irish Sea by Chetoceras subtile and Rhizosolenia semispina. Certain species of most of the genera we are dealing with are commonly regarded as “oceanic” in the sense that they are characteristic of the open sea, although they may occasionally be carried in shore and so form a part of the coastal plankton ; while other species are “neritic,” having their origin and their home in coastal waters and not being found normally in the open ocean. For example, Chetoceras densum, C. boreale, Coscinodiscus radiatus, and Rhizoso- lenia semispina are supposed to be oceanic; while Biddulphia mobiliensis, Chetoceras subtile (and other species), Coscinodiscus concinnus, Rhizosolenia setigera, and Lauderia borealis are probably neritic. Other species of these genera are of doubtful position in this classification, or it may be are “panthalassic”’ or equally at home in both regions. BIDDULPHIA. This is a winter and early spring group of Diatoms, generally appearing in September or October, becoming more abundant in November and reaching a maximum in March or April. It dies off during May, and is practically absent in June, July, and August. The species, or forms, that occur in our plankton are biddulphia mobiliensis (2B. regia) and B. sinensis (fig. 3). This is one of the cases where there is some difference of opinion as to the validity of species. Whether B. regia is only a form of B. mobiliensis, and what is the exact relation of B. sinensis to the others, has been and may still be a matter of discussion. Specimens can be picked out that seem distinct and characteristic, but others occur in nature that are intermediate and possibly abnormal *. B, sinensis is an exotic, oceanic form which, according to Ostenfeld, made its appearance at the mouth of the Hlbe in 1893, and spread during successive years in several directions. It appeared suddenly in our plankton gatherings at Port Erin in November 1909, and has been present in abundance each year since. Ostenfeld, in 1908, when tracing its spread in the North Sea, predicted that it would soon be found in the English Channel. Miss Lebour, who has recently examined the old plankton gatherings at the Plymouth * See forms figured in Trans. Biol. Soc. Liverpool, xxvii. (1918) p. 210. 102 PROF. W. A. HERDMAN ON THE DisTRIBUTION Laboratory, finds that as a matter of fact this form did appear in abundance in the collections of October 1909, within a month of the time when according to our records it reached Port Brin. Whether or not this species was brought accidentally by a ship from the far East, there is no doubt that it was not present in our Irish Sea plankton gatherings previous to 1909, but has been abundant since that year, and has completely adopted the habits of its English relations—appearing with B. mobiliensis in late autumn, @ ~ on a. x : — | Fic. 3.—Photo-micrograph of a plankton preparation showing (a) Biddulphia mobiliensis, forma “regia,” and (b) B. sinensis. persisting during the winter, reaching a maximum in spring, and dying out before summer. Biddulphia is generally the first Diatom to show a marked increase in early spring, and is responsible for the moderate rise in the curve which takes place in February or March. Our largest records of Biddulphia are as follows :— From three hundred thousand to over six hundred thousand per haul on several occasions between middle of March and middle of April in 1910 and 1911, and towards end of April and beginning of May, 1916. From three hundred thousand to over seven hundred and fifty thousand on ten occasions between November 10th and 27th, 1911, 1914, and 1915. The highest monthly averages in the early spring months occur as follows :— February ... in 1907 and 1912. March ... in 1910, 1911, 1914, and 1915. April Ss in 1908, 1909, 1913, and 1916. A second, usually slighter, maximum occurs in November, when the numbers are higher than in December and January ; but on three occasions OF DIATOMS AND COPEPODA IN THE IRISH SEA. 103 (1911, 1914, and 1915) in our ten years the November monthly average is the highest in its year. Moreover, two of these November records, viz., 341,231 in 1911 and 406,100 in 1914, are the highest in the whole series. So Biddulphia may be regarded as characteristic of the late autumn (November) as well as the early spring plankton. CoscINODISCcUs. The more abundant species that our records deal with are Coscinodiscus concinnus, C. Grani, and C. radiatus. They are mainly winter and spring forms, the maximum of the genus occurring always in our experience in either March or April. Coscinodiscus (fig. 4) agrees very closely with Biddulphia in its distribution throughout the year, beginning in late autumn, maintaining its position throughout the winter, increasing in January or Fic, 4.—Photo-micrograph of an early spring phytoplankton consisting mainly of Coscinodiseus and Biddulphia. February, and then more rapidly in March and April, and dying away before the height of the summer; but it sometimes continues to be present in small quantities further into the summer months than is the case with Biddulphia. The presence of Cose’nodiscus in quantity in the spring plankton is easily seen through a glass jar with the unaided eye, as the little rounded drum- like cells give a characteristic granular appearance to the gathering. Our largest records of Coscinodiscus are :— From four to five hundred thousand per haul on several occasions between middle of March and middle of April, 1911, 1912, 1913, 1915, and 1916. About nine hundred thousand on several occasions at end of April 1914. From one million to close on five millions on two occasions in middle of April 1915. So Coseinodiscus may outrival its companion form Biddulphia at the time of the vernal maximum, but does not attain to such high numbers in late 104 PROF. W. A. HERDMAN ON THE DISTRIBUTION autumn. The October and November monthly averages never, in our series of years, come anywhere near the averages for March and April. Both Biddulphia and Coscinodiscus seem to be slightly later in their maxima in the Irish Sea than in the English Channel, judging from the Plymouth records. In the years 1907 to 1912, inclusive, the highest monthly averages (March and April) for Coscinodiscus run in general about 100,000. In 1913 and 1916 they are close on 200,000, in 1914 over 300,000, and in 1915 there is a sudden jump, in April, to over 840,000. The other months of that year do not show unusually high numbers. CHA@TOCERAS. The chief species of this genus in our Irish Sea plankton are Chetoceras boreale, C. contortum, C. criophilum, C. debile, C. decipiens (fig. 6), C. densum, April 8 millions vu 3 RS v 3 I Sune ' iA t \ Gf g} \ 6 a BHM uv ~! 9; \ ua) \ a! \ : Ny ‘ <} 4 Rei \ u 1 ' \ ' 1 ' ‘ ! 1 t 1 ! 1 4 ' ' ' \ ! \ 1 ! ‘ I ' 1 ? 1 ! i 1 ! if \ i) ’ \ a 1 Ha \ / 1¢ ¥ \ i> is \ H 2 1 A \ 1 i iy \ ‘ pe OF ' Lae aa \ 19 \ \ ! \ M . . \ | million ! \ \ \ Chaeto¢eras Jan. Feb. Mar, Apr. May June July Aug. Sept. Oct. Noy. Dec. Fie, 5.—Curves of the more important Diatoms of the April and June maxima. C. diadema, C. sociale, and C. teres. Of these, C. boreale, C. criophilum, and C. decipiens are Arctic oceanic forms, C. densum is a temperate oceanic or OF DIATOMS AND COPEPODA IN THE IRISH SEA. 105 Atlantic species, and all the rest may be classified as temperate neritic. Some of these (C. eriophilum, C. debile, C. decipiens, C. sociale, and C. teres) are spring forms with a maximum in April or May, while C. boreale and C. denswn are autumn species having their maxima in September or October. Consequently the genus is well represented throughout a con- siderable part of the year, and the numbers are very high in April and May, and sometimes also in September and October (see curve, fig. 5). A few of our highest records for the genus, giving the nearest million in each case, are :—151 millions on May 4th, 1914, 95 millions on April 29th, 1912, 68 millions on May 16th, 1911, 49 millions on April 22nd, 1910, and 44 millions on May 19th, 1911. The highest record we have for the autumn species is 30 millions on September 26th, 1912. On May 16th, 1911, C. debile contributed 30 millions and C. sociale 12 millions to the total in the haul ; and on May 4th, 1914, C. debile gave 148 out of the 151 millions 2 LN. £ oe Zi {2 in BE pig Fu: ie aa mile (7 g 2S, BY op) °F IG Prin Pg Se ens) ei \ Sp! Wi PNK ba \ & "y, \ 2 Fie. 6.— Chetoceras decipiens, showing the active winter growth. From a photo-micrograph by A. Scott. present. If we examine the records of the separate species for the year 1914 as an example, we find that C. contortum has an average of 62,700 per haul in May, C. debile an average of 867,878 in April and 18,972,800 in May (the record), C. decipiens an average of 821,811 in April and 321,050 in May, C. sociale an average of 1,229,500 in May, C. teres an average of 577,867 in April ; while of the autumn species C. boreale has an average of 53,200 in September and 54,644 in October, and C. densuwm has 151,120 in September and 100,624 in October. C. debile and C. decipiens were also very abundant that October. The highest monthly averages for Chetoceras fall as follows during our ten years :— March... in 1907. April ... in 1909, 1910, and 1912. May ... in 1908, 1911, 1913, 1914, 1915, and 1916. 106 PROF. W. A. HERDMAN ON THE DISTRIBUTION The highest average is nearly 23 millions, in May 1911. None of the September and October averages run as high as those in spring, and only two reach millions, viz., 3,956,047 in October 1911, and 7,702,658 in September 1912. The years 1911 and 1912 had high numbers of Chetoceras throughout many of the months*. There are no montlis in the ten years when Chetoceras was totally unrepresented ; but July and August show the lowest averages— the lowest of all being only six individual cells in August 1907. LAUDERIA. We have only the one species, Lauderia borealis, Gran: (fig. 7), in our records. It is a late spring or early summer form, occurring generally from March or April to June or July, with a later, smaller, occurrence in autumn. It is sometimes present in large quantities, e. g., 20,064,000 on April 22nd, 1910; 123 millions on April 29th, 1912; 3,600,000 on May 4th, 1914. The maximum is towards the end of April or beginning of May, when Lauderia helps, along with Chetoceras, to form the main crest on the vernal Diatom curve (see fig. 5). Fic. 7,—Photo-micrograph showing a chain of Lauderia borealis. THALASSIOSIRA. The only species of this genus that are of any importance in our records are 7. gravida, Cleve, and T. Nordenskioldi, Cleve (fig. 8). Apparently T. gravida is the only one common at Plymouth, but 7. Nordenskioldi, along with Chetoceras contortum and C. debile, helped to constitute the vernal maximum at Port Erin in 1907, and has been still more abundant on several occasions since. 7. Nordenskioldi is in the main a neritic, arctic or Scandinavian species, and probably its occasional occurrences in quantity are to be regarded as invasions of some arctic water and northern plankton into our British seas. In April 1917, it was abundant-at Port Hrin along with Chetoceras teres, C. debile, and C. decipiens. All our high records (over a million per haul) for Thalassiosira lie between late in April and late in May, and the two highest are six millions on April 29th, 1912, and six and a half millions on May 16th, 1913. Other high * The largest hauls of Diatoms as a whole, all species taken together, were in May of 1912 and 1913 (see Table on p. 188). OF DIATOMS AND COPEPODA IN THE IRISH SEA, 107 records have occurred during these same months, in 1910, 1911, 1912, 1913, 1914, 1915, and 1916. The highest monthly averages for Thalassiosira are 721,000 in April 1912, 898,000 in May 1913, and 976,000 in May 1915. The genus is totally absent from our records throughout the ten years in December, Fic. 8,—Photo-micrograph of a phytoplankton consisting mainly of Thalasstosira Nordenskioldi. January, and February, and again in July and August, and is very poorly represented in several other months. In fact, it is only really abundant in April and May and not always then. In this distribution over the months of the decade this genus forms a marked contrast with Chetoceras, which is so constantly present. RHIZOSOLENIA. The species that occur most abundantly in our records are Rhizosolenia semispina, R. Shrubsolei, R. Stolterfothi and, less abundantly, R. setigera and R. alata (rarely). The two last named are temperate Atlantic forms, and R. semispina is a typically arctic oceanic species. They are all summer or autumn species, the genus being often quite unrepresented in certain of the winter months. It usually begins to appear in the nets in February and reaches its maximum in June. The numbers are sometimes enormous, giving the water in a glass jar a characteristic silky or fibrous appearance and causing the June crest that is usually present as a distinct elevation on the Diatom curve (see fig. 5, p. 182). The most abundant species in the Irish Sea is Rhizosolenia Shrubsoler Plymouth it is ®. Stolterfothi ; but R. semispina (fig. 9) is sometimes present in abundance at Port Erin in autumn, causing a second crest or maximum in September or October. In September 1907, thirteen millions and sixteen millions of this species were taken in two hauls of the surface-net at Station ITI. (three miles off Port Hrin), while the following year, at the same time and locality, it was almost absent, at 108 PROF. W. A. HERDMAN ON THE DISTRIBUTION We have had much larger numbers, of R. Shrubsolei, in June; and amongst our largest records of the genus taken as a whole are :—184 millions on May 30th, 178 millions on June 3rd, 59 millions on June 8th, 43 millions on June 13th, all in 1912, and 17 millions on June 21st, 1915. Fre. 9.—Photo-micrograph of a phytoplankton consisting mainly of Rhizosolenia semispina. The highest monthly average is in June in all of our years except 1907 when it was in May, and 1913 and 1916 when it was in July. The greatest average recorded is over 40 millions in June 1912. The autumn records are very irregular and sometimes show no rise whatever. Even when present it is slight compared with the June maximum (see fig. 5, p. 182). The highest monthly average in autumn is 117,122, in September 1912. GUINARDIA. Only the one species, Guinardia flaccida, occurs in our records (fig. 10). It is a summer form occurring mainly between April and July, with the maximum almost invariably in June. Our largest records are 22,800,000 on June 8rd, 1912; 18,000,000 on May 30th, 1912; and 8,773,000 on June 11th, 1910. We have several records of over 7,000,000 early in June, one of over two millions on July 8th, 1913, and one of over a million as late as July 16th, 1916. Guinardia, when present, makes up along with /hizosolenia the June elevation, which is sometimes so marked towards the end of the vernal Diatom maximum. This last summer (1917) it was unusually late, showing a maximum of five millions on 23rd July, along with one million of Rhizosolenia. The highest monthly averages are all in June with the exception of 1916, when it was in July. The greatest recorded average is close on six millions for June 1912. The lesser, secondary, increase in autumn has its highest OF DIATOMS AND COPEPODA IN THE IRISH SEA. 109 point in September—the greatest recorded monthly average being 10,967 for September 1912. It is interesting to notice that according to Pavillard Guinardia flaccida is abundant in the Gulf of Lyons as late as October. Fic, 10.—Guinardia flaccida, from a photo-micrograph by A. Scott. TABLE showing the monthly averages of the selected Diatoms throughout a year. 1912. | Biddulphia. Chetoceras. otk ee Rhizosolenia. peal Guinardia. | Lauderia. Jan..... 24,920 7,342| 9,877 | 22 0 257 189 Feb. ..| 36,885 | 10,301) 10,034 12 | 0 105 0 Mar. ..| 21,176 3,977,292 | 95,446 525 | 25 162 337 |Apr. ..| 21,869 | 18,365,750 | 100,619 36,464 721,338 3,158 |1,622,478 May ..| 1,072 | 2,844,861] 14,586 | 22,008,744 | 41,088 | 2,789,488 | 170,878 June ..| 0 67,543 0 | 40,883,771 0 | 5,919,429 July ..| 0 39,527 56 1,228 | 0 | 158 | 0 Aug. ..[ 9 11 694,961 0 13 0. O 0 Sept. ..| 4,131 7,702,658) = 1,677 117,122 | 3,978 | 10,967 | 8,944 Oct....:| 5,827 | 214,421] 11,914 | 820 57 | 2,176 | 29 Nov. ..| 25,714 9,476) 5,436 0 | 0 21 0 Dec. ... 8,059 | 1,106] 1,578 | 0 0 | M1 0 The above table shows very clearly how these important genera reach their maxima at different times, and how, for example, Chetoceras in spring (April) gives place to Rhizosolenia and Guinardia in early summer (May and June). Curves can be drawn mentally from a glance at the columns of figures which will demonstrate the waxing and waning of the several types, 110 PROF. W. A. HERDMAN ON THE DISTRIBUTION The following table showing the largest hauls of total Diatoms and of total Copepoda in each of the ten years brings out well how greatly the Diatoms outnumber the Copepoda—in some cases in the proportion of about a thousand to one. Diatoms. Copepoda. Largest haul. Date. Largest haul. Date. OO Tire 698,350 | April 6th 28,900 Aug. 17th GOST ee: 5,746,300 | May 28th 68,015 Sept..14th 1909 ee 10,358,300 April 22nd 71,010 | Oct. 18th | 1910) 5.3 70,128,400 April 22nd 119,650 Sept. 19th | TOU ead 69,982,500 May 16th 248,045 July 18th 1912 ....| 202,993,600 May 80th 223,789* May 20th 1913 ....) 205,814,700 May 16th 118,660 July 21st 1914 ....; 155,288,000 May 4th 217,571 Nov. 9th 1915... | 18,893,300 | June 15th 117,840 Aug. 2nd TSIGy a5 24,260,800 | May 25th 118,524 Sept. 25th OM 7rerer 64,339,250 | May 3rd | 147,706 | July 19th The relatively high records for Diatoms in 1912 and 1913 are due in the former case to the usual June maximum of hizosolenia and Guinardia, and in the latter case to a quite exceptional invasion of Asterionella japonica— the only occasion in our records when this genus has been abundant in the Trish Sea. We regard it as quite a rare form here. It is exceedingly abundant to the south of Iceland (Ove Paulsen) and also in the Gulf of Lyons (Pavillard). COPEPODA. The following table shows the total numbers recorded of our six.important species of Copepoda in the years 1909 and 1910 :— 1910. 1909. Oithona helgolandica ... 872,678 465,066 Pseudocalanus elongatus 308,326 309,973 Acartia elausi ee 340,631 63,373 Paracalanus parvus ... 217,633 54,120 Temora longicornis 386 147,043 62,659 Calanus jinmarchicus ... 15,418 21,412 * An exceptionally large haul at such an early date—due apparently to several moderately large swarms of different Copepoda occwring together, viz., Calanus, Pseudocalanus, Temora, and Acartia, Calanus are shown separately in figure 12. OF DIATOMS AND COPEPODA IN THE IRISH SEA. IAL The actual detailed numbers are of no importance except as indications of the relative abundance of the species. It is clear that Octhona and Pseudocalanus (in 1909) far outnumber the others. Acartia shows a notable increase in 1910. The Copepoda as a whole are a summer and autumn group, all the crests of their annual curves being found between May and October. Figure 11 shows curves of occurrence of the five most abundant of the above-named 8s 30,000 ny in ee est . Di at oma Ne 20,000 a i\ / ’ i ‘ jl i i t Ps i I 7 H 10.000 I fi 1 la] vt ine ; Vt ne. ! ve Avie D Ee fo ean eo Ce ore Dene 9 cain Send Me Fa pr ant hal a a a he at foe ! fz ant ml oa (aes Mav. Apr. May June July Aug. Sept. Oct. Nov. Dec. Fia. 11.—Curves of five most abundant Copepoda at Port Erin in 1912. species for the year 1912, which may be regarded as typical. Curves of The remaining one of our six selected species, Calanus finmarchicus, although much the largest individually and probably one of the most important from the fisheries point of view, only occasionally occurs in very large quantities in the Irish Sea, and its total table shows. numbers in a year are much below those of the other species, as the following LINN. JOURN.—ZOOLOGY, VOL. XXXIV. 9 PROF. W. A. HERDMAN ON THE DISTRIBUTION 99% S81 0g | S&F | PPG ors Gg ZS I6L 11 LLG “* nvq dod eSvieay 066693 | 1g0‘8T GIL'EL | Er sr esl'es | 63r'GL SF8'C ISF‘GL | SIF'ls | OFO'BT | cesIS | 1 smunpng Ih \aeerecs | ——. leSera= == oe a en | | ————— — ———— oe —— FEST | 8ou'T OLTT ZIT | G6S'T | SFI | £66 GFL. | 689 | 0&6 ocs‘T | * neq aed oder0ay | | | | | PIGOOST || FLUOFL | SIZ‘SIL | ZoZ‘9FT | FzO'FFT | BFS‘OIS’ | 6ss‘90L | SFO‘LZPT| 629'9 | G6r'0S | EGO'FOT|**°* °° °°" Blowar _—— | __ —_—| —_—— —_ = pue spejoy | | | | | 4 916°C || SPrT Osl'‘T 898'¢ | 290°6 gIZ'e 183'¢ relia | esF |= 086 ~—s| cee ~——s|** ynery aod oSuraay | | | | poesl'c | sep‘set | 26L‘6IL | 6c9's0¢ | Fos‘o6r | Tes‘s9s | SBO'TSe | E89°LTG| OGI'FS | ZPI‘I6 | LeL'ce |'°*"*” Snumpnoning | = eS See Sl ee ee 11% | oes | 090s | errs VGGF 009‘F go's =| ons"6 «| 999 | 8za'T peo eee qed oSvtaay gL9'F19% | eaz'scs | Ott'6oe | ose‘9es -| sos‘tor | szZ‘ocr | seo‘ece TegOre GLe'e9 ec6'ST | LIGEG fo BEwop eee'r 910'L pos’, | 920G | 0269 goe'9 | O6F‘E geu's | soz'e | Goel | czrT | [neq sed esuroay sricar'r | 6gc'619 | PLI'ss. | eseZer | og9'00g | gosta | eg6'coe | 9GE‘BE CLE;G0E GES'LLT) LEEELI |" * Smumpmopnesd = — —_—— | — | | | | IL1'8 F218 yge‘g | zoz‘ar = |.-«LOS‘OL =| 29201 C6201 ogg | eel = Szo'e— cogte =" [ney tod odwr0ay | | 116'T66'L || FET'6SS | 6EB'98 | SF6‘EST'T | cgc‘266 | SIz'eso'l | 8OT‘Sel‘T | 829°@Z8) 990°G9F| FEO‘TSz| GLL°ENS |” POYRO eee —— is —|- a ates | | —— ; = ee B1GR (|lsge- 06m | LOL, |G emia 86 ZO = | S260 i Gir G6 eo Le. S| ee SBN IOSON | | Ee eee | | ee | | | | | | | ‘savok OT | | | | | SER EREING 9T6T | “ST6T PIGT SI6T | ‘ZIGT | ‘TI6T | ‘OT6T ‘606T ‘806T “LOGT | | | | | 112 *sunali uaz ay) Ut pap.Lovae npodado,) paz0a]78 ay} Lo suaquine 1040} YF Pumoys ATA I, OF DIATOMS AND COPEPODA IN THE IRISH SEA. 113 CaLANts. Calanus sinmarchicus (Gunner) * is present throughout the year in the Trish Sea, and may be taken in small quantities at almost any time and in larger numbers on different occasions in different years. In 1907 it was fairly abundant (a few thousands in a haul) in April, and again at the end of July and the beginning of August, and in October. In 1908, on the other hand, the thousands per haul were not reached until later in the year, and the numbers remained relatively high from September to December (2,850 on December 23rd). In 1909, in addition to occasional thousands in May, 30 BOO Ba Calanus a Le \ —" Taam veut a Manama pr “Flay” Sune "Say “Tag” Sept "Ost Now "Lee Fic. 12.—Curves of Calanus for three years (1907-9) at Port Erin, showing successive waves of invasion, 2 or 3 in each summer, June, July, and October, we have evidence of an unusually large swarm that entered the bay on July 17th and 19th, when hauls estimated at 20,000 each were obtained. The average for these three years shows maxima in July and October (see fig. 12). In 1910 there were no very large hauls, and the thousands were only reached at the end of July and the beginning of August—perhaps the most usual time for swarms to appear in the bay. In 1911 the numbers in general were low, but two of the customary large swarms appeared in summer, one on July 4th and the other on July 18th. In 1912 the numbers were higher again, and thousands occurred on various dates from April to October. The very high record of 50,720 was obtained on May 17th, unusually early for such a large swarm. * T use this older specific name as I am inclined to agree with Wolfenden, Esterly and others that the characters used in the attempt to separate “ finmarchicus” and “helgolandicus” as species are too slight and inconstant to be relied upon. g* 114 PROF. W. A. HERDMAN ON THE DISTRIBUTION In 1913 the numbers were lower, and the maximum was 4000 on July 28th. In 1914 the maximum was 7320 on August 6th. In 1915 again the numbers were low, the highest being 3232 on July Ist. In 1916 a swarm appeared in the bay on July 11th, when over 12,000 were taken in a haul. A few days before a similar haul gave only 10 specimens, and a few days after 200. The average per haul at Port Erin over the ten years is only 266. On the whole this record agrees well with that at Plymouth, which is “Common on and off from the end of April to the beginning of November, generally present in small numbers at other times” (Miss Lebour). Fie. 13.—Calanus finmarchicus, from a photo-micrograph by A. Scott. Calanus jinmarchicus (fig. 13), as its specific name suggests, is a northern or Scandinavian form with a wide distribution through the colder waters of the North Atlantic. According to Ove Paulsen its home and centre of dis- tribution lies to the south of Iceland. In Loch Fyne on the west coast of Scotland, off Skate Island, Calanus appears to be present throughout the year in great quantities in deep water, at or near the bottom, along with MHucheta norvegica and Nyctiphanes norvegica. For example, in July 1907, off Skate Island in 104 fathoms, we caught 13,000 Calanus in one vertical haul, and on another occasion off East Loch Tarbert, in 76 fathoms, we got 10,000. We have also obtained OF DIATOMS AND COPEPODA IN THE IRISH SEA. 1LaLG} from the yacht at various localities in the deep lochs on the west coast of Scotland large hauls of Calanus by means of vertical hauls when no specimens were obtained at the surface. In fact this Copepod seems to be permanently present in the deeper waters of these lochs. In addition to these deep water specimens of Calanus we find on occasions, on the west coast of Scotland, large swarms on the surface, which may be the result of invasions from the ocean, and there is a considerable amount of evidence that such swarms are the cause of local mackerel fisheries. Some years ago (1909) Dr. E. J. Allen and Mr. G. E. Bullen published * some interesting work, from the Plymouth Marine Laboratory, demonstrating the connection between mackerel and Calanus and sunshine in the English Channel; and Farran f states that in the spring fishery on the West of Ireland the food of the mackerel is mainly composed of Calanus. In the summer of 1913 we had an experience at Tobermory, in Mull, which I shall quote from two letters written from the yacht at the time, and published in ‘ Nature.’ “S.Y. ‘Runa,’ Tobermory, July 12th, 1913. “On arriving in this bay last night we found that the local boats had been satching abundance of mackerel close to. We bought some for supper (good fish for a halfpenny each), and on dissection found that the stomachs of all of them were crammed full of fresh-looking Calanus (the individual Copepods being for the most part distinct and perfect), along with a few immature Nyetiphanes and larval Decapods. Professor Newstead and my daughter then noticed, while fishing over the side of the yacht, about 8 p.m., that the gulls in the bay were feeding in groups around patches of agitated water evidently caused by shoals of fish. On rowing out to these we saw distinctly the mackerel, large and small, darting about in great numbers in the clear water, and we also noticed every here and there on the smooth surface of the water—it was a beautifully calm evening—innumerable small whirls or circular marks which, looking closely; I found to be caused by large Copepoda close to the surface. “ About twenty years ago I sent a note to ‘ Nature,’ from the yacht * Argo,’ in regard to large Copepoda (I think it was Anomalocera on that occasion, and the locality was further north, off Skye), splashing on the surface so as to give the appearance of fine rain; and this present occurrence at once reminded me of the former occasion, but here the Copepod was Calanus finmarchicus of large size and in extraordinary abundance. They could be clearly seen with the eye on leaning over the side of the boat, a small glass collecting jar dipped at random into the water brought out twenty to thirty specimens at each dip, and a coarse grit-gauze tow-net of about 30 em. in diameter caught about 20 cubic centimetres of the Copepoda in five minutes. * Journ. Mar, Biol. Assoe. vol. viii. (1909) pp. 394-406. + Conseil Internat. Bull. Trimestr. 1902-8, Planktonique, p. 89. 116 PROF. W. A. HERDMAN ON THE DISTRIBUTION The mackerel were obviously darting about, occasionally leaping to the surface (which gave the gulls their opportunity) where the whirls, caused by the Copepoda, were thickest, and an examination of the stomach-contents of the fish on the yacht afterwards, showed us that the amount in one mackerel was about the same quantity as that caught by the tow-net in five minutes. Professor Newstead and I have made a count of 8 e.c. of the tow-net gathering, and estimate that it contains about 2400 specimens of Calanus. This would give about 6000 Copepods in the stomach of an average mackerel, or in a five-minutes’ haul of the tow-net, on this occasion. Tic. 14.—Photograph of exceptionally large hauls (about 1000 ¢.c. in a jar) of Calanus taken from the yacht ‘Runa’ in 1918 on the west coast of Scotland. The largest haul was estimated to contain at least » half a million individuals, “Tt may be added tbat these mackerel were evidently not being nourished in accordance with the views of Piitter, and were clearly able to fill their stomaclis from the plankton around them... .. The following note, written some weeks later, records the conclusion of the matter, so far as that summer’s observations went :— “S.-Y. ‘Runa,’ off Island of Higg, August 12th, 1913. “On getting back to Tobermory on Saturday, we found the plankton to be in marked contrast to its condition four weeks ago. The vast swarm of OF DIATOMS AND COPEPODA IN THE IRISH SEA. 117 Calanids has gone, and there are now no signs of mackerel feeding in the bay. In fact, the change has been noticeable for some days in the seas outside, and we have not been getting lately the large plankton catches that were usual in the latter half of July. On July 14th a haul of the large surface tow-net, in the open sea off Ardnamurchan, gave such a huge eateh of Calanus (about 1000 c.c.) that we promptly took a second similar haul, and had it cooked as a sort of potted ‘shrimp’ confection for tea (sumpled by ten persons, including the crew, who were much interested to try this new edible ‘ fish’); while on August 11th a haul of the same net, taken at the same spot, gave only a small catch of some 15 c.c., containing very few Calanids, along with the usual scanty summer zooplankton.” The importance of Calanus as a food for migratory pelagic fishes such as the mackerel cannot be doubted (see fig. 14). PSEUDOCALANUS. Pseudocalanus elongatus (Boeck) is widely distributed in the North Atlantic and is present in the [rish Sea throughout the year (fig. 15). It reaches its lowest level in January and February, and bas its maximum in late summer and autumn (June to October in our records, generally September or October). } ; S e e Fia. 15, Pseudocalanus elongatus, from 2 photo-micrograph by A. Scott. This Copepod comes next after Oithona as the second most abundant species in the Irish Sea, its average per haul over the ten years being 4583. Some of the most important records are :—65,200 on May 20th, 1912; 91,960 on October 21st, 1912; 68,120 on June 24th, 1913 ; 60,600 on May 5th, 1916 ; 59,600 on September 25th, 1916; 58,200 on October dist, 1912; 54,350 on July 31st, 1911; and many other records between 35,000 and 50,000 in these same months. 118 PROF. W. A. HERDMAN ON THE DISTRIBUTION PARACALANUS. Paracalanus parvus (Claus) is a southern form (fig. 16) which extends to the tropics and has a wide distribution. Jt is present in quantity only during autumn and winter in the Irish Sea, being practically absent between February and July, and having its maximum in September or October. The largest hauls are in autumn, for example, 59,460 on September 7th ; 138,300 on September 10th; 53,500 on September 15th, all in 1914; 61,930 on August 24th, 1911; 49,800 on September 8th, 1910; 49,390 on September 7th, 1912; 44,060 on October 24th, 1912; while other large hauls in the same months range from 25,000 to 40,000. ® Fre. 16.—Paracalanus parvus, from a photo-micrograph by A. Scott. This is a species that has varied considerably in quantity from year to year, its average per haul ranging from 325 in 1907 to 5863 in 1914. In the present year, 1917, it was especially abundant at Port Erin in September, along with Oithona. OITHONA. Oithona helgolandica (=O. similis), Claus, is the most generally abundant Copepod throughout the year in the Irish Sea; but the months when it is takey in greatest numbers are June to November (inclusive) with the maximum generally in July. There may, however, be a second maximum later, in October or November. The following may be quoted as exceptionally large hauls :—The highest, 225,450, is on July 18th, 1911, and the next highest, 199,300, is on November 9th, 1914; then we have 93,580 on July 28th, 1913, 87,530 on September 26th, 1912, and 83,550 on June 14th, 1910. We have twenty other records of over 44,000 each scattered over the months June to November. This is the Copepod which has the highest average number per OF DIATOMS AND COPEPODA IN THE IRISH SEA. Aai9 haul over the ten years, viz., 8171 (see table, p. 190). Although this isa small species (fig. 17), still from its very great abundance it must be of considerable economic importance as a food-matter in the sea. Fie. 17.—Oithona helgolandica, from a photo-micrograph by A. Scott, ACARTIA. Acartia clausi, Giesbr., is a summer species widely distributed over the North Atlantic, and present in greatest abundance in the Irish Sea from May to October (inclusive), and having a maximum generally in June, with another slighter rise in autumn, September or October. Fie. 18.— Acartia clausi, from a photo-micrograph by A. Scott. The following are some of our highest records :—59,490 on June 27th, 1912 ; 59,360 on August 24th, 1911 ; 52,200 on May 20th, 1912: 50,100 on June 2nd, 1913 ; 44,000 on September 28th, 1913 ; and 41,950 on June 17th, 1912. The average of this species per haul throughout the ten years 120 PROF. W. A. HERDMAN ON THE DISTRIBUTION is 2571. Swarms of this oceanic species (fig. 18) in summer may possibly be connected with an inflow of Atlantic water ; but probably also large numbers appearing locally in enclosed areas may be the result of reproduction and development i sitw. In some eases Acartia has been found in unusual abundance quite close to the shore, and even in water of low salinity. Towards the end of September, 1917, we emptied for cleaning purposes the large open-air fish-ponds at the Port Erin Biological Station, and when most of the water had been pumped out we noticed that what remained was swarming with small Copepoda. Every bucket and jarful that one removed as densely, in fact, as an for inspection was found to be densely crowded average jar of plankton representing the contents of a standard haul. Some buckets of these Copepoda were used for feeding the lobster larvee and some of the smaller fish and other animals in the Aquarium, and a small sample that was preserved was found to bea practically pure gathering of Acartia clausi. One specimen of Centropages was present in the tube amongst many thousands of Acartia, when Mr. Scott took the photo-micrograph shown in fig. 18. TEMORA. Temora longicornis (Miill.) is a very local species, which may be found on occasions swarming in great abundance in small areas of the sea, generally close to land. It is a summer species and its maximum is generally in June Fig. 19.—Temora longicornis from the stomach of a Mackerel. From a photo-micrograph by A. Scott. or July, but was in May in 1912, when the large Lauls of 50,400 and 83,400 were taken on the 17th and 20th respectively. Other large hauls of this OF DIATOMS AND COPEPODA IN THE IRISH SBA. aah species in our records are :—45,530 on July 11th, 1916 ; 43,800 on June 11th, 1913; and 34,665 on July 26th, 1910. The average for this species per haul throughout the ten years is 1234, Temora (fig. 19) is characteristic of coastal as opposed to Atlantic water, and is the only “neritic” form in our series. ‘The other five are all usually classed as ‘‘oceanic,” but as I have pointed out above most of them are really present throughout the year in the Irish Sea. Temora longicornis is on oceasions one of the most abundant of our Copepoda in the Irish Sea and must be of considerable importance as a food for fish and especially for the herring in summer. In the latter part of July and the first half of August, 1917, the shoals of herring to the west of the Isle of Man came unusually close to land and even penetrated into bays and creeks ; and during this time they were feeding mainly, if not wholly, on Fig. 20.—Temora longicornis from the “red patches.” From a photo-micrograph by A, Scott. Temora. Late in July this Copepod was so abundant that its presence caused large patches of a red colour on the surface of the sea off Port Brin and around the Calf Island. These red patches were noticed by the fishermen, and were spoken of amongst them as being “ fish-food” or “spawn.” A large jarful from such a red patch, obtained by one of the fishermen, was brought to the laboratory and found to be swarming with small Copepoda, which on examination proved to be almost wholly Temora longicornis (fig. 20). About one-fourth part of the contents of the jar was preserved, and on being counted later on was found by Mr. Andrew Scott to amount to 50 c.c. of Copepoda, consisting of 33,340 Temora and 2 Calanus. Mr. Scott estimated the oil present in 9 c.c. of the dried Temora at 2°47 per cent. of the weight, which was 0°925 gramme. During this same time the men were catching herring in quantity unusually close inshore in the neighbourhood of the red patches, and on examining, in 122 PROF. W. A. HERDMAN ON THE DISTRIBUTION the laboratory at the Biological Station, the stomach-contents of a number of these herrings, I found in every case that the stomach contained a mass of red material which was obviously, under the microscope, the broken-down remains of Copepoda. A few Crab zoea were recognisable, but the bulk of the material consisted undoubtedly of the Copepoda. Mr. Scott examined 5 c.c. of the stomach-contents for me, and found that it contained 975 easily recognisable specimens of 7emora. A photograph (fig. 21), which Mr. Scott t Fie. 21.—Temora remains from the stomachs of the Herring. From a photo-micrograph by A. Scott. has made from one of the microscopic preparations, shows appendages that undoubtedly belong to this Copepod, while here and there in the stomachs complete specimens of Temora are to be seen. It is not possible to doubt that during these weeks, at the height of the summer herring fishery in the Irish Sea, the fish were feeding mainly upon this species of Copepod. We recorded a similar occurrence off the Lancashire coast a few years ago, when in July 1913, at the time of an abundant mackerel fishery off Walney Island, the stomachs of some of the fish were found to be full either of Temora alone or of Temora mixed with Jsias and a few other Copepoda (see fig. 19, p. 198). A few herrings from the Port Erin fishery of Juiy 1916 were found by Mr. Scott to be feeding mainly on Calanus. CONCLUDING REMARKS. Many food-fishes are known to feed upon Copepod plankton during at least some portion of their life. The Loch Fyne herrings are frequently at the time of a fishery found to have their stomachs filled with Mucheta or Calanus. Mackerel, in the English Channel and to the S.W. of Ireland and elsewhere, have been recorded as feeding on Calanus. It has been shown in this paper that in Hebridean Seas the mackerel and in the Irish Sea herrings, OF DIATOMS AND COPEPODA IN THE IRISH SEA. 123 at a time when they are present in great abundance, are feeding on some prevalent form of Copepod, such as Calanus or Temora. Other similar cases could no doubt be quoted and are known to marine biologists. Then, as to demersal fish—young plaice, after their metamorphosis, feed chiefly on the smaller Copepoda, while in younger stages the post-larval plaice feeds upon Diatoms. We have noticed at the Port Krin Biological Station the post-larval plaice with its stomach showing of a golden brow n colour from the Diatoms with which it was filled, and we ieee watched in a shallow pond the metamorphosed young plaice darting backwards and for- wards pursuing, catching, and devouring the individual Copepoda. It is known that these Copepoda in their turn feed in part at any rate on Diatoms, so our two main constituents of the plankton are undoubtedly concerned in the nourishment of either young or adult fishes useful to man. The association of shoals of fish with abundance of plankton is the result of the fact that, in order to get an adequate quantity of planktonic food, the fish must seek out and capture the Copepoda. In other words, the fish must yo where the plankton is abundant and must in its moyements follow the movements of the shoals of Copepoda. It is the very poverty of the plankton in some sea-areas, insisted on by Pitter, Lohmann and others, which makes it necessary for plankton-eating fish to move about in search of more abundant supplies. Consequently it is of importance to show, as we now can, that in our coastal seas at least, where the fisheries we are interested in take place, the plankton is not uniformly distributed. Many of the Copepoda oceur very definitely in local swarms, and various localities and depths are characterised at the different seasons by purticular assemblages of plankton. It is therefore reasonable to believe, in view of the facts given above as to the association of fish and plankton, that these variations in the distribution must have a marked effect upon the presence and abundance of at least such migratory fish as herring and mackerel, and also of the shoals of post- larval young of many of our other food-tishes. No less than three masses of sea-water of different origin ent character may enter or affect the British seas in varying quantity, viz.:—(1) Arctic water such as normally surrounds Iceland and the east of Greenland, and may extend further southwards and eastwards towards Norway, the Faroes, and Shetlands ; (2) Atlantic (“ Gulf-stream Drift”) water which impinges on the western shores of Ireland, and may flood the English Channel and extend round the Shetlands or down into the North Sea ; and (3) “ Coastal” water such as flows out from the Baltic and, mixed with the other waters, bathes the coasts of N.W. Europe generally, and to a large extent surrounds the British Islands. The Irish Sea may be regarded as primarily an area of coastal water, which is, however, liable to be periodically invaded to a greater or less extent 124 PROF. W. A. HERDMAN ON THE DISTRIBUTION by bodies of warmer and salter Atlantic water (re-inforced possibly by por- tions of a deeper outflowing Mediterranean current) carrying in oceanic plankton, and more rarely perhaps by Norwegian or Arctic water causing an invasion of northern organisms. The variations which we find in different years in the nature and amount of the plankton at the same localities no doubt depend to some extent upon the volume and period of such southern or northern invasions ; but they may depend also upon other factors, such as the weather (temperature, sunshine, rainfall, wind, etc.) at the time, and previously. Of the six Copepoda discussed above only one—7emora—is a neritic form ; the others are all usually regarded as oceanic, that is as having their true home and centre of distribution somewhere to the north, west, or south in the open Atlantic. The following list gives an approximate indication of what is supposed to be the source of these five oceanic Copepoda :— Calarus.—N. Atlantic, about Iceland (‘“ Boreal oceanic’’). Paracalanus.—Southern, temperate and tropical Atlantic. Pseudocalanus.—N. Atlantic (“‘ Boreal oceanic”). Oithona.—N. Atlantic ( Boreal oceanic ”’). Acartia.—N. Atlantic (“ Northern styli-plankton ”’). Some no doubt live on during the year in the Irish Sea, but these indi- genous populations are probably reinforced by waves of immigration from outside. In the case of our Diatoms some of the species of a genus may be neritic and others oceanic, as is shown in the following list, where (N.) stands for neritic and (O.) for oceanic, and a (?) indicates that the evidence seems to me conflicting or inconclusive *:— Biddulphia mobiliensis (N.), B. sinensis (O.). Coscinodiscus radiatus (O. ?), C. concinnus (N.), C. Grani (N.). Lauderia borealis (N.). Chetoceras boreale (O.), C. criephilum (O.), C. decipiens (O?), C. densum (0.), C. contortum (N.), C. debile (N.), C. diadema (N.), C. sociale (N.), C. subtile (N.), C. teres (N.). Thalassiosira gravida (N.), 7. Nordenskioldi (N.). Rhizosolenia alata (O.), R. semispina (O.), Ef. setigera (N.), R. Shrubsolet (N.?), BR. Stolterfothi (N. ?). Guinardia flaccida (N. ?). It is remarkable how small a number of forms make up the bulk of the macro-plankton throughout the year. These half-dozen kinds of Diatoms and half-dozen Copepoda are the all-important organisms upon which our * The matter was discussed more fully some years ago in a paper by Herdman and Riddell in Trans. Biol. Soc. Liverpool, xxv. (1911) p. 178, OF DIATOMS AND COPEPODA IN THE IRISH SEA. 125 fate depends so far as concerns food from the sea, That shoals of migratory fish such as herrings and mackerel are attracted in summer and autumn by dense swarms of certain Copepoda such as Calanus and Temora can no longer be doubted ; and there is abundance of evidence that earlier in the year the young stages of other food-fishes, such as plaice, are nourished first by Diatoms and when larger by the smaller Copepoda. These conclusions, however, do not finish the matter. There remains in the sea the much more minute Protozoa and Protophyta of the Nannoplankton, which to a large extent escape through the meshes of our plankton nets and which are doubt- less of great importance as the food of Copepoda and other large organisms which in their turn nourish fish. There are several other interesting lines of further investigation which open up as collaterals from plankton investigation. One of these is the origin of the great vernal phytoplankton maximum. In the spring there is an awakening of the plant-life of the sea comparable to the growing of the grass and the budding of the trees on land. This cannot be due to any rise in temperature, as the sea at the time that the Diatoms start their active growth and reproduction is at its coldest. We have series of observations extending over more than ten years showing that the water of Port Erin bay is generally of lower temperature in March than in December or January. Of the various causes for the rise in the Diatom curve in these cold waters of early spring I have for some years* regarded with greatest favour the view suggested first by Sir John Murray, viz., the increase in sunlight at that time of year. In that case it would be a photosynthetic phenomenon— the increased solar light energy enabling the Diatoms to obtain from their environment by photosynthesis the materials required for their growth and reproduction. The view that the spring increase in plankton is due to changes in the alkalinity of the water does not in my opinion conflict with the photo- synthesis theory but is supplementary to it. The position in regard to the relation between variations in alkalinity and in the plankton, in our district, is as follows :— The sea around the Isle of Man is a good deal more alkaline in spring (say April) than it is in summer (say July); and during the years 1912-14 Professor Benjamin Moore, by examining samples of sea-water periodically at the Port Erin Biological Station, was able to show + that the alkalinity, which gets low in summer, increases somewhat in autumn, and then decreases rapidly, to disappear during the winter; and then once more, after several months of a minimum, begins to come into evidence again in March and rapidly rises to its maximum in April or May. This periodic change in alkalinity will be seen to correspond roughly with the changes in the living * See “Intensive Study, &c.” Part IIL. p. 260, 1910, Trans. Biol. Soc. Liverpool, xxiv, + Trans. Biol. Soc. Liverpool, xxix. (1915) p. 233, {26 DISTRIBUTION OF DIATOMS AND COPEPODA IN THE IRISH SEA. microscopic contents of the sea represented by the phytoplankton annual curve, and the connection between the two will be seen when we realise that the alkalinity of the sea is due to the relative absence of carbon dioxide. In early spring, then, the developing myriads of Diatoms in their metabolic processes gradually use up the store of CO, accumulated during the winter and so increase the alkalinity of the water, till the maximum of alkalinity, due to the reduction in amount of carbon dioxide, corresponds with the crest of the phytoplankton curve in, say, April. Prof. Moore has calculated that the annual turn-over in the form of carbon which is used up or converted from the inorganic into an organic form probably amounts to something of the order of 20,000 or 30,000 tons of carbon per cubic mile of sea-water in the Irish Sea; and this probably means a production each season of about two tons of dry organic matter, corresponding to at least ten tons of moist vegetation, per acre—which shows that we are still very far from getting from our seas anything like the amount of possible food-matters that are produced annually. Testing the alkalinity of the sea-water may therefore be said to be merely ascertaining and measuring the results of the photosynthetic activity of the great phytoplankton rise in spring due to the daily increase of sunlight, Other possible causes, more or less related to the above, have been suggested —such as Brandt’s hypothesis that the fluctuations in the phytoplankton depend upon the accumulation, and then the exhaustion, of necessary inorganic food-matters in the water, such as nitrogen or phosphorus com- pounds or silica ; and the view of Nathansohn, Gran and others that vertical currents, carrying up food-matters from the deeper water, have a powerful effect upon the seasonal development of surface plankton. These may be contributory causes or may be effective locally, or on occasions; but it seems probable that a widespread phenomenon of enormous amount such as the vernal increase of phytoplankton must depend upon an equally widespread and powerfully-acting cause such as the rapid increase in the amount of solar light energy which marks the lengthening days of the year in early spring. SLADEN TRUST EXPEDITION TO ABROLHOS ISLANDS. 127 THE PERCY SLADEN TRUST EXPEDITIONS TO THE ABROLHOS ISLANDS (Inpran OcEAN). Under the Leadership of Prof. W. J. Daxty, F.L.S., F.Z.S. Report I,—Introduction, General description of the Coral Islands forming the Houtman Abrolhos Group, the Formation of the [slands. By W.J. Daxiy, D.Sc., F.L.S., Professor of Biology, University of West Australia. (PLaTEs 10-14, and 12 Text-figures.) [Read Ist February, 1917.] INTRODUCTION. SHortLy after my arrival in Western Australia in 1913 my attention was drawn to the interesting position of certain coral islands known as the Houtman’s Abrolhos Islands. I determined to visit the group at the earliest possible date, and to this end applied to the Trustees of the Percy Sladen Trust for a erant in aid. An expedition was planned and left the Australian Coast in November 1913 for the islands. On this occasion I was fortunate in having a very able colleague in Mr. W. B. Alexander, M.A., of the West Australian Museum, and the grant of the Perey Sladen Trustees was supplemented by aid from the University of West Australia and the West Australian Museum. Part of our equipment was conveyed direct to the islands from Fremantle, but the personnel with the rest of our apparatus and stores, ete., travelled by train to the port of Geraldton, where the fishing lugeer ‘Queen,’ a boat of 22 tons, awaited us. The ‘Queen’ is one of a fleet of fishing-boats which makes Geraldton its headquarters, and fishes the waters round the Abrolhos Islands, and as far north as Shark’s Bay. Geraldton is a port of about 3500 inhabitants, situated on the coast of Western Australia in latitude 28° 46'S. Although only a small town it ranks as the second or third port in Western Australia, and is the centre for a large agricultural and pastoral area, as well as the Murchison goldfield. It marks the most northerly point on the coast which can be reached by train from Perth, and it will be seen on reference to a map that a great length of the coastline of Western Australia remains to the North and can only be visited with difficulty. There are a few ports which are reached by regular steamers from Perth, but between these places the coast might almost still be termed unexplored ; this certainly holds good, from the biological point of view, for the entire coast north of Geraldton with the exception of Shark’s Bay. The Abrolhos Expedition of 1913 extended over a period of three weeks, during which time we worked the ‘Queen’ from islet to islet, and combined shore collecting with biological and geological observations on the islands, and some dredging in the lagoons and between the island LINN, JOURN.—ZOOLOGY, VOL. XXXIV, : 10 128 PROF. W. J. DAKIN: EXPEDITION TO THE groups. Extensive collections resulted and many observations of interest and importance were made. As a result it was felt highly desirous that a supplementary expedition should be arranged as part of a scheme embodying the investigation of the North-Western Coast. Fortunately the authorities at home came once more to the rescue, and with the aid of a further grant from the Percy Sladen Trustees, together with help from the Royal Society Grant Committee and the British Association, a second expedition was Text-figure 1. HOUTMAN eSegrpe Baek ROCKS See aMid Reef Champion Bay Hummack 1. Lang 14° East from Greenwich General Map of the Houtman Abrolhos Islands. made possible. On this occasion I was assisted by my colleague, Mr. A. Oayzer, B.Sc., Assistant Lecturer in Biology, University of West Australia. The fishing lugger ‘Ada’ was chartered and we left Geraldton for the Abrolhos in October 1915, four weeks being spent amongst the islands. I was fortunate in having more hydrographic equipment at my disposal on the second expedition, and a rowboat fitted with a motor was of invaluable service in connection with work in the lagoons, I can heartily recommend HOUTMAN ABROLHOS ISLANDS. 129 this very cheap type of motor craft for use in sheltered waters such as those of the lagoons. It was even possible to dredge with a small instrument 2 feet long and weighing, without sinkers, 12 Ibs., in a most satisfactory manner. Much more time was spent in dredging on the second expedition, the small dredge being worked from the rowboat in the lagoons, whilst an Agassiz trawl was used from the lugger in the more open waters. On both expeditions the skippers and crew were Scandinavians and they helped enthusiastically in the work. On both occasions we took with us camp equipment. The working plan was as follows. The lugger was sailed to a convenient and_ sheltered anchorage where we intended to work, and usually quite close to an islet (large or very small as the case might be). A camp was then set up on shore, where my colleagues and myself slept, whilst the crew remained on board. Two or three days might be put in at one anchorage in this way, or we might shift lugger and camp every day according to the work carried out. On some occasions the shore collecting, surveying, and work in the lagoons with the dinghy, occupied all our efforts, so that the lugger was left undisturbed for several days. On other occasions when the weather was suitable for dredging in the open sea, we would start early in the morning and leave our camp on shore, to return in the afternoon with the spoil, which needed sorting, labelling, and preserving. One of the great difficulties in connection with work at the Abrolhos Islands is to find the right kind of weather. For the greater part of the year the winds and sea are too unruly for continued work in these waters, except from a large vessel, and a large vessel could not be used amidst the coral islets. It appears that the best times are between the two well-marked seasons, Summer and Winter (dry and wet seasons), 7.e. the months of October and November on the one hand and April and May on the other. Both of our expeditions were carried out between winter and summer, but we were not nearly so fortunate on the second occasion, although longer at the islands. On one occasion we hung with two anchors out whilst a gale blew for twenty-four hours ; a nasty coral-reef, just submerged, was situated ouly about twenty feet behind our stern. No accidents happened, but our work was delayed for several days owing to the bad weather. On another oceasion, however, the ocean swell was so reduced to leeward of one of the islands that we were able to follow the outer reef and examine it from the motor dinghy. Reversing thermometers were used on the second expedition in conjunction with the Ekmann water-bottle, but we were too short of workers to initiate a scheme for the complete oceanographical investigation of this region and were forced to be as economical as possible. Our plankton hauls were made chiefly with the object of comparing the plankton of the lagoons with that of the ocean outside, and for this purpose we usually aimed at taking hauls in Oe 130 PROF. W. J. DAKIN: EXPEDITION TO THE both places on the same day, and about the same time. The catches have not yet been examined in detail, but it was fairly obvious, each time, that the plankton of the lagoons contained large quantities of dead organisms and much debris. On the second expedition blasting gelignite was tried as an experiment for the capture of fish. A detonator was pushed into a stick of gelignite, and, after lighting the fuse, the whole thing was thrown overboard. It only worked satisfactorily over the reefs in moderately shallow water, but we could often throw it amidst shoals of coral-haunting fish in these regions. The burning fuse appeared to attract the fish nearer to the explosive rather than frighten them away. After the explosion, numerous fish would appear on the surface. They were not usually dead, but swimming just as if the gas- bladder were distended. Some species did not rise at all, and it would appear that the ascent of the fish immediately after the explosion is not due to death, but to physiological conditions following the shock. Another experiment was the use of chloride of lime as a poison in rock- pools. This was suggested to me by Professor Starr Jordan and worked very well, but owing to the tides being rather poor and the wind high on many days when we could have utilized this, it was only tested on a few occasions. The Houtman’s Abrolhos le about 40 miles away from the coast of Western Australia, on the very edge of the continental shelf in latitude 28° 40' S. (see map, text-fig. 1). The depth of water over the shelf between the Abrolhos and the mainland is very uniform and averages about 25 fathoms, whilst the same depth occurs between the different groups of islets. A few miles west of them the sounding-line gives depths of hundreds of fathoms. Unfortunately, the Abrolhos Islands are poorly charted and small maps on a reasonable scale deal only with two of them, together with a small area of lagoon in each case which was probably considered to be well sheltered and providing good anchorage. There are no lights of any kind on the islets, and in consequence the fishermen, who have spent years about this area, will not approach close when the sun has set. As a matter of fact, navigation in the lagoons and about the small islets is only safe when the sun is high, and a hand stationed in the rigging can spot the submerged coral- reefs which oceur here and there. Coastal steamers give the Abrolhos a wide berth, but more than one ship has finished her life on these barren reefs. In fact, the history of the islands almost commences with a wreck (1629) and the coral reef of the Pelsart Island is to-day dotted with the scattered remains of a steamer, the ‘ Windsor,’ which became a total wreck a few years ago. Islands in such a position as the Abrolhos—in close proximity to a continent—are full of interest to the biologist, and this is especially the case when they are coral islands, It is not surprising then to find that they HOUTMAN ABROLHOS ISLANDS. 131 have already been visited by one or two collectors. They were discovered by the Dutch navigator Houtman in the year 1619, but the name Abrolhos is a contraction of a Portuguese phrase meaning “keep your eyes open.” They are now more frequently called the Abrolhos Islands. The West coast of Australia was known before this date, and Saville Kent is quite incorrect in stating that it was Pelsart’s discovery of the Abrolhos (1629) that led to the “ earliest recorded discovery of the great island-continent of Australia.” The wreck of Pelsart’s ship (one of the Dutch East Indian Co.’s vessels) in 1629, led to one of the most romantic episodes in the early history of West Australian exploration. Pelsart in command of the ‘Batavia’ left Texel on October 28th, 1628, for the East Indies in company with several other ships, all of which were equipped and commissioned by the Directors of the Dutch Hast India Company. The ships became separated and the ‘ Batavia,’ continuing her course alone, got among the coral reefs of the Abrolhos Islands and struck before sunrise on June 4th, 1629. Trouble followed thick and fast, and the tribulation of the passengers was accentuated by the drunkenness and disorderly conduct of the soldiers and sailors. The ship’s company were, however, eventually landed on two of the small islands of the lagoon in the Pelsart Atoll, and much of the provisions and treasure was also salved. Unfortunately, there was a great lack of water, and as we have already seen, little or none is to be obtained on these islands. This trouble accentuated the spirit of unrest which had broken out, and Pelsart finally decided to set out and seek for water. A section of his crew appeared to he very ripe for mutiny. For a few days the islands were investigated for water with little success, and then Pelsart set off to the mainland. The mainland, however, seemed to offer no better prospects, and for some time the men were unable to land owing to the surf. Drifting northwards some of the crew eventually made a landing by swimming ashore, but an inhospitable country dispelled all hopes, and ultimately Pelsart decided to make for Batavia in order to seek assistance and report to the Governor the misfortunes which had befallen them. He reached that place safely on July 5th, after a Journey of upwards of 1700 miles in an open boat. Tragic events had in the meantime taken place at the Abrolhos Islands. The Supercargo of Pelsart’s ship, a thorough villain, named Cornelis, resolved to take up the life of a pirate on the high seas. To this end, he and his accomplices determined to murder in cold blood all the ship’s company (220 souls) with the exception of about 388, and then capture the rescue vessel which Pelsart was expected to bring back with him. By this time the passengers, soldiers, and crew were occupying three of the lagoon islands, some of them having moved to the third in search of water. Cornelis and his company were on the largest island. The murders were carried out on 132 ' PROF. W. J. DAKIN: EXPEDITION TO THE two of the islands, only a few boys and some women being spared. A few men escaped, however, from this hell and reached the third island with the news, whereupon the party (now forty-five in all) led by a man named Weybehays resolved to defend themselves. Cornelis, who had assumed the title of Captain-General, sent two expeditions against them, but these being defeated le determined to gain his ends by more subtle means. Unfortunately for himself, he was hoist with his own petard, taken prisoner, and some of his men were killed. The rest of the mutineers remained on their islands awaiting the arrival of Pelsart, who had been given assistance and a ship at Batavia. He reached the Abrolhos on September 13, eagerly awaited by the two opposing parties, but four of the third island defenders managed to reach him first and acquainted him with the sad state of affairs. When the mutineers arrived to capture the ship, Pelsart and his men were ready for them. The boarding party was captured, and the rest of the ruffians on their island experienced a like fate. After an investigation of the tragedy, Cornelis and his associates were tortured and put to death—they had murdered over 120 innocent souls,—two other prisoners were marooned on the coast, the survivors of the wreck and mutiny eventually reaching Batavia in safety. The disaster to Pelsart’s ship was not the only one experienced by the Dutch at the Abrolhos—for which reefs their ships seemed to have an unfortunate liking. The ‘ Luytddorf’ is supposed to have been lost here in 1711, whilst in 1727 the ‘ Zeewyk’ was wrecked on the western reef of the Pelsart Atoll. Pelsart described some of the forms of life met with on the islands and in so doing gave a most interesting account of the wallaby (Adacropus eugenii, Desm.) which still abounds on two of the islands. This was one of the first written descriptions of a member of the kangaroo family to be given to the world. The diagnosis runs as follows :—“ Besides we found in these islands large numbers of a species of cats which are very strange creatures. They are about the size of a hare, the head resembling the head of a civet cat, the fore paws are very short, about the length of a finger, on which the animal has five small nails or fingers resembling those of a monkey’s fore paw. ‘The two hind legs on the contrary are upwards of half an ell in length, and it walks on these only on the flat of the heavy part of the leg so that it does not run very fast. Its tail is very long like that of a long-tailed monkey. If it eat it sits on its hind legs and touches its food with its fore paws just like a squirrel or monkey. “Their manner of generation or procreation is exceedingly strange and highly worth observing. Below the belly the female carries a pouch into which you may put your hand. Inside this pouch are her nipples ; we have found that the young ones grow up in this pouch with the nipples in their mouths. We have seen some young ones lying there which were only the HOUTMAN ABROLHOS ISLANDS. 133 size of a bean, though at the same time perfectly proportioned, so that it seems certain that they grow there out of the nipples of the mamme from which they draw their food until they are grown up and ready to walk. Still, they keep creeping into the pouch even when they have become very large and the dam runs off with them when they are hunted.” It is worthy of note that the Abrolhos Islands were charted by H.M.S. ‘Beagle’ under Wickham and Stokes a few years after Darwin’s famous voyage in that ship. The vessel was some time at the islands and obser- vations were made on the tides and specimens of the fauna were collected. Most of the names of the islets and passages were given at this time. Darwin himself refers to the Abrolhos Islands in his famous work on Coral Reefs. He had, however, very little information concerning them, beyond that reported by the surveying ships. The only serious biological investigations of the Abrolhos prior to our visits were made by naturalists who were attracted by the bird-life. One marine biologist, however, had visited the islands, namely Saville Kent. Whilst making a survey for the Government of Western Australia in 1894, this keen naturalist paid a flying visit to the Abrolhos. He was evidently impressed by them, for in the preface to his work ‘The Naturalist in Australia,’ he writes (referring to the development of the scientific poten- tialities of the country) :—‘‘ As an indication of the leading position Western Australia is eligible to oceupy with relation to one important biological subject, reference may be made to that chapter which deals with the Houtman’s Abrolhos. As there demonstrated, very exceptional facilities prevail at that place for the conduct of reef-boring operations and for the prosecution of all methods of investigation relating to coral and coral-life.” A considerable number of terrestrial vertebrates, particularly lizards, were obtained by Gilbert, who visited the islands on behalf of Gould the ornitho- logist. These specimens have been described in the British Museum Catalogues on the Lizards, Snakes, and Batrachia respectively. After Gilbert’s collecting excursion, the next naturalist to visit the Abrolhos was A. J. Campbell, F.L.8. A paper on the zoology of these islands by this worker appeared in the Reports of the Australian Association for the Advancement of Science, Melbourne, 1890. Campbell visited this group in December, 1889. Beyond a few remarks about the Mammals and Reptiles he deals only with the bird-life. Campbell states that the Abrolhos form the greatest “rookery for sea-birds in Australia, and by reason of their geographical position in the sub-tropics, perhaps afford suitable breeding grounds for a greater number of species than any other distinct or limited spot in the world.” The next paper to appear dealing with the Abrolhos was a short article by Helms, who visited the islands in 1898, This author names several plants, speaks in a general way of the corals and fishes, and devotes most attention 134 PROF. W. J. DAKIN : EXPEDITION TO THE again to the birds. Another paper on the bird-life was written by Chas. G. Gibson and published in the ‘ Emu,’ 1908-1909. It will be seen that whilst the mammals, birds, reptiles, and amphibia have been referred to by three or four visitors to the Abrolhos, nothing in any detail has been published on the marine fauna beyond the work of Saville Kent. Kent considered that the marine fauna of the Abrolhos was essentially a blend of tropical and temperate species. ‘The tropical species were not to be found on the adjacent coast. Rich coral growths were to be seen every where (the islands were stated to be almost entirely composed of coral), but the living species were not the extra-tropical forms to be seen on the coast in this latitude, but were tropical types. The Percy Sladen Trust Expeditions to the Abrolhos Islands in the years 1913 and 1915 have been conducted, in the first place, for an investigation - into the structure and formation of these coral islets. and in the second place for the purpose of collecting information regarding the fauna and marine flora of this region of the Indian Ocean and West Australian coast. In connection with the latter aim we were particularly interested in the statements made by certain authors (Saville Kent, and Michaelsen of the German Expedition to South-West Australia) that the marine fauna of the Abrolhos was entirely different from that of the coast in the same latitude and only 40 to 50 miles away. We have collected some data with regard to the Hydrography of these regions, but it is extraordinary how little is known even of sea temperatures on the Australian coast, while numerous problems await investigation in connection with the tidal phenomena. It is a pity that the means of communication along the West - Australian coast should be difficult and costly, but probably in the future, when this large State has its present vast spaces occupied by energetic emigrants, these difficulties will disappear. The time may even come when the scientific investigation of these coastal waters, with their valuable products of pearl-shell and fish, will be recognised as of just a little importance. The present paper is concerned with a genera! description of the Abrolhos Islands and includes an account of their structure and formation. Other papers will follow dealing with the collections obtained on the two expeditions. One paper on a new species of Hnteropneusta has already been completed. A short account of the Vertebrates, by Mr. W. B. Alexander, M.A., is also ready. The Sponges are now in the hands of Professor Dendy, and Professor W. A. Herdman has the Ascidians. The Holothurians are being worked up by Dr. J. Pearson of the Colombo Museum, the other Echinoderms by H. L. Clarke. The remaining groups have not yet been distributed owing to the dislocation of work consequent on the war. The study of the marine fauna of the Abrolhos Region and the North- west coast of Australia forms an interesting sequence to biological work HOUTMAN ABROLHOS ISLANDS. 135 at Ceylon, and in particular to that of Stanley Gardiner at the Laccadive and Maldive Islands. We have found the reports of Stanley Gardiner’s expedition, both on the fauna and the formation of the islands, of very great service. GENERAL DESCRIPTION OF THE ISLANDS. The islet group of the Abrolhos extends between 28° 15’ and 29° §. lat. and consists of a large number of islands, the smallest of which are only a few square yards in area and generally uncharted. There is urgent need of a survey, for the maps of the islands leave much to be desired and we often had evidence of the inefticiency of the charts. The archipelago (for such it might be termed) is about 50 miles in length, and the islands oceur in four groups. The most northerly group consists, however, of a single island, North Island, whilst the others (the Wallaby, Easter, and Pelsart Groups) comprise a number of islets which according to our views are closely related to one another. These four collections of islands are separated by channels whose depths are approximately the same as those between the entire Abrolhos Group and the mainland. It may be noted as a characteristic feature, that all the islets occur ona line running roughly N.N.W.-S.S.H. Furthermore, if an island is much longer than it is broad (and this is frequently the case), the long axis usually runs almost north and south. The submerged reefs are often found to be running in the same direction. It is usual to leave Geraldton about midnight or early in the morning, so that the islands will not be reached before the sun is well up. It is difficult to observe submerged reefs if the sun’s rays are striking the water very obliquely. As a matter of fact, on each occasion that I have visited the islands the night start was quite unnecessary, for unfavourable winds and rough sea prevented us sighting the islands until late in the following afternoon, Owing to the low elevation of the islands (usually about & feet above sea- level), with the exception of some to be mentioned presently, they are not seen until one is quite close. The picture presented is somewhat uninteresting, for the vegetation covering them is very scanty and consists of low bushes. No palm trees oceur whatever, the largest plants being mangroves, but these are by no means common and eceur on the lagoon flats, so that they are rarely visible from the sea. The gorgeous pictures made by the living coral below the glass-like surface of the water of the lagoons more than makes up for the uninteresting appearance of the surface of the islands. Coral is everywhere in evidence and appears to be growing luxuriously, although the Abrolhos are situated on the extreme southern limit of coral-reef formation. The islands would be quite uninhabited were it not for a few guano workers who come over for the summer months only, and frequent some 136 PROF. W. J. DAKIN : EXPEDITION TO THE small island or part of a larger one. The fishing luggers are, however, very frequent visitors. These boats are chiefly manned by Italians or Scandinavians and sail from Fremantle or Geraldton. They may fish in the deeper waters round the islands, or send one or two small boats amongst the reefs, whilst Text-figure 2. Long. 113° 40’ East of Greenwich 113° 50’ WALLABY GRouP 2 North lv @. 1 28°20 Ke) he a) (Behp *. Wallaby | keeping the lugger off during the day. At night, the numerous sheltered regions between the islets are available, and in case of storms a ready harbour is at hand, although, as previously pointed out, it is useless trying to enter the lagoons after sunset. = HOUTMAN ABROLHOS ISLANDS. 137 North Island (see map, text-fig. 2) is the most northerly of the Abrolhos Group. It is situated about nine miles north of the Wallaby (or Wallabi) Isles, and is about one square mile in area. The Wallaby Group proper (see map, text-fig. 2) consists of two large islands, the Hast and West Wallaby Islands, and a number of small ones which are either unnamed or bear local names given by the fishermen. The West Wallaby Isle is the largest of the Abrolhos Group. About 14 miles south of the Wallaby Group is situated the Easter Group (see map, text-fig. 1, p. 128). The largest of this collection of islands is Rat Island, which, however, is much smaller than either the East or the West Wallaby Islands. The Zeewyk Channel (about seven miles across) separates the most southerly islands of the Easter Group and the northern reefs of the Pelsart Group. The largest island of the latter group is Pelsart Island itself (calied Long Island by the fishermen), an islet of considerable length extending to the extreme south of the Abrolhos archipelago. Although far exceeding all the other islands in length, being roughly eight miles long, Pelsart Island is in most places only a few hundred feet across, and much of this consists of heaped coral fragments. The Pelsart Group takes much more closely than any of the other groups the form of an atoll, a somewhat triangular atoll, with the apex directed towards the south. Pelsart Island forms part of one side of this triangle. A distinct resemblance is to be traced in the Easter Group to the atoll, but this question will be considered in detail when describing the structure of the different islands. Saville Kent stated that ail the islands of the Houtman’s Abrolhos were coral formations with the exception of certain of the Wallaby Group. In this region plutonic rocks were said to occur corresponding to those of the mainland and having an elevation of some 30 or 40 feet. It is difficult to understand to what particular rocks Kent referred, but in any case his remarks are incorrect. The Wallaby Islands were carefully examined and, although they are much larger than the other islets and attain the greatest elevation, there is no trace of any rock other than recent limestone. All the Abrolhos islets are coral formations. The nearest rock of Kent’s plutonic type on the mainland is granite. This has been met with in Geraldton at a depth of 420 feet and it also occurs down the coast. It may be pointed out here that Michaelsen and Hartmeyer in their introductory section to the ‘ Fauna Siidwest- Australiens’ speak of the coastal limestone near Fremantle as being uplifted coral-reef. This is also quite incorrect, and it is curious that these German workers failed to notice it. The limestone of the Fremantle coast has been formed by the action of water on vast accumulations of drift sand containing much calcium carbonate. The uplifted coral limestone of the Abrolhos is quite different from this seolian limestone of the more southern coast. 138 PROF. W. J. DAKIN: EXPEDITION TO THE The Abrolhos Islands are favourite nesting places of sea-birds, the chief breeding season being about the months of October and November. During these months enormous numbers of Noddy Terns, Sooty Terns, and Lesser Noddies frequent certain islets and the bushes are covered with their nests. They rise in hundreds as one passes across the islets, although many stop in their nests and refuse to budge unless pushed off. Wherever the coral surface is covered with deposits of sand, etc. the excavations of Mutton- birds abound. One falls through into them at every other step. As might well be imagined, vast accumulations of guano have been formed on the Abrolhos and this was removed on a large scale in the early nineties. Wooden jetties were built and long trackways put down on the Wallaby Islands. Moderately sized vessels called for the fertiliser, which was shipped out of the State. Up to 1898 the records show that 55,000 tons of guano had been excavated and exported. The work is still carried on during the summer months by Mr. Fallowfield of Geraldton, who has the concession from the Government for this purpose, but the amount now shifted is small indeed compared with that of earlier days and it is not allowed to be exported out of the State. The guano deposits have covered the coral surfaces of the islands with a layer a foot or more in thickness. The first step in the process of collecting consists in the removal of all plant-erowths from the area being worked. Large loose coral biocks are then picked: out of it and stacked, and the rest is shovelled away and screened so that all the small stones are removed. The residue is trucked to the jetty and bagged for shipment to the coast. When the guano deposits are removed in this way, the material is stripped so that the limestone surface of the island is exposed. This is often quite flat and smooth, for the action of the rain, the guano deposits, and the coral has resulted in a compact and rather hard surface limestone. The appearance of such an island afterwards is often curious. The surface is brushed clean of all sand and deposit, whilst walls appear to have been constructed in all directions—they represent the stacked coral blocks picked out of the guano. The invasion of the islands by guano workers has had an appreciable effect upon the land flora, which consists almost entirely of xerophytic and halophytie shrubs, and many of the plants have been introduced from the mainland. So far as the fauna is concerned the guano workers do not seem to have affected it very much, if at all, except that rats became a plague on Rat Island and cats were introduced to keep down the pest. The rats no longer exist, but a few cats occur in a wild state and probably bave a very happy time during the nesting season of the terns. In concluding this brief general description, it may be pointed out that any expedition visiting the Abrolhos Islands has not only to carry full supplies of provisions, but also fresh water. All the water used by the guano workers has to be carried in tanks from Geraldton and is then laboriously bailed out with kerosene tins and stored in small tanks. In the rainy HOUTMAN ABROLHOS ISLANDS. 139 season (winter months July-September) there may be enough rain water, but rain which has fallen heavily whilst I have been on the islands has speedily disappeared. One or two wells occur on the larger islands—EHast and West Wallaby Isles, Rat Island and Pelsart Island, but the water is poor and brackish and often contains too much decaying organic matter to be pleasant. METEOROLOGICAL AND HyproGRsPHICAL CONDITIONS. Records of sea temperatures are very badly wanted on the coast of Western Australia and up to date but little is known. It would seem strange that the Meteorological Department, although so desirous of fore- casting weather, has not arranged long ago to take sea temperatures at their coastal stations, especially where these are only a few yards from the sea, What data I have included here has been obtained from a few records made on our expedition and from the work of Mr. J.J. East of Perth. This gentleman collected the sea temperatures recorded by the engine-room staffs on mail and coastal steamers in Australian waters. The temperatures apply to the sea water as pumped into the condensers. Unfortunately this source of our knowledge only covers a period of a few months and goes but little way to fill up a big gap. The results so far are quite interesting. From the appended table it will be seen that the coast of Western Australia is washed by water the temperature of which ranges from about 15°6° C.-19°4° C., on the south, to 23°3° C.-30°5° on the north-west. The seasonal change at most places is not very great, and the highest sea tem- peratures are recorded in February and March, the lowest in August and September. Sea Temperatures (°C.) * observed in the offing along the West Coast of Australia (chiefly based on the engine-room log of 8.8.‘ Paroo’ during the years 1908-12, voyaging between Fremantle and Singapore, and various coasting steamers, 1912). Jan. Feb. Mar, Apl. May June July Aug. Sept. Oct. Nov. Dee. Derbyauirieet: ae S207 830:01 28:9: 127°8. 25:9) 25:6) 25:91 25:6) a. ow 10:0 Broome ...... 30°77 30°0 30°0 29° 27-7 247 248 24-4 23:3 244 98:3 29-4 Hedland...... 29:4 30:0 29:7 27:8 26:7 24:6 22:5 99-7 29-7 95:3 27:2 98:6 Cossack ..\.:. 98°3, 29'6 27:8 .28:9 27:2 29:5 29-7 91-4 99'7 94-6 961 979 Onslow a... 96°38. 28:4 282 28:6 26:7 22:4 23:9 21:8 93:3 244 93:8 97-8 Carnarvon.... 244 25° 25:9 25:2 24°9 23:5 22:9 90:8 20:6 21:7 21-7 23:3 Geraldton .... 22:26 23:6 °° 23:7 23:3 21:8 214 204 197 19:5 19°5 20:4 99-7 Fremantle . 22:2 99:7 93:3 24:5 203 20°83 18:9 171 17°83 19:5 20:0 31:9 (10-80 miles N. of por) ALAN ys seca ve us re .. 19:0: 717:8' 2160 Norr.—Except in the case of Fremantle, the figures are the mean of the observations taken during the watch (10-40 miles run) after leaving, or when approaching, tke port named.—Our own temperature observations taken with certificated _ thermometers indicate that the winter sea temperature at Geraldton is lower than that given in the above table, and the above figures must be taken there- fore as only approximate. [W.J. D.] * T am indebted to Mr. J. J. East of Perth for the compilation of this table, 140 PROF. W. J. DAKIN: EXPEDITION TO THE At Geraldton, the nearest port to the Abrolhos and in the same latitude as the islands, the sea temperature ranges from about 16°7° C. in winter to 23°5° C. in the summer. [For purposes of comparison the air temperatures taken in the shade at Geraldton during the three years 1913, 1914, and 1915 are appended. Iam indebted to the Common- wealth Meteorological Bureau for this information. ] Shade Temperatures at Geraldton in °F. Year. Jan. Feb. Mar. Apl. May June July Aug. Sept. Oct. Nov. Dee. Mean max. 1918. 839 856 83:2 787 789 72:6 687 67:9 72:9 72:5 76:4 775 Highest .. 99:3 1065 107-2 94:6 90:0 82:0 80:0 74:2 95:0 861 104-0 103°3 Mean min. 645 668 65:1 61:0 655 556 486 51:7 51:0 555 586 62:9 Lowest .. 56:0 543 545 514 470 47:0 37:3 43:0 39:0 46:3 47:0 52:8 Meanmax. 1914. 850 83:3 826 75:1 72% 71:0 676 78:2 767 77:0 77:1 79°6 Highest .. 109:8 108'3 105°8 84:0 898 83:8 742 840 95:0 97:8 95:0 101'8 Mean min. 656 643 626 563 52:2 51:5 500 52:0 53:8 606 625 65:0 Lowest .. 57°5 545 47:0 440 408 45:0 39:0 43:0 42:0 49:0 53:0 66-4 Mean max. 1915. 86:3 844 840 827 76:0 714 685 694 67:3 729 819 866 Highest .. 100:0 106-8 99:0 97:2 848 81:8 746 783 69:8 96:0 1046 115-0 Mean min. 693 69:0 641 642 585 58:5 545 544 535 56:3 61:0 66:0 Lowest .. 59:0 60°38 542. 53:0 508 530 41:6 420 46:2 46:0 51:0 546 If sea temperatures are taken a few miles away from the coast in the region between Geraldton and Shark’s Bay, it will be found that these off- shore waters are somewhat warmer than those close to shore in the same latitude. Now this is an important point, for statements have been made to the effect that the marine fauna of the Abrolhos Islands contains so much of a tropical element that one must travel up to the Tropic of Capricorn before reaching a similar fauna on the coast. Whether this be correct or not *, the fact remains that the Abrolhos are coral islets, and no coral reefs are met with on the coast for many miles north of this latitude. The statements about the tropical character of the Abrolhos fauna were first put forward by Saville Kent, after his very short stay at the islands in 1894. He sug- gested also as an explanation that an ocean current existed which set in from the equatorial waters of the Indian Ocean and brought down floating larvee without impinging on the adjacent coast. In support of this, Kent stated that temperatures as low as 56° F. (13°3° C.) were recorded at Geraldton when at the same time the temperatures at the Abrolhos were 14° F. higher (21:1° C.). This is a very considerable difference—14° F. between two and it isa great places 40 miles apart with oceanic conditions prevailing pity that Kent did not give more details concerning his figures. * According to Mr. Alexander, who has just completed an account of the Fishes from the Abrolhos, the greater number of species are distinctly Southern types (extratropical forms). HOUTMAN ABROLHOS ISLANDS. 141 On our first expedition to the Abrolhos, leaving Geraldton on November 9th, 1913, the temperature of the sea at Geraldton was 16°7° C. when we started. On our arrival at the islands we found the temperature to be 20° C. Arrangements had been made for the Harbour Master to take some tempera- tures at Geraldton whilst we were at the islands. The following is a comparison of the figures, the temperatures being taken at the same hour and early in the morning at the two places. Geraldton. Abrolhos Isles. Nov. 9th,1913_... NGenesO: ne 202 ©: py Othe os he es72. ©: ee 20°C: Pee Oth ne 6s Nese. “a 19°75° C. sie LOU ys oe 19-4° ©. ee 20° C. c6 cllBy Ang oe ee 19:49: C) Re BAe The temperatures at Geraldton were certainly lower than those at the Abrolhos Isles, but a remarkable increase happened to take place within a week at the former place whilst the sea temperatures at the islands remained constant. This was undoubtedly due to a very local heat-wave which visited the mainland causing the temperature to rise to 104° F. in the shade! It would quite easily affect the waters of the Geraldton bay. Iam afraid this Text-figure 8. Carnarvon West AUSTRALIA 20 Abrolhos” 9 Islands. Geraldton 17-2\0Fremantle illustrates the futility of drawing conclusions from isolated records. At the same time the early temperatures record a difference of about 3° C. between the waters bathing the Abrolhos and those of Geraldton. This is not such a great difference as that noted by Kent, und | am inclined to regard the low temperature of 13°3° C. recorded by him at Geraldton in July 1894 as also abnormal (granting the possibility of error in making the record). 142 PROF. W. J. DAKIN: EXPEDITION TO THE Further information is, however, to be obtained from the engine-room observations already referred to. If isotherms are plotted from the readings of June 1911, it will be seen that a definite tongue of warmer water extends down the coast of West Australia, the tip reaching to about the Abrolhos Islands (text-fig. 3). During this time, a southern movement of equatorial water was taking place, and it was separated from the coast by a distinct zone of cold water which was still more obvious during the following month and extended from the Leeuwin northwards. The evidence so far collected goes to show, then, that there is a general tendency for the temperature of the sea at the Abrolhos to be slightly higher than that at the coast, or at all events for certain periods, and that an equatorial current may account for the phenomenon. The difference between the Abrolhos Islands and the mainland may not however lie in the same direction at all seasons of the year. It is to my mind noteworthy that it is during the winter that we are sure of a certain higher degree at the Abrolhos than at Geraldton, and it is not the actual amount of difference that I regard as important but the fact that the temperature at the Abrolhos Islands is pulled up during the winter months, when it probably rarely falls below 20° C. Indications of a tropical current reaching the Abrolhos in the manner described, do not appear on the American Pilot Charts of the Indian Ocean, but the following quotation from Otto Kriimmel’s ‘Oceanographie,’ 2te. Aufl. ii. 675-6 (1911), shows that other investigators have noticed these West Australian conditions. “Tm Vergleich zur homologen Benguelastrémung ist die westaustralische nicht durch die gleichen niedrigen Temperaturen an ihrem Kiistenrande ausgezeichnet, wie wir sie oben darlegen konnten. Wir haben als Ursache dieses Verhaltens die abweichende Kontiguration des australischen Festlands zu bezeichnen, welche einem von Norden und Nordosten her kommenden warmen Strom entlang der Kiiste einen Weg nach Siiden und somit in den Riicken des Stidostpassats gestattet. Dadurch ist also eine ausreichende Kompensation an der Oberfliiche von Norden wie von Siiden her erméglicht ; man kann schon aus dein Auftreten von Riffkorallen dei den Houttmanischen (283° S.B.) auf dauernd warmes Wasser schliessen. Hin warmer Strom kommt insbesondere im Siidsommer aus der ‘limorsee und, in den Buchten Nordwestaustraliens Neerstréme entwickelnd, geht er nach Siidwesten, um anscheinend bei der Dirk-Hartoginsel, dem westlichsten Punkte des Festlandes, nach Siiden umzubiegen und nach dem Befund der Gazelle- Expedition 16 Seemeilen in 24 Stunden nach Siidosten zu laufen.” We have seen that Saville Kent emphasized the tact that the marine fauna of the Abrolhos differed in an extraordinary manner from that of the adjacent coasts, and he accounted for this by the assumption of a current, the existence of which is now supported by a considerable amount of evidence. The Hamburg Expedition of Michaelsen and Hartmeyer did not visit the HOUTMAN ABROLHOS ISLANDS. 143 Abrolbos Islands, but Michaelsen adds a note concerning them in a report *, apparently as a result of an examination of collections in the West Australian Museum. His report is somewhat singular for he makes no reference to the only marine biologist who had visited the islands, and adds: ‘An exact description of these islets and of their physiographic faunal relations has been given by Helms. After studying the zoological materials of the Abrolhos in the West Australian Museum in Perth, I can affirm Helms’ assertion about the marine fauna of the Abrolhos.” Now, the first mention of the tropical character of the marine fauna of the Abrolhos is due to Saville Kent. Helms states that this naturalist visited the islands but does not refer to his paper, and indeed does not enter into any discussion on the character of the marine fauna or on the southern position of these islands. Yet Michaelsen states that he can support the remarks of Helms on the character of the marine fauna, and then goes on to account for the tropical character of this fauna. His explanation includes an exactly similar theory to that propounded ten years before by Kent. This is put forward without even the temperature records that Kent had made, and without reference to Admiralty Charts for confirmation. We must consider in some detail the further remarks of Michaelsen to the effect that a “Kalter Auftrieb” rising from the depths of the sea may also explain the difference between the marine fauna of the Abrolhos and the coast. The term “ Kalter Auftrieb” is used by the German hydrographers to denote the rising cold waters which are met with along a lee shore, and are due to the removal of the more superficial waters to windward. This phenomenon accounts very well in many places for a great difference in temperature between the coastal waters and those some little distance out, and excellent examples are to be met with in the Baltic Sea and the Sea of Azof, as well as along the coasts of some continents when strong offshore winds are blowing. ‘This vertically moving colder water is also held to account for the lower sea temperatures in the Atlantic and Pacific Oceans along the Western coasts of Africa and South America in the region of the trade winds, one of the results of which is the well-known absence of coral reefs along the West coasts of these continents. If, however, we examine the following diagram (text-fig. 4), which shows the position of certain isotherms between the land masses of Australia, South America, and South Africa, we shall see that there is a difference between the west coast of Australia and that of these other two continents. “The inflowing and outflowing waters probably find their way not only through the wide and deep channels between the islets, but to some extent through the body of the reef, which in its upper portions would appear to be almost as open and pervious as a sponge.’—Judd. ‘‘ Materials sent from * “Die Tierwelt Siidwest-Australiens und ihre geographische Berichtegen.” Mitteil. Geog. Ges. Hamburg, Bd. xxii. LINN. JOURN.—ZOOLOGY, VOL. XXXIV. all 144 PROF. W. J. DAKIN: EXPEDITION TO THE Funafuti.” Report of Coral Reef Committee of the Royal Society, Section X. p. 176 (1904). The mean annual surface tempera- ture of 21:1° C. is much nearer the equator on the west coasts of South America and South Africa than it is on the west coast of Australia. In other words, the water washing the coast of © West Australia is warmer than that ; washing the west coasts of South America and South Africa in the same \ latitude. ° 10 20: 30° aa* $0 -267°C 2u°C 18-38° +rtt + 53884 _ We really require much more data from this coast before theorising on the ‘ effect of such’ movement of water feu as a “Kalter Auftrieb.” It would . 1) appear, however, from the following that Michaelsen’s suggestions can scarcely hold good for the region under discussion. Hast winds are stated to prevail on the west coast of West Australia for a considerable part of each day. ‘‘ These push the warmer surface water seaward, and so cause the rising of the colder bottom water to replace the driven water at the surface. The rising of the colder bottom water, ‘the cold swell,’ takes place quite close to the coast and of course in- fluences mostly the line of the coast and its fauna. On the surface the originally colder water soon gets a higher temperature, so that by the time it has been driven by the east winds over the space between the coast and lee: the Abrolhos it has grown warm.” We shall refer later to the winds on this coast; let us look again at the sea temperatures. If a cold water up- welling, especially of any extent, takes place along the west coast of Australia it is caused by the trade winds (not an offshore local breeze), and it would probably occur along the edge of the co AOI ahs oe aioe Text-figure 4. HOUTMAN ABROLHOS ISLANDS. 145 continental shelf and not against the coast. The water between the Abrolhos and the mainland is only about 20-25 fathoms deep. Now it is hardly likely that one would find on this shallow coastal strip of water lying over the shelf a much colder zone nearer the bottom consisting of water running in towards the land with a well marked layer of warmer water running out on the surface. The temperatures, so far taken by me, give no indication of such a state, the waters of the shelf appearing homothermic. Itmust be remembered that there is frequently a considerable sea running in this region, and with depths of only 20 fathoms it is likely that homothermic water will result. The following figures are taken from those of the last Abrolhos Expedition. Unfortunately they are few because we were usually close to the islands working the lagoons or else to the leeside (the Hast). Thursday, 28th October, between Wallaby and Easter Groups towards open ocean. Surface cee20sbe (C: 15 fathoms ... 19°87° C. Sunday, November 7th, between Haster Groups and Pelsart Group. Surface see 20200 C: 20 fathoms ... 20:3° C. Friday, October 22nd, outside Haster Group. Surface Son bg Rely (0) Bottom ee ligeoe C: Many readings were taken in the lagoons, but they are useless from the above point of view—the surface and bottom temperatures were usually the same except that the extreme surface layer, on an absolutely calm day, was often higher as a result of the direct heating effects of the sun. If we examine the prevailing winds at the Abrolhos we shall see that there is less evidence still in favour of a warm surface current blown out from the coast accounting for the Abrolhos Fauna. Winds, Geraldton District and Abrolhos. Summer—From the middle of September to the middle of May, Southerly winds are characteristic. Rest of Year—Winds variable, with occasional N.W. to W. gales of moderate force and of usually short duration. Summer Winds—During the summer the Southerly wind blows almost without a break and resembles the trade wind in character and in force. In the early hours of the morning it is certainly in the form of a land breeze, but is E.S.E. to 8.H. and is light. About 9 a.m. it hauls to the S. or S.S.W. and commences to freshen, blowing its strongest during the afternoon and continuing very frequently until after midnight. It then falls light and works back to the land breeze again. Very occasionally, lib. 146 PROF. W. J. DAKIN : EXPEDITION TO THE during the summer, the wind with a falling barometer comes away fresh from the E. to N.E. at sunrise, but it drops at noon and is followed by the Southerly as usual. Winter Winds—During the winter the winds are variable. Southerly winds are frequent though of lighter force than those during the summer. N.. winds are also of frequent occurrence in the mornings, but there is a calm or Southerly during the afternoon, or perhaps a strong N.W. blow. Taken altogether there is far less wind during the winter than summer. Now, at the Abrolhos, 30 to 50 miles away from land, the winds are much the same, except that the land breeze is not nearly so evident and often quite absent. There is not much support here for a prevailing Hast wind blowing out the surface water. Further information may be obtained from actual observations of the currents. Here again I have had to depend upon information culled from the fishermen who are out at all seasons of the year. They state that during the winter, a southerly current is usually experienced between Geraldton and the Islands. Its velocity is about one knot per hour, but this may be increased with N. winds to 2-3 knots. This is quite in favour of the tropical current theory. During the summer southerly set has also been experienced, but as calms are very seldom it has not been noticed nearly so definitely. TIDES. The tides on the West Coast of Australia present some great problems, and very little is known of these phenomena at present, although the question is of importance to shipping. For the greater part of the coast from the Leeuwin northwards (as far as Shark’s Bay), the rise and fall of the tide is very small, the mean spring rise being only 2 ft. 6 ins. at Fremantle and Geraldton, and only 2 feet at Bunbury. At Carnarvon the mean spring rise is 5 feet, but in the north-west the conditions are very different, and a mean spring rise of 34 feet has been recorded at Derby. The mean neap rise at Geraldton is only 1 ft. 6 ins. Tt must not be thought, however, that tidal phenomena on this coast are only peculiar in that the amplitude varies enormously along the coast, and is so small at Fremantle and Geraldton as to be practically insignificant. Tie greatest problem is associated with tidal irregularity. Thus, at Fremantle the tides are usually diurnal, roughly 24 hours elapsing between the two high tides. Occasionally, however, for a succession of a few days, the tides become semi-diurnal. The diurnal tides are associated with irregularity, and it is extremely difficult to predict them with any accuracy. At the Abrolhos, as at Fremantle, there is but little rise and fall of the tide, but it is of more use for collecting purposes than at the latter place, and seems to be of somewhat greater amplitude. We should say that the HOUTMAN ABROLHOS ISLANDS. 147 mean spring rise was nearer three than two feet. On both occasions when we were at the islands the tides were diurnal, and it came as a surprise on our first expedition to find that not only was there but one tide every 24 hours, but that low water came at practically the same hour each day. On both expeditions low water occurred always between 6 and 8 a.m. The harbour master at Geraldton stated that during the summer months the tide at that port was usually high about 9 p.m., with low water at 9 a.m., whilst in the winter the converse was the case. Whether this be so or not at the Abrolhos is impossible to say. We can only affirm that all our shore-collecting was carried ont in the early hours of the morning during the last week of October and the month of November. Stokes gives the following rather surprising details which we can neither confirm nor deny; such conditions certainly did not prevail whilst we were at the islands. The high-water at full moon and change occurred at 6 p.m. During the night there was a short flood lasting six hours, the water rising seven inches, and a shorter ebb of only two hours’ duration, the tide falling five inches. During the day, however, there was a flow of 8 hours 20 min., and an ebb of 8 hours 5 min., the rise and fall being 25 and 26 inches respectively. There were further variations at the different island groups. It appears quite certain that the tides are very irregular. TUAINFALL. No statistics exist, or are likely to exist in the future, concerning the rainfall at the Abrolhos Islands. The nearest figures that can be utilised are those for Geraldton, they will suffice to show the type of rainfall experienced in this region. The rainfall is stated in ‘ points,” 100 points equalling an inch. rainfall at Geraldton. Year. || Jan. Feb. Mar, April May June July Aug. Sept. Oct. Nov. Dec. \| Total. [9ISIMy sO eens 7) d07) eae 946) 56 1826 7 631 268 118 7 8 1914 ml 3 38 Nil 29 1138 291 2738 185 5 87: 93 Nil |) 1032 1915 ....] 18 459 28 78 241 790 529 3382 159 97 Nil 18 |) 2749 Averages for the last 38 16 27 383 87 272 454 392 301 185 66 25 13} 1821 years. J Norr.—The figures indicate “ points,” one hundred points of rain equal 1 inch. It will be seen that the average annual rainfall for the last 38 years is 18 inches.. The dry months are October, November, December, January, February, and March, the average fall recorded for this period of six months being only 1°8 inches. 148 PROF. W. J. DAKIN : EXPEDITION TO THE NORTH ISLAND. North Island was only visited on our first expedition to the Abrolhos, and our observations have not been so detailed there as at the other places. This island does not appear to have been visited so frequently as the others, being due in all probability to its greater distance from the coast. North Island is almost a square mile in area. It is not surrounded by smaller islets, and differs from the other groups we shall consider, in being without any real lagoon region. In fact, the part above water at all states of the tide rises from a coral flat’ which fringes it on all sides, but it is very much wider on the west than on the east. On the Hastern shore, a fringing reef lies quite close to the island, and a few breaks or channels occur where fishing luggers can obtain some shelter and anchorage in bad weather. The channels, which are very irregular, are about 5-6 fathoms in depth, with sandy bottoms on which large holothurians may be easily seen through the clear water. On the West side of North Island the margin of the reef lies just over a mile away from the shores of the island. No boat channel or lagoon exists between it and the shore. A broad reef-Hat extends from the shore out to the reef-margin and is covered by a foot or two of water at high tide. One can walk out for the whole distance, but care must be taken, for the coral surface is rotten and full of holes, with larger hollows here and there. There are no living corals growing upon the surface of this flat. The same reef-flat extends north and south of the island for a greater distance and is continued as a submerged reef for a considerable stretch towards the Wallaby Islands. The surface of the flat is covered with living Vermetus, and, as a matter of facet, this gasteropod is the common animal of all the reef-flats of this region. The shells puncture neat but exceedingly painful holes if anyone carelessly places an unbooted foot on such ground. The rock-pools swarm with young fish which must pass through their larval stages with some rapidity in such heated water (the temperature rises to remarkable points whilst the sun is shining). A blue crab, Thalamita sp. (probably stimpsoni, an Indian form), was very abundant in holes in this coral, and also the sea- urchin Echinometra mathe. We shall meet with this species elsewhere. North Island itself and the two large islands of the Wallaby Group are the only ones to attain any height above 8-14 feet. The highest point on North Island is 42 feet above sea-level. All this high ground of North Island consists of blown sand, which rises in the form of dunes quite close to the eastern shore margin. Other hills lie irregularly behind these, so that a distinctly high region occurs close to the eastern side. The height of the island falls off rapidly as the centre is approached. At the south-eastern corner of the island the margin takes the form of a low coral-limestone cliff instead of sand-hills, and the coral-reef flat, already referred to above, runs into the foot of this. Asa matter of fact, in many places it is undereut by HOUTMAN ABROLHOS ISLANDS. 149 the sea and overhangs very considerably (PI. 14. fig. 6). This limestone conglomerate forms the margin of the island for a little distance along the western shore, but gives place to a sandy beach again as the northern shore is reached. The sand dunes are not so high here as on the eastern side. The coral limestone, which is exposed at certain places along the shore, can be traced inland and evidently forms the foundation of the island. Its surface is about 6 or 8 feet above sea-level, and it agrees in structure with that to be referred to later in connection with the Wallaby Isles. The central part of the island is quite low and flat, and only a few feet above sea-level (a salt lake, dried up at the time of our visit, occupies part of it). No rock exposures are visible once the shore is left, the whole being covered with sand which, as we have remarked above, is responsible for the height of the island above sea-level. The sand-hills are thickly covered with low halophytic bushes of a hard and brittle nature, and the terrestrial fauna appears to be poorer than that of the southern islands. The lack of fresh water is probably responsible for this, Shore-collecting was not very successful at North Island, and as the other islands appeared likely to offer better opportunities for all-round work we spent very little time there. A number of specimens of eterodontus phillipi were captured whilst we lay at anchor close to the shore, together with the Parrot fish (Coris auricularis, one of the Labride); Groper (Acherodus gouldit) ; Skipjack (Caranex platessa) and Buttalo Bream (Carangide). The latter looks very much finer than it tastes! A number of birds were noted and are enumerated in Mr. Alexander’s report. Further reference will be made to the structure of North Island in the section dealing with the formation of the four islands. We may remark here that the wide expanse of reef-flat represents the result of erosion and that North Island will probably be cut down still further in the future. THE WALLABY ISLES. This group of coral islets is, in some ways, the most interesting of all, for it is on certain of these that the land fauna requires, as an explanation of its character, the assumption of land connection with the mainland of Australia. The group consists of two large islands (the East and West Wallaby Islands) together with a considerable number of small islets, many of which are not depicted at all on the small-scale charts. The Hast Wallaby Island lies really north-east of West Wallaby Island, from which it is separated by a very shallow channel about a mile across. At first sight it appeared somewhat difficult to piece out the former history and the structure of the irregular collection of islands, reefs and lagoons of the Wallaby Group. It became much more clear after our second expedition, and I now consider that the group represents a stage in the development of an atoll form. 150 PROF, W. J. DAKIN: EXPEDITION TO THE The two large islands form a kind of central mass. To the west of them lies a more or less shallow lagoon with an outer reef margin four miles or so to the west of Hast Wallaby Isle. To the eastward of the two Wallaby Isles is another lagoon with irregular coral reef bounding it, and small islets exist here and there where the ree? rises above the sea-level. When sailing from the coast, one first sights the small islets of the reef on the eastern side, especially the rather long narrow islet (known as Long Island). The elevation of the Wallaby Islands, however, is sufficient to make them more conspicuous than is the case with any of the southern groups, so that in reality they are picked up from the masthead practically as soon as their eastern outposts. On our first expedition we ran into Turtle Bay (see map, text-fig. 5) first, as this shelter was easily reached from the open sea. After a few days we moved down Recruit Bay until a position was reached close to Hast Wallaby Island and almost opposite the channel between it and West Wallaby Island. On the second expedition we reached the Wallaby Group from the south, and after passing through the Channel close to the east side of Long Island (see map, text-fig. 2) we entered the lagoon and anchored close to the latter. Later on we changed to an anchorage of the former expedition in order to investigate the larger islands over again. Dredging was carried out north of the Wallaby Group between the East Wallaby and North Islands, in Recruit Bay, in the Channel east of Long Island, north of Morning Reef, in the lagoon, and some distance south of the group. We shall consider the large islands first. The Hast and West Wallaby Tslands appear to have been only recently separated, and the separation does not amount to much at the present time. The channel between them, which has a width of about a mile, is only a foot or so deep at high tide, and large areas are exposed at low water. It was regularly crossed by members of the party when they wished to reach West Wallaby Island, and if the tide were up and one were near the middle of the channel, with a fresh breeze stirring up the water, the situation was distinctly peculiar— one seemed to be standing in the centre of the sea! The floor of the channel is a flat of coral, here and there covered with a thin layer of mud, and bearing very distinct evidence of solution and erosion by the sea. The only additions are being made in odd places by the ever-present Vermetus. Beyond a few sponges, crabs, and some fish which abound in the large hollows of the flat, little is flourishing here. Hast WALLABY ISLAND. The East Wallaby Island is about 14 miles from north to south and a little more from east to west measured along the northern coast. Its longest diameter runs N.E. to S.W. and juts out considerably in the form of a north-eastern cape, Fish Point, which separates Turtle Bay from Recruit Bay. The highest part of East Wallaby Island is 40-50 feet high and this occurs right on the eastern coast, which is cf some little elevation from the north- HOUTMAN ABROLHOS ISLANDS. alfsya eastern extremity down to a point where it turns more abruptly westward toward the entrance to the Wallaby Channel. By walking along the beach (which is a reef-flat) one can obtain a view of an excellent section showing the structure of the island where its height is greatest (see photo, Pl. 18, fig. 4). Text-figure 5. Recruit Nag Aye E7 2 eo East Wallaby Island. partly From Admiralty Ch Reef Hat area dotted. A&B. indicate position of Sections illustrated (figs.6 & 7). Spots photographed indicated by letter Pe fig,no, (Not all of the Photographs have been reproduced; those numbered 6 and 9 are now 4 and 5 on Plate 13.) The two text-figures (text-figs. 6 and 7) indicate sections at right angles to the line of coastal cliff. They were taken about a } mile apart. A coral-limestone cliff rises direct from the reef-flat and overhangs somewhat. Scale | Sea Mile 152 PROF, W. J. DAKIN: EXPEDITION TO THE It is washed by the sea at high tide. This cliff is only from 4—7 feet high and presents a flat terrace of varying width, bere only a foot or so, there a couple of yards (see text-figs.). From this, another cliff-surface rises to a varying height, 14 feet in the one case shown above, to probably upwards of 25 feet in the other. This is marked by another limestone surface, and it is possible to notice from the sea how this surface is not at the same height along the stretch of coast, but varies slightly. The complete height of Text-ficure 6. i4fr4ains 2eft 14 ft ifr Elevation Coral U7 rock Yj; Terrace i about +ft high” (7 YY Highhde level Reet Flat < 53 leet Scie thnx 60fr > Text-figure 7. EI evation 4 feat of hill Tas Terrace 4ft high —> Reef Flat High tide level 22 feet in the one section and 40 feet in the other is brought about by sand. On East Wallaby Island there is nowhere a thickness of 40 feet of coral above sea-level. Fourteen to twenty feet is probably about the maximum, and this is topped with sand. This condition, however, exists only along the margin of the Hast coast which runs in a straight line N.H.-S.W. If a traverse is made inland ina westerly direction (see map of East Wallaby Island, p- 151) the sand-hills become lower, until at a distance varying from }-} a mile inland they disappear, and one reaches an almost perfectly flat area the HOUTMAN ABROLHOS ISLANDS. 153 surface of which is limestone. No sand or guano hides the rock in this area now, but a few bushes are rooted in the crevices and cracks in the weathered limestone surface. This flat area occupies the greater part of Hast Wallaby Island, and sand-hills are not met with again until one reaches the extreme west and northern coasts. It might be expected that the flat area we have described with the coral-limestone surface would be as many feet above sea- level as the highest limestone exposed on the Hast, coast. This is not the case, it is only about 10 feet above sea-level. It is only on the East coast of Hast Wallaby Island that any limestone higher than this is found. At the same time, the land rises once more on the western margin, so that the greatest heights are found on the East and West coasts. If, however, the eastern elevations are examined, they will be found to be composed entirely of sand, and the exposures are quite interesting. One of these eastern sand-hills is marked on the chart with an elevation of 35 feet—a well is indicated about 200 yards to the N.E. of it. Now just north of this sand-hill the wind has cut an extraordinary pass right through the hills and eroded it down to the limestone surface. The sides of the pass are perfectly vertical and there is abundant evidence that the whole is due to the wind. It is quite obvious here that the high land is due chiefly to blown sand. The southern extremity of East Wallaby Island towards the channel is the lowest part of the island, and a sand-heach presents a marked contrast to the cliffs of the eastern shore. This sandy beach extends up the western shore to the north-western corner of the island where cliffs again appear, but only 3-4 feet in height and have a sandy beach at their base. The northern beaches are also of sand, with sand-hills behind, until Fish Point is reached on the east of Turtle Bay. Samples of the shore-sand from the north and west beaches of Hast Wallaby Island have been examined. The sand consists almost entirely of calcareous matter :—foraminifera, shell- fragments, coral particles, and echinoid spines. The foraminifera are all shallow-water forms, the predominant genera being Orbitolites, Polystomella, and Peneroplis. Other common forms are species of Afiliolina, Planorbulina, Globigerina, and Pulvinulina. A reef-flat exists all round the island, but it presents some points of difference on the various coasts. On the east coast (Recruit Bay), it is narrow, 100-200 feet or so in width, and extends to the base of the lime- stone cliff. It is covered with living Vermetus, and broken up with irregular hollows every here and there. To the south it forms a great expanse con- necting East Wallaby Island with West Wallaby Island. On the west of the island the reef-flat is wider and extends out for at least a quarter of a mile from the beach, which, as we have already noted, is of sand. The same aspect presents itself again on the north-west, and there is much sand and weed growths here. A considerable quantity of decaying weed is conspicuous on the eastern and northern shores. 154 PROF. W. J. DAKIN : EXPEDITION TO THE Let us return to the limestone. The best exposures illustrating its structure are to be seen in Turtle Bay on the west side of Fish Point (Pl. 10). The cliff is collapsing here, but the surface is not weathered in the same way as at some other places where its structure is hidden. The limestone is a conglomerate of coral and mollusc shells (PI. 10). In parts it has quite a stratified appearance due to the presence of great flat coral colonies of a species common at present in the lagoons and allied to Madrepora corymbosa, if not this species itself. Between the corals is a hard mass of coral-mud, or secondary deposits of calcium carbonate, which cement everything together, and in which lamellibranch and gasteropod shells are imbedded. These shells show little or no signs of weathering or of abrasion. {See photo, Pl. 10.) The most common of the molluscs are Turbo pulcher, Reeve. Conus magus, Linn. Cytherea reticulata, Linn. Septifer bilocularis, Linn. There is every evidence of elevation to bring this coral limestone into its present position, and the corals appear for the most part to have grown in situ. We shall refer to this elevation later after describing West Wallaby Tsland. It has been said that plutonic rocks occur on the Wallaby Islands. We have now made traverses across these islands in every direction, and we have also closely examined the wells which occur in one or two places. Nothing but limestoue is to be seen anywhere, and we are quite sure that this rock only is present. There are no references to the actual discovery, or examination, of plutonic rock in the literature on the islands. What statements are made appear to be due to a casual remark which has been handed on by people who have not investigated, and in some cases not even visited, the islands. West Watuasy IsuAnp. There is no need to describe in any detail the structure of West Wallaby Island, for it is clearly the same as Hast Wallaby Island and obviously part of the same elevated limestone mass. West Wallaby Island is of irregular shape and is the largest island of the Abrolhos. From north to south the greatest length is just short of three miles. A few points of interest should be noted. The highest part of West Wallaby Island does not occur on the eastern side (cf. East Wallaby Isle), but on the southern mile of the western coast, and this elevated region is remarkably like that already described as occurring on the N.E. coast of Hast Wallaby Island. The highest point, for example, is again quite close to the shore, and a steep cliff rises at once from the HOUTMAN ABROLHOS ISLANDS. EO reef-flat. The cliffs themselves rise to a height of 30 feet, and the greatest elevation, only a little distance inland from these, is upwards of 50 feet. This south-western corner of West Wallaby Island presents the finest scenic picture of the Abrolhos Islands, the cliffs are of rugged weathered limestone, and the margin of the outer reef swings in at this place until it comes quite close to the extreme S.W. point. The full swell of the ocean breaks, therefore, quite close to the island, and the waves rush across the reef- flat wearing deep pot-holes and excavating the coast. The photographs illustrate the type of coast at this point (Pl. 14. fig. 7). It is very interesting to note that the mile of coast on the south-western shore of the island should resemble closely the stretch on the north-east of East Wallaby Island. Everywhere else on the Wallaby Islands the margin is much lower and with either sand or low-lying limestone. Another point of great interest on the West Wallaby Island is the occurrence of a well-marked terrace on the western cliffs, about 6-8 feet above the level of the reef-flat. A similar terrace, slightly lower in places, is found, as already stated, on the N.E. shore of East Wallaby Island. We shall see later that there is evidence on the Abrolhos Islands of a recent uplift of 8 feet or so. The terrace represents the old sea-level before the last elevation. A considerable part of the area of the West Wallaby Island is low-lying, being about 6 feet above sea-level. This applies to all the north-eastern portion (about a square mile) and the southern part of the Hast coast. The north-eastern region presents a surface which is largely flat limestone— resembling in appearance that photographed by Fryer on Picard Island in the Indian Ocean and described as “ platin.” It is covered with the usual stunted halophytie bush. In most places the limestone is completely hidden by a layer of guano and sand which looks as if inhabited by thousands of rabbits. The excavations are, however, due to mutton-birds, and as they are not very deep it is usual for anyone trying to walk over this area to fall through with oae foot or both at every other step! The pavement-like surface of the limestone has been probably produced by successive solutions and redepositions of calcium carbonate by rain-water—the guano aiding. In past times a large quantity of guano was removed from West Wallaby Island and the ruins of an old camp and about 14 miles of tramway still exist as relics, whilst a wooden pier runs out for a considerable distance on the west coast into the waters of the lagoon. We devoted some little attention to the limestone exposed on West Wallaby Island, especially on the West coast where there was a thickness of 25-30 feet above sea-level. The lowest parts at this place consist of a compact rock, granular in appearance, fairly hard and without any traces of corals or molluscs. Microscopic examinations, kindly carried out for me by Professur Woolnough, of the University of West Australia, reveal the 156 PROF. W. J. DAKIN: EXPEDITION TO THE fact that it is entirely composed of Foraminifera, Nullipore fragments, and Echinoderm spines, with a considerable amount of secondary deposition of carbonates round the individual grains. It is in fact an indurated beach- formation. No trace of land-derived sediment exists. It is this rock which forms the lowest part of the cliff, a layer of coral rock infiltrated with secondary carbonates appears at a higher level. A reef-flat extends round the base of the West Wallaby Island. On the eastern side it extends out for some distance and smaller islets rise from it. This area will be dealt with again below. To the south the flat is narrow, and lies close up against the outer reef at this point. On the south-western side, at the foot of the high cliffs already described, the reef-flat is wide again, and here one meets with plenty of evidence showing that the flat has been formed by erosion. Pinnacles of rock are left standing, overhanging on every side (PI. 14. fig. 7), and great pot-holes occur on the flat itself. Tue ReEer AND SMALLER IsLETS OF THE WALLABY GROUP. It is convenient after discussing the character of the chief land masses of the Wallaby Group to consider as a whole the remaining small islets and the coral reefs surrounding them. The “tout ensemble ” is much more irregular than in the southern groups to be presently considered. A glance at the chart (text-fig. 2, p. 136) will show that a mass of coral reef exists several miles to the west of West Wallaby Island and protects a lagoon with 3-4 fathoms as an average depth in the northern portion. Towards the south-western point of West Wallaby Island this reef swings in, until it joins the reef-flat which fringes both the Wallaby Islands and forms the inner shore of the lagoon. Towards the same point, the broad lagoon becomes more and more shaliow until only one or two feet in depth. The outer reef, against which the ocean swell beats, remains close to the coast of West Wallaby Island on the south as a fringing reef, and then leaves the §.H. point of the island in the form of another broad sweep which extends eastwards, but in a more irregular manner, and can be traced past Noon Reef and Morning Reef, where it turns northerly (text-fig. 2, p. 136). Inside the large expanse sheltered by the reef there are numerous coral growths and remains of islands, the whole thing being the result of solution, denudation, and active coral growth. The two Wallaby Islands are connected, as we .have already pointed out, by a broad and shallow reef-flat. This reef-flat extends over the shaded area in text-fig. 2, and several small islands rise from it, three of which (the more northerly ones) are shown on the map of Hast Wallaby Island (text-fig. 5, p. 151). The most northerly of these is called Pigeon Island by the fishermen, who like to enter Recruit Bay as far as they can and anchor with perfect shelter between Pigeon Island and East Wallaby. We camped on it several times. All the islands on this reef-flat agree in type. They are usually much longer than broad and with the long axis running nearly HOUTMAN ABROLHOS ISLANDS. 157 north and south. They are all flat-topped, and possess vertical or over- hanging cliffs about 8 feet in height (they are highest on Pigeon Island, 8-10 feet). Again, there is ample evidence that they have once been continuous. Pigeon Island is gradually becoming smaller, the cliffs overhanging in places for 20 feet or so forming caverns. Here and there great masses have broken off and are gradually being removed. The structure of the coral limestone of which they are composed is similar to that of the Wallaby Islands. Whilst a glimpse of these islands with the lagoon waters presents certain pleasing features, they make a very uninteresting photograph. No palms or tall plants break the monotony, and the elevation of the islands is so small that from a little distance they appear simply as streaks in a photograph. It may be taken for granted that al] these islets have been cut away from one mass, and they are still being reduced in size by ihe action of the sea. The reef-flat connecting them together is all that remains of the elevated limestone of which they once formed part. We are now left with one or two islets much further away from the Wallaby Islands on the outer reefs to the east, a narrow island known as Long Island by the fishermen and several islets not marked on the chart, some of which are only a few yards across. Many of these islands are obviously the result of blocks of coral heaped up by the waves, and their loose structure is quite different from the more compact limestone of the central mass. We are in fact dealing with structures more like the islets of the rim of a coral atoll. Let us glance at Long Island, it being the largest islet in this situation. It is about one mile long and only a few hundred feet across. Its surface consists of small and loose coral fragments, and in fact the whole island is made up of such coral fragments. They prevail above all else, and one cannot consider coral as playing any subordinate part here to calcareous algee, molluses, or other calcareous structures in the formation of this mass. In places, the weathered fragments have been more or less cemented together to form a more definite kind of conglomerate. I consider that a slight elevation (not more than 6 feet) has played some part in the formation of Long Island as well as the heaping up of the coral fragments by the waves. The straight line taken by the reef on which the land stands is no doubt due to the current, which scours the eastern side, and has produced the channel referred to later. A small lake exists at the northern extremity of the island and there are one or two other similar depressions at other places. We shall see that these are quite common on the islets of the rim of the Abrolhos Groups. This is rather an interesting point, and I have found it very advantageous and informing to compare the descriptions given by Stanley Gardiner of the “faros” of the Maldives with these islands of the outer rim possessing lakes 158 PROF. W. J. DAKIN: EXPEDITION TO THE or lagoonlets. The water of the lakes is salt, and often disappears altogether at low tide. In other cases it is permanently present, and deep, but rises and falls with the tide outside. One in long Island is only a few yards across, the banks are of coral conglomerate and loose blocks, and the floor of the lake or lagoonlet of coral with a deposit of mud. We shall refer in greater detail to the formation of these lakes when describing the same on the islets of the aster Group. Mention must be made here of the channel to the east of Long Island. It is very much used by fishing boats passing north and south, for it has a depth of 20 fathoms, and that quite close to Long Island itself. The channel seems to have broken through the reef, cutting off Morning Reef to the east of it from the coral masses on the west. Perhaps Long Island itself owes its formation to the development of this channel. As one would expect from the depth, a strong scour runs through the channel which bears every appearance of erosion. We dredged several times here and with excellent results. Nothing more remains to be said about the reefs except that coral growth is active at the present time in the lagoons wherever the water appears to be free from mud. To the north-west of the coral marked Noon Reef on the chart is an expanse of water almost completely surrounded by coral reef. It can be entered by a channel to the west of Long Island, and by another one to the S.W. indicated by the arrow on the map, text-fig. 2, p. 136. This expanse of water is called “‘ The Lagoon ” by the fishermen (no other area is spoken of as lagoon by the fishermen anywhere in the Abrolhos Group). Depths of 6-12 fathoms occur, and the bottom is almost everywhere of coral. We investigated this area with the motor dinghy on many occasions, and were always delighted with the gorgeous growths of coral seen perfectly through the clear water. The most abundant coral is a species of Madrepora with a most beautiful blue-purple colour. This occurs everywhere, rising from the bottom in the form of great “shrubs.” The more compact corals like Meandrina and Astrea, ete., do not seem to flourish here. Pocillopora is common in places. We have never obtained a Fungia at any of the Abrolhos Islands, and have seen no traces of such, not even fragments. Several times we tried the dredge in these lagoons and pulled it forcibly over the bottom, tearing away the more delicate corals at the risk of losing the dredge and cable, but we never obtained anything but coral and the small crustacea and worms which live in the crevices. It would seem to us that the outer margins of the coral reefs fringing the Wallaby Group are extending seawards by active coral growth, at least in places, but that such growth is very slow. In the lagoons we see coral growth in certain areas, and sandy deposits with no coral growth in other parts. Here active solution and erosion is taking place, whilst there we find that coral growths are forming, or coral fragments are being heaped up by HOUTMAN ABROLHOS ISLANDS. 159 currents and wave action. In the latter case small islets are formed and are often capped with a little sand. They only rise perhaps one foot above high- tide mark and are known by the fishermen as “ Sandy Islands.” THe Fauna oF THE WALLABY GROUP. The name of the group suggests the first member of the fauna that deserves mention. The Wallaby A/acropus eugenti, Desm., occurs on both East and West Wallaby Islands and in very large numbers, especially on the West Wallaby Island. I have seen far more individuals in one afternoon at these islands than of all species of marsupials seen on the mainland in three years! Some surprise was exhibited last year when I stated that four were shot on our first expedition without moving from the one place—one shot having killed two. On our second expedition we beat this by actually catching four specimens alive, by merely running them down! It is no uncommon sight to see half a dozen leaping before one at the same time. Naturally it is a point of the greatest interest to find fifty miles away from the coast, on a small coral island with only halophytic bushes, an animal of this kind in such large numbers. The large Wallaby Islands are the only ones inhabited by a marsupial. Traces of other mammals also occur, and a much decomposed rat was found (and left) in one of the water-holes on West Wallaby Island. Reptiles abound on these islands and were even common on the smaller ones which we have already referred to as rising from the Wallaby Reef flat. One of the most interesting is the Carpet Snake (Python spilotes), which occurs in very large numbers on West Wallaby Island. We could hardly ever walk down the two miles of the old guano tramway on the occasion of our first expedition without meeting several specimens lying across it, and usually they averaged about 7 feet in length. They were very sluggish. Curiously enough we did not see one in 1915; probably they hibernate in winter, and the winter of 1915 was particularly severe and extended well into Octoher. No specimens were met with on any other but the West Wallaby Island. Two other species of snakes are supposed to occur in the Wallaby Isles, one of which was seen but not Reference to the Lacertilia will be made in Mr. Alexander’s captured. Seventeen species are recorded altogether from paper on the Vertebrates. the Abrolhos Isiands, and probably most of these occur only on the West and East Wallaby Islands. ‘Two or three species are common on the smaller islands of this group, but the larger southern islands of the Pelsart and Easter groups have only one or two. ‘Two species of Amphibia, frogs, are recorded from the islands, but we were unsuccessful on both occasions in finding any. On our second expedition we thought we should be more successful owing to the wet winter and our earlier arrival at the islands. We failed, however, to see any trace of them, LINN. JOURN.—ZOOLOGY, VOL. XXXIV. 12 160 PROF. W. J. DAKIN: EXPEDITION TO THE Several species of birds were nesting on the Wallaby Islands, and an account of the species recorded is given in Mr. Alexander’s paper. The following remarks may be made here in reference to the photographs taken by the author. A large number of birds had congregated for nesting pur- poses on the west shore of the West Wallaby Island, several species occurring together. At one place there were several hundred nests of the Pied Cor- morant (Hypoleucus varius hypoleucus, Brandt), all built of broken twigs and branches from some neighbouring bushes, and erected in close contact on the sandy shore. Our arrival was the signal for the departure of the birds, but the less shy Gulls, undisturbed by our presence, took the opportu- nity of feasting upon the eggs. Close to the Cormorant colony, Caspian Terns, Pacific Gulls, and Silver Gulls were nesting. The eggs were simply laid in slight hollows in the sand amongst stray pieces of mollusc shells. Reference has already been made to the Mutton-bird excavations. The birds must occur in thousands on the Abrolhos Islands wherever the rock is covered with sand in which they can burrow. Several species of insects were captured on the islands, but the collection has not yet been worked up. The ant keeps up its West Australian reputa- tion and is present everywhere, on the smallest islands as well as on the larger. Toe Marine Fauna oF THE WALLABY GROUP. The first place to be examined was Turtle Bay on the north of Hast Wallaby Island. The bottom was for the most part sandy with groves of Alge, species of Codium being particularly abundant. Zostera also occurs here in large quantities. Except for a few Amphipods, both Algee and Zostera were very barren. The shore sand contained a large percentage of Foraminifera of which Orbitolites was the most common. The very characteristic Heterostegina of the Ceylon sands appeared to be absent altogether. Nothing of particular interest was obtained on the reef-flat on the east side of Hast Wallaby Island. Vermetus was very abundant, a few sponges and crustacea occurred, and the coral limestone was bored by Lithodomus and Gephyrea. We used our chloride of lime for capturing fish in the rock-pools with some success here, but unfortunately our work was cut short by bad weather which caused so much disturbance of the water on the reef-flat, that it was impossible to poison tke fish—they moved with the waves to other pools! Our best shore-collecting was carried out on the lagoon side of a small island (loose coral fragments heaped by waves) on the outer reef, about 15 miles to the east of the extreme S8.E. point of West Wallaby Island. A long spit composed of small blocks of coral runs out from this island and is entirely submerged at high tide. The lagoon bottom to the north of the island was of sand, but with very fine growths of coral here and there, the two species most abundant being the HOUTMAN ABROLHOS ISLANDS. 161 purple Madrepora already mentioned, and tree-fungus-like growths of a species allied to M/. corymbosa, forming great flat expansions 38-6 feet across. Several large species of Holothurians abounded on the shore of the island, and could be picked up by wading. Anemones, Ascidians, and Sponges (calcareous and siliceous) abounded, and several Turbellaria were captured under the stones. In fact we were able to make a very good general collection here. A better place still on our first expedition was the eastern shore of Long Island. The beach consists of large coral slabs thrown up by the waves, and at low-water mark almost each slab was covered on its under surface with a gorgeous array of organisms. Rarely have I met such variety and colouring. A straggling brilliant lemon-yellow sponge was one very evident specimen. After turning several stones over and noticing what appeared to be pieces of this sponge falling off, it was discovered that the falling pieces were nudibranchs resembling the sponge in both colour and general appearance: The species belongs to the genus Votodoris, this being the first record for the genus on the Australian coast. It was instituted by Bergh for a-single specimen of .V. citrina from Rarotonga. Two other species have been made by Eliot for specimens from Zanzibar and the Maldive Islands. All three are yellow in colour. It is stated in Eliot’s report that nothing is known of the habits of these animals and further— “With Mr. Gardiner’s specimen is a piece of hard yellow sponge. There is no note, but as the colour and consistency of the sponge closely resemble those of the Nudibranch it is highly probable that the latter frequent it.” It is interesting to find that this is actually the case and in an entirely different region. Our specimens were always found associated with the sponge. The Nudibranch moves about very slowly, so far as could be observed, and apparently lives in the dark under coral blocks. A large number of different species of Brachyura occurred on the Long Island shore, amongst which may be mentioned the characteristic species Liolphus platissimus. This peculiarly shaped and heautifully marked crab was exceedingly common. It is an Indian form. Several specimens of the Stomatopod Gonodactylus chiagra were also obtained under stones. Small macrurans of a species as yet not identified were common, and specimens of at least two species of Alpheus were captured. The Echino- derm collection contained chiefly large crinoids and asterids. The gorgonids were represented by one species. On the western shores of West Wallaby Island collections were also made on the reef-flat. Five or six species of anemones were to be seen here in great numbers. The most abundant molluscs were Astralium sp., with beautiful blue operculum (the species is common on the coast both north and south of the Abrolhos latitude), and Vermetus. Large numbers of the Siphonophore, Porpita, came in over the reef one afternoon. is 162 PROF. W. J. DAKIN: EXPEDITION TO THE Hauls with the dredge were made several times just outside Recruit Bay, with a depth of 20 fathoms. We finished up to the N.E. with a sandy bottom and an abundance of seaweed. Large hauls of sponges were made here, and compound Ascidians and Echinoderms, including Synapta sp., were present in considerable numbers. LHchinoderms were always well to the fore so far as individual numbers were concerned. Crustacea were only moderately represented, being chiefly Brachyura and Alpheids, which abounded in the crannies of the coral rocks brought up in the dredge. Macrura were not common on the whole. A few dredge hauls were made east of Evening Reef in a rather rough sea. The bottom was of sand and small fragments with much weed. A few Hchinoderms were obtained, including a fine Astrophyton. THE EASTER GROUP. The Easter Group differs in several respects from the group of islets we have just been considering. In the first place it has much more the aspect of an atoll with an island in the middle of the lagoon, the total area of the atoll measuring about 35 square miles. The central island is Rat Island, (2 or 3 small islets which lie to the south and are separated by narrow channels a few yards across—dry at low tide— may be considered as part of it), and a glance at the chart (text-fig. 8) will show that the surrounding reefs encircle it in quite a regular manner at a distance varying from 2-5 miles. A gap occurs in the reefs to the south of the group, and they are not well developed to the north-west and north. In other words, the larger entrances to the lagoon are to the north and north-west. We shall see later that this agrees with the Pelsart Group. It will be remembered here that encircling reefs were similarly best developed to the south of the Wallaby Group, whilst the chief entrances to the lagoons opened to the north. The encircling reef to the west of Rat Island does not bear any islets above sea-level at high tide. Thereare several islets, however, on the eastern rim forming a chain upwards of eight miles in length. This development also agrees perfectly with the character of the more southern group of the Abrolhos. Hach group of islets is evidently being modelled by the same forces and conditions. The depths of water in the lagoon of the Haster Group range from 2-22 fathoms, there being quite considerable depths on the eastern side of Rat Island between it and the islands of the rim. Before considering, however, the lagoon and the coral reefs, let us glance at the central islet—Rat Island. This is the largest island of the Easter Group, but it is smaller than either of the two Wallaby Islands, being only a little over # mile long (from N. to 8.) and not half a mile broad. The small islets to the south of Rat Island are four in number and are HOUTMAN ABROLHOS ISLANDS. 163 obviously part of it. They rise from the same reef-flat and show distinct signs of having been cut off comparatively recently by erosion. All agree in formation and in appearance. Rat Island itself is flat-topped and almost Text-figure 8. 28° 20' PELSART GROUP everywhere exactly 8 feet high. The margins take the form of an undercut cliff, which overhangs very considerably in most places. Here and there are great masses which have fallen and now are being removed by the sea at 164 PROF. W. J. DAKIN: EXPEDITION TO THE high tide. It has been turned over pretty thoroughly by the guano workers, so that there is little material left over the rock and one can examine the latter quite easily at any point. It is a coral limestone which has been so compacted that a very hard homogeneous mass has been produced. Professor Woolnough, who kindly made sections of this material for me, states that it is coral rock completely infiltrated with secondary carbonates and completely solid. I consider that the coral limestone of the central islets is slightly older than any rocks exposed as part of the outer reef islets. The surface of the island is covered with the usual bushes standing about four feet high at most. Many introduced plants occur, the result of the residence of guano workers and fishermen. Numerous holes due to subaérial denudation exist, and two or three of these may be termed wells— they probably contain fresh water throughout the year. It is neither satisfactory, nor plentiful enough, for the few guano workers at present on the islands, who have all their supplies brought in tanks from the mainland. Surrounding the whole of Rat Island is a reef-flat which is increasing in width, for the island is being rapidly eroded to this level. Close to the edge of the reef-flat, just before the deeper waters of the lagoon are reached, is a small boat channel about four feet deep. The section (text-fig. 9) shown illustrates this character. Text-figure 9. Rat Isle Reef Flat Channel High tide level Lagoon The rim islands of the Easter Group and the eastern area of the lagoon have been very closely examined—more so than other parts of the Abrolhos. There are upwards of a dozen islets of varying size, some of which can hardly be said to be separated from one another, for at low tide one can cross with- out wetting the feet, whilst at high tide it is quite easy to wade across. Others are, however, separated by deeper channels. A channel, with 6-10 fathoms of water, exists where we entered the lagoon of the Easter Group (see chart), called Haster passage by the fishermen. Another one occurs just south of Wooded Island where the charts are marked ‘Channel reported.” Between these two passages there ure no deep channels, and one could pass more or less easily from islet to islet, a distance of roughly 5 miles. All these islets agree, as might be expected, in character. They are largely composed of heaped up blocks and fragments of coral. For the greater part these still remain loose and uncemented, or but slightly so. In places, however, a more solid limestone is exposed, as for example on Wooded Island and on the southern end of the islet immediately to the north of it. It is a conglomerate composed chiefly of weathered fragments of coral. The island immediately south of the Easter passage is also largely composed HOUTMAN ABROLHOS ISLANDS. 165 of cemented coral fragments. At no place is the height of these islets of the outer rim as great as that of Rat Island (viz. 8 feet), and, as we have already pointed out, where the land does consist of cemented coral fragments the limestone produced is nothing like so hard or compact as that of Rat Island. The eastern beaches of the islands in every case consist mainly of loose coral blocks, and in places one can see how these are piled in such a manner as to form very definite strata by the regular heaping of these blocks on the northern shore of Wooded Island. In many places, perhaps in most, the surface of the coral blocks is covered with sand and guano. The surface of Wooded Island appears to be of quite clean coral fragments with a sparse covering of shrubs and Mesembryanthemum. I find, however, that the guano workers shifted some tons of guano several years ago from the island, and it is quite probable that they also cleared this material from the general surface. There is a great tendency towards the formation of small or large lakes on these islands. This has proceeded to such an extent that Wooded Island has quite a character of its own ; a lagoon of such extent exists that the island might be called a miniature atoll. Moreover, this internal lagoon attains a considerable depth, for we sounded six fathoms in the centre. The water of all these hollows, whether small or large, is salt, and furthermore is in communication with the sea. Some idea of the mode of formation of the comparatively large lagoonlet on Wooded Island can be deduced from the smaller ones on the other rim islands. Thus, on the island imme- diately south of Haster passage is a pit with six feet of water in it at the deepest part. It is 58 feet wide across the widest part and 100 feet long, being roughly oval in shape. The sides of this hollow overhang and are obviously being eroded at the water-level. On one side a large slab of the cemented coral formation has broken off owing to this undercutting. There is ample evidence that these hollows are produced by the combined action of percolating sea-water and the weather, and in most cases a/ter the formation of the land. By means of a few simple surveying instruments, we were able to show that the level of the sea-water within the hollow was either the same as that of the sea-water outside the island, or within a foot or so of it. Whilst the tide was rising, the level of the water within the hollow was somewhere about one foot below the sea- level. When the tide was falling, the level of the water within was higher than sea-level by about the same amount. The communication with the sea is therefore evident, and also the “lag” due to the resistance experienced by the percolating water. Another much larger ‘“‘lake” existed on the island we have just referred to, but it is not so large as that on Wooded Island. In character it is similar to the one already described, and it is obviously produced by the extension in area of such a small one, or the union of several small ones. 166 PROF. W. J. DAKIN: EXPEDITION TO THE The floor of the hollows referred to on the more northerly island was of hard coral, with a small deposit of calcareous mud. Ulva and one or two species of mollusea abound in the water and a fish (a species of Cobbler) was captured in the smaller cavity. It is impossible to say whether these fish could pass in after having attained the size which they measured—7 to 8 inches. They may have reached the interior at a much earlier stage and developed there. In any case they must have passed at some stage through a considerable length of subterranean channel, for it is very unlikely, if not impossible, that human hands could have placed them where they were found. Text-figure 10. Remains of Jetty We have surveyed the rim islet known as Wooded Island and a rough map of the same is given (text-fig. 10) which shows the extent of the remarkable lake or “velu.” * Now this velu is of very considerable depth (6 fathoms), and from certain points of difference between it and the smaller ones mentioned, and also from the fact that very similar ‘‘ holes” exist on the reef-flat in the lagoon close by, I am inclined to believe that this lagoonlet or “velu”’? had developed before Wooded Island rose above high-tide level. In any case, the original velu has been enlarged since, by erosion similar to that taking place in the smallest pits, and probably guano workers have shifted deposits from it where the mangrove trees are found, One very frequently finds mangrove trees growing in these hollows—in fact they are restricted to this position on the islands of the outer rim of the Kaster Group. Several small thickets occur on the shores of the velu of Wooded Isle, and a thicket occurs in a similar depression (which, however, * The term “velu” is used by Stanley Gardiner for the lagoonlets or deep pools some- times found on the islets of the Maldive atolls. Ihave not used the term throughout for the hollows referred to above, owing to an uncertainty as to whether it would be correct to apply it here. The lagoonlet on Wooded Island seems, however, to agree both in formation and appearance with the “ velus” of the Maldives. HOUTMAN ABROLHOS ISLANDS. 167 is without sea-water at any state of tide) on the island immediately north of it. This latter hollow (Pl. 12) has almost been filled up with guano deposits resulting from the hosts of the Lesser Noddy, which breeds only on Wooded Island and the adjacent one. The amount of guano produced by the birds is so great that it is actually causing the death of these mangrove trees on the northerly of the two islets, and quite half the thicket consists of leafless and bleached branches. The nests of the Lesser Noddy are only to be found in the mangrove trees. Since the guano workers meditate an attack on this deposit, and their working months coincide with the breeding season of the birds, it is quite possible that the Lesser Noddy may leave these islands in the future. Ture LAGOON AND SUBMERGED REEFs. The usual reef-flat occurs round the islets of the eastern rim. In fact, this is a continuous structure running at least from Easter passage to the passage south of Wooded Islet, and the islets are only elevations upon it. On the seaward side the reef-flat varies in width, but it has the same character throughout. The surface is fairly smooth, not eroded leaving sharp points, as in some places in the Wallaby Group, where it is more exposed to still water and rain than to the constant wash of the sea at high tide. Large hollows occur here and there, and form shore pools at low tide in which animal life abounds. As a matter of fact, this reef-flat was probably our best shore-collecting ground in the whole of the Abrolhos. On one occasion the sea was calm enough for us to examine the outer margin of the reef from the motor dinghy. At the particular place visited the reef-flat was about 100 feet wide. It rose slightly, not more than a foot, at its extreme margin and then descended to a depth of 153 fathoms in Text-figure 11. Bizz 06 2 == <= ‘a +o oF IN S30 <0 ' J j--3 Sea level ----4-- esofect fromedge of reef Flat ihe depth was 15k fathoms 100 yards. It was not easy to fix our position and make soundings, being shorthanded, and there was just enough swell to make very careful handling of the boat a necessity, especially when near the edge of the reef-flat ; some figures were obtained, hewever, and the diagram indicates the nature of the slope. 168 PROF. W. J. DAKIN: EXPEDITION TO THE Corals were growing actively on the outer edge, and there is evidence that the reef is extending seawards although but slowly. There are signs that the islands of the outer rim on the eastern side of the Haster Group are becoming joined up. There was certainly a deep channel at one time between Wooded Island and the one immediately north of it. At least I judge so from certain peculiar holes and channels amidst the reefs in the lagoon at this point. Now, one can walk across even at high tide, although if a strong wind blew from the east it might be difficult. We shall see later that in the Pelsart Group, which seems to represent a stage to which the Easter Group is leading, there are very few channels indeed and a single long island forms the eastern rim. The lagoon reef-flat differed in character from that of the seaward side. In some places it was covered and hidden by sand or mud. In other places it was eroded in such a way as to be honeycombed everywhere, leaving but rotten coral, so that one had to be prepared when walking for at least one’s foot going through. This was the case along the lagoon side of Wooded Island. Usually the reef-flat ended with an abrupt drop to the depth of the lagoon, instead of the slope to the sea-bottom which is seen on the outside. In some places, as a matter of fact, it overhangs. There were luxuriant coral growths along the lagoon edge of the reef-flat on the shore of Wooded Island, the more massive corals being present as well as several species of Madrepore, Poceillopora, and Montipora, but I do not think the reef-flat is extending lagoonwards at this point. Coral growths occur on the floor of the lagoon between Wooded [sland and Rat Island, although not very luxuriant, and often nothing but dead fragments come up in the dredge. North of this latter region, 7.e. between the islets immediately north of Wooded Island and Rat Island, the bottom of the lagoon consists of fine impalpable coral-mud, and dredge hauls were exceedingly barren. Further north still, between Rat Island and the island immediately south of Haster passage, luxuriant growths of coral occur once more, and in many places submerged rods rise sufficiently high to bea menace to boats drawing 8 feet of water. One of these reefs, situated just to the left after entering the lagoon by the Easter passage, was rather interesting, and some notes may not be out of place here. The reef in question rises up to the surface so that it is exposed at low tide. It is ring-shaped in form, with a broad opening on one side—that directed N.E. towards the nearest island of the outer rim. The sketch given in the text (fig. 12) indicates roughly the form of the reef, its dimensions, and the depths of water about it. The bottom of the small enclosed lagoon is about 30 feet deep, and living coral abounds. It is also to be found on the floor of the entrance, and along the outer margins of the reef. It would appear that this structure resembles the “faros” of the Maldive Islands. It has developed by active coral vrowth from the floor of the lagoon, and it has taken up its present form as HOUTMAN ABROLHOS ISLANDS. 169 a result of natural phenomena related to the growth of reef-building corals. Neither elevation nor subsidence are required to account for it. The only other possible explanation is that this ring-shaped reef is all that remains of a former mass perhaps at one time part of the outer rim islets some distance away from it. In that case erosion would play a more prominent part in the explanation of the present form, but would be governed by phenomena of normal coral growth. One can find other cases in the Pelsart Group where Text-figure 12. 55% afethoms i si fathons small coral growths are taking the form of miniature atolls, and neither subsidence nor the heaping of debris by waves and currents are responsible for the general architecture of these structures. Just as the one mentioned above occurs where coral growths are most luxuriant in the Easter Group lagoon, i.e. near its northern end, so in the Pelsart Group the northern open side of the atoll is the site of the greatest lagoon coral growth. In the latter case it is obvious that the general trend of water in the lagoon is from south to north, and the water is more free from sand and debris te the north where the lagoon is open to the sea. Faunistic Norses on tHe Waster Group. No large mammals occur on Rat Island except those which have been introduced, and mention has already been made of the cat which was intro- duced by the guano workers to keep down the rats. The island has been noted by many ornithologists for the vast number of Noddy Terns which are found regularly every year during the months of October and November nesting upon the bushes. The nests are practically nothing but “stations” on the tops of the shrubs where the birds lay their eggs (PI. 12). Very often one meets with shrubs almost covered with the nests of these birds, the owners refusing to leave them when dis- turbed, unless actually pushed off. Large numbers of Sooty Terns also nest ‘ 170 PROF. W. J. DAKIN: EXPEDITION TO THE on this island, but they prefer the ground just below the bushes. They are more timid than the Noddy Terns, and leave their nests on the approach of an intruder. Mention has already been made of the Lesser Noddy which inhabits the mangrove trees of Wooded Island and the islet to the north of it. In the case of the latter, the mangrove thicket presents a curious appearance at the breeding season. The branches of the trees are covered with picturesque nests of seaweed, the straggling ends of which hang down; the birds are not at all shy, as may be guessed from the photograph (see Pl. 11). Reptiles are much less common in the Easter Group than on the islands of the Wallaby Group, and no snakes are found. Two small species of lizards were seen on Rat Island but not captured. Rat Island was one of the few places where we were troubled with Diptera—another relic of human occupation. The reef-flat surrounding Rat Island did not prove to be a very good collecting ground. Vermetus was common as usual, and rock pools con- taining small fish and some crabs and Macrura were abundant. A living specimen of Physalia was captured in one of them, having drifted in with the previous high tide. Corals and gorgonids occurred on the submerged blocks of the little jetty which runs out for a few yards on the north of the island. Wooded Island coral flats—both the lagoon and seaward flats— proved extremely interesting, and were the home of a rich shore fauna. ‘lhe margin of the lagoon flat, which as we have already pointed out presents an almost vertical or overhanging wall dropping to about 10 fathoms of water, is a beantiful picture of coral growth, a considerable number of species occurring side by side. A very large Crinoid lives in enormous numbers amongst these coral growths, and in many cases half a dozen or so could be picked up in a mass. Their colour was a gorgeous mixture of green and yellow, but unfortunately it left the animal’s body with the greatest ease on preservation, and coloured everything else a rather dirty brownish green ! Another animal which was extraordinarily abundant on this little stretch of lagoon flat (only on the margin) wasa large and beautiful Nudibranch, almost certainly new and allied probably to Dendronotus. The singularity of its occurrence is accentuated by the fact that notwithstanding its abun- dance here, not a specimen was captured anywhere else at the Abrolhos, and on our second visit in 1915 it was just as common at this place as two years before. Large simple Ascidians and a large species of Serpulidee were marked features of this reef-margin. On the seaward side several species of Chitonidee were obtained, Cryptoplax sp. being fairly common. Sea- Urchins (chiefly Lchinometra mathe) occur in large numbers, each individual in a little hollow in the reef-flat which it has excavated and into which it fits. Nudibranchs and Tectibranchs (Aplysia sp.) were HOUTMAN ABROLHOS ISLANDS. 171 extremely common. We could have obtained hundreds of specimens of Aplysia by merely picking them up as we waded in the shallow water. Under the larger coral blocks was a rich ascidian and sponge fauna, and one or two cidarids were obtained. Alpheids and other Macrura abound. Reference must be made also to our dredge hauls. We dredged in the lagoon, but obtained little except on certain patches where we procured several species of Ascidians (simple and compound) not met with elsewhere, and some coral blocks with Brachiopods adhering to them. Against the edge of the reef-flat (lagoon) we obtained little else but fragments of rotten coral, and, as already pointed out, further north we obtained coral mud and nothing else. In the passage south of Wooded Island the dredge brought up great quantities of weed. We have already referred to the alge as being common at other places. As a matter of fact, it appears strange to the writer to find so much coral growth and a rich algal flora in close proximity. Unfortunately I have not yet been able to see a coral reef in real tropical waters—the comparison with the Abrolhes would be interesting in many respects. On our last expedition to the Abrolhos we made a good collection of these algee and they now await examination. THE PELSART GROUP. We have emphasized the fact that the Abrolhos Islands are extremely far south for a rich growth of reefs. It is interesting to note that it is the Pelsart Group—the most southern group of the Abrolhos Islets—which most resembles an atoll, and which has given us our best collecting. It is also the historically interesting region of the Abrolhos. The Pelsart “Group” consists of a more or less triangular lagoon bounded on two sides by a continuous coral reef, and more or less open to the sea on the north. There are three small islets in the lagoon—Gun Island, Middle Island, and Square Island—together with some smaller islets not all properly charted. Deep water is found closer to the encircling reef than to the outer islets of any of the other Abrolhos Groups, and, on the ocean side, depths of over 100 fathoms ure quite near, whilst even on the eastern side the depths are at least two or three fathoms greater than to the east of the more northerly islets. The Western margin of the Pelsart Atoli is formed of a long reef extending for upwards of 14 miles without any breaks, and without any heaping of coral fragments above high-tide level to form islets. This reef rises about one foot so above low-water mark. On the Eastern side of the Pelsart Lagoon there is a typical rim islet-— Pelsart Island. This is the longest in the Abrolhos Group and extends for about eight miles. It is quite narrow, being in some places only a few hundred feet across. At the extreme southern end of Pelsart Island (which 172 PROF. W. J. DAKIN : EXPEDITION TO THE runs roughly N.N.E.-S.S8.W.) the reef runs 8.W. for a little distance, and then curves round to continue in a north-westerly direction as the western part of the encircling reef. Thus, there is no entrance into the lagoon over the encircling reef between the northern extremity of Pelsart Island on the east, and the northern end of the Western Reef (see text-fig. 8). On the North, however, the lagoon is open to the channel between the Pelsart and Easter Groups (the Zeewyk Channel), and the encircling reef is represented by a large number of scattered reefs and islands. There is quite a collection of these at the north-eastern corner of the lagoon (the Mangrove Islands of the Chart). On our second expedition we anchored the lugger amidst these islets and used the motor dinghy for work in the lagoon. On the first expedition we sailed the lugger through the lagoon (beating all the way against the dominant southerly wind) to the southern extremity of Pelsart Island, where we anchored not far from the ruins of a wooden jetty. When still some distance away from the encircling reef one hears the continued thunder of the breaking waves, for the ocean swell comes in unhindered till it meets this coral barrier. One can often see the white breakers when several miles away owing to mirage effects. This effect was noted by Kent, whose observations at the Abrolhos Islands were contined chiefly to the Pelsart Islands, and a half-tone production of a photograph showing these breakers is given in his work ‘The Naturalist in Australia’ (page 132). A steady wash of water is apparently the usual occurrence over the southern and south-western part of the encircling reef into the lagoon, and consequently a current of some strength is met with in the latter, the water running from south to north. As a result, the lagoon water to the north is more free from debris than that towards the southern part of the lagoon. This factor determines to a large extent the position of lagoon coral growths. The lagoon is roughly 11 sea miles across where it is widest, and about 8 miles from north to south. One does not meet with the same depths that are found in the Easter or Wallaby Groups, and over a very considerable part of the area of the lagoon the depth varies round two, three or four fathoms. This applies to the whole of the southern end where the depth is only 6-3 feet, except for a channel which is indicated on the chart and runs S.S8.E. towards the southern end of Pelsart Island. In this channel depths round 10 fathoms are recorded. There is a very considerable difference indeed between the floor of the lagoon in different parts. At the southern end it is flat, with a floor of level coral covered with but little sand, except in patches. Probably the current is too strong for deposition of debris except in certain places. No live coral growths were observed here. The bottom of the lagoon near Pelsart Island, and about opposite the middle of its length, is somewhat similar but with a greater deposit of sand. To the north-east of the lagoon (the part shaded on the map, text-fig. 8, between HOUTMAN ABROLHOS ISLANDS. Weis Square Island and the northern extremity of Pelsart Island), the lagoon is broken up in the most extraordinary way by reefs. The best illustration of their character, that I can think of, is to refer the reader to the well-known “ maze’? constructed of hedges that one meets with in certain of our famous English gardens. Our attempts to find a way through the lagoon at this place, with the motor dinghy, were absolutely the same as those of an unfortunate stranger in a “maze.” The “hedges” of the maze are represented by coral reefs rising vertically from the bottom of the lagoon and having flat tops which stand a little above low-water mark. They are circular, semi- circular, straight, and of all manner of shapes. The water between them is often of considerable depth, 13 to 16 fathoms in fact, and one careers along happily for a little distance only to find perhaps a blind end in front. This means a return and another trial along some other channel. Rich coral growths are found in this north-eastern corner of the lagoon, and, as a matter of fact, this is the only place in the Abrolhos Group where we found fields of Madrepora so close to the surface that they were exposed at very low tide. This observation fits in with Saville Kent’s “chromo” made twenty years ago, but Kent did not know that he had pictured about the only spot in the Abrolhos Group where such a growth of coral prevailed. I am convinced that the greater part of these lagoon reefs in the north- eastern area of the Pelsart Atoll are due directly to coral growth from the bottom of the lagoon, and not to the erosion of an elevated area. Their sides rise almost vertically from the bottom, but growths extend outwards from the top, and frequently overhanging portions break off and fall to the bottom to be overgrown by fresh coral. Wherever coral growth is not taking place one finds erosion, deposition of sand, or a temporary state of equilibrium. The atoll is the result of no single factor, but of a complex interaction of vital, physical, and chemical components. The larger islets of the lagoon were unfortunately (owing to lack of time) but cursorily examined, but some of the smaller ones on the north were visited several times and Pelsart Island itself was examined more closely. The smaller islets consist of coral rocks and debris heaped up by the waves. Some of those in the north, however, seem to have suffered an elevation of a few feet, and this also applies to Pelsart Island. The coral fragments do not remain loose but are compacted together by secondary deposition. A well near the southern end of Pelsart Island enables one to obtain a glimpse of 10-12 feet of rock, all of which is coral conglomerate. Much of Pelsart Island is composed of loose fragments still, and in some places there are enormous quantities of gasteropod shells which have been thrown up. The island is 6-8 feet above sea-level, and sand and guano deposits occur here and there. Guano workers have, however, visited this island, and con- sequently the surface deposits have been somewhat tampered with, 174 PROF, W. J. DAKIN: EXPEDITION TO THE Mangroves rise from the lagoon flat at many points on the shore of Pelsart Island, and they also occur on some of the Northern Islets. This coral flat, however, is not to be found forming the lagoon shore of Pelsart Island along its entire length ; there are many stretches of sandy beach, the sand forming a covering over the limestone flat. There remains for further consideration the encircling coral reef to wind- ward of the lagoon. On approaching this barrier at low tide, and from the lagoon, one draws near to a kind of great natural weir over which the ocean water flows almost constantly. The reef is two or three hundred feet across and the surface is smooth and well worn by the water. Few blocks of loose coral are to be found on it and practically no stones to turn over, except where they lie in a hollow—a rock-pool at low tide. Besides these small pools, however, there are large cavities—great pot-holes. Near the lagoon margin in one or two places were also irregular excavations which were perfect pictures at low tide, for amidst the scattered coral growths were long needle-spined echinoids and thousands of fishes—huge Gropers (Acherodus gouldii) and other species which probably enjoyed the seclusion of these cavities so well supplied with food. The antennz of the large Crayfish Panulirus penicillatus projected everywhere from the crannies. We have already remarked that there are no islets on the western margin of the atoll. It would appear as if the ocean swell was too powerful to allow any heaping of coral fragments to take place. Broken pieces of coral must be constantly thrown on the reef from the outer face and yet there is no collection of this on the reef itself. The lagoon is noticeably more shallow near this windward side than against Pelsart Island to leeward. Some idea of the force of the waves is apparent from the fact that half the engine-room and some other sections of a ship, the ‘ Windsor,’ now lie littered on this reef not far from the southern extremity of Pelsart Island. They have been picked up from the bottom outside the reef and thrown on top of it. Tur Fauna oF THE PELSART GROUP. There is little to add regarding the terrestrial fauna of the islands. As might be expected, it is poorer than that of the islands of the more northern groups. The bird fauna alone is abundant, and many species were found nesting in considerable numbers, including Mutton-birds, Sooty Terns, Noddy Terns, and Ospreys. Beyond a small lizard, which occurred also on Rat Island, no other vertebrates were observed. The marine fauna, on the other hand, was decidedly interesting. In this connection we may mention certain dredge hauls taken in the Zeewyk Channel just north of the Pelsart Group. The bottom was sandy with lumps of worn nullipore and shells. The dredge brought up hundreds of Holothurians of two or three species, one being particularly common. Large quantities of Cidarids were common at the same place, but there were few HOUTMAN ABROLHOS ISLANDS. 15 star-fish and these practically all Ophiuroidea. In contrast to the hauls at almost all other places, we-obtained but few sponges and fewer ascidians. Our most interesting finds were obtained by shore-collecting on the outer reef, starting from the southern extreme end of Pelsart Island and working south and west. At this spot we obtained the first specimens of any Enteropneust to be recorded from the West coast of Australia. They were found in the coarse sand which lies in a few places, sheltered in shallow hollows on the reef-flat close to the lagoon. On both expeditions we obtained them by sifting the sand through the fingers, and nowhere else but here on the Pelsart Reef were we successful in finding any. The animals were translucent and of a pale yellow colour. They varied in size, one of the larger specimens measuring 12 centimetres when somewhat con- tracted. After fixation very great .contraction takes place and dimensions of fixed specimens are of little value. The species has been described by the author in another paper, and is regarded as a new species of the genus Ptychodera allied to P. flava. It has been called Ptychodera pelsarti after the famous voyager whose ship was wrecked on the group now bearing his name”*. On the same reef, in close proximity to the spot where Ptychodera occurs, we found a large species of Pagurid particularly common—each shell bearing three or four large anemones. The gephyrean Boneilia was found for the first time on the coast of West Australia. It has since turned up several times on the coast, off Fremantle, but the species has not yet been investigated. The Pelsart Atoll seems to bea particularly happy ground for Echinodermata. Several species of Hchini occur on the floor of the lagoon, each occupying its own area. A very fine species, resembling the English chinus esculentus in size, occurs in immense numbers on the floor of the lagoon at the southern end of the atoll. In another region not far from this we passed over large numbers of Asterids. Very few Aleyonaria have been obtained from the Abrolhos, and so far as our collections go these were all obtained on the Pelsart Group. Some were collected on the reef near its extreme southern point, whilst others were abundant on the northern shores of the islets termed the Mangrove Islands on the chart. Some of these islands, together with the coral reefs, form a small well- sheltered area known by the fishermen as Whale’s Bay. It is entered by a narrow channel from the north, and, as it is quite close to the open sea and in a very suitable position for departures to Geraldton, it is well frequented by the fishermen, who anchor there for the night. I was informed that the spot was well known to the fishermen as a place to which whales resorted to scrape off their attached barnacles. The skipper of our lugger said that they had on more than one occasion had a whale scraping itself against the anchor * Journ. Linn. Soc., Zool. xxxiii. (1916) pp. 85-100, pls. 10, 11. LINN. JOURN.—ZOOLOQY, VOL, XXXIy. 165} 176 PROF. W. J. DAKIN : EXPEDITION TO THE chain. I was rather sceptical at first about these stories, but am quite certain now that whales do resort to this small enclosed area for this purpose, for we had visitors whilst I was there. As it is not easy to find the entrance, and the area is but small, any other good reason for the frequency of whales in this particular spot seems to me entirely wanting. The visitors whilst we were there were specimens of the Humpback. A detailed report on the general faunistic characters of the Abrolhos Islands will be written up after the various collections have been investigated by specialists. CONCLUSION. The Coral Formations of the Abrolhos Islands. It will be convenient if our observations on the coral reefs of the Abrolhos Archipelago are brought together in the form of a theory relating to the origin of these islands. To this end it will be advisable in the first place to repeat in summarised form certain observations that have been made. They are as follows :— 1. The islands consist entirely of limestones for the greater part composed of recent corals cemented together by secondary deposits of carbonates. In places, as, for example, the lowest rock exposures on West Wallaby Island, the limestone consists of foraminifera, echinoid spine fragments, and nullipore fragments, cemented together by secondary carbonates into a compact rock. The corals and molluses are all recent and shallow-water species. No signs of any plutonic rock, or other rock older than the limestone mentioned above, are to be found. 2. The more northerly islands, 2. e., the large central islands of the Wallaby Group, show signs of considerable elevation, cliffs of limestone rising to heights of 30 feet. The islands of the southern groups are not nearly so high. 3. All the island groups show evidence of a more recent elevation of about 8 feet. This is marked by the terrace on the Hast and West Wallaby Islands, and by the uniform flat worn surface of other islands which rise to about the height of this terrace, i.e, Rat Island, some parts of Wooded Island, Gun Island, Pelsart Island (part), and some of the Mangrove Islands. 4. The most southern group of the Abrolhos Islands, the Pelsart Group, takes the form of an atoll. The lagoon depths are moderate, 17 fathoms being the greatest recorded, and the depths on the outside of the encircling reef are not very great, averaging probably about 30 fathoms, but not less on the west side where the depths are considerable a few miles away. This absence of depth is due to the fact that the atoll rises from the continental shelf, which is uniformly level over a considerable area. Had this sea- bottom sloped more rapidly, there is no doubt that the outward growth of the HOUTMAN ABROLHOS ISLANDS. 177 encircling reefs with talus formation would have resulted in greater depths on the western side of the encircling reef. In this connection it is worthy of note that Darwin in an appendix to his famous work, ‘Coral Reefs,’ p. 230, refers to the Abrolhos, about which it is true he had very little information, in the following terms :— ‘“Houtman’s Abrolhos (lat. 28° S. on west coast) have lately been surveyed by Captain Wickham (as described in Naut. Mag. 1841, p. 440) ..... From the extreme irregularity of these reefs with their lagoons, and from their position on a bank, the usual depth of which is only 30 fathoms, I have not ventured to class them with atolls, and hence have left them uncoloured.”’* The reefs of the Pelsart Atoll cannot be said to be arranged irregularly, and the fact that the islands rise from a bank is no reason against terming any of the groups an atoll, unless we argue that every atoll must be formed according to Darwin’s views. 5. I consider that the Wallaby and Haster Groups represent stages in the formation of such a type as the so-called Pelsart Group. Haster Group, for example, can be compared with an atoll having still a large island in the centre of the lagoon. The Wallaby Group is least like an atoll and consists of a group of islets, the remains of a large limestone mass, surrounded by irregularly arranged reefs. North Island is still practically only an elevated limestone topped with sand dunes and rising from a reef-flat which extends seawards. Lagoon formation is only commencing on its eastern side, and encircling reefs cannot be said to exist. It is possible that North Island will develop along the lines indicated by the other groups. There are several interesting points of resemblance between the Wallaby, Easter, and Pelsart Groups. In the first place, the large openings into the lagoons are always found to the north. This is the only way into the Pelsart Atoll. Both the other groups can be entered from the south, but careful navigation of the channels is required. We find a long rim islet without a break, in the Pelsart Group, and we have noticed that there is evidence of confluence of the rim islets in the Easter Group to bring about a similar condition. More primitive conditions prevail in the Wallaby Group and the rim islets are not even regularly placed. It will be noticed that rim islets are developing on the eastern, the leeward, side zn each group. 6. Judging from the presence of certain elements in the fauna of the Wallaby Islands—the Wallabies, and the Amphibia—it is extremely probable that land connection existed between the mainland of Australia and this part of the Abrolhos Group at least. The foundation of the Abrolhos Groups is probably Tertiary limestone, but of this we have no yery direct evidence. It is worthy of note, that north of * Uncoloured on the map given by Darwin. 178 PROF, W. J. DAKIN: EXPEDITION TO THE the Abrolhos Islands the sea is deeper at a similar distance from the coast. In other words, the Abrolhos Islands rise from a bank connected with the coast. We must start therefore from the stage when limestones formed largely of corals, foraminifera, mollusca, and calcareous algse were elevated in this region and united to the mainland of Australia. In fact, the west coast of Australia may be said to have been brought much further west. This elevated limestone was weathered and eroded, and it is very likely thatthe channels between the present groups of islands bear some relation to the presence of small rivers, the Hutt River, the Bowes River, and the Chapman River, opening on the adjacent coast. During this period the Wallabies and Amphibia inhabited the extreme western region of the coast now represented by the Hast and West Wallaby Islands. This period of elevation was followed by one of depression or subsidence, and of this there is ample evidence on the present West Australian coast. Jutson* states : “In the vicinity of Perth the coast-line was formerly farther west, as indicated by the drowned valley of the Swan River.” Further examples occur in the south-west and also to the north. Thus the same author states, “The North coast is the most broken in West Australia. Deep sounds and bays run far into the land, and are manifestly the drowned continuations of the adjacent rivers. This region has been deeply dissected, probably to maturity, and then submerged. It thus affords excellent illustrations of drowned valleys.” This subsidence may be considered to have separated the Abrolhos Islands from the mainland. It is also possible that a fault running approxi- mately N. and §. occurs between the islands and the mainland, and has helped to bring about separation, for there are several faults running parallel with the coast in this region. In any case subsidence together with erosion resulted in the separation of probably four land masses from the mainland, the ancestors of the present four groups. We have now four islands of coral limestone, and we may assume that either the altitude of the more northern ones was higher than that of the others before subsidence, or else that the subsidence was not uniform and the southern island was depressed most. This is to account for the much greater altitude of the Wallaby Isles to-day. Around the original islands encircling reefs have developed, whilst ai, the same time erosion has taken place and tended to remove them, lagoons taking their place. Such erosion has produced least effect so far in the Wallaby and North Island Groups, probably owing to their greater elevation to start with. Evidence of this erosion is everywhere present. The East and West Wallaby Islands have been separated, and a series of small islets to the east of them are all remnants of the original central mass. An excellent * Jutson. “An Outline of the Physiography of Western Australia.” Bull. 61. Geol. Surv. West. Australia. Perth, 1914. JOURN. LINN. Soc., ZOOL. VOL. XXXIV. PL. 10. DAKIN AILYNL “44110 ANOLS3WIT Py em wt ie JOURN, LINN. Soc., ZOOL. VoL. XXXIV. PL. 11. DakIN. “SSADYONVIA NO ONILSAN AGGON Y3SS37 Wem So) ena a ane ZOOL. VOL, XXXIV. PL. 12. JOURN. LINN. SOC,, DAKIN, “SNYAL AGGON DAKIN. JOURN. LINN, Soc., ZOOL. VOL. XXXIV. PL. 13. Fre. 4, ]et@eee, ABROLHOS ISLANDS. od Dakin. JOURN. LINN. Soc., ZOOL. VOL. XXXIV. FL. 14. oh 23 3h, See Fra. 6. ABROLHOS ISLANDS. HOUTMAN ABROLHOS ISLANDS. 179 illustration of this erosion is left on the reef-flat to the west of West Wallaby Island and is depicted (Plate 14. fig. 7). It consists of a limestone mass which overhangs on all sides, and agrees entirely in structure with the rock of the adjacent island a few hundred feet away. A few more years and it also will have gone. Erosion has done much more in the Haster Group, where the original central mass is now represented by Rat Island. The lagoon is deep, particularly on the eastern side, and there are several islets on the encircling reef to the east, of which Wooded Island is an excellent example. These islands consist for the greater part of heaped coral fragments, loose or cemented into a conglomerate, and very frequently lagoonlets are present on them. In the Pelsart Group the central mass is now represented by only small islets, and they are few in number. The outer reef is much more regular than in the other groups, and is practically unbroken on two sides. So far, we have emphasized the action of the sea in producing the lagoons. We must add that in some places the reefs are progressing seawards on the outward margins by coral growth, whilst the area of the lagoons is increasing. Small reefs occur irregularly in all the lagoons. In some places they represent portions of the original limestone mass; in others, as for example near the Easter Passage and at the N.H. end of the Pelsart lagoon, they have grown up from the floor of the lagoon. Thus in the formation of the lagoons at the Abrolhos Islands the greatest factor has been that of erosion and solution. Coral growth has modified this action in places, and the heaping of coral blocks by waves and currents must be also taken into consideration. Subsidence has played practically no part in lagoon formation. As to the future, the East and West Wallaby Islands and their adjoining fragments will be probably still further redaced. Rat Island in a similar manner will be further eroded, and the Easter Group become still more like a typical atoll. The depth of the Pelsart lagoon at the extreme southern end will probably decrease owing to deposition of material washed over the encircling reef. APPENDIX: In the concluding section immediately preceding this, a short summary of the observations on the coral formations of the Abrolhos Islands has been given. To this we may add a short summary of other data. 1. The Perey Sladen Trust Expeditions to the Abrolhos have been arranged for the purpose of obtaining information regarding the formation of the islands, the conditions permitting of coral-reef formations in this southern 180 SLADEN TRUST EXPEDITION TO ABROLHOS ISLANDS. latitude (the Abrolhos are probably the most southern coral islands in the world), and the fauna of this region of the southern Indian Ocean. 2. There is evidence that the temperature of the sea at the Abrolhos is usually some degrees higher than at the adjacent coast, and that as a result the temperature during the winter months rarely falls below 20° C. 3. In consequence of this hydrographical condition, the temperature of the water at the Abrolhos only very exceptionally falls below a point that is detrimental to the active growth of reef-building corals. 4. The difference in temperature between the coastal waters and those 50 miles out, is due toa tropical current from the North and North-East, which passes down the coast but is deflected away from it, leaving a strip of colder water against the land. 5. Any phenomenal preponderance of tropical species in the marine fauna of the Abrolhos region of the Indian Ocean is due to this tropical current. It remains to be seen, when the collections are worked up, to what extent the marine fauna is a mixture of Tropical and Southern types*. 6. A great blank in our knowledge of the distribution of marine organisms in the southern Indian Ocean will be filled up by the study of these collections from the Abrolhos region in conjunction with those to be obtained from an expedition to the north-western coast of Australia, which is to follow as part of this investigation. The collections so far examined give every indication of this, as for example the record of a new Enteropneust, and many species of other groups for the first time on the West Australian coast. EXPLANATION OF THE PLATES. Puiate 10. [ Fig. 1.] Face of Limestone cliff, Turtle Bay, East Wallaby Island. PuareE 11, [Fig. 2.] Lesser Noddy, on nest in Mangrove Swamp. Pratt 12, [Fig. 8.] Noddy Terns nesting on Rat Island. PLATE 13. Fig. 4. North-eastern coast of Hast Wallaby Island. Fig. 5. Limestone cliffs, Turtle Bay, East Wallaby Island. PLATE 14. Fig. 6, Undercut Limestone, North Island. Fig. 7. Limestone “ Pinnacle” on reef-flat, West Wallaby Island. * The fishes, at least, appear to be chiefly Southern forms. RESTORATIONS OF THE HEAD OF OSTEOLEPIS. 181 Restorations of the Head of Osteolepis. By Hpwin 8. Gooprica, F.R.S., Zool. Sec. L.8., Fellow of Merton College, Oxford. (With 6 Text-figures.) [Read 17th January, 1918.) THE genus Osteolepis is one of the commonest and best-known of the early Devonian Teleostomes. Nevertheless the exact disposition and homology of the superficial bones of the head are still but imperfectly understood. Two recently published restorations of the head, one by Gregory (6) and the other by Watson and Day (14), differ so remarkably from each other that it seemed advisable to reinvestigate the subject; for the Osteolepidee are a very interesting and important group. In many respects, as for instance in the form of the paired fins and in the cosmoid structure of the scales, they approach the Devonian Dipnoi such as Dipterus (Goodrich, 4, 5) ; while, on the other hand, the skull shows undoubted affinity with the early Amphibia (Stegocephali). Huxley, I think unfortunately, in his valuable work on the fishes of the Devonian epoch (7), included Ostgolepis and its fossil relatives together with Polypterus in the one group Crossopterygii. Polypterus, however, as I have endeavoured to show elsewhere (5), really differs funda- mentally in its structure from the Osteolepids, and is almost certainly an aberrant Actinopterygian preserving some primitive characters. But, how- ever this may be, there can be little doubt that the Osteolepidée have departed less from the structure of the common ancestor of the Teleostomi and Tetrapoda than any other known fish. A thorough understanding of the structure of the skull of Osteolepis is, therefore, of the greatest importance for the elucidation of the homologies of the bones in the higher Vertebrates. The most complete restorations of the skull of Osteolepis yet published are those given by Pander in his well-known monograph (10). But, beautiful as are his figures and excellent his reconstructions, they can by no means be trusted in every detail. Since then, Traquair has contributed a very good restoration of the whole fish (13), without detailed figures of the head ; and Smith Woodward in his text-book (15) has given a figure of the roof of the skull, which is in all essentials correct. As already mentioned above, in his interesting discussion of the origin of the Tetrapoda (6) W.K. Gregory has figured restorations of Osteolepis microlepidotus, admit- tedly based on Pander’s work ; while D, M. 8. Watson and H. Day, in their valuable paper on “‘ Paleeozoie Iishes” (14), restore Osteolepis macrolepidotus. Three sets of these figures from Pander, Gregory, and Watson & Day are here reproduced for comparison (figs. 2, 3, 4, & 5). 182 MR. FE. 8S. GOODRICH : RESTORATIONS OF For my own restorations (figs. 1 A & B) I have not only had the benefit of the work of my predecessors, but have been able to study a large number of excellent specimens in various collections. I have to thank the authorities of the Museum of Practical Geology, Jermyn Street, for affording me every facility for the study of the fine series in their keeping, also Prof. Sollas for the loan of specimens in the Geological Department of the Oxford Fie. 1A. 2 soe Restoration of the head of Osteolepis macrolepidotus, Ag. Dorsal view, enlarged. (For lettering see fig. 1 B.) University Museum; but more especially Dr. A. Smith Woodward for his unfailing kindness and helpfulness during my frequent visits to the British Museum of Natural History. As for the general disposition of the skull-plates covered with cosmine the figures speak for themselves, and only certain points of doubt or disagree- ment need be discussed in detail, First of all it may be mentioned that, THE HEAD OF OSTEOLEPIS. 183 although my restorations apply more particularly to O. macrolepidotus 8 pply 1} ’ et I find no important difference between that species and O. microlepidotus. I t ] The most careful scrutiny has convinced me that the supposed transverse ) ny series of small plates behind the parietals figured in O. microlepidotus by Pander, and accepted by Gregory, are not separate elements. They are ) gory, y; parts of the supratemporals end parietals, and the often incomplete lines Fie. 1B. Restoration of the head of Osteolepis macrolepidotus, Ag. Left side view. a, angular; c, main lateral-line canal; d, dentary ; e, ethmoid included in rostral shield; fr, frontals fused in middle line and enclosing the pineal opening; zoc, infra-orbital canal; it, intertemporal; 7, jugal; /, lacrymal; dg, lateral gular; mdc, mandibular canal; my, median gular; mx, maxilla; 7, nostril; za, nasal included in rostral shield ; 0, orbit; op, opercular; p, pineal opening ; pop, preopercular ; pope, preopercular canal ; ptf, postfrontal ; pto, postorbital ; 1s, rostral shield ; sc, scale-like plate overhanging hyo- mandibular; so, supraorbital; soc, supraorbital canal; sop, subopercular; spo, dermal supraoccipital or postparietal ; sq, squamosal; sge, squamosal canal; st, supratemporal or pterotic; sfe, supratemporal canal; tb, tabulare ; toc, transverse occipital canal ; vw, ventral paired gular. which were supposed to be sutures marking them off from these bones are merely superficial grooves involving only the cosmine layer, and possibly indicating the presence of rows of small pit-organs or some other sensory structures. I can find no evidence of the existence of such a transverse series in any other Osteolepid ; and an examination of the under surface LINN. JOURN.—ZOOLOGY, VOL. XXXIV. 14 184 MR. E. §. GOODRICH : RESTORATIONS OF shows in this region the bone stretching uninterruptedly from three centres of ossification on each side, and three only : the parietal, the intertemporal, and the supratemporal. Grooves similar to those just mentioned are found on the frontals, diverging from the hinder border (as figured by Pander, see fig. 5), on the squamosal, and on some of the trunk scales running between the lateral-line scales and the mid-dorsal line. The diverging lines on the frontals do not mark off the postfrontals as suggested by Gregory ; these bones are distinct but small, harbour the lateral-line canal, and seem just to reach the edge of the orbit. A distinct intertemporal is present. A suture separates the two parietals; but the fronials are fused superficially, although the suture can be distinguished on the inner surface. In one specimen a small plate and a minute nodule, possibly transparent in the living, were situated in the pineal foramen. Jaekel figures several such plates in Diplopterus (8). In front of the frontals the snout is covered by a shield formed of small plates the outlines of which can still be distinguished Fig. 2. re. 8. Fig. 2. Restoration of the dorsal, A, and right side, B, of the head of Osteolepis micro- lepidotus, Ag., from Gregory (6). Fig. 3. Restoration of the dorsal, A, and right side, B, of the head of Osteolepis macro- lepidotus, Ag., from Watson & Day (14). Lettering for figs. 2 & 3:— § D, dentary ; D.So., dermal supraoccipital; F & Fr, frontals; Gu, gular; Lop., inter- opercular; Jt, intertemporal ; Ju, jugal ; Za, lacrymal; Ma, maxilla; m, nostril: NA, nasal; OP, opercular; PA, parietal; Pf, prefrontal; P.O. & Po.O., postorbital ; Po.Fr. & Pof., postfrontal; P.P., postparietal; Q.j., quadratojugal; Sp, splenial ; Sq, squamosal; S.OP., subopercular; S.7, supratemporal; TAB & 7%, tabulare. to some extent on the inner surface. Unlike Watson and Day, I find here a fairly large median ethmoid between the nasals. Almost at the extreme lateral edge of the rounded rostrum can usually be seen an external nostril, which in some specimens is certainly surrounded by bone; while the position of the lower or posterior nostril is perhaps indicated by a notch on the THE HEAD OF OSTEOLEPIS. 185 recurved ventral border of the shield. At the hinder corner of the skull, between the squamosal, tabulare and supratemporal, is a small triangular plate, figured by Pander and Gregory but not by Watson and Day. This scale-like plate seems to have been attached to the supratemporal and possibly overhung a spiracular opening. Below it passed the head of the large curved hyomandibular (traces of which can often be seen) to articulate with the supratemporal. The lateral aspect of the head is by far the most difficult to interpret since the bones are here almost always much crushed and displaced, especially near the articulation of the jaws. . Nevertheless, there can be no doubt of the existence of a large curved squamosal covering the whole of the cheek behind the postorbital and jugal. It has been well figured by Watson and Day, but ' Pander seems to have considered that it consisted of three plates, being partly misled by the curved superficial groove shown in figure 1 B. The squamosal reaches down to the articulation of the jaws, covering the quadrate region. Just behind its hinder edge, and overlying the end of the hyomandibular, is a small plate, the preopercular. This element is not figured by Watson and Day, but is possibly the one drawn by Pander and designated quadratojugal Fig. 4, Fia. 5. Fig, 4. Diagram of the bones of the skull of Osteolepis macrolepidotus, from Pander. Fig. 5. Restoration of head of Osteolepis microlepidotus, from Pander. by Gregory. If so, its true position has not been correctly determined by these authors. For a long time I doubted the existence of this separate element which is almost always considerably displaced ; but after very careful examination of the most favourable material was finally convinced of its presence. ‘The conspicuous large opercular bones pass ventrally into a series of lateral gular or bramchiostegal plates, which end in a pointed anterior element wedged in between the ventral gular and the lower jaw. Pander’s restoration of the lower jaw appears to me much more correct than that of Watson and Day. No trace could be found of the series of intradentaries figured by the latter authors. The dentary and angular are obvious; an ‘opercular’ or prearticular plate seems to have covered part of the inner 186 MR. E. S. GOODRICH: RESTORATIONS OF surface, while further forward there are indications of a splenial showing on the lower outer surface. Of the existence of this element, however, I could not make certain. Special attention was devoted to the course of the lateral-line system. It is not included in the figures of Watson and Day ; but Pander studied it with considerable success. Recently it has again been figured by Collinge (3) without, however, adding much to Pander’s results. These authors seem to have indicated in their restorations the distribution of the lateral-line pores rather than of the canalitself. The distribution of the two is by no means always the same, since the pores often stray far from the canal with which they are connected by delicate branches, and may be dotted about somewhat irregularly. The double broken line in my figures indicates the course of the canals only, as they are often beautifully revealed in the fossils. The main lateral-line canal passing from the body scales enters the hinder region of the tabulare (often called supratemporal), runs forward through the supratemporal (pterotic), intertemporal, and postfrontal. About the middle of the tabulare it gives off a transverse occipital branch which joins its fellow from the opposite side in the median dermai supraoccipital (post- parietal). That portion of the canal which lies between the origin of the transverse occipital and the origin of the infraorbital branch may be called the temporal canal. 1t is generally considered to belong to the infraorbital. In the postfrontal the canal branches intw an upper supraorbital and a lower infraorbital canal. The former proceeds along the margin of the frontal to the rostral shield, where it describes an elegant curve and appears to end close to the nostril. I could find no anterior commissure; if such exists it must be on the ventral surface of the snout. The hitherto unrecognized jynction of the supraorbital canal in the postfrontal bone with the infraorbital canal which runs up through the postorbital from the jugal I have been able to trace quite clearly. In the jugal a horizontal branch is given off from the infraorbital canal; it. passes backwards across the squamosal, and apparently joins the preopereular canal which enters the angular, and proceeds forwards to the front end of the dentary. These lateral-line canals are of special importance when comparing the skull of fishes with that of the primitive Tetrapoda. Pollard (11), Baur (2), Allis (1), and especially Moodie (9), have already made use of the canals in Amia and Polypterus in comparison with the Stegocephali, where the course of the lateral line is often marked by grooves. But these modern fish are specialized in many respects, and the pattern of Osteolepis agrees more closely with the Stegocephatian (see Moodie’s figures, 9). However, it is not proposed to enter into a detailed discussion of the comparison in this paper, but it may be pointed out that the horizontal squamosal canal mentioned above is characteristic also of the Stegocephali. It is possibly THE HEAD OF OSTEOLEPIS. 187 represented by the ‘hyomandibular canal’ of Selachians. A further study of this region in the Stegocephali might enable us to determine the homology of the lateral bones of the skull, especially of the quadratojugal, which is still very obscure. Judging from modern forms, the preopercular canal must have been supplied by a postspiracular branch of the hyomandibular nerve ; we should not, therefore, expect to find it and the bone containing it in the skull of an Amphibian. Although a canal is known to have been present in the lower jaw of some Stegocephalians, its connexion with the more dorsal canals does not seem to have been made out. The settlement of this important point might help us to determine whether the preopercular is really represented in the Amphibian skull or not. In conclusion attention may be drawn to the remarkable uniformity in the structure of the skull among the Osteolepide. Thursius, Diplopterus, and even Megalichthys seem to differ in no important respect from Osteolepis in the number and disposition ef the bones on the head. The Rhizodontidee also closely resemble the Osteolepids ; and, except for the presence of infra- dentaries and for the subdivision of the squamosal, the restorations of Rhizodopsis made by Traquair (12), that most accurate of observers, would serve almost equally well for Osteolepis, as may easily be seen by comparing his figures with mine. List of References. 1. Auuis, E. P.: On Certain Homologies of the .... Bones of Amia calvu. Anat. Anz. vol. xvi., 1899. 2. Baur, G.: The Stegocephali. Anat. Anz. vol. xi., 1896. 3. Cottince, W. E.: The Sensory Canal System in some Fossil Fishes. Proc. Birmingham Phil. Soc. vol..ix., 1895. 4. Goopricu, E.8.: On the Scales of Fish. Proc. Zool. Soe., 1908. 5. —— Vertebrata Craniata, Part 9 of a ‘Treatise on Zoology,’ 1909. Grecory, W. K.: The Problem of the Origin of the Tetrapoda. Ann. New York Acad. Sci. vol. xxvi., 1915. 7. Huxuny, T. H.: The Systematic Arrangement of the Fishes of the Devonian Epoch. Mem. Geol. Survey, 1861. : 8. JAEKEL, O.: Ueber die Epiphyse u. Hypophyse. Sitz. Gesell. Naturf. Fr. Berlin, 1908. 9. Mooniz, R. 8.: Lateral Line of Extinct Amphibia. Journ. Morph. vol. xix., 1908. 10. Panprr, C. H.: Die Saurodipterinen.des Devon. Systems, 1860. 11. Pontarp, H. B.: On the Anatomy and Phylogical Position of Polypterus Zool. Jahrb., Abt. Anat. vol. v., 1892. 188 RESTORATIONS OF THE HEAD OF OSTEOLEPIS. 12. Traquair, R. H.: Cranial Osteology of Rhizodopsis. Trans. Roy. Soe. Edinburgh, vol. xxx., 1881. 13. —— HExtinct Vertebrata of the Moray Firth Area, ‘Fauna of Moray Basin,’ 1896. 14. Warson, D. M. §., and Day, H.: Notes on some Paleozoic Fishes. Mem. Manchester Lit. Phil. Soc., 1916. 15. Woopwarp, A. SurrH: Vertebrate Paleontology, 1898. LINNEAN SOCIETY OF LONDON. RULES FOR BORROWING BOOKS FROM THE LIBRARY. 1. No more than Six volumes shall be lent to one person at the same time without the special leave of the Council or one of the Secretaries. 2. All books shall be returned before the expiration of Six weeks from the time of their being taken out, but if not required by any other Fellow, they may, on application, be kept for a further period of Six weeks. 3. All books lent shall be regularly entered by the Librarian in a book appropriated for that purpose. 4. No work forming part of Linneus’s own Library shall be lent out of the Library under any circumstances. Norze.— Certain other works are included im this prohibition, such as costly illustrated works, volumes belonging to sets which could not be replaced if lost, und unbound parts of serials. 5. All carriage charges on books to be defrayed by the Fellow borrowing the same. 6. All borrowed books when returned to be sufficiently wrapped and protected. Borrowers will be held liable for any damage incurred owing to improper packing. 7. Newly published books shall not ordinarily leave the Library until after the expiration of one month from the date of their being received. A GENERAL INDEX to the first twenty Volumes of the Journal (Zoology) may be had on application, either in cloth or in sheets for binding. Price to Fellows, 15s.; to the Public, 20s. NOTICES. Tue attention of the Fellows, and of Librarians of other Societies, is requested to the fact that TWO volumes of the Journal (Zoology) are in course of simultaneous issue, as follows :— Vou. 33. Nos. 222 and 223; No. 224 is in the Press, and will complete the volume. Vou. 84, Nos. 225 and 226, the present number. Authors are entitled to 25 copies of their communications gratuitously, and may obtain another 75 by payment, as shown on the — printed slip which accompanies the proof. If more than 100 copies are wanted, application must be made to the Council. Abstracts of the proceedings at each General Meeting and Agenda {;, for the next are supplied to all Fellows, B. DAYDON JACKSON, General Secretary. Marcu 30, 1920. Price 14s. THE JOURNAL OF THE LINNEAN SOCIETY. Vout. XXXIV. ZOOLOGY. No. 227. CONTENTS. Page I. A Post-Puerulus Stage of Jasus lalandii (Milne Edw.), Ortmann. By J. D. F. Ganountst, M.A., D.Sec., Ph.D., F.L.S., Professor of Zoology in the University of Cen Town. (Plates 15 & 16, Giniel Ula) AES er SGA) AessaodandeonbestadeacoRqhen nae toncacoorasoacnnusse 189 II. On the Linnean Species of Non-Marine Mollusca that are repre- sented in the British Fauna, with Notes on the Specimens of these and other British forms in the Linnean Collection, By A. 8. Krennarp, F.G.S., and B. B. Woopwarp, F.LS. ......... 203 III. The Haogonee. By W. A. Haswett, M.A., D.Sc., FR.S., F.L.S., Wmeritus Professor of Biology, University of Sydney. (Plates eda Srands 2 Moxten ouness) ee. cenasit icone cer toss omeeeceesen. V6. 10.—Mandible of post-puerulus, Fra. 11.—Mandible of puerulus. p., mandibular palp; ¢. 1, ¢. 2, ¢.38, teeth of mandible. The first and second of these, counting from the side next the mandibular palp, are usually close together, though, in some cases, all three are about equidistant. The third being situated somewhat internal to the other two, though still forming part of ihe cutting-edge, I formerly suggested (4) might prove to become he molar or canting part of the adult mandible. An examination of the post-puerulus does not, however, entirely confirm this. ‘The changes which have taken place are illustrated by figs. 10 and 11, which are camera lucida drawings of the shed cuticle of a puerulus and the mandible of the post-puerulus stage, which takes its place, both from the same specimen. The mandibular palp (p.) is now fully formed and is pro- vided with three segments and sete. The beginning of the mandibular spine is also indicated but only by a clear line in the cuticle, which will form POST-PUERULUS STAGE OF JASUS LALANDII. 197 the hinge of the mandible. The three prominences or teeth of the mandible are more marked, and a cutting-edge on which there is a series of small denticles of irregular shape, has appeared between the second and third. A few of these appear also on the first and second tooth in some specimens, not in all, and, at their base, a large one was in all cases found, ending in three sharp points. These denticles are readily distinguished from the rest of the thickened cuticle, on which they lie, as they are of a dark brown colour. Doubtless their appearance is associated with the change in the feeding habits of the animal. They disappear in the subsequent stages, in which the mandible has become calcified. The change in the third tooth is marked. From a mere undulation of the cuticle in the puerulus it has now become prominent and conical, with a cleft apex, and is situated behind the cutting denticulate outer margin of the mandible. The part which it takes in the formation of the adult mandible is now apparent, for in the latter the main part of the molar surface is formed by a cleft or V-shaped prominence, which is apparently the transformed third tooth. The inner limb or ridge of the prominence forms the thick posterior margin of the molar surface. The whole of the molar surface is, however, not formed from the third tooth; for, behind the second tooth, there is a ridge running backwards, which fades off into a rounded prominence, thus forming the rest of the molar surface. The second tooth therefore undergoes a transformation similar to that which occurs in the third, for it becomes a V-shaped ridge, one limb of which forms the outer cutting-edge of the mandible and the other the molar ridge just mentioned. The bifurcate characters of the second and third teeth are better seen in the early calcified stages than in the adult where, however, they can readily be made out. Thus in a small crawfish 27 mm. in length, in which the mandibles only are calcified, and in a later stage 32 mm. in length, in which both mandibles and cuticle of the body generally are calcified, these two V-shaped structures are readily distinguishable. (7) Antennular screen. Another marked change in the transition from the puerulus to the post- puerulus is the disappearance of a bundle of long feathered sete, arranged as a fan-shaped structure, on the upper side of the distal end of the first segment of the antennule. It is not readily seen in a surface view, as it lies at right angles to the axis of the antennule, and it is obscured by the thick antennz in a lateral view. If, however, the antennules are removed and viewed laterally, the setse are very obvious. There are about half a dozen of them all inserted close together in a short transverse furrow, thus forming a sort of screen in front of the auditory pit, which lies at the other end of the same segment. The auditory pit is vs yet widely open, and is protected by 198 PROF. J. D. F. GILCHRIST ON A short feathered sete round its margin, so that it seems not an unreasonable conjecture that this screen is for the further protection of the organ at this stage, the large antennze forming effective protections at the sides. Its function may, however, be sensory, but, whatever it be, the structure would appear to be necessary only at this stage, as it is absent in the phyllosoma, and very much reduced in the post-puerulus. Fig. 13 shows the structure in the shed cuticle of the puerulus and fig. 12 its reduced condition in the antennule, from which it was cast off. The setze have, in the second stage, become very much shorter, so that, whereas they could reach beyond the Sy N \ anf fi v --4-.-5eg. |. Fie. 12.—Reduced antennular screen Fre. 18.—Antennular screen in post-puerulus, in puerulus. seg. 1, first segment of antennule; seg. 2, second segment of antennule. distal end of the.second segment in the puerulus when bent forwards, they now could searcely reach to the distal end of the first segment. In older specimens the setze become further reduced, though, even in the largest crawfish, they are found in this position. The furrow, however, in which they are lodged is well marked, though relatively shorter, being in the adult about a fifth of the diameter of the antennular segment. It is some- what semilunar in shape, the concavity being directed forward. It has a number of pit-like depressions for the setee, and these are often separated by marked prominences in the cuticle. In front of this semilunar furrow is a shallow depression in which the sets lie. The persistence of this organ, in the adult in such a definite form, seems to suggest some function, which of course cannot be, in this case, the protection of the auditory cavity. POST-PUERULUS STAGE OF JASUS LALANDII. 1) (8) Colour. The colour of the post-puerulus stage has been described above. This, as pointed out, appears in the later puerulus stages, and is accompanied by the appearance of the new spines of the succeeding stage. To be quite accurate therefore, if we are to regard the increase of spines as a characteristic of the second. stage, we must also consider the coloration to be so, the puerulus heing strictly colourless, except for the bright spots on the under surface, found also in the phyllosoma. Neither in the puerulus nor post-puerulus does the median rostral tooth meet the antennular segment as in the adult, a fact already noted in some specimens in the British Museum, from Stewart Island, New Zealand, and briefly described by Calman (2), who was the first to throw light on the real significance of the genus Puerulus. In a specimen with well calcified cuticle and 32 mm.in length it just touches it. In a larger specimen of about 40 mm. two slight projections of this segment were seen on each side of the tooth, but not clasping it, as in the adult. The same is true of the relative length of the antennular and antennary peduncle, the former being markedly the shorter both in puerulus and post- puerulus, and even in much more advanced specimens, about 55 mn. in length, they are only about equal, though decidedly longer in the adult. This feature therefore cannot be regarded as characteristic of the puerulus only. A comparison of the earlier stages, the naupliosoma and phyllosoma, of the Jasus lalandii of the Cape with that of New Zealand is of interest, That this first stage is also found in the New Zealand crawfish is now established. In the year 1906 Mr. G. M. Thomson (8) described a stage of Jasus lalandii which resembles the naupliosoma (4), but the antennules, not the antennie, are described as bearing plumose sete. A further discrepancy is the presence of five pairs of pereiopods, instead of three found in the Cape phyllosoma. Mr. Thomson kindly sent me a copy of his note, which is readily overlooked as it is included in a paper by Mr. Anderton on “ Obser- rations on New Zealand Fishes.” I suggested a re-examination of the New Zealand larva, in view of what I had found in that of the Cape crawfish, and in the year 1916 Archey (1), without however any previous knowledge of Thomson’s paper, redescribed and figured these early stages of the New Zealand crustacean, and concluded that the first larval form is identical with that which I described in 1913, There are, however, some slight dis- crepancies: thus, for instance, he figures free sete on the endopodites of the first and second pereiopods. This may be accounted for by the fact that the cuticle of the naupliosoma is first shed in these parts, and some of the subcuticular setee may have thus become free, this part of the limb representing, therefore, part of the phyllosoma limb. 200 PROF, J. D. F. GILCHRIST ON A I may add that though T have here spoken of a post-puerulus stage, this is only for convenience. All such “stages” are artificial, and strictly there are as many stages as there are ecdyses before the adult form is reached. Much remains still to be done before a fairly complete account of the early stages of the Cape crawfish can be given. Thus, the transition from the phyllosoma to the puerulus is not known, and the embryonic development is untouched. The first will be a matter of a lucky haul of the nets, the second is now being carried out. The material consists of stages from the first formation of the blastoderm to the time of hatching. It has not yet been completely examined, but I may state in connection with the present paper, that the nauplius stage in the egg was found to occur forty-two days before the time of hatching and setting free of the first larva. Summary. 1. Several pueruli of Jasus lalundii have been kept alive to succeeding or post-puerulus stage, which shows the following changes :— a. The red spots on the underside of the body disappear, and the whole of the upper parts become coloured. b. The spines of the carapace (of a definite number in the puerulus) become much more numerous. c. The antennules lose the antennular screen of the pnerulus. d. The mandibles resemble those of the puerulus in having three teeth, but differently disposed, and a cutting-edge has appeared. Denticles appear between and on the first two teeth, as well as along the cutting edge. The palp has become three-segmented. e. The exopodites of the pereiopods have disappeared. jf. The exopodites of the third maxillipedes are longer, by the addition of a flagellum. g. The cervical groove is well marked at the sides of the body. h. The pleopods have become smaller, especially the endopodite, which is now devoid of setee ; the coupling hooks and seta of the appendia interna have disappeared. i. The telson has additional spines. 2. The post-puerulus differs from succeeding stages in the following respects :— a. Cuticle unealcified. >. The incisor part of the mandible is provided with denticles on the margin of its thick cuticle. ce. The pleopods show no sexual differentiation. d. The third maxillipedes are separated at their bases. 3. The occurrence of a naupliosoma stage has been confirmed in the New Zealand Jasus lalandii, and the phyllosoma and puerulus are similar, Linn. Soc. ZOOL. VOL. XXXIV. PL. 15. Journ. Gitcnrist. eee ee - 4 i Grout, sc. & imp G. L. Birbel, de). JASUS LALANDIL GILCHRISY. JOURN. LINN. Soc., ZOOL. VOL. XXXIV. PL. 16. JASUS LALANDII. i (1) (2) (3) (4) (5) (6) (7) (8 YS POST-PUERULUS STAGE OF JASUS LALANDII. 201 References. ArcHey, GILBert.—* Notes on the Marine Crayfish of New Zealand.” Transactions of the New Zealand Institute, vol. xlviii. pp. 396-406 ; LOGS Bouvinr, E. L., Transactions of the Second Entomological Congress, 1912, pp. 78-89. Catman, W. T.—* The Genus Puerulus and the Post-larval Develop- ment of the Spiny Lobsters (Palinuride).” Ann. & Mag. Nat. Hist. (5) vol. iii. pp. 441-446 ; 1909, Gitcarist, J. D. F.—“ A Free Swimming Nauplioid Stage in Palinurus.” Linnean Society’s Journal, Zoology, vol. xxxii. pp. 225-231 ; Oct. UBL. Gitcurist, J. D. F.—Larval and Post-larval Stages of Jasus lalandii (Milne Edw.), Ortmann.” Linnean Society’s Journal, Zoology, vol. xxxiil. pp. 101-125 ; Nov. 1916. Gruvet, A.— Contribution & Pétude systématique des Palinuride.” Comptes-rendus Acad. des Sciences, T. 152, pp. 1350-2; 1911. GruveL, A.—* Mission Gruvel sur la cdte occidentale d’Afrique (1909, 1910). Resultats scientifiques et économiques (Palinurides).” Ann. Inst. Océanogr., T. ili. fase. iv. Tuomson, G. M.—* Note on the Development of Palinurus edwardsii ” (incorporated in a paper by Anderton on ‘ Observations on New Zealand Fishes”). Trans. New Zealand Inst. vol. xxix., 1906. EXPLANATION OF THE PLATES. PuatE 15. Drawing showing natural colours of the post-puerulus stage. Puiarr 16. Four pueruli, showing various stages in colouring. From a photograph of specimens kept alive for six days. fest THE LINNEAN SPECIES OF BRITISH NON-MARINE MOLLUSCA. 203 On the Linnean Species of Non-Marine Mollusca that are represented in the British Fauna, with Notes on the Specimens of these and other British forms in the Linnean Collection. By A. 8. Kennarp, F.G.S., and B, B. Woopwarp, F.L.S. {Read 21st November, 1918. ] THANKS to the active researches of Mr. C. D. Sherborn, A.L.S., for the ‘Index Animalium,’ prosecuted for now some thirty years, of Mr. Iredale, Mr. Reynell and others, our knowledge of the work of the earlier writers has been so greatly increased, especially of late, that there now seems some chance of approximate finality being attainable in the matter of nomenclature on the basis of priority—at least in the case of the British post-Pliocene non-Marine Mollusca with which we are particularly concerned. Accordingly we are attempting a more thorough revision of their synonymy than essayed by us in 1903 (Journ. of Conch. x. pp. 852-867) and 1914 (List of the British Non-Marine Mollusca, 8vo, pp. 12). Naturally one of the first steps in this undertaking has been to re-investigate the Linnean types, which have not been scrutinized since Hanley wrote his account of the whole collection of shells (Ipsa Linneei Conchylia, 1855), save for the incompleted investigation of Mr. J. W. Taylor and Mr. W. D. Roebuck in May, 1914, when they ‘‘ examined and isolated all the British land and freshwater shells ” (Taylor, Monograph, iii. p. 17). Now it had oceurred to us that Linné’s contemporary and correspondent, the celebrated O. I. Miiller, was likely to have known more about the great Swedish Naturalist’s work than was usually credited, despite the fact that many of Linné’s species do not find place in Miiller’s ‘ Vermium Historia’; and we accordingly proceeded to test the supposition. The first point of importance in this connection proved to be that Miiller in a paper on Fungi (Efterretning ... om Svampe 1 seer Ror-Svampens velsmagende Pilse. Ato. Kibbenhavn, 1763) described the slugs that fed on them, identifying three with Linnean species and describing, though not naming, two others. To one of these last Linné in the twelfth edition of his ‘Systema Nature,’ gave the name of Limaw albus, citing Miiller’s paper (a copy of which is in his library) as his authority. Miiller, however, omits in his ‘ Vermium Historia’ (ii. p. 4) to quote this although elsewhere in its pages he refers abundantly and solely to this edition. We then sought further evidence in the letters from Miiller to Linné in the Society’s possession, and under the kindly guidance of Dr. B. D. Jackson discovered the following postscript to his letter of 21 Novembr. 1767 :— 204. MESSRS. A. 8S. KENNARD AND B. B. WOODWARD ON THE Ms eee i shee ; : Testacea Sveciee, si Tibi supersint mecum communicare precor, qu sequuntur mihi nondum obvia. Mya lutraria. margaritifera. Bulla hypnorum. Turbo perversus. muscorum, Helix albella. Nerita littoralis. lacustris.” complanata. spirorbis. —— arbustorum. fragilis. —— limoga. Beneath the address on the outside of the letter when folded, and in the left-hand lower corner, was written : — ““avee un paquet marqué C. de L.” An identical memorandum also occurs in the same position on another letter dated 21 May, 1773. Another of Miiller’s letters, dated 4 Martii, 1768, begins :— “ Litteris tuis etsi & verbis & re satisferi, mittendo Tibi Stirpes, testacea ac dissertationes quasdam Italicas, adjecta prece ut mihi de his sententiam Tuam quam primum diceses, jam tamen per trimestre & quod excurrit sollicitus & incertus maneo.” Thus it is evident that Miiller sent mollusean specimens to Linné, but his request for information and specimens in return to help him in his work met no response as he complains in his Verm. Hist. (Tom. ii. p. 161). Our next discovery was that the numbers on the little packets containing shells in the Linnean collection that are alluded to by Hanley seemed in- dubitably to be in Miiller’s “N.14 handwriting. The list of these is as follows :— minima containing 2 Carychium minimum, Mill. cavend. reper.” CNS citrina p * % > {indecipherable | “No. 27” containing ‘No. 40” 3 “No. 44”, “No. 52 E NO./0d)” 3 1 Polita radiatula, Alder, and 1 Zonitoides nitidus (Miill.). bh) 1 Polita cillaria (Miul.) and 1 P. Draparnaldi (Beck). 1 Hygromia hispida (Iuinn.) and 1 Helicella sp. ? 1 Pupilla muscorum (Linn.) [edentulous variety |. a number of miscellaneous species of Clausilia. 10 Ancylus fluviatilis, Miill. LINNEAN SPECIES OF BRITISH NON-MARINE MOLWLUSCA. 205 In the box with the contents of No. 27 was a label “ Helix cellaria,” and in with No. 53 one “ Patella lacustris.” Both these labels are in an un- recognized handwriting. There are likewise in the collection, but without any original information, specimens of :—2 Pyramidula ruderata (Studer) ; 8 Helicella virgata (DaC.); 20 Cochlicella acuta (Miill.) ; 1 Vallonia ewcentrica, Sterki (labelled by some one “ Helix pulchella”) ; 2 Helieodonta obvoluta (Mill.) ; 3 Helia aspersa, Miill., ticketed in the unknown handwriting “ Helix grisea,” and 10 others including one with the remains of a label in script apparently of the date of the collection ‘ Helix [**] 2” in two lines; 2 adult and 7 juvenile specimens of Helie pisana, Miill.; 2 Clausilia, ticketed ‘ bidentata,” Strém’s name evidently being meant; 4 Suecinea pfeiferi, Rossm.; 1 Limnea palustris (Miill.) ; 38 Segmentina nitida (Miill.), which had been wrapped in a scrap of paper bearing German print on both sides ; a packet, the paper of which bore Swedish print on one side and Latin on the other, containing 3 Planorbis umbilicatus, Mill. [= Pl. planorbis, Linn.], 1 Pl. carinatus, Miill., 1 PZ. albus, Miill., with a juvenile example of Jelicigona lapicida (Linn.), these several species have been distributed in different glass-topped boxes, but careful record given of their former association ; 4 Valvata piscinalis (Miill.); 20 examples of Pomatias elegans (Miill.)*; and finally a specimen of Unio tumidus, Retz., on a tablet with single valves of two other species, with the sole comment in pencil ‘3 species.” It will be observed that with the exception of Pyramidula ruderata, Helicella virgata, Clausilia bidentata, Suceinea pfeiferi,and Unio tumidus all are Miillerian species, for the Vallonia was evidently sent as //eliv pulchella, Mill. Putting the above facts together it seems clear that cotypes of some at all events of Miiller’s species are included in the Linnean collection, and that the recognition of their presence explains some of the doubtful points that have attended the attempts to identify certain of Linné’s species. A Botanist first and last, the great Naturalist does not appear to have attached much importance to his collection of shells, nor to have preserved those obtained for and described in the ‘Fauna Svecica.’ Although these last may perhaps have suffered in the tribulations the collections underwent even in Linné’s time, as recorded by Dr. B. D. Jackson (Proc. Linn. Soe., Sess. 125, suppt., pp. 1 & 2). In passing under review the various species with which we are ut present concerned it will be convenient to take them in the order of classification and under the nomenclature most generally accepted to-day, as follows, an * being prefixed to those which were recorded by Linné as present in his collection (cf. list by Dr. B. D. Jackson, Proc. Linn. Soc., Sess. 125, LINN. JOURN.— ZOOLOGY, VOL, XXXIV. 16 206 MESSRS. A. 8. KENNARD AND B. B. WOODWARD ON THE suppt. pp. 38-43), whilst the number of specimens is added in (_) after the name :— Limax Maximus, & —— FLAVUS. No doubt attaches to the identification of these two species which are established by the references given to Lister’s figures. Acriotmmax (Limax, L.) AGRESTIS. Linné’s reference to Lister, who remarks on its peculiar white mucus, suffices to show the species intended. Arion (Limax, L.) arer. As shown by the figures cited in Lister this well-known form cannot be mistaken, nor its colour varieties rufus, Linn., and albus, Miill. in Litt. of the 12th edition of the ‘ Systema.’ According to Collinge (Conchologist, ii. 1892, p.59) Pollonera identified the British form with the A. empiricorum of Férussac, and held it distinct from the LZ. ater of Linné, overlooking Linne’s reference to Lister, but no other malacologist has subscribed to this opinion. HELIX GOTHICA. The diagnosis of this species might apply equally to either Pyramidula rotundata (Miill.) or P. ruderata (Studer), whilst the young stage of Heliw lapicida, Linn., has also been suggested. As previously mentioned two examples of P. ruderata are in the collection, but there is no means of connecting them with Linné’s diagnosis, conse- quently the name cannot be used. *HeELicetta (Hetix, L.) raza (4). Jinné’s number in his own handwriting still visible on one of the speci- mens places the identity of this species beyond doubt. *CocHLICELLA (Heurx, L.) BARBARA. ~ Hanley (Ipsa Linneei Conch. p. 384) states that Dillwyn identified this form with the Bulimus acutus of Miiller and that the supposition was confirmed by the presence of specimens in the Linnean cabinet that alone agreed with the author’s diagnosis. Unfortunately Hanley overlooked the fact that Dillwyn (Cat. ii. p. 956) quotes Miiller’s Helix acuta for the English shell and that under f/eliv barbara (p. 960) he merely criticises Chemnitz’s supposition that /7. carinula might be the species meant, and proceeds “ but the description answers better to a diminutive specimen of HH. acuta.” Dillwyn, therefore, made no definite statement on the subject, and conse- quently one is not surprised to find in the. Hist. Brit. Moll. of Forbes and LINNEAN SPECIES OF BRITISH NON-MARINE MOLLUSCA. 207 Hanley the Linnean name is only quoted as “probably” a synonym for y A y Miiller’s //. acuta. Linne’s description, he unfortunately quotes no figure, leaves one in doubt whether his specimen may not have been one of the My smaller, closely allied species such as ventricosa, for he describes it as ‘“* Hordei semine paulo major.” Miiller makes no allusion to Linné’s species, and we consider that in all ) probability the specimens of acutus now in the Linnean collection had been received from Miiller, and that Linné’s name owing to its uncertainty must be allowed to lapse. Hyeromia (Hextx, L.) nisprpa (1). The specimen in the Linnean collection that corresponds with the descrip- tion is, as Hanley points out (p. 368), identical with the H/. concinna of Jeffreys. It now forms one item of the little packet marked “ No. 40” of those we believe Miiller sent to Linné. It may have become enclosed in the packet by error, for Taylor (Monograph, iii. 1916, p. 17) states that though he saw it in the collection thirty years before, he failed to find it when he looked in company with Mr. Roebuck in 1914, and Hanley does not mention its enclosure in paper. Whether this particular specimen be a Linnean type or not there seems no reason for displacing the name which so many authorities from Miiller onwards have attached to the form that now bears it. *Hevicicona (Hexix, L.) uaprcrpa (4). The identity of this species docs not seem ever to have been called in question. Artanta (HELIX, L.) arBUsTORUM (7). The interpretation of this species also has met with universal acceptance. It may be noted in passing that the reference given in the 10th and 12th editions of the ‘Systema’ to “ List....conch. i. n. 52” is wrong, no such number existing in that work: it is corrected by Gmelin in his edition to ‘© 53,” which is correct. The further reference to ‘“ Argeny. conch. t. 32. f. 8” is quite wrong, the figure in question depicting H. nemoralis: this citation is rightly omitted by Gmelin. HELIX GRISEA, Although Hanley (p. 378) found the box in the Linnean cabinet so marked to be filled with examples of //elia aspersa, Miill., he does not venture to advocate the adoption of the name for Miiller’s species, although Dillwyn (Cat. ii. p. 943) does. Turton in his translation of Gmelin’s edition of the ‘Systema’ (vol. iv. p. 530) says of grisea that it “resembles H. pomutia.” L6* 208 MESSRS. A. Ss KENNARD AND B. B. WOODWARD ON THE Linne’s cited figure for his grisea in Gualtieri (Index Test. Conch. pl. i. f. C) so inscribed in his own copy of that work, is obviously a pomatia, whilst the next figure on the plate (J) which represents H. aspersa, has not been marked by Linné, who also when citing correctly for pomatia ‘“ Argeny. t. 32. f. 1,” omits all reference to f. 2 & 3 which are good figures of aspersa. We think, therefore, that he was unacquainted with the later shell. As already mentioned, among the examples of aspersa in the collection is one with the remains of a label (Linné always wrote numbers on his specimens), and it seems possible that some if not most of these specimens were received from Miiller. HELIX POMATIA (8). This is another of those species concerning which there has never been any question, the figures instanced leaving no room for doubt. *HELIX NEMORALIS (26). Hanley, we consider, was quite correct in his decision that Linné’s species was that which now passes under the name. The reference to “ Lister conch. no. 54,” where the figures are unmistakable, is correct in the 10th edition of the ‘Systema’; but by error the number has been altered in the 12th edition to ‘ 53,” though this is corrected by Gmelin in his edition. Westerlund’s contention (Fauna Svecie, i. 1873, p. 101), that the . nemo- ralis, L., is the AH’. hortensis, Miill., because the latter form alone he states is met with in Oliind, where Linné appears to have first observed banded snails of this description (Oliindska Resa, 1745, p. 127), seems to us beside the mark. Linné’s name anyhow dates from the 10th edition of the ‘Systema’ (p. 773), where the extended habitat of ‘ Kurope” is given, but furthermore Linné in the ‘ Oliindska Resa’ cites “ Petiv : mus. 5.n. 14” and the latter quotes the figures in Lister’s ‘ Hist. Conch.’ i. no. 54, which as already noted are unmistakably HH. nemoralis. That Linné did not discriminate between nemoralis and hortensis is pro- bable, but this was rectified by Miiller, from whom possibly some of the specimens in the collection may have come. “HELIX LUUCAS. Hanley records (p. 361) finding in a paper certain specimens of immature Heliv pisana, Miill., which in his opinion coincided with Linné’s description of his /Z. leucas and which he thought were in all probability the originals. Two adult and seven immature individuals of A. pisana, Miill., are present in the collection to-day but no trace of the paper in which according to Hanley some of the latter were preserved. These specimens may have come from Miiller. Linné’s diagnosis is not sufficiently clear to enable definite pro- nouncement to be made, and the name cannot, therefore, well be revived in LINNEAN SPECIES OF BRITISH NON-MARINE MOLLUSGA. 209 lieu of Miiller’s better-known one. It is of interest at the same time to note that Beck (Index Moll. 1837, p. 14) queries under his Theba leucas “an HT, pisana, Bhrbg.? Sav. Egpt. m. ii. 15-16?” The figures thus indicated are undoubted H. pisana, Miill., whose species, however, follows next as a distinet one in Beck’s list. HELIX ocTona. Miiller (Verm. Hist. ii. p. 150) makes this a synonym of his Buceinum acicula regardless of the fact that Linné described the species as having “apertura subrotunda,” whilst of his own he states “apertura oblonga seu fusiformis.” Nilsson (Hist. Moll. Svec. 1822, p. 92) calls it a Paludina, and says it is not dissimilar from Draparnaud’s Cyclostoma acutum. Hanley (p. 381) agrees with this conclusion. The cited figure (Gualtieri, pl. vi. f. BB), however, suggests the fry of Limnea glabra (Miill.) with which it has been identified by both Pennant (Brit. Zool. iv. 8vo. ed. p. 138, with a ?) and Fleming (Hist. Brit. Anim. p. 274). Pulteney included it in his “ Cata- logues,” but Rackett in the later edition eliminates it on the ground that it is not English. Linné’s name, therefore, cannot well be connected with any of our known species and must remain in abeyance. HELIX SUBCYLINDRICA. This also was included by Pulteney in his ‘ Catalogues,” but rejected by Rackett as “ not of English growth.” Dillwyn (Cat. ii. p. 952) and Moquin- Tandon (Hist. Moll. France, ii. p. 304) identify this with Miiller’s WZ. lubrica, although the habitat is given by Linné as “in aquis dulcibus Huropze borealis.” Hanley (p. 379) gives as his opinion that only Truncatella Montagui of Lowe [ = 7. truncata, Mont.] agrees with the description of the shell, forgetting that Truncatella is a southern not a northern form. Possibly Linné’s shell was the fry of some freshwater form. Anyhow the name will have, we consider, to be left out of account. Poriuza (Torso, L.) woscorum (1). The identification of this species rests mainly on tradition, Draparnaud alone having applied the name to a different form from that which generally bears it. There is an example, noted by Hanley, of an edentulous specimen in the Linnean collection that alone accords as he says with the description, but this is in packet “ N. 44” of those we now believe to be of Miiller’s sending. Miiller be it noted (Verm. Hist. ii. p. 105) also describes the species, which he puts under /eliw, as “ Apertura edentula,” and the name can, we think, be accepted on Miiller’s confirmation. 210 MESSRS. A. S. KENNARD AND B. B. WOODWARD ON THE Baga (Turso, lL.) pervEersA (10). Nothing in the description distinguishes the currently accepted interpreta- tion of this species from the young form of one or two species of Clausilia and tradition and the occurrence of undoubted specimens in the Linnean collection alone support its identity. Miller took it for the young form of Clausilia rugosa, Drap., as shown by his description of the clausium (Verm. Hist. ii. p. 119). We would propose to retain the name for the species which has so long borne it, although Draparnaud’s trivial name of fragilis has been almost equally used in the past. *Sucotnea (HeEtix, L.) purris. The figure cited from Lister and the specimen in the collection leave no doubt as to the identity of this species. As Hanley points out, and as already noted, examples of S. p/eijfer?, Rossm., are also present in the collection and it is possible that Linné did not discriminate between the two forms. AcroLoxus (PatEetia, L.) LACUSTRIS. Although the form which to-day bears this name is not in the collection whilst examples of Miiller’s Ancylus fluviatilis are, we do not agree with Hanley that the latter is Linné’s species, but regard them as part of the Miillerian contribution to the collection, since the ten individuals were included in packet ‘ No. 53.” Linné’s definition is ‘‘ testa integerrima ovali membranacea, vertice mucro- nato reflexo ” (Faun. Svec., ed. 2, p. 534), showing that the oval shape and prominent mucro had struck him. Contrast this with his description of Patella pellucida, which in form comes nearest to Miiller’s fluviatilis, and which follows in the ‘Systema’ (10th ed., p. 783), where the term used is ‘“obovata.” * Ovali” on the other hand is applied to elongate forms such as Patella compressa and P. lutea. Then the locality ‘‘ Lacubus . . . foliis insidens subaquis” is more applicable to Acroloxus than to Miiller’s species, which is chiefly found in swiftly running water, though at times the two have been taken together on water plants. Lister’s figure cited is of course more applicable to Miiller’s species, as also is the habitat he quotes, but in this respect Hanley’s caveat (p. 6) as to Linné’s citation of somewhat similar figures must be borne in mind. The same remark applies equally in the following. *Limna#a (Heiix, L.) aurrcunarta (3). Linné’s diagnosis “ spira brevissima apertura ampliata ” is more convincing than the figures pressed into his service. These are various. Thus “ Lister angl. t. 2. f. 23” seems an inflated form of ZL. pereger (Miill.). “ Argenv. conch. t. 31. f. 7. bona” shows two shells on either side of the “7” which will pass for auricularia, whilst the one to the left of them is an inflated LINNEAN SPECIES OF BRITISH NON-MARINE MOLLUSCA. 211 pereger. “ Klein ostr. 54. t. 3. f. 69” is a bad copy of the figure in Lister’s ‘Hist. Anim. Angl.’ pl. ii. f. 21 (and is Klein’s type of his Awricularia stagnorum) or the figure already ‘cited by Linné as his Helix stagnalis. The “ Acta Helv. 5. t. 3. f. 27, 28” is evidently awricularia. The specimens in the collection include auricwlaria vera (3) and inflated forms of pereger (4). Miiller’s segregation of the latter is therefore correct. HELIX LIMOSA. Nilsson (p. 72) identified this with Limnea truncatula (Mill.), whilst Hanley (p. 387) conjectures it may have been a narrow form of L. pereger (Miill.) and Reeve (Land & Fresh-w. Moll. Brit. Is., p. 157) adopt the name for that species. The figure cited by Linné in the 12th edition of the ‘Systema’ (p. 1249) from Gualtieri (pl. v. £. H) and marked on the margin of the plate in his own bandwriting, is suggestive of the fry of possibly Suceinea putris. In the ‘Fauna Svecica’ first edition (p. 367. no. 1314) it is described as “anfractibus quinque,’ an item omitted in both the 10th and 12th editions of the ‘Systema’; whilst the final ‘“ Deser.” reads ‘“ Preecedentibus duabus duplo minor ; operculo etiam clauditur.” The two preceding species of the Ist edition of the ‘Fauna’ are named in the 2nd Helix vivipara and Helix tentaculata. Linné’s Helix limosa was therefore an operculatet, which cannot now be identified, so that the name must be allowed to lapse. HELIX BALTHICA. The “ rugis elevatis ” noted in the description of this species (Fauna Svec. 2nd ed., p. 532) and amplified in the “ Descr.” to “superficies teste striis elevatis remotis versus aperturze marginem oblique descendentibus ” does not accord with a Limnea, to which genus Nilsson (pp. 64-65) refers it. The form appears indeterminate and likely to remain so. *Limyaza (Hoecrx, L.) sraenatis (1). By diagnosis and figures the identity of this species is confirmed. Also that Linné’s HZ. fragilis is a synonym. *PrLANoRBIS (HeEttx, L.) corneus (4). This is, fortunately, one of those species concerning whose identity there never has been any doubt. *“PranorBis (Henix, L.) pLanorpis (8, including 3 probably sent by Miiller as examples of his P. wmbilicatus). Why there should have been any question concerning this form it is difficult to understand, for the reference to Lister’s figure (Anim. Angl. pl.ii. f..27) clearly establishes its identity with the Planorbis umbilicatus of + This we find has already been pointed out by Mérch (Synop. Moll. Danis, 1864, p. 43, note). 912 MESSRS. A. Ss KENNARD AND B. B. WOODWARD ON THE Miller. Miller includes in his synonymy (Verm. Hist. ii. p. 160) the Helix complanatus of Linné, but three pages further on queries whether the latter be not the same as his own P. nitidus, a point we discuss later. Linné’s name for this species having priority must be restored to literature. Prianorpis (NAutiLus then Turso, L.) crista. Despite Linné’s doubt as to the zoological position of this form, referring it as he does first to Nautilus and then with the changed trivial name of nautileus to Turbo, there never has been any hesitation as to the species meant. Pranorpis (HELIX, L.) vortex (1). The figures referred to in Lister and Gualtieri show clearly the group to which Linné’s species belongs, and the description in the ‘ Fauna Svecica’ (2nd ed., p. 527, no. 2173), “ margo teste in illa extra testam exseritur, at in hac ipse testee angulus est acutus,” points to the correctness of the accepted interpretation, which is further borne out by the single example in the collection, Pranorsis (HELIX, L.) SPIRORBIS. No figures are cited by Linné and the description is rather too vague to enable one to accurately discriminate the species. _Linné’s specimen must have been bleached since it is given as “‘ Testa alba.” No example is in the collection. Miiller, who adopts the species (Verm. Hist. ii. p. 161), further defines it as having “anfractus quatuor...margine tereti absque ulla carina. Apertura rotundata sublabiata ; margo saltem summus intus albus, crassius- culus.” This seems to indicate the commonly accepted species. ‘The closely allied rotundata, Poiret, has a whorl more and a more angular mouth, *Pranorsis (HELIX, L.) conrorTus. Although no figures are cited, the description in the ‘Fauna Svecica’ (2nd ed., p. 528, no. 2181) and the former presence, vouched for by Hanley, of a specimen in the collection, that, however, we have not seen, serve to establish the identity of this species with the commonly accepted form. PLANORBIS (HELIX) COMPLANATUS. Hanley’s arguments and conclusions concerning this species are correct, given the fact that he in common with most other observers of his period identified Miiller’s nztidus with Lightfoot’s fontanus, whereas it is now known, and we have the authority of Dr. A. C. Johansen for the fact, that all Miiller’s specimens belong to the form called lacustris by Lightfoot and placed to-day in the genus Segmentina of Fleming. Some of Miiller’s specimens are not, Dr. Johansen states, well preserved and this led LINNEAN SPECIES OF BRITISH NON-MARINE MOLLUSCA. 213 Dr. Gwyn Jeffreys to conclude from his observation of them that both species were included in Miiller’s nit¢dus. Linné’s description ‘“ subtus plana omnino, sed parum cava versus centrum” is more applicable to Lightfoot’s fontanus than to his lacustris. Linné’s name must, we think, be restored and the two forms be respec- tively known as Planorbis (Hippeutis) complanatus (Linn.) and Segmentina nitida (Miill.). As previously mentioned, three individuals of this last are now in the collection, given, as we believe, by Miiller. Paysa (Buia, L.) FoNTINALIS & Puysa (Bua, L.) #YPNoRUM are recognizable from the descriptions in the ‘ Fauna Svecica’ (2nd ed., p. 523, no. 2160 and p. 522, no. 2159 respectively). No example of either is in the collection, but no doubts have ever been raised concerning their identity. Patoupestrina (Herix, L.) staqnatis. Linné in the 12th edition of the ‘Systema’ (p. 1248, no. 697) introduced the name “ //elia stagnalis” for Baster’s Turbo stagnalis. As Hanley points out, he afterwards noticing that he had already used the name for the well- known pond-snail, altered it in manuscript in his own copy to Basteri*, whilst Gmelin (Syst. 13th ed., p. 3653, no. 119) changed the trivial name to stagnorum. Baster’s species has been held to be identical with Pennant’s Turbo ulve, and consequently his name being prior has been used in lieu of Pennant’s. We have shown, however (Proc. Malac. Soc. Lond. xii. 1907, p. 124), that the species abounding at Baster’s locality was identical with a form con- sidered at the time on the authority of Dr. Johansen to be the Paludina minuta of Totten. Totten’s species, however, proves to be quite different, and there is now reason to believe that we are here dealing with the Turbo ventrosus of Montagu, vera, the form commonly passing under ihe name being distinct. But this requires further research. Biraynta (Hetix, L.) rexracuLara (19, wrapped in a scrap of paper bearing German print). We agree with Hanley that the description in the ‘ Fauna Svecica’ (2nd ed., p. 531) and the reference to Lister's figure sufticiently determine this species to be the form commonly accepted as such. In the first edition of the ‘ Fauna’ (p. 376, no. 1813) Linné contrasts it “cum precedenti” (No. 1312) the Helia vivipara of the 2nd edition, but neglected to make the necessary correction in the 2nd edition where five other species have been intercalated. * Miiller, Verm. Hist. ii. p. 132 note, also calls attention to the double use by Linné of Helix stagnalis, 214 MESSRS. A. S. KENNARD AND B. B. WOODWARD ON THE Linné was most unfortunate in his selection of further illustrations for the 12th edition of the ‘Systema,’ those from Bonanni, Lister’s ‘ Hist. Conch.’ and Klein’s copies of these last, being all well-marked marine trochoids ; whilst the reference to Adanson is hopelessly wrong. *Viviearus (Hewrx, L.) vivipara (14). Linné’s description “ H. testa imperforata” definitely determines which of the two closely allied forms he had in mind, but of the figures referred to, Lister’s “ exere. 2. p. 17. t. 2 [£.5],” which is Miiller’s fasciata, shows that he had not discriminated between them. Any possible uncertainty is, however, set at rest by our discovery the other day that two of the speci- mens bore the number “ 603,” that of the species in the 10th edition of the ‘Systema’ in the master’s own handwriting +. There are now, in the same glass-topped box, two specimens of /asciata, obviously from quite a different gathering, and these we are inclined to think may have been sent him by Miiller, whose separation of the two forms is correct, and whose name has priority over contecta, Millet, which has of late been generally used. TURBO REFLEXUS. We do not agree with Hanley that because examples of Pomatias elegans (Mill.) were found in the box marked for Linné’s 7urbo reflexus that Dillwyn’s surmise was correct. Like other Miillerian species they may have come from that author. Linné’s phrase “apertura reflexa” and “ Habitat in Europa australi ” dispose to our mind ot the possibility of its identity with Miiller’s elegans, but we decline to speculate on its correct interpretation. *THEopoxus (Nuria, L.) FLUVIATILIS & LACUSTRIS. There is, fortunately, no hesitation as to the correct ascription of Linné’s name fluriatil’s to the form that has always borne it; whilst we further agree with Hanley in believing that the /acustris was only a variety of the same polychromatie shell. *Unio (Mya, L.) prororum. Although the diagnosis is too meagre to show which of the two forms pictorum or tumidus was meant, the figure first cited by Linné (Lister ang. app.-t. 1. £.4) is that of the commonly accepted pictorum, whilst the inscribed specimen in the collection certainly is. There is an individual also in the cabinet of tumédus but without any history. Theadded reference in the 10th edition of the ‘Systema’ to “ Bonan. recr. 2. t. 41,” which is a copy of + After this paper was written and handed in Mr. J. W. Taylor has placed on record (Naturalist, 1918, p. 249) that this fact had been observed by him and Mr. Roebuck, to whom belong the credit of having detected specimens that escaped us some years ago. Or LINNEAN SPECIES OF BRITISH NON-MARINE MOLLUSCA. Pak Lister’s “Anim. Angl., pl. ii. f. 30,” rather suggests tumidus. Retzius, however, as next reviser separated the two species on lines ever since followed and no objection has ever been raised to his decision. It is commonly overlooked, however, that the name Unio was Linné’s own, as Retzius acknowledges, its characters having been communicated to him by Linné’s pupil Acharius, *MARGARITANA (Mya, L.) MARGARITIFERA (1). No uncertainty has ever existed concerning the identification of this well- known species, supported as it is by Lister’s figure (Anim. Angl. append., jell me te IL). *Anoponta (Mytrizus, L.) cy@nna (1). Of the two species now once more recognized as valid, the inscribed and numbered specimen in the collection with the figures cited (Lister angl. app. t. 1. f. 3 and Gualtieri t. 7. f. F, the latter written against by Linné himself in his copy of the work) quite establish its identity. The further quotation of Lister “conch. t. 193. f. 8” in the 10th edition of the ‘Systema’ (p. 706) should, as Hanley remarks, be ‘* 153,” but both Tinné and Hanley overlooked the fact that the figure in question is a reprint of Lister’s Anim, Angl. append., pl. 1. £. 2 adduced by Linné in illustration of his anatinus. This same figure copied by Klein (Ostr. t. 9. f. 26) is correetly cited by Gmelin under anatinus. The correct reference to Lister’s Conch. should have been ‘ 156” which is the copy of the “app. t. 1. £. 3.” *Anoponta (Myinus, L.) anatrna. Lister’s figure (Anim. Angl. append., pl. i. f. 2) originally cited in the first edition of the ‘ Fauna Sveciea, with the enlarged description in the ‘Systema’ (10th ed., p. 706, no. 219) and in the second edition of the ‘Fauna Svecica’ (No. 2158) convince us that the customary identification of this species is correct. The figure in Gualtieri added in the 12th edition of the ‘Systema’ may or may not have been intended for a Unio, but lacks any hinge-teeth. This may have caused Linné to include it and possibly may account for lis puzzling observation in the 10th edition of the ‘Systema “— “similis Myze pictorum, sed fragilior & cardine distinctissimus.” We do not agree with Hanley that Lister’s “f. 2” shows “an ordinary example of the Anodonta cygnea,” nor with his selection of the specimen figured by him in illustration, which, as the umbonal rugze show, belongs to the genus Pseudanodonta, Cf. P. grateloupiana (Gassies) or P. normandi (Dupuy). It is only fair to add, however, that this genus had not in his days been separated off from Anodonta. *SpHarium (Teviina, L.) corneum. Happily, though we were unable to trace the specimen in the collection, this is another of the undisputed species and calls for no comment. PROF. W. A. HASWELL: THE EXOGONEA. 217 The Exogonee. By W. A. Haswett, M.A., D.Sc., F.R.S., F.LS., Emeritus Professor of Biology, University of Sydney. (Puatres 17 & 18, and 2 Text-figures.) {Read 5th December, 1919. j INTRODUCTION, Iw the introductory part of his ‘Recherches sur les Syllidiens’ (1913), Malaquin has embodied a very complete history of the development of our knowledge of the family. Malaquin’s own work is certainly of the first im- portance in connexion with this development, on account both of the original observations recorded therein and of the systematising of work previously published. But Malaquin concerns himself almost exclusively with the Eusyllidea, the Syllidea, and the Autolytea, having few actual observations on the Lzogonew to record: and out of three hundred and fifty or more figures with which the work is illustrated only three refer to the Lwogonec. Thus in his account of the alimentary canal, the nephridia and gonads, and the development, it is almost assumed that the HLwogonew resemble the larger forms but for the difference in size. And this is quite true up to a certain point. But beyond that, in a number of important details the Kxogonev, or certain of them, present very special features. The material on which the following observations were based was collected almost exclusively in Port Jackson, either between the tidal limits or a little below low-water mark. In the structural part I have not given any general account of the mor- phology ; but have confined myself to certain sets of organs with regard to which the Lwogonew exhibit peculiarities of a strongly-marked character—the integumentary and related glands, the alimentary canal, and the nepliidial and reproductive organs. Tn the account of the embryology which follows there are various obvious gaps. I hope to fill some of these later by the adoption of certain methods which I have not hitherto utilized ; the detailed history of the cell-lineage, for example, can only be followed out by keeping large numbers of living specimens under observation. As it is, however, the study of my extensive collection of fixed material has resulted in the observation of a number of facts which have hitherto, so far as I have been able to ascertain, been unrecorded, I have pleasure in acknowledging my great indebtedness to Professor Melntosh for literature, specimens, and information furnished ; to Professor Benham for the loan of his copy of Malaquin’s work ; and to Dr. Pierantoni 218 PROF. W. A. HASWELL! for sending me at my request a copy of his paper, “ La gestazione esterna,” which was otherwise inaccessible to me. For assistance in collecting specimens I owe thanks to Professor 8. J. Johnston, Mr. Thomas Whitelegge, and Miss B. M. Somerville. e SYSTEMATIC. The only published work in which there is any reference to the Mvogonee of the Australian coasts is Augener’s ‘ Polycheeta’ of the ‘Fauna Sudwest Australiens.’ In this the following are described :— Exogone heterocheta, McInt. Spherosyllis hirsuta, Ehl. Spherosyllis perspicaw, Ehi. Grubea kerguelensis, McInt. Grubea quadrioculata, n. sp. Grubea furcelligera, n. sp. EXoGONE FUSTIFERA, n. sp. (Plate 17. figs. 1-6.) This is a larger form than the European /2. gemmifera, reaching a length of 7 mm. as compared with the 3 to 4 mm. of the latter. The maximum number of segments observed to occur was 43 as against 33 in L. gemmijera; but one complete specimen had only 82. The palpi are completely united, but separated ventrally by a median cleft. The almost completely semicircular outline of the united palpi may be broken by a slight median notch, but this is frequently indistinguishable. The prostomium and peristomium are very closely united together and scarcely distinguishable superficially. The three prostomial tentacles are nearly equal in length, the median one slightly longer than the other two, projecting ~ slightly beyond the end of the palpi. ‘They are very slender except towards the end, where there is a marked dilation. The eyes, as is usual in the genus, vary greatly in size in different phases. The peristomium bears a pair of extremely rudimentary button-like tentacles, which are placed comparatively far forwards, so that in fixed specimens they may be on a level with the interval between the anterior and posterior eyes. Close to each is a ciliated pit. The parapodia (P1. 17. fig. 2) are short —in length less than half the breadth of the body, narrowed distally, the extremity simple, rounded. In each are normally a single stout aciculum and five sete, of which the most dorsally and the most ventrally situated are simple, the rest compound. The dorsal simple seta (figs. 4, 5) ends in a curved conical extremity which may be obscurely notched close to the apex, the notch sometimes, though rarely, bearing a fine hair-like appendage. Often this seta is absent in some of the parapodia or does not project on the surface. The ventral (fig. 6) is more strongly hooked, and is bifid at its apex. The three compound sete (fig. 3) have the terminal appendages greatly reduced, in the form of simple slender THE EXOGONEA. 219 curved rods without serrations or teeth. Sexually ripe individuals have in addition bundles of long capillary natatory setze on the dorsal side of the parapodia of most of the segments. As in other Syllids, the setigerous sac in which new sete are developed is quite separate from the bundle of mature setee, and attached firmly to the aciculum on its dorsal side. The dorsal cirri are greatly reduced, shorter than the parapodia, oval or spherical in shape. The second segment is always devoid of dorsal cirri. The ventral cirri are also rudimentary, but more cylindrical in form. The anal cirri are longer than the parapodia, slender and cylindrical, not swollen at the ends. In the retracted condition of the proboscis the pharynx runs through the first three segments, the proventriculus lies in the fourth and fifth, and the ventriculus and ceca are situated in the sixth. Exogone fustifera is by far the commonest of the Port Jackson Evegoner, and occurs abundantly among finely branched Alow of all kinds about low- water mark. I also found it in Broken Bay among Algze and Sponges growing on the piles of a wharf; and in material collected at Port Stephens by Prof. S. J. Johnston. Viguier (84) * identifies the species of Hvogone which he found most abundant in the Bay of Algiers with that observed by Pagenstecher (26) at Cette, and named by him /. gemmifera. Viguier was not in any doubt on this point. He writes :—“ C’est bien le type étudié par cet auteur que j’ai pu observer & mon tour.” But a comparison of the descriptions and figures of the two authors leaves some doubt of the correctness of the identification. Viguier states with regard to the tentacles: ‘ Elles sont leg¢rement renflées un peu au-dessous de leur base, et décroissent jusqu’au sommet, qui est arrondi et muni de quelques petites soies raides” (p. 77); while Pagenstecher refers to the same appendages as ‘nur wenig Kolbig anschWellende Fiihler,” and his figure represents them as distinctly swollen at the ends. I think it probable that Pagenstecher’s . yemmijera is Girsted’s LZ. natidina (25) and that Viguier’s /. gemmifera (Pagenstecher), though identical with Claparéde’s iH. clavigera, as Viguier himself maintains, is not Pagenstecher’s species. Should this suggestion prove to be correct, it would be to 2. naidina (Cirsted) that /. fustifera would most closely approximate rather than to EE. clavigera (Claparede). EXOGONE VERRUGERA, Claparéde. (Plate 17. figs. 7-10.) Pedophylax veruger, Claparéde, 2, p. 215, pl. 12. fig. 5. A second species of Ewogone occurring in Port Jackson is nearly related to Claparéde’s Padophylax verrugera from the Gulf of Naples. Like the pre- ceding it is a small form, and does not exceed.7 mm, in length. There are about 37-40 segments. * The thick figures refer to the Bibliography at the end. 220) PROF. W. A. HASWELL : The palpi are completely united, without any median notch (even in the embyro) and are comparatively long, so that in the living animal, when fully extended, they may be three times the length of the prostomium and peri- stomium together. The tentacles are usually reduced to three similar oval rudiments, but in some the median tentacle is a good deal longer than the others. The single peristomial tentacle is very small, though not so rudi- mentary as in £2. fustifera. The simple seta on the dorsal side of each parapodium (fig. 9) has the curved terminal part tapering to a fine point close to which is a small denticle. There are five compound setee (fig. 8), all of which have terminal appendages of the normal kind, that of the first (most dorsal) (text-fig. 1) long and slender, while those of the others, though completely formed, are quite rudimentary. The most ventral seta is usually simple, curved at the end and bidentate, like the corresponding seta in /. fustifera. The aciculum (figs. 10 4, 10 B) is curved at the end and slightly expanded. Both the dorsal and the ventral cirri are small and approximately oval like the tentacles. Text-fig. 1. w\\ ii Setee of Exogone verrugera. Dorsal cirri are always present on the second segment. The eyes in most. cases bave the appearance represented in fig. 7. In one specimen they are both large and irregularly shaped, the anterior pair being rather the larger ; in another the posterior pair of eyes had the pigment remarkably drawn out into a long narrow process projecting backwards into the peristomium behind the ciliated groove. In several others the lenses of both the anterior eyes are produced into horn-like pointed bodies. The median tooth, when the proboscis is retracted, lies in the first setigerous segment. The gizzard is much more elongated than in /. fustifera, - extending from the fifth to the seventh segments in the retracted condition, and containing about 25 rows of muscle-columns. In most respects this species, which is fairly common in Port Jackson, approaches very near Claparéde’s LZ. verrugera. But the ditference in the palpi seems to be very marked—those of L. verrugera being, to judge from Claparéde’s figure, distinctly notched terminally. lo a THE EXOGONE.. 2: EXOGONE HETEROSETOSA, McInt. (Plate 1%. figs. 11-17.) Exogone heterosetosa, McIntosh, 20, p. 205, pl. 38. figs. 15 & 16; pl. 544, fig. 11. Exogone heterosetosa, McInt., Ehlers, 8, p. 51, pl. 3. figs. 61-65. ? Exogone heterosetosa, McInt., Gravier, 15. Exogone heterocheta, McInt., Augener, 1, p. 247. The specimens of this species which I have before me are all small worms, not more than 4 mm. in length. There are about 40 segments. The palpi are elongated, longer than the prostomium and peristomium, closely united, but separated distally by a distinct notch which is well marked in late fixed embryos. The unpaired tentacle extends as far as, or slightly beyond, the extremity of the palpi. The lateral tentacles are not more than half the length of the median. As regards the peristomial tentacles and the dorsal cirri, there are two types. In the one all these appendages take the form of stout cylinders. In the other they are all oval. The dorsal cirrus of the second segment in all my specimens is absent. The ventral cirri are smaller than the dorsal. The anal cirri are long and tapering, similar in shape and size to the median tentacle. Of the sete the most dorsally placed (PI. 17. fig. 12) is a simple, curved, pointed form similar to the corresponding structure in L. fustifera. The next (figs. 13 and 14) is a compound seta of a characteristic form. The shaft is expanded at the end into a broad plate with an oblique, finely crenulated edge. With this articulates the terminal appendage, which usually takes the form of a very thin triangular plate sometimes replaced by a slender tapering rod. The remaining compound sete (figs. 15 to 17), three or four in number, are all similar to the corresponding sete of /. verrugera, with small but well-formed appendages, or one may partake to some extent of the special characters of that just described. Sexually mature specimens have in addition bundles of capillary setae on the segments from the twelfth or thirteenth backwards. In the most complete specimens the proboscis is partly protracted and the proventriculus lies in the 3rd, 4th, and 5th segments. In another, apparently retracted, it lies in the 4th, 5th, and 6th. The proventriculus is similar to that of /. verrugera, and has about 20 rows of muscle-columns. Augener refers to Ewogone heterocheta, McInt. (obviously a slip for FE. heterosetosa) as one of the commonest species in South-west Australia, and states that his determination of the Australian specimens was confirmed by comparison with southern-antarctic specimens. But Iam not sure that he has not confounded this species with the one which I have set down tentatively as L. verrugera. In E. heterosetosa the presence of the notch between the palpi (which is well sbown in McIntosh’s original figure), and the absence of the dorsal cirri of the second segment appear to be constant features (recognizable even in LINN. JOURN.—ZOOLOGY, VOL. XXXIV. 17 222 PROF. W. A. HASWELL ! attached larvee) distinguishing this species from the last, apart from the differences in tentacles and sete. E. heterosetosa occurs in Port Jackson much more sparingly than /. fusti- fera. GRUBEA PUSILLOIDES, n. sp. (Plate 17. figs. 27, 28.) This small Syllid, not much over 2 mm. in length in the case of the largest specimens, is not uncommon in certain situations between tide-marks in Port Jackson. It is nearly allied to G. pusilla (Dujardin) ; but, unless the published accounts err in certain important particulars, it is quite distinct from that species, which has been found hitherto only in the Mediterranean and in the English Channel. In describing G. pusilla Dujardin fell into the error of regarding it as hermaphrodite, the rod-like bodies developed in glands in the dorsal cirri being supposed by him to be spermatozoa. This misconception was corrected by Claparéde (8. p. 44: 4. p. 89) who pointed out the true nature of these bodies. More recently Baron de St. Joseph (3). p. 79) has observed males with natatory setee and spermatozoa and females with embryos on the ventral side. The most interesting point about the Australian species is that, whereas its very near relative was originally supposed (erroneously) to be hermaphrodite, and proved when critically examined to have the sexes separate like the vast majority of the Polycheeta, the new form actually has the sexes united. The details of this condition will be given in the general part of this paper. A complete specimen of Grubea pusilloides is about 2 mm. long and contains some 25 to 33 segments. The palpi, united dorsally at their bases, are separated distally by a cleft which may be widened to a distinct gap: they are approximately equal in length to the prostomium, The prostomial tentacles are dilated at the bases, narrow distally, tipped with non-motile cilia: they project a little distance beyond the ends of the palpi, the median tentacle being slightly longer than the others. There are two pairs of small eyes, the anterior pair the larger. The peristomial tentacles are similar in character to the prostomial, the upper of the two on each side being some- what longer than the lower. The parapodia (see text-fig. 2, p. 235; Pl. 17. fig. 28) are simple, pointed, their length about a third of the breadth of the body. Of the sete, in all but the first few segments the one situated most dorsally is simple, almost straight, gently curved towards the end, with a row of very minute fine teeth along its concave border *. All the rest are compound, with long and slender terminal appendages, The dorsal cirri are, like the sete, very similar to those of G. pusilla. * The seta in G, pusilla corresponding to this is described nearly as above by De St. Joseph, and the figure he gives agrees well with the corresponding seta in G. pusilloides. But he erroneously refers to it as ventral instead of dorsal, THE EXOGONEA. 225 They are sub-cylindrical, relatively thick, truncate distally, the edge of the terminal face finely crenulate and with a few long non-motile cilia. The most striking feature of the dorsal cirri is the presence in each of a pair of spindle-shaped bundles of thread-like objects situated towards the distal end, each enclosed in a sae which is produced into a narrow prolongation or duct ending at the free end of the cirrus. These, as pointed out by Claparéde, are the equivalents of the ‘ glandes & bitonnets” (rhabdite glands) found more deeply placed in some allied Syllids such as Sphwrosyllis hystria. In close opposition with these is a rounded sae with granular contents, having like them a prolongation or duct opening at the free end of the cirrus. The ventral cirri are short, cylindrical, not nearly so long as the parapodia. The anal cirri are like the tentacles in shape and are half as long again as the dorsal cirri ; there is no unpaired process between them. The proventriculus in the retracted state of the proboscis lies in the 4th and 5th or 5th and 6th segments. Its walls contain about twenty rows of muscle-columns. GRUBEA KERGUELENSIS, Melntosh. (Plate 17. figs. 18-20.) Salvatoria kerguelensis, McIntosh, 20, p. 188, pl. 380. fig. 4; pl. 85. fig. 1; pl. 154. figs. 11 & 12, Spherosyllis Macintoshi, Mhlers, 8, p. 46. Grubea kerguelensis, Augener, 1, p. 252, text-fig. 57 ; Taf. 3. fig, 23. ? Grubea limbata, Claparede, 2, p. 208, pl. 13. fig. 4. A species of Grubea which is fairly common in Port Jackson corresponds for the most part to Claparéde’s description and figures of his G. limbata, with certain differences, the most important being in the shape of the sete. I think the Port Jackson species may be the same as one found common in the collections from South-west Australia studied by Augener, and identified by that author with G. herguelensis (McIntosh). I do not think that sufficient grounds for that determination aye adduced, but here accept it provisionally. Meanwhile I give some figures of the Port Jackson form which may help to solve the question. It is to be noted that Claparede gives as one of the differences between his G. limbata and his G. tenuicirrata that in the latter the second dorsal cirrus is longer than those that follow. But this can only be a matter of degree since in his figure of the former species he represents the second dorsal cirrus as distinctly longer than those behind. GRUBEA QUADRIOCULATA, Augener. (Plate 17. figs. 21-26.) Grubea quadrioculata, Augener, 1, p. 254, text-fig. 58; Taf. 3. fig. 31. This Grubea, which Augener found rare in the South-west Australian collections, occurs, also somewhat rarely in my experience, in Port Jackson, It is most clearly distinguishable from the preceding by the absence, in the ict 224 PROF. W. A. HASWELL: majority of specimens, of the frontal eyes, and the greater size of the proventriculus. Augener’s fig. 31, though it doubtless represents his material accurately, does not correctly represent the head-end as it appears in living or well-fixed specimens, and I give here a new figure as well as one of the parapodium and the sete. The two species here referred to as G. kerquelensis and G. quadrioculata do not differ very widely from one another. The latter is a larger and more robust form than the former, has a cumparatively large proventriculus, and usually only four eyes. But six eyes are present in a considerable proportion of cases. SPHEROSYLLIS HYSTRIX, Claparede. (Plate 18. figs. 32-35.) Spherosyllis hystrix, Claparéde, 3, p. 45, Taf. 13. figs. 36, 37, Spherosyllis hystrix, Marenzeller, 23, p. 25. Spherosyllis hystrix, McIntosh, 21, p. 156, pl. 59. figs. 3, 4,8; pl. 70. fig. 1; pl. 78. figs. 11-18. Tn describing S. pirifera (Ann. Chet.) Claparede refers to a few points of difference from S. Aystriv previously described by him. The cirri and tentacles he states are relatively shorter and more dilated at the base and always curved or hooked at the ends. Marion and Bobreizky (24) regard S. hystrix and S pirifera as identical or at least not specitically distinct. The nearness of the two forms is emphasized by De St. Joseph. The distinctive character, he states, is the existence of the “glandes 4 batonnets” in S. hystrix. A Spheerosyllis which is extremely common in Port Jackson conforms closely to Claparéde’s account of S. pirifera as supplemented by Viguier (Arch. de Zvol. exp. et gén., 2 sér., tome ii. (1884)). But it possesses the “ olandes 4 bAtonnets,” and should thus, by De St. Joseph’s interpretation, be assigned to the ‘older species. This, however, is contradicted by De St. Joseph’s statement that in the latter the yellow glands are absent. Moreover Viguier states that S. hystriv remains always transparent and does not become encrusted with foreign bodies. As the Port Jackson form, like the specimens collected by Pierantoni in the Gulf of Naples, frequently possesses both the rhabdite glands and the yellow glands, I have followed the last-named author in bis conclusion that the presence of the structures in question or their apparent absence is not to be relied on as a specific distinction. I am inclined to believe that both sets of structures are present in all S. hystria and S. pirifera specimens, but. that one or other, owing to its physiological or developmental condition, may sometimes not be readily recognizable. As the Australian specimens differ in some minor points from the published descriptions, [ think it desirable to describe and figure this common Australian form pretty fully, while still retaining for it for the present the name of S. hystriv. THE EXOGONE®. 995 The length of a mature specimen is, on the average, about 5 mm. with a breadth of about *5 mm. The namber of segments varies from about 25 to over 40. The length of the segments is approximately half the breadth. The colour is grey or reddish- or yellowish-grey, and is regulated by the amount of adherent débris. The colour, when present, is due to diffused irregular dots of red pigment. The palpi, which equal prostomium and peristomium in length, are, as in other species of the genus, closely united together, separated by a median groove and a terminal notch. The distal part is contracted, the narrowing usually taking place rather abruptly ; the narrower terminal part is often bent sharply downwards so that it becomes invisible when the animal is looked at from above. Mést specimens have the palpi entirely devoid of papillee, but in a few cases two symmetrically placed pairs of very distinet papillee are present on the dorsal surface of these anpendages. ‘The tentacles are of equal length, scarcely extending as far as the distal end of the palpi. The basal portion of each is dilated, the distal narrow—the dilated and narrow parts being of nearly equal length. Under some conditions, or in some individuals, the dilated parts of the tentacles (and the same holds good of the cirri) become, almost spherical and the narrow terminal parts greatly reduced. There are never more than four eyes, the “frontal” pair present in some nearly-related species being absent. The size of the eyes varies greatly, being greatest in individuals with ripe sexual products. The single pair of peristomial tentacles are similar to the prostomial. The dorsal surface of the peristomium and the segments of the body are ornamented with papillee. The surface of the body is very usually covered with a layer of granular débris adhering to the surface, and this may com- pletely hide the papillee, but, when the latter are distinguishable, their arrangement is extremely regular, about fifteen pairs being symmetrically placed on each segment. On the ventral side also papillae are present, but these are smaller and less numerous than those of the dorsal surface. The dorsal cirri closely resemble the tentacles in form and size. They are about equal in length to the corresponding parapodia, but may project beyond them. They bear each a small number of papillee, but these are never so prominent as those on the body and parapodia and are sometimes very obscure. The dorsal cirri of the second setigerous segnient were absent in all the specimens examined for them. The ventral cirri are smaller than the dorsal, and not definitely dilated at the base. The anal cirri do not differ notably from the largest of the dorsal cirri and resemble them with regard to the papille. There is no posterior median process, but three pairs of papillae, rather larger than those on the general surface, lie between the anal cirri. The parapodia (P1, 18. fig. 33) are in length about half the breadth of the body, of compressed conical shape. Hach bears two prominent papill towards its extremity. The single aciculum has its terminal part bent 226 PROF. W. A. HASWELL: nearly at right angles with the rest of its length in the manner described by Viguier (34), by St. Joseph (31), and by Pierantoni (28). The most dorsally placed seta is simple, gently curved towards the end, and pointed ; along the concave edge of its terminal part runs a row of exceedingly small hair-like teeth. The remaining sete, usually six in number, are all com- pound and of the same general character, with well-developed, though short, terminal appendages. In sexually matured specimens, but not in females bearing external ova or embryos, there are bundles of long capillary sete (natatory or puberty sete) beginning on the 12th to the 14th segment and ending on the 28th to the 35th. The proventriculus (gizzard), when the proboscis is fully retracted, lies in the filth and sixth setigerous segments. Its wall contains about 15 rows of muscle-columns. Yellow glands and “ glandes a batonnets”’ are both present. The eggs and embryo are borne in pairs on the ventral side. SPHAROSYLLIS PERSPICAX, Ehlers. ? Spherosyliis kerguelensis, Mclutosh, 20, p. 206, pl. 29. fig, 5; pl. 53. fig, 10; and pl. 15 a. fig. 22. Spherosyllis perspicax, Khlers, 11, p. 66, tab. 6. figs. 1-3. ? Spherosyllis antarcticus, Gravier, 15, p. 12. Spherosyllis perspicax, Augener, 1, p. 250. The Spherosyllis which I refer to the above-named species is compara- tively very rare in my experience in Port Jackson. Augener found it also rare in the South Australian collection of the Hamburg Expedition to South Australia. It differs from the common form which I have referred to S. hystrix, (1) in having the palpi relatively shorter, not produced into such a long narrow anterior prolongation, and beset with well-formed papilla; (2) in the presence of the third or frontal pair of eyes ; (8) in the acicula being straight to the end; (4) in the much greater length of the proventriculus, which extends nearly throuzh four segments, and has about 20 rows of muscle-columns in its walls; and (5) in the embyros being borne on the dorsal side, four on each segment. SPHAROSYLLIS HIRSUTA, Ehlers. Spherosyllis hirsuta, Ehlers, 8, p. 48, Taf. 3. fies. 58-60. Spherosyllis hirsuta, Ehlers, Annelid. d, Valdivia-Exped. Band xvi. p. 66. Spherosyllis hirsuta, Ehlers, 10. Spherosyllis hirsuta, Augener, 1, p. 249. This species appears to he common in South Australia, but I have not hitherto succeeded in finding it in Port Jackson. The embryo-bearing females should be readily distinguishable from those of S. hystriw by the dorsal position of the embryos and their occurrence in fours. From S. perspicax the absence of the frontal eyes seems to be the only important distinction, THE EXOGONEA, ho no ni INTEGUMENTARY AND PEDAL GLANDS. In E. fustizera the integumentary glands display a remarkable develop- ment. In each segment near its posterior limit they are arranged in a broad zone (PI. 18. fig. 49) which is much more strongly marked on the dorsal side than on the ventral. The secretion has a strong affinity for hematoxylin, and in many stained specimens the result is the appearance of a very characteristic pattern of transverse bands. The individual cells are irregular in shape and of small size for the most part; but in the middle ventrally, closely applied to the nerve-cord (see Pl. 18. fig. 48 and tig. 50) are two or three of considerably larger size—these groups of larger cells having the appearance, in entire well-stained specimens, of dividing the nerye-cord into lengths corresponding to the segments. In series with this zone of integumentary glands, but on a deeper level and projecting into the ccelom from the body-wall, there are on each side in the posterior part of each segment three larger rounded bodies (Pl. 18. figs. 44 & 45, i.gl.; Pl. 18. figs. 41 & 42), each made up of a group of uni- cellular glands with ducts opening on the surface about the lateral border of the segment behind the parapodia. A feature of the glands last referred to which may be of physiological significance, is the development within them, or some of them, of excessively minute granules which pass out from the substance of the gland-cells into the celom as definite corpuscles (PI. 17. fig. 29). The pedal glands are referred to by Malaquin (22), but only in the most general terms. In Exogone fustifera the pedal glands are two pairs in each segment— a dorsal and a ventral (Pl. 18. fig. 49, v.p.g.), each: placed in close relation to the corresponding cirrus and partly contained within it. Of these the ventral, situated for the most part just internal to the ventral cirrus, but extending also into the interior of the latter, is of special interest since in the female it is the gland secreting the viscid substance by means of which the ova when discharged become attached to the parent. In £. verrugera the integumentary glands are not condensed into transverse zones as in /. fustifera. They are fine convoluted tubules which are most abundant in the anterior region of the body. The pedal glands (Pl. 18. fig. 36) are very similar to those of /. fustifera. In Spherosyllis hystrix the secretion of the integumentary glands forms a resistant layer in which small particles of grit become thickly embedded. This may be heaped over and around the papille in such a way as to give rise to the appearance of a system of comparatively large tubercles regularly distributed over the dorsal surface. This does not seem to be of the nature of a permanent layer since it varies greatly in thickness, and specimens are sometimes met with in which it is entirely absent. This point has already been referred to in the systematic part of this paper. 228 PROF. W. A. HASWELL ! The ventral pedal glands (PI. 11%. fig. 30) are usually conspicuous in S. hystriw. In the female their secretion has, as in Hwogone, the function of effecting the attachment of the eggs to the ventral surface. In Grubea kerguelensis and G. quadrioculata the integumentary glands are diffused and inconspicuous. “Capsules a batonnets” or rhabdite glands are not universal in the Ewogonee. Among the species dealt with in this paper they occur in Spherosyllis hystria, Grubea pusilloides, and G. kerquelensis. In Spherosyllis hystrix each is an oval sac situated on the dorsal and posterior side of the corresponding bundle of setze and opening on the surface on the dorsal side of the parapodium: the slender rhabdites are often to be seen protruding through the external aperture. In Grubea pusilloides, as already described, there are two of these bodies in each of the remarkably modified dorsal cirri (PI. 1%. fig. 27). In G. kerguelensis they have almost exactly the same position as in Spherosyllis hystrix and are found in all the segments from about the eighth backwards. There is little doubt that where they occur these rhabdite glands represent a modification of the dorsal pedal glands of other members of the group. In the absence of direct observation it may be conjectured that the main functions of the system of integumentary glands are protective: the viscid matter which they secrete becoming strengthened by the inclusion of foreign gritty particles may form a temporary investment as in Hwogone, or may become closely adherent to the integument as in Spherosyllis. With such a protective function is doubtless associated the entangling of animals seized upon as prey, permitting of their being effectively attacked by means of the proboscis. The pedal glands, dorsal or ‘ventral, as the case may be, assume in the female the special function of secreting the viscid matter by the agency of which the ova are caused to adhere firmly to the surface: since the same plands are equally developed in the male, it seems probable that in the latter sex they also have a special function—perhaps connected with fertilization. The rhabdite glands, which in part replace pedal glands in some cases, doubtless bave a specialised function in connection with the capture of prey and the warding off of enemies. ALIMENTARY CANAL. The pharynx in Lxogone fustifera has the usual single tooth and the circlet of conical papille, which are eight or ten in number. In the living speci- mens in some cases droplets of clear liquid are to be detected exuding from apertures at the ends of the papille, and in connection with these are traceable fine ducts which are the ducts of the glands of the papillae (pharyn- geal glands). These (PI. 17. fig. 1, p.g.) are about ten in number, in the THE EXOGONEA, 229 shape of narrow cylinders, which extend backwards parallel with the pharynx, encircling it closely and becoming intimately attached to it behind. Hach of these cylinders appears to be of the character of a group of unicellular glands each terminating orally in a slender duct. These glands were observed by Claparéde in various Syllids. Thus in his account of Grubea limbata (2. p. 208) he says :—‘* Tout autour (de la trompe) sont disposés des boyaux Capparence glanduleuse.” In his figure of Pado- phylaw claviger (pl. 13. fig. 2) the letter g points to a narrow body at the side of the pharynx which is referred to in the explanation as “ glandes de la trompe.” But this, though he seems to have mistaken it for them, has nothing to do with the “boyaux d’apparence glanduleuse,” being situated within, not without, the wall of the pharynx, and probably is an indication of the body to be described presently under the designation of anterior proventricular glands. He overlooked altogether the occurrence of the structures in question in the larger Syllids, and only referred to them in connection with Spherosyllis, Padophylax (Exegone), and Grubea. The definite connection with the papillee he appears not to have ascertained. De St. Joseph (31), on the other hand, is very clear on both of these points. Malaquin (22. p. 196 ef seg.) gives a very complete account of these pharyngeal glands and I have nothing to add to his observations. With regard to the functions which they may be supposed to discharge he remarks (p. 198), “Le contenu de ces glandes, se déversant par les papilles, a pour bout de les faire adhérer fortement a la proie, et, d’un autre cdte, il a peut- étre un role spécial dans l’absorption des aliments ; il se pourrait encore qwil ait des propriétés toxiques destinées a neutraliser les mouvements de cette proie.” g The pharynx has the usual thick cuticle with an epithelial layer of ill- defined cells which contain numerous yellowish granules. Posteriorly this layer swells out to form the anterior proventricular gland (PI. 1%. fig. 1: PI.1%. fig. 81: Pl. 18. figs. 37 & 38). This is a mass of irregular cells with, here and there, large vacuoles containing the secretion. From these a number of ducts, some wide, some narrow, pass backwards through the thick epithelium of the anterior part of the proventriculus, to open eventually into the lumen of the latter. The structure just referred to seems to have escaped the notice of previous observers. The only reference to it which I have been able to find is the following statement of Malaquin’s (22. p. 198): ** Dans la région postérieure de la trompe pharyngienne |’épithelium devient alvéolaire, et prend partout lapparence d’un épithelium sécrétant, c’est qu’en effet, l’accroissement de la trompe et par conséquent de la chitine se fait par cette extrémité, d’avant en arricre.” This however, though describing the superficial appearance of the part, overlooks its real character and the special destination of the secretion. 230 PROF. W. A. HASWELL: The structure of the muscular organ which I have called the gizzard, but for which the name proventriculus used by Malaquin and others will be employed here, is similar in all essential respects to that of the corresponding organ in the Syllide. The proventriculus is a thick-walled cylinder, the lumen of which is reduced in the passive condition of the organ to the form of a vertical slit. Along the dorsal and ventral median lines opposite the ends of the slit run dorsal and ventral raphes along which the thickness of the musculature is greatly reduced. The greater part of the thickness of the wall is taken up by radiating muscular columns arranged in annular rows. Malaquin (p. 213) states that the wall of the proventriculus consists of the following layers :—(1) peritoneal layer; (2) external circular muscles forming a thin layer; (3) radial muscle-columns separated by transverse diaphragms; (4) internal circular fibres; (5) the general columnar epi- thelium of the digestive tube ; (6) the cuticle. How far this is correct in relation to the Syllidea, Husyllidea, and Auto- lytidea will not be discussed in this paper. In Hwvogone fustifera (and the other Lwogonew which I have studied) there are no layers of circular muscle —the only circular fibres being those in the so-called diaphragms. That the radial muscle-columns are in the Syllidee striated musele-fibres of a primitive type was pointed out by me (16) in 1886, and the subject was further elaborated in 1839 (17). © Malaquin (22), who was unacquainted with my second contribution to the subject, arrived independentiy at very similar results. In Heogone fustifera, as in all the other Mwogonew which I have examined, the muscle-columns (PI. 18. figs. 387, 39 & 40) are non-striated : they contain no doubly-refracting substance and are devoid of transverse networks. In other respects they closely resemble the muscle-columns of Syllidea. They have a cortex of muscle-substance made up of fibrillee, and a core of a granular protoplasmic material. Near the squarish outer end of the column is a nucleus usually single in the middle of the core ; and in this region the protoplasm of each muscle-cell communicates with that of its neighbours in the same row by means of narrow processes. In addition to these peculiar muscle-columns with their protoplasmic cores, the wall of the proventriculus also contains muscular fibres of the same character as those occurring in other parts of the body. These are arranged in narrow annular bands corresponding to the annular rows of muscle- columns. Each of these in Hwogone fustifeva (Pl. 18. figs. 39 & 40) runs through the muscle-colunms of the corresponding annular row close to their outer ends. The muscle-columns of /. fustifera are divided by the circular muscles, THE EXOGONEZ!. 23 usually unequally, the circular ring passing through at the level of the single nucleus of the column or deeper (according to the condition of contraction ?). The cireular fibres have their own nuclei at intervals. The internal epithelium and cuticle of the proventriculus, both very thin, present no special features of importance, except that the former is censider- ably thickened both in front where the ducts of the proventricular glands open, and behind in the neighbourhood of the entrance to the ventriculus. The proventriculus is invested in a thin but resistant membrane which 1s doubtless derived partly or wholly from the splanchnic layer of ccelomic epithelium. It contains flattened nuclei at long intervals. In sections stained with iron-hematoxylin, it shows a fine irregular network of fibrils *. The rest of the Ewoyonew examined by me resemble Lwogone fustifera in the structure of the pharynx and proventriculus ia all essential respects. The pharyngeal and proventricular glands are present in all. In all with one exception the rings of transverse fibres perforate the outer ends of the radial muscle-columns, dividing them either equally or unequally. The only exception is an undetermined species of Grubea in which the rings are arranged regularly between the annular rows of muscle-columns. The ventriculus of Hwogone fustifera, somewhat shorter than the proventri- culus, is nearly as broad as the latter in front, but narrows behind towards the junction with the intestine. The wall of the organ contains radiating muscle-columns which are more widely spaced than those of the proven- triculus and of simpler structure—solid and without protoplasmic core. Circular rings of fine muscular fibres run between the rows of columns close to the outer surface. The internal epithelium does not form a definite layer, but its cells lie regularly among the muscle-colummns. Here and there in the wall of the organ lies a unicellular gland. ‘Ihe ducts of these, the posterior proventricular glands, do not open into the lumen of the ventriculus but run forwards and pierce the e;ithelium to open into the post-rior part of the proventriculus. In Spherosyllis hystrix the ventriculus has a definite epithelium. In Grubea kerguelensis and G. quadrioculata the ventriculus is so greatly reduced as to be practically obsolete. The other Laogonee studied resemble Lwogone Justifera in the structure of this part of the alimentary system. * With regard to the relations between the circular and radial muscular fibres in the wall of the proventriculus, it is to be remarked that Malaquin takes a view which dues not accord with the account above given. [He regards the annular diaphragmis, complete or iucom plete, as always separating the rows of radial fibres, and he is thus led to the conclusion that such an arrangement as that ocemring in Zvogone fustifera (which is common also in the Syllidea) means that the radial mus le-columns of each row are arranged‘ back to back.” But since it is exclusively with the Sydl/dea that he is dealing, it will be best to defer for the present entering further into this subject. 232 Pror. W. A. HASWELL ! NEPHRIDIA. | Characteristic of the nephridial system in Mvogone fustifera is the fact that in both sexes the nephridia of each pair unite completely towards the time of sexual maturity. In immature stages they are represented by pairs of narrow ciliated tubes of the usual character, with ciliated funnels. But in this and subsequent stages the walls of these organs are composed, not of few and comparatively large cells, as observed by Goodrich (14) in other Syllids, but of numerous small cells, as noticed by Claparéde (2. p. 213, pl. 13. fig. 2A) in the mature male of Hxogone clavigera. In the female in the earliest stage of sexual differentiation observed (PI. 18. fig. 48), each of the segments from the 12th backwards contains a pair of small ova, each enclosed in a sae with small-celled walls having a tubular outgrowth opening at the nephridiopore, he ciliated funnels at this stage appear to be entirely separate from the sacs. The derivation of the ova is unknown. In specimens bearing embryos, in the segments on which embryos are situated, and in several in front and behind, nephridia are present having the following features. Each is a small oval sac with small-celled walls opening on the exterior through the nephridiopore, and of a short narrow canal leading from it to the adjacent nephridial funnel on the other side of the septum. The sac is contractile, the eontractions being rhythmical though irregular, and the rhythm corresponding roughly with that of the contrac- tions of the ventral vessel. The contractions have the effect of drawing the walls of the organ towards the attached external end. In sections it is seen that a transverse branch from the ventral blood-vessel ends blindly imme- diately in front of the nephridial sac, and it is probable that it is the contractions of this branch vessel that stimulate the contractions of the wall of the sac. At the stage now referred to, the sac does not contain an ovum —merely an irregular coagulum. But specimens without embryos were found as already stated having the nephridia in almost exactly the condition just described except that each sae contains a small ovum. The ovum increases in size, distending the enclosing sac till iis wall becomes reduced to a very thin membrane—the tubular efferent part being permanently retained with little alteration. As the two ova in each segment increase in size their enclosing sacs grow inwards towards one another. fiventually they meet in the middle line (PI. 18. fig. 49) and unite to form a single spacious sac enclosing the two ova and having on each side the original communication with the exterior. When the two sacs communicate, the to ova, instead of lying one on either side of the segment as at first (tig. 50), take up a new position one in front of the other, each extending transversely right across the cavity of the segment (fig. 51). When the ova are discharged from the nephridial sacs each at once becomes attached by one end to the small area of the ventral surface (immediately THE EXOGONEA, 233 internal to the base of the corresponding ventral cirrus) on which the ducts of the ventral pedal gland open. A mass of the secretion of the gland in question has collected within the gland : this must be pressed out, presumably as a result of the same contraction which leads to the discharge of the ova, and forms a viscid plug to which the ovum becomes fixed. The position assigned by Viguier (34) in his description and figure (p. 89, pl. 4. fig. 19) to the discharged ova of EL. yemmifera corresponds very closely to what I find to be the arrangement in L. fustifera. He is not right, how- ever, in the statement that the point of attachment corresponds exactly to that of the opening of the ‘‘vésicule séminale’’ of the male: it is dis- tinetly behind that point. The sequence of events in the male is more complicated, and I have not been able to follow it out completely. The original simple, nearly straight, obliquely directed tube of the nephridium begins to become complicated in the manner described below. At this stage there is as yet no appearance of sperms. Later, when a mass of sperms has appeared, the pair of nephridia are represented by much looped tubes with walls of the chacteristic small cells. Subsequently the sexual segments have their ccelomic cavities (en- larged by the extreme attenuation of the intestine), with the extensions into the parapodia, filled with sperms—the nephridia in each segment now being represented by a single sac (Pl. 18. fig. 45), apparently containing only a watery liquid with which it remains distended, and communicating with the exterior laterally by the nephridiopores. Sperms subsequently enter the nephridial sacs—presumably through the ciliated funnels. From the fre- quency with which this stage is met with it would appear to last for some time—the huge stock of accumulated sperms gradually being drawn into the nephridial sacs and becoming passed to the exterior. ~ When the male nephridia first appear they are developed as hollow in- growths from the surface in the position of the nephridiopores. Hach of these grows obliquely forwards and inwards, and becomes completed by passing through the mesentery and terminating in the ciliated funnel, which seems to be independently developed. The nephridium has now the character of a straight, or nearly straight, tube running obliquely from the nephridio- pore on the ventral side of the segment a little in front of the parapodium to the nephrostome opening into the cavity of the segment next in front. Its wall, as in the case of the female, is composed of a single layer of small cells. The terminal part of the tube is dilated into a contractile sac as in the female. The first change that takes place is the giving off of a branch from the sac not very far from its external opening (PI. 18. figs. 41 & 42). This grows backwards and becomes thrown into a loop. At the same time the original tube is drawn out into a loop which extends inwards and nearly meets its fellow of the opposite side in the middle line on the ventral side of the intestine. 234 PROF. W. A. HASWELL ! Meanwhile the rudiments of the testes have appeared as masses of narrow tubes surrounding the nephridia closely and extending round the intestine towards the dorsal side. The origin of these testicular masses is difficult to trace. They are developed in close association with the nephridial tubes and are composed of cells at first very similar to the cells composing the walls of these tubes. But they are also from the first, or from an early stage, closely associated with the peritoneal lining of the sides of the intestine. Before the testes are formed, or while they are still small, there is to be seen in the living animal in front of the nephridium a colourless thin-walled sac which contracts at intervals. This must correspond to the dilated end of the vascular cecum described by Malaquin (p. 383) as forming the axis of the testis in the Syllide. It becomes hidden in Ewogone fustifera when the testes develop, and I have been unable to trace it in sections. Its equivalent in the female is referred to above. Later, when masses of sperms have been set free in the ccelom, the nephridia come to be represented by two relatively wide tubes (P1. 18. fig. 43), one on either side, still connected with the nephridiopore and provided with a ciliated” nephrostome. Each becomes a sae with narrow lumen and thick wall com- posed of numereus small cells. The lumen of each sac increases in extent (P]. 18. fig. 44) and the cavities of each pair at length unite, though partly separated for a time by a septum which at last disappears, the wall of the unpaired sac (PI. 18. fig. 45) thus formed becoming at length thinned out till it assumes the character of a single layer of small cells. The presence in /2. clavigera of “ vésicules séminales ” which were in all probability formed each by the fusion in the middle line of a pair of nephridia was observed by Claparede (2. p. 212). He does not describe a corresponding fusion in the female. Viguier (34. p. 86) agrees with Claparede regarding the unpaired pouches of the male, but describes them as opening behind and not in front of the ventral cirri. In the ripe female he describes (p. 88) the two ova in each segment as increasing in size till they meet below the intestine, and one may pass in front of the other. From the results of compression he concludes that the apertures of egress are in front of the base of the parapodium. De St. Joseph (31. p. 86) confirms Claparede’s account as regards the male. With regard to the female he states :—“ Une femelle de 25 segments a dans les 10e.a 23e. segments une masse yitelline grise sans noyau, entourée Wune membrane et commencant a se segmenter en deux; située sous le ventre entre l’intestin et la paroi du corps elle occupe presque toute la largeur du segment. Hst-ce une poche renfermant deux ceuts dont je n’ai pas vu le noyau et que commence a se séparer en deux? UWst-ce un ceuf unique a son premier stade de segmentation ?” Grubea pusilloides has the sexes united. ‘lhe occurrence of this condition, so rare in the Polycheeta, has been recorded by Du Plessis in another species HE EXOGONEA. 235 of the same genus—G. protandrica. In G. pusilloides there is no protandry, so far as I have been able to observe—active sperms and maturing ova occurring together. In a sexually mature specimen, before oviposition takes place, one large ovum is found in each of the 11th or 12th to the 16th or 17th segments (text-fig. 2). Hach of these is lodged in a thin-walled sac formed by the union of the two nephridia of the segment and opening on the exterior by the two nephridiopores, situated just in front of the parapodia on the ventral side. Text-fig. 2. Development of ovum in Grubea pusilloides. d.c., dorsal cirrus; 7., Intestine ; 2.¢., ventral neryve-cord ; np., nephridiopore ; ov., ovum ; v.c., ventral cirrus. The testes are developed as small irregular bodies in close connection with the nephridia in the 9th and 10th or 10th and 11th segments. There is no union between the nephridia of opposite sides. The sperms in the 9th seg- ment collect in the posterior part of the segment and doubtless are taken up by the ciliated funnels, and passed to the exterior through the nephridia of the 10th segment. Sperms still remain active in specimens with three externally attached ova. In specimens with the full number (five) of external ova attached to the 10th to 15th segments sperms were still to be found in the male segments in front. De St. Joseph states (31. p. 79) that in G. pusilla the female has two e attached to each segment from the 10th to the 26th. In Spherosyllis hystria in the male the early development of the nephridia and testes takes place very much as in /aogone fustifera. Later the inner o0S 55” portions of the testes unite in the middle line. In mature males with the celont full of sperms, the two nephridia of each seement, which remain quite separate from one another, have assuined the form of coiled thick- walled tubes with degenerate epithelium. In the female the two ova developed in each segment (PI. 18. fig. 16) are in early stages of growth, each enclosed in a sac with a wall composed of a single layer of cells and apparently corresponding to the nephridial sac of other Exogonew. But in specimens with full-grown ova this investment has completely lost its cellular character and has quite the appearance of a cuticle. 236 PROF. W. A. HASWELL : In young females of Grubea kerguelensis there are four small ova on each side in each segment. Two of these occupy a position in the segment corresponding to that occupied at first by the only two in Exogone fustifera, lying nearly transversely—the other two are situated laterally. Each of the transverse ova is situated in a thin-walled sac (nephridium) opening in the usual position, and there can be little doubt that the investment of each of the other ova is a prolongation of this sac. So far as the material at command goes, Grubea quadrioculata corresponds closely in the female reproductive apparatus to G. kerguelensis. TWMBRYOLOGY. For our earliest information on the development of the Hxogonew we are indebted to Ersted (25), who gave a brief account of some of the stages as seen in his Exrogone naudina. Viguier (34) notes some points in the development of Ewogone gemmifera. The segmentation stages figured do not, the author acknowledges, throw any light on the mode of formation of the germinal layers, but he thinks it probable that there is an invagination the orifice of which becomes the anus. The buceul invagination appears somewhat late. When the larva becomes detached it possesses five setigerous segments. De Saint Joseph (81) describes various stages in the later development of EF. clavigera and makes an interesting observation on the difference between’ the development of Grubea clavata and G. pusilla which will be referred to later. : The only other recorded observations on the embryology of the Exogonew are those of Pierantoni (28). This author made a noteworthy contribution to our knowledge of the remarkable phenomenon of epigamy; but the objects which he aimed at in his research did not include a complete study of the embryology. His results as regards the embryology of Spherosyllis hystria may be summarised as follows :— The ovum divides into two blastomeres of which one divides rapidly into two, four, and eight, while the other divides much more slowly. From the successive divisions of the former blastomere there result a great number of micromeres (ectoderm) and from those of the other a very small number of macromeres (endoderm) which ultimately become enclosed within the micro- meres ; the former at this stage are only eight in number. i Both micromeres and macromeres increase in number, but, as regards the latter, after the s age of eight cells it is no longer possible to distinguish the individual elements, the whole appearing as a mass of red yolk-matter. In Evogone fustifera the ova when discharged are fixed to the ventral surface just internal to the ventral cirri—the attachment being effected, as already stated, by means of the secretion of the ventral pedal glands. A paiv is attached to each segment from the 11th or 12th as far back sometimes as the 43rd or 44th. bh THE EXOGONES. 37 The earliest stages of development closely resemble corresponding stages in the Polycheeta in general. Complete segmentation results in division into four cells from which the four primary micromeres are derived by further division. But I have not attempted to follow out the details of the cell- lineage—a task which, in view of the minuteness of the ova and the small number of stages available, would be one of peculiar difficulty. I can only record the increase in the number of the micromeres until, while the four macromeres still remain unaltered, they come to form a cap of small cells on one side of the embryo. I have not succeeded in observing a stage when the micromeres have completely enclosed the mass of macromeres. Or perhaps it would be more correct to say, that when such an investment has taken place, some differ- entiation has already been effected. Since no stage intermediate between those represented respectively by figs. 52 and 53 of Pl. 18 has been observed, it would appear probable that a process of differentiation goes on part passu with the extension of the investing layer of micromeres. This differentiation results in the formation of two centres or areas of development—the dorsally situated dorsal plate (Pl. 18. figs. 54-57, d.pl.) and the ventral stomodal tract (Pl. 18. figs. 54-57 and figs. 58 & 59, st.). The former is early distinguishable into anterior and posterior lobes, and consists at first of only a single layer of cells. The latter at the outset appears as a small rounded group of cells irregular in shape and arrangement in the middle of the surface of the embryo destined to be ventral, and about midway between the two poles. The dorsal plate sends off posteriorly a pair of outgrowths which eventually meet and unite at the posterior end of the body. These outgrowths, the germinal bands (q.b.), are separated from one another both on the dorsal and on the ventral surface by areas composed of ectoderm (enclosing the mass of yolk). Sections of embryos at this stage (Pl. 18. fig. 59) show that each germinal band is composed of two layers, a superficial with smaller nuclei —the ectoderm,—and a deeper with larger nuclei—the mesoderm. The stomodzeal area has in the meantime been undergoing differentiation. A depression has appeared on the surface which deepens and penetrates into the mass of cells as a sharply-defined cylindrical pit. Latterly the stomodzal rudiment sends out a string of cells which extend to the angles of the dorsal plate, or more exactly, of the posterior lobe of the latter. The destiny of those clearly defined bands passing out right and left of the stomodzeum and joining the apical plate dorsally, has not been followed. They very soon lose their distinctness. Each germinal band now undergoes a process of transverse segmentation (Pl. 18. figs. 62-65). In the region at the posterior end in which the two have become united they give rise to the rudiment of the pygidial segment. LINN, JOURN,—ZOOLOGY, VOL, XXXIV. 18 238 PROF, W. A. HASWELL : In front each germinal band divides into five segments, the first smaller than the rest, which are subequal. Sections show that, like the germinal bands, each segment is composed of two layers, a superficial, ectodermal, and a deeper, mesodermal (PI. 18. fig. 63). The area between the two rows of somites continues to be covered dorsally by a very thin ectoderm, Ventrally, on the other hand, at the stage when seomentation has been completed, is a thicker layer of closely-set small cells divided along the mid-ventral line by a raphe or suture (PI. 18. figs. 62 & 64). This is obviously the rudiment of the ventral nerve-cord. Whether it is formed from the thin ectodermal layer originally present in this position or by lateral ingrowth from the two germinal bands has not been actually determined. But the latter mode of origin, especially in view of the presence of the median raphe, is by far the more probable. At this stage the rudi- ment of the nerve-cord presents no trace of segmentation. The dorsal plate has thickened materially and extended over the anterior (free) end of the embryo (PI. 18. figs.58-64). The stomodeeum has extended and assumed a rounded form. Posteriorly it gives off an extension—the rudiment of the proventriculus—the original stomodzeum becoming the pharynx. The epithelium of the former has formed around it a simple layer of cells which tend to elongate radially, these are destined to form the radial muscular fibres of the proventriculus. External to this and closely, applied to it are flattened cells which may form a continuous layer. The next change of importance is the appearance of the first rudiments of the tentacles (figs. 64 & 65). The free or head end of the embryo had already developed a solid thickening—the rudiment of the closely united palpi. On the dorsal side of this appear two, to which a third (median) is soon added, small knobs—the beginnings of the cephalic tentacles. About the same time appear the rudiments of the anal cirri, The peristomial tentacles do not appear till considerably later. At this stage there are six pairs of mesodermal somites, the members of each pair widely separated from one another. When the rudiments of the parapodia first appear as slight lateral pro- jections they are developed from the somites. It appears to be clear that the coelom cannot be formed in the same way as in Polycheeta in general. At the stage when rudimentary setz first make their appearance in the parapodia the stomodaum ends behind in a mass of tissue in which the large yolk-granules are embedded. This tissue, derived as has been seen, from the macromeres, no longer retains the character of distinct cells, but has all the appearance of such a syneytium as that presented by the enteric or endodermal mass of the ‘“ Acclous ” Turbellaria or certain of the Rhabdocoles. In these the food is digested in larger or smaller vacuoles in the syncytium. In the embyro of Hwogone fustifera the yolk-granules, which take the place of food, are similarly THE EXOGONEA. 239 contained in vacuoles in the syncytium. This syncytial tissue with its contained yolk-granules completely fills the interior of the body and is continuous externally with the body-wall; there is at this stage neither mesenteron nor ccelom. The changes which result in the formation of these—the mesenteron and coelom—take place in the last stages before the young worms become free. In such a late larva (fig. 66) it is found that the syneytium has undergone a profound modification. A cylindrical space has been formed through the mass of yolk-containing tissue along the axis of the larva towards the dorsal side. This becomes continuous with the lumen of the stomodeeum in front and that of the proctodieum behind. The most advanced larvae observed have five setigerous segments. Ewogone verrugera and EK, heterosetosa have the ova arranged exactly as Li. fustifera, in pairs from the 10th or 11th segment backwards. The stages in the development are essentially the same, and the oldest fixed larvee have each five setigerous segments. In Grubea quadrioculata the eggs, which are ‘12 mm. in long diameter, are attached dorsally to the surface between the dorsal cirrus and the para- podium—a pair on each segment from the 13th or 14th to about the 30th. A peculiar feature of this form is that there are invariably two, sometimes even three, more or less widely separated stages occurring together. The arrangement of these is various, but usually all the eggs of the same stage are grouped together; sometimes, however, there is more or less mixing. The means of attachment are not very obvious; but from the position occupied it seems highly probable that the secretion of the dorsal pedal glands performs this function. The early stages are not in any noteworthy manner different from the corresponding stages of /. fustifera. The early stages of seementation were observed and the formation of the ectodermal cap. A stage was obtained in which the ectoderm as a thin, irregular layer encloses an endoderm of about a dozen large clearly-defined cells. The chief peculiarities which appear in later stages are associated with the fact that the embryo, instead of growing straight out, the egg-membrane either disappearing or becoming converted into the cuticle, is developed flexed dorsally within the egg-membrane (PI. 18. fig. 68), from which it only escapes when the tentacles and the parapodia of the primary segments have appeared. As in Lwogone, the development of the embryo proceeds from two separate centres—the dorsal plate and the ventral stomodeeal rudiment. The former early becomes divided into two by a deep constriction, the anterior part forming the head plate and the posterior the body plate. From the latter the germinal bands (PI. 18. fig. 67) grow more directly downwards (ventral), in accordance with the destined flexure. The head plate extends 240 PROF. W. A. HASWELL : earlier over the anterior end of the egg than in Hwogone. The stomodeal rudiment appears as a plug of cells which becomes extended transversely and is connected with the dorsal plate by a pair of narrow cords. The plug becomes invaginated by a narrow transverse slit which later becomes a deep rounded pit. In G. kerguelensis the eggs, which are only half the size of those of G. quadrioculata (‘06 in long diameter), are attached dorsally, typically four on each segment from the 7th to the 17th. The course of the development is essentially the same as in G. qguadrioculata. The peculiar mode of development of Grubea quadrioculata and-G. ker- guelensis, flexed within the egg, appears to be shared by at least one other member of the genus, viz. G. clavata. In comparing the development of that species with that of G. pusilla, De St. Joseph remarks (p. 80) :—‘* Chez la premiere, les embryons se développent dans loeuf sur le dos de la mere, dont ils se séparent dés quwils sont murs en gortant de Voeuf; chez la deuxieme ils se développent sous le ventre de la mére et y restent fixés apres qwils sont sortis de Poeuf” (p. 80). In G. pusilloides there are few, not more than five, eggs attached loosely to the ventral surface, one on each of the segments from the 11th to the 16th. The mode of attachment has not been observed. The embryo is straight ; the stages of development observed resemble the corresponding stages of Evogone fustifera rather than that of the other species of Grubea. In Sphwrosyllis hystria the eggs ave attached to the ventral surface of the parent by means of the secretion of the ventral pedal glands (Pl. 17. fig. 30). Sometimes this secretion, or that of the neighbouring integumentary glands, is poured out so abundantly that it forms an investment for each ovum, and this may have embedded in it numerous minute gritty particles which render it very opaque; frequently this investment is not present, and the egy appears quite clear, being enclosed only in the thin vitelline membrane. Regarding the mode of attachment of the eggs in this form, Pierantoni states (p. 243): ‘* Lo strato di sostanza mucosa que involge luovo é sotti- ‘issimo, salvo nel punto di attacco ove ¢ piti spesso e di consistenza quasi gelatinosa, e prende l’aspetto di sottili cordoncini trasparenti che corrono dall’ uovo alla parete esterna del corpo dell’ animale, intorno al punto di immediato contatto.” In addition to the delicate filaments referred to in the above quotation, certain other structures become visible in the supporting plug or stalk, when sections of it are stained with eosin (Pl. 18. fig. 47). These are a series of rod-like bodies arranged quite regularly parallel with one another and at right angles to the surface of the worm. These must be formed from the eosinophilous granules which, as already pointed out in the part of this paper dealing with the integumentary glands, forms a part of the secretion of the ventral pedal glands in this species. THE EXOGONEA. 241 My preparations quite definitely contradict Pierantoni’s account of the early stages in one point. One set consists of a 4-celled stage in which the cells are quite equal and symmetrical with the nuclei approximated towards ‘the animal pole evidently in preparation for the formation of the first four micromeres, as in the other Hwoyonew and in the Polycbeta in general. Later stages observed also correspond quite closely with those of Exogone. SUMMARY. The following are the most important of the results embodied in this paper :— 1. Descriptions and determinations of Australian species of Lxogonew. 2. Demonstration of the role of the pedal glands in producing the secretion by means of which the ova are attached after extrusion. 3. Description of the proventriculus with its non-striated muscle-columns, and of hitherto undescribed glands—the proventricular glands—the ducts of which open into it. 4. The following out of the changes undergone in Ewogone by the nephridia in both sexes associated with the development of the sexual elements. ~ 5. The hermaphrodite condition in Grubea pusilloides. 6. Description of stages of the development not previously investigated. BIBLIOGRAPHY. 1, AucrEnreR, H.—Polycheta I, Errantia. Die Fauna Siidwest-Australiens, Band iy. Lief. 5. 1913. 2. CLAPAREDF, R.—Annélides chétopodes du Golfe de Naples. Mém. Soc. Phys. Hist. nat. Geneve, tome xix. 1868, 3. CLaparipE, R.—Beobachtungen iiber Anat. und Entwickelungsgeschichte wirbelloser Thiere, ete. 1863. 4, Craparépr, R.—Glanures Zootomiques parmiles Annélides de Port-Vendres. Mém. Soc. Phys. Hist. nat. Genéve, tome xvii. 1864. 5, Dusarpry, F.—Note sur un Annélide (Lvogone pusilla) qui porte a la fois ses ceufs et ses spermatozoides. Ann. Sci. Nat, 3 série, tome xv. 1851. 6. Du Piessis, G.—Un cas de protandrie chez les Syllidiens. Rey. Suisse Zool. vol. xvi. pp. 321-328; pl. 16. 1908.* 7, Enters, E.—Die Borstenwiirmer. 1864. 8. Eu ters, E.—Polycheta der Hamburg. Magalhaens. Sammelreise. 1897. 9. Exters, E.—Die Polycheten. Sammlung der deutschen Siidpolar Exped. 1901-1905. 1915. ‘s Only known to me through the account given by Potts (29). 242 PROF. W. A. HASWELL : 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22, 23. 24, 25. 26. 27, 28. 29. 30. Eutrrs, h.—‘ Siboga’ Polycheta. Eutrrs, .—Die bodensiiss. Anneliden d. Valdivia-Iixped. xvi. 1908. Faces, L.—Modifications et role des organes segmentaires chez les formes epitoques dannélides polychétes. C. R. t. exli. pp. 6164, 1905. Favuve., P.—Annélides polychétes des Iles Falkland. Arch. Zool. exp. et gén., tome ly. 1916, Goopricu, E, $.—On the Nephridia of the Polycheta. Part 38. Quart. Journ. Micros. Sci. (n.s.), vol. xliti. 1900. Gravier, Cu.-—Les Annélides polychétes. Exp. antarct. francg. 1907. Haswevi, W. A.—On the structure of the so-called Glandular Ventricle of Syilis. Quart. Journ. Micros. Sci. (n.s.), vol. xxvi. 1886. Haswert, W. A.—A comparative study of Striated Muscle. Quart. Journ. Micros. Sci. (n.s.), vol. xxx. 1889. Lancgrruans, P.—Ueber einige Canarische Anneliden. Nova Acta Acad. Cvs. Leopold.-Carol. Germ. Nat. Cur., Halle, Bd. xlii. 188]. LANGERHANS, P.—Die Wiirmfauna v. Madeira. Zeitsch. wiss. Zool., Bd. xxxii. 1879. McIntosu, W. C.—The Polycheta. Rep. ‘ Challenger,’ Zool. vol. xii. 1885. McInrosu, W.C.—A Monoevaph of the British Annelids. Vol. I. Part i. Ray Soe. 1908. Maxaquin, A.—Recherches sur les Syllidiens. Mém. Soc. Sci. Lille. 1898. Marenzetier, I. von.—Zur Kenntniss der Adriatische Anneliden. Sitzb. der Ix. Akad. der Wissensch. Wien, Bd. lxis. 1874. Marion ET Bopretzky.—Ktudes des annélides du Golfe de Marseille. Ann. Sci. Nat. 6™ série, tome ii. 1875. OrrstEepD, M.—Ueber die Entwick. der Jungen bei einer Annelide, ete. Arch. f. Naturg. 1845. Pacenstrecuer, H. A.—Untersuchungen tiber niedere Seethiere aus Cette. 1, Lxvegone gemnufera, ete. Zeitschy. wiss. Zool. Bd. xii. 1863. Preranront, U.—Una nova maniera di gestazione esterna della Pionosyllis pulligera Krohn. Monitore zool. Ital. x.; Ann. Mus. zool. Ital. (n.s.) vol. ii. 1905. Prerantoni, U.—La gestazione esterna. Arch. Zool. Napoli, vol. i. pp. 2381-252, tab. 10-11. 1903. : Ports, F. A.—Methods of Reproduction in the Syllids. Ergebnisse u. Fortschritte der Zoologie, Bd. iii. 1911. Prevor, G.—Sur les modifications et le réle des organes segmentaires des Syllidiens a Yepoque de la réproduction. C. R. t. exxxiv.. pp. 242-245. 1902. . St. JosepH, Baron px,—Les Annélides Polychétes des cotes de Dinard. Ann. Sci. Nat. 7° sér. tome i. 1887. . SCHNEIDER, J.—Zur Entwickelung der Pionosyllis pulligera Langerhans. Zool. Anz. Ba. xliv. No. 13. 1914. . SourHERN, R.—Achiannelida and Polychieta, Clare Island Survey. Proc. R. Irish Acad. xxxi, 1914, . Viegurpr, C.—Sur l’ Exoyone yemmifera (Pagenstecher) et quelques autres Syllidiens & ; d 8 gestation. Etudes sur les animaux infériewrs de la baie d’Alger. Arch. Zool. expér, et gén., 2 sér. t. i. pp. 69-110, planches 3-5. 1884. Haswell. Caleb aebel ith, JNUOP SPIRO AS Journ. Linn. Soc. Zoon. Vor. XXX TV. Pr.1'7. C Hodges & Son.imp | EXOGONEAE: Haswell. Journ. Linn. Soc. Zoor. Vor. XXX TV. Px.17. C.L.Birbel. lith. C Hodges & Son.imp AUSTRALIAN xs © COM EAE a ee Rani e f Bans Haswell G.L. Birbel. lith. AUSTRALIAN Journ Linn Soc Zoo Vor XXXIV. lenatsy pon Ss & ye aT He ade = c Journ. Linn. Soc Zoon. Vor. XXXIV. Pu. 18. Haswell _ GL. Birbel. lith. C. Hedges & Son. imp AUSTRALIAN EXO GONEAE : a Saneoite - Fre THE EXOGONEA. 243 EXPLANATION OF THE PLATES. LETTERING. ae. aciculum, | ne. nephridium. ane. anal cirrus. np. nephridiopore, apg. anterior pro-ventricular glands. ov. ovum. b. rhabdite glands. p: pharynx. @. ceca, pg. plaryngeal glands. ef. ciliated funnel. | pp. pharyngeal papille. eg. ciliated groove. pr. proctodeeum. pi.m. protractor muscles of proboscis. pv. proventriculus. rm.ec. yadial muscle-columns, em. circular muscles. cu. cuticle. dc. dorsal cirri. d.py. dorsal pedal gland. 80. somites. d.pl. dorsal plate of embryo. | ss. setigerous sac. ep. epithelium. | st. stomodeeum. if funnel, ti pharyngeal tooth. g-b. — germinal bands. vc. ventral cirrus. a. intestine. vp.g. ventral pedal gland. z.gl. integumentary glands. vv. ventral vessel. ne, ventral nerve-cord. y. yellow gland. All the figures from drawings by the author. PLATE 17. 1, Lvogone fustifera. Anterior region, dorsal view, magnified, with outline of alimentary canal, ete. LE. fustifera. Dorsal view of parapodium, x 1040. E. fustifera. Samples of compound sete. E. fustifera. Most dorsal seta, ordinary form. E. fustifera. Most dorsal seta, form with appendage. E. fustifera. Most ventral seta, strongly bidentate form, x 1040. E. verrugera. Anterior region, LT. verrugera. Second compound seta, x 1040. LE. verrugera. Most ventral, simple seta. x 1040. Figs.10a & 100. £. verrugera, Hnd of aciculum. x 1040. Fig. 11. £. heterosclosa. Headend. x 440. Fig. 12. £. heterosetosa. Most dorsal. simple seta. > 1040. Figs. 13 & 14. Z. heterosetosa. First (most dorsal) compound seta, two varieties. Figs. 15, 16,17. £. heterosctosa. Ordinary compound seta, Fig. 18. Grubea kerguelensis. Anterior end. x 440. 19. G. kerguelensis. Most dorsal, simple seta. x 1040. 20. G. kerguelensis. Compound seta. 21. G. quadrioculata. Anterior region. x 440. Vies. 22 & 23. G. quadrioculata. End of aciculum. Figs. 24 & 25. G. quadrioculata. Compound sete. Vig. 26. G@.quadrioculata. Pavapodium and cirri, ventral view. x 440. All the compound sete are not represented. ODD vP ge 10 27. G. pusilloides. Dorsal view of anterior region, x 300. Only one of the two rhabdite glands in each parapodium is represented. 28. G. pusilloides. Parapodium from the dorsal side. x 1040. 29. Section of one of the groups of granule-producing glands of LZ. fustifera. x 800, Fig. 40. PROF, W. A. HASWELL : Spherosyllis hystrix, 2. Outlines of transverse section showing the position of the ventral pedal glands. x 300. Evogone fustifera. Outlines of proventriculus from stained and cleared specimen, showing the anterior proventricular glands and their ducts. x 440. PLATE 18, Spherosyllis hystrix. Anterior region from the dorsal side. x 240. Setze only represented on one side. S. hystrix. Parapodium and cirri. x 440. S. hystrix. First (most dorsal) compound seta. x 1040. S. hystrix, Anal cirri. Exogone verrugera. Ventral view of a portion of a stained and cleared specimen of a female with early ova, to show the ventral pedal gland with its mass of secretion. £. fustifera. Horizontal section of proventriculus, showing ducts of anterior proventricular glands. , E. fustifera, Sagittal section in the region of the junction of the pharynx and proventriculus, showing anterior proventricular glands and ducts. E. fustifera. Muscle-columns as seen in horizontal sections of the lateral wall of the proventriculus. E. fustifera. Transverse section through outer ends of muscle-columns as seen in horizontal series. Figs. 41 & 42. ZH. fustifera. Corresponding parts of two successive longitudinal (hori- Fig. 43. 44, ig. 48. 49, zontal) sections of a male specimen showing a nephridium at the point where it opens on the exterior, with the bifurcation of the main duct to form two loops. x 440. Y E. fustifera. General view of nephridium plus testis at an earlier stage than that represented in Figs. 41 and 42. From a stained and cleared specimen. x 1040. L. fustifera. Segment of a male specimen in which the nephridia plus testes are far advanced in development, but have not yet begun to unite mesially. As in the following figure, the sete are merely indicated. E. fustifera. Semi-diagrammatic view of a segment of a ripe male with the median vesicle surrounded by a mass of sperms. Spherosyllis hystrix. Ventral view of left portion of a segment with young ovum enclosed in its follicle (nephridium) and the ventral pedal gland with the mass of secretion by means of which the ege becomes attached after it has been discharged. S. hystrix. From a section passing through the point of attachment of an embryo (ov.) to the ventral surface with the plug of secretion by means of which the attachment is maintained. x 440. All the figures refer to Ewogone fustifera. Ventral view of two segments of female with early ova. From stained and cleared specimen. x 440. View of a segment of a female in which the two ovisacs (nephridia) have become fused together across the middle line. Shows also the positions of the ventral pedal glands and the zone of integumentary glands. x 440. Segment of a female with the two ova fully grown lying side by side in the common follicle or nephridial sac. x 440. Segment of a female with the two ova lying one in front of the other. From stained and cleared specimen. x 440. THE EXOGONES. 245 Fig. 52. Stage in the epiboly in which the four original cells (macromeres) still maintain their individuality : the nuclei of these are marked with small crosses. x 440. 58. Lateral view of embryo with dorsal plate and stomodal rudiment, before the appearance of the germinal bands. 54, Lateral view of an embryo at a somewhat later stage than that represented in fig. 58, with the pair of narrow bands running from the stomodieal rudiment to the apical plate. 55. Ventral view of an embryo of about the same stage as that represented in fig. 54, but with the stomodzal rudiment more extended laterally and the germinal bands beginning to appear. 56. Lateral view of a similar stage to that represented in fig. 55, 57, Lateral view of a somewhat more advanced stage with the dorsal plate extending forwards nearly to the free end. All the figures, except 67 and 68, refer to Lvogone fustifera, . 58. Ventral view of an embryo at the stage in which the germinal bands meet posteriorly : the stomodeeal rudiment, transversely elongated and connected with the dorsal shield by a pair of narrow strands, is not yet definitely invaginated. x 440. 59. ‘Transverse section of embryo at about the stage represented in fig. 58. x 440. so. here points to the germinal band. 60. Ventral view of an embryo with the stomodeal invagination completed, the germinal bands well established, meeting behind to surround a proctodieal area. x 440. 61. Lateral view of the embryo represented in fig. 60. x 440. 62. Ventral view of embryo in which the germinal bands have become segmented and the stomodzum, destined to become the pharynx, has given off an offset which will develop into the proventriculus, x 440. 63. Horizontal section of stage similar to that represented in fig. 62. x 1000, 64. Stage of the first appearance of the tentacles. x 440. 65. Stage with well developed prostomial tentacles and without rudiments of the parapodia. x 440. 66. Horizontal section of advanced embryo with cae segmunts and parapodia with sete. x 440. 67. Grubea kerguelensis, Stained and cleared early embryo, viewed somewhat obliquely from the ventral side. The dorsal plate deeply constricted into head plate and body plate, and the latter giving off ventrally the germinal bands. x 440. 68. Grubea quadrioculata. Late embryo stained and cleared: outline, to show the ventral curvature. x 300. LINN. JOURN.—-ZOOLOGY, VOL. XXXIV. 19 ON THE ABUNDANCE OF SOME COMMON MARINE ANIMALS, 247 Sprott RuntrAva.—I1V. Notes on the Abundance of some Common Marine Animals anda preliminary Quantitative Survey of their Occurrence. By W. A. Herpmay, D.Sc., LL.D., For.Sec.R.S., F.L.S., Professor of Oceanography in the University ot Liverpool. (With 8 Text-figures.) {Read 11th December, 1919. ] For many years I have taken any opportunities that offered, both on dredg- ing expeditions in the ‘Runa’ and also when tramping the sea-shores of various parts of the world, of making observations on the abundance and mode of occurrence of the common marine animals and plants. Two years ago I brought before the Society some statistics and conclusions as to the occurrence ad distribution of the commoner planktonic organisms, Diatoms and Copepoda, throughout the year in the Irish Sea *, with the object of showing that the supply of such food-matters to our plankton-eating fishes depends upon a surprisingly small number of species which are present in enormous abundance. Some half-dozen kinds of Diatoms and about the same number of Copepoda are the all-important organisms upon which our fate depends so far as concerns our food from the sea. I desire now to place on record some notes as to the occurrence of a few of the commoner fixed or sedentary bottom-living animals of the sea-shore or shallow-water—animals which are also, like the plankton, of great practical importance in nature as the chief food of some of our most valuable fishes. Our knowledge of the number of animals living in different regions of the sea is for the most part relative only. We know that one haul of the dredge is larger than another, or that one locality seems richer than another, but we have very little information as to the actual numbers of any kind of animal per square foot or per acre in the seat. Some years ago Hensen attempted to estimate the number of food-fishes in the North Sea from the number of their eggs caught in a comparatively small series of hauls of the tow-net, but the data were probably quite insufficient and the conclusions may be erroneous. It is an interesting speculation to which we cannot attach any economic importance. All biologists must agree that to determine even approximately the number of individuals of any pacicalay species living in a known area is a contri- bution to knowledge which may be of great economic value in the case of the edible fishes ; but it may be doubted whether Hensen’s methods, even * “Spolia Runiana—II1.,” Journ. Linn. Soc., Zool. xxxiy. p. 95 (1918). + Professor McIntosh, in his book ‘The Resources of the Sea’ (1899), gives a vivid picture of the abundance of life of all kinds in the sea, but does not venture upon any numerical estimates, 19* 248 PROF. W. A. HERDMAN ON THE ABUNDANCE with greatly increased data, will ever give us the required information. Petersen’s method, of setting free marked Bisice and then assuming that the proportion of these soucarnyaht | is to the total number marked as the fishermen! s catch in the same district is to the total population, will only hold good in circumscribed areas where there is practically no migration and where the fish are fairly evenly distributed. It is difficult to imagine any method which will enable us to estimate any such case as, say, the number of plaice in the North Sea*, where the individuals are so far beyond our direct observation and are liable to change their positions at any moment. But a beginning can be made on more accessible ground with more sedentary animals, and that is what I have been attempting to do. Dr. C. G. Joh. Petersen, of the Danish Biological Station, has for some years been pursuing the subject in a series of interesting Reports on the “ Evaluation of the Sea” ft. He uses a bottom- sampler or grab, which can be lowered down open and then closed on the bottom so as to bring up asample square foot or square metre (or in deep water one-tenth of a square metre) of the sand or mud and its inhabitants. With this apparatus, modified in size and weight for different depths and bottoms, Petersen and his fellow-workers have made a very thorough examination of the Danish waters, and especially of the Kattegat and the Limfjord, have described a series of “animal communities ” characteristic of different zones and regions of shallow water, and have arrived at certain numerical results as to the quantity of animals in the Kattegat expressed in tons—such as 5000 tons of plaice requiring as food 50,000 tons of “useful animals” (Mollusca and Polychet worms), and 25,000 tons of starfish using up 200,000 tons of useful animals which might otherwise serve as food for fishes; and the dependence of all these animals directly or indirectly upon the great beds of Zostera, which make up 24,000,000 tons in the Kattegat. Such estimates are obviously of great biological interest, and, even if only rough approxi- mations, are a valuable contribution to our understanding of the metabolism of the sea and of the possibility of increasing the yield of local fisheries. But on studying these Danish results in the light of what we know of our own marine fauna, although none of our seas have been examined in the same detail by the bottom-sampler method, it seems probable that the animal communities as defined by Petersen are not applicable on our coasts, and that the estimates of relative and absolute abundance may be very different * Heincke, however, has attempted to estimate the adult plaice at 1500 millions, of which 500 millions are caught annually. + See ‘ Reports of the Danish Biological Station,’ and especially the Report for 1918, “The Sea Bottom and its production of Fish Food.” Professors L. Joubin and Guérin-Ganivet have published a series of papers, entitled “ Gisements de Mollusques comestibles des Cotes de France” (Bull. Mus, Ocean. Monaco, from 1906 onwards), surveying the shell-fish beds . and fisheries, but giving no estimates of numbers present. OF SOME COMMON MARINE ANIMALS. 249 in different seas under different conditions. The work will have to be done in each great area, such as the North Sea, the English Channel, and the Irish Sea, independently. This is a necessary investigation, both biological and physical, which lies before the oceanographers of the future, upon the results of which a rational conservation and exploitation of our national sea- fisheries may depend. My own contributions to the subject so far deal only with the shore and shallow-water animals of the littoral and the top of the Laminarian zone, and I shali give here merely a few examples from different groups of animals and plants, and from different localities, mainly to demonstrate the enormous abundance of some of the commonest, and therefore the most important, of the animals and plants on what are sometimes called our “barren” sea- shores. I shall first take the cases of a worm, a crustacean, a mollusc, and an ascidian all forms that are free-swimming when young but fixed in the adult condition, and all of value as food of marketable fishes. SABELLARLA. The gregarious Polychat Annelid Sabellaria alveolata is present in great abundance on many parts of the coast of North-West Europe, generally on Vie. 1.—Sabellaria alveolata, from Tilbre Id., nar. size. shores where stones or patches of rock crop out in proximity to sand and where strong tidal currents disturb the sand and carry the suspended grains along in quantity. The worms stick the sand-grains together to form the tubes in which they live and which adhere to one another so as to build up solid masses of a porous, crisp and brittle material, which crumbles to a certain extent when walked upon, but which is constantly being renewed and has its injuries repaired by the living worms within, and must, therefore, have a very considerable effect in some places in protecting the rocks or shore 250 PROF. W. A. HERDMAN ON THE ABUNDANCE generally from the erosive action of the sea. On Hilbre Island, in the Hstuary of the Dee, for example, Sabellaria covers the soft red triassic sandstones with encrusting sheets, rising in places to form massive hummocks and outstanding reefs many yards in extent; and the same reef-like formations are to be seen on a still larger scale on the coast of Normandy, near Granville, Mount St. Michel Bay, and elsewhere. I have measured and counted numbers of samples of these Sabellaria tubes, and find that the diameter of the mouth of the tube is generally between an eighth and a tenth of an inch. Of course, there are also smaller ones, those of the young worms in the mass, and a few larger, but a fair average size is one-tenth of an inch. A photograph (fig. 1) of such a mass of Sabellaria tubes shows that in a surface of about three square inches there are from 65 to 75 mouths of tubes. Taking 65, a square foot would have 3120 and a square yard some 28,080 of the worms. Hence, forty square yards would contain over a million, Now, there are very many square yards of Sabellaria on the shore at such a locality as Hilbre Island, containing, therefore, millions of worms, each from one to two inches in length, and such Polycheet worms are a favourite food of flat-fish like the plaice and sole, and can be easily obtained by crunching up the brittle sand-tubes. Roughly, about half of a mass of Sabellaria tubes consists of sand and the rest is formed of the nutritious worms. Sabellaria is not found on the shore only, but extends beyond low-water mark for some distance. I have dredged lumps of it from depths of about 10 fathoms, off the Lancashire coast, near Fleetwood, and in some places it it so abundant at the bottom, and comes up so frequently in the fishermen’s nets, that it is familiar to them, and is known by the local name of ‘“ knarrs.” That flat-fish are present on this trawling ground is probably due in part to the abundance of the Sabellaria masses on the bottom. BALANUS. The common rock-barnacle Balanus balanoides is probably the most abundant fixed animal on the rocky shores of North-West Europe. On many sea-cliffs the barnacles form a layer covering practically every inch of rock between the tide-marks. Bradda Head at Port Erin is a good example, and at low tide, when seen from a distance, it looks as if a horizontal band of whitewash had been painted along the base of the cliff. When examined in detail, it is found that the barnacles—young and old—form a continuous layer, roughening the whole surface of the rock and leaving scarcely any spots of stone exposed. In fact, the individual barnacles in many cases have not room to grow to the normal size. They are overcrowded, and by mutual pressure are caused to assume vertically elongated prismatic forms, like columnar basalt. A quarter of an inch square is an average size for an adult OF SOME COMMON MARINE ANIMALS 251 Balanus, and there are 2304 square quarter inches in a square foot. But on one square foot of-rock at Fleshwick Bay, near Port Erin, 2940 barnacles were counted (fig. 2). Fig. 3; Part of a mussel bed at Morecambe Bay. Balanus must be of value as a food-matter, not only directly if crunched up or sucked out of its calcareous covering, but also indirectly through: its free-swimming larvee. In early spring it sets these free in the Nauplius stave in such abundance that the water close to the barnacle zone, or in any shore-pools of the neighbourhood, looks muddy, and when dipped out in a glass jar, is seen to be crowded with dense swarms of the minute larvie, 252 PROF. W. A. HERDMAN ON THE ABUNDANCE They form a conspicuous feature of the plankton for a few weeks in March and April, and constitute, no doubt, an important food-matter at this time when there are comparatively few Copepoda available. Myriuvs. Probably the edible mussel (A/ytilus edulis) is the most abundant molluse in the seas of North-West Europe, and it is certainly the most generally useful to man, both directly as a food and as a fisherman’s bait and also in- directly as the food of marketable fishes. Although mussel ‘“ beds,” “ scars,” or “scalps,” as they are variously called, are regularly worked over by the Tie. 4..—Mussel “ spat” on sea-weeds, magnified. fishermen, they cannot be regarded as in any sense artificial. The mussels have settled down and grown upon them naturally. It is only very rarely that any transplanting takes place, and there is nothing in this country comparable with the artificial “bouchét” system of mussel culture which takes place in the bay of Aiguillon on the west coast-of France. In a mussel bed (fig. 3) not only are the mussels so closely placed as to be touching their neighbours all round, so as in some cases to cause distortion and prevent proper growth, but they are even piled on the top of one another, it may be several layers deep, the interstices below being filled up with mud and the byssus fibres of the molluses in the layers above. At the time when the mussel “spat” is settling down from the free- swimming laryal condition, sea-weeds, zoophytes, rocks, and other objects OF SOME COMMON MARINE ANIMALS. 253 between tide-marks may become covered with a layer7of the little black dots as if they had been thickly peppered over (fig.*«4); and in favourable localities the young mussels grow in a profusion which may cause their own destruction. Ihave seen the rocks at Hilbre Island covered with a layer of small mussels from a quarter to half an inch in length, which are so closely placed as to be absolutely continuous, and to be tightly united to one another by matted byssus fibres, while they have little or no attachment to the rock beneath, the result being that in the first storm the waves roll up the sheets of mussels and wash them ashore on the neighbouring sandy beaches in masses as large as pillows and bolsters. In such sheets and masses I have counted from 80 to 100 mussels visible on the surface of a square inch, and in the case of slightly older ones from 50 to 60 (fig. 5). At the rate of 100 there would be about 129,600 in a Fie. 5.— Young mussels on Laminaria stem, nat. size. square yard, and there are very many such square yards around our coast. No doubt the majority of these young mussels never grow to maturity. They are killed by storms, smothered by their neighbours, or eaten by star- fishes or by plaice and other fishes. In the latter case they are not lost as a food matter, and even in the former their remains will be eaten by some- thing which will indirectly feed man. Nothing is lost in the sea, and everything ultimately in the metabolic cycle contributes to man’s harvest. Dr. James Johnstone has estimated” that a Lancashire mussel bed may have 16,000 mussels to every square foot, and that the proteid contents of the mussel flesh is comparable with that of a lean farm animal, while per unit of area the mussel bed produces nearly a hundred times the amount of flesh for food that is produced by cultivated land. The sea when cultivated, * Conditions of Life in the Sea,’ Cambridge University Press, 1908, 254 PROF. W. A. HERDMAN ON THE ABUNDANCE or when populated under favourable conditions as on a mussel scalp, seems to be more productive than the land *. STYELOPSIS. One of the most densely-covered rock surfaces that I have seen is in the large caves near Spanish Head at the south end of the Isle of Man. These caves have to be entered by boat at low tide, and are of considerable extent, and their walls between tide-marks and as far down as can be seen in the clear deep water are covered with an assemblage of animals, consisting of: the Tetractinellid sponges Pachymatisma johnston’ and Stelletta collingsi, several species of small white calcareous sponges, Halichondria panicea, Tubularia indivisa, Sagurtia miniata, and enormous quantities of the small red Ascidian Styelopsis grossularia. The Ascidians cover by far the greater part of the rock, and are so closely placed that their bases are continuous, and masses can be peeled off in small sheets (fig. 6). On the average of a Fig. 6.—Two aggregations of Styelopsis grossularia, nat. size. number of measurements they are three-quarters of an inch in height, and there are from 10 to 30 in the square inch according to size, over 50,000 to the square yard. Although Styelopsis is not generally supposed to be a common animal, the numbers present in these caves, and no doubt in many other similar places on rocky shores, must be very large, and must afford a very considerable amount of highly nutritious food. In many parts of the world Ascidians closely related to our Styelopsis (Cynthiade) are used as food by man. OrgER ComMoNn SHORE ANIMALS. Tt may be of interest to add here the results of countings of a few other common shore animals, although I shall not discuss them further. Along * As Spenser seems to haye known, or imagined, more than four hundred years ago :— “O what an endlesse worke have I in hand, To covnt the seas abundant progeny, Whose fruitful seede farre passeth those in land,” &e.—- Faerie Queen, Bk. TV. Canto 12. Aa OF SOME COMMON MARINE ANIMALS. DO with Bulanus on the rocks (fig. 7), the most abundant shell-fish are Patella vulgata, Purpura lapillus, and several species of Littorina. Ten counts of Patella gave an average of 29 to the square foot, the highest record being 37. Two counts on square feet where Purpura seemed numerous on the limestone rocks at Port St. Mary gave 72 and 101 respectively. The method adopted in all these cases was to plant a wooden frame measuring one square foot internally down on the rock and count or remove all that lay inside it. In half-a-dozen small shallow rock-pools on the limestone at Port St. Mary the common red anemones (Actinia equina) gave respectively 30, 47, 60, 49, 55, 37 to the square foot, taking the pools in order from the highest down- wards towards low water. The average is over 46, and the range (30 to 60) Tia. 7—Rock surface at Port Erin covered with Balanus, Patella, Ke. is not great—sugvesting that such little pools having «a surface of about two square feet and a depth of a. few inches, and generally lined by Nullipores and Corallina, can only support some 40 or 50 anemones each. On the sand at Port Erin the most conspicuous marks are the the burrowing worm 4renicola marina (the fisherman’s ** Lug-worm”), and from our ccunts we find 11,800 worms in a rectangle measuring 100 yards “easts” of by 30 yards at low tide—that is about four to the square yard ; while a little further up the shore, in a square of 10 yards each way, we found 653, or 64 worms to the square yard—and this, moreover, on an area where both the fishermen and the visitors were constantly digging the worms for bait. Very much larger numbers than these—up to 30 or 40 to the square yard— have been counted on some beaches, and Ashworth has estimated that at Musselburgh, in a zone at low tide about a mile in length, there are from 3 to + million worms*. * 1. M. B.C. Memoir xi, Avenicola; 1904. bo Or PROF. W. A. HERDMAN ON THE ABUNDANCE Some ALG. On rocky or stony shores the most conspicuous organisms after the barnacies are the large brown sea-weeds—the various species of Fucus and its allies in the littoral zone and those of Laminaria at and below low-water mark. At Port Erin and Port St. Mary we have made many counts *, and some weighings, of the plants of Fucus and of Laminaria cut off from sample square feet in different regions, and also of the epifauna associated with the Algee. Fucus is most variable, ranging from 2 to 352 stalks to the Fig. 8.—Diagram showing distribution of Zaminaria stalks in one square foot. The black - square in centre shows the area they would occupy if placed together. square foot. Laminaria is more constant; our counts range from 6 to 3€ ; but in a typical area of Laminaria-covered rock, such as on the lower limestone reefs at Port St. Mary, the latter figure or anything about 30 stalks to the square foot is usual. As the stalks are on the average half an inch in diameter, this distribution means that they are scattered from 1 to 2 inches apart over the square foot (fig. 8). The fronds of the Laminaria springing from the one stalk may be wide- spreading, and anything from 3 to 6 feet in height. Moreover, they may support an epifauna and epiflora of many molluscs, worms, polyzoa, hydroids, and smaller Algze. We have made a number of measurements + I am indebted to one of my students at Port Erin, Miss Catherine Mayne, B.8c., for much help with these counts and for a series of diagrams like figure 8 illustrating the density of distribution. “a OF SOME COMMON MARINE ANIMALS. bo Or and weighings at Port Erin, but the results vary greatly with the amount of moisture retained by the Algzwe. The general conclusion, however, is that a very large amount of organic food must be present in the region of the coarse brown sea-weeds, and especially in the Laminarian zone, and therefore it is not surprising that shoals of young fishes are found feeding there. ZOSTERA, The ‘“ Grass-wrack” (Zostera marina) is a flowering plant allied to the sedges, which grows in muddy sand at and below low-water mark. It is enormously abundant in certain places, and in the South of Australia I have seen banks of it several feet high extending for miles along the sandy beach. Petersen says that about 2000 square miles are covered by Zostera in Danish seas, that in the Kattegat alone it amounts to 24 million tons, and that it produces annually four times as much dry vegetable matter as all the hay on the land of Denmark. It is evidently the fundamental plant- food for fishes and other animals in the Danish seas, and no doubt it serves the same important purpose on a smaller scale in our seas. There is a small Zostera bed in Port Hrin Bay, and three patches at the mouth of Port St. Mary Harbour which are just accessible by wading at the lowest spring-tides, and there are much larger beds elsewhere further north on the shores of the Isle of Man. A Zostera bed always supports a large fauna or epifauna associated with its roots, rhizomes, and leaves. Characteristic animals are the anemone Anthea cereus (= Anemonia sulcata), various species of Trochus, Littorina and Lacuna, some nudibranchs, many worms, zoophytes, polyzoa, and compound ascidians, especially the transparent Diplosoma. In addition, there are delicate filamentous Aloe (Confervee) and enormous masses of Diatoms attached to the older withered or decaying leaves. Even after the Zostera leaves have decayed and gone to pieces, they contribute to the important organic detritus on the bottom, upon which many creeping animals on the surface of the sand are nourished. So that both directly and indirectly, through the organisms it attracts and supports, the Zostera bed is an important source of food to fishes and to many invertebrates. The long blades or leaves, which may extend from 3 to 5 or 6 feet above the sand, arise in shoots from a creeping rhizome, and in my counts I find from 4 to 6 leaves in a shoot and about 20 shoots in an area of 3 inches by 1 inch—that is, about 1000 shoots in the square foot. The waving forest so produced, clothed in turn with many other organisms large and small, is one of the densest masses of living food-matter that I know of in the sea. The value of the Zostera bed in nature, both directly from the food that it furnishes to the animals living on it, and also indirectly from the enormous quantities of Diatoms which cover its decaying leaves, is a part of the bo t PROF. W. A. HERDMAN ON THE ABUNDANCE metabolism or domestic economy of the sea—the computing of income and expenditure so as to find out whether we have a balance to the good, and whether the sea could be made by further cultivation to produce still more food for the service of Man. CONCLUSIONS. From these and similar data which can be readily obtained, it is not difficult to calculate totals by estimating the number of square yards in areas of similar character between tide-marks or in shallow water. And from weighings of samples some approximation to the number of tons of available food may be computed. But one must not go too far. Let all the figures be based upon actual observation. Imagination is necessary in science, but in calculating a population of even a very limited area it is best to believe only what one can see. Countings and weighings, however, do not give us all the information we need. It is something to know even approximately the number of millions of animals on a mile of shore and the number of millions of tons of possible food in a sea-area, but that is not sufficient. All food-fishes are not equally nourishing to man, and all plankton and bottom Invertebrata are not equally nourishing toa fish. At this point the biologist requires the assistance of the physiologist and the biochemist. We want to know next the value of our food matters in proteids, carbohydrates, and fats, and the resulting calories. Dr. Johnstone, of the Oceanography Department of the University of Liverpool, has already shown us how markedly a fat summer herring differs in essential constitution from the ordinary white fish, such as the cod which is almost destitute of fat, as follows:— | Herring. | Cod. Proteids Pein tees PR Mcartens 21 167 SH AtSie is fa. cjates net evotehs ets Oslo Ten 22 | 0:3 Arsh) (2 Salt), autem onniesie cere Baten 9 13 | Water (+ traces) .......+..6- 48 | 81:7 Prof. Brandt, at Kiel, Prof. Benjamin Moore, at Port Erin, and others, have similarly shown that plankton gatherings may vary greatly in their nutrient value according as they are composed mainly of Diatoms, of Dino- flagellates, or of Copepods; and Prof. Moore has kindly analysed for me, OF SOME COMMON MARINE ANIMALS. 259 recently, samples of zoo-plankton (Calanus) and phyto-plankton (Diatoms) which I had collected in the Irish Sea for the purpose and transferred in the living condition into absolute alcohol. His results are as follows a | Paine 5 Carbohydrates Fats x Broteids. | (by difference). (+-pigment). el Phyto-plankton ba 90.7 O44 a5 (Chetoceros)... ae eal oo | cae Zoo-plankton .. | 53] | 18-2 176 | Wl (Calanus).... § And, no doubt, the animals of the benthos—the common invertebrates of our shores—will show similar differences in analysis +. It is obvious that some contain more solid flesh, others inore water, in their tissues, others more calcareous matter in the exoskeleton, and that therefore weight for weight we may be sure that some are more nutritious than the others; and this is probably at least one cause of that preference we see in some of our bottom- feeding fish for certain kinds of food, such as Polychet worms, in which there is relatively little waste, and thin-shelled Lamellibranch Molluses, such as young mussels, which have a highly nutrient body iv a comparatively thin and brittle shell. I am aware that [am presenting a somewhat incomplete investigation, but my object is to direct attention to what seems an obvious and useful exten- sion of faunistic work, for the pnrpose of obtaining some approximation to a quantitative estimate of the more important animals of our shores and shallow water and their relative values as either the immediate or the ultimate food of marketable fishes. * These figures differ somewhat from those given by Brandt, and this may possibly be due to differences between these types of plankton at Kiel and in the Irish Sea—which raises the interesting question whether the same organism may have different nutritive values under different hydrographic conditions. + Moore and others have made analyses of the protein, fat, &c., in the soft parts of Sponge, Ascidian, Aplysia, Fusus, Echinus, and Cancer at Port Erin, and find considerable differences—the protein ranging for example from 8 to Ol per cent and the fat from 2 to 14 per cent (see Bio-Chemical Journ. vi. p. 255). f EIN NEANGS ©. Glen Yeo. £ © Nip @ IN: RULES FOR BORROWING BOOKS FROM THE LIBRARY. 1. No more than Six volumes shall be lent to one person at the same time without the special leave of the Council or one of the Secretaries. 2. All books shall be returned before the expiration of Six weeks from the time of their being taken out, but if not required by any other Fellow, they may, on application, be kept for a further period of Six weeks. 3. All books lent shall be regularly entered by the Librarian in a book appropriated for that purpose. 4. No work forming part of Linneus’s own Library shall be lent out of the Library under any circumstances. Nore.— Certain other works are included in this prohibition, such as costly illustrated works, volumes belonging to sets which could not be replaced if lost, and unbound parts of serials. 5. All carriage charges on books to be defrayed by the Fellow borrowing the same. 6. All borrowed books when returned to be sufficiently wrapped and protected. Borrowers will be held liable for any damage incurred owing to improper packing. 7. Newly published books shall not ordinarily leave the Library until after the expiration of one month from the date of their being received. A GENERAL INDEX to the first twenty Volumes of the Journal (Zoology) may be had on application, either in cloth or in sheets for binding. Price to Fellows, 15s.; to the Public, 20s. NOTICES. Tuer attention of the Fellows, and of Librarians of other Societies, is requested to the fact that the parts of the Journal (Zoology) of the present volume have been issued as follows :— Vou. 34. Nos. 225, 226, and 227, the present number. Authors are entitled to 25 copies of their communications gratuitously, and may obtain another 75 by payment, as shown on the printed slip which accompanies the proof. If more than 100 copies are wanted, application must be made to the Council. Abstracts of the proceedings at each General Meeting and Agenda for the next are supplied to all Fellows. B. DAYDON JACKSON, General Secretary. DrEcEMBER 7, 1920. Price 14s. THE JOURNAL OF THE LINNEAN SOCIETY. Vou. XXXIV. ZOOLOGY. No. 228. CONTENTS. I. The Germ-Cells, Fertilization, and Early Development of Grantia (Sycon) compressa. By J. Bronté Garensy, B.A., B.Sc., D.Phil., Senior Demy, Magdalen College ; Senior Assistant in slope Lecturer in Cytology, Univer sity College, London. (Communicated by Prof. E. S. Goopricu, F.R.S., Sec.L.8.) (Blatese19=23 -eandi A lexta tures) emer eetese ees nasene sesh 261 II. The Fragrance of Calcinean Sponges and the Shoat of Guancha coriacea and Sycon raphanus. By G. P. Brpper, Se.D., F.L.S8. (Plate 24.) 2.0... ccee eeteee eee eee eee ete 299 III. Synerypta spongiarum, nova. By the same (with Text-figures)... 305 IV. Notes on the Physiology of Sponges. By the same (with Text- IMeADTESI)) nS aBeAs On oadooo’ das RadbacmerndsohontonoBadbnunycttasasmunmsepun nossa 315 V. On certain Nuclear Phenomena in the Oocytes of the Gall-tly Neuroterus. By Lancexor T. Hoasen, B.A., B.Se. (formerly Frank Smart Prizeman in Zoology, Cambridge). (Communi- eated by Dr. E. Drazsiey, F.L.S.) (With 2 Text-figures.) 327 VI. Notes upon the Reproduction of Asellus aquaticus. By ERNEst E. Unwin, M.Se. (Leeds). (Communicated by the Rey. T. R. R. Srespine, M.A., F.R.S., F.L.S. (Plates 25, 26.) ............ 335 LONDON: SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON HOUR we UO PICCADILLY, W.1, j ‘ AND BY i JAN 13 9% LONGMANS, GREEN, AND CO., ea AND 8tien ai WILLIAMS AND NORGATE, 1920 LINNEAN SOCIETY OF LONDON. LIST OF THE OFFICERS AND COUNCIL. Elected 27th May, 1920. PRESIDENT. Dr. Arthur Smith Woodward, F.R.S. VICE-PRESIDENTS. E. T. Browne, M.A. Horace W. Monckton, F.G.S. Prof. J. B. Farmer, F.R.S. R. I. Pocock, F.R.S. TREASURER, Horace W. Monckton, F.G.S. SECRETARIES. H. 8. Goodrich, F.R.S. | Dr. A. B. Rendle, F.R.S, GENERAL SECRETARY. Dr. B. Daydon Jackson. COUNCIL. Edmund G. Baker, Esq. Gerald W. E. Loder, M.A. Prof. Margaret Benson, D.Sc. Horace W. Monckton, I'.G.S. EK. T. Browne, M.A. R. I. Pocock, F.R.S. Henry Bury, M.A. Dr. A. B. Rendle, F.R.S. Stanley Edwards, F.Z.S. The Lord Rothschild, F.R.S. Prof. J. B. Farmer, F.R.S. Dr. E. J. Salisbury. Prof. E. 8. Goodrich, F.R.S. C. E. Salmon, Esq. Capt. A. W. Hill, F.R.S. Miss A. Lorrain Smith. Dr. B. Daydon Jackson. Lt.-Col. J. H. Tull Walsh. C. C. Lacaita, M.A. Dr. A. Smith Woodward, F.R.S. LIBRARY COMMITTEE. ‘The Officers ea officio, with the following in addition :— Dr. W. T. Calman. C. C. Lacaita, Esq. E. J. Collins, B.A., B.Sc. Prof. E. B. Poulton, F.R.S. L. V. Lester-Garland, M.A. Miss E. M. Wakefield. Dr. R. R. Gates. | F. N. Williams, Esq. Dr. A. D. Imms. THE GERM-CELLS OF GRANTIA COMPRESSA. ; 261 The Germ-Cells, Fertilization, and Early Development of Grantia (Sycon) compressa”. By J. Bronte Gatensy, B.A., B.Se., D.Phil., Senior Demy, Magdalen College, Oxford ; Senior Assistant in Zoology, Lecturer in Cytology, University College, London. (Communicated by Prof. E. 8. GoopricH, F.R.S., Sec.L.8.) (PLaTeEs 19-23; and 4 Text-figures.) [Read 11th December, 1919. ] CONTENTS. Page En troductoryesemcsata tase cecteces meaisea Aedes ale i seetseeie aie alle ctevae la hoarse cena teea lees 262 2. General Statement of the Problems investigated .................. cee, 262 STEELE VIOUSHLW OF aon seman iereic aoe Pome a ceacien tac cetebcws macteom lattes siole,t race acces 263 ApMaterialzandeNlethod serena nctacauacrets cure vctereers, cewek a sisi saree eels ont ate 264 5. General Account of the Development of a Sycon, up to the Formation of the Free-swimming Amphiblastula Larva ...0. 50.02. 0c esse sess sw eeseccces 265 6, The:Structure‘of the Amphiblastula Marva, i302. .c.c00ces...+-censa.e. 266 7. The Minute Cytology of the Gastral Epithelial or Collar Cell of Grantia.... 268 SyeAmosboidyCell-elements: of Granite aie nam crre ect tercieles 1a ) 9. The Question of the Origin of the Germ-cells. That ova and spermatozoa ultimately arose from collared cells was the conclusion of Haeckel, but Schultze and others claimed that they arose from wandering ameeboid cells of the mesoglea, which they supposed were unrelated to the collar-cells. Poléjaeff and Gorich also consider that germ- cells arise from wandering mesogleal cells, which are not connected with the gastral epithelium. Jérgensen likewise derives his germ-cells from the so-called mesoderm (mesogleal) cells, either resting stellate connective- tissue cells, or amcebocytes. Dendy, as before mentioned, in 1915 published a paper in which he disinterred Haeckel’s “ heresy,” and courageously stated that in his opinion collar-cells did metamorphose into germ-cells. In the present paper I am finally obliged to follow Dendy’s unpopular view, simply because I find myself unable to overlook directly confirmatory evidence, derived from examination of my own and other workers’ sections. It is my opinion that not only do collar-cells migrate into the mesoglea and subse- quently become germ-cells, but also that collar-cells zn situ may metamorphose into oogonia. Schultze, Poléjaeff, Jorgensen, and others may be correct in considering the germ-cells as originating from mesogleal ameeboid cells, but I claim to have shown here that collar-cells are the main source of the above-mentioned ameeboid cells, so that in the long run Haeckel and Dendy are correct. 10. How Collar-cells become Amaboid Cells of the Mesoglea. , In PI. 22. fig. 24 I have drawn by means of the camera lucida a part of a sponge showing the migration of collar-cells into the mesoglea. The cells marked a1, a2, and a3 are three stages in the process; in Pl. 19. figs. 7 and 8 are two later stages showing the matamorpbosis of an amceboid cell into an oogonium. ‘The cells drawn in PI. 22. fig. 24 can only be doing one of two things—either they are migrating into the mesoglea or they are migrating from the mesoglea to the gastral cavity. That they are not doing the latter seems indicated by the fact that Just near the figure a2 in Pl. 22. fig. 24 are gaps in the epithelium showing that cells have left their place, and the cell marked by the figure a3 is opposite one of these gaps. When the collar-cells begin to migrate inwards they lose their collar and flagellum and become ameeboid. As Dendy and Carter showed, collar-cells teased out alive often begin to thrust out pseudopodia. Many amoeboid cells of the mesoglea contain similar cytoplasmic inclusions to those of the ordinary collar-cells, but later their mitochondria often become more marked, as in PJ. 19. fig. 7. Most of the pieces of sponge I have sectioned exhibit appear- ances such as drawn in fig. 24 of Plate 22, and I am led to believe that this behaviour of collar-cells is of constant occurrence during the life of the sponge individual. EARLY DEVELOPMENT OF GRANTIA COMPRESSA. 271. 11. Formation of Oogonia in situ from Collar-cells. In Pl. 22. figs. 23, 25, and 26, I have drawn with the camera lucida parts ot the gastral epithelium to show the metamorphosis im situ of collar-cells into oogonia. The first stage is in Pl. 22. fig. 26; at PC, a collar-cell has sunk below or become partly covered by its immediate neighbours. That this cell, which is larger than ordinary collar-cells, is really a lately metamor- phosed gastral epithelial cell is indicated, firstly because its nucleus resembles that of the collar-cell, and secondly because its cytoplasmic inclusions are closely similar to those of its neighbours. In Pl. 22. fig. 25, the cell marked OG has a cytoplasm just like that of the collar-cell except that its size is greater; the nucleus of this cell is like the nucleus of the collar-cell, except for the size of the karyosome. In PI. 22. fig. 23 at PC and OG are two stages which are similar to those already described. The cell OG is now an oogonium, and is about at the sume stage as that drawn in Pl. 19. fig. 8. In Pl. 22. fig. 27 at OTH is a still later stage. The figures 23-26 of Plate 22 were drawn carefully with the camera lucida, from the edge of a sponge where a large number of young oocytes were found. All stages up to the formation of larger oocytes were present, and I believe there can be no doubt that such a cell as that in Pl. 22. fig. 25 is a stage in oogenesis, and is derived from a collar-cell. 12. Position of Germ-cells in Grantia compressa. In Pl. 19. fig. 5 is part of a sponge showing an egg (OTE) on the left and a group of spermatids (SPT) on the right. These lie beneath the collar- cells, and tend to abut into the gastral cavity (GCAV). Jorgensen (8a) mentions that oogonial divisions may take place in the gastral cavity beneath the collared epithelium, but more often in the mesoglea. The growing oocyte always occupies a position beneath the basement-membrane of the collar-epitheliam, and I believe that it may come to lie in this position, in many cases by a simple sinking inwards as shown in PI. 22. figs. 25 and 26. Between the stages drawn in the latter figures and in Pl. 22. fig. 27, OTE, Jérgensen finds oogonial divisions and the prophases of the first maturation division. These latter have occurred in Pl. 22. fig. 27, OTE, which is well advanced in growth. Jérgensen would possibly interpret the cells marked PC and OGA in PI. 22. figs. 23, 25, and 26, as secondary oogonia which had migrated from the mesoglea to their present position ; this I would dispute, for I find all stages intermediate between such cells and the collar-epithelium cell. The migrations of the oogonia are very peculiar, and I cannot claim that this account is at all exhaustive. The problem is one which will need special study and a larger amount of material than at my disposal ; the main point is to note that collar-cells do not always remain as such, but may grow and metamorphose into other elements. 2A2 DR. J. B. GATENBY ON THE GERM-CELLS AND 13. Spermatogenesis in Grantia compressa. About twenty-five pieces of sponge were sectioned by me, and of these only one contained any obvious stages of spermatogenesis. The sperm- elements of Grantia consist of small groups of cells lying in the interstitia between the gastral chambers. Each group lies enclosed in a special Trxt-Fie. |. Camera-lucida drawing of part of sponge showing oocytes (OTE) and nests of spermatids (SPT) with tails. Compare with PI. 19. figs.5and5a4. The entire substance of this sponge contained patches of spermatids and isolated growing oocytes ; it was a slightly protogynous hermaphrodite, as some of the oocytes were being fertilized. capsule of mesogleal cells, the whole arrangement being unlike anything described by previous authors. It is a moot point as to whether each group of sperm-cells should be called a testis; I think that the cells are too much scattered to form a distinct gonad (text-fig. 1). EARLY DEVELOPMENT OF GRANTIA COMPRESSA, 273 In Pl. 19. fig. 5 is depicted a part of a hermaphrodite individual showing one gastral cavitry, on the left an oocyte about to be fertilized, and on the right a nest of spermatids (SPT), enclosed by their nurse-cells (NSC). The latter are pale cells, cubical, and contain faint granulations ; they are not ciliated and the epithelium they form may be partly interrupted by absence of cells at any one side. The spermatids lie inside the cavity of the nurse-cell chamber (SPA), and their tails can easily be seen. In my material I was in addition able to recognise larger cells also enclosed by a nurse-cell chamber—these I believe to be spermatocytes, as they are of the expected size. These aggregations of sperm-cells probably arise at a certain period of the vear from amoeboid elements of the mesoglea, and the nurse-chamber would be formed subsequent to the inception of spermatogenesis. 14. General Description of the Inclusions of the Cytoplasm in Ooyenests. The young oocyte of Grantia contains the usual three categories of cytoplasmic granules or bodies, 7.e. yolk-spheres, mitochondria, and Golgi apparatus. In PI. 19. fig. 2 is a young oocyte containing three Golgi elements (GAO), with yolk and mitochondria stained yellowish. The apparatus is like that of the flagellated cells, being formed of an inner protoplasmic (archoplasmic) area, surrounded by a peripheral rind of Golgi material. My Kopsch preparations were not suitable to study the Golgi apparatus in all stages, because those which were successful did not contain oocytes older than’ that drawn in Pl. 19. fig. 2. The mitochondria I worked out fully: they are the “chromidia”’ of Jérgensen. Without clouding this present section with controversy with regard to the origin of the germ-cells, it can be stated that those cells (in fact, all cells of the sponge) from which the gametes ultimately arise contain mitochondria in their earliest stages. There can then be no question as to the origin of the sponge mitochondria from chromatin. The mitochondria in the young oocyte (PI. 19. fig. 8) may be extremely large and coarse, and they lie in the endoplasmic region. As the ooeyte grows larger, the mitochondria become more numerous and gene rally smaller at first, but later they may grow very large and dense. As soon as the oocyte has grown so large that the crowding of the granules is less marked, vacuoles begin to appear in the endoplasm, The yolk in the Grantia compressa oocyte cannot be demonstrated properly by short fixation in chrome-osmium, as is suitable for mitochondria. The best method for demonstrating yolk is the unmodified Kopsch (7). In Pl. 19. fig. 9 is an oocyte just before fertilization, showing the mito- chondria (black) and the very fine numerous yolk-granules as lightly stained bodies. The yolk-bodies of the egg of Grantia are so delicate that only 274 DR. J. B. GATENBY ON THE GERM-CELLS AND prolonged osmication enables them properly to withstand contact with alcohol and xylol in the subsequent embedding and staining. I have not hitherto met with such delicate yolk, though that of the mollusc eggsis often very easily destroyed. No other author has described the true yolk of sponge- eggs. While the yolk granules are beautifully fine and regular in size and shape, the mitochondria are often extremely coarse, and of irregular shape, and may occasionally be abnormally large (PI. 20. fig. 13, M). In the unfertilized and unmatured oocyte, the yolk granules and mito- chondria are evenly distributed in the endoplasm, as depicted in Pl. 19. fir. 9, and, as will be seen, are subsequently evenly sorted out among the blastomeres during cleavage of the ovum. 15. Ectoplasmic and Endoplasmic Regions in the Egg. Even in the youngest oocytes one may notice that at an early stage a clear ectoplasmic zone becomes differentiated from an inner or endoplasmic zone. The ectoplasmic zone (Pl. 19. figs. 5 and 9, Pl. 20. figs. 10-13, Pl. 21. figs. 18-20) contains few or no vacuoles, is smooth, and is often drawn out in the form of blunt pseudopodia or filamentous dendriform threads. The inner or endoplasmic zone is vacuolated completely and has a fine frothy appearance ; it is in this region that the cytoplasmic inclusions lie, granules in the ectoplasm of the oocyte being rare or never found. Occasionally, in preparations fixed in mixtures containing alcohol or acetic acid, the vacuoles collapse, and the egg comes to have a curious radiation of fibres around the nucleus (see Jérgensen’s figures, 8a). Hees treated with silver nitrate solution show the endoplasm browner than the ectoplasm, and in some cases the difference in quality of ectoplasm and endoplasm is as marked as that shown in Pl. 19. fig. 5, END, ECT. In favourable cases not only young oocytes, but amoebocytes, may be seen to possess ectoplasmic and endo- plasmic zones (Pl. 19. figs. 7 and 8). As in the case of the older oocyte the cytoplasmic inclusions lie in the endoplasm, while the pseudopodia consist mainly of ectoplasmic material. During development of the egg, all the blastomeres in most cases come to have an equal portion of the ectoplasm. The ectoplasm can be traced through cleavage up to the formation of a blastula (Pl. 21. fig. 21), but it soon either becomes absorbed or is invaded by endoplasmic substance. In Pl. 21. fig. 22, a young blastula is shown, which still has distinctly marked ectoplasmice regions. In a rare number of cases, it was found that in young blastulee the cells of one side had less ectoplasm than those elsewhere, but in no example could I show that this inequality had any relationship to the formation of the flagellated and non- flagellated parts of the amphiblastula. EARLY DEVELOPMENT OF GRANTIA COMPRESSA. Dil 16. The Problem of the Polarity of the Oocyte, the Ovum, and the Embryo. This important problem is very difficult to understand properly, because the granules in the Grantia ege are evenly disposed, and one cannot identify a vegetative or animal pole by this means. In fact, one is obliged to consider that the egg of Grantia is a primitive amoeboid body, without visible signs of a polarity such as one finds in the amphibian egg (PI. 19. fig. 9). In most cases the egg of Grantia lies beneath the gastral epithelium in such a way as to be compressed into an oval form, as shown in Pl. 19. fig. 9 ; the nucleus itself is rarely quite spherical, being oval in the same direction as the egg. While some oocytes at this stage are quite spherical, and their nuclei likewise, it is true that by far the majority of Grantia oocytes possess this form of bilateral polarity. While the oocyte nucleus is most commonly oval or elliptical, the pronuclei are often quite spherical. Though it is difficult to come to a conclusion, I fear that the above observations on the gross polarity of the oocyte are of only small value. The ectoplasm, too, lies around.the entire periphery of the egg, and lends no clue to the solving of this problem. Polar bodies are most often given off towards the gastral side of the egg, but I know of one case where a polar body was passed out on that side of the egg lying towards the mesogiea. I do not consider that much significance can be attached to this. I have been forced to the conclusion that, apart from the bilateral polarity caused by pressure, the oocyte shows no permanent signs of definite polarity. Now during fertilization the sperm always, in my experience, enters at that side of the egg nearest the gastral cavity (Pl. 19. fig. 9, 8), and in by far the greater number of examples the penetration of the sperm causes a partial flow of cytoplasmic granules to that side of the oocyte on which the sperm entered (PI. 20. fig. 13). The egg thus comes to have a definite polarity temporarily conferred by the entry of the sperm. The question then arises as to whether the place of entry of the spermatozoon marks the animal pole, and whether it is the aninial pole which later becomes the flavellated hemisphere of the blastula ? There are several established facts which we may notice at once: The flagellated half of the amphiblastula larva almost always becomes formed from that hemisphere of the undifferentiated blastula abutting against the gastral layer ; the sperm always enters on this same side, and the polar bodies are most often given off on this side; in the differentiating blastula (Pl. 22. fig. 27), the line of demarcation between future flagellate cells and future granular cells is sharply distinguishable. I may say, nevertheless, that I am aware that the above evidence is insufficient fully to settle the question of the polarity. One may, however, follow out this line of reasoning to its fullest extent, though it should be 276 DR. J. B. GATENBY ON THE GERM-CELLS AND noticed that the constant position of entry of the sperm may only be due to the fact that the gastral side is the nearest path which the sperm could take to the egg. The fertilized egg soon loses that temporary polarity conferred by the disturbing influence of the spermatozoon, and soon the granules so dis- arranged flow back and become evenly disposed (Pl. 21. figs. 18 and 19). No methods used by me succeeded in demonstrating any differences in cells of the two halves of the early blastula (Pl. 21. fig 21), and it should be mentioned that in several cases it seemed to me that the future flagellate hemisphere, instead of being turned towards the nearest gastral cavity, was turned away from it, as drawn in PI. 22. fig. 27, where I presume the spermatozoon entered near the letters SPX. The conclusion which seems most likely to be correct is that the oocyte lies with its animal pole towards the gastral cavity, and that the sperm enters at this pole; one may also suppose that the animal pole gives rise to the flagellated cells or histocytes, while the vegetative hemisphere gives rise to the granular cells. It may be noted, too, that the oocyte cytoplasm shows, on a microscopical basis, no signs of regional differentiation. IT believe that while the latter remark is correct, it may be true that the ground-plasma of the oocyte is more or less perfectly differentiated into two sorts of substances, which may be the forerunners of the two different categories of cells of the amphiblastula larva. 17. Preliminary Outline of Events leading to the Union of Sperm and Egg. The following account of the stages leading to the union of sperm and egg is derived only partly from assumption, but mainly from actual observa- tion of microscopical preparations. The spermatozoa are carried into the gastral cavities, presumably being borne along by the inhalent currents. Passing to that region of the gastral epithelial carpet overlying a ripe oocyte, one spermatozoon enters a collar- cell (P1. 19. fig. 9). After the entry of the sperm into the choanocyte or collar-cell, the latter loses its collar and flagellum, becomes amoeboid, and gradually passes through the basement-membrane of the collar-cell epithelium (Pl. 20. fig. 14), and penetrating all obstacles finally comes to lie on what I have tentatively presumed to be the animal pole of the egg : generally the sperm-carrying collar-cell makes a special depression on the surface of the oocyte. Protoplasmie continuity is established between oocyte and sperm- carrying cell, and the spermatozoon is passively borne into the egg by means of protoplasmic currents (Pl. 20. fig. 12). In short, this account would mean to say that the differentiated flagellated collar-cell may become modified, in a short space of time, into an amoeboid non-flagellated cell, whose function is to carry the spermatozoon to the egg. EARLY DEVELOPMENT OF GRANTIA COMPRESSA. 277 I consider that I have found all stages confirming this peculiar happening, not only in Grantia, but also in part in Sycon ciliatum, Jobrgensen’s figures (8a), misinterpreted, show that the same process takes place in Sycandra raphanus. 18. Evidence as to the True Identity of the Sperm-carrying Cell. In by far the greater number of cases, the earliest stages of fertilization that I have found are those at a time when the sperm-carrying cell has reached the egg (PI. 20. figs. 11 and 14). In many such eases it is quite clear that the nucleus of the sperm-carrying cell has a somewhat clearer sub- stratum, and may be bigger than that of the collar-cell. Did one know only these special cases, one might be led to consider that the sperm-carrying cell was some sort of amcebocyte from the mesoglea. Nevertheless, I have con- cluded definitely that the sperm-carrying cell is, or was, a collar-cell, and for the following reasons :— ; 1. In certain few but definite examples the ripe spermatozoon has been found in a collar epithelial cell, which reposed in place alongside its fellows (text-fig. 1, camera lucida). 2. Ina very large number of examples I found spermatozoa in cells in the position indicated in text-fig. 3—that is to say, still in the epithelium, but more or less covered by its neighbours. 3. In a very large number of cases, I believe the majority, the nucleus and the cytoplasmic inclusions of the sperm-carrying cell exactly resembled those of the collar-cell or choanocyte, while the size of the former approxi- mated closely to that of the latter (see page 279). 4. The position of the point of entry of the sperm-carrying cell into tissue of the sponge, on its journey to the oocyte, is remarkably constant (compare text-fio. 4) ; yet when one examines ripe oocytes before the advent of the spermatozoon, no special cells can be found in the epithelium at the region where the sperm-bearing cells later appear. These reasons lead me to believe that the sperm-carrying cell is a modified choanocyte ; I consider that any one of the statements given in the above four paragraphs is sufficient to indicate the true nature of the sperm-carrying cell, and all the facts together constitute a clear and indisputable proof. The occasional slight change in the density of the nuclear substratum (linin) I consider explicable under the assumption that the presence of the spermatozoon exerts an influence leading to this change, and it is important to note that this change does not always occur. The matter is further discussed on page 283. 19. Behaviour of Sperm after Lntry into Collar-cell. In my own preparations I never found the last stages in spermatogenesis, so that it is impossible for me to say whether the nucleus of the sperm is altered by its entry into the collar-cell. In Pl. 19. fig. 5.4, I have drawn LINN. JOURN.—ZOOLOGY, VOL. XXXIV, pal TPXxT-Fias. 2-4, Letters CHC=collar-cell of gastral epithelium. CT and KND=ectoplasmic and endo- plasmic regions of oocyte. M=mitochondria. SP=spermatozoon. Figs. 2 and 3.—Two stages in the fertilization of Grantia ; in fig. 2 the spermatozoon (SP) has just entered a collar-cell; in fig. 3 the collar-cell has lost its collar and is sinking below its fellows. Subsequent stages are drawn on PI. 20, figs. 12, 16, and 17. Fig. 4.—Part of an oocyte and the covering-wall of collar-cells (CHC). The black dots represent the comparative positions of twenty-five spermatozoa in that number of cases examined in the stage drawn in fig.3. In all I have over fifty oocytes in stages at or between Text-figs. 2 & 5 and Pl. 20, fig. 11. EARLY DEVELOPMENT OF GRANTIA COMPRESSA. 279 the oldest spermatids I could find, and in PI. 19, fig. 6 I have drawn the sperm possibly just after it has entered the collar-cell. The tail of this sperm has been added to my drawing. From what evidence I gained by examining a number of ripe sperms lying in collar-cells, I believe that the only change undergone by the sperm in this position is the loss of its tail or axial filament. The nucleus of the sperm, after entry into the collar-cell, might swell a little, but I have no evidence that such is the case*. It is only when the sperm is drawn into the egg that the former breaks into parts, and the nucleus begins to swell. The collar-cell has no such effect on the sperm: this is a very interesting point. Gérich, however, in Spongilla draws the sperm quite differently from that of Grantia, and if the free sperm of the latter resembles that of Spongilla it is certain that some change comes over the Grantia sperm after entry into the collar-cells. Gorich’s account I do not trust in details. 20. Changes in the Collar-cell brought about by the Entry of the Spermatozoon. While most sperm-carrying cells at the stages drawn in text-fig. 3 closely resemble the collar-cell, it is true that by the time the former reach the oocyte some undoubted changes have taken place, especially within the nucleus. The example drawn by Professor Dendy (2) on Pl. 26. fig. 52 of his paper is a fairly typical case, such as I have occasionally noted in my own material ; a similar example is drawn on PI. 20. fig. 16 of this paper. I find that a good deal of variation in the general appearance of the nuclei of the sperm-carrying cell can be noted, and some nuclei are exactly similar in size, shape, and staining-power to the collar-cell nucleus, while others lack especially that dense substratum peculiar to collar-cell nuclei. I consider that this partial clearing-up of the dense nuclear substratum is caused by the prolonged presence of the sperm, and is in some way connected with the amoeboid phenomenon of the metamorphosed collar-cell. In addition to this occasional change, all sperm-bearing collar-cells lose their collar and flagellum, and become amoeboid. As would be expected, most sperm-bearing cells are a little larger than the unmodified collar-cell : this is due to the presence in them of the sperm. 21. The Fate of the Sperm-carrying Cell after the Sperm passes into the Oocyte. The sperm-carrying cell lying upon the surface of the oocyte is a most conspicuous object, yet in the cleavage-stages subject to fertilization this cell could not any longer be made out. Having examined a number of early * T have lately procured evidence that the sponge sperm is filiform and flagellate, so a great change comes over it :fter entry into the collay-cell (see Journ. Roy. Mier. Soc. 1920). 21* 280 DR. J. B. GATENBY ON THE GERM-CELLS AND cleavage-stages I have come to the conclusion that the sperm-carrying cell either wanders away after the spermatozoon has passed out of its cytoplasm, or in some way effaces itself by degenerating or shrinking in size. I found only one case suggesting that the sperm-carrying cell subsequently degene- rates: in Pl. 21. fig. 18, the cell at SCC is obviously unhealthy. At present I am led to consider that the sperm-carrying cell subsequently wanders away through the mesoglea. 22. Behaviour of Sperm and Sperm-bearing Cell explicable as a Chemotactic Phenomenon. In text-fig. 4 is drawn diagrammatically a part of an oocyte and the neigh- bouring row of collar-cells. Twenty-five oocytes were examined at the stage of fertilization drawn in text-fig. 3, and the position of the sperm was indicated in the diagram in fig. 4 by a black dot. This diagram, therefore, gives a graphic representation of the fact that the swimming spermatozoon is definitely attracted in some way towards the egg. If this were not so, one would not find the sperm almost invariably entering cells neighbouring the egg. I have not found sperms in collar-cells further away from the ege than those drawn on the extreme left and right in fig. 4. Chemotaxis as an explanation of the attraction of any sperm by any egg has been abandoned to a great extent by zoologists, but surely here is a definite case of attraction between egg and sperm? The egg of the sponge seems able to give forth some substance which causes the sperm to swim to that region of the gastral epithelial carpet which just overlies an oocyte. I cannot think of any other explanation which accounts for text-fig. 4 so satisfactorily. ; 23. The Entry of the Spermatozoon into the Egg and the Lffect on both Gametes. The sperm is drawn into the unmatured oocyte probably by synchronous flowing movements of the protoplasm of both sperm-carrying cell and the oocyte. That the sperm itself has no immediate part in the process seems indicated by the fact that it is able to pass into the egg with its nucleus forwards or its middle-piece forwards (Pl. 20. fig. 11) or sideways. There is undoubtedly a complete protoplasmic continuity established between the cytoplasm of the carrier-cell and that of the oocyte ; here the vitelline membrane of the egg is interrupted as shown in PI. 20. figs. 12 and 14. When the sperm has passed into the egg, the first change noticed in it is the breaking-down of the sperm-cell wall (Pl. 20. fig. 16). The acrosome (AC) becomes faint, and is just disappearing in the last figure. The sperm cytoplasm becomes furred at its edges where it touches the egg-cytoplasm ; soon a flowing movement, apparently initiated by the entry of the sperm, tends to disturb the arrangement of mitochondria, which partly pass to that hemisphere of the oocyte at which the sperm entered (PI. 20. figs. 10, 12, 13). EARLY DEVELOPMENT OF GRANTIA COMPRESSA. 2 oral — Vacuoles collect near the region of the sperm, and the path of the latter into the egg becomes marked by an area formed of more liquid substance (Pl. 20. figs. 13 and17}. The fine yolk-spheres keep closely around this area, but do not invade it, as is shown in PI. 20. fig. 13. Subsequent to the entry of the sperm, the oocyte undergoes maturation giving off two polar bodies. 24. The Fate of the Middle-piece or Sperm Macromitosome (‘‘ Nebenkern’’) during Fertilization. In Ascaris and Phallusiait has been shown that the mitochondrial substance in the middle-piece of the spermatozoon does not degenerate during fertili- zation, but retains its faculty for growth and division and apparently takes some minor part in the phenomenon of fertilization (14). With Champy-Kull technique the middle-piece of the Grantia sperm stains a bright red, while the egg-cytoplasm is a greenish or golden-grey colour ; it is, nevertheless, a somewhat difficult matter to ascertain exactly what becomes of this body during fertilization. Between the stages of Pl. 20. fig. 12 and fig. 13, the middle-piece in its form of a bun-shaped bead completely disappears. In Pl. 20. fig. 17, MAM, it is quite clear that the middle-piece is breaking up into a cloud of small granules which adhere to the growing male pronucleus like the tail of a comet. In subsequent stages I never found any granules which I could identify for certain as having originated from the macromitosome, but in Pi. 20. fig. 13, at w, are certain grains which might be the remains of the original middle-piece. While I do not wish to deny that there may be activity of the male mitochondria subsequent to the stage of Pl. 20. fig. 17, I consider that such is unlikely. The evidence either way is not complete enough to allow me to express a definite opinion. 25. The Mitochondria in the Development of Grantia compressa: In certain Ascidia it is established that the mitochondria become unevenly segregated into different blastomeres, the myoplasm being formed mainly from these granules. The two regions of the sponge amphiblastula do not contain granules in equal quantity, and it is interesting to ascertain whether in Grantia there is any special segregation of cytoplasmic granulations into special blastomeres. Subsequent to fertilization in Grantia compressa, the mitochondria become more or less evenly spread out in the endoplasm (PI. 21. fig. 21), and during cleavage the distribution of the mitochondria between the blastomeres is equal, as shown in PI. 21. figs. 19, 20, 21, and 22. In Pl. 21. fig. 22, the blastula is found to be formed of cells, all of which contain approximately equal numbers of mitochondrial grains. It is only much later during the differentiation of the two main cell-layers in the 282 DR. J. B. GATENBY ON THE GERM-CELLS AND embryo that any differentiation of granules appears, and these differences seem due to slow growth and not to any segregation during cleavage. It is apparent that pari passu with the cleavage of the egg into smaller and smaller blastomeres, the mitochondria either shrink considerably in size or break up into smaller fragments (compare the size of the granules in PI. 21. fig. 19 and fig. 22). It is not true that the mitochondria (chromidia of Jorgensen) disappear in the late blastula-stage, as was thought to be the case (8 a). 26. Differentiation of the Nuclet of the Blastula-wall heralding the Formation of Two Definite Tissues. On Pl. 21. fig. 21 I have drawn a Grantia blastula at a stage when all the cells appear alike ; the nuclei and the inclusions of the cytoplasm of these cells seemed to me to be similar throughout the entire blastula. On Pl. 22. fig. 27 is a blastula fixed at the period when the nuclei of the future flagellated hemisphere have begun to be easily distinguishable: the earliest signs of differentiation of the future flagellated cells is constituted by a tendency for the ceils of one hemisphere to divide more rapidly, and to become more regularly arranged: the peculiar denseness of the nuclei of this layer is a subsequent and rapid differentiation. In PI. 22. fig. 27 the granules in the cytoplasm are not different in either region, and it seems certain that nuclear differentiation precedes or instigates special cytoplasmic activity. Soon after the stage depicted in Pl. 22. fig. 27, the character of the yolk- spheres becomes altered ; this is indicated by the fact that the yolk-spheres of the embryonic tissues are more persistent and less easily destroyed than those of the egg or earlier blastula. Owing to lack of suitable material, there were several interesting problems which I did not examine, but it is hoped to investigate these at some later period. 27. Discussion. (a) Organ-forming Substances ? In this study the behaviour of yolk and mitochondria has been followed out in the development of the sponge Grantia compressa. The introduction of specific and selective staining and fixing methods has enabled me to identify mitochondria, yolk, and Golgi apparatus. The sponge falls into the group of animal forms which have been shown to possess the two protoplasmic inclusions. The main result of this research, in so far as it touches on the mitochondria, has been to show that they pass through development without losing their individuality, and that they are distributed evenly during cleavage to the blastomeres. No sort of segregation of materials in special blastomeres could be detected with the methods used by me, and I have come to the conclusion that such differentiation of cell-layers and materials as one finds in the EARLY DEVELOPMENT OF GRANTIA COMPRESSA. 283 amphiblastula larva is of gradual growth in the stages between the end of cleavage and the final stages of the formation of the free-swimming larva. I could not find any evidence for such segregation of materials as described by Conklin (1) in Cynthia, and Ihave come to the conclusion that, so far as fixed inaterial goes, no visible organ-forming areas can be demonstrated in the egg of Grantia compressa. It is possible, however, that delicate easily destroyed pigments may be present in the fresh ripe egg, and these might mark out ground-protoplasmic areas not to be distinguished in fixed material. Of this I have no evidence. (b) Chromidia and Mitochondria. Jorgensen and Dendy both describe the mitochondria under the name “chromidia,’ implying that they have a true chromatinic nature and are related to the nucleus. If the egg is treated by Champy-Kull’s method the mitochondria are red (fuchsin) and the nucleus is blnish. In the Bensley-Cowdry stain the mito- chondria are red (fuchsin) and the nucleus is green (methyl-green). In Carnoy’s fluid the mitochondria are dissolved away and the nucleus remains. The mitochondria of Grantia are not “ chromidia” in the proper sense of the word, nor are they derived from the nucleus at any period of oogenesis *. All embryonic cells contain mitochondria, and there is no question of the latter appearing from the nucleus at any stage of the oogenesis of egg-cells derived from the ameeboid or other elements of the mesoglea or collar-epithelium. In a late paper on Apanteles I described true cytoplasmic chroimidia in the egg of this form; such chromidia stain like chromatin, and ultimately metamorphose into secondary nuclei. In the Hymenopterous egg the secondary nuclei (chromidia) and the mitochondria are easily distinguishable and can be stained in different colours. (c) Fertilization. In this paper I have given the first correct interpretation of the peculiar stages already described by Gorich, Jorgensen, and Dendy. It has been demonstrated beyond a doubt that the spermatozoon of Grantia is brought to the oocyte by a carrier-cell. On page 277 I have given my reasons for sup- posing that the sperm-carrying cell is really a collar-cell which has lost its collar and flagellum, and become ameeboid. Should subsequent work confirm tnis description (and I have little doubt that there has been no mistake) this will constitute one of the most remark- able occurrences in the fertilization of any animal. It is remarkable that the collar-cell temporarily containing the spermatozoon should desert its position in the gastral epithelium and wander into the * T now abandon this view and admit Dendy’s description of nucleolar extrusions to be correct. This matter is treated at length in my new paper in the Journ. Roy. Micr. Soc. 1920. 284 DR. J. B. GATENBY ON THE GERM-CELLS AND mesogieal tissue. That the sperm is the instigator of this unusual behaviour seems certain, and apparently any collar-cell may transform into ‘a sperm- carrying cell. From examination of Dr. E. 8. Goodrich’s small series of Sycon ciliatum slides, I believe that in this form also the collar-cell acts as a sperm-carrier, and Jérgensen’s account of similar stages in Sycandra raphanus, all misinterpreted, lead me to consider that the process described by me for Grantia compressa will apply to other forms of sponges. Ii is evident, as Professor Dendy has himself pointed out to me, that two distinct processes have been confused. The young oocyte is fed by a process of phagocytosis. It is definitely established by Professor Dendy that certain large ameeboid elements of the mesoglea carry other cell-elements to the young sponge-oocyte, and the latter engulfs the food-matter so offered. The cell-elements which bring such cell-food to the growing oocyte are unlike the cell which later brings the spermatozoon to the oocyte ; moreover, this process of feeding only takes place in small oocytes. After the latter have grown to half their ultimate size, ingestion of other cells is rarely found to occur, and, moreover, ingestion of other cells by the growing oocyte takes place on any part of its surface, whereas the process of fertilization occurs on one definite side of the oocyte. There can be little doubt that not only is the process of phagocytosis in the oocyte different from the stages leading to fertilization, but also that the cells engaged in the two processes are not so alike as to Jead one to believe that they are derived from the same source. The cells engaged in bringing food-matter to the young oocyte are nearly always much bigger than those which bring the sperm to the egg, and the cell-contents of the two cells are not alike. (d) The Sperm-carrying Cell considered as a Mesogleal Amebocyte? The sperm-carrying cell has been assumed by me to be a modified choano- cyte or collar-cell. There may be critics of this view who would prefer to take the other alternative, and look upon this peculiar cell as an amoebocyte from the mesoglea: according to such critics, the amcebocyte would be much more likely to behave in the manner of the sperm-carrying cells ; moreover, it is true that some of the latter have nuclei rather unlike those of the collar- cells, and similar to the true mesogleal amcebocytes. In the first place, one never finds amcebocytes near unfertilized but ripe oocytes in the position in which the sperm-carrying cell later appears (text- figs. 3 and 4). Secondly, I have never found spermatozoa in amcebocyte elements of the mesoglea, and I have found several cases where sperms lie within what I presume to be choanocytes (text-fig. 2, SP). Thirdly, there is the fact that the position of the sperm-carrying cell is remarkably constant (text-fig. 4): now, if mesogleal amosbocytes carry the sperm to the egg, why is this position so constant? Presuming that sperms swim through into the bo (os) Ou EARLY DEVELOPMENT OF GRANTIA COMPRESSA. mesoglea and get. into amobocytes, we are then forced to believe that each sperm-bearing amcebocyte, whatever its primary position, always carries the sperm to one side of the egg ; there would then undoubtedly be cases where the ameeboeyte would have to perform a comparatively long journey with the opposite pole of the egg as its goal. Against this amcebocyte view there is a fourth reason: cells carrying sperms can never to my knowledge be found on the short “ side” of the egg, as marked by the star on the right of text- fig. 3. Had the sperm-bearing cell been a mesogleal amcebocyte I would surely have found one with a sperm, in positions other than the gastral side of the oocyte (text-fig. 4). Reference to Professor Dendy’s pl. 24. figs. 50 and 51, and comparison with fig. 52, will show that the cell which carries food to the oocyte is not only bigger than that which carries the sperm, but has different cytoplasmic contents. In Prof. Dendy’s fig. 52 the sperm-carrying cell (NC) is about the same size as a collar-cell. To explain my text-fig. 4 satisfactorily on the basis of the amcebocyte view, one would need to assume that there was a row of waiting amoebocytes ready in the position in which the sperm is later found. No author who has drawn sponge-oocytes has figured such amcebocytes in this region, and I have never seen an ameebocyte waiting in this part of the gastral epithelium, To assume that the sperm pushes its way into the mesoglea till it finds an ameebocyte to carry it to the egg is preposterous, for if the sperm can push into the mesoglea, there would be no need for another cell to carry it to the egg. I think it will be obvious to all that the sperm does penetrate into a collar- cell, cannot itself pass through to the oocyte, but is carried by the collar-cell, which, during the process, may occasionally become somewhat altered in appearance. its way, “‘ under its own power,’ (e) Germ-cells and Sex in Sponges. My observations lead me to consider that all Grantia individuals are probably potential hermaphrodites, but that the occurrence of sponges with sperm-stages is somewhat rare. In only one case out of twenty-five that I know was the sponge a positive hermaphrodite, and in all the other examples I examined only oogenesis stages were discovered. I am led to consider that Grantia compressa is either a simultaneous hermaphrodite, or female, both sorts of individuals being found. The sponges which I found to contain oogenesis stages alone might have been either protandrous or protogynous hermaphrodites, but this seems rather improbable, for at the time I fixed my material the bulk of the oocytes were ripe and being fertilized, while there was a general absence of sperm-stages, except in one individual, which contained both spermatids and oocytes being fertilized. 286 DR. J. B. GATENBY ON THE GERM-CELLS AND One small point I may mention with regard to any possibility of self- fertilization in Grantia: the sole individual in which I found unripe spermatids also contained oocytes just being fertilized, showing that the sperms must have come from some other sponge, or otherwise one would have found later stages in spermatogenesis. As a matter of fact, the whole question of sex in sponges is far from being definitely settled, for I doubt whether certain previous observers of marine sponges have really identified the true sperm-stages. (£) Spermatogenesis in Grantia. > just under Haeckel in his ‘ Kalkschwamme ’ figures small “ sperm-balls ’ the gastral epithelium, much in the same position as in Dendy’s figures. Neither Haeckel nor Dendy have produced convincing evidence as to their “ sperm-balls.” Subsequently to my writing the account of Grantia spermatogenesis given on page 272, I had the opportunity, through Professor Dendy’s kindness, of seeing not only his own “ sperm” stages, but also those of Dr. Poléjaeff. In figs. 86-93 of Professor Dendy’s plate 26 are drawn what this observer describes as “sperm-morule,” with spermatozoon heads; he has also described the derivation of these morul from other cells which he believed to be spermatogonia and spermatocytes. On pl. 26. fig. 94 he gives figures from Dr. Poléjaeft’s slides. After carefully reading Dendy’s account, and having seen his preparations, I have little hesitation in claiming that what he has described are certainly not sperm-stages of any sort; I believe that these bodies he describes are parasites or inquilines of some kind, probably of plant nature. Dr. Poléjaeff’s sperm-balls are, I consider, nothing more than some protozoan, while the granules inside the cell are not individual cells, but discrete cytoplasmic aggregations. Moreover, neither Dendy nor Poleéjaeff figure tails on their spermatozoa, while I find tails on my spermatids *. I believe that there is no doubt that the cells I have described are true stages of spermatogenesis, and I sectioned twenty-five sponges before I found them. I wish to emphasize this fact, for other observers have found the same scarcity of spermatogenesis stages in sponges, and this is one reason why two observers (and probably three) identified either parasites or inquilines as sperm-stages. Wilhelm Gérich in Spongilla fluviatilis describes what are undoubtedly stages in spermatogenesis ; his description of the formation of the spermato- zoon from the spermatid is quite detailed, but he mentions neither mitochondria nor Golgi apparatus. In Spongilla the sperm-stages lie inside a wall of nurse-cells (‘‘Cystenzellen”’), somewhat as in Grantia. In Sycandra raphanus * At the meeting of the Society, after I had read this paper, Prof. Dendy, in reply, abandoned his previous views with regard to his “sperm-morule,” and stated that in two other sponges he had found stages like those which I have described in this papev. EARLY DEVELOPMENT OF GRANTIA COMPRESSA. 287 wd he describes the formation of the “ Spermatocyste ’’ by means of cell which approaches the spermatogonium (‘‘Spermazelle”), and surrounds it com- pletely; he figures and describes the sperm-cell dividing inside the cyst-cell. If Gérich’s account of the formation of the nutrient-cell capsule is correct, it would mean that in Grantia probably all the cells in Pl. 19. fig. 5, marked NSC, would have originated from a single nurse-cell. I doubt this, but am unprepared at present either finally to reject or adopt Gorich’s account. In Gorich’s spermatid and spermatozoon there is no nucleolus. (g) The Collar-cells as the Dominant and most Characteristic Tissue of the Sponge. Dendy (2) remarks, ‘The one constant and characteristic feature about sponge histology is, of course, the collared cell, and that is only constant in the sense that its typical form is that which possesses a collar and a flagellum. The sponge is, after all, not very much more highly advanced in organisation than a colony of choano-flagellate Protozoa.” This paragraph aptly describes the views, at which I have myself arrived by my independent observation. The collar-cell is the dominant cell in the sponge, and I feel that Haeckel was correct in tracing the origination of germ-cells from collar-cells. Collar-cells are continually migrating into, and reinforcing, the mesoglea of the growing sponge ; the collar-cell is really very litile differentiated except for the flagellum and collar, which, we are now quite sure, can readily be withdrawn, The collar-cell then becomes an active ameeboid cell- individual, whose subsequent fate possibly may be to metamorphose into a germ-cell. One is obliged to remark that in the discussion on the origin of germ- cells—not only in sponges, but in other animals—there has been far too much reliance set upon theoretical conceptions, and not enough on simple observa- tion. It does not suffice for critics to state that they ‘do not think that collar-cells can metamorphose into germ-cells”’ ; they must bring forward something more definite before one can give their views adequate attention. So far as I am able to understand, the crities of Haeckel’s view as to the origin of sponge germ-cells base their assumptions on the illogical and unnecessary extension of Weismann’s views to sponge embryology. According to these observers, germ-cells are supposed to be derived from non-differentiated embryonic elements, and the collar-cell, having become differentiated, is debarred from entering the charmed circle of cells which alone can become germ-cells. Speaking for myself, I am quite unable to see why a collar-cell cannot differentiate, especially when one remembers that an embryonic cell can become a collar-cell. If the stimulus can be provided to make a cell undergo certain changes in one direction, why should not other stimuli cause the changed cell to pass back along its former path? ‘The collar-cell is not a highly ditterentiated cell, such as a metazoan ganglion or gland-cell. 288 DR. J. B. GATENBY ON THE GERM-CELLS AND (h) Final Remarks. Professor Dendy has given me permission to state that he accepts my explanation that what he described in his plate 24. fig. 52, as feeding of the oocyte by nurse-cells, is really a stage in fertilization. The question of the true nature of the sperm-carrying cell is less well established, but Professor Dendy considers that in all probability my interpretation will be found to be the correct one. The problem of the spermatogenesis. is a moot point, but I believe that few will doubt that what I have described in text-fig. Land PI.19. figs. 5 and 5 a are the true stages in Grantia. Haeckel, Poléjaeff, and Dendy have failed to bring forward convincing evidence as to the spermatogenesis, and no cytologist acquainted with these questions would identify the ‘ sperm- balls” of these workers as authentic stages in the formation of spermatozoa in any animal. Attention is drawn to the remarkable facts which other authors and I have established with reference to sponge gametogenesis :— 1. The young oocyte bodily engulfs large cells and feeds at their expense. 2. The oocyte and ovum are both amoeboid and capable of moving through the mesoglea. 3. Spermatozoa are developed in special nurse-chambers, possibly formed of mesogleal cells. 4. The spermatozoa are carried to the oocytes. by special carrier-cells, which are very probably collar-cells. 5. The larva is nourished by a special nutrient chamber formed of mesogleal cells. 6. The pupal stages comprise many peculiar events, which lead to the formation of the sponge. Probably in no other order does the gametogenesis and early development present so many extraordinary peculiarities. This paper merely touches on a few of these peculiarities, and I hope in subsequent work further to pursue the subject. 28. SUMMARY. Choanocyte or Collar- Cell. 1. The nuclear background contains a fairly dense basophil substance, which is either a dense linin,network or a chromophil karyoplasm—more probably the latter (PI. 19. fig. 1). The collar of the choanocyte bas two thickenings, an inner and an outer ; the latter one is very well marked ; both thickenings are more easily seen in the partly withdrawn collar (PI. 19. fig. 1, B). 3. The collar consists of delicate protoplasm which is best preserved by EARLY DEVELOPMENT OF GRANTIA COMPRESSA. 289 prolonged osmication ; the collar-cell is often drawn out into two regions, an outer “neck” and an inner “ base,” containing all the granules and the nucleus (PI. 19. fig. 4); the Golgi apparatus and centrosome generally lie in the “neck,” other granules in the “ base.” 4, Collar-cells contain three cell-elements other than the nucleus: the Golgi apparatus (and centrosome), the mitochondria, and metaplasmic store- granules. 5. The mitochondria are irregular, but fine, and lie around the nucleus CEIN TON tome Metiow3: VE) 6. The metaplasmic or yolky store granules are much larger than the mitochondria, but lie in the same region. (PI. 19. figs. 1 and 4, Y). 7. The Golgi apparatus lies in the “neck” of the bottle-shaped cell, either upon the nucleus (Pl. 19. fig. 1, A) or separately (PI. 19. fig. 1, B). 8. In certain definite cases the centrosome is seen to lie inside the Golgi apparatus (PI. 19. fig. 1, A) and the flagellum originates from the centrosome. In other cases the centrosome appeared separate (Pl. 19. fig. 4, C7). Oogenesis. 1. The young oocyte contains the three cytoplasmic inclusions: yolk granules, Golgi apparatus, and mitochondria (PI. 19. figs. 2 and 9). 2. The oocyte consists of two well-marked regions: a peripheral smooth ectoplasm and an internal frothy endoplasm, in which all the cytoplasmic granules lie (PI. 19. fig. 9). 3. The Golgi apparatus consists of isolated multiple elements formed of several rods arranged in angular formation, much resembling that of the molluse oocyte (Pl. 19. fig. 2, GAO). 4, The mitochondria are fairly large, often irregular bodies, lying in the endoplasm (PI. 19. fig. 9, M). In the ripe oocyte they are often abnormally large, and are never very numerous. The mitochondria constitute the “ chromidia” of Jérgensen and Dendy. In their histochemical reactions they differ from chromatin. 5. Yolk-spheres are very fine and delicate, entirely filling the endoplasm and lying in the trabeculee between the endoplasmic vacuoles. The yolk is only properly preserved by Kopsch’s method. Yolk-granules are formed in and by the ground-cytoplasm, and have no connection with the mito- chondria (PI. 19. fig. 9 ete., Y). 6. The egg is amoeboid and ectoplasmie pseudopodia are common (PI. 19. fig. 9, ECT). 7. The full-grown oocyte is oval or elliptical in shape, as is the nucleus, but the granules and yolk are evenly disposed, and there is no sign of a definite vegetative or animal pole. 8. No organ-forming areas, or visible aggregations of cytoplasmic materials, was made out. 290 DR. J. B. GATENBY ON THE GERM-CELLS AND Spermatogenesis *. 1. The spermatozoon is probably a (flagellated) ovoid body, containing a pun-shaped nucleus. The spermatid is flagellate, so we know that the spermatozoon must be flagellated, though one was never found outside a sperm-carrying cell. In Pl. 19. fig. 6 is drawn a sperm from a sperm- carrying cell, the tail being added from observation of spermatids. The sperm has an acrosome in front, which lies on the edge of the cell, but which is not pointed. The nucleus of the spermatozoon is bun-shaped, reticulate, and contains a nucleolus (karycsome) ; the mitochondrial apparatus is in the form of a bun-shaped fuchsinophile, macromitosome (‘‘ Nebenkern’’), or middle-piece. The centrosome lies partly between the nucleus and macro- mitosome, but more in the latter than in the former (Pl. 19. fig. 6). 2. Spermatogenesis takes place inside a definite capsule of pale, cubical, granular, non-flagellated mesogleal cells. The sperm-stages are aggre- gated into little patches here and there throughout the entire area of the hermaphrodite sponge; these patches lie beneath the flagellated gastral epithelium, and stain more darkly than their surroundings. Both spermatocytes and spermatids have been identified ; the latter are flagellated (text-fig. 1 and PI. 19. fig. 5). 3. The number of spermatids in one patch is generally above fifty. Fertilization (1st Part). ‘ 1. The ripe sperm, after having entered into the gastral cavities, is carried from the collar-epithelium to the oocyte by means of a carrier-cell. The sperm lies inside the carrier-cell, which hears its passenger to the oocyte. 2. The carrier-cell is assumed to be a modified collar-cell or choanocyte, for the following reasons :— (a) There are two cases in which I have found a sperm inside a collar- cell (text-fig. 2, Pl. 20. fig. 10). (b) There are a large number of cases in which I have found a sperm inside a cell which closely resembles a collar-cell, except for absence of collar and flagellum, and which lies still within the collar- epithelium in the position drawn in text-fig. 3. (c) There are about fifty cases in which I found the spermatozoon contained within a cell which always lay in the region indicated by the black dots in text-fig. 4. No sperm-carrying cells or spermatozoa were found in any part of the mesoglea or at the short sides of the egg. (d) I have never found, nor has any other author figured, mesogleal amoehocytes lying near unfertilized oocytes, in the invariable position in which the sperm-carrying cell later appears—only collar-cells are found there ; the gastral cavities were not found to contain sperm-carrying cells—the collar-cells alone can be the source of the sperm-carrying cells. * See foot-note on page 279, EARLY DEVELOPMENT OF GRANTIA COMPRESSA. 291 3. The entry of the spermatozoon into the collar-cell does not cause the former to break up into its constituent parts. 4. The entry of the sperm into the collar-cell or choanocyte causes the latter to lose its collar and flagellum. Subsequently the nucleus of the collar-cell may lose that characteristic smooth chromophil background, and become more open and reticulate—this does not always occur. The sperm- containing choanocyte sinks below its fellows in the epithelium (text-fig. 3), passes through the basement-membrane of the choanocyte epithelium, and comes to rest on that surface of the ovcyte nearest the collar-epithelium. Occasionally, the sperm-bearing cell makes a hollow in the egg, into which it fits (Pl. 20. fig. 11). 5. The wall of the sperm-carrying cell and of the oocyte vitelline membrane come into direct contact ; they become interrupted, protoplasmic continuity is established, and the spermatozoon flows passively into the egg (Pl. 20. figs. 11, 12, 14, and 16). 6. The sperm may enter the egg frontways, backways, or sideways (P1.20. s. 11 and 12). 7. The sperm does not appear to undergo the special rotation noted in fig other forms. 8. The sperm-carrying cell, after the sperm has passed away, does not immediately change, for it can be found during early cleavage. Later on it becomes increasingly difficult to find this cell, and one is led to believe that it subsequently wanders away from the scene of its former activity. In one case I thought the sperm-carrying cell had begun to degenerate. 9. From the fact that the spermatozoa are always found ina circumscribed area of the flagellated cavities, just where the eggs lie, it has been concluded that this is a definite example of chemotaxis (text-fig. 4), Fertilization (2nd Part). 10. The spermatozoon, after entry into the egg, first of all loses its cell- wall, which degenerates. 11. The acrosome next becomes chromophobe, while the edges of the sperm-cytoplasm tend to spread into the surrounding egg-cytoplasm, being for some time recognizable by its coarse stringy appearance, but later it can no longer be noticed (PI. 20. figs. 16 and 17). 12. The sperm nucleus now begins to grow, and other nucleoli put in an appearance (PI. 20. fig. 13). The nucleus becomes spherical. 13. The sperm-middle-piece or macromitosome breaks up into a cloud of fine granules, which adhere to the growing male-pronucleus like the tail of a comet (Pl. 20. fig. 17, MAM). Ima later stage these granules could not be identified. 14. Soon after the entry of the sperm, one finds that in the majority of oocytes there is a flowing towards the region of entry of many of the bo 92 DR. J. B. GATENBY ON THE GERM-CELLS AND mitochondrial spheres, so that the hemisphere away from the gastral layer is partly bereft of mitochondria (PI. 20. figs. 10 and 13). 15. The path of entry of the sperm is marked by an area formed of a chromophobe vacuolated substance, probably intra vitam of a liquid nature. 16. The oocyte nucleus breaks up and forms two polar bodies during these changes in the spermatozoon (PI. 20. fig. 13, PB), and finally the male and female pronuclei grow to the same large size (Pl. 21. fig. 18). Cleavage and Development. 1. During cleavage and early development the yolk and mitochondrial granules are sub-equally sorted out among the blastomeres (Pl. 21. figs. 19, 20;,.22, and 211): 2. During the histogenesis of the two main tissues of the sponge-larva (Pl. 28. fig. 28, FC & GC), these mitochondria and yolk-grains (and probably the Golgi elements), though sorted out previously into subequal quantities between the blastomeres, later become more or less numerous, or specially modified or grouped in special regions, as the tissues develop and differentiate. 3. Such differentiation is not, so far as the granules themselves are con- cerned, in any way traceable to a special segregation during cleavage. 4. The fixing and staining reactions of the yolk-spheres of the egg are different from those of the amphiblastula larva, for during differentiation or histogenesis the yolk-spheres become denser and less delicate, especially in the flagellated cells. The same applies, though less well, in the case of the mitochondria. The Amphiblastula Larva. 1. The larva is formed of three sorts of cells: anterior flagellated histo- cytes, posterior “ granular ” archeocytes, and inner amcebocytes. All these originate and differentiate from a unilaminar blastula formed of similar cells (Pl. 21. fig. 21). For larvee see Pl. 28. figs. 28 and 32. 2. The larva in later stages of its formation becomes partly or wholly surrounded by a layer of squamous mesogleal (maternal) cells, which are marked by their containing peculiar irregular mitochondria. This nutrient capsule is best developed near the posterior “ granular-cell” pole of the larva, and often interrupted on the flagellated hemisphere (PI. 23. figs. 28 and 32). 3. In many cases these maternal cells penetrate and lie among the posterior “ oranular cells” (Pl. 28. fig. 28, LGC) ; when the larva breaks away from the sponge the nutrient capsule is left behind. 4. The histology of the three cell-elements of the larva is as follows :— The flagellated hemisphere is formed of flagellated cells, marked especially in the live animal by the possession at their posterior or inner pole of a dense group of yolk-spheres which are pigmented (brownish). There is mixed up EARLY DEVELOPMENT OF GRANTIA COMPRESSA. 293 with these yolk-granules, but more towards the posterior or inner pole, a number of mitochondria. At the outer or anterior pole of the cell lie a group of special granules doubtfully identified as of true yolky nature. The nucleus is small, extremely basophil, and contains a karyosome. Lying on the nucleus is a distinctly marked capsular body, from which the flagellum arises ; this body is thought to be the Golgi apparatus (Pl. 28. fig. 33). The large centrally situated “granular cells” (GC in Pl. 28. figs. 28 and 32) are formed of a nucleus which stains palely, and contains a coarse karyosome and a cytoplasm in which lies a number of granules, some mitochondria, others yolk, and certain others (GX in PI. 28. fig. 31) possibly Golgi apparatus. In most cases these large granular cells contain fewer yolk-spheres than the flagellated cells and about the same number of mitochondria. 5. The smaller inner granular amoeboid cells (or mesenchyme) are apparently derived by an inward immigration from the flagellated cells, and are distinguished by the fact that they are crammed with mitochondrial granules and possess very little, if any, yolk. In Grantia compressa the amphiblastula, at the period drawn in Pl. 28. fig. 28, generally contains only two or three such cells (Pl. 23. tig. 30). University College, London, October 30, 1919. 29. BIBLIOGRAPHY. 1. Connuiy, E. G.—*'The Orientation and Cell-Lineage of the Ascidian Kee.” Journ. Acad. Philadelphia, vol. xiii., 1905. 2. Denpy, A.‘ Gametogenesis of Grantia compressa.” Quart. Journ. of Micro. Science, vol. Ix., 1915. 3. —— ‘The Pseudo-gastrula Stage in the Development of Caleareous Sponges.” Proceed. Roy. Soc. Victoria (Australia), 1898. 4. GatenBy, J. Bronté.—‘* The Modern Technique of Cytology.” Quart. Journ. cf Micro. Science, no. 63, 1919. “ The Identification of Intracellular Bodies.” Journ. Royal Mier. Society, 1919. 6. Gorton, W.—*‘ Zur Kenntnis der Spermatogenese bei den Poriferen und Coelenteraten nebst Bemerkungen iiber die Ovogenese der Ersteren.” Zeit. f. wiss. Zool., Bd. Ixxvi. 7. Hanoxet, E.—‘ Ueber die sexuelle Fortpflanzung und das natiirliche System der Schwiimme.” Jenaische Zeit. f£. Med. u. Naturwiss., Bd. vi. 8. Hixscatter, Jan.— Ueber Plasmastrukturen in den Tunicaten-, Spon- vien-, und Protozoenzellen.” Anat. Anz. vol. xlvii., 1914-15. 8a. JORGENSEN, Max.—‘‘Beitrige zur Kenntnis Hibildung Reifung, Be- fruchtung, und Furchung bei Schwiimmen (Syconen).” Arch. fiir Zellfor., Bd. iv., 1910. 9. Mincuty, A.—Lankester’s ‘ Treatise on Zoology, Part IT. LINN. JOURN.—ZOOLOGY, VOL. XXXIV. 22 294 DR. J. B. GATENBY ON THE GERM-CELLS AND 10. Maas, O.— Ueber Reifung und Befruchtung bei Spongien.” Anz., Bd. xvi., 1900. 11. MacBrivg, BE. W.—‘ Text-Book of Embryology, Part I.” London, 1914. 12. Morean, T. H.—‘ Regeneration.’ Anat. Macmillan, Maemillan Company, N.Y., 1901. 13. Ropinson, M.—‘*The Division of the Collar-Cells of the Calcarea Heterocela.” Quart. Journ. of Micro. Science, vol. lvii., 1920. 14. Gatensy, J. Bronriéi.—* The Gametogenesis and Early Development of Limnea stagnalis, ete.” Quart. Journ. of Micro. Science, vol. Ixiii., OM: DESCRIPTION OF THE PLATES. EXPLANATION OF LETTERING. al,a2,a3=stages in passage of collar-cell into mesoglea. AC=acrosome. AS=archoplasmic sphere. BM=basement - membrane of collar epithelium, C=centrosome. COL=collar of collar-cell. ECT=ectoplasmic zone of oocyte or ege. END=endoplasmic zone of oocyte. FC=future histocytes or collar-cells of larva. FL= flagellum. FPN =female pronucleus. GA=Golgi apparatus. GAO=Golei apparatus of oocyte. GAM=eranular (mitochondria) amcebo- cyte. GC =eranular cells of larva. GCAV=gastral cavity of sponge. GOX= future granular (and lower cells?) of larva. GE=gastral or collar-cell epithelium. GX=large granules possibly of Golgi nature. IT=inner thickening of collar. LGC=lower granular cells of larva, as distinguished from inner granu- Jar cells. M=mitochondrium, M &NG=mitochondria and nutrient (yolk) oranules. MAM=macromitosome (‘‘ Nebenkern ”) or middle-piece of sperm. MES=mesogleal cells. MPN=male pronucleus. N=nucleus. NCH=nucleus of collar-cell (choano- cyte). NK=drawn-out “neck” of collar-cell formed of hyaline protoplasm. NSC= nurse-cells associated with sper- matogenesis. OG=outer granules of flagellated cell (possibly yolk of some sort), OGA=oogonium (young oocyte) formed tn situ from collar-cell. OT = outer thickening of collar. OTE = oocyte. PB= polar body. PC=first stages of metamorphosis of collar-cell into germ-cell. PE=thickening of protoplasm on peri- phery of collar-cell, inside the collar. S=sperm. SCC=sperm-carrying cell (metamor- phosed collar-cell). SN =sperm-nucleus. SPA=space around, forming spermato- zoa or nurse-cavity. SPT=spermatids. VAC=vacuolated region of cytoplasm caused by entry of sperm and flow of material to the region of entry. «=supposed remains of macromito- some (MAM). Y=yolk-sphere or nutrient meta- plasmic granule. With regard to fixation and staining, the following abbreviations have been used :— F.W.A.= Flemmine’s fluid without acetic acid; H.W.A.=Hermann’s fluid without acetic EARLY DEVELOPMENT OF GRANTIA COMPRESSA. 295 acid; K.=Kopsch; I.H.=Heidenhain’s Iron Alum Hematoxylin; Ch.K.=Champy-Kull ; Alt.=Altmann’s stain. All figures have been drawn with camera lucida by means of a ; semi-apochromatic oil immersion and compensating eye-pieces. Figs. 2, 3, 5a, 6, 7, 8, 9, 11-22, and 29-31, and 33 were drawn with an 18 compensating eye-piece. Figs. 5 and 23-28 and 382 with an 8 compensating eye-piece. Figs. 1 and 4 were not drawn to scale. Fig. 10 was drawn with a 12 compensating eye-piece. The Plates have been reduced, but a scale is given at the side of Plate 21, which applies to all figs. drawn with an 18 compensating eye-piece. Scale to fig. 10 is alongside the drawing. Scale to all the figs. on Plate 22is given at the left side, excepting to figs. 29-31 and 33, whose scale is found on the right. While Pl. 19, figs. 1,4, and 9 are drawn with the camera lucida, they are a combination, of both Kopsch, Champy-Kull, and Iron Alum Hematoxylin preparations. The same applies to Pl. 20. figs. 13, 14, and Pl. 21. figs. 18, 22. Every care has been taken, however, to see that the strictest accuracy has been preserved, corresponding stages in Champy-Kull or F.W.A. and I.H. being united with those in Kopsch preparations. All the other figures are drawn from one preparation only. In each case the arrow points to the nearest gastral cavity (compare Pl. 19. fig. 5 to illustrate this). PLATE 19, Fig. 1A & B. Two collar-cells or choanocytes drawn and combined from two preparations, one by Kopsch’s method (fig. 2), the other by chrome-osmium and Iron Hematoxylin. The smaller granules are mitochondria, the larger yolk-spheres (reserve granules). In fig. 1 A, the nucleus lies on the outer part of the cell, in fig, 1 B in the middle; in each the Golgi apparatus is shown at GA; in fig. 1 A the flagellum comes out of the latter, in fig. 1 B it passes through or to one side of the Golgi apparatus, and reaches to the ceutrosome, which lies on the nucleus. The collar in fig. 1 B is partly retracted and shows an outer (OT) and an inner thickening (IT); in fig. L A the collar is stretched out and the thickenings are not so marked. On the periphery of the cell, at PH, is a denser region of the protoplam. K. and I.H. Fig. 2. Row of choanocytes and a young vocyte by Kopsch’s method. Fig. 3. Same, fixed in Hermann-without-acetic acid and stained in Iron Hematoxylin. Fig. 4. Choanocyte by Kopsch’s method, to show characteristic neck to cell (NIX). This consists of hyaline protoplasm. The Golgi apparatus (GA) is very clearly marked, and consists of archoplasm (AS) surrounded by a peripheral layer of darkly staining matter (GA). Fig. 5. Part of a hermaphrodite sponge which contained ripe oocytes and ripening spermatids. The latter (SPT) lie in a special cavity (SPA), lined by uon- flagellate cells (NSC) without collars. These cells are probably formed in the mesoglea. At GCAYV is the gastral cavity lined by choanocytes (GE), and at OTHE is the oocyte showing ectoplasmic (ECT) and endoplasmic (END) regions ; this oocyte is just about to be fertilized (S). Note the tails (axial filaments) of the spermatids. Fig.5a. A group of three spermatids drawn at a high power showing nucleus with karyo- some, the centrosome, and the axial filament. The mitochondrial substance was washed away. Cajal. . 6, Spermatozoon drawn from an example which has passed into a collar-cell as in fig. 9,8. The tail has been added from observation of ripening spermatids, and this figure is thought to represent the ripe unchanged sperm of Grantia compressa. At AC is the acrosome. Ch.K. (See footnote on page 279.) Fj 8 296 DR. J. B. GATENBY ON THE GERM-CELLS AND Fig, 7. Ordinary amceboid cell of the mesoglea of Girantia; it lies in the mesogleal fibrous tissue (MES). At M are its mitochondria, There is a distinct ectoplasm. Ch.K,, 1-H. Fig. 8 An oogonium of the mesoglea; the arrows point to the nearest gastral cavities. Ch.K. Fig. 9. Full-grown oocyte of Grantia with nearest accompanying choanocytes, in one of which (at S) a sperm has entered. This oocyte shows the typical endoplasmic region with the granular inclusions and the ectoplasmic part; on the left is a pseudopodium. At M the mitochondria, and at Y the yolk-granules, are evenly distributed; no animal or vegetative pole could be identified by reason of unequal distribution of the cell-granules. Ch.K., L.H., & K. PLATE 20. Fig. 10. Oocyte at lower magnification to preceding, showing large sperm just after entry. The cell-granules tend to approach the region of entry of the sperm. Figs. 11 & 14. Gastral side of oocyte showing sperm-carrying cell lying upon the former ; in both sperm-carrying cells the nucleus is flattened on one side owing to the pressure of the contained sperm; in fig. 11 the sperm will enter middle-piece forwards, in fig. 14 middle-piece backwards. The ectoplasm in the region of entry of the sperm is interrupted. In fig. 11 yolk is not shown. Ch.K. Fig. 12. Sperm just entering, acrosome first. Shows protoplasmic continuity between sperm-carrying cell and oocyte. Ch.K. Fig. 18. Later stage, polar body at PB, and female pronucleus growing (FPN). Sperm- path marked by pale chromophobe area (VAC). Note large size of mitochondria (M). Macromitosome of sperm disappeared, but may be represented by granules Bybee CHR Lee ica. Fig. 14. See description to fig. 11. Ch.K., 1-H, & K. Fig. 15. Flattened sperm-bearing cell, which was compressed by egg, and so looks abnor- mally large. Its nucleus is of the open type. Ch.K. Fig, 16. Sperm just after entry ; its cell-wall has disappeared, macromitosome still complete, and sperm-cytoplasm furred at edges but still distinguishable from egg-cyto- plasm by its greater chromophility and stringiness. Acrosome pale. Nucleus neatly spherical and has grown. Sperm-carrying cell at SCC, with open nucleus. Fig. 16. Stage in fertilization after that in fig. 12. The sperm cell-wall has disappeared and its stringy cytoplasm is still distinct from that of the egg. The acrosome (AC) is becoming faint, Ch.K. Fig. 17. Later stage just after breaking up of sperm middle-piece or macromitosome (MAM). This preparation was fixed by Champy-Kull and stained by Heidenhain’s [ron Alum Hematoxylin. Note that nuclei of sperm-carrying cell (SCC) and those of choanocytes (NCH) stain alike. Ch.K., LH. Puats 21. Fig. 18, The two pronuclei side by side prior to last stage of fertilization. Note even dis- posal of inclusions (M & Y). The sperm-carrying cell (SCC) is degenerating in this case. Yolk added from another preparation at a similar stage. Ch.K.&K. Fig. 19. Two-cell stage, showing even disposal of mitochondria. Upper blastomere not cut across middle. Ectoplasm evenly divided. Ch.K., IH. Fie. 20, Four-cell stage, showing continued subequal distribution of the mitochondria and of ectoplasm. H.W.A., IH. Gatenby. JOURN. LINN. Soc. ZOoL VoL. XXXIV,PL.19 ECT. MES. J.B.G.del Huth,Lond GRANTIA (SYCON) COMPRESSA. Gatenby. JOURN.LINN. Soc. Zoot VoL.XXXIV, PL.20. J.B.G.del. Huth, London. GRANTIA(SYCON) COMPRESSA, Gatenby. Linn. Soc. Journ, ZOOL. VoL. AXXIV.PL. 21. J.B.G.del. Huth, London GRANTIA(SYCON) COMPRESSA. JOURN.LINN. Soc. ZooL. VoL. XXXIV. PL. 22. Gatenby. Huth, London J.B.G. del. GRANTIA(SYCON) COMPRESSA. JouRBN.LINN.Soc.Z 00L VoL. XXXIV, PL.23. Gatenby. J.B.G.del GRANTIA(SYCON) COMPRESSA. EARLY DEVELOPMENT OF GRANTIA COMPRESSA. 297 Fig. 21. Blastula just before differentiation of flagellated hemisphere (see Pl. 22. fig. 27). The fine mitochondria, still visible but staining less well, lie around the nuclei. The yolk is not shown. At GAM isa granular amceboid cell. Jndging from the contour of this blastula, the future flagellated hemisphere will differentiate on the right, and away from the nearest collar-cell. Ch.K., IH. Fig. 22. Younger blastula showing even segregation in each blastomere of yolk-spheres and mitochondria. Phe ectoplasm is slightly larger in quantity on the right than on the left. The yolk-spheres have been added from another preparation prepared by Kopsch’s method. Ch.K., L.., & K. PLATE 22, Figs. 23, 25, & 26. Parts of flagellated gastral epithelium, showing at PC and OGA two stages in metamorphosis of collar-cell into oogonium. Ch.K., L.A. Fig. 24. Part of flagellated gastral epithelium showing immigration inwards of collar-cells, al, a2, & a3, into the mesoglea. Ch.Is., IH. Fig, 27. Differentiating amphiblastula, showing future flagellated cells (FC) and future lower and granular cells (G@CX). At OTE is a young oocyte in growth-stage. PLATE 23. Figs. 28 & 52. Two amphiblastula larvae, for description of which see page 266. Ch.K., LH.. and Ch.K., Alt. Fig. 29. Maternal nutrient capsule-cell (LGC in fig. 28), showing irregular and numerous mitochondria (M). Ch.K., LH. Fig. 30. Inner amoeboid granular cell (GHM in fig. 28) showing dense numerous mito- chondria. Fig. 31. Granular cell (GC in fig. 28) showing scattered yolk-spheres (Y), mitochondria (M), and what are possibly Golgi elements at GX. Ch. K., I... Fig. 32. (See legend to fig. 28.) Fig. 33. Flagellated cell or histocyte showing four cytoplasmic elements. Mitochondria (M), yolk (Y), Golgi apparatus (GA ?), and outer “yolk” granules at OG, ne pallebe t cu ra qb cf enisdee yn aN, iia 4 ae mesogloeal phagocyte *. The evidence is that the ‘‘ masses of granules’ are the last products of digestion, and within a phagocyte, or group of phagocytes, may be taken as showing where the combat with an intruding organism has been victorious and complete. This speculation has partly originated in the attempt to explain the cell drawn in fig. D, alluded to at the meeting of the Linnean Society on Dec. 11th. With a superficial resemblance to a pylocyte, it differs in its thick body and large vesicular nucleus ; in each of these particulars it resembles Dendy’s fig. 81. Also the plane of its crescent is radial instead of tangential to the chamber into which it projects, instead of framing an aperture in the wall. The appearance of the cell, either in fig. 1 D or in Dendy’s fig. 81, is unquestionably that of a gonocyte. In describing my feeding experiments, and confirming my previous observations that the gonocytes feed on the basal spherules of the collar-cells, which are stores of digested food (1892 0, p. 474; quoted 1893, p. 220; rediscovered 1914, pp. 328, 344), I wrote :—‘ A large number of gonocytes are in contact with collar-cells which contain plentiful carmine ; in only two of them I found carmine-grains, and it is tempting to deduce that vacuoles and undigested food do not pass into the gonocyte ” (1895, p. 380). Cotte (1908, p. 448) uses similar language :—‘‘ Les coupes de Sycandra montrent quelques grains de carmin ou de charbon englobés par des amibocytes ; c’est la, on peut le dire, une véritable exception.” But our experiments show only that the amcebocytes of Calcaronea do not ingest particles which reach the choanocytes from the water of the chamber. This negatives in no way their power to engulf foreign bodies which they may find in the mesoglea, And beside the classical experiments of Metschnikoff (1879), Cotte has shown that they have this power in Reniera (p. 448) and Spongelia (p. 455); while Dendy (1914) has given a most interesting description of phagocytosis by odcytes and other amcehocytes in Grantia compressa. I interpet fig. 1D as an amoebocyte which has digested some foreign substance encountered in the mesoglw@a, and is dragging the excreta to be cast out into the efferent water of the flagellate chamber?. And Dendy’s fig. 81 indicates that the substance partially digested there, wholly digested in my fig. 1 D, is a lightly-staining sphere of 6 diameter (with nine spots, * [Since this paper was read I have seen van Trigt’s interesting work (1919, p. 162), in which he records similar feces formed by the amceboid cells of Spongilla and by the walls of the efferent canals. Van Trigt considers that this sponge feeds upon and digests the cells of the green alga which give its colour. ] t (Cf. van Trigt, 2. ¢.] 322 DR. G. P. BIDDER: NOTES ON THE staining more darkly, in fig. 81). The indications seem to me that these figures, Dendy’s fig. 85, and many others in his very interesting plates—all record stages in the war against Synerypta, which appears to be carried on successfully i in Grantia aithion by the gonocytes themselves, or by other amcebocytes which have this function * Most interesting in this connection is Poléjaeff’s statement (1854, p. 72) as to “spermospores” in Verongia :—‘‘ When ripe they recall so vividly the corresponding formations in Sycon raphanus (with the sole distinction that while in this latter instance the nucleus of the covering cell in quite ripe spermospores is in most cases indistinct, in Verongia I tind quite empty capsules, nevertheless, provided with it).” All the “covering cells” represented in Verongia (pl. 10) contain only masses of granules; and similar masses are in Carteriospongia in an endothelial chamber, as described by Schulze (and verified by Poléjaeft) for Oscarella, Aplysilla, etc. I suggest that the nucleus of the ‘ covering cell” is indistinct where the contained Synerypta has killed the amcebocyte which ingested it, and that those in Verongia are the nuclei of amcebocytes which —as shown by their granular contents (as in my fig. 4)—have destroyed the Syncrypta, themselves uninjured. An endothelial capsule occursin Carterio- spongia, instead of a covering cell as in Verongia, because the heap of granules is 40 ~ x 15 w instead of 9m in diameter, showing that the digested substance was too large for one phagocyte to contain. ~ Such conclusions merely assume a process wholly comparable with the destruction of Oscillatoria by the amcebocytes of Spongelia which is described by Cotte (l.c. p. 455). As Vosmaer said (1887, p. 412) :—‘The structures which many spong- ologists have set forth more or less positively as spermatozoids may have Been so or may not; in many cases it remains doubtful.’ Norn H.—Origin of Sponges. If the hypothesis be adopted that the granules of porocytes attract spermatozoa, a possible history of the origin of sponges is suggested. If we construct our imaginary ancestral sponge on the lines of the Sycon amphiblastula, illumined by Protospongia, we find a colony of flagellate collared cells whose function is to obtain nourishment, surrounded by their associate archeeocytes, which receive the digested nourishment and alike * Tam unaware how far, by those who have worked at these problems, phagocytosis and conjugation are considered cognate or opposite. My friend Miss E. R. Saunders tells me that in the revival of a plasmodium of Myxomycetes from the sclerotial condition, as each sclerotial cyst is enveloped by the advancing plasmodium, it is uncertain for a moment whether a vacuole will be formed round its protoplasm and this be digested, or whether it willzevive rapidly enough to burst to the moisture, and unite alive with the syncytium. This alternative absorption as food, or as a living partner, suggests that conjugation may arise from attempted phagocytosis. PHYSIOLOGY OF SPONGES. 323 excrete in a manner possibly protective for the colony * and are charged with its reproduction. At this stage we may conceive the secretion of an attractive odour to have become permanently associated with their nitrogenous excretion, with the advantage of more frequent reinvigoration of the stock by conjugation with locomotor units from other colonies. Development led to differentiation of purely reproductive cells from those charged with excretion, the reproductive cells being placed in greater safety by a covering of the excretory cells. The latter would still attract the locomotor units, which, when engulfed, they would pass on for nuclear union with the gonocytes. This absorption from the exterior and transmission to the interior would be simplified (as the reverse process has probably been in nephridial tubes) to perforation (¢f. 1892, p. 182). Misdirected perforation, placing the excretory surface of the sponge in communication with the flayellar surface instead of with an odcyte, would sporadically occur, and when so occurring would lead to a sudden increase in the hydraulic efficiency of the flagella as a food-catching apparatus, supplies being thus brought to centrally placed choanocytes which otherwise could only be reached by water which had passed over their neighbours. Such misdirected perforation would therefore be perpetuated by the increased prosperity of those colonies in which it happened, their form being now in essence that of the rudimentary Olynthus ; from this upwards “all transition from more to less primitive canal systems exhibits an increase in the ratio” that determines the mechanical efficiency of the canal system (1888 ; 1895, p. 29). The cercids are fragmented archzeocytes (see Minchin 1897, p. 500) which (ante, p. 300) break out into the surrounding water and effect conjugation with the archeoeytes of other colonies. My excuse for printing such a fanciful story is the extreme difficulty of forming any conjecture as to possible value to the organism of the intermediate stages in the evolution of a hydraulic system supplied by intracellular pores (cf. MacBride, 1918, p. 52). If the pores were multicellular, we might conceive them as originating as grooves or furrows in the flagellate surface, with the advantage in every stage of bringing better supply to the more central flagella. Such an evolution would fit in with many embryological observations, and there is so much evidence that sponges are polyphyletic that it is not impossible that both developments have occurred}. But in both Calcinea and Calcaronea we know that the pores are intracellular, * The nitrogenous excretion of the archzeocytes may be beneficial to the colony by beiug either repulsive to foes or attractive to friends. But, in view of the known defzecation by collared cells, these probably get rid of their own nitrogenous poisons, and I do not now believe that in this sense either the primitive archzeocytes or the specialised porocytes excrete for the whole colony. Loisel’s criticism on this point (1898, p. 203) is quite just. + In the Donatiidie and the Hexactinellida there is no evidence of the presence of pore- cells, and it is doubtful if in either of these groups there is any sexual reproduction. J24 DR. G. P. BIDDER: NOTES ON THE Minchin (1900, p. 49) suggests that the pore-cells may primarily be nutritively phagocytic in function, and therefore have come to surround pre-existing afferent pores. The hypothesis does not seem a priori per- suasive; and it does not fit well with the facts, since we have no evidence of pre-existing pores, nor, in these sponges, of any prosopyles which do not perforate porocytes *. [ Posrscrirr. Nore F.—Odgenesis in Sycon. Dendy’s derivation of the odcytes from collar-cells (1914) in Grantia compressa is not in harmony with my own views; but on re-examination of my preparations I cannot claim that they disprove it. His igs. 2, 38, 12, 13, 14 (as also 62, 66, and 68) are from a sponge which had been a week in the aquarium circulation. It appears possible, therefore, that he observed the early stage of the pathological changes in the flagellate chambers, of which I have recorded the condition after a month in the aquarium (1895, p. 27). It is remarkable, however, that in my pathological specimen, though the highly gelatinous mesogloea contains many cells of the size and appearance of collar-cells, there are scarcely any recognisable gonocytes or cells which resemble gonocytes in size or character. I agree with Dendy that the ova are fonnd free in the cavity of the chambers, strange as this appears, and that such is the interpretation of the “larger nucleate cells, possibly Protozoa, partly enveloped by the distended collars, sometimes more than one cell converging on them,” which I described in 1895 (p. 31), and which are well indicated in his figs. 11 and 74 (in the latter case the ovum is called a “primary spermatogonium”). I found in sections of S. raphanus unmistakable ova in the lumen of the flagellate chamber and one just passing through the wall, and I sketched at Plymouth (April 1920) such an ovum alive in the cavity of a flagellate chamber of G. compressa, seated with pseudopodia stretching over the free ends of half- a-dozen collar-cells, in outline ludicrously Jike an octopus. Between the vesicular nucleus and the free ovoid surface was a comet-shaped streak or are of minute green fragments, suggesting that it had used its freedom to feed on and digest an alga, thus supporting Dendy’s suggestion on pp. 323, 335 (1914). Nore G.—WMigration of the Nucleus in Collar-cells. In fig. 1 C the nucleus is visible in two cells, and is seen to have moved downwards to allow of the escape of the faeces through the aperture in the * Cotte (1903, p. 434) considers that there is a general power of cell-perforation in the pinacocytes of Renrera, and quotes Topsent for similar perforations of the contractile cells of this species and of Cliona. PHYSIOLOGY OF SPONGES. 325 iris. Such displacement must also occur when any large mass of food is ingested, and Synerypta spheres commonly do not pass the nucleus (¢/. fig. C). Hence, Hammer (1908) found the nucleus often not distal, and his fig. 85 shows such a nucleus, with the agglomeration of ingested foreign bodies which have forced it down. He correctly quotes me as having watched the living nucleus moving in the protoplasm, but I found its common situation apical. The fact that his observation of a usually basal position in life was contradicted by his stained sections indicates that in the living cell a food-vacuole was mistaken for the nucleus—in like manner as I described the nucleus at first as a vacuole (1895, p. 18).] REFERENCES to works cited in the three Papers by G. P. Bidder. 1849. Carrer, H. J. Aun. & Mag. Nat. Hist. vol. iv. p. 98. 1857. BowrrsBanx, J.S. Reports of the British Association for 1856 (p. 15 of reprint). 1858. e Reports of the British Association for 1857, p.121. (Quoted also in 1864.) 1864. “6 . ‘Monograph of the British Spongiade,’ vol. i. p. 121. 1872. Haxrcxst, E. ‘Die Kalkschwimme.’ 1879. Merscunrkorr, E. Zeits. fiir wiss, Zool. xxxii. p. 349. 1879 6. Scuuuze, F. E. Zeits. fiir wiss. Zool. xxxiii. p. 25. 1882. PouésaErr, M. Sitzungsb. d. math.-naturw. Cl. d. k. Akad. d. Wiss. Wien, Ixxxvi. p. 277. 1883 *. ¥) ‘Challenger’ Calcarea, p. 33. 1884. ) ay 5 Keratosa, p. 72. 1887*. Vosmarr,G.C.J. Bronn’s ‘ Porifera,’ p. 412. 1888. Bripper, G. P. Proc. Phil. Soc. Cambridge, vol. vi. pt. 4. 1891. Dernpy, A. Quart. Journ. Mier. Sci. vol. xxxii. fig. 26. 1891 b. : a Trans. Roy. Soc. Victoria, vol. iii. 1891 ¢c. Brppsr, G. P. Quart. Journ. Mier. Sci. vol. xxxii. p. 627 (p. 3 of reprint). 1892. Mincuin, E. A. Zool. Anz., No. 391, vol. xv. p. 182. 1892 6. BrmpeEr, G, P. Roy. Soc. Proe. vol. li. p, 474. 1892 c. Mincury, E. A. Quart. Journ. Mier. Sci. vol. xxxiii. p. 477. 1892 d. BrppER, G. P. Roy. Soe. Proc.‘vol. lii. p, 184. 1893. Drnpy, A. Quart. Journ. Mier. Sci. vol. xxxv. p. 159. 1894. Grernwoop, M., & Journ. Physiology, vol. xvi. p. 449. Saunpers, E. R. 1895. Brpper, G. P. Quart. Journ. Mier. Sci. vol. xxxviii. p. 9. 1897. Muincuin, KE. A. Quart. Journ. Mier. Sci. vol. xl. p. 487, p. 499, etc. 1898. LorsEL, G. Journ. Anat. et Physiol. vol. xxxiv. p. 203. 1898 6. Vosmaer,G.C.J.,& PEKELHARING,C.A. Verhand. Konink. Akad. Wet. Amsterdam, D. vi. No. 3, p. 10. * These are the dates on the title-pages. I have been given the honour of superintending the publication of Vosmaer’s Bibliography of Sponge Literature, and in this these works are entered under the years in which they were actually issued—the ‘Challenger’ Calcarea in 1884 (note date of its preface) and Bronn in 1886. 326 NOTES ON THE PHYSIOLOGY OF SPONGES. 1898 c, Brippmr, G. P. Roy. Soe. Proc. vol. lxiv. p. 61. 1900. Mrncutn, E. A. “Sponges.” Lankester’s Treatise on Zoology, pp. 29, 30. 1903. Corrs, J. Bull. Sci. de la France et la Belgique, tome xxxviil. p. 420. 1903 6. Urpan, F. Verh. Ges. deutsch. Naturf. lxxiv. ii. 1, p. 159 (seen only in abstracts). 1906. Harroe, M. Cambridge Nat. Hist. vol. i. p. 127. 1907. Awnanpdatp, N. Records of the Indian Museum, vol. i. Part 4, No. 27, p. 387. 1908. Muincury, EH. A. Quart. Journ. Mier. Sci. vol. lii. pp. 325, 354, fig. 71. 1908 6. Hamumr, E. Archiv fiir Biontologie, Berlin, Bd. ii. p. 328. 1910. Roserrson, M.,& Quart. Journ. Micr. Sci. vol. lx. pp. 689, 640, figs. 31-35, 48. Mincuin, E. A. 1910 6. Parxer, G. H. Journ, Exp. Zool. vol. viii. No. 1, p. 784. 1910 c. Urngan, F. Internat. Rey. Ges. Hydrobiol. vol. iii. p. 42. 1913. Drnpy, A. Linn. Soc. Trans. Ser, 11. Zool. vol. xvi. Part 1, pp. 4-6. 1913 6. Drnpy, A., & Proe. Zool. Soc. London, p. 718. Row, W. HH. 1914. Drnpy, A. Quart. Journ. Mier. Sci. vol. Ix. p. 313. 19146. MacBripr, EK. W. Text-Book of Embryology, p. 52. 1919. Cxurcu, A. H. Journ. Botany, Supplement II. p. 5. 1919 6. van Trier, H. “A Contribution to the Physiology of the Fresh-water Sponges.” Brill, Leiden. 1920. Orvron, J. H. Journ. Mar. Biol. Assoc. vol. xii. pp. 340, 341. ON THE OOCYTES OF THE GALL-FLY NEUROTERUS. 327 On certain Nuclear Phenomena in the Oocytes of the Gall-fly Newroterus. By Lanorxor T. Hoary, B.A., B.Sc. (formerly Frank Smart Prizeman in Zoology, Cambridge). (Communicated by Dr. EH. Drassie, F.L.S.) (With two Text-figures. ) [Read 6th November, 1919. ] Despite the existence of a very considerable literature dealing with insect cytology, there is still the most meagre knowledge of the history of the nucleus in the oocyte, particularly in the case of the Hymenoptera, though this group has attracted many investigations of a cytological nature on account of the frequent occurrence of parthenogenesis among its members. Recent work by Doncaster on the process of maturation, and by Hegner, Martin, Silvestri, and Gatenby on the earlier stages of oogenesis in certain of the Hymenoptera parasitica, have revealed somewhat unique phenomena which call for further elucidation. The present communication contains some observations of a preliminary nature which shed light upon the character of the mitotic figures which occur during the formation of polar bodies in this group. Gametogenesis in the Gall-fly Neuroterus has already been made the subject of three memoirs by Prof. Doncaster, to whom I am indebted for the materia] of the present note. In describing the maturation of the ova, attention was directed to the fact that an unique type of ookinesis occurs, in which, to quote Doncaster, “the first division takes place by the drawing out of threads (probably double) on each side of the nucleus ; the reticulum becomes absorbed in these threads, which form two groups of parallel chromosomes on a spindle. These chromosomes then divide, probably longitudinally, giving rise to the group which forms the egg and three polar groups of chromosomes.” An apparently similar type of maturation division was described nearly twenty years ago by Henking in the Cynipid genus Rhodites. Quite recently Hegner and others mentioned above have made cursory references to the existence of a precocious maturation spindle of a normal type in certain species of Chalcids, Braconids, and Cynipids ; and, according to these authors, instead of preceeding to pass into anaphase, the chromosomes, in the form of end-to-end pairs, condense into a compact nucleus like that found in the full-grown egg of Rhodites and Neuroterus. It was therefore thought desirable to attempt to correlate these phenomena if possible. Oocytes of the agamic generation were chosen for study, the larve of the spring brood being entirely females. A few pupze of Newroterus lenticularis 328 MR. L. T. HOGBEN ON CERTAIN NUCLEAR PHENOMENA were collected by Prof. Doncaster, but as it was difficult at the time to procure many, the closely-allied species N. nwmismatis was also studied, late pupse and egg-masses of newly-hatched flies being preserved in addition to the November material. For the latter Flemming was used, egg-masses being preserved in Bouin or Gilson’s fluid. Sections were cut 4-6 m. in thickness and stained with Heidenhain’s iron hematoxylin and eosin. The preparations were examined with a Zeiss *2 immersion and apochrom. ocular No. 12. The Somatic Mitoses. Fifty counts were made of the dividing nuclei of the developing wing, nervous system, and follicular epithelium of Newroterus numismatis. The average was 19-2, 26 giving 20 and the rest 18 or 19. Some of the figures Fie. 1. 1-6. Pairing of the chromosomes and formation of the abortive maturation spindle in Neuroterus numismatis. giving 20 were very clear, and there is no doubt that the diploid number for this species is the same as for JW. lenticularis, viz. 20. Very clear figures of abnormal mitoses were found in the hypoderm cells of larvee of WV. lenticularis. Doneaster has figured in his second paper the anaphase of one of the giant cells which lie below the epidermis showing at least 50 chromosomes at each pole. In several cases of cells of ‘the IN THE OOCYTES OF THE GALL-FLY NEUROTERUS. 329 hypoderm itself I noted the occurrence of 30 chromosomes at each pole. In one case this number (30) was quite unmistakable. In three nuclear divisions in the same region the haploid number occurred ; and it is perhaps of some significance that the two numbers are complementary. It is thus possible that the occurrence of the haploid figure may be due to a difference of polarity—pathological maybe—in the dividing nuclei. The Development of the Oocyte. As there are two kinds of ova produced by the agamic generation in N. lenticularis (viz. those which undergo maturation and give rise to males, and others which do not form polar bodies and are female producing), it appeared possible that differences might occur in different individuals during the synapsis stages. No points ef contrast between the very young oocytes of different agamic females could be detected either in Veuroterus lenticularis or in WV. numismatis, although the ovarioles of more than twenty individuals were examined. Synapsis occurs in the upper part of egg-tubes of ovaries in which the former have just become differentiated. The cells of this region are arranged in groups of about 8 to 16, and synizesis nuclei can be seen before it is possible to distinguish nurse-cells from oocytes. Subsequently the chromatin material of certain cells, which are to become nurse-cells, under- goes fragmentation into very numerous granules. Their nuclei later exhibit one or two large karyosomes of irregular shape: these at a late stage disintegrate and are ejected into the cytoplasm. During the growth period it is not possible to detect definite chromosomes in the oocyte nucleus, the chromatin content of which greatly loses its staining eapacity. One or two karyosomes and a plasmosome are found often lying on the outer side of the nucleus. As in Andricus, there is no oosoma at the posterior end of the egg like that described by Hezner in Diastrophus or the body referred to by Weissmann in Phodites. There are no secondary nuclei. The Maturation Spindle. Later phases were studied exclusively in V. numismatis. The oocytes of the March pups no longer display the diminished staining capacity de- seribed above. The diploid number of chromosomes is seen to be dispersed throughout the nucleus, and the plasmosome is undergoing disintegration. If, as is quite likely, there is a preliminary syndesis that synchronises with the synaptic contraction in the young oocyte, this stage may be interpreted as that of diakinesis. It is followed by an end-to-end pairing of the chromosomes. In sections of the egg-mass taken from newly-hatched flies at the beginning of April the chromosomes of the oocyte nucleus exhibited 330 MR. L. T. HOGBEN ON CERTAIN NUCLEAR PHENOMENA a considerably increased staining capacity, at the same time assuming a rod- like appearance. Later they approximate, each with one tip in contact with that of another (fig. 1, 3). It was possible to count the pairs so formed, ten eee Ww WL WI Fie. 2. Illustrating maturation phenomena in Neuroterus. a, b,c, d. The maturation figures (after Doncaster). e, f, 9, h, t. Suggested interpetation of the maturation prophases IN THE OOCYTES OF THE GALL-FLY NEUROTERUS. dol being present. At first these chromosome pairs are scattered irregularly ; but subsequently they take up a definite arrangement, which is in all respects like the metaphase of a mitotic figure. They place themselves in a parallel series around a central axis with the point of contact of each conjugant indicated by a constriction in the equatorial plane. The chromosome com- plex so formed then condenses to form the oval compact nucleus, which is described by Doncaster in the eggs after they have been deposited @Gigol, 4,15, 6): Hegner has given full details of similar phenomena, as already mentioned, in Copidosoma ; and on most points his account agrees closely with that given above for Neuroterus. In the former, according to Hegner, “ the spireme becomes more and more open, and finally breaks up into thin chromosomes of irregular shape. These chromosomes then become shorter and thicker, and appear to unite near their ends. At first the pairs are scattered about in the nucleus, but they soon straighten out and become arranged in a parallel series with their points of union lying in the equator. Spindle fibres could be seen, but apparently no centrosomes nor asters are present.... The mitotic figure then passes through the stage of condensation.... The chromosomes gradually get closer together and become shorter and thicker. Where these ends meet at the equator a ridge appears which causes the complex to resemble a Maltese cross. Soon the spaces between the chromosomes are entirely obliterated and a homogeneous mass of chromatin results.” Elsewhere Hegner refers to the arrangement of the chromosome pairs on a ‘ Maturation spindle” in Apanteles, and in Andricus, the Oak-knot Gall-fly closely allied to Newroterus. In Neuroterus the junction of the conjugating chromosomes is not indicated, as in these, by a thickening, but a constriction. General Remarks. In a postscript to his final memoir on gametogenesis in Neuwroterus, added on receipt of Hegner’s ‘ Protoplasmic Differentiation in the Oocytes of certain Hymenoptera,’ Doncaster expressed the belief that the top-shaped nucleus, described by him in certain ova prior to maturation, is a stage in the disentangling of a compact nucleus formed in some such manner as in Copidosoma. This has been shown to be the case. The polar mitosis described by Doncaster represents the type of figure that might be predicted if the preparation for the homotypic split begins during the first division and extends from the junction of the conjugants to their free distal ends. The double threads drawn out in the formation of the first polar body from the reticulum so formed would be equivalent to univalent chromosomes precociously and incompletely split. The formation of the second body then consists presumably in separating the two halves, 332 MR. L. T. HOGBEN ON CERTAIN NUCLEAR PHENOMENA If this interpretation prove to be correct, the separation of the polar chromosome groups in Veuroterus is the completion of a mitotic process that begins in the oocyte before it is laid. The maturation spindle in the late oocyte is a precocious and interrupted metaphase, preceded by an end-to- end pairing of univalents, and only differing from the normal type in consequence of its previous history and in the absence of asters or centro- somes. According to Doncaster, however, there are neither asters nor centrosomes in the polar and segmentation mitoses of the Saw-flies. In this case the solid and top-shaped nuclei of the full-grown egg in Neuwroterus are not nuclei in the strict sense of the word. The only account, as far as | am aware, of a comparable prolongation of the maturation process is that of Histriobdella, where Shearer describes a protracted prophase lasting for a week while the cytoplasm of the ege was stiJ] in its growing phase. It follows that in those Hymenoptera where similar maturation spindles have been described in the late oocyte, irregular methods for the extrusion of polar nuclei should prevail also. This appears to be the case in Martin’s account of Agehiaspis, though details are lacking. Gatenby has described the formation of polar bedies in the Chaleid Trichogramma by what he terms ‘“amitosis.” It is, of course, of immense importance from the standpoint of the chromosome hypothesis to know whether amitosis occurs in germ nuclei ; and if it is accepted that the nucleus of the full-grown egg in the parasitic Hymenoptera is in reality a condensed equatorial plate, an apparent con- striction, simulating amitosis, does not imply an unequal distribution of chromatic elements, seeing that these are definitely orientated lengthwise in the complex. I have no hesitation in saying that the formation of polar bodies in Trichogramma cannot be justifiably contended to detract from the accepted views as regards the persistent individuality of the chromosomes. Mention has been made of Henking’s description of the maturation of the egg in Rhodites, the solid nucleus of which is reminiscent of that of Neuroterus, the separation of the polar chromosomes being also similar. A comparison of Schleip’s figures of the chromosomes in the late oocyte with the same stage in Andricus, as shown by Hegner’s plates, suggests very forcibly that they are in reality doubles and not univalent as Schleip believed*. I must, however, refrain from further comment on this head, seeing that I hope shortly to publish a full account of the chromosome cycle in this form. T have to thank Dr. Doncaster for providing me with material for these observations, as well as for advice. June 1919. * T have since found that such is actually the case, IN THE OOCYTES OF THE GALL-FLY NEUROTERUS. 333 Literature. Doncaster, L.—* Gametogenesis and Sex-determination of the Gall-fly Neuroterus,” I-III. Proc. Roy. Soc. vols. Ixxxii., Ixxxiii., [xxxix., 1910, 1911, 1916. Gatensy, J. Bronré.—* Tho Segreeation of the Germ Cells in Zricho- gramma evanescens.’ Q.J. M.S. 1918. Heener, R. W.—“ Studies on Germ Cells, LV. Protoplasmie Differentia- tion in the Oocytes of certain Hymenoptera.” Journ. Morph. 1915. Heyxine, H.—‘ Untersuchungen iiber die ersten Entwicklungsvorgiinge in den Hiern der Insekten,” III. Zeit. Wiss. Zool. liv., 1892. Mart, F.—“ Zur Entwicklungsgeschichte des polyembryonalen Chaleidiers Ageniaspis.” Zeit. Wiss. Zool. ex., 1914. Scuiere, W.—‘‘ Die Reifung des Hies von Rhodites rose.” Zool. Anz. xxxv., 1909, SHearer, CRessweLL.—‘“‘On the Anatomy of HWistriobdella homari.” Q.J. M.S. 1910. Stivestri, F.—‘ Prime fasi di svillupo del Copidosoma buyssoni Imenottero Calcidide.” Anat. Anz. Ixvil., 1914. Weismann, A.—Beitriige zur Kentniss der ersten Entwicklungsvor im Insektenei. 1882. ange 0 to) LINN. JOURN.—ZOOLOGY, VOL. XXXIV. no Or THE REPRODUCTION OF ASELLUS AQUATICUS. 335 Notes upon the Reproduction of Asellus aquaticus. By Ernest E. Unwiy, M.Sc. (Leeds). (Communicated by the Rey. T. R. R. Steppine, M.A., HRS: Es) (PLaTRs 25, 26.) [Read 5th June, 1919.] INTRODUCTION. Some years ago when beginning an enquiry into the life and structure of the terrestrial Isopoda, I started a general examination of the order; and the fact that Asellus aquaticus was common in all ponds and streams and easily kept in captivity, led me to spend not a little time upon its life and structure. I have kept it more or less under observation since. Many iv-vestigators have paid attention to this fresh-water isopod. Leydig, Schneider, Rosenstadt, Kimus, and more recently Wege and Kanlbersz, have published valuable papers *. Kimus deals exclusively with the struc- ture of the branchial pleopods in a comparative study of the branchize of aquatic isopods. Wege uses Asellus for experiments upon regeneration of appendages. Kaulbersz, in a long and interesting paper, describes a large number of experiments dealing with the behaviour of Asel/us under varying conditions of light and chemical stimuli, as well as some notes upon regeneration, reproduction, and moulting. The notes which follow are a contribution to a fuller knowledge of reproduction in the Lsopoda. REPRODUCTION. There is an intimate relation between the processes of reproduction and the change of cuticle. Both copulation and egg-laying are heralded by the ecdysis of the female cuticle. It will be best, therefore, to describe this in brief. The cuticle is always cast in two pieces: first of all the hind-half and then, after a short interval, the fore-half. This interval has usually been about 24 hours, but I have records varying from 8 hours to 3 days. Both Kaulbersz and Wege confirm the normal interval; the latter writer has compiled an elaborate table in which a day is the usual interval, and he mentions that the normal time is from 12 to 24 hours. The splitting is always between the 4th and 5th thoracic segments. I have examined a large number of cast ‘cuticles and have not, as yet, found an exception to this. * See “ Bibliography,” p. 342. 336 MR. E. E. UNWIN: NOTES UPON THE The old cuticle is split transversely in the thin cuticle joining the terga of the 4th and 5th segments, and then quickly the split extends laterally and ventrally. The creature holding firmly on to the weed with the first 4 pairs of legs sways the hinder part of its body up and down. This action seems to help in the withdrawal of the legs and other appendages from the old cuticle. J have also noticed that the casting of the hind part is assisted by slowly walking forwards a short distance; the friction of the legs of the old cuticle against the weed of the ground helps the withdrawal. I have not been able to detect any other splitting ; each half is a perfect east of the part of the body from which it comes. Asellws does not, as a rule, eat its cast cuticle. This is the usual procedure in the Oniscoidea. Both Wege and Kaulbersz give tables of the interval between two succes- sive moultings, and discuss the effect of various factors, such us food, age, temperature, mutilation, upon this interval. With poor food the intervals average about 24 days; Wege has records varying from 8 to 30 days. The average time under normal conditions is about 3 weeks. The moulting in relation to the sexual processes will be discussed below. The sexes are distinct. Full-grown males are much larger than sexually mature females. With a lens the males can be readily distinguished by the presence of the modified pleopods (second). During the summer some females proclaim their sex by the presence of a brood-pouch containing developing eges or young Aselli. I have records of egg-laying showing considerable variation ; the earliest date on which I have captured a female with eggs in the brood-peuch is January 24th, and if the weather is fairly mild they can be found in this condition as late as the end of November. In the North of England from April to June is the most active period for reproduction. Male. There are three pairs of testes lying in the middle region of the thorax. Each testis is an ovoid body attached by a very short duct to the vas deferens of its side. ‘The two vasa deferentia, fairly thick tubes, extend from about the 3rd or 4th segment to the 7th. Each opens separately by a spout-like aperture, which, protruding from the last thoracic segment, lies justin front of the small rudimentary first pleopods (P]. 25. fig.6). They will discharge in the immediate vicinity of the second pair of pleopods. These appendages (PI. 25. fig. 7) are modified for the purpose of passing the mass of sperinatozoa to the female. The structure can be seen from the figure. It is the endopodite which is specially modified, forming something of a tube. The general shape is triangular, and its attachment to the protopodite allows great freedom of movement. The distal end is knob-like, with a pointed recurved process not far from the end. REPRODUCTION OF ASELLUS AQUATICUS. 337 The fourth pair of walking-legs are accessory organs. They will be described later. Female. The ovaries are paired and lie dorsal to the alimentary canal and hepatic ceca. In an immature specimen they are small and show little sign of eggs; but in a female about to lay eggs, the ovaries are very large and conspicuous, extending the length of the thorax, and have the form of thin-walled bags packed with piicrical egos. The external openings of the oviducts can only be seen at a particular period. When the female has passed through the ecdysis which releases the oostegites, the openings can be seen on the ventral surface at the base of the fifth pair of legs. They are narrow slit-like openings with thickened lips. By means of serial sections through specimens about to lay eggs after being released from the marriage-clasp, I found that these apertures lead into a fay large spherical cavity in which a mass of spermatozoa could be seen. It is clear that these cavities function as receptacnla seminis. ‘They appear to be lined with chitin and communicate with the ovary of the corresponding side (PI. 26. fig. 12). Marriage-clasp and fertilization —It is a very common thing to find specimens of Asellus in the spring and early summer in pairs, a larger one above carrying a smaller one. These couples are always male and female male above, female beneath. This is the marriage-clasp, and the larger male always holds the female by a special pair of legs, the 4th pair, which are smaller and slightly modified for this purpose. The 7 pairs of legs normally increase in size from before backward ; but in a mature male this iniddle pair is an exception to this (PI. 26. fig. 3). The three distal segments are modified to form a hook. The carpopodite, with a flattened surface fringed by special spines, is the dorsal arm of the hook, and grips the dorsal surface of the female’s thorax. The propodite and the dactylopodite act as the ventral arm of the hook, which bends round the edge of the tergum and grips the ventral surface. The gripping surface of these two segments is similarly flattened and roughened with spines. Each of the legs is able to bend inwards at the joint between the basipodite and the ischiopodite (PI. 25. fig. 3); this gives the legs a Z-like arrangement, and the female can be slung in the space between the ventral surface of the thorax and the legs of the male. By varying the angle between the basipodite and the ischiopodite, the female can be raised or lowered for locomotion or feeding. Kaulbersz does not mention these legs. The position of hooking is not quite constant, but is usually near the middle of the thorax. The female does not seem to mind the inconveniences of her position, but resigns herself to the inevitable. When the male walks forward slowly, 338 MR. E. E. UNWIN: NOTES UPON THE she walks as well; but if he is moving rapidly, she allows herself to be dragged along. In this case she bends her head and fore-part of the thorax downwards so as to present her rounded back to the rush of water and to the obstacles that may be in the way. When the male stops she lifts her head and continues her feeding and breathing actions. The male is not inconvenienced by the presence of the female, as his larger size enables him to reach forward in front of the female for food. Tf a male and female are put together in a dish of water, the process of capture and hooking can be easily watched. When they meet, the male quickly seizes the female; sometimes he has a long chase before he can secure her. Kaulbersz gives some observations, noting that if several males are present there are fights for the female, and that yellow-coloured males usually come out victorious. The female is secured by the Ist pair of legs, which are prehensile. In the male these legs are especially large and powerful (PI. 25. fig. 2). The figure shows the modification. The propodite is much enlarged, forming one side of the claw, and containing the powerful muscles for working the dactylopodite down upon it. Both of the segments have the opposable faces roughened with spines, and, in addition, there is a process carrying a few large spines near the proximal end of the propodite. The carpopodite is very much reduced. When the male grasps the female, he turns her quickly over, as though inspecting her condition. If she has been impregnated already, or has the brood-pouch developed, ire releases her at once; if she is in the right condition, he quickly turns her about until he can hook his 4th pair of legs over the edges of her thorax. I do not find that he prefers to le on his right side as Kaulbersz suggests; indeed, he more often lies on his back so as to bring all his legs into use. Having secured the female, the male rights himself and walks off with the female slung beneath. The real meaning of this association of male and female has become clear by a number of observations made in the spring of 1908 and verified several times since. Most observers refer to this association and to the varying times of ity duration; but they have not appreciated the real significance of it or the reason for the longer or shorter time of association in this marriage- clasp. The length of the period of association depends upon the nearness of the next female ecdysis. Impregnation and egg-laying cannot take place until achange of cuticle releases the large oostegites for the brood-pouch and exposes the external openings of the oviduct. It is the casting of the hind- half of the female cuticle which gives the signal for the attentions of the male, and this casting is always accomplished before copulation. If one thinks of the short time which elapses between the casting of the cuticle and egg-laying, as well as the delicate condition of the creature at the time, it REPRODUCTION OF ASELLUS AQUATICUS. 339 will be obvious that the association is of real value. The chance of male and female not meeting at the right time, and the fact that in all probability the female would be in hiding, makes this marriage-clasp most important. When the ecdysis is about to take place, the male gives his attention to the female, and seems to assist her in working off the hind-half by walking slowly forwards a few steps and by holding her in such a position that the waving motion can be performed with ease. When the cuticle is cast, the female is ready for the attentions of the male. He need not wait until the fore-part is cast, for the oviducal openings are on the 5th thoracic segment, and this segment is exposed by the casting of the hind-half. The male now adjusts the position of the female so that her 5th thoracic segment is immediately below his 2nd pleopod. Kaulbersz mentions that the impreg- nation takes place with the creatures belly to belly. I have never seen this. All that I could see of this operation was a sideways sliding of the body of the male, so that the ventral surface of his body pressed against the side of the female’s thorax. It looked almost like a hugging of the female, and the arrangement would certainly bring the 2nd abdominal appendages into the immediate neighbourhood of the female reproductive openings. After occupying this position for some time on one side, he changed sides and the operation was repeated. I have kept couples under observation during the whole of this period, that is to say from the casting of the hind-half to the release of the female, and have never seen the two individuals with ventral surfaces opposed. The length of this active copulation period varies somewhat. In one case it only lasted an hour. The hind-half of the female’s cuticle was finally cast free at 8.30 p.m. after two hours being spent on this operation. The male waited about 20 minutes before beginning copulation, and this was completed and the female released by 9.30 p.m. This was the shortest time recorded, tlle average being from 2 to 3 hours. In one case it lasted 5 hours. After separation from the male, the females had a mass of spermatozoa in the expanded receptacles just within the openings on the 5th thoracic segment. I have verified this by dissection and by paraffin sections (Pl. 26. fig. 12). The female is now released, as the purpose of her long association with the male has been achieved ; and she seeks some convenient hiding-place, under a leaf or stone or in a hollow stem, to await the completion of her ecdysis, and with it the release of the oostegites. This completion of the eedysis takes place usually about 24 hours after the first part, but I have a record of 8 hours only. The oostegites are now released. Up to this they have been represented by small club-like processes; but if one of these is examined, the much larger plate-like oostegite is seen folded up within, like a leaf in a bud waiting 340 MR. E. E. UNWIN: NOTES UPON THE for its release at the time of ecdysis. Hach oostegite is a large free oval plate, overlapping its neighbours and its fellow of the other side to form a very safe brood-pouch for the reception of the eggs. The first 4 pairs of legs carry these oostegites attached to the coxopodites. The general structure can be seen by reference to Pl. 26. figs. 10 & 11. It is interesting to find that the detailed structure of the oostegites is very like that of the branchial plates. Both surface observations of living oostegites and transverse sections of fixed material reveal a very similar arrangement to that found in the 4th pair of pleopods. It is easy to see in living specimens the flow of blood through the oostegites, and they show blood- spaces and “pillars” of much the same structure as those described by Kimus in connection with the branchiz of Asellus. When the oostegites are released, the next operation is the passage of the eggs from the ovary to the exterior by way of the very short oviduct and receptaculum seminis. The fertilization probably takes place at this time, and the eges are passed out through the aperture at the base of the 5th pair of legs and conducted into the brood-pouch. The 4th pair of oostegites can, by a slight bending of the body, be made to overlap the apertures through which the eggs are passed out. In this way, and by the help of the last two pairs of thoracic legs, the eggs are passed into the breod-pouch. This operation does not take place directly the brood-pouch is available : a period of from 5 to 12 hours elapses before the brood-pouch is filled with the eggs. The number of eggs laid at a time varies somewhat, the smallest number which I have found being 39 and the greatest 62; the average of many specimens is 46. The succession of events can now be summarized :— 1. Male and female in association until the hind-half. of the female cuticle has been cast. 2. Impregnation of the female by the male. 3. Female released by the male from the marriage-clasp. 4, Fore-half of the female cuticle cast; the oostegites are released thereby. 5, Fertilization and passage of eggs into the brood-pouch. Development period.—The eggs are well supplied with yolk, and the whole of the development takes places inside the brood-pouch, which is, as we have seen, external to the body and contained between the ventral body-wall and the overlapping oostegites. The problem of a constant supply of fresh water is solved in several ways. The movement of the legs in walking will cause a certain amount of movement of the oostegites and thus cause a slight changing of the water ; REPRODUCTION OF ASELLUS AQUATICUS. 341 but there are definite and very effective methods of doing this. The first one to be noted is the lowering and raising of the oostegites themselves. The plates slowly lower a little way. The appearance is as though the brood-pouch was swelling. Then the plates return rather quickly to the normal position. A considerable change of water is caused by this method, and the eggs are seen to shake about during the operation. If the animal is undisturbed, this operation may be repeated about 8 times a minute for 10 minutes or longer ; and then a period of rest before restarting another set of pulsations. By the use of powdered carmine the effectiveness of the method was easily demonstrated, and it was clear that the flow of water was from behind forwards. The carmine was sucked in at the hinder end of the pouch and, after several movements of the plates, was entirely removed. The 1st pair of legs is carried bent inwards at the joint between the basipodite and the ischiopodite, to form two elbows which repeatedly jerked backwards into the brood-pouch whilst the plates were being lowered and raised. By fastening the legs so that they could not do this, I found that they acted as guards at the exit of the brood-pouch. By their action they prevent eggs escaping with the outgoing rush of water. Quite a number of eggs escaped whilst the legs were not in action, but under normal conditions I have never seen them escape (Pls. 25 & 26. figs. 2 & 10). There is a further method of changing the water in the brood-pouch. It is a less vigorous method than the one just deseribed, and is not seen during the movements of the oostegites. It is in the periods of rest between the pulsations of the brood-pouch that the flapping movements of the maxilli- pedes are noticed. The maxillipedes are broad and plate-like, and are situated just in front of the 1st pair of oostegites. By their flapping movement they draw a steady stream of water through the brood-pouch. PI. 25. fig. 5 shows one of these maxillipedes taken from a female with a brood-pouch packed with eggs. If it is compared with a maxillipede of a non-eeg-carrying female or of a male (Pl. 25. fig. 4), it will be seen that a coxal Jobe is present. This lobe is covered with feathered sete and projects a little way into the brood-pouch (PI. 26. fig. 10). The addition of this feathered lobe will make the maxillipede more effective as a current producer; but their position and structure make them effective guard against the escape of eggs during this operation. The Ist pair of legs only come into action during the more vigorous method. This coxal lobe is only found on the maxillipedes of a female with brood- pouch. Ina female taken from the marriage-clasp, and about to cast the cuticle, the new chitinous cuticle, with the coxal lobe, could be seen under the old cuticle of a normal maxillipede, ready to be pushed out when the old cuticle was cast. The eggs take about a month in their development. The tiny Aselli LINN. JOURN.—ZOOLOGY, VOL, XXXIV. 26 342 MR. E. E. UNWIN: NOTES UPON THE remain for some days in the brood-pouch, eating the débris and generally getting used to life before struggling out of the pouch at the hinder end. There is no general escape of the young; they push out by twos and threes, and even after a week or ten days one or two young Aselli can be found in the pouch. Here are three typical periods of development :— Eggs laid Feb. 18th ; young escaping Mar. 20th. be a Wiles TAOS as ee ra\yayesy( (telNs > gs | Mans2 Othe: May Ist. br) 3° Soon after the escape of the young, the female passes through another ecdysis and returns to the condition similar to that of the normal female. The large oostegites are replaced by small processes. Noie upon a variation of the 1st pleopod of the female.—In Pl. 25. fig. 9 is shown a curious variation of the Ist abdominal appendage. The normal one is seen in Pl. 25. fig. 8. It approximates to the 2nd abdominal appendage of the male (Pl. 25. fig. 7). I discovered it whilst watching the circulation of blood in the oostegites of a living female with egys. The unusual appearance of the appendage attracted my attention. I examined at. once every available female, 35 in number, but did not find another with any such variation. Ernest Hwarr UNWIN. BIBLIOGRAPHY. KauLBersz, G. J.—‘‘ Biol. Beobach. an Asellus aquaticus.” Zool. Jahrb.; Jena (Abt. f. alle. Zool.), 1913, Bd. 33. Kimus, J.—“ Recherches sur les Branchies des Crustacés.” La Cellule, t. xiv. 2efase. Leypic, ¥.—* Uber Amphipoden und Isopoden.” Zeitschr. £. wiss. Zool. xxx. Bd. Suppl. 1878. Rosensrapr, B.—* Organ. von Asellus aquaticus.” Biol. Centralbl. viii. 1888-9, Sars, G. O.—Histoire naturelle des Crustacés d’eau douce de Norvége. 1867, pp. 93-123. ScuNnemeEr, R.—“ Hin bleicher Asellus in d. Gruben vy. Freiberg im Hrzge- birge.” Sitzb. Kel. Pr. Akad. Wiss. Math. u. naturwiss. Mitth. Berl. 1887. Weer, W.—* Morph. u. Exp. Studien. Asellus agquaticus.’ Zool. Jahrb. (Abt. f, allg. Zool.), xxx. 1911. UnwIy. JOURN. LINN. SOC., ZOOL. VOL. XXXIV. PL. 25. ASELLUS AQUATICUS. Unwin. JOURN. LINN. SOC., ZOOL. VOL. XXXIV. PL. 26. EEUnwin. del, ASELLUS AQUATICUS. =) le} JQ” 2 dQ” dg” REPRODUCTION OF ASELLUS AQUATICUS. EXPLANATION PLATE 25. 343 OF THE PLATES. 1. Last thoracic leg (7th), for comparison with the specialised legs in Pl. 25. figs. 2 &3. x30. 2, Ist thoracic leg of S. x 30. eth Gs. oy ee we: 4, Maxillipede of a female during the non-reproductive stages. x 30. De °F _ ; A », brood-pouch stage. X30. (Notice the coxal lobe ep.) 6. 1st abdominal appendages of g¢ with the spout-like terminations of the vasa deferentia. X 20. 7. 2nd abdominal appendages of dg. x 30. . lst abdominal appendages of 2. x 30. . Ventral view of female. . Variation of lst abdominal appendage of 9. PLATE 26. been cut off. . An oostegite, ventral view. . coxopodite. . basipodite. Lh. . ischiopodite. References. MUP. x 50. x 10. All the thoracic legs, except the Ist left, have x 40. MACRUROUS DECAPODA OF THE SUDANESE RED SEA, 377 antennular peduncle, while the antennal scale reaches to the level of the distal end of the second joint of the antennular peduncle. The lateral spine of the basicerite is very acute and as long as the first joint of the antennular peduncle, therefore slightly shorter than the stylocerite. The spine at the upper angle of the basicerite is well developed, acute, about one-third as long as the lateral spine. The telson (Pl. 28. fig. 5) is two and a third times as long as wile at the posterior margin, Its posterior angles are acute but not spinous, and immediately inside of them the posterior margin bears two spines on each side, the inner of which is slightly the longer. The usual two pairs of dorsal spines are present, rather robust in form, the posterior pair situated about the middle of the telson. The larger cheliped (Pl. 28. fig. 2) is of the general form found in S. biunquiculatas. The merus is triangular in cross-section, each angle of the triangle marking a carina running longitudinally down the joint and ending distally in acute spines. The carpus is acutely spinous at its lower distal corner. Tbe ehela is rather more than 24 times as long as broad, quite smooth and oval in form. The anterior margin of the palm ends in an acute but not spinous tubercle. The fingers are about one quarter of the length of the whole chela. The merus of the smaller cheliped is three times as long as wide, unarmed at the apex, with numerous lone sete on the inner margin. The whole chela is three times as long as broad, the palm being two-thirds of the total length and the fingers one-third. The movable finger is tapering and furnished with stiff sete at its tip. The merus of the second pair of legs (PI. 28. fig. 3) is very nearly five times as long as wide. The carpus is about equal in length to the merus, the first joint equal in length to the succeeding four, the last joint longer than the combined length of the third and fourth. The chela is very slightly shorter than the first joint of the carpus, with the fingers one and a half times as long as the palm and fringed with long tufts of setee. The third pair of legs (Pl. 28. fig. 4) are of relatively stout form. The merus is rather more than three times as long as broad and unarmed. The carpus is four-ninths and the propodus two-thirds as Jong as the merus, the carpus being unarmed and the propodus bearing seven short spines on its inner margin. The dactylus is very short with the secondary nail well developed. Length of the only male 13 mm., of the female 13-20 mm., the smallest as well as the largest of which are ovigerous. This species falls within that group of forms belonging to the birmmgucu- latus division of the genus, in which the posterior margin of the sixth abdominal somite is armed with teeth. De Man has described six species in the Siboga Report, belonging to this group: S bispinosus, S. triacanthus, LINN, JOURN.—ZOOLOGY, VOL. XXXIV, 29 378 DR. W. M. TATTERSALL ON THE STOMATOPODA AND 8. quadridens, S. quadrispinosus, 8. trispinosus, and S. septemspinosus, the specific names of which refer to the number of teeth on the posterier margin of the sixth abdominal somite. Following this nomenclature, the specific name of the new species here described indicates at once the main point of difference from the species described by De Man. Among De Man’s species, S. quinquedens is at once distinguished from S. triacanthus, S. trispinosus, anid S. septemspinosus by the character of the rostral plate. In the first two species the central spine is exceedingly long, much longer than the lateral teeth and as long as or longer than the first joint of the antennular peduncle. In S. septemspinosus the central prong of the rostral trident is longer than the laterals, whereas in S. guinguedens all three are equal. 8. trispinosus and S. septemspinosus are further distinguished from S. quinguedens by having the merus of the third legs armed with seven or eight spinules. Of the other three species, S. quinquedens approaches most closely to S. quadri- spinosus, differing only in the extra spine on the sixth abdominal segment. This group of species is most nearly related to the type-form of the division of the genus to which they belong, S. biunguiculatus, and in fact, but for the armature of the sixth abdominal somite, would be difficult to separate from that species. Genus ALPHEUS, Fabricius. Macrochirus group. ALPHEUS GRACILIS, Heller, 1861, p. 271, Taf. 3. figs. 19, 20. See De Man, 1911 4, p. 387, pl. 14. fig. 60. Localities. Station VII. B, 1 g, 16 mm., 1 2 with eggs, 17 mm. Station IX. A, 4 ¢, 9-17 mm., 4 2, 9-16 mm., three of which, 12-16 mm. in length, were carrying eggs. Distribution. The type-form has so far not been met with outside the Red Sea, from which both Heller and Coutiére have recorded the species. The var. alluaudi, Coutiére, is known from Mahé and the var. luciparensis, De Man, from Lucipara Island in the Dutch East Indies. ALpHevs ventrosus *, H. M.-Hd., 1837, p. 352. A, ventrosus, Coutiére, 1905, p. 882. A. ventrosus, De Man, 1911 8, p. 359. Localities. Station VII. B, six specimens. Station V. H, several. Distribution. Recorded previously from the Red Sea by Heller, Paulson, Kossmann, De Man, Miers, Nobili, and Coutiére ; widely distributed through- out the Indian and Pacific Oceans. * Stebbing (1915) identifies this species with the earlier described Alpheus lottini, Guérin, which name it should accordingly bear, MACRUROUS DECAPODA OF THE SUDANESE RED SEA. 379 Crinitus-obeso-manus group. ALPHEUS MICRosTYLUS, Sp. Bate, 1888, p. 566, pl. 101. fig. 6. A. microstylus, Coutiére, 1905, p. 884, pl. 76. fig. 23. A. microstylus, De Man, 1911 4, p. 844. Loeality. Section V. EH, 2 g, 20 and 22 mm. Distribution. Previously recorded from the Red Sea by Coutiére from Djibouti and Maseat ; widely distributed in the Pacific Ocean. Crinitus-crinitus group. ALPHEUS ALCYONE, De Man, 1902, p. 870, Taf. 27. fig. 61. A. alcyone, Nobili, 1906 4, p. 32. A. aculeipes. Coutiére, 1905, p. 892, pl. 79. fig. 31. A. alcyone, De Man, 1911 3, p. 351. Locality. Station V. C, one specimen, 10 mm. Distribution. Previously recorded from the Red Sea by Couti¢re from Djibouti ; widely distributed in the Indian Ocean, ALPHEUS sp. ? Locality. Station V. A, 1 g, 10 mm. Remarks, This specimen approaches A. aleyone very closely, but differs from it mainly in the fact that the median rostral process is absent or obsolete so that the rostrum presents the form of an emarginate plate. The dactylus of the last three pairs of legs has a very small accessory tooth so that it is obscurely bi-unguiculate, There is no prominent spine on the distal corner of the lower margin of the carpus of the third legs, but there are two spines on the median portion of this margin. The second joint of the carpus of the second pair of legs is about three times as long as the first. The other characters are exactly as in A. alcyone. ALPHEUS BUCEPHALOIDES, Nobili, 1905 b, p. 238. A, bucephaloides, Nobili, 1906 6, p. 29. Localities. Station V. E,1 9,9 mm. Station IX. A, 1 ¢, 10 mm. Remarks. It is with a considerable amount of reserve that I refer these specimens to Nobili’s species. They differ from the latter in not having the movable finger of the small cheliped broadened and fringed with sete, and in having the second joint of the carpus of the second legs 1°3 times as long as the first, instead of 1°6 as in A. bucephaloides. The carpus of the third legs has only three spinules on the lower border in addition to the terminal one and has no spines on the upper border. There 99* 380 DR. W. M. TATTERSALL ON THE STOMATOPODA AND is only one spine on the outer uropod at the suture. These small differences may be due to immaturity, as the specimens are in otherwise close agreement with A. bucephaloides. Distribution. This species is new to the fauna of the Red Sea and is other- wise only known from Nobili’s record from the Persian Gulf. ALPHEUS ConsoBRINUs, De Man, 1908, p. 101. A, consobrinus, De Man, 1911 4, p. 360, pl. 16. fig. 75. Localities. Station V. A, 1, 19, 14mm. Station V. D,1 9,9 mm. Station V. BE, 29,14 mm. Station VIT. C, 1 g,192, 10mm. Station IX. At, Wigs 222.9 mm. Remarks. This species is very closely allied to A. bucephalus, Coutiere, but differs in having the fingers of the small chela of the male expanded and hairy. The differences between the two species do not seem to me to be clearly established, but by reason of the above character I refer my specimens to De Man’s species. Distribution. New to the fauna of the Red Sea; otherwise only known from the waters round the Dutch East Indies. ALPHEUS PACHYCHIRUS, Stimpson, 1869, p. 30. A, pachychirus, De Man, 1911 8, p. 366, pl. 16. fig. 77. ; - Locality. Station IX. A, 1 ¢@, 15 mm. Distribution. Recorded once previously from the Red Sea by Coutiére from Djibouti ; widely distributed throughout the Indian and Pacific Oceans. Crinitus-insignis group. ALPHEUS INsIGNIS, /Teller, 1861, p. 269, Taf. 3. figs. 17-18. A, insignis, Coutiére, 1905, p. 899. Locality. Station TX, A, 1 ¢?. Remarks. This specimen agrees very well with the descriptions of A. insignis except that the lateral lobes of the rostrum are not setiferous. Distribution, Recorded previously from the Red Sea by Heller, Paulson, De Man, Nobili, and Coutiére ; otherwise widely distributed in the Indo- Pacific Ocean. ALPHEUS PARACRINITUS, Miers, 1881, p. 365, pl. 16. fig. 6. Locality. Station VII. C, four specimens, 11-12 mm. Distribution. Recorded by Couti¢re from the Red Sea at Djibouti; other- _ MACRUROUS DECAPODA OF THE SUDANESE RED SBA. 381 wise known from the original record of Miers from Goree Island, Sene- gambia. A variety of this species, bengalensis, Coutidre, is known from Minikoi. Edwardsi group. ALPHEUS AUDOUINH, Coutiére, 1905, p. 911, pl. 87. fig. 52. A. audouinti, De Man, 1911 4, p. 414, pl. 28. fig. 100. Locality. Station V. HE, 1 $6, 21 mm.,1 9,19 mm. Distribution. This species has hitherto been confused with A. edwardsii, Audouin, so that the previous records for the Red Sea are not easy to determine. It has, however, been certainly recorded by Nobili and Coutiére from this area. ALPHEUS STRENUUS, Dana, 1852, p. 543, pl. 34. fig. 4. A, strenuus, Coutiére, 1905, p. 913, pl. 87. fig. 63. A. strenuus, De Man, 1911 4, p, 423. Localities. Station I. A, one specimen, 28 mm. Station V. I, two speci- mens, 17 and 24 mm. Station VII. C, one specimen, 22 mm. Distribution. Previously recorded from the Red Sea by De Man, Nobili, and Couticre ; a widely distributed Indo-Pacific species. ALPHEUS BOUVIERI, A. J/.-Ed., var. HULULENSIS, Coutiére, 1905, p. 908, pl. 85. fig. 46. Locality. Station V. C,1 ?, 10 mm. Distribution. Recorded from the Red Sea by Couti¢re from Djibouti ; otherwise only known from the Maldives. ALPHEUS PARVIROSTRIS, Dana, 1852, p. 551, pl. 35. fig. 3. A, parvirostris, Coutiére, 1905, p. 906. A, parvirostris, De Man, 1911 4, p. 432, pl. 22. fig. 106. Localities. Station V.A, 2 6, 12 and 13 mm., 5 ?, 9-14 mm. Station V.-B, 14 specimens. Station V.C,2 ¢,8and10 mm. Station V. E, 2 ¢, 1Zand 14 mm., 2 ovig. 29,12 and 14 mm. Station VII. C, 2 6,12 and 13 mm., 2 ovig. ?, 12 and 13mm. Station IX. A, 3 (including 1 ovig. 2), 5-9 mm. ; Distribution. Recorded from the Red Sea by Heller and Coutiére ; other- wise widely distributed in the Indian and Pacific Oceans. In addition to the above-named species the collection contained two ovigerous female Alpheids, from among coral in one fathom of water at Mersa Ar-rakiya, which were too much damaged to identify with certainty. 382 DR. W. M. TATTERSALL ON THE STOMATOPODA AND Family HIPPOLYTID A. Genus Saron, Vhallwitz. Saron necLectus, De Man, 1902, p. 854, pl. 26. fig. 58. S. neglectus, Coutiére, 1910, p. 71, figs. pp. 73, 78. S. neglectus, Kemp, 1914, p. 87. Locality. Section LX. A, two specimens, 25 mm. Previously recorded from the Red Sea by Coutiére. Genus THor, Kingsley, 1878. THOR PASCHALIS (L7/eller, 1861, p. 276, pl. 3. fig. 24). Paschocaris paschalis, Nobili, 1906 a, p. 38, pl. 3. fig. 1. Thor paschalis, Kemp, 1914, p. 94, pl. 1. figs. 6-10. Locality. Station V. CU, two specimens, 9 mm. Remarks. The carpus of the second perepods is six-jointed and the telson bears four pairs of spinules. Distribution. Previously recorded from the Red Sea by Heller and Nobili. For further distribution see Kemp (loc. ct.). Genus Hippotyren, Leach. Hirpotyre prorevs, Paulson (1875, p. 109). (Pl. 28. figs. 10-12.) See Nobili, 1906 a, p. 83. Locality. Exact locality uncertain, The label bears the number W. 12, and I think refers to specimens captured in Suakin Harbour. Remarks. Two small specimens of Lippolyte, 9 and 11 mm. in eee I refer to Paulson’s species, with some doubt. Nobili (1906 a) has published a translation of Paulson’s original description, and to it I am indebted for information on this form. My specimens belong to category A of Paulson and to the first group of that category. The rostrum (PI. 28. fig. 10) is equal in length to the antennular peduncle and much shorter than the antennal scale. The upper edge bears three teeth, the first two placed at one- third and two-thirds of the way along the rostrum and the third and smallest tvoth near the tip. The larger specimen has two small teeth on the lower edge in the anterior (distal) third, the smaller specimen only one small tooth. The antennal scale (PJ. 28. tig. 11) is about three and a quarter times as long as broad. The carpus of AN second pair of legs (Pl. 28. fig. 12) las the first joint twice as long and the third joint one and a third times as long as the second. The latter is one and a half times as long as broad. MACRUROUS DECAPODA OF THE SUDANESE RED SBA 383 Nobili regards the category B specimens of Paulson as synonymous with H, orientalis, Heller, which Kemp (1914) suggests is possibly a synonym of H. ventricosus, M.-Ed. From the latter as redescribed by Kemp, my speci- mens differ in having a much shorter rostrum. In //. ventricosus the rostrum is equal in length to the antennal scale and much longer than the antennular peduncle. In my specimens the rostrum is only as long as the antennular peduncle and much shorter than the scale. There are also differences in the proportions of the joints of the carpus and in the antennal scale, and it seems useful to indicate these by the following table, in which I have incorporated measurements made from Kemp’s figures. HT. ventricosus. HT, varians. HH, proteus. Antennal Scale— Tere DBe isusshee eat cos re cesar: 3:0 35 3°25 Joints of Carpus1 ........ 3:0 2°6 2:0 a pep een ete 10 10 10 Be Be Un Ob stimeradcets NY 14 ss 2nd joint of CarpusL:B .. 10 2:0 15 At the same time it is only just to point out that my specimens are not fully grown, and the proportions of the joints of the carpus may change with age. The rostrum is well known to be of very variable form in species of this genus. I record my specimens under the name //. proteus, rather with the idea of indicating their structure, than from any conviction that the species is really distinct from H. orientalis, Heller, or H. ventricosus, M.-Ed. Family PALASMONIDAS. Subfamily Poxntoniin#®, Kingsley, 1878. Genus PAL&MONELLA, Dana, 1852. PALEZMONELLA TENUIPES, Dana, 1852, p. 582, pl. 38. fig. 3. P. tenwipes, Nobili, 1906 a, p. 70. P. tenwipes, Borradaile, 1917, p. 358. Localities. Station V. BR, one specimen, 10 mm. Station VII. C, three specimens, 10-17 mm Remarks. I think this is certainly the species recorded by Nobili as P. tenwipes, Dana. The main points of difference from Dana’s description and figures are:—(1) The relatively longer palm to the chelze of the second pair of legs; (2) the merus is armed with a spine at the distal extremity of the lower border only ; (3) Dana makes no mention of the minute teeth on the fingers of the chelee noted by Nobili and present in these specimens. 384 MR. W. M. TATTERSALL ON THE STOMATOPODA AND The four specimens in the present collection have the rostrum armed with seven or eight teeth above, two of which are on the carapace, and two below. Two of the specimens still retain the second pair of legs, and these, measured, give the following proportions, taking the carpus as unity. These measurements are compared with those given by Nobili in his description of this species and with similar measurements taken from Dana’s figure and from Borradaile’s figure of P. tridentata (1898 a, p. 1007, pl. 64. fig. 8). T have added the proportions of the joints of the first legs, derived from the same sources, | Crossland Red Sea Tons | Borradaile. specimens. Nobili: Dana. P. tridentata. | Ist leg. 2nd leg. | 2nd leg. | Ist leg. “2nd leg. | Ist leg. | 2nd leg. me nes [ea Sal a Soe Oe tes Merus ...... VSO Sea aoe nea ase dee 9 7 Carpus: sete: | 1:0 10 10 10 =| 10 NO) sie 20) 1:0 Palma asses ‘4 eG eal 15 | +625 ie | 5 15 Fingers ...... st) | 10 10 8 625 8 | “4 ‘15 From this table of measurements it will be noted (1) that in Dana’s specimens the palm of the second leg is relatively shorter than in either Nobili’s specimens or mine, and (2) that my specimens have relatively longer fingers than either Dana’s or Nobili’s, though Nobili gives the measurements of the palm and fingers of another specimen in which the fingers are relatively longer than in the other limb measured and approach more closely to the present specimens. As regards the first pair of legs, Dana’s specimens have the arm and hand both longer than the wrist, while in mine the arm is equal to the wrist and the band shorter. Palemonella tridentata, Borradaile, is very closely allied to P. tenwipes. Borradaile gives the following points of difference :— (1) There are three teeth on the underside of the rostrum, instead of two. (2) The inner edges of the fingers of the second pair of chelee are armed with teeth. (3) The distal end of the merus is rounded in profile, but provided with alarge spine below at ashort distance from the end. In P. tenwipes it is acute in profile and without the tooth. (4) The arrangement of teeth on the inner ramus of the mandible is different in the two species. Of these differences, number one is merely an individual variation, and numbers two, three, and four apply also to the present specimens and to MACRUROUS DECAPODA OF THE SUDANESE RED SEA. 385 Nobili’s as distinct from Dana’s. The most important difference between the present specimens and P. tridentata is brought out in the table of measure- ments of the second pair of legs given above. In my specimens, and also-in those examined by Dana and Nobili, the merus is longer than the carpus. In Borradaile’s figure of P. tridentata it is shown as considerably shorter than the carpus. On the other hand, the proportions of the joints of the first legs in P. tridentata agree closely with the present specimens. As a result of the above considerations, it follows that the present specimens belong to the same species as those examined by Nobili and referred by him to P. tenuipes, Dana. I accept Nobili’s identification. P. tridentata, Borradaile, is closely allied to P. tenwpes and doubtfully distinct. Distribution, Previously recorded from the Red Sea by Nobili; Sulu Sea (Dana) ; Japan (Ortmann) ; Maldives (Ortmann) ; Amboina (De Man). Genus PERICLIMENES, Costa. PERICLIMENES PETITLTHOUARSLI (Audouin, 1826). Palemon petitthouarsi, Audouin, 1826, Descr. Egypte, Hist. Nat. I. iv. p. 91; Savigny, Atlas, Crust., pl. 10. fig. 5. Periclimenes petitthouarsti, Borradaile, 1898 6, p. 381; and 1917, p. 369. Localities. Station V. A, one, 16 mm. Station V. C, two, 13 and 15 mm. Station VII. EH, one, 11 mm. Station VII. C, four, 10-15 mm. Sulan Coast, no definite locality, forty, 7-18 mm. : Distribution. Previously recorded from the Red Sea by Audouin, Paulson, Kossmann, and Nobili, and by the latter from the Persian Gulf, PERICLIMENES CALMANI,n. sp. (PI. 27. fig. 11; Pl. 28. figs. 14-15.) Locality. Sudan Coast, four specimens, 13-17 mm, Description. The carapace is smooth, without supra-orbital spines. Antennal and hepatic spines are present, the latter a little below the level of the former. The rostrum (PI. 27. fig. 11) reaches almost or quite to the apex of the antennal scale and considerably beyond the apex of the antennular peduncle. It is concave from the base and bears dorsally 8-9 teeth, the proximal tooth well behind the orbit, the second tooth placed over the orbit, and the remainder more or less regularly spaced with a longer interval between the 5th and 6th tooth. The lower edge bears 4-5 teeth, the proximal one below the 5th tooth of the upper edge. The third maxillipeds reach to the end of the antennal peduncle. The exopod is longer than the antepenultimate joint. The latter bears six spinules and a few setee on its outer margin. The first legs (Pl. 28. fig. 14) extend 386 DR. W. M. TATTERSALL ON THE STOMATOPODA AND forwards to the apex of the antennal scale. The carpus is about one-sixth longer than the merus and one and a half times the length of the hand. The palm is equal in length to the fingers. There are no ischial, meral, or carpal spines. ‘The fingers are without teeth on their cutting-edges. ‘I'he second legs (PI. 28. fig. 15) extend beyond the antennal scale by one-half of the carpus plus the whole of the hand. The proportions of the joints are as follows :—Ischium 2°0 mm., merus 2°3 mm., carpus 2°55 mm., palm 1°8 mm., finger 1°6 mm. There are no ischial, meral, or carpal spines. Both the fixed and immovable fingers have on their inside margins a shallow oval pit bounded by small teeth proximally and distally. This species clearly belongs to the subgenus Falciger, and if we attempt to run it down with the aid of Borradaile’s key to the species of the subgenus we find that it would fall in Section I. by reason of the oval pits in the fingers of the large chela, a peculiarity only noted in two of the species, P. spiniferus, De Man and P. petitthouarsi (Audouin). From the former species, P. calmani is distinguished by the absence of supra-orbital spines and from the latter by the long and slender form of the great chela, the much greater length of the carpus of these limbs, and the absence of spines on the first and second legs. it is possible, however, that the oval pits have been overlooked in other species of the genus. If, therefore, we ignore Section |. of Borradaile’s key and proceed to Section II., we find the nearest ally of P. calmani in P. sey- chellensis, Borradaile, from which it is distinguished by the shorter rostrum and the longer pair of second legs. P. calmani is very closely similar to P. demani, Kemp, 19154. The general form of the body and the proportions of its various parts are very much alike. But P.demani has a supra-orbital spine and has meral and carpal spines on the second legs. Judging from Kemp’s figure, P. demani also has oval pits on the fingers of the chelze of the second pair of legs. PERICLIMENES sp. ? Localities. Station V. C, two, 7-10 mm, Station VII. A, one, 7 mm. Uncertain, one, 13 mm, These specimens belong to the subgenus Fulciger and to that group of the species having a supra-orbital spine. The rostrum is longer than the antennular peduncle and subequal to the antennal scale. The formula is - = and the rostrum is concave from the base. Unfortunately the second legs are missing in all the specimens. I think they are almost certainly the form recorded from the Red Sea by Nobili under the name of P. ensi- frons, Dana, but as the second legs provide the characters for specific determination it is not advisable to name these specimens. MACRUROUS DECAPODA OF THE SUDANESE RED SEA. 387 PERICLIMENES sp. ? Locality. Uncertain, one specimen, broken, labelled “* @ommensal, P.O.13.” Remarks, | cannot identify the single mutilated specimen with any described species and I think that it represents a new form. As the second pair of legs and the posterior half of the abdomen are missing, I refrain from giving it a name. Among described forms it appears to be most closely allied to P. borradailei, Rathbun = P. tenuipes, Borradaile, nec Holmes, and to P. kolumadulensis, Borradaile, 1915. The rostrum, which is slender and slightly recurved at the tip, extends beyond both the antennular peduncle and the antennal scale, and is one and a quarter times as long as the carapace measured dorsally. The rostral Derails formula is The carapace bears supra-orbital, antennal, and hepatic spines. Such of the legs as remain still attached to the specimen are exceedingly long and slender. The first pair extend beyond the apex of the rostrum by the whole of the chela. The carpus is one and a quarter times as long as the chela. The fifth leg reaches forward as far as the apex of the antennal scale. This specimen differs from both P. borradailei and P. kolu- madulensis in the shorter rostrum and different rostral formula and in the presence of a supra-orbital spine, but resembles both in the slender form of the legs. PHRICLIMENES sp. ? Locality. Station VII. C, one specimen, 9 mm. Remarks. This specimen cannot be identified with certainty as the second pair of legs is missing. It belongs to the subgenus Falciger and to that eroup of species having antennal and hepatic, but no supra-orbital, spines on the carapace. The suborbital angle of the antero-lateral border of the carapace is acute but not spiniform, and the lower angle sub-rectangular. The rostrum is long, equal in length to the dorsal line of the carapace from the border of the orbit, and extends forward to the same level as the tip of the antennal scale and far beyond the antennular peduncle. It bears seven teeth (including the terminal one) on the upper margin and two teeth on the lower. All the upper teeth are situated on the rostrum itself, the first one immediately above the orbital border, and there are none on the carapace. The antennal scale is about three and a half times as long as broad at its widest part, and the external margin ends in a strong spine which extends beyond the apex of the scale. The antennular peduncle reaches forward to the level of the fifth tooth of the rostrum. It has one spine only on the external distal corner of the basal joint and a prominent spine on the ventral surface of this joint near the middle of the outer margin. The last three thoracic legs seem unusually stout. 388 DR. W. M. TATTERSALL ON THE STOMATOPODA AND The species of this genus are difficult to determine in the absence of the second pair of legs, and | have not attempted a specific identification in this case. PERICLIMENES sp. ? Locality. Station VIL. A, one, 5 mm. Remarks. This small specimen belongs to the subgenus Cristiger and is _most closely allied to P. potina, Nobili. The rostral formula is Fi and all the tecth are on the rostrum, none on the carapace behind the orbit. There is no supra- orbital spine. Genus Harpizius, Dana, 1852. The type species of the genus is Harpilius lutescens, Dana, and Borradaile, on the evidence of Sollaud (1910) that the third maxilliped in this genus has no arthrobranch (although Sollaud does not say what species he has examined), coupled with the remarkable form of the second maxilliped tigured for the type species by Dana, has instituted a new genus Harpiliopsis to include two species, H/. beaupresi and H. depressus, which have the second maxillipeds of normal form and arthrobranchs on the third maxillipeds. The validity of the genus Harpiliopsis seems to me to be questionable. _The addition of a single line to Dana’s figure (the line showing the contour of the antepenultimate joint) will give a form of second maxilliped not unlike that depicted by Borradaile for HH. depressus. It is more reasonable to imagine that this line has been accidentally omitted from Dana’s figure than to suppose that HH. lutescens really has the remarkable form of second maxilliped actually figured. 7. lutescens is otherwise so closely similar to H. depressus as to be doubtfully distinct. The latter species has normal second maxilJipeds, and on Borradaile’s evidence as well as my own, has an arthrobranch on the third maxilliped. So that if my suggestion as to Dana’s figure be ‘accepted, Harpiliopsis at once becomes a synonym of Harpilius. On the other hand, specimens which I have referred to H. gerlachet, Nobili, do not appear to possess an arthrobranch on the third maxillipeds, and they further differ from all the other species of the genus in the absence of a hepatic spine. It would appear therefore that if a new genus is required, it must be instituted to receive H. gerlachei. Until more evidence is forth- coming as to what species of Harpilius Sollaud has examined with reference to the presence or absence of arthrobranchs on the third maxillipeds, it would be premature to define a new genus for /. gerlachei and I prefer, at present, to include all the species in one genus. MACRUROUS DECAPODA OF THE SUDANESE RED SEA. 389 HARPILIUS BEAUPRESI (Andowln, 1826, Deser. Egypte, Hist. Nat. I, 4, p- 91, pl. 10. fig. 4). (Pl. 28. fig. 8.) HT, beaupresi, Borradaile, 1898, p. 886; and 1917, p. 379. H. beaupresi, Nobili, 1906 a, p. 68, Locality. Suakin, Station VIT. A, five specimens, 10-16 mm. el OwD: ve 3 a these specimens have a spine on the lower surface of the basal joint of the Remarks. The rostral formule of the specimens are Lol or > and antennular peduncle. Nobili was unable to find a similar spine in the speci- mens he examined. //. beaupresi is at once distinguished from the other species of the genus by the extreme leneth of the spine on the outer corner of the joint from which the antennal scale springs. The form of the dactylus (Pl. 28. fig. 8) of the last three pairs of legs is characteristic. It is stout, slightly curved, swollen at the base, and equal in breadth at its base to the propodus to which it is attached. Savigny’s original figure shows the form of the dactylus very well, but I refigure it here to compare with other species of the genus. Istribution, Previously recorded from the Red Sea by Audouin, Heller, Paulson, and Nobili; Persian Gulf (Nobili) ; Hast Indies (De Man). HARPILIUS DEPRESSUS (Stimpson). (Pl. 28. fig, 7.) Harpiliopsis depressus, Borradaile, 1917, p. 380. Locality. Station VIL. C, 1 9,18 mm. Remarks. The identification of this specimen is based on Borradaile’s monograph (1917), but Iam doubtful of the distinctness of this species from IT, lutescens, Dana. Except for the fact that this specimen has two teeth on the lower margin of the rostrum, [ can find no marked character in which it differs from Dana’s species. The discrepancy in the form of the second maxilliped I have already attempted to explain. The present specimen has an arthrobranch on the third maxillipeds. It is to be noted in this connection that Nobili records Dana’s species from the Red Sea, without comment. The species differs from H. beaupresi in its more robust form, in the shorter spine on the outer corner of the joint bearing the antennal scale, in the shape of the antepenultimate joint of the third maxilliped, and in the stouter form of the last three pairs of thoracic legs and their dactyli. The latter are short, stout, and curved, and apparently capable of being almost retracted into a socket at the base of the propodus (PI: 28. fig. 7). The whole arrangement recalls the claws of the carnivora and is found in FH. lutescens, H. depressus, H. gerlachei, and H. consobrinus. In H. beaupresi the last thoracic legs are much more slender, and there does not appear to be a socket at the apex of the propodus for the retraction of the claws. On the ventral surface of the thorax, between the bases of the first pair of 390 DR. W. M. TATTERSALL ON THE STOMATOPODA AND legs there isa strong median forwardly directed spine, and in front of the bases of the third pair of legs are two transverse chitinous ridges or plates, one on each side, meeting in the centre and separated by a notch. These latter are exactly in the situation occupied by the thelycum of the Penzide, but I have no idea whether they function as such in this species or are even homologous. I have noticed similar structures in H. gerlachet. Distribution. Indo-Pacific, in corals. New to the fauna of the Red Sea. HT, lutescens is known from Tongatabu (Dana) and the Red Sea (Nobili). HARPILIUS GERLACHEL, Wobili, 1905 a, p. 160. (PI. 28. fig. 9.) HT, gerlachei, Nobili, 1906 4, p. 45, pl. 6. figs. 10, 10a; Borradaile, 1917, p. 381. Locality. Station V. H, 3 2 with eggs, 14-18 mm. Remarks. These specimens are in substantial agreement with Nobili’s description and figures. The rostrum reaches to about the level of the apex 4-5 1 ’ rostrum and none on the carapace. This species is characterized by the absence of a hepatic spine, which together with its tumid form and the shape of the antennal scale and has the formula all the teeth being on the of the dactyli of the last three pairs of legs serve to render it easily recog- nizable. The dactylus of the last three pairs of legs (Pl. 28. fig. 9) is short, stout, strongly curved, and much narrower at its base than the distal extremity of the propodus, which is swollen slightly and appears to have a socket at its apex into which the dactylus can be retracted. é Distribution. New to the fauna of the Red Sea. Otherwise only known from the Persian Gulf (Nobili). Genus CorALiiocaris, Stimpson, 1860. CORALLIOCARIS SUPERBA (Dana, 1852, p. 573, pl. 37. figs. 2 a-f). See Borradaile, 1898 4, p. 885, and 1917, p. 383. Locality, Station V. EB, 2 9 with eggs, 18 and 23 mm. Remarks. This species has no hepatic spine on the carapace. Distribution. Previously recorded from the Red Sea by Paulson and Nobili; Hast Indies; Tongatabu ; Tahiti. CoRALLIocaRis LucINA, Vobili, 1901, p. 5. C. lamellirostris, De Man, 1902, p. 842, pl. 26. fig. 55. C. lucina, Nobili, 1906 a, p. 57 ; Borradaile, 1917, p. 384. Locality. Station V. H, two specimens, 17 and 18 mm. Remarks. These specimens agree closely with the descriptions given by De Man and Nobili. The only difference I can find from De Man’s figures is that there is a greater interval between the fifth (penultimate) tooth of the MACRUROUS DEGAPODA OF THE SUDANESE RED SEA. 391 rostrum and the small sixth (last) tooth, which is much nearer the apex of the rostrum in my specimens. The rostrum extends forward to the apex 6 of the antennal seale and has the formula me All the teeth are situated on the rostrum and none on the carapace. This species possesses both antennal and hepatic spines on the carapace. Distribution. Only known from the Red Sea (Nobili) and Ternate (De Man). Genus Ancuistus, Borradaile, 1898 b. Ancutstus miErRsI (De Man, 1888, p. 274, pl. 22. figs. 6-10). See Borradaile, 1917, p. 388. Locality. Station VII. G, 16 specimens, g, 15-21 mm., ? , 21-30 mm. Distribution. Previously recorded from the Red Sea by Nobili; Persian Gulf (Nobili) ; coasts of India (Henderson), and the Mergui Archipelago (De Man). ANCHISTUS INERMIS (Miers), 1884, p. 291, pl. 32. fig. B. (PI. 27. fig. 4.) A, inermis, Borradaile, 1898 b, p. 887; and 1917, p. 388. Locality. Station VII. G,1 go, 24 mm. Remarks. The first legs present a feature not hitherto noticed in this species. The chela has the appearance of a somewhat deep spoon or scoop, the edge of which is fringed with somewhat long sete (Pl. 27. fig. 4). This appearance is brought about by the expansion of the propodus and the folding inward of its margin. This character is possibly sexual. Nobili (1906 a) in recording Pontonia pinne, Ortmann, from the Persian Gulf, ascribes to his specimens two characters which do not agree with Ortmann’s original description. He says that the rostrum extends almost to the end of the antennular peduncle, and that the fingers of the great chela are a little longer than the half of the palm, rather more than one-third of the total length of the hand. Ortmann shows the rostrum to be con- siderably shorter than the antennular peduncle and, while giving no pro- portions in his short description, figures the fingers of the great chela of the second pair of legs as scarcely more than one-third of the length of the palm. It is precisely in these two characters that Anchistus inermis, Miers, differs from Pontonia pinne, Ortmann. LT think it is very probable that Nobili’s specimens from the Persian Gulf, recorded as Pontonia pinna, Ortmann, should be referred to Anchistus inermis, Miers. There remains the question as to whether Pontonia pinne, Ortmann, is really distinct from Anchistus inermis, Miers. ‘The differences between the genera lie entirely in the characters of the third maxilliped. In Anchistus, the last 392 DR. W. M. TATTERSALL ON THE STOMATOPODA AND two joints are narrow, while in Pontonia they are broad. Ortmann makes no mention of the form of the maxillipeds in his specimens. No one except Nobili has recorded his species since. I have given reasons above for sup- posing that Nobili’s specimens were really referable to A. inermis, Miers. The two characters of the rostrum and large chela I have already mentioned provide the only points of difference between the two species. Are these sufficient for specific differentiation ? An examination of Ortmann’s type is necessary to clear up this point. Distribution. New to the tauna of the Red Sea; hitherto known from the Indian Ocean. 2 Genus Concnopyres, Peters, 1851. CoNCHODYTES MELEAGRINAE, Peters, 1851. C. meleagrine, Borradaile, 1917, p, 393. Locality. Uncertain. 40 specimens, 12-30 mm., labelled “ Commensal 1e45(O),, dba"? Distribution. Previously recorded from the Red Sea by Nobili; Indo-Pacific in Meleagrina and.occasionally in Tirzdacna. Subfamily Pan #MONIN”. Genus LEANDER, Desmarest. LEANDER TENUICORNIS, Say, 1818, p. 249. Leander natator, Nobili, 1906 a, p. 74, Leander tenuicornis, Stebbing, 1914 a, p. 288. Locality. Station IT, 1 2, 37 mm. Remarks. My specimen agrees very closely with figure 6, pl. 128, of the ot) ‘Challenger’ Report except that the rostral formula is =, two of the dorsal teeth being situated behind the orbit. The rostrum extends just beyond the apex of the antennal scale and is of the deep ‘“Jatirostris” form. The con- fusion which at present exists amony the species of this family is well exemplified in the species here in question. Stebbing (Joc. cit.) has identified the Leander natator of Milne-Edwards and subsequent authors with the earlier described Palemon tenuicornis of Say. At the same time Stebbing points out that this species agrees with Leander squilla in having the palp of the mandible two-jomted. His description runs as follows: ‘the mandibles have a very slender two-jointed palp, the second joint much the longer.’ My specimen agrees exactly with this description. It should, however, be pointed out that specimens in the British Museum labelled Leander natutor have, according to Calman (Kemp, 1910, p. 130, footnete), MACRUROUS DECAPODA OF THE SUDANESE RED SEA. 393 a three-jointed mandibular palp. Spence Bate (1888, p. 784) makes no mention of the number of joints in the palp but merely remarks on its “extreme tenuity.”” The question is of great importance in connection with the characters of the genus Leander, The type of the latter genus is L. erraticus, Desmarest, identified by Spence Bate and later authors with L. natator, M.-Ed., which, as mentioned above, Stebbing has shown to be identical with L. tenuicornis of Say. Stimpson also. gives L. nutator as the type species. It follows, therefore, that the type species of the genus Leander has a two-jointed palp, and if the number of joints in the mandibular palp be considered of generic importance, it is the three-jointed palp species which must be transferred to a new genus. Palemon torensis, Paulson, cannot be identical with Z. natator, M.-Fd., as surmised by Nobili, since Paulson distinetly figures the mandibular palp as three-jointed. On the other hand, Palemonella gracilis, Paulson, is a species of Leander with a two-jointed mandibular palp. These remarks will serve to show the pressing need for a revision of the genera and species of this family. Distribution. Widely distributed in the Atlantic, Mediterranean, Indian, and Pacific Oceans. LEANDER CONCINNUS (Dana, 1852, p. 587, pl. 38. fig. 10). Leander longicarpus, Ortmann, 1891, p. 516. Leander concinnus, De Man, 1897, p. 765. Leander concinnus, De Man, 1902, p. 807. Locality. Station I. A, 23 specimens, the largest 38 mm. in length. Remarks. The rostrum in the majority of the specimens has the form shown in Dana’s figure 106. There is in all the specimens but a single tooth behind.the orbit. This is followed by from 4-6 teeth more or less equidistantly placed on the proximal part of the rostrum. The distal part is upeurved slightly, and usually devoid of teeth with the exception of a small one almost at the extreme apex, giving the latter a bifid appearance. In two specimens the apex is trifid, and in one of the remaining specimens there is a single tooth on the distal part of the rostrum half-way between the proximal teeth and the apex. Below, the rostrum bears from 8-5 teeth. ‘The rostral formula may therefore be represented as follows :— 1+ (4-6) + (0-1) + bifid (trifid) apex, 3-5 This species, like the preceding one, has the palp of the mandible two- jointed. Distribution. New to the fauna of the Red Sea. Fiji (Dana) ; Wast Indian Archipelago (De Man). LINN. JOURN.—ZOOLOGY, VOL. XXXIV, 30 394 DR, W. M. TATTERSALL ON THE STOMATOPODA AND Super-family Crangonoida. Family PROCESSID. Genus NixoipeEs, Paulson, 1875, p. 98. NIKOIDES sp. ? Locality. Station VIII. C, one specimen, 8 mm. Remarks. I am unable to determine this specimen with any degree of certainty, and as the left leg of the first pair and both legs of the second pair are broken it is not possible to give an adequate description. It differs from the description of WN. dane, Paulson, as given by Nobili in the following particulars :— (1) The rostrum is very much shorter, hardly extending as far forward as the proximal margin of the cornea of the eye ; it is acnte with a single small acute dorsal tooth quite near the apex ; on each side of the base of this tooth there springs a single strong seta which extends forward to the apex of the rostrum. (2) The exopod of the first pair of walking legs is very much shorter and barely extends beyond the ischiopodite. (3) There are no spines on the ischius and merus of the last three pairs of legs, which are otherwise in close agreement with Nobili’s description. ; In the last two characters my specimen approaches JV. maldivensis, Borra- daile, 1915, p. 209, but I am unable to institute a comparison with that species in the characters of the first legs, and Borradaile’s description of the rostrum of WV. maldivensis does not agree with what I have observed in the present specimen. The chelate leg of the first pair, which is present in my specimen on the right side, agrees closely with Nobili’s description of that -appendage in NV. dane. Suborder REPTANTIA. Tribe ANOMURA. Super-family Thalassinidea. Family AXIIDA, Bate, 1888. Genus Axtopsis, Borradaile, 1903, p. 538. Axtopsis £THtopica, Nobili, 1904, p. 235. A. ethiopica Nobili, 1906 a, p. 93, pl. 6. fig. 1. Locality. Station V. E, one specimen, 19 mm. MACRUROUS DECAPODA OF THE SUDANESE RED SBA, 395 Remarks. This specimen agrees completely with Nobili’s careful descrip- tion, and I have nothing further to add to his account. Distribution. At present only known from the Red Sea (Nobili). Family CALLIANASSIDA, Bate, 1888. Subfamily UrpoGrBiin &, Borradaile, 1903, p. 542. Genus Upoaesra, Leach. Upocesia (CALLIADNE) SAVIGNYI, Strahl, 1862, p. 1064. U. savignyt, Nobili, 1906 a, p. 98. Localities. Station VII. D, 29 specimens, 5-27 mm. Station I. U, always in pairs, a large ? and small ¢ in yellow sponge, 14 specimens. Station V.G, one specimen, 9 mm. Uncertain. 1 juv., 5 mm., labelled “ Crust. fr. 44.” Remarks. This species seems to be usually, if not always, associated with sponges. Two specimens were infected by a Rhizocephalan parasite on the under side of the abdomen. Upoaesta (UPOGEBIA) PSEUDOCHELATA, n. sp. (PI. 28. figs. 16-22.) Locality. Station VII. C,1 3,6 mm., 1 ¢, ovigerous, 12 mm. Description. The rostrum is quite short and does not extend beyond the eyes. In dorsal view it is triangular in shape with an obtusely rounded apex, and its margins are not provided with teeth or spines. The lateral tooth on each side is almost obsolete, but it marks the anterior termination of a strong lateral tubereulated ridge which runs backward on each side to the well-marked cervical groove. This ridge bears about a dozen small tubercles. The central portion of the dorsal surface of the rostrum and carapace is provided with numerous small obscure tubercles arranged irregularly in six rows, more numerous and prominent nearer the rostrum, becoming obsolete or absent towards the cervical groove. Between the tuberculated portion of the carapace and the lateral ridge on each side, and running paratlel to the latter, is a linear groove or impression, devoid of tubercles and likewise becoming obsolete as it nears the cervical groove. The rostrum itself is provided with a dense tuft or mass of short sete, and the remaining part of the carapace and the body is adorned with scattered hairs of varying lengths. The antennular peduncle (PI. 28. fig. 17) is shorter than the antennal peduncle, equal in length to the first three joints of the latter, and extending beyond the eye by the whole of the last two joints and the narrow distal portion of the basal joint. The third segment is three times the length of the second. One flagellum is thirteen-jointed and the other ten. The longer flagellum is about one-fifth longer than the peduncle, the shorter flagellum equal to the peduncle in length. 30* 396 DR. W. M. TATTERSALL ON THE STOMATOPODA AND The antennal peduncle (PI. 28. fig. 16) is composed of four segments with no prominent spines on any of them, Between the second and third seoments there is the articulated remnant of the antennal scale, consisting of asmall triangular plate with two small apical teeth. The second joint is furnished with a row of very long setee on the whole of its lower margin. This row of setee is continued across the outer face of the third joint and terminates in a dense brush on the upper distal corner of the joint. The first pair of legs (PI. 28. fig. 18) are equal in size and subchelate in both sexes. The merus is equal in length to the propodus and double the length of the carpus. The propodus or palm is three times as long as wide, oblong in shape with parallel sides. The movable finger is about half as long as the palm, and the fixed finger rather less than half the length of the movable one. The fixed finger bears five small tubercles on its proximal half and impinges on a prominent tcoth on the movable finger so that a false chela is thereby formed. There are no prominent spines on any of the joints but the limb is richly provided with setee. The merus has a row of very long setee on its inner and lower margin, while the carpus has a dense fringe of small hairs on its upper margin. The palm has its lower margin fringed with long sete, and there is a dense row of shorter hairs on its outer face besides other scattered hairs. The movable finger is well provided with setee. The second leg (Pl. 28. fig. 19) has the merus about equal to the propodus and carpus combined, the proportions of the joints being merus 3°75, carpus 1°75, propodus 2, dactylus 1. The propodus is two and a half times as long as broad. There are no specially prominent spines on these legs, but the merus, carpus, and propodus are well provided with long setee, The third and fourth legs (Pl, 28. fig. 20) are specially noticeable for the form of the dactylus. Tts outer margin bears five to seven tubercles and the inner margin a dense saw of ahout 14-16 finely pointed teeth on the distal part. The fifth legs (Pl. 28. figs. 21-22) have the usual subchelate arrangement due to a prolongation forward of the lower edge of the propodus. This prolongation is about half as long as the dactylus and terminates in a prominent tooth. The dactylus bears a saw of fine teeth on its inner margin. The telson is broader than long, its lateral margins parallel, the lateral corners evenly rounded and the posterior margin straight. It is of the same length as the uropods. Length of an ovigerous female, 12 mm., of the only male, 6mm. The male specimen agrees fully with the female, but Iam unable to say whether there is the marked sexual dimorphism in size which the two specimens suggest. Among described species, U. pseudochelata approaches most nearly to U. heterocheir, Kemp, 1915 a. The Jatter has the same pseudochelate form ance ASS TSS SS = WMT Del. - =. STOMATOPODA AND MACK mnie Reid, Lith Edint QUS DECAPODA FROM RED SEA. ee ana 7, VWrl yy) 7 Tattersall Journ. Lisw. Soc. Zoon. Vol XXXIV PL.27 WMT. Del. P : JT Rennie Reid Lith Kaine STOMATOPODA AND MACRUROUS DECAPODA FROM RED SEA. STOMATOPODA AND MACRU Journ. Linn. Soc. Zoo. VolLXXXIV PL23 JT.Rennie Reid, Lith Edin? S DECAPODA FROM RED SEA Tattersall Journ. Linn: Soc. Zoou. Vol XXXIV F128 I H WMT Del : ST-Rennie Reid, Lith Edin? STOMATOPODA AND MACRUROUS DECAPODA FROM RED SEA. MACRUROUS DECAPODA OF THE SUDANESE RED SEA. SiS)tl of the first pair of legs in the male only, the female having the first pair of legs simple. In this character the species are distinct. also agree in the character of the dactyli of the third and fourth pairs of lees. U. pseudochelata differs from U. heterocheir in the following points :— (i.) the much shorter rostrum and the almost complete absence of lateral teeth ; (ii.) the presence of tubercles on the rostrum ; (iii.) the relatively longer flagella to the antennule; (iy.) the absence of spines and teeth on the limbs and other appendages; (v.) the different form of the »seudochelate These two species limb and the fact that it is common to both sexes; (vi.) size. Jn general facies and structure the two species are closely allied, but the differences already mentioned may be considered of specific value. EXPLANATION. OF THE PLATES. PLATE 27. Vig. 1. Pseudosquilla megalophthalma, Bigelow. Dorsal view of anterior region. 2, i 5 Lateral view of eye to show the real 7 ’ fs proportion between cornea and peduncle. 3. ' Dorsal view of sixth abdominal seg- te) ? t=} Anchistus tnermis (Miers). o 7. Penaopsis stebbingi (Nobili). 5. Gonodactylus brevisquamatus, Paulson. ment and telson. Chela of leg of the first pair of the male. Dorsal specimen 28 mm. Dorsal view of sixth abdominal seg- ment and telson «f the same specimen, Carapace and rostrum. view of anterior region of 8. Fa . Merus of last thoracic leg of a male to show ; notch. 9. 3 5 Thelycum. 10, as 5 Petasma. 11. Periclimenes calmant, n. sp. 12. Pencopsis vaillanti (Nobili). 3. dthunas crosslandi, n. sp. Lateral view of anterior end. Thelyeum. Lateral view of anterior end to show the relations of the rostrum, extra- and infra- corneal spines, and eyes. Vt; ; 3 Antennular peduncle. Akay, ; 5 Antennal peduncle and scale. 16, ; x Leg of the first pair. 17. ” ” Leg of the third pair. PLATE 28. Fig. 1. Synalpheus quinguedens, n. sp. Dorsal view of anterior end. 2, ” » Large chela. 3. # rf Leg of the second pair. 4, ss i Leg of third pair. or “I STOMATOPODA AND DECAPODA OF THE SUDANESE RED SEA. Synalpheus quinquedens, n. sp. Lysiosquila multifasciata, W.-M. Harpilins depressus (Stimpson). beauprest (Aud.). 3 gerlachet, Hippolyte proteus (Paulson). ” ” ” ted Pencopsts stebbingt (Nobili). ” ” ” ” Upogebia pseudochelata, n. sp. ” ’ ” ” ? ” ” b} ” ’ ” 3 ’ Athanas dimorphus, Ortmann. ” ” Athanas djitoutensis, Coutiére. Dorsal view of sixth abdominal segment and telson. Raptorial claw. Third leg (distal joints) to show the form of the dactylus. Distal joints of third leg to show form of dactylus, Distal joints of third leg to show form of dactylus. Rostrum. Antennal scale. Carpus and chela of the second pair of legs. Lateral view of petasma to show teeth. First leg. Second leg: Antennal peduncle. Antennular peduncle and flagella. Leg of the first pair, Leg of the second pair. Leg of the third pair. Leg of the fifth pair. Distal end of leg of the fifth pair. Chela of the right leg of the tirst pair in the male. Chela of the left leg of the first pair in the male. i Chela of the right leg of the first pair in the male, ON THE SELENARIADA, CONESCHARELLINIDA, ETC. 399 Observations upon the Relationships of the (Bryozoa) Selenariade, Conescharellinide, etc., Fossil and Recent. By Arruurn Wu. Waters, ince GS. (PLATES 29, 30.) [Read 19th June, 1919.] CONTENTS. Page IINDRODUCTIONM sertacitertsimtterure fuels utciian eaters cle st. steer ace 599 CUplGnuuicanaiienstsw DUS carat ayers eke siete eet ete teat 410 5 ower wD USKes. wares ect Badsyaiere de lactertve ota reae ah eteree av eee Bes 412 + LORTISONU NS USI aes gecacctetsac tate 4 waste ncet tes sadats renee dha seis aonianeve 413 iD UIMNOCUALG MMELTANCEL, a0) ste etarei ve severe aieveuemirtuesel soe seis oue 414 ISCLEN CIN CONCUIUI Gael ms NVCOOG Se ate) cast edajatanate cus 2tete avai eieieMeviy slays. saa Patel 415 DMM DUTCLELCLs Lisa WV OOG Sew eyec tot syohe euchsis cstaley such slaves ple ops aTeloy cues 416 rf GA CULAEH ASUS). monetegtienn te states vere S irscataee i atic aabonder dualons 417 TAU ULLESTCU) NLU SIS US ln muusntetitets steps etaeensesfotsecerie cea om aeraret Ms fo oe ere = 417 Lrochoporay VOrDiony Wye rasta tage sis ew aisesaccateusaasateaee Hess ene aes 418 Hehodonua anuyreeatas, CAlVetie is terse )steiste'e cles state's sleet 25 vl hc ae 419 GoneschapellmarcancellatG,wOUske aie cerelatsrsterereca eraiestere eiccs «tees thes 419 PINUeppurertsts ) THUS es sceee sees atersceiscerstosele ates eieesessctaie' sr 420 "7 B/labellai;ts = ne VINSC Mata ye ogni emer si tee apes pean 2 5 GNGULO DONC Vem VO OCS trereratenstjsanapetey= eietersieieis svayreete 422 5 CONIC LASWELL eesycie ttateit not crees tye Fotis anee srcvenmia ays 425 fi eoccend,| Nevianin sai etc tw yi an vars on canes enatntaey « 424 * In preparing a list of Bryozoa from Oran (Algiers) a few specimens of Cupularia Johnsoni, Busk, have led me to more extended studies for a better understanding of the whole of the group ; however, a large number of spirit or fresh specimens are required for a complete study of this little understood group, which is one of the most interesting. This communication makes no pretension to dealing exhaustively with it, being principally remarks on some species in my collection. I had previously received, from the late Canon Norman, a few spirit specimens of three species of Cupularia, collected off Madeira, namely, C. canariensis, Busk, C. Johnsoni, Busk, C. Lowet, Busk, each of which revealed new structures, in stained preparations and sections. Of what have been considered as Selenariade all are more or less cup- shaped or flat, and though now the shape alone would not be considered * Shortened references are used in the text for the sake of economy, and the explanations of these contractions are given on p. 409. 400 Mr. A. W. WATERS ON THE RELATIONSHIPS sufficient for classification, yet it is important to see how far other characters run through all or most species. The basal* structure of Cupularia canariensis has a series of parallel chambers (Pl. 29. fig. 1) filled with a granular substance and having a connection from each chamber to its neighbours through rosette-plates. The lateral walls are parallel with the axis of the zoarium. These chambers are partly shown by Busk in his L. canariensis, ‘Crag Polyzoa,’ pl. 2. fig. 2 e, and would have been called kenozocecia by the late Professor Levinsen, while Reuss and others have named the chambers at the base of Batopora “abortive cells.’ In Cupularia the walls of these chambers, as well as those of the zoccial and vibracular chamhers, are lined by large square flat cells with a small nucleus (Pl. 29. figs. 2&3). These large ceils seem to cecur generally in the Selenariade. In C. Lowei, Busk, looking at decalcified and stained preparations, with the opercular wall in focus, a number of small bundles of muscles are seen, which pass from the frontal membrane through the large frontal calcareous pores (PI. 80. fig. 1). Similar muscles occur generally in the Microporide. Looking at the same preparation from below, the zocecial chambers are seen separated by a considerable free space, and to about the median line of the zocecial chamber there is a row of singlet muscles, or sometimes a small bundle, passing from the lower surface of the zocecial chamber to the lower surface of the zoarium (PI. 30. figs. 2. & 3). No structure at all comparable to this has been mentioned as occurring in the Bryozoa ; however, the basal calcareous wall gives an indication of their occurrence, by a groove along the line of the muscles, with a pore at the end (PI. 80. fig. 4) ; and there is no doubt that fossil C. wmbellata, Defr., and some other species, had the same muscular arrangement. [urtber examination of spirit specimens may reveal points which could not be distinguished in the very limited material available. Whether a new genus should be made before more material and species have been examined is perhaps an open point, though it does not seem that species haying such different structures as C. eanariensis and C. Lowei can remain together. If found advisable, C. Lowe’, C. umbellata, etc., might be * T shall throughout the paper, so far as possible, consider the colony as seen in tie position in which it first grows on its support, so that the opercular wall is the upper surface, while the lower or base rests upon the support in the first stage. In doing this no opinion is pronounced on the suggestion that some mature forms may live in a reversed position, and for the same reason the base of the cone is always shown below, even though possibly it might be more correct to reverse it in the mature forms of some species. t Ina very good specimen the threads to the dorsal surface are fine lines, about the same size as the frontal opesial muscles, but in a specitaen in which some changes have taken place they are large and alternately light and dark, that is striated, which may be a histological change. At each end the muscle widens out.(as in Pl. 30. fig. 28). OF THE SELENARIADA, CONESCHARELLINIDA, EYC. 401 placed in a new genus. Itisthe Miscoporella of V@Orbigny with C. umbellata as the type, but the name having long been used, in mistake, for a Cyclostomatous genus it is better dropped. These two structures having been briefly referred to, the Selenariadee can be considered, and an attempt made to understand the relationship and limitations of the group. In al! the growth is from the apex, and in Cupularia, Selenaria, and Lunulites it is known that under the apex there is often a grain of sand, a flake of shell or stone, or very frequently the shell of a foram. This was, so far as I know, first pointed out by Defrance, but has since then been frequently and almost amusingly rediscovered, being alluded to by Michelin, Goldfuss, Reuss, Stoliezka, Busk, Beissel, Gabb & Horn, Waters, Maplestone, Robertson, Canu, and Harmer, though frequently it has been thought to be a characteristic of some one species or genus (see Pl. 29. figs. 1, 7, 8, 9, 10, and Pl. 30. figs. 10, 12). The calcareous shell of the Selenaridan colony usually entirely encloses the hard base on which the growth started, showing no sign of the support either above or below, though sometimes the growth is on a much larger stone, as in some specimens of Cupularia canariensis from Petit Tahou, Liberia (PI. 80. fig. 12). It would seem impossible for a colony so heavily weighted to float, nor can we think it could float in a reversed position. In Vibracella trapezoidea, Reuss, there is in several specimens a small projection on the under surface below the apex, caused by the presence of a very young foram, but a few are found growing on much larger specimens of this Orbitoides stellata, Giimb., or, as Iam told it ought now to be called, Ortho- phragma (see Pl. 29. fig. 10). This Vibracella grows on the foram, and then grows beyond it, like an inverted cup, though in the specimen figured perhaps simultaneous growth of both organisms had taken place. When I first. described * WVibracella trapezoidea, Reuss, only small flat pieces were available, but since then a number of disk-like forms, from Bocea di Sciesa, Colle Berici, N. Italy, have been cleaned up, closely resembling in size and shape Cupularia canariensis. From the same locality I bave a few specimens of Mamillopora bidentata, Rss., also with the Orbitoides attached to the lower surface, but not centrally, nor have the majority anything attached, so that the few cases are probably accidental. No sand or other support has been mentioned as occurring in Mamillopora, Batopora, or Conescharellina, Returning to Cupularia and Selenaria, the larva settles on the support mentioned, and round the primary zocecia there are in Cupularia, and usually in Selenaria, eight zocecia which are much smaller and shallower than the peripheral zocecia, and often in these smaller zoccia there is no sign of ‘there ever having been an opercular opening. In’ Lunulites there are * “North Italian Bryozoa,” Quart. Journ. Geol. Soc. vol. xlvii. p. 11, pl. 1. fig. 23 (1891). 402 MR. A. W. WATERS ON THE RELATIONSHIPS six or eight zoccia surrounding the primary. The growth in the surrounding zoccia in Cupularia is shown in my Pl. 80. fig. 11, and Busk has figured them in C. Johnsoni *, Busk; Selenaria maculata seems to have six zocecia round the primary f. Canu mentions eight zocecia round the primary in Lunulites levigata, Canu ; but some Lunulites as L. sella, Marsson and L. Gold/ussi, Hag., have only six. Ihave figured Selenaria concinna, Busk (P1. 80. fig. 9), showing only five zocecia, but as there are three vibracular chambers these may be taken as representing three zocecia. There are various Bryozoa in which the primary is surrounded by eight zoccia, as Flustrella hispida, Fab., from Oban, and as shown by Barrois there are eight zocecia round the primary iu t Microporella impressa, Aud., and six in § Membranipora pilosa, L. in AM. lineata, L., there are also six, as is the case in M. nitida, Johnst., M. Dumerilii, Aud., AL. tenuirostris, Hincks, Microporella diadema, MacG., M. ciliata, Pall., from Oban, and Jf, Aalusii, Aud. These may be taken as typical figures, even if there is some variation. In Cupularia, Selenaria, and Lunulites, the younger zoccia, sometimes even to three rows round the primary, have a calcareous perforated front, just as we have seen in the ordinary zocecia of Cupularia Lowei, covered on the front by a membrane, to which bundles of muscles passing through the frontal pores are attached to draw it down, so that the structure of these is Microporidan, while the outer zocecia have a Membraniporidan operculum. : In some Selenariadee the opercular opening is more or less closed by a tongue-like calcareous projection to which I have referred as occurring in || Selenaria petaloides fossil from New Zealand, but in S. concinna (Pl. 80. fig. 9) it commences near the proximal end until at last the aperture may be entirely closed. It seems that this closure must be compared with those which I {{ have described over the operculum of Schizoporella unicornis and other Schizoporellide, though it takes a rather different form. Maple- stone ** has already referred to it. * Quart. Journ. Mier. Soe, vol. vii. pl. 23. fig. 3. + I have made sections of the small (3-6 mm.) fossil Cedlepora globularis, Rss, from the Bartonian of S, Urbano di Mte. 5. Greve, Vicentine, a species described by Reuss from Val di Lonte, but in his first description other species were included. [Enclosed in the cclony, but not in the middle, is a grain of sand without any small zocecia on it, so that there is no reason for considering that it grew upon the sand or for making any comparison with Selenariade. There are two small oral avicularia in the peristomial tube, and this shows it is either Cellepora or Lagenipora, but no ovicells have been seen. { Emb. des Bry, pl. 16. fig. 2 (1877). § Loc. cit. pl. 15. fig. 9. || Quart. Journ. Geol. Soc. vol. xxxix. p. 442, pl. 12. fig. 11 (1883). 4] Journ. Linn. Soc., Zool. vol. xxxiv. p. 15 (1918). ** “ Vict, foss. Selenar.,” vol. xvi. p. 217, pl. 25. fig. 8 (1904). OF THE SELENARIAD#, CONESCHARELLINIDA, ETC. 403 APPENDAGES. The appendages of the group have not received the attention they deserve, and avicularia have frequently been in a wholesale way called vibracula asin Conescharellina, which always has avicularia. The vibracula of Selenaria are of a type quite different from anything known elsewhere, for they move in a ring external to the vibracular chamber. In Lunulites there are various forms of appendages all of which have been generally spoken of as vibracula, whereas vibracula only occur in a limited number of forms in the family, such as Vibrucella trapecoidea, Rss., and Lunulites mitra, Marsson (probably Vibracella). There are ‘‘ onycho- cellaires”’ as figured by Beissel in L. Groldfussi,* Hag., also they occur in L. sella, Mars., and L. salebros, Mars., and probably in L. Beisseli, Mars., in which there is an elongate chamber, usually broken down. L. cupulus, Busk =L, gibbosa, Busk, also closely related to L. patelliformis, Maplestone (non Marsson), has a long tapering seta divided at the end (PI. 380. fig. 16), which is different from anything we are acquainted with, and probably must be considered as avicularian. Lunulites of the radiata-type are the most common, and the avicularian chamber has a projection on each side, but being rarely well preserved they have consequently seldom been sufficiently figured. ‘he name Lunulites will probably ultimately be confined to this group, which will include such species as L. Hagenowi, Bose (non Beissel), L. transiens, Gregory, &e. I+ have on several occasions said that the real difference between avicu- laria and vibracula consisted in the ayicularia only having motion in one direction, whereas vibracula have motion in all directions, and we must look at the base of the appendages, or the chamber containing tlie muscles, Kc., to decide which of the two we are dealing with; whereas the length of the appendage has little classificatory value in the Cheilostomata, for short, round, or triangular avicularian mandibles may be replaced by whip-like mandibles, as in Microporella coronata, &e. In Cupularia canariensis the central setee are very short and small, though quite mature, increasing in size to the periphery, where they are very long. Correct appreciation of the differences between the two organs would haye led to their being more considered in classification in the present group, as well as in other recent and fossil forms. Conescharellina has avicularia, Cupularia and Selenaria have vibracula, Lunulites, as it has been understood, * “Ueber die Bry. der Aachner Kreidebildung,” pl. 2. figs. 22, 24. Nat. Verh, Holland, Maat. Wetenschappen, vol. xxil., Haarlem 1865. + “Bryozoa,” Résultats du Voyage du 8.Y. ‘Belgica’ in 1897-8-9, p. 27 (1904). 404 MR. A. W. WATERS ON THE RELATIONSHIPS has avicularia, vibracula, and onychocellaires. Mamillopora probably always has avicularia. In Cupularia, the museles are frequently attached to the vibracula by broad bands (PI. 30. fig. 29), instead of by narrow tendons as is generally the case in avicularian mandibles. OVICELLS. No ovicells are known in Cupularia, but ovaria and embryos are found in the zowcia near the periphery (Pl. 80. fig. 25), nor are any known generally in Se/enaria, though the late C. M. Maplestone, in a letter, wrote that in specimens from Queensland he has sven the ovicells of S. coneinna resembling those of Conescharellina philippinensis. In Conescharellina philip- pinensis there is a raised globular ovicell as described by Whitelegge *, and I have some in my collection, though, out of some hundred specimens, ovicells have only been seen in two or three. The ovicell, like the oper- culum, is directed to the centre of the zoarium in a puzzling manner. Conescharellina eocena, Neviani, also has ovicells. In Mamillopora simplex, Kosch., the ovicell is a raised inflation and is widely open, as in Haloporella. In Orbitulipora petiolus, Lonsd., the ovicell is also globose. Smittt figures it in Mamillopora cupula, Sm.; Reuss and I have described it in Batopora multiradiata, Rss. “ Cupularia bidentata,” t Rss., also shows an ovicell, but from the figures it was not clear to what genus it belongs, but now it is found to be Mamillopora, closely allied to M. simplex, Kosch. Most of the published figures of the earlier authors gave the Selenariade upside down, whereas in such genera as Cupularia, Selenaria, Lunulites, the distal end with its Membraniporidan aperture should be shown at the top. Maplestone |], when dealing with Conescharellina, says, in accounting for the position, “the free edge of the operculum is directed towards the apex, but it is not the distal edge. The fact is that the operculum is hinged at the distal end and not at the proximal one; so that in the conical forms not only are the zocecia upside down but the operculum is also upside down.” Difficult as it is to understand this reversal it seems to be the case in Con- escharellina, but not in Cupularia, Selenaria, or Lunulites. Whitelegge §, Haswell 4, and Maplestone**, call attention to the semi- * “ Australian Poly.” p. 842. 1887; see also Maplestone’s “ Bipore,” p. 6, pl. 1. fig. 2 (1910). + Floridan Bryozoa, pl. 2. fig. 33; pl. 7. figs. 146, 147 (1873). } Waters, ‘ Batopora,” p. 87, pl. 6. figs. 7, 11 (1919). | “ Growth and Habits of Bipore.” § “ Australian Polyzoa,” p. 389. 1877. 4 “ Polyzoa from the Queensland Coast,” Proc. Linn. Soc, N.S. W. vol. v. p. 42 (1881). ** (© Bipore,” p. 5. 1910. OF THE SELENARIADA, CONESCHARELLINIDA, ETC. 405 lunar slits * which seein to occur quite generally in C. philippinensis, and at any rate in most species of Conescharellina. They are found in quite young zoaria, and in older ones they may occur in various positions ; sometimes in acircle, that is at equal distance from the apex. I have seen in a colony two such circles with several slits, sometimes the slits oceur along the radial line of the zoaria, in others between two radial lines, Whitelegee’s theory, of these slits being rudiments of intercalated new zocecia, was declared by Levinsen to be practically impossible, and from the position of these slits it does not seem that they can indicate new zowcia, also the chambers are smaller and simpler than those of the zocecia, nor are there so many connections as in the zoocia. Having a considerable number of good dry specimens of C. philippinensis, it was hoped that examination, ani sections in various stages, would give an explanation of their function, but this was not easy, though the explanation now offered will, I fully expect, be confirmed when living or spirit specimens are examined. Conescharellina angulopora (Woods) and C. flabellaris, Lev., have a pro- jecting growth at the apex (Pl. 29. fig. 16; Pl. 30. fig. 19) giving much the same appearance as the basal end of Cellaviu and Tubucellaria, both of which are directly attached by tubular radicles to the substratum. The slits, if my theory is right, indicate the radicle chamber; and in a large number of Bryozoa radicle chambers may oceur with great frequency, even to each zowcium, though the radicles may be developed in very small numbers. In a previous paper allusion has been made to the radicle chamber in Catenaria Lafontiit, Aud., in which a round spot, on the dorsal surface, was shown to each zoucium by Savigny, Busk, and others, without it being suspected that this was the opening of a dorsal radicle chamber, until I found a specimen with a few radicles,—althouzh I have seen only a few, since they are very rare. When a radicle is formed in Conescharellina probably the disk closing the chamber is absorbed, and then a semicircular opening occurs as in Pl. 29. fig. 19.s.s. Levinsen, p. 310, says “these superficial chambers seen in the whole colony seem to be in mutual connection with each other.” They are in communication through rosette-plates. Some rather important suggestions { have been made that the mature * Levinsen proposes to speak of luncecia instead of semilunar slit, but when we have a good name, why do we want to change it, even before the function of the structure is understood? We must resist the tendency to change the name of each minute structure we find. Canu and Bassler say that the lunccia are openings of ‘special compensation zomciules”: no proof is given, and I doubt whether it will be found to be the case. See Early Tert. Cheil. Bry. p. 76. 1917. + “Bry. from Rapallo, &e.,” Journ, Linn. Soe., Zool. vol. xxvi. p. 15 (1896), t Maplestone, “ Bipore,”‘p. 8. D'Orbigny, Pal. Frang. p. 447, suggested that the young colonies of Conescharellina were perhaps fixed by the conical extremity, and in many cases, at any rate, this seems the most probable, 406 MR, A. W. WATERS ON THE RELATIONSHIPS zoarium floats with the apex at the bottom, and we are awaiting proofs as to how this takes place, as it is difficult to understand. Whitelegge*, in a postscript, says that he has had C. philippinensis under examination, and that “nearly every specimen possesses a pair of tubular filaments inserted on each side of the zoarium’’; also he thinks ‘the tube appears to grow out of an aviculurium.” Wil] it not be found that they grow from the semilunar slit ? Harmer +, speaking of the Selenariadze, says that he has some evidence . that they may be attached to the ooze by means of very delicate flexible rooting processes, but he does not indicate the species or genus to which he is referring. From fossils and dried specimens of Conescharellina the con- clusion come to independently, is that there have been radicle processes, but in Cuputaria, Selenaria, and Lunulites there is no indication of anything similar. CLASSIFICATION. We may now turn to he classification, as these investigations were made to find a natural one, and certainly we are brought up against a most difficult problem. Generally in Cupularia, Selenaria, and Lunulites there is a thick under surface, through which long tubes may pass, or there may be a series of chambers; further, the arrangement of the back as well as the front is radial, all of which seems to be quite different from anything known in other.genera. In many cases in the Cheilostomata, such differences as whether the zoccia are uni- or bilateral, or whether they are adnate or erect, are purely zoarial characters, of no or but slight value in classification, but the characters on the under surface, now dealt with, are not zoarial in the same sense, but are in connection with the zocecia. Looking at Cupularia and Selenaria with their similar opercula, similar lower surface, in most cases with central small zocecia, often closed, with vibracula in both genera, there can be no hesitation in placing them in the same family ; in Lunulites, the lower surface is radial and thick, long pores pass through it as in the last two genera, the opercula are similar, but there is a more solid caleareous frontal wall without perforations, though it clearly belongs to the same family. Lunulites, as generally understood, requires separating into several genera, as already indicated. Conescharellina differs from Cupularia and Selenaria in many particulars, such as the shape of the separable operculum, the reversal of the position of the operculum, the semilunar slit, absence of radial under surface, although the zocecia are placed radially ; on the other hand, in the lower part there are anumber of vertical chambers, which seem comparable with the horizontal ones of Cupularia canariensis, and it would be strange that species not * “ Australian Polyzoa,” p. 347, 1887. } Presidential Address, Brit. Assoc. Zool. Section, p. (9). 1908, OF THE SELENARIAD®, CONESCHARELLINID®, ETC, 407 closely allied should develop into a conical form with small central zocecia, and should have a series of chambers below the zoccia. Put shortly, are there a number of Bryozoa from different families with quite different zocecial characters which have taken on the same way of growth and sub- basal characters, or have related forms with similar growth gradually assumed more divergent characters ? Batopora has small chambers at the lower part, which Reuss and Stoliezka call abortive cells, and these may be homologous with the chambers of Cupularia canariensis and Conescharellina. Gregory * would place batopora under Conescharellina, but Batopora has an oral aperture with a more or less straight lower edge, and a large widely open ovicell—in fact, in most characters it resembles Holoporella, showing no reversal of the position of the zowcia, also the zoaria are more globular than conical. I am not sure that I understand what Gregory meant about the aperture of Conescharellina clithridiata, Greg., which seems to be Holoporella or Cellepora. Trochopora, VOrb. has the whole of the base filled in by a enleareous growth through which pass very long pore tubes (PI. 80. fig. 17). This character may not be universal and is discussed on page 418. The base has radial divisions just like those of Lunulites, Cupularia, &e., and in making sections these radial divisions are seen at every stage of the preparation (P]. 29. fig. 18). A specimen of Selenaria nitidat, Maplestone, in the British Museum, from 22 miles EH. of Port Jackson, has the cone entirely filled in with a solid calcareous mass, and in some cases shows radiating lines at the base just as in Trochopora. Canu & Bassler { place Trochopora and Otionella, Can. & Bassl. under Membraporina §, which belongs to Malacostega, but Lunuaria they place with Opesi lidee, that is Coilostega. Tf there were sufficient reason for this it would indicate that the new elassification has some weak points, but I fail to find sufficient grounds for separating Tro- chopora from Lunulites. My conclusions are, that of the forms with discoid or cupuliform growth there are two main divisions with one subgroup. (1) Those with the operculum in the frontal membrane, a radial base, and usually no ovicell, including Cupularia, Selenaria, Lunulites, with Selena- riopsis, Maplestone, Trochopora, d’Orb., Otionella, Can. & Bassl., Heteractis, Can. & Bassl. (2) Those with usually a fairly large operculum fitting into the Lepralioid or Holoporellidan oral aperture; with a large ovicell widely open in front, as * “Brit. Paleog. Bry.,” Trans, Zool. Soc. London, vol. xiii. pt. 6, p. 251 (1893). + “Results of Deep Sea Investigations,” Records of the Australian Museum, vol. vii. p- 271, pl. 77. fig. 8 (1909). { “Tert. Cheil. Bry.,” U.S. Nat. Mus. Bull. 96, p. 10 (1917). § Levinsen, ‘‘ Morph. Cheil. Bry.” p. 144. 1909, 408 MR. A. W. WATERS ON THE RELATIONSHIPS in Mamillopora*, Spheropora, Haswell +, Kionidella, Kosch., as well as Batopora and Stichoporina t, Stol. (non Kosch.) with smaller oral apertures. At the base the zoaria may be radial, or there may be a growth of zocecia over the radiate zocecia §. (3) Asa subgroup, those with a small, nearly oval oral aperture, with a separable operculum, having muscular dots fairly near together. The zocecia are directed to the apex, there are semilunar slits and a small raised globular ovicell, the avicularia are adventitious, whereas in Cupularia they are vicarious. Only Conescharellina is known in this subgroup, The first division, if only the zocecial characters were considered, would come under the Membraniporidee of Levinsen||, which is a very large family, or rather a casual ward for the homeless. Canu{] places Lunulites under Onychocellidee, but only a part have ? while some have vibracula, and others have avicularia, “ onychocellaires,’ as proved by the symmetrical avicularian chambers ; Cupularia he places under the Opesiulidze, in which family he puts Micropora, as he recognized in the fossils, what most others had not appreciated, that the frontal pores, where they exist, are opesiules. The muscles passing through these pores have now been seen in my stained preparations. Besides the genera mentioned, the question of the relationship of the following must be considered :— 1846. Stichopora, Hagenow. 1847. Prattia, d’ Archiac, Mém. Soc. Géol. de la France, vol. iii. p. 407. 1851. Discoporella, d’Orb., is Cupularia. 1851. Discoflustrella, VOrb., is Cupularia. 1851. Discoflustrellaria, d’Orb., Marsson says only worn Lunulites. 1851, Cymbalopora, Hag. This is much like Conescharellina. The under- side has been mistaken for the upper side. 863. Discoescharites, Roemer, Nord. deutsch. Tert. Poly. p. 21, is Lunulites. 864. Bicupularia, Reuss, Sitzb. Akad. math.-naturwiss. vol. 1. p. 9, pl. 3. fig. 2 (1864). 1882. Ascosia, Jullien, must be placed under Mamillopora. 1887. Bipora, Whitelegge, is Conescharellina. * See Neviani, “ Nuova specie fossile di Stichoporina,” Riv. Ital. di Paleont. pp. 1-4. 1895. Proc. Linn. Soc. N.S. Wales, vol. v. p. 42, pl. 8. figs. 5, 6 (1880). Stoliezka, “ Oliz. Bry. von Latdorf,” Sitz. Akad. der Wissensch. Wien, vol. xly. p 92. This group is dealt with more fully in Waters, ‘ Batopora and its Allies,’ Ann. Mag. Nat. Hist. ser. 9, vol. iii. 1919. | “ Morph. Cheil. Bry.” p. 148. @ “Bry. foss. de Argentine,’ Anal, del Museo Nacional de Buenos Aires, vol. xvii. p- 275 (1908). at + / OF THE SELENARIADA, CONESCHARELLINIDA, ETC. 409 1893. Biselenaria, Gregory, nom. nov.= Diloptavis,* Reuss (a name pre- ? oD ? ? ‘e viously employed elsewhere). From Reuss’s figure this looks like Cupularia with the frontal growth continued on the under side of the zoarium, meeting in the middle. Spheropora fossa, Haswell, in the same way, grows from the upper side over to the under side (see ‘* Batopora,” p. 80). 1897. Ennalipora, Gabb & Horn. Tt is put by Yves Delage (with a?) in the Selenarina, but this is clearly a slip. 1913. Selenariopsis, Maplestone. Species of these groups have been dealt with in a large number of works, but special reference is made to structure or relationship in the following :— Busk, Brit. Mus. Catal. pt. 2, p.97. 1854. + (B.M. Cat.) Busk, Fossil Polyzoa of the Crag, p. 78. 1859. t (Crag.) Stoliczka, “ Oligoc, Bry. von Latdorf,” Sitz. Ak. der Wissensch. Wien, vol. xlv. p. 71. 1862. Tenison Woods, “On some Recent and Fossil Species of Australian Selenariade,” Trans. Phil. Soc. Adelaide, vol. iii. p. 1. 1880. + (Selenariadee.) Waters, ‘“Bryozoa from N.S. Wales,” Ann. Mag. Nat. Hist. ser. 5, vol. xx. pp. 199-202. 1887. + (N.S. Wales.) Whitelegge, ‘Notes on some Australian Polyzoa,” Proc. Linn. Soc. N.S. Wales, vol. ii. pp. 337-347. 1887. + (Austr. Poly.) Maplestone, ‘Notes on the Victorian fossil Selenariide,” Proc. Roy. Soc. Vict. N.s. vol. xvi. p. 207. 1904. + (Vict. Selenar.) Maplestone, ‘On the Growth and Habit of Bipore,”’ Proc, Roy. Soc. Vict. N.s. vol. xxiii. 1910. + (Biporee.) Maplestone, “The Exped. of H.M.C.S. ‘ Miner’: Polyzoa,” Rec. Austr. Mus. vol. vii. 1909. Maplestone, Supplement, op. ect. vol. viii. 1910. Waters, ‘ Batopora (Bryozoa) and its Allies,” Ann. Mag. Nat. Hist. ser. 9, vol. iii. p. 79. 1919. + (Batopora.) The following also are referred to under the contractions + given :— Cann, “Contributions a l’étude des Bry. foss. des Terrains du Sud-Ouest de la France,” 3ull. Soc. Géol. de France, various volumes. + (Bry. foss. France.) Levinsen, Morph. and Syst. Studies on the Cheil. Bryozoa, 1909. + (Morph. Cheil.) Manzoni, Bri. foss. del Mioe, d’Aust. ed d’Ungheria, Wien. Denks. xxxvii. 1877. + (Mioc. d’Aust.) Manzoni, “ Bri. plioe. Ital.,” Sitz. Wissensch. Akad. Wien, lix.txxi, 1869-71. + (Bri. plioce. Ital.) Norman, “ Polyzoa of Madeira and neighbouring Islands,’ Journ, Linn. Soc., Zool. vol. xxx, + (Poly. Madeira.) Waters, “ Bryozoa from N.S. Wales,” Ann. Mag. Nat. Hist. ser. 5, vol. xx. + (N.S. Wales.) * “Ueb. einige Bry. aus dem deutsch. Unterolig.,” Sitz. Akad. Wiss. math.-nat. Cl. vol. ly. o.7 p. 231, pl. 2. figs. 5-7 (1867). + Contractions used in the text. LINN. JOURN,—ZOOLOGY, VOL. XXXIV, oe 410 MR. A. W. WATERS ON THE RELATIONSHIPS CUPULARIA. The name Cupularia was first suggested by Lamouroux for the fossil “ Lunulites urceolata” Lamk., which Lamouroux subsequently figured for the first time, though neither from the description nor figure will it ever be possible to know what species was intended. D’Orbigny, accepting Lamouroux’s suggestion, described the genus Cupularia, and Busk, apparently forgetting that d’Orbigny had done so, again introduced it. Canu * figures a species as Limulites urceolata, Cuv. and says that this Paris basin species has been known for a century by geologists as L. wrceolata, and so long as it is not considered as the species of Lamouroux, this seems to cause least difficulty, though Cuvier gave no description, but then we must not call L. wreeolata, Cuv., as figured by Canu, a synonym of ZL. urceolata, Lamk. and Lamx., which it certainly is not. Lamouroux considered his species to be the unfigured L. wreeolata, Lamk., and his description is verbally copied from Lamarck, to whom Lamouroux submitted much of his material. Dr. Alice Robertson describes + Cupularia under incrusting Cheilostomata, and says “touching the substratum only on the rim of the colony.” It is certainly incrusting in its earliest stages, but there is never an attachment confined to the rim. Nearly all forms of Bryozoa are incrusting in the earliest zocecia, but beyond this, Cupularia is not so in the sense in which it has been used for many Cheilostomata, 2 Cupularia is represented at present by a few species from temperate and tropical regions, and was abundant in the European tertiaries, with some in the cretaceous formation. CUPULARIA CANARIENSIS, Bush, (PI. 29. figs. 1-5: Pl. 30. figs. 11, 12, 21, 22, 25.) Cupularia canariensis, Busk, Q. Journ. Mier. Se. vol. vii. p. 66, pl. 28. figs. 6-9 (1859); “Orage,” p. 87, pl. 15. fig. 2 (1859); de Angelis, “ Anthos. y Brios. plioc. de Cataluna,” R. Acad. de Cien. y Artes de Barcelona, p. 53, pl. B. figs. 6-9 (1900); Waters “N.S, Wales,” p. 201 (1887); Waters, Zool. Chall. Exp. Suppl. vol. xxxi. p. 36, pl.3. fig. 2 (1889); Q. Journ. Geol. Soe. vol. xli. p. 808 (1885); Robertson, “Ine. Cheil. Bry.,” Univ. California Pub. Zool. vol. iv. p. 314, pl. 24. figs. 90, 91 (1908); Manzoni, “ Bri. Plioc. Ital.” pt. i. p. 10, pl. 2. fig. 17 (1869); “ Plioc. sup. de Rhodes,” Mém. Soc. Géol. de France, ser. 3, vol. i. p. 67 (1887); Seguenza, “ Formaz. Terz. Reggio,” p. 371 (1879); Pergens, ‘“ Plioc. Bry. von Rhodos,” Ann. k.k. nat. hist. Hofmus. vol. ii. p. 31 (1887); Neviani, ‘Cont. alla. Conose. dei Bri. foss. Italiani,’ Bull. Soc. Géol. Ital. vol. x. p. 130 (1891) ; “ Bri. foss. della Farnesina,” Pal. Ital. vol. i. p. 101 (1895); “ Bri. Neoz. di aleuni Loe. d'Italia,” Bull. Soe. Rom. per gli Stud. Zool. vol. iv. p. 238 (14), p. 243 (1895); op. cit. vol. v. p. 121 (1896) ; * “Bry. des Terrains Tert. des Environs de Paris,’ Ann. de Paléont. vol. ii. p. 26, pl. 4. figs. 4-8 (1907). + ‘“Incrust. Cheil. Bry.,” Uniy. California Pub. Zool. yol. iv. p. 314, pl. 24. figs. 90, 91 (1908). OF THE SELENARIAD®, CONESCHARELLINIDA, ETC. 411 op. ett. vol. vii. p. 88 (5) (1898); Bri. Neog. della Calabrie, p. 168 (1900); Bri. Terz. ed Postterz. p. 362 (1900); “ Bri. foss. di Carrabare,” Bull. Soe. Géel. Ital. vol. xxiii. p. 552 (1905) ; Canu, Bry. foss. de Argentine,” Ann. del Mus. Nac. de Buenos Aires, vol. xvii. p- 275. pl. 5. figs. 8-10 (1908); “Bry. foss. France,” vol. xiii. pp- 124, 128 (1913); op. cit. ser, 4, vol. xvi. p. 187, pl. 3. figs. 4, 5, 6 (1917). Membranipora canariensis, Smitt, Floridan Bry. pt. 2, p. 10, fies. 69-71 (1873). Cupularia guineensis, Busk, B. M. Cat. pt. 2, p. 98, pl. 114 (1854); Zool. Chall. Exped. vol. x. p. 206, pl. 14. fig. 6 (1884); Norman, “ Poly. Madeira,” p. 289, pl. 37. figs. 2-6 5 (1909) ; Osburn “ Bry. of the Tortugas Isl.” Pub. Carnegie Inst. 182, p. 194 (1914). Lunulites canariensis, Manzoni, ‘* Mioe. d’ Aust.” p. 72 (24), pl. 17. fig. 56 (1877). Cupularia canariensis, as shown by stained sections, has at the base a series of parallel chambers (PI. 29. figs. 1, 2, 4) filled with granular substance, and with a connection from each chamber to its neighbours, through rosette- plates (see page 400). These chambers must surely be homologous with those of Conescharellina, as seen in C. philippinensis, angulupora, ete., even though the shape is somewhat different. In whole stained preparations of C. canariensis these chambers can be seen at the base forming squares or rectangles (Pl. 29. fig. 5), yet in many cases the calcareous zoarium shows no sign of these squares, but only the radial lateral ridges meeting in the centre of the lower wall. Jareful examination, however, often shows the cross lines in some places, where only the more distinct radial lines are seen at first, and in some fossil forms the squares can clearly be distinguished, so that this is a character which must be dealt with cautiously. No ovicells are known in Cupularia, but there are ova and large embryos shown in sections, from which it is clear that they emerge directly without passing into any external ovicell. In C. canariensis there are 14 tentacles, the same as in C. Johnsoni. I have on several occasions maintained that C. canariensis and C. quineensis are synonyms, and Norman*, agreeing with me in their identity, wishes to take Busk’s earlier name of guineensés; but Canu, considering that canariensis has been universally used for half a century, and that for both these the author is the same, considers we should retain the name canariensis. Canu also thinks that the figure of canariensis was good, whereas that of guineensis was bad. Having several times compared the British Museum specimens, it does not seem that we must speak of a bad figure, though in the specimens there may be in parts the structure as figured in canariensis as well as that of guineensis, and Lagree with Canu in thinking we should retain the name canariensis, which has often been well described. Should other characters be found in guineensis, it will be open to re-consider the name. Loe. Madeira ; Canaries, 80-250 met. (Calvet) ; Florida (Smitt); Tortugas (Osburn); Cape Blane (Calvet); Liberia; Philippine Is.; New Guinea; * Linn, Soc. Journ,, Zool. vol, xxx. p. 289 (1909). 412 MR. A. W. WATERS ON THE RELATIONSHIPS off Cape York, Australia; S. Pedro, 4 fath., and 8. Catalina, California (Rob.) ; Oran (Algiers). Fossil. Miocene: Austria and Hungary (J/anzoni); Pliocene: Crag, Italy, Spain, Rhodes, Monte Mario, Rome; Pleistocene: California (Zob.) ; Quaternary: Italy; Tertiary: Bahia-Blanea, Argentine (Canu) ; Aldinga, Australia, CupuLarta Lower, Busk. (PI. 30. figs. 1-6, 26-29.) Cupularia Lowei, Busk, B. M. Cat. p. 99, pl. 116. figs. 1-6 (1854); Norman, “ Poly. Madeira,” p. 290, pl. 87. figs. 7-12 (1909); Osburn, Bry. of the Tortugas Islands, Florida, p. 194 (1914). Canon Norman kindly gave me a few spirit specimens, from Madeira, and of one I have prepared a most interesting whole stained decalcified pre- paration. Looked at with the frontal membrane in focus, a number of small bundles of muscles (PI. 80. fig. 1) are seen which pass from the membrane through the frontal pores ; then focussing lower down, as if the membrane had been removed, we see what must perhaps be called a semicircular opesium through which the polypide protrudes (fig. 2). The zocecia are connected by means of tubes in which there are septa or rosette-plates, where the neighbouring tubes join. The zoccial chambers are seen as quite separate sacs. - Looking at the zoarium from below, the zocecial chambers are separated by a considerable space, and to about the median line of the zocecial chamber there is a row of muscles which pass from the lower surface of the zocecial chamber to the lower surface (fig. 3) of the zoarium. The muscle threads are usually single, though occasionally two or three occur together, and there are here and there similar muscles to other parts of the zocecial chamber. ‘I'his hydrostatic system is different from anything yet recorded, and having in the lower part of the zoarium this muscular system and no basal chamber as in C. canariensis, Busk, it seems questionable whether they can remain in the same genus. The calcareous dorsal surface, especially if rubbed down a little, shows short grooves, with frequently a pore at the end (Pl. 80. fig. 4). These of course indicate where the row of bundles of muscle occur. The peripheral zocecia on the dorsal surface have small granulations, but the older parts have elongate slight protuberances. The dorsal pore and grooves occur very distinctly in Cupularia denticulata fossil from Veletri, near Lorenzano, on the hills near Pisa, in my collection, and less distinctly on fossils from San Gemignano, near Siena, as well as from the Antwerp Crag (all collected by me). The same structure evidently OF THE SELENARIAD 2, CONESCHARELLINIDA, ELC. 413 occurs also in C. Oweni, Busk of the B.M. Cat., but this may be wmbellata, and it may also be seen in C. denticulata of the Crag. This species is much like C. wmbellata, Dely., but the lobed or irregular growth of C. Lowet, as described by Busk and Norman, has not been mentioned in wmbellata, though it ovcurs in Cupularia deformis” Busk, MSS. in the British Museum, which however is closely allied to C. denticu- lata, with the denticles very wide and solid. The zoaria grow in irregular shapes, often lobed just like C. Lowe? ; and Reuss in his manuscript after- wards published by Manzoni* speaks of the growth of C. Haidingeri, Rss., being excentric, and from the figures it has the groove mentioned as occurring. in C. Lowei. The “deformis” occurs from the Porcupine Expedition, 45 fath. (B.M.) ; Tangier Bay, 35 fath. (B.M.); Ras el Amourh, 45 fath. (B.M.) ; Cape Sagras (B.M.), Portugal; and in my collection from Mer el Kebir, Oran, 50 metres ; I propose to place it under denticulata. The polypide of C. Lowei is about double the size of that of C. canariensis. The frontal membrane has no trabeculee comparable with that cf Selenaria maculata, but there is a minute chitinous curve from the base of the oral aperture to the boundary of the zocwcium (PI. 80. fig. 1). We thus know Cupularia with regular, and others with irregular zoaria of the Lowe: form, having large frontal pores ; also the same zoarial forms with denticulate lamina. ioc. Madeira (Busk), common in 50-70 fath. (Norman); Tortugas Islands, 12-22 fath. (Osburn) ; North Carolina (Verrill § Osburn); Oran. CupuLariA Jounsoni, Busk. (PI. 29. fig. 17; Pl. 80. figs. 23, 30, 31.) Cupularia Johnsoni, Busk, Q. Journ. Micr. Sci. vol. vii. p. 67, pl. 28. figs. 1-5 (1859) ; Norman, “Poly. Madeira,” p. 290, pl. 38. figs. 1-6 (1909); Canu, “Bry. foss. France,” vol. xvi. p, 139 (1917). Cupularia Reussiana, Manzoni, “ Bri. plioc. Ital.” p. (11), pl. 2. fig. 19 (1869); “Bry, plioe. sup. de Ile de Rhodes,” Mém. Soe. Géol. de la France, ser. 3, vol. i. p. 67 (1877) ; Waters, ‘Bry. from the Plioc. of Bruccoli,” Trans. Manchester Geol. Soe. vol. xiv. p. 480 (1878) ; Seguenza, ‘ Formaz. terz.” pp. 131, 208 (1879) ; Neviani, “ Bri. Neoz. di alcune loc. in Italia,” Bull. Soc. Rom. vol. iv. p. (7), 115 (1895): op. ect. vol. iv. p. 243 (19) (1895 ); “Bri, Neog. della Calubrie,” Pal. Ital. vol. vi. p. 169 (1900). Discoflustredla doma, W’Orb, Pal, Fy. vol. v. p. 561 (1850-2). Cupularia doma, Smitt, “ Floridan Bry.” pt. 2, p. 15, pl. 3. figs. 81, 84 (1878). The Madeira specimens, though in spirit, had evidently been dead some time before preservation, and externally had grains of sand adhering to the membranes, while internally diatoms were numerous, so that satisfactory preparations were impossible. The large cells lining the zoccial and other walls are very marked, and the %® “ Mioc, d’Aust.” p. 71, pl. 16. fig. 54 (1877). 414 MR. A. W. WATERS ON THE RELATIONSHIPS embryos (about 0°5 mm. long) nearly fill the zocecial chambers. They are about the same size as in C. canariensis. The under surface of the Oran specimens are spinous (PI. 29. fig. 17), but in a fossil from Rhodes the under surface is uneven, rather mamillate than spinous, and a fossil from Monte Mario, near Rome, is almost plain underneath. This was, no doubt, first met with by d’Orbigny, who called it Discoflustra doma, but it was not figured, and the description was insufficient for recognition, though now having specimens from the same locality as d’Orbigny, with the ‘‘dessous trés rugeux comme epineux,” we may feel practically certain as to the identity, though as Busk first gave it a recog- nisable description it must be called Johnson. More than one mistake has been caused through d’Orbigny creating the genus Discoflustrellaria, with the species doma (probably Lunulites), and also Discoflustrella with the species doma. Loc. Madeira; Mediterranean ; Oran, Algiers ( Waters § Canw), 152 fath.; Benzert Road, Tangier, and Ras el Amourh, 45 fath. (all Brit. Mus.). Fossil. Castelarquato, Bruceoli, Rhodes (Pliocene); Ravagnex (Nev.), Amato (WNev.), Benestare (Seg.), Tortonian; Terreti (Seg.) ; Leognan, Gironde, France (Burdigalien) (Canu). : CUPULARIA UMBELLATA (Defrance). Lunulites umbellata, Defrance, Dict. des Sc. Nat. vol. xxvii. p. 861 (1828). Lunulites urceolata, Blainville, Man. d’Actin, ou de Zoophytologie, p. 449, pl. 72. tig. 1 (1854). Discoporella umbellata, VOrb. Pal. Fy. p. 478, pl. 717. figs. 1-65 (1850-2). Discoporella Beradana, VOrb. loc, cit, p. 474. Lunulites Hatdingeri, Reuss, Foss, Polyp. des Wien. Tert. p. 58, pl. 7. figs. 26, 27 (1847). Cupularia Haidingert, Manzoni, ‘ Mioc. d’ Aust.” p. 71 (28), pl. 16. fig. 54 (1877) ; Canu, “ Bry. foss. France,” vol. xiii. pp. 124, 125, 128 (1913); p. 3820 (1915) ; op. ect. vol. xvi. p. 188 (1917). Cupularia umbellata, Manzoni, “ Bri. Plioe. Ital.” pt. 1, p. 26 (10), pl. 2. fig. 16 (1869) ; “ Bri, del Plioc. Ant. di Castrocaro,” p. 39, pl. 5. fig. 67 (1875); Smitt, “ Floridan Bryozoa,” p. 14, pl. 3. figs. 75-80 (at least 79, 80) (1878) ; ? Hincks, “‘Poly. and Hyd. of the Mergui Archipelago,” Journ. Linn. Soe. vol. xxi. p. 125 (1887) ; Seguenza, “ Terz. Reggio.” pp. 131, 296, 371 (1879); Calvet, Expéd. Se. du ‘ Travailleur’ et du ‘ Talisman,’ vol. viii. p. 3893 (1907) ; Canu, “ Bry. foss. de Argentine,” Ann. del Mus. Nac. de Buenos Aires, vol. xvii. (ser. 3, vol. x.) p. 275, pl. 5. figs. 4, 5 (1908); “ Bry. foss. France,” ser. 4, vol. ix. p. 448 (1909) ; vol. xiii. p. 180 (1918); vol. xiv. p. 322; vol. xv. p. 332 (1916) ; vol. xvi. p. 187 (1917); Neviani, Boll. Soc. Rom. per gli Studi Zool. vol. iy. p. 243 (1895); # “ Bri. foss. della Farnesina” (pars), Pal. Ital. vol. i. p. 101 (1895) ; Canu, “Bri. Helv. de Egypte,” Mém. Inst. Egyptien, vol. vi. p. 205 (1912). When the frontal calcareous wall is partly broken down there is resemblance to C. denticulata, which has sometimes caused confusion between * Tam unable to accept all his synonyms. OF THE SELENARIAD, CONESCHARELLINIDA, LIC. 415 the two, but specimens of fossil wmbellata from the Pliocene of S. Gemig- nana, near Siena (one of whichis flat and must have been at any rate 25 mm, in diameter, while another is more raised and is 8 mm. diameter), show that the proximal part of the oral aperture is a calcareous bar or wall, often directed upwards, while near each end there is a small ridge or tooth, and the frontal wall has a row of large pores near the border, with smaller ones in the central portion. A part of the zoarium, not quite at the centre of the large piece, has a thick wall over the front with few small pores and sometimes with no opercular aperture. This structure is the same as that of the central zooecia. Specimens I collected from the Antwerp Crag, and consider to be denttculata, have no calcareous wall up to the proximal end of the oral aperture, except in a few cases near the centre of the zoarium, where there are 4-5 yery large pores and not a row round the border. The denticles are large and stout, whereas in the broken down wmbellata mentioned they are much more numerous and smaller, C. umbellata*, as described, differs from C. Lowe? in having regular zoaria instead of irregular and lobed-shaped ones, as well as in some zocecial characters. A new genus must probably be created for C. Lowe, C. umbel- lata, ete. Loc. Florida (Smitt) ; Canaries (Canu); Cape Verde Islands, 1900 met. (Calvet) ; Mergui Archipelago (Hincks). Fossil. Oligocene and Miocene: France; Miocene: Austriaand Hungary ; Pliocene : Crescentino, Bordighera, Siena (4. W.) ; Rhodes.* Burdigalian, Helvetian, Tortonian, Plaisancian, Astian, Sicilian, and Quaternary of Italy. Bahia-Blanea, Argentine (Canw). SELENARIA CONCINNA, Tenison Woods. (PI. 80. figs. 7-10.) Selenaria concinna, T. Woods, dicaas Phil. Soc. of Adelaide, vol. iii, p. 10, pl. 2 figs. ll a-1l ¢ (1880); Waters, “N.S. Wales,” p. 201, pl. 5. fig, 11 (1887); MacGillivray, “ert. Vict.” p. 48, pl. 7. fig. 15 Bae At the distal end of the vibraculum there is an incomplete ring ers by a kind of stalk at one side of the vibraculum. This is seen in fig. 8, with the zoarium somewhat tilted so as to look into the opening. The base “of the seta works upon this incomplete ring and the muscles pass through the ring. I+ spoke of this ring in S. concinna, Busk as a tubular projection and also figured it. * Pergens in his “ Bry. von Rhodos,” p. 30, gives the synonyms of wmbellata, and on the next page after “non” gives a list of species not to be placed under wmdellata. Miss Jelly unfortunately seems to have overlooked the ‘non,” and in her Catalogue makes Pergens responsible for placing five of these species under wmdedlata. + “N.S. Wales,” p. 201, pl. 6. fig. 11 (41887), 416 MR. A. W. WATERS ON THE RELATIONSHIPS The primary zocecium is surrounded by five secondary zowcia, and there ure three vibracula, so that we get eight chambers, which is the number of surrounding zocecia in so many of the group, that it seems almost general. The earlier vibracula are much smaller than the later ones, which are also relatively much wider. MacGillivray expresses some doubt as to whether the specimens which I described were the species of T. Woods, who described the avicularian chamber as having the edges studded with very fine teeth. 1 concluded that Woods’ specimens had the front broken down, when iastead of the pores described by me we have irregular teeth. It does not seem that: there can be any doubt as to the recent and fossil forms being the same species. The opesia are much larger than the operculum, except in the central zocecia, where the operculum is the same size as the calcareous opening. In Lunulites the opening is similarly opesial. Loc. Off Port Stephens, New South Wales, 25 fath. Fossil. Muddy Creek, Gellibrand, and Lake Bullenmerri, Victoria (A/acG.). SELENARIA PuNCTATA, Tenison Woods. (PI. 29. fig. 7.) Selenaria punctata, 'T. Woods, Trans. Phil. Soc. Adelaide, vol. iii. p. 9, pl. 2. fig. 8 (1880 Waters, ‘N.S. Wales,” p. 201 (1887) ; Maplestone, Proc. Roy. Soc. Victoria, n.s. vol. xvi. p- 208, and described again on p. 212, pl. 24. fig. 2 (1904). S. fenestrata, Haswell, Proce. Linn. Soc. N.S. W. vol. v. p. 42 (1880). My recent specimens from Port Stephens and from Princess Charlotte Bay, N.k. Australia, are a trifle larger than the measurements given by Maplestone, the zocecia being 0°23 mm. wide and long. I* called a fossil from Muddy Creek punetata, though mentioning the size of the aperture as 0:21 mm., whereas in the recent form it is only about 0:09-0:14 mm. Later MacGillivray + placed the fossil as 7. punctata and called attention to the large size of the zocecia. Maplestone $ would call the fossil S. magnipunctata, though it is an open question as to whether the mere difference in size between a tertiary fossil and recent forms is sufficient to necessitate a new species. However, in 1887 I was describing recent forms which are S. punctata, and therefore Maplestone has made a slip in placing these under the synonyms of magni- punctata. The S. fenestrata, Hasw., presented by the Australian Museum to the British Museum, is the magnipunctata of Maplestone. Loe. Off Cape Three Points (Woods) ; Holborn Island (Haswell) ; Princess Charlotte Bay, Port Stephens, N.S.W.; Port Jackson. * Quart. Journ. Geol. Soc. vol. xxxix. p. 440 (1888). + “ Tert. Poly. Victoria,” p. 47, pl. 7. figs. 8, 9 (1895). t “ Vict. foss. Selenar.’”’ p. 212, pl. 24. fig. 2 (1909). OF THE SELENARIADH, CONESCHARELLINID A, ETC. 417 SELENARIA MACULATA, Busk. (Pl. 29. fig. 8; Pl. 30. figs. 13-15.) For synonyms see Miss Jelly’s Catalogue and add: MacGillivray, Tert. Poly. Victoria,” p- 47, pl. 7. figs. 5, 6, 7 (1895). As mentioned in my Supp. Rep. Zool. Chall. p. 38, there are spreading round the oral aperture, and ending somewhat lower than the operculum, trabecule in the frontal membrane, reminding us of the trabeculze described by Busk in Cellaria, and which also occur in Onychocella angulosa, Rss. Busk says that the trabeculee ‘appear to lie beneath the common epitheca and not to form mere thickenings of it.” Where I have examined them they seem to form part of the membrane. In Onychocella angulosa the ends are raised and thicker, and seem to be for the attachment of the muscles to the membrane. The different shapes of the trabeculee seem to give useful characters wherever they occur. ‘The vibracular sete are spinous on the one side (fig. 13). Loc. Holborn Island, Queensland ; Barnard Island, N.E. Australia, 10 fath. Fossil. Muddy Creek, Bird Rock, Schnapper Point, Belmont, and Cape Bullenmerri (Victoria) ; River Murray Cliffs, 8. Australia. LUNULITES cupuLUs, Busk. (PI. 30. fig. 16.) Lunulites cupulus, Busk, Voyage of the ‘ Rattlesnake,’ p. 1, pl. 1. figs. 13, 14 (1852) ; to) B.M. Cat. p. 100, pl. 112. figs, 1-6 (1854). Lnnulites gibbosa, Busk, B.M, Cat. p. 100, pl. 111. figs. 1-6 (1854). Judging from the chamber the appendage is apparently avicularian, that is to say it has movement in one direction only, but without seeing spirit specimens it is best not to speak too definitely. The mandible or seta is long and large, gradually diminishing to the apex (fig. 16), and near the apex a small branch grows from the side. The figure is taken from a preparation in the British Museum, made by Busk, and only labelled ‘ hairs of Lunulites,’ but comparison with the type-specimen shows that it is certainly from ZL. cupulus, B. This is clearly allied to Lunulites repandus *, Maplestone, and L. patelli- formist, Map|. (non Marsson), but without examining specimens it is not advisable to say more. The fossil L. patelliformis, Marsson { (non Mapl.) and 1. semilunaris, Marss.§, seem also to belong to this group. Busk changed the name Lunulites to Lunularia, a change which has not * “ Victorian Selenariide,” vol. xvi. N.s. p. 216, pl. 26. fig. 7 (1904). + “ Vict. foss. Selenar.” p. 216, pl. 25. fig. 6. { “Bry. Weiss. Schreibkreide des Insel Riigen,” Pal. Abh. vol. iy. p. 79, pl. 7. fig. 11 (1887). § Loe, cit, p. 78, pl. 7. fig. 10. 418 MR. A. W. WATERS ON THE RELATIONSHIPS seemed to me necessary, though it might have been better in the first instance. However, the modification did not make a new genus, so that under no circumstances could Lunulites cupulus, as proposed by Canu and Bassler, be the type of the genus, for being mentioned in the ‘ Challenger’ Reports in the re-named genus, it does not replace the type, which is, whatever Lamouroux meant for his Z. radiata, an abundant form under various names. In all probability L. cupulus will have to be placed in a new genus, partly based on the long zocecia and the stout seta with a simple base. Trocnopora, d’Orbigny. (PI. 29. fig. 18; Pl. 30. fig. 17.) Although I have seen many specimens of Vrochopora, the state of pre- servation has not allowed «a satisfactory examination of the avicularian (?) chambers, nor are they shown clearly in published figures. The fossils readily break both along the radial and annular lines, and change of the shell has taken place so that the minute structure cannot be studied. Near the tubes passing through the calcareous base, and also near some of the annular lines, the calcareous matter has become chalky, showing in section white against the more transparent parts. In making horizontal sections the radial divisions, as in fig. 18, are seen in every part of the solid interior, that is to say, there is a succession of the same structure. i The interior of the cone is filled in with a solid calcareous mass in species described as Trochopora, but I have not been able to see that it should be separated from ZLunulites, and finding that L. conica, as described and determined by Busk, from the Crag, is sometimes solid and sometimes hollow, as stated by Busk, has confirmed my opinion. Specimens 6722, 6723, 6724, British Museum, are filled in below and are solid, while 6718, from the Red Crag of Sutton, are hollow cones about the same size, with similar annular divisions and similarly worn. In the British Museum there are also specimens named by yarious collectors Lunulites conica, many large like the Crag specimens; and in boxes from numerous European localities the solid and hollow forms occur together. There seems ample reason for considering that Lunulites and Trochopora cannot be separated generically, and further, though very difficult to understand, it does not seem that a specific distinction can be made between those that are solid and those that are hollow. Also we must see whether the large form sometimes called urceolata and the small one, as figured by Michelin, are distinct. Further studies with freshly gathered material may help us. A specimen of Selenaria nitida, Maplestone, in the British Museum has the under surface filled in just as in Zrochopora ; however, the vibracuiar OF THE SELENARIADA, CONESCHARELLINID®, ETC. 419 chambers are very small and narrow, and at first they were overlooked, but a few are made out clearly. We thus get this structure in two quite distinct groups. HeELIopoMA iImPLicatTa, Calvet. Heliodoma implicata, Calvet, Bull. Mus. Hist. Nat. p. 157 (1906); Expéd. Se. du ¢ Trayail- leur’ et du ‘ Talisman,’ vol. viii. p. 396, pl. 25. figs. 7, 10 (1907). In the British Museum there are some specimens, 1253, to which Busk had given the manuseript name Cupularia minima. They grow on grains of sand, shell, or portion of stalk, and the single spiral form of growth can be followed, though it is not so marked as figured by Calvet, and does not end abruptly. The interesting point is that the auricular process is always on the same side; now in the Cupularidee generally there is a co-relation between the position of the polypide and of the vibraculum, so that a row having the polypide turned to the right (as shown by the cxecum being to the right) will have the auricular process also on the right. In other Cupu- laridee there may be one, two or more rows with all turned to one side ; when a change takes place, the position of both the polypide and the vibraculum alters together. Although in the small specimens only a single spiral is visible I should hesitate to make a specific separation, though further material may make this necessary. The spiral growth is most interesting and perbaps nothing quite similar is known in other Bryozoa. Loc. Cape Verde Islands, 1900 met.; Canaries, 3709 met. (Calvet) ; Aigean Sea, 130 fath., collected by Spratt. CONESCHARELLINA CANCELLATA (Busk). (PI. 29. fig. 22.) Lunulites cancellata, Busk (pars), b.M. Cat. p. 101, pl 115. fig. 5 (non 6) (1854); Waters, ‘ Bry. from Bairnsdale,” Q. Journ. Geol. Soc. vol. xxxviii. p. 512, pl. 22. figs. 10, 11 (1882). Bipora cancellata, Whitelegec, “ Austr. Poly.” p. 340 (1887); MacGillivray, “ Tert. Victoria,” p. 89, pl. 12. fig. 1 (1895). Conescharellina cancellata, Waters, “ Bry. N. S. Wales,” p, 200, pl. 4. fig. 24; pl. 6. figs. 13, 18 (1887). In the British Museum specimens thus named by Busk belong to three or four species. Specimens from Busk’s own collection so named are C. angustata, WOrb., of which species I figure a specimen (PI. 80. fig. 18) from China, sent to me thus named by Jullien. The specimen, pl. 113. fig. 6 (as cancellata) in the British Museum Catalogue, has round avicularia as in C. philippinensis without any notch in the outer zocecia, but the minute pore by the proximal end is just visible. The cone is somewhat higher than in C. philippinensis in my collection, though no doubt it is that species, as are also the other two small specimens. The larger specimen on the same slide 420 MR. A. W. WATERS ON THE RELATIONSHIPS has the base about twice the diameter of C. philippinensis. It has small round or oval avicularia, but the preservation is not very satisfactory. The under surface of C. cancellata has some chambers with one large pore in the centre of the basal wall, sometimes surrounded by small pores, as in C. angulopora, Haswell (PI. 9. fig. 21) and C. flabellaris, Lev., while generally there is the large opening of the chamber and smaller openings round it as in Pl. 29. fig. 22. My specimen is from Port Stephens, New South Wales, and,.although the front surface is not very.well preserved, I think it is the same as Busk’s large specimen, for which the name can be retained. My specimen is 4 mm. at the base. - The mistake made, when the type was described, by Busk led me astray when my New South Wales and other specimens were compared in the British Museum, in consequence the specimens then called cancellata by me are seen to be philippinensis. Loc. Philippine Islands (Bush); Port Stephens, N.S. Wales (4. W. W. coll.). Fossil. Bairnsdale, Gippsland. CONESCHARELLINA PHILIPPINENSIS: (Busk). (Pl. 29. figs. 11, 12, 13; Pl. 80. fig. 24.) Lunulites philippinensis, Busk, B.M. Cat. pt. ii. p. 101, pl. 113. figs. 1, 2, 3 (1854). Bipora philippinensis, Whitelegge, ‘ Austr. Polyzoa,” vol. ii, p. 341 (1887); MacGilli- vray, “ Tert. Poly. of Vict.” p. 89, pl. 12. fig. 2 (1895); Maplestone, “ Bipore,” p. 3 (1910) ; Levinsen, Morph. Cheil. Bry. p. 309, pl. 24. fig. 1 (1909). Conescharellina cancellata, Waters (pars), “N.S. Wales,” p. 200. pl. 4. fig. 24; pl. 6. figs. 13, 18 (1887). The comparison of Busk’s type-specimens led me astray, on a previous occasion, as he had placed more than one species under L. cancellata, one of these was C. philippinensis, but more preparations and more material of the Selenariadee has enabled me to feel more certain of my position. Unfortu- nately of philippinensis I have only dry specimens, and of all the Selenariade have only seen the three spirit specimens mentioned. I have a considerable number of philippinensis ranging from very young specimens, from under 1mm. in diameter, up to mature forms nearly 3 mm. ‘The zoarium is usually rather watch-glass shape, that is convex above and concave below, but it may be flat below; in either case there area number of small raised avicularia with a semicircular mandible in the middle of the wall of the avicularian chamber. The oral aperture is oyal with a sinus in the part nearest to the periphery of zoarium, there is a distinct notch in the secondary aperture of the outer zocecia, and a minute pore by the other end of the oral aperture. Numerous round avicularia occur on the front of the zoarium, similar to those on the OF THE SELENARIADA, CONESCHARELLINIDA, ETC. 421 under surface, and typically there is one on each side of the zocecium, and one below the aperture. On the under side of the zoarium there are a number of chambers (PI. 29. fig. 11), the youngest ones being avicularian. These chambers, apparently, must be compared with those of Cupularia canariensis, and are found in Conescharellina philippinensis, C. angulopora, C. cancellata, C. flabellaris, C. conica. Whitelegge speaks (loc. cit. p. 431) of the zocecia taking their origin from the cancellated structure, but it is now clear that the zocecia grow first, and from them the cancellated structure. The semilunar slit to which Whitelegge called attention, and which Maplestone has also studied, occurs in this species. These semilunar slits have not a definite position in relation to the zocscia, being most abundant near the apex of the zoarium, and sometimes several are found in the same circle somewhere near the apex, others in a circle half-way between this and the periphery, thus there is considerable irregularity and variation. In some Conescharellinee there seem to have been radicles from near the apex, where there is a considerable calcareous cap with pores (PI. 29. fig. 16), but this is not the case in C. philippinensis, in which there is no such calcareous growth, but on p. 405 it is suggested that radicles grew from the semilunar slits, and I expect with living or spirit specimens we shall see confirmation of deductions made from dried specimens. However, I have found no indica- tions suggesting radicles, except in Conescharellina, and perhaps Batopora. In the young philippinensis the under surface is flat or slightly concave, whereas in older ones the under surface is concave. The younger zowcia are very much smaller than the older ones, and the secondary apertures are also smaller. A figure (Pl. 80. fio. 24) of a very young zoarium is added, as it shows the small size of the zocecia, and also the typical position of the avicularia at each side of the zocecia, which is not so easily seen in more mature forms, as the avicularia are then more irregularly placed. The raised and globular ovicell, described by Whitelegee and figured by Maplestone, is directed towards the apex. A specimen with ovicells was given to me by Maplestone, but altogether only very few have been seen. Loc. Philippine Islands (Busk) ; Katow, New Guinea, 7 fath.; Darnley Island, Torres Straits, 10-30 fath.; Princess Charlotte Bay, N.E. Australia, 13 fath.; Port Stephens “from weeds on sandy bottom,” dredged by Brazier; Moreton Bay, N.E. Australia (Whitelegge in lit.) ; Port Jackson. Fossil. Curdies Creek, S.W. Australia ; Schnapper Point (AMacG.). CONESCHARELLINA FLABELLARIS, Levinsen. Conescharellina flabellaris, Levinsen, “ Morph, Cheil. Bry.” p. 312 (1909). Conescharellina elegans, Waters, “N.S. Wales,” p. 200, pl. 5. figs. 13-17 (1887). Bipora (?) elegans, Whitelegge, “ Austr. Polyzoa,” vol. ii, p. 346 (1887). In my “Notes on sume Recent Bryozoa in d’Orbigny’s Collection,” * * Ann, Mag. Nat. Hist. ser. 7, vol. xv. p, 3, pl. i. fig. 5 (1905), 422 MR. A. W. WATERS ON THE RELATIONSHIPS T said that the examination of d’Orbigny’s specimens showed that his Flabellopora elegans is not what I described as Conescharellina elegans, and since then Levinsen has suggested the specific name flabellaris for the latter, and I have adopted it. One specimen (fig. 17, loc. cit.) which I figured is quite bilaminate, while others (like figs. 15, 16, /..) are much wider, so that the section of one specimen is very much like that of angulopora or of coniea (PI. 29. fig. 16). The under surface has chambers with a large round pore in the centre and five or six small pores surrounding it, Just as in C. angulopora, T. Woods (Pl. 29. fig. 21). The avicularia on the upper surface are small and round, whereas in C. angulopora and C. conica they are large and triangular. The semilunar slit occurs in various places, but perhaps most often in the region of the apex and the border. The slit is materially wider than the aperture of the ordinary zomcia, which, as I have suggested before, seems to indicate that the slit could not be for the formation of a zocecium or zocecial aperture. In other species there are typically a pair of avicularia by the slit, but there is no rule in this species. Loc. Port Stephens, N.S. Wales, 7-8 fath.; Port Jackson (Whitelegge & Lev.) CoNfSCHARELLINA ANGULOPORA (Tenison Woods). (PI. 29. figs. 6,19, 21; PI. 80. fig. 19.) Lunulites angulopora, Woods, ‘ Austr. Selenariade,” vol. iil. p. 7, pl. 1. figs. 3a-8¢ (1880); Whitelegge, “ Austr. Polyzoa,” vol. ii. p. 348 (1887). Lunulites incisa, Hincks, Ann. & Mag. Nat. Hist. ser. 5, vol. viii. p. 127, pl. 4. figs, 1-8. Conescharellina angulopora, Levinsen, “ Morph. Cheil.” p. 311, pl. 23. figs. 7 a-7f (1909). 2 Conescharellina depressa, Haswell, “ Poly. from Queensland Coast,” Proc. Linn. Soe. N.S. Wales,” vol. v. p. 41, pl. 3. fig. 4 (1880). There are two very similar species from Australia, but the present one has the larger zoaria with the base of the cone more spread out ; the zocecia and avicularia are also larger, the peristomes at the sides of the oral aperture are more raised, as is also the avicularian chamber and beak. However, the most important character is furnished by the under surface, as the basal wall of the cancelli has a large central perforation with smaller ones round it (fig. 21), as a rule without avicularia, though I saw one case in which there was an avicularian bar to the larger perforation. These perforated walls of the cancelli are mentioned by Tenison Woods in his angulopora, and by Haswell for his Conescharellina depressa, and although Haswell’s figures do not correspond closely with my specimens, there is the probability of both being the same species. The under surface of C. flabellaris, Lev., is quite similar. The semilunar slit is larger than in other species examined, being wider than the zocecial aperture, which would seem to show that it is not for the formation of a zocecial aperture; the slit may occur either in the row OF THE SELENARIADA, CONESCHARELLINIDE, ETC. 423 of the zocecia or of the avicularia, and often the disk of the semilunar slit has disappeared leaving only a semicircular opening. There is a pore-tube in the wall, at the proximal end of the oral aperture, and a similar tube oceurs in C. philippinensis, Busk, C. flabellaris, Lev., C. conica, Hasw., and has heen referred to by Levinsen (loc. cit. p. 309), and by Whitelegge (loc. cit. p. 339). The dorsal surface of the smaller allied form (see C. conica p. 423) has moderate-sized pores with usually a few small triangular avicularia scattered about, though in some specimens none are found. A specimen from Port Stephens, which I think must be C. cancellata, Busk, has fairly large round openings at the base with smaller round openings near the larger ones, sometimes surrounding them, in other cases irregularly placed. The base of C. philippinensis has numerous small round avicularia, so that the basal surface of Conescharellina gives most useful characters. Woods’ figure is very unsatisfactory and might represent either of two or three species ; though from it together with the description we seem justified in using his specific name. McGillivray* gives this with a ? as fossil, but it really seems as if both his description and figure have got in the wrong places, at least T cannot understand them. Loc. Port Stephens, N. 8. Wales (Woods), and 25 fath. sandy mud bottom (sent by Brazier) ; Bass’s Straits (inchs). CONESCHARELLINA Conca, /Zaswell (non Hantken). (PI. 29. figs. 16, 20.) Conescharellina conica, Haswell, Proc, Linn. Soc. N. 8. Wales, vol. v. p. 42, pl. 3 figs. 7, 8 (1880). Conescharellina incisa, Waters (pars), N.S. Wales,” p. 199, pl. 6. fig. 26 (1887). In deseribing C. angulopora, Woods (p. 422) I have said that there were two very similar Australian species. This smaller one has, however, the axis of the cone relatively much longer, while the larger one has the base relatively much larger. The zomcia and avicularia are somewhat smaller as is also the semilunar slit, but the most important difference is in the character of the base, for on the under surface of conica there are moderate- sized pores, with usually a few small triangular avicularia scattered about, though in some cases none have been found, I have a specimen of undoubted C. conica, Haswell, from Holborn Island, sent to me by Haswell, and which no doubt was determined by him, though, as it is a long time since it was received, more cannot be stated. At one time it seemed that the name coniea could not be retained, as it has been used in several closely allied genera, but as we are getting more definite ideas about the genera, there is not as much force in the objection as there was then. Loe. Holborn Island (aswell) ; N.1. coast of Australia, 23 fath. (sent by Brazier). * “ert. Victoria,” p. 46, pl. 8. fig. 1 (1895). 424 MR. A. W. WATERS ON THE RELATIONSHIPS CoNESCHARELLINA HOCHNA, Neviani. Conescharellina eocena, Neviani, “ Bri. Koc. del Calcare nummulitico di Mosciano presso Firenze,” Bol. Soc. Geol. Ital. vol. xiv. p. 122 (6) fig. 5 (1895); Waters, “ Batopora,” p. 85, pl. 6. figs. 8, 12 (1919). Batopora conica, Seguenza, ‘Form. Terz.” p. 42, pl. 4. fig. 10 (1879). Batopora conica, Hantken (non Haswell). This at first was taken for Batopora, but the shape of the cone is the same as in various Conescharellinee, while sections show its relationship, as the interior of the cone is filled in by chambers longer in the direction of a line from the apex to the base (fig. 8, loc. cit.). The zocecia are in irregular longitudinal lines, and are hexagonal or round, with large pores between the neighbouring zocecia. The visible aperture, which is probably only the secondary aperture, is round or oval. Near the apex of the zoarium there are only small openings with a larger one at the apex, though this is smaller than the pits of Spheropora and Batopora. On cther Conescharellina I have often seen small openings in the secondary growth over the apex, but have never found a central pit. The base of the zoarium shows large pores much like those of C. conica, Haswell (PI. 29. fig. 20). The ovicell is wide, slightly raised, and only occurs near the base of the zoarium. This seems to be the only species of Conescharellina known from the Ttalian tertiaries, and in some respects it varies from most Conescharelline. Hantken has frequently mentioned Batopora conica, Hantk.; but attempts made by various people to find where he has described it have failed, and I have in vain looked through most of his papers. Finding specimens presented by Hantken in the British Museum (B. 3724), is therefore most interesting. It is conical, about 5 mm. high, there is considerable swelling at the apex, and one of these shows a hollow surrounded by very small pores, with larger ones over the rest of the swelling. The zocecia have a nearly round aperture and smaller openings round them. One specimen is cut longitudinally through the middle, showing the cancellate structure, just as in my figure 8, loc. cit. The preparation referred to was no doubt made by Hantken. Loc. Fossil: Mosciano near Florence (Nev.); Spiassi, Monte Baldo, N. Italy ; between Grotte and Sarego near Lonigo, N. Italy (A. IW.) ; Tongrian: Antonimina, Calabria (Seg.); Buda Pesth, Lower Clay, Szaboi beds (Hantken) ; a'so from Schonthal, Festungsberg, Klein, Schwabenberg, and Urém (Hantken). an Fig, oo OL 10. Ills 20. Or OF THE SELENARIAD®, CONESCHARELLINID®, ETC, 42 EXPLANATION OF THE PLATES. PLATE 29, Cupularia canariensis, Busk, X 25. Stained transverse section showing the shallow central zocecia (¢.z.), which have grown upon some substance removed in decal- cification, and the large zocecia (/.z.), as well as the vibracular chamber (v.) shown near the periphery of the zoarium. The lower part of the zoarium is formed by a series of parallel chambers filled with granular contents and connected by rosette-plates. Do. do. x 200. Lower part of chambers with contents. A and B are connected by a rosette-plate, whereas B and C are close together bat not connected at this level. Transverse wall (¢.2v.). Lateral wall of the series (/.w.). Do. do. x 250. Transverse section showing contents of the chamber and the large cells. Do. do. xX 25. Transverse calcareous section showing chambers. Do. do. x 25. Decalcified base of the zoarium, with muscles of one vibraculum. Conescharellina angulopora, T, Woods, x 25. Calcareous section, showing the chambers at the base of the cone. @ x 3. From Port Stephens, N.S. Wales. . Selenaria punctata, T. Woods, X 12. Showing a piece from which the central supporting flake has disappeared. From Port Stephens, N.S. Wales. Selenaria maculata, Busk, X 12, Showing a piece of shell on which the colony has commenced to grow, It will be noticed that the early central zocecia are very small. From Holborn Island. Selenaria petaloides, VOrb., x 6. Starting on shell. Fossil from Wanganui, New Zealand. Vibracella trapexoidea, Reuss, X 12. Colony growing on Orbrtotdes stellata, Giimbel. Fossil from Bocca di Sciesa, Colle Berici, N. Italy. Conescharellina philippinensis, Busk, X 25. Basal view showing small chambers with avicularia and others either broken down or incomplete. From Port Stephens, N.S.W. Do. do. x Do. do. x lunar slits. 5. Section showing zocecia and basal cells. 5. Quadrant of upper surface showing some of the semi- Lo bo . Cupularia canariensis, Busk, X 85. Showing connecting tubes to the zocecia and to the vibracular chambers. From Madeira. . Cupularia Lowet, Busk, x 250. Chamber of peculiar body of the vibracula showing two small glands (g/.). . Conescharellina conica, Haswell, X 25. Calcareous section showing subsequent calcareous growth over the apex. From N.Is. coast of Australia. . Cupularia Johnsoni, Busk, x 25. Under side of the cone showing spinous pro- cesses, From Madeira. Trochopora conict, Def., X12. Lower surface. From Salles (Gironde) ; Helvetian. . Conescharellina angulopora, ‘l. Woods, x 25. Upper surface. There are two semilunar slits without the plate (s.s.). From off Port Stephens. Conescharellina conica, Waswell, x 25. Lower surface. From N.E. coast of Australia. LINN, JOURN.—ZOOLOGY, VOL, XXXLY. 32 99 fl 15. 16 17. 18 19. 20 23. D4 MR, A. W. WATERS ON THE RELATIONSHIPS Conescharellina angulopora, T, Woods, x 25, Lower surface. From Port Stephens, N.S. Wales, Conescharellina prob, cancellata, Busk, x 25, Lower surface. From off Port Stephens, PLATE 30, Cupularia Lowe’, Busk, x 50. Decalcified preparation, looked at from the front. Throveh the membrane the bundles of muscles attached to.it can be seen and they pass through the frontal pores, Do. do. x 50. The same preparation focussed at a lower level. The circular opening is shown through which the polypide pases and the tubular connections from this opening to the neighbouring zocecia, The polypide is faintly shown in the right-hand zocecium. Do. do. x 50. The same preparation seen from the dorsal surface. A line of muscles (2.) reaching down to the zocecial chamber is seen, and the poly- pides are usually alternately right and left in each radial row of zocecia. Do. do. x 25. Dorsal surface showing pore at the end of the groove. Do. do. x 85. Somewhat pressed down, so that the row of muscles are seen laterally. Do. do. Lateral section; diagrammatic, showing rows of muscles at- tached to the lower membrane (d.m.) and to the zocecial chamber (z.c.). . Selenaria coneinna, T. Woods, X 25. From Port Stephens. Do. do. x 50. Vibraculum. Selenaria concinna, Busk, X 25. Showing central zocecia and a few of the border zocecia. From off Port Stephens. ? Do. do. x 25. Section showing double expansion of a pore-tube under the supporting flake. Cupularia canariensis, Busk, X 25, Central zocecia with eight surrounding zocecia, From Petit Tahou, Liberia. Do. do. x 15. Under surface of same with iarge sand grain. 13. Selenaria maculata, Busk, x 325, Portion of vibracular seta showing spinous fringe on one side. Do. do. x 150° Operculum from inside showing trabeculie, and below the operculum two muscles attached to the frontal membrane. Do. do. x 85. Size for comparison with previous figures, Lunutites capulus, Busk, x 85. Vibracular seta. Trochopora conica, Def, X 25. From Salles: Fossil. Conescharellina angustata, VOrb., X 25. From China (4. W. coll.). Conescharellina angulopora, T. Woods, X 6. Showing growth over the apex. Conescharellina philippinensis, Busk, X 85. Decalcified to show the position of the zoce cial (ze.) and avicularian (avc.) chambers. Cupularia canariensis, Burk, X 150. Base of seta and front wall of the vibracular chamber. Do. do, x 150, Operculum seen from inside together with the frontal membrane showing trabeculz. Cupwlaria Johnsont, Busk, X 150. Operculum from inside with the bordering ridge or trabecule united to the zocecial border. Conescharellina philippinensis, Busk, x 25, Young colony. Waters. AW.Waters, del. BI Journ. Linn. Soc., ZooL. Vou. XXXIV PL 29, CambridgeUnivorsity Press \ZOA. Waters. Journ. Linn. Soc. Zoon. Vou. XXXIV PL 29, AW.Waters, del. CambridgeUnivorsityPress. BRYOZOA. Waters. BRY AW.Water#, del. Journ. Linn. Soc.,Zoou.Vov. XXXIV P1380. aI oe mawinigeees OES GR ice a eal H ‘ ' & oO ss pon ene te ee ee ee ee eee ee = | \ Kel x180 ; \ lr 1] | fj \ tal f V4 / ie /, X250 | | xee | WAG i ‘f / Bocas } (O) Vm i es ani (=. y, CambridgeUniversity Press. A. Waters. BRYOZOA. Journ. Linn. Soc., Zoon.Vor. XXXIV P1380. CambridgeUniversityPress. ue : Ae Fig, OF THE SELENARIADA:, CONESCHARELLINIDA, ETC. 427 25. Cupularia canariensis, Busk, & 25, Ovum in zocecial chamber. Portion of remains of polypide (p.). 26. Cupularia Lowest, Busk, x 85. Vibracular chamber, showing the peculiar body ending at the thin circle in the membrane. One bundle of long muscles, xs well as the short ones, is shown. 27. Do. do. x 85. Vibracular chamber above the zocecial chamber. 98. Do. do. x 550. Muscle of vibraculum. 29. Do. do. x 150. Muscles leading to fascia, which is attached to the base of the seta. 30. Cupularia Johnsoni, Busk, x 250, Base of seta. 31. Do. do. x 85. ‘Seta. 39% ON THE STRUCTURE OF MAXILLULE IN INSECTS. 429 On the Structure and Occurrence of Maxillule in the Orders of Insects. By Atwen M. Evans, M.Sc. (Manch.), Dept. of Medical Entomology, Liverpool School of Tropical Medicine. (Communicated by Dr. A. D. Imus, F.L.8.) (PLatE 31, and 17 Text-figures.) {Read 5th June, 1919.] CONTENTS. Introduction, p. 429; Apterygota, p. 480; Dermaptera, p. 453; Orihoptera, p. 434; Ephemeride, p. 435; Perlaria, p. 436; Psocide and Mallophaga, p. 457 ; Neuroptera, p. 439 ; Coleoptera, p. 442; Lepidoptera, p. 445; Trichoptera, p. 447; Hymenoptera Tenthredinee, p- 448; Diptera, p. 451; Summary of Conclusions, p. 452; Bibliography, p. 454; Explana- tion of Plate 31, p. 456. INTRODUCTION. A QUARTER of a century ago Hansen (1893) originated the theory that the pair of appendages associated with the hypopharnyx of AMachilis and other Apterygota is homologous with the mavxillule of Crustacea (21)*. The theory has since been raised to the status of an established fact, chiefly by the embryological researches of Folsom. In the paper in which his results are recorded, Folsom (17. p. 116) stated that “The superlinguce (maxillulee) should hereafter be recognised as morphologically important struciures, and be searched for in even the most specialised haustellate Od ersile ys, ees | Since this forecast was made its value has been proved by reasearches on the constitution of the hypopharynx in Coleoptera and Lepidoptera. In the larvee of certain genera of these orders the presence of structures of undoubtedly maxillular nature has been demonstrated. The present paper is a record of investigations whose object has been to determine as far as possible the occurrence of structures comparable with maxillulee in the various orders of insects. In the haustellate orders only the mandibulate larvee are considered. The Hemiptera, Mecaptera, and Thysanoptera are not dealt with. Some attempt is made to compare the form of maxillule characteristic of different orders and, where enough material has been available, to study the distribution and variation of these structures within the order. I wish to express my gratitude to Professor 8. J. Hickson, F.R.S., who * My attention has recently been called to the fact that Lubbcck in his monograph of the Thysanura and Collembola recognises the maxillule as appendicular structures, referring to them as “a second pair of maxille.” 430 MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE has kindly allowed me to work in the Entomological Laboratory of the University of Manchester, I find it impossible to express how greatly Lam indebted to Dr. A. D. Imms for the invaluable assistance he has rendered in innumerable ways. To him I owe the suggestion of the subject of investigation. I am glad to have this opportunity of thanking Dr. Imms for the immense amount of trouble he has taken in the furtherance of this work. Terminology.—Folsom in 1899 furnishes a complete list of the various terms employed up to that date to designate the hypopharynx and maxillule. To these latter terms, proposed by Hansen, he objects on the grounds that they imply unproved homology. He proposed to call the median element of the compound “ hypopharynx” the “lingua,” and its lateral appendages the “ superlingue.” His objection to Hansen’s terms can no longer hold after his own proof of the correctness of the assumption of this homology. Investigators on the morphology of the mouth-floor in Hndopterygota speak of the median portion as the hypopharynx and the lateral lobes pro- jecting from or situated upon it as maxillule. These terms possess the advantage that they are free from the possibility of confusion with others used to designate portions of the labium. ‘They will for these reasons be employed in the present paper. APTERYGOTA. Machilide.—-The maxillule are figured by Oudemans (85) under the name “ Paraglossee.” They are described in some detail by Hansen (21. p. 31). They are, he says, independent of the hypopharynx near the base of which they are articulated to the skeleton of the head. He notes the “small single jointed palp,” and observes that each appendage shows, towards the tip, a ten- dency to cleavage into two lobes. A detailed figure is given by Carpenter (6). He refers to the ‘spicules, ridges and pits” which occur on their surface, and states that the outer lobe of each is articulated with the basal sclerite. They have, he says, “all the appearance of a reduced pair of jaws.” Petrobius sp.—P. oudemansi, Carpenter (?). The headless remains of the specimens were kindly identified as far as possible by Professor G. H. Carpenter. Since these appendages are in Machilidw in a less rudimentary condition than in any of the other insects considered in this paper, their structure in this species has been studied with some care. Observations made on simple dissections were supplemented by examination of pre- parations treated with potassic hydrate, and others, which in addition were stained with acid fuchsin. The maxillulee, which lie between the mandibles and first maxille, are seen, after the removal of these pairs of jaws, to stand on the floor of the month, above the base of the hypopharynx. O€ this organ they obscure all but the OF MAXILLULZ IN THE ORDERS OF INSECTS. 431 extreme tip. Pl. 81 is drawn from a preparation in which the maxillule were pressed apart so as to expose the hypopharynx. This latter measures. -56 mm. in length. The length of the left maxillula is 59 mm. and its greatest width 1:9 mm. A pair of long peduncles (p.e.) support the hy pophyarynx, The cavity of the maxillula was found to be undivided; the sutures separating the lobes consisting of no more than ridges and grooves in the upper wall only. Of these ridges, that which marks off the outer lobe, arises beyond the palp. At its base is the rounded, chitinous structure (a.) sug- gestive of an articulation. A short distance behind the apex of the inner lobe there projects from the depression between this and the outer lobe the curious structure (/,), armed with six curved tecth, which point towards the surface of the maxillula ; the portion of the palp (p.) beyond the lateral wall measures ‘14 mm. in length. Campodea lubbocki. Dorsal aspect. From a preparation stained with acid. fuchsin. Pressure has been applied and the maxillule are displaced outwards. x 250, a. Articulation of maxillula. 4%. Hypopharynx. m. Maxillula. s. Supporting skeleton. Campodew.—In his paper on the Campodeze, Meinert (32) described the lingua and paraglossee of Japya and of Campodea. The lingua, he says, is small in the former genus and the paraglosse large and bilobate. In the latter the appendages are large and flat, entirely covering the dorsal surface of the oval lingua. Hansen (21) states that the two genera are closely allied as regards the structure of the mouth-parts. He agrees with St. Traunfels that-in Japyx the maxillule consist of an inner lobe, an outer lobe, and a three-jointed palpus. In dealing with these structures, in Lepidocampa Jimbriatipes Carpenter (8) states his belief that this outer lobe and palpus were rightly referred to the maxilla by Borner in 1908. Specimens of Campodea lubbochi (Silvestri) (for the identification of which I am indebted to Mr. R.S. Bagnall) were collected at Fallowfield, Manchester. 432 MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE The structure of the hypopharynx and maxillule is illustrated in text-fig. 1. The hypopharynx measures ‘09 mm. in length and ‘12 mm. in greatest breadth. In its natural condition it is almost obscured by the maxillulee (m.), which are articulated at (a.) to a sbort branch of the skeletal support (s.). They are laminar, and can be removed entirely from the hypopharynx. The distal portion is triangular and the pointed apex furnished with sharp teeth directed inwards. The dorsal surface is covered with sete, those of the anterior and inner areas being borne by scales arranged in rows. Hach maxillula measures ‘11 mm. in length and °05 in total width. Lepismide.— Hansen (21) states that in the structure of the hypopharynsx Lepisma stands between Machilis and the Orthoptera. His statement is referred to by Carpenter (8. p. 13), who described the tongue of Jsolepisma and the maxillule borne by it. These latter have “a roughened apex, with —S Lok Lepisma saccharma. Dorsal aspect of tongue. 4. Hypopharynx. m. Its maxillular surface. p.e. Peduncles. 7. Longitudinal ridge marking inner edge of maxillula. 7, Anterior transverse bar of skeletal system. fine short bristles,” and there is “ no differentation int lacinia and galea.” The tongue of Lepisma saccharina (text-fig. 2) is a broad membranous structure closely applied to the surface of the labium. It is supported distally by a system of sclerites, composed of elements corresponding to those of Isolepisma. The width at the base is *2 mm. and the distance from the transverse bar (¢.) to the anterior edge 12 mm. The maxillulee have become completely fused with the tongue, of the upper surface of which they form the lateral portions, being separated from the median scaly portion by longitudinal ridges (7.). Collembola.—Maxillule appear to be present in all Collembola as a pair of delicate lamellee intimately associated with the hypopharynx. A frequently OF MAXILLULZ IN THE ORDERS OF INSECTS. 433 occurring feature is the presence of a fringe of tooth-like projections or of setee along the inner margin. Isotoma palustris was regarded by Carpenter (6) as possibly one of the least specialised Collembola. He described at the apex of the maxillula an indication of cleavage into two lobes. Of the Achorutidee Anurida maritima is the subject of Folsom’s well-known work of 1900. Reference has already been made to his couclusions with regard to the nature of the superlingue. An exact account of their structure is to be found in the Monograph of Imms (24). DERMAPTERA. Forjicula.—In his paper on Hemimerus talpoides, Hansen (22) devotes considerable attention to the hypopharynx and maxillulee which he describes and figures in detail (pp. 70, 71, fig. 10). I have also been able to examine the hypopharynx of specimens of Hemimerus hanseni, Sharp, kindly deter- mined by Professor Carpenter und collected by Dr. J. W. 8. Macfie at Text-fig. 3. Forficula auicularia. Dorsal aspect of tongue. x 80. 6. Brush-like sclerite. # Asym- metrical fold in dorsal wall of hypopharynx. %, Hypopharynx. m. Maxillula. n. Chitinous plate in inferior wall of pharynx. p. Transverse skeletal bar. p.m. Chitinous rod in outer wall of maxillula. s. Small sclerite. Accra, Gold Coast. I found it essentially similar to that of Foryicula. The main differences concern the degree of fusion of the maxillulee with the hypopharynx, and the details of construction of the transverse elements of the skeletal support. We find in Forjicula a roughly oblong tongue (text-fig. 3) lying above the labium. It is produced distally into a median (h.) and two lateral (m.) mem- branous lobes, corresponding to the hypopharynx and maxillule respectively 434 MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE of Hemimerus, although each is considerably shorter in proportion to its width than are the corresponding organs in this latter genus. In ene adult specimen they measured ‘02 mm. in length and their width at the base was ‘(02 mm. The hypopharynx extended beyond the level of their bases to a distance of ‘03 mm. and its greatest breadth was ‘07. Proximally the floor of the mouth bears the pair of asymmetrical sclerites (n.) from the side of which the plates (e.) extend downwards to be closely associated, at their ventral extremities, with the inner angle of the mandible. These sclerites have an exact counterpart in HZ. hanseni, Sharp, although in the latter case there is no sign of asymmetry. Anteriorly we find in Forjicula a chitinous arch consisting of two asymmetrical sclerites (p.), the small piece (s.), and the curious setiferous median sclerites (6.), which will be referred to when we come to consider the Psocidee and Mallophaga. In this position in H. hanseni, Sharp, however, we find only a single slender arch of chitin immediately above which the maxillulz arise. In all the specimens examined the dorsal wall at the base of the hypo- pharynx exhibited a tendency to form the curved fold (7) whose apex lay considerably to the left of the median line. A similar fold at the base of the left maxillula gave this organ a bilobed appearance. The ventral surface of the tongue is stiffened by a continuous plate of yellow chitin. This plate extends forwards to a short distance behind the base of the hypopharynx. It is associated at its anterior lateral corners with a slender chitinous rod (p.m.) which supports the external wall of the maxillula. In this position I found a more plate-like sclerite in 7. hansen, Sharp. It is probable that Jorficula is derived from a form in which the hypopharynx and its lateral appendages originated as far back as the transverse sclerite. Such a condition would resemble more ciosely that of Hemimerus. This supposition is supported by the presence of a faintly marked suture (y.) extending back from the inner angle of the left maxillula to this skeletal bar. ORTHOPTERA. Folsom (17. p. 115) gives an account of the hypopharynx of Orthoptera. The lingua he says ‘corresponds exactly with the lingua of Apterygota.” He finds that the chitinous supporting stalks described by Miall and Denny are comparable to those in Apterygota. In Melanopus femoratus, Folsom discovered “ superlingue ” as “large dorso-lateral rounded lobes intimately united with the lingua.” He notes the evidence of the presence of lingua and superlingue in Packard’s figure of Anabrus, Packard (36). The tongue of Blattidee is described by Mangan (29. p. 3, pl. i. fig. 1). “The free tip,” he says, “is furnished at the sides with a pair of elongate plates (Z) which OF MAXILLULA IN THE ORDERS OF INSECTS. 435 carry bristles, and are continuous behind, as thin rods, round the opening of the salivary duct;... The position of ...(Z) is conformable with the idea that they may represent a pai of maxillule.” Schimmer (39) describes the hypopharynx of the Gryllid M/yrmecophila. On the ventral surface are borne a pair of elliptical plates covered with fine setee. Mjoberg (34) has no hesitation in homologising these plates with the so-called “lingual glands” associated with the hypopharynx of Psocide. The nature of these latter structures will be discussed below, but here it may be remarked that Enderlein (15) regards them as maxillule. EPHEMERIDA. As the well-known figure of the lingua and superlinguee of Heptagenia Vayssiere (45) makes evident, these structures attain a very well-developed condition in the Ephemeridee. No attempt had at that time been made to homologise the lateral appendages with maxillule. Vayssicre was, however, Text-fig. 4. Chloéon dipterum. Tongue of nymph. x 90. a. Articulation of maxillula. *. Hypo- pharynx. m. Maxillula. mp. Plate in ventral wall of maxillula. p. Transverse sclerite at base of maxillula. 7, ez. Setose ridges at base of hypopharynx. so much impressed by their appendicular appearance that he proposed to regard the Ephemeridz as insects provided with a secondary labium, con- sisting of the median “langue” together with its ‘deux appendices.” Eaton, in his monograph of 1883, figures the ‘ tongue” and “ paraglosse ”’ of the nymphs of a large number of Ephemeridee. Though the latter organs show considerable variation in size relative to the former, they are never wholly absent. Hansen (21) refers to them as maxillule. This view of their nature is supported by the figure of the head of a young larva of Ephemera, in which Heymons (23. Taf. ii. fig. 29) clearly shows that they are originally distinct from the hypopharynx. From his description (p. 22) it is evident that Heymons regards the appendages as lateral portions of the hypopharynx. 436 MISS A. M. EVANS ON THE SI'RUCTURE AND OCCURRENCE Folsom (17. p. 116) is convineed of their homology with the lingua and superlinguee of Anurida, which in a certain stage of their development bear a great resemblance to Heymons’ figure of Ephemera. Chloéon dipterum.—The oval h¥popharynx (text-fig. 4) measures *3 mm. in length and +21 mm. in breadth at the widest part. The maxillulze (m.) resemble this structure in shape and size, and in the membranous nature of their walls. Hach measures ‘2 mm. in width and extends *3 mm. beyond the sclerite (p.). Each maxillula is supported ventrally by a curved chitinous plate (m.p.), which is proximally merged with the chitinised posterior portion of the wall of the hypopharynx. The inner edges of the maxillulee are continuous with a pair of convergent setose ridges (r.), which extend backward along the upper surface of the hypopharynx. Outside these occur a second pair of ridges (e2.) also bearing sete, whose distal portions lie on the upper surface of each maxillula. PERLARIA. “an unusually Packard (86) describes the hypopharynx of Perlide as large tongue-like mass nearly filling the buccal cavity.” I have examined this organ in the adult of a species of Perlodes closely resembling P. dispar, and find that it answers to the above description. No signs of maxillulze were discovered. The larvee of the two sub-orders Plecoptera filipalpia and Plecoptera seti- palpia Klapalek (27) are distinguished by the structure of the labium. This in the former group is apparently less specialised than in the latter. An examination of the hypopharynx in the two genera Leuctra and Nemura of - the Filipalpia and Perlodes, Chloroperla, and Perla of the Setipalpia suggests that the form of this organ offers a further distinction *. Nemura.—In the larva of an undetermined species of this genus the hypo- pharynx (text-fig. 5) is a much-rounded structure convex dorsally. The surface is thin and membranous. Tt measures ‘13 mm. in length and 2 mm. in width in a specimen with length of abdomen 2°88 mm. From its dorsal surface project two large lobes (m.) measuring ‘09 mm. in length and ‘05 mm. in width, which extend laterally some distance beyond the hypopharynx, and whose cavities are confluent with that of the latter. Ventrally a chitinous plate (p.m.) supports the free portion of each lobe. Proximally the dorsal surface of the hypopharynx (or ventral wall of the pharynx) bears a pair of setose ridges (r.). -Hach ridge is distally continuous with the inner margin of the lateral lobe. Thus the position of these ridges agrees strikingly with * The only traces of maxillule associated with the large fleshy tongue of larval Setipalpia were a pair of very small setiferous lobes, but in the Filipalpia these organs were large, and well developed and similar in both genera examined. OF MAXILLULA IN THE ORDERS OF INSECTS. 437 that of a similar pair which have been described in Chloéon, if we regard the structures im.) as the homologues of the maxillule of this Ephemerid. Further comparison gives additional grounds for this assumption. The lateral lobes are in both cases fused with the hypopharynx on its dorsal surface. They extend beyond it laterally. In both cases they are supported ventrally by a chitinous plate, and a transverse bar of chitin (text-figs. 6 & 7) marks the line of junction to the upper surface of the hypopharynx. The difference existing between Nemura and Chloéon in respect of these organs appears therefore to be concerned merely with the degree of fusion of their adjacent surfaces, Textefig. 5. Nemura sp. Tongue, dorso-lateral aspect. x 250. h. Hypopharynx. m. Maxillula. I gue, It yPopnary p. Transverse sclerite at base of maxillula, p.m. Plate in ventral wall of maxillula, PSOCIDA anp MALLOPHAGA. The tongue of Psocidee (text-fig. 6) and Mallophaga bears a pair of oval, or sole-shaped, structures (p.) as to the nature of which considerable con- troversy exists. They were described by Burgess (5) in the Psocide and termed “lingual glands,” but Burgess suggests that this term might not correctly express their nature. The chitinous cord (¢.c.) proceeding from each, which unite, and pass backwards as a single cord, to be attached to the “bonnet-shaped” skeletal structure (0.s.), are referred to as “ ducts” of these glands. Among later writers this interpretation has been followed by Snodgrass (43) and Shipley (41). The “glands” are, however, considered to be chitinous thickenings of the hypopharynx by Grosse (19), Mjoberg (34), 438 MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE and Cummings (10). Hnderlein (15) expressed the opinion that the “glands” are appendages homologous with the maxillule of Collembola, while Borner (4) calls them the “fulturee” or lingual stalks, and says that the maxillee are represented by a pair of membranous lobes attached to the tongue. The mouth-parts of British Psocidee are, owing to their minute size, exceedingly difficult of dissection, but, through the kindness of Dr. Imms, Text-fig. 6, Large Psocid. Tongue, dorsal aspect. The plates m. and chitinous cord e¢.c. are seen through the membranous dorsal surface of the hypopharynx and maxillula. x 250. b. Brush-like sclerite. e¢.c. Chitinous cord = “duct.” h. Hypopharynx. @ Lobe of hypopharynx. m. Maxillulay lobe of hypopharynx. p. Chitinous plate=“ lingual gland.” 0.s. Gisophageal sclerite. s,m. Sclerite rising from apex of plate (p.) and associated with m. ¢, Tendon. I have been able to examine specimens of a very large species which is a native of India. The hypopharynx, whose walls are of very delicate mem- brane, is figured (text-fig. 6) from the dorsal surface. The ventral wall is thickened laterally to form the paired, convex, sole-shaped plates of yellow chitin (p.) (seen by transparency), the so-called “lingual glands.” The size of the specimens made it possible to dissect out the hypopharynx and open up OF MAXILLULZ IN THE ORDERS OF INSECTS. 439 the dorsal wall, so that there was no doubt at all as to the exact position of these plates. Chitinous sclerites (s.p.1) and (s.p.2) connect the outer edge of each with the upper surface of the hypopharynx. Distally arise sclerites (s.m.) reaching to the apices of the delicate lateral lobes (m.). These lobes bear the same relation to the hypopharynx as do the maxille of Forjicula, and it seems highly probable, as Borner (4) suggested, that they represent these appendages in a somewhat modified condition. A pair of membranous lateral projections (/.) arises behind them and may possibly represent a basal lobe of the maxillule. Attached to the base of each plate (p.) are the forks of a strong tendon (t.), whose presence strongly suggests that the chitinous structures associated with the hypopbarynx are mechanical in function. In order to settle the question whether or not gland tissue is to be found associated with the plates (p.), nymphal specimens of a small British species of the Psocina group were sectionized. The fixatives employed were Carnoy’s fluid (cold) and Gilson’s fluid (hot), and the sections were stained with Heidenhain’s iron hematoxylin in some cases and Grenacher’s hematoxylin and eosin. The tissue lying between the plates and the upper surface of the tongue was indistinguishable in structure from that seen in many parts of the other mouth appendages. In transverse sections through the tongue no histological difference could be detected between the median portion and that lying above the plates. This investigation, therefore, lends support to the assertion that these latter structures are not in any way glandular or asso- ciated with glands. They are to be looked upon as thickened parts of the ventral wall of the tongue, which may possibly have originated as lingual stalks, or as the basal portions of maxillulee which have become incorporated into the hypopharyngeal wall. NEUROPTERA. Raphidia.—Adult forms of a species of Raphidia were examined and in them the hypopharynx revealed some features of great interest. Text-fig. 7 (p. 440) represents the labium and floor of the mouth as seen when viewed from the dorsal surface; (p.) is the labial palp and (/.) the bifid, dorsally flexed, extremity of the single median endite borne by the mentum. See Westwood (46. pp. 45-56). The opening of the salivary duct occurs above the labium and ventral to the median fleshy projection (h.) in the floor of the mouth. This rounded lobe I regard as the hypopharynx or its distal portion. It is almost obscured by two others of similar membranous texture (m.). Each lateral lobe measures ‘14 mm. in length and +11 mm. in width. In texture and relative positions these three lobes bear a decided resemblance to the median hypopharynx and lateral maxillulee of Forficula. Since the Neuroptera are amongst the less specialised Kndopterygota it is possible that a true homology underlies this similarity. 440 MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE In the larvee of Sialis and Chrysopa the hypopharynx was not found to bear any distinct indications of maxillulee. Sialis.—The labium of the adult Sialis is described by Westwood (46. vol. ii. p. 50). He states that the true labium, “ligula,” does not extend beyond the palpi but is “internally dilated.” Packard (36. p. 73) describes the lingua of Sialis. It is, he says, short rounded and little developed. He notes the presence of sensory hairs at the edge similar to those borne by threads of the maxillary and labial palpi. Text-fig. 7. Raphidia sp. Tongue, dorsal aspect. 90. a.m. Chitinous bar arising from mentum. h. Hypopharynx. 2. Ligula. m. Maxillula. p. Labial palp. In text-fig. 8 (p. 441) this structure together with the hypopharynx (h.) is shown as viewed from before. (U.) is the anterior border of the labium (ligula). At each side the Jabium is connected internally to the palp (p.), with a large rounded lobe bearing small fine setee and at the end a row of long stiff hairs: apparently those referred to by Packard. Viewed from the dorsal surface these lobes present much the same appearance as the similarly located structures in Raphidia. They may therefore be maxillule. This view as to their nature is based partly on the fact that here as in Raphidia their texture resembles that of the mouth-floor and differs from that of the “ligula,” OF MAXILLULZ IN THE ORDERS OF INSECTS. 441 Hemerobius.—The appearance of the mouth-floor of an adult specimen of Hemerobius sp. as seen trom the dorsal surface is illustrated in text-fig. 9 (p. 442). Ibis seen that from the membranous hypopharynx (1) arise a pair of lateral lobes (m.), the surface of which, like that of the hypopharynx, is covered with minute sete (not skown in the figure). These lobes are, according to my interpretation, mayillulee which have assumed a somewhat ventro-lateral position with regard to the hypopharynx. Later a fresh specimen of Chrysopa flava was examined. The labium and hypopharynx resembled those of Hemerobius in general characters and also in their relative position. Prominent lateral plates were present at the sides of the hypopharynx, entirely clothed with long fine sete. They could be reflected outwards to a considerable extent. It was hoped that investigation of the hypopharynx in larval and adult Neuroptera might throw light on the origin of the condition in which it Text-fig. 8. Sialis lutuaria. Labium and hypopharynx viewed from before. h. Hypopharynx. /. Ligula of labium. m. Maxillular lobe. p. Palp. o.s. Opening of salivary duct. s, Sete bordering maxillula ventrally. occurs in some of the mandibulate forms of Endopterygota. Now, let us suppose that the labium of Forjicula were so reduced as to bear but a single median endite, and its sides fused with those of the tongue. The resulting structure would bear a distinct resemblance in constitution and appearance to the condition in Raphidia. These considerations suggest that in the latter such reduction and fusion have taken place. The maxillule in Forjicula would, moreover, oceupy precisely the same position with regard to hypo- pharynx and labium as do the lobes (m.) in Raphidia. It is therefore probable that these latter are wholly or partially the persistent maxillule. Ina later part of the present paper I have attempted to demonstrate a LINN. JOURN,—ZOOLOGY, VOL. XXXIV. oD 442 MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE close correspondence between the construction and arrangement of the hypo- pharynx and Jabium in a Tenthredian larva and that of Sialis. I have further suggested a possible mode of derivation of the condition in larval Lepidoptera and Trichoptera from one less specialised such as we found in larval Tenthredineze. If there is any foundation for these suggestions then the Neuropterous types studied have afforded some clue to the mode of development of the highly specialised labio-hypopharyngeal structure of Lepidoptera from the free tongue and generalised labium of Forficula. They have, further, made it apparent that the possession of a pair of lobe-like projections is.a feature of the hypopharynx in the Neuropterous forms studied, In dealing with the Tenthredinew, reasons are given for assuming Hemerobius sp. Labium and hypopharynx, dorsal aspect. &. Hypopharynx. . Dorsally flexed ligula, dp, Labial palp. om. Maxillula. | s, Sclerite supporting hypopharynx. that the corresponding structures (m.) in that family are homologous with the maxille of Lepidopterous larvee I suggest, therefore, that in Neuroptera we find a condition of hypopharynx and mavxillule intermediate between that of Forficula and larval Lepidoptera. COLEOPTERA. In the larva of //elodes Carpenter and McDowell (9. pp. 373-96) have deseribed and figured a. pair of articulated appendages, one at each side of the hypopharynx. OF MAXILLUL& IN THE ORDERS OF INSECTS. 443 Homologous though reduced structures were found in Dascillus. These again were represented in a further reduced condition in the larva of the Lamellicorn Geotrupes. In a similar situation on the left side of the hypo- pharynx of the larval Phyllopertha occurred a very small lobe bearing teeth, a scarcely recognisable maxillula. In each of these cases it appears that the hypopharynx is membranous and is supported by a system of chitinous sclerites. It covers almost the whole of the surface of the labium to which it is fused or sutured in front. Of a totally different character is that of the larval Pterostichus described by Carpenter (7. p. 213, fig. 6). The anterior border is here broad transversely, set with long sete and not united to the labium. Where this border comes into close proximity to the base of the maxilla it bears a small setose projection. This is regarded as a Text-fig. 10, Nebria brevicollis. Wypopharynx and labium, dorsal aspect. x 90. h. Wypopharynx. @. Labium. m, Maxillula (left). maxillula, agreeing as it does in position with the process described by Magan (1912) in larval Dytiscidee. This Jatter type of hypopharynx I have found to occur in a modified form in larvee of Staphylinide, Hlateridw, and the Malacoderm Telephorus litur- atus. In other larvee examined, Coccinellidee, Tenebrionide, Chrysomelide, Cerambycide, and Curculionidee, the structure was of the soft membranous character supported by sclerites. In the former group the labium extended considerably beyond the hypopharyns, and was not fused therewith anteriorly as was the case in the latter group. 444. MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE In the Carabid, Nebria brevicollis (text-fig. 10), the larval hypopharynx (h.) is produced at the corners into two prominent lobes (m.) identical in position with the maxillulee of Pterostichus. Apparently they are homologous with these appendages. Each measured ‘09 mm. in width, and the combined width of maxillule and hypopharynx was *6 mm. No distinct suture separating the maxillula from the hypopharynx could be detected. The maxillulee here were more prominent than in the larvee of Pterostichus and Carabus violaceus which were examined. In various larval Elateride the hypopharynx was examined and great uniformity of construction was found to exist. The condition seen in Campylus linearis will serve to illustrate the type. The hypopharynx (text-fig. 11), which is of stiff chitin, is supported by a Text-fig. 11. Text-fig. 12. Campylus linearis. Labium and Anatis ocellata, Labium and hypopharynx, hypopharynx, dorsal aspect. dorsal aspect. x 55. a, Anterior skeletal x 90. hk, Hypopharynx. system. f. Wypopharynx. dp. Labial palp. 1. Labium. m. Maxillula. m. Maxillnla. yp. Posterior skeletal system. stronger transverse chitinous plate. Its anterior border is independent of the underlying surface of the labium. It bears a pair of triangular plates (m.), measuring at the base (05 mm. These I regard as vestigial maxillule from considerations based on a comparison with those of Pterostichus. The com- bined width of the hypopharyux and maxillule is *24 mm. Hach of the latter carries long, fine hairs which are mingled with similar but shorter hairs borne by the Jabium and by the basal Harton of the maxille. The larva of Telephorus lituratus is described by Miss O. M. Payne (38). From pl. 2. figs. 14, 15, in which the hypopharynx is shown, it seemed that this structure was very similar to that of Hlateridee. Examination of the mouth-parts of a specimen confirmed this supposition. I discovered hidden Of MAXILLULE IN THE ORDERS OF INSECTS. 445 by the long hairs which clothe the hypopharynx and labium a pair of very minute triangular projections situated at each side of the anterior border of the hypopharynx. The similarity in character of the hypopharynx and maxillulse and their relations with the labium in species of the Carabidee, Dytiscide, Elateride, and Malacodermidee gives support to the conclusions reached by Gahan (18) regarding the affinities of the Coleopteran families. This author states that the Adephaga, including the Carabide and Dytiscide, is the most primitive group of the Coleoptera, and that the Malacoderms are a primitive family of the Polyphaga, and gave rise to the Elaterids: to which they are closely related. Tam indebted to Dr. Imms for specimens of the Coccinellid larva, Anatis ocellata. The hypopharynx (text-fig. 12) possesses a membranous, greatly convex, surface covered with minute spines. It is strengthened by an anter‘or (a.) and a posterior (p.) skeletal system. Supported in the angle between the two anterior arms of the latter arises on each side a rounded lobe (m.). Hach measured ‘11 mm. in width, and.the greatest width of the hypopharynx is ‘32 mm. These paired lobes may be compared in form (though they do not bear spines), and position upon the hypopharynx with the vestigial maxillulee of Geotrupes. A condition very similar to that last described was met with in the larva of Doryphora decemlineata where, however, the lobes were of much larger extent, practically covering the surface of the hypopharynx. The hypopharynx in Zenebrio mollitor and Otiorhynchus sulcatus was found to possess indefinite paired dorsal lobes, which, however, could not with any certainty be looked upon as maxillule. LEPIDOPTERA. De Gryse (12. pp. 173-179) records the results of exhaustive investigations on the hypopharynx in a large number of Lepidopterous larvee. On p. 174 he makes the generalisation that in larval Lepidoptera the maxillule “essentially present the appearance of protruding fleshy lobes, covering the floor of the buecal cavity wholly or in part only,..... These lobes are generally clothed with flexible lashes, with hairs or with rows of strong SPINecane sicace: In many instances the lobes are also furnished with chitinous bladesi.2 275. A feature common to the types described is the presence of a chitinous bar or “‘arm” arising from the mentum and at its distal end associated with the fleshy lobe. Ihave examined the hypopharynx of a number of species of Lepidoptera belonging chiefly to the families Nymphalidee, Pieridee, Geometride, Noctuide, Lasiocampidee, Pyralidee, and Tineidee. In most cases there was little that called for comment. 446 MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE All the Noctuze observed were found to agree remarkably closely with the condition found in Mamestra persicarie. The right maxillula of MM. persicarie (text-fig. 13) measures *51 mm. in length from the point (@.) Irom this level to the apex of the hypopharynx the distance is ‘88 mm. A thinly chitinised plate (p.) supports the outer wall of the maxillula to which it is closely applied. It is fringed with freely projecting teeth. The greatest width of the left lobe is*2 mm. The mental arm (a.m.) supports the chitinous plate for the greater part of its length. The Rhopalocera investigated were not found to be strikingly different in the character of the hypopharynx from the Heterocera. In Vanessa urtice Text-fig. 13. | Mamestra persicarie. Labium and hypopharynx, dorso-lateral aspect. x 55. a.m. Mental arm. hk. Hypopharynx. Jp. Labial palp. om. Maxillula. p. Plate supportiing outer wall of maxillula. s.p. Spinneret. this structure was of a form commonly occurring amongst the latter group. This type, represented by Lagoa crispata, De Gryse (12. pl. 17. fig. 1), is found in both micro and macro-lepidoptepa (p. 176). The fleshy lobes are ‘covered with spines but devoid of all traces of blades.” In Pieris brassice and Philosamia cynthia the maxillular lobes are very large in proportion to the hypopharynx, the proximal half of which they completely covered. Considerable attention was devoted to examination of the condition of the OF MAXILLUL IN THE ORDERS OF INSECTS. 447 hypopharynx in Hepialus humuli. Although the family to which this species belongs is in many respects primitive the hypopharynx is apparently con- siderably specialised. It does not seem to afford any clue to the original form of the maxillu'a in Lepidopterous larve. The whole of the mouth- floor is remarkably soft, rounded and membranous, and the areas of chitin- ization very small. A pair of lateral membranous protuberances were too indefinite in character to be homologised with maxillule. TRICHOPTERA. The hypopharynx has been investigated in eruciform larve of Phryganea and Limnophilidz and in the campodeiform Rhyacophila and Hydropsyche. It was found in most cases to bear a general resemblance to that of Lepido- ptera. Its dorsal wall, forming the floor of the mouth, is produced distally to Text-fig. 14. Rhyacophila sp. Labium and hypopharynx, dorsal aspect. x 250, am. Mental arm, h. Hypopharynx. lp. Labial palp. m. Maxillular lobe. p. Plate supporting outer wall of lobe. s.p. Spinneret. form a conical spinneret. Liaterally its walls are continuous with those of the labium. With the exception of Phryganea all the examples investigated were found to be constructed on a common plan, which can be best described by a consideration of the condition in a species referable to the genus Rhyacophila, 448 MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE The hypopharynx (text-fig. 14) measures ‘25 mm. in length and +22 mm. in greatest width in an almost fully-grown larva. In its median portion the upper surface is thrown into a pair of convex ill-defined lobes (m.) set with stout curved spines. Each is supported externally by a chitinous plate (p.) measuring ‘05 mm. in length and -03 mm. in breadth. The free edge of this plate is furnished with long hook-like spines. It arises from a chitinous lateral expanse (a.m.). This latter is seen to be continuous ventrally with the thickened portion of the mentum and is apparently bomologous with the chitinous “arms of the mentum” in Lepidoptera. (In the larva of a Limophilid a narrow bar of chitin is found in this situation.) The condition of the mouth-floor in Hydropsyche sp. was similar to that of Rhyacophila sp. It has been shown that pxired lobes associated externally with chitinous plates occur in widely separated families of Trichoptera. This evidence seems to suggest that the possession of these structures is a feature characteristic of Trichopterous larvee with certain exceptions. The hypopharynx, when these structures are present, is comparable in constitution to that of the more generalised Lepidoptera. In addition to the sctiferous lateral lobes, it is characterised in both orders by (1) the possession of a conical spinneret, (2) the presence of chitinous arms extending from the mentum and associated dorsally with the exterior of the base of the lobes, and (3) the membranous rounded nature of the floor of the mouth. Thus the hypopharynx of Trichoptera is seen to be constructed on a plan fundamentally similar to that of Lepidoptera. The maxillule of the latter order are therefore in all probability represented in Trichoptera by the similarly located lobes (m.). HYMENOPTERA TENTHREDINEA. A vertical section through the head of Hylotoma rose Berlese (1. p. 522, fig. 595) shows that the duct of the labial or spinning gland rans above the labium to open above its extremity. Dorsal to the duct is seen a fold which constitutes the mouth-floor. The distal portion of this fold is, in all probability, the hypopharynx in Tenthredinian larvee. In a recent paper on the immature stages of the Tenthredinoidea McGillivray (31) fully describes and figures the mouth-parts of Pamphilius dentatus. It is stated that in Pamphilius, “On the dorsal surface of the ligula and laterad of the labial palpi, there is on each side a protuberance, a paraglossa. Between the labial palpi and arising on the ental surface, there is a prominent Jobe which represents the glossa. The glossa has been modified into a spinneret (figs. 9, 13, sp.) for the opening of the duct of the silk glands.” In fig. 13 the spinneret is seen as a structure projecting beyond the ventral surface of the ligula. It is therefore probably more highly developed than in the forms to be described. OF MAXILLULA IN THE ORDERS OF INSECTS. 449 This structure has been examined in various species, and considerable diversity was found to exist. Perhaps the simplest condition was that met with in Selandria siaii. The anterior portion of the labium (text-fig. 15) bears the three-jointed palps, and is scarecly differentiated into a median (1.) and two lateral portions (p.g.). The labium forms the ventral wall of a wide depression (f.) at the base of which is the end of the spinning duet. Dorsally this funnel is bounded by a rounded chitinous protuberance (.), which is connected laterally with a similar pair of lobes (m.) closing in the sides of the funnel. They are the “ glossa” and * paraglosse” of McGillivray (31). Each lateral lobe measures "18 mm. in length and ‘16 mm. in breadth, From their position in the floor of the mouth, and above the labial duet, I conclude that the median structure is the hypopharynx, respectively Text-fig. 15. Selandria sivii, Labium and hypopharynx, dorsal aspect. x 55. h. Hypo- pharynx. 7. Ligula, m, Maxillula. 2p. Labial palp. p.y. Paraglossie. and that the laterally associated pair may be maxillule or, at least, partly maxillular in constitution. : The mouth-parts of the larva of Wematus Lrichsonii are typical of those of the majority of Tenthredines: investigated. The surface of the hypo- pharynx and its dorso-lateral lobes is membranous and of rounded indefinite form. The larva of Fenusa melanopoda was found mining the leaves of alder- trees. When mature it measures 9 mm. in length. The internal structure of the head bears a striking resemblance to that of the young Acrocecrops Heinrich and De Gryse (loc. cit. pl. 3. figs. 1 & 2). The mouth-parts are in both cases reduced and are associated with slender skeletal rods traversing the entire length of the head. The mouth-floor of Fenusa melanopoda, viewed from above, presents the 450 MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE appearance shown in text-fig. 16. Between the vestigial labial palpi (m.p.) projects the rectangular flat membranous structure (h.). Since this structure lies above the labial duct it is here considered to be the hypopharynx. It is supported laterally by two arms of the skeletal system. ‘The length is ‘08 mm. and the greatest width ‘09 mm. ‘The dorsal surface is markedly convex and bears two large convex plates (m.) covered with rows of minute spines. I regard these plates as maxillulee, which have become flattened in accordance with the dorsi-ventral compression of the head. A comparison between the anterior aspect of the labium and hypopharynx of Sialis and Selandria (text-figs. 8 & 15) reveals a close agreement with regard to the arrangements of the constituent elements. The salivary opening in both cases causes a slit-like depression above the border of the labium. Text-fig. 16. Fenusa melanopoda. Hypopharynx and maxillule, dorsal aspect. The maxilla is indicated on the left side only. x 400. A. Hypopharynx. m.p. Labial palp. m. Maxillula. ma. Maxilla. Dorsal to it is a median lobe, the hypopharynx, whose lateral lobe-like expansions form the lateral walls of the depression. These latter are asso- ciated with the labium at each side behind the base of the palps. In Stalis, however, the ligula of the labium exists as a definite setose plate which its not represented in Selandria. This comparison points to the conclusion that the labium and hypopharynx of Neuroptera and larval Tenthredinez are constructed on the same fundamental plan. In discussing the labial gland of Tenthredinian larve, Berlese (1. p. 522) described its termination “sotto la lingua in una specie di filiera.’ This leads us to enquire how far the condition here may be compared with that of Lepidoptera. We have noticed that the labium in larval Tenthredinez is in a con- siderably reduced state, while in Lepidoptera it has become almost OF MAXILLULZ IN THE ORVERS OF INSECTS. 451 unrecognisably modified to form the basal portion of the spinneret. Now in larval Trichoptera the spinnerct is intermediate in development between that of Lepidoptera and the incipient condition of Tenthredinese. It there- fore appears likely that in the two orders in which a spinneret occurs it is derived from such raw materials as are afforded by the state of the hypo- pharynx and labium of Tenthredineze. Assuming this to be the ease, the hypopharynx of the latter order is homologous with the proximal portion of that of Lepidoptera. We should then expect that the simple lobes borne at the anterior lateral border of the hypopharynx in Tenthredineze would be represented in a modified form in Lepidoptera, where they would occur at the sides of the hypopharynx behind the base of the spinneret. In this position we find the paired lobes considered by De Gryse as maxillule. It, therefore, appears to me probable that the lateral lobes we have described in the mouth-floor of Tenthredinian larva: are homologous with the maxillule of Lepidoptera. Text-fig. 17. Bibio pomone. Hypopharynx, dorsal aspect. The outline of the labium is indicated beneath the surface of the hypopharynx. /. Hypopharynx. i. Apical processes of labium. m. Mavxillula. s, Sclerite supporting mayillula externally. DIPTERA. Bengtsson (2) announced the discovery of maxillee associated with the endolabium in Tipulid Phalacrocera, stating that the latter structure was innervated by a pair of endolabial nerves arising from the sub-cesophageal ganglion dorsal to the mandibular nerves. Traces of maxillulee were sought for in the larvee of various genera of Nemoceran diptera. These included Chironomus, Tanypus, Bibio, Muyceto- phila, Tipula, Ptychoptera, and Simulium. In the majority of cases the views of authors as to the exact limits of labium and hypopharynx were so various as to render impossible any definite conclusion as to the presence of maxillulee associated with the latter. 452 MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE In the case of Chironomus, however, we have the statement by Miall and Hammond (33. p. 29) that ‘The salivary duets pass forwards to open above the mentum and behind a minute projection in the floor of the mouth (lingua).” When viewed from the dorsal surface the lingua (hypopharynx) of Chironomus sp. is seen to consist of a broad plate bearing at each side a wing-like projection which is inclined at right angles to the floor of the mouth. I have compared this condition with that found in two species of Bibio— B. pomone and B. Johannis. Specimens of the latter species were kindly given me by Mr. H. M. Morris, M.Se., and my description refers to B. pomone. The only differences discovered between this species and B, Johannis with regard to the hypopharynx were those of size. Morris (344. p. 96) describes the mouth-parts in his paper on the larval and pupal stages of this Nemoceran. I found that the salivary duct ran immediately dorsal to the thick bidentate structure, which is considered to be a portion of the labium. I therefore conclude that it represeuts the whole labium. In lateral view it appears not as a simple plate, but of composite nature. Above the end of the salivary duct is the membranous anterior portion of the mouth-floor. This has been described as the upper plate of the labium. Its position with regard to the Jabial duct appears to me to be conclusive evidence of the fact that it corresponds to the lingua or hypo- pharynx of Chironomus. Laterally it bears a pair of forwardly projecting pointed processes (text-fig. 17, m.)*2 mm.in length .These are so disposed that together with the hypopharynx they form a membranous crescentic structure, The exterior surface is supported by the arm of a chitinous sclerite s. The ventral portion (v.) of this sclerite is associated by means of a chitinous conuective with the lateral angles of the labium. The form of the hypopharynx and its lateral projections recalls in some features that of Zlibius, one of the less specialised Dytiscids, which belong to the most primitive suborder of Coleoptera. Moreover, Morris (1917, p. 104) gives reasons for'regarding the larva of Bibio as primitive amongst Diptera. Thus, if maxillule exist in Dipterous larvee, we should expect to find them present in such a form as this. The suggestion is therefore put forward that further investigation may demonstrate the maxillular natare of the processes borne by the hypopharynx in the larva of Bibio, and of the similarly located projections in Chironomus. SUMMARY OF CONCLUSIONS. Briefly, then, the presence of maxillule as a pair of appendages more or less closely associated with the hypopharynx is a characteristic feature of Apterygota. They are in the least modified condition in Machilis. I have emphasised the great degree of reduction which they present in Lepisma. They are well developed in Forficula though partially fused with the OF MAXILLULA IN THE ORDERS OF INSECTS. 453 hypopharynx. In the Ephemeridze the maxillule lie between the mandibles and maxilla, though they are articulated to the base of the hypopharynx. In larvee of certain Perlaria Filipalpia I find structures projecting from the sides of the hypopharynx, which bear comparison with the maxillule of Ephemeride. In the section Setipalpia these appear to be represented by small setiferous convexities on the surface of the hypopharynx. My investi- gations do, I believe, tend to support the view that partially fused maxillulse may be recognised in the membranous lobes associated with the hy popharynx in Psocidee and certain Mallophaga, and that the so-called “lingual glands” ave chitinizations of the ventral wall of the hypopharynx. In the Endopterygota the presence of maxillulee is established in four families of Coleoptera. I have deseribed structures which I regard as possibly homologous with these appendages in Elaterids, Staphylinidee, Telephoridee, Coccinellidee, and Tenebrionidee. Reference has been made to others of a more doubtful nature in Doryphora and Otiorhynchus. The hypopharynx of Trichopterous larvee bore comparison with that of Lepidoptera. On its dorsal surface were structures apparently homologous with the maxillule: of Lepidoptera. Larvee of Sialis and Chrysopa did not appear to possess these appendages. In the adults of species of these two genera of Neuroptera and in Raphidia the hypopharynx bore lateral strue- tures which, from comparison with Forjiewla, seem to be wholly or partially maxillular in nature. Similarly located lobes, though modified in form, oceur in larval Tenthredinesze. Attention has been called to the presence of a pair of projections associated laterally with the hypopharynx of two Nemoceran Diptera, Bibio and Chironomus. It is doubtfully suggested that these are of the nature of maxillule. A comparative review of the various conditions of the hypopharynx and maxillulee of insects from the Apterygota to the most highly specialised mandibulate Pterygota, points to the existence of a tendency to modification along certain lines, as follows :— (1) Greater association of the maxillulze with the hypopharynx. (Com- pare Lepisma with Machilis, Blatta, and Forjficula ; Perlodes with Vemura). (2) Reduction in the size of the maxillulze. (3) In Exopterygota the formation of a compound “ tongue” composed of the hypopharynx and maxillulee carried by a forwardly pro- duced portion of the mouth-floor, specially developed for the purpose, This is seen in an incipient condition in Chlocon and Nemura. It is well formed in Forjicula, Blattidee, and probably in Psocidee. 454 MISS A. M. EVANS ON THE STRUCTURE AND OCCURRENCE (4) In the Endopterygota fusion of the distal portion of the mouth-floor with the labiam. This occurs in the more specialised Coleoptera. Lateral fusion has taken place in the Neuropterous forms examined, The Tenthredinese, Trichoptera, and Lepidoptera represent, consecutive stages in the formation of a spinneret, by prolongation of the intimately fused labium and hypopharynx around the opening of the labial duct. Finally, I wish to lay emphasis on the tentative character of the “con- clusions’’ here reached. In certain of the orders and families considered maxillule were not hitherto known to be present. It is hoped that sub- sequent investigation into the embryological development and morphology of allied forms will reveal the truth with regard to this interesting question. (1) 1909. (2) 1897. (3) 1906. (4) 1904, (5) 1887. (6) 1903. (7) 1912. (8) 1916. (9) 1912. (10) 1913. (11) 1915. (12) 1915. (13) 1915. (15) 1903. BIBLIOGRAPHY. Bertesr, A.—“Gli Insetti.” Vol. Primo, 1004 pp.; 1292 figs., 10 pls. Milan. Benarsson, S.— Bidrag till Kannedomen an larven af Phalacrocera replicata (Lin.).”. Acta Regie Soe. Physiog. Lund. Lunds Universitets Arsskrift, xxiii. pp. 1-102, pls. 1-4, ——— ‘Zur Morphologie des Insekten Kopfes.” Zool, Anz., Bd. xxix. Nr. 15, pp. 457-476, Borner, C.—“ Zur Systematik der Hexapoden.” Zool. Anz. xxvii. pp. 511-583. Burerss, E.—“The Anatomy of the Head and Structure of the Maxilla in Psocidee.” Proc. Boston Soc. Nat. Hist. vol. xix. pp. 291-296. Carpenter, G. H—‘“ On the relations between the Classes of Arthropoda.” Proc. Roy. Irish Acad. vol. xxiv. B. pp. 820-860. ——— “The presence of Maxillulee in Beetle larve,” Trans. 2nd Entom. Congress, pp. 208-215 ; 6 text-figs. —— “The Apterygota of the Seychelles.” Proc. Roy. Irish Acad. vol. xxxiii. Sect. B. No. 1, pp. 1-70, pls. 1-18. Carpenter, G.H., “The Mouth-parts of some Beetle larvee (Dascillidee and & McDowrii, M.C. Scarabeeidee), with special reference to the Maxillule and Hypopharynx.” Q.J.M.S. vol. lvii. pp. 873-396, pls. 85-37. Cummines, B, F.—“On some points in the Anatomy of the Mouth-parts of the Mallophaga.” Proc. Zool. Soc. Lond. pp. 128-141. ——— “Descriptions of five new species of Anoplura and Mallophaga.” Bull. Ent. Research, vol. v. pp. 155-177; 8 figs. Dr Gryssr, J. J.—‘‘ Some Modifications in the Hypopharynx of Lepidopteran Larvee.” Proc. Ent. Soc. Wash. vol. xvii. pp. 173-179; 3 pls., I text-fig. Hetneica, C., & “ Acrocecrops strigifinitella.” Proc. Ent. Soc. Wash. Dr Gryse, J. J. vol. xvii. pp. 6-14; 9 pls. . Baron, A. E.—“A Revisional Monograph of the Ixisting Ephemeride.” Trans. Linn. Soe. vol. iii. ENDERLEIN, G.—‘‘ Ueher die Morphologie, Gruppierung und systematische Stellnng der Corrodentien.” Zool, Anz, xxvi. pp. 424-428. (16) 1889. (17) 1900. (18) 1911, (19) 1885. (20) 1915. (21) 1893. (22) 1894. (23) 1896, (24) 1906. (25) 1913. (26) 1896, (27) 1909. (28) 1801. (29) 1908. (30) 1912. (31) 1913. (32) 1867. (33) 1909. (34) 1910. (344) 191 (35) 1887. (36) 1903. (37) 1886. OF MAXILLULA IN THE ORDERS OF INSECTS. 455 Fousom, J. W.—‘ The Anatomy and Physiology of the Mouth-parts of the Collembolan Orchesella cincta.” Bull. Mus. Comp. Zcol., Harvard, vol. xxxv. No. 2; 40 pp., pls. 1-4. ——— “The Development of the Mouth-parts of Anurida maritima. Bull. Mus. Comp. Zool., Harvard, vol. xxxvi. No. 5, pp. 87-157; 8 pls. Ganan, C. J.— On some recent attempts to Classify the Coleoptera in accord- ance with their Phylogeny.” The Entomologist, p. 121. Grosse, F.—‘ Beitrage zur Kenntniss der Mallophaga.” Zeitschr. Wissensch. Zool, xlii. pp. 536-540. Happon, K.—“On the Methods of Feeding and the Mouth-parts of the Larva of the Glow-worm, Lampyris noctiluca.” Proce. Zool. Soc. Lond. Pile ppy qi-82; plead. Hansen, H. J.—“ Zur Morphologie der Gliedmassen und Mundteile bei Crus- taceen und Insekten.” Zool. Anz. xvi. pp. 198-8, 201-12. (Transl. in Ann, Mag. Nat. Hist. (6) xii. 1893, pp. 417-432.) —— “On the Structure and Habits of Hemimerus talpoides, Wall.” Ent. Tidskr., Arg. 15, pp, 65-93, pls. 2, 3. Heynons, R.— “Grundziige der Entwickelung und Anhang des Korpersbau von Odonaten und Ephemeriden.” Abh. Kongl. Preuss. Akad. Wiss., Berlin; 66 pp. and 2 Taf. Inus, A. D.—“ Anurida.” L.M. B.C. Memoirs xirr. 99 pp. and 7 pls. “‘ Contributions to a Knowledge of the Structure and Biology of some Indian Insects. II. On Lmbia major sp. noy. from the Himalayas.” Trans. Linn, Soe. vol. xi. pp. 167-195, pls. 36-38. Kextog, V. L.--“ New Mallophaga.” Proc, Calif. Acad. Sci. ser. 2, vi. pp. 31- 168, pls. 2-15. KLapALEK, F.—“ Plecoptera” in “Die Siisswasserfauna Deutschlands ” heraus- eegeben von Braver, Heft 8. Jena. Korscustt, E. U., “Lehrbuch der vergleichenden Entwicklungs-geschichte & Hriper, K. der Wirbellosen Thiere,” xii. ; 1509 pp., 889 figs. Jena. Manean, J.—“ On the Mouth-parts of some Blattide.” Proce. Roy. Tish Acad. vol, xxvii. Sect. B., 10 pp. and 3 pls. —-— “The presence of Maxillule in Larvae of Dytiscide.” Mem. and Proc. of the Mane, Lit. & Phil. Soc. vol. lvi. Part IL. ; 6 pp., 2 figs. McGinurvray, —. “The Immature Stages of the Tenthredinoidea.” 44th Ann. Report of the Entomological Soc. Ontario, Metert, F.—(Translation from 1865) “On the Campedew, a family of Thy- sanura.” Ann, Mag. Nat. Hist. ser. 3, vol. xx. pp. 861-378 ; 3 figs. Mratr, L.C. “The Structure and Life History of the Harlequin-fly (Chiro- & Hamnonp, A. ft. nomus).” Oxford; Clarendon Press. 196 pp., 1 pl., 127 figs. Msonera, E.—“ Studien tiber Mallophagen und Anopluren.” Arkiv for Zoologi, Bd. vi. No. 13. 7. Morris, IT. M.—‘‘ On the larval and pupal stages of Bibio johannis L. Ann App. Biol. vol. iv. pp. 91-108; 1 pl., 12 figs. OvuprMans, J. T.—“ Bijdrage tot de Kennis der Thysanuren und Collembola.” 104 pp., 3 pls. Amsterdam. Packarp, A.S.—“A Text Book of Entomology.” 729 pp., 654 figs. New York. ——— “A new arrangement of the Orders of Insects.” The American Naturalist, xx. p. 808. ON THE STRUCTURE OF MAXILLULZ IN INSECTS. Payne, O, G. M.—“ On the Life History and Structure of Telephorus lituratus, Fall.” Journ, Zool. Researvh, vol. i. Pt. I.; 2 pls., 18 text-fiys. . ScuimMer, I*.—“ Beitrag zu einer Monographie der Gryllodeengattung Myrme- cophila Latr.” Mit Tafel xxii-xxiv. und 26 fig. im Text. Zeitschrift f. Wiss. Zool. pp. 410-531. . Scuioprr, J. C.—“ De Metamorphosi Eleuteratorum Observationes.’ Natur- hist. ‘Tidsskrift, Pars L., Tab. ix. . Surpiey, A. .—“The Ectoparasites of the Red Grouse (Lagopus scoticus).” Proc. Zool. Soc. Lond. pp. 309-334, pls. 35-47. . SNoperass, R. E.—In “ New Mallophaga, II.” Kelloge, Proc. Calif. Acad. Sci., ser. 2, vol. vi. pp. 484-457, . ——— A Revision of the Mouth-parts of the Corrodentia and the Mallo- phaga.” Trans. Amer. Ent. Soc. pp. 297-305, . Tintyarp, R. J—“The Biology of Dragon-flies (Odonata or Neuroptera).” Cambridge Zool. Series. 396 pp., 4 pls., 188 figs. . Vayssinrp, A.—“ Recherches sur Vorganisation des larves des Ephémérines.” Ann. Sci. Nat., ser. 6, Zool., tome 13 ; 137 pp., pls. 1-11. . Westwoop, J. O.—“ An Introduction to the Modern Classification of Insects, ete.” Vol. Il. London. Pp, 46-56, figs. 64-66. . WiLLEM, V.—“ Recherches sur les Collemboles et les Thysanura.” Mém. Couronnés Acad. Roy. des Sciences de Belgique, lviii.; 144 pp., pls. 1-17. EXPLANATION OF PLATE 31. Petrobius sp. Dorsal aspect of Hypopharynx and Maxillule. cs From a preparation stained with acid fuchsin. a. Indication of articulation. h. Hypopharynx. 7@. Inner lobe of maxillula. Z,. Comb-lilke projection trom surface of maxillula. m. Left maxillula, p. Palp. p.e. Peduncles of hypopharynx. 7. Chitinous rods in ventral pharyngeal wall. Evans Journ. Law. 50c. ZOOL. VoL XXXIV, Py-3l) Sores A < a Lf NaN NSIK SP ) KN) Dye? Span A.M.E. del. C. Hodges &Son.lith. Bees PETROBIUS sp. LINNEAN SOCGIEIY OF LONDON. RULES FOR BORROWING BOOKS FROM THE LIBRARY. 1. No more than Six volumes shall be lent to one person at the same time without the special leave of the Council or one of the Secretaries. 2. All books shall be returned before the expiration of Six weeks from the time of their being taken out, but if not required by any other Fellow, they may, on application, be kept for a further period of Six weeks. 3. All books lent shall be regularly entered by the Librarian in a book appropriated for that purpose. 4. No work forming part of Linneus’s own Library shall be lent out of the Library under any circumstances. Norn.— Certain other works are included im this prohibition, such as costly illustrated works, volumes belonging to sets which could not be replaced if lost, and unbound parts of serials. 5. All carriage charges on books to be defrayed by the Fellow borrowing the same. 6. All borrowed books when returned to be sufficiently wrapped and protected. Borrowers will be held liable for any damage incurred owing to improper packing. 7. Newly published books shall not ordinarily leave the Library until after the expiration of one month from the date of their being received. 4A GENERAL INDEX to the first twenty Volumes of the Journal (Zoology) may be had on application, either in cloth or in sheets for binding. Price to Fellows, 15s.; to the Public, 20s. NOTICES. THE attention of the Fellows, and of Librarians of other Societies, is requested to the fact that the parts of the Journal (Zoology) of the present volume have been issued as follows :— Vou. 84. Nos. 225, 226, 227, 228, and 229 the present number. Authors are entitled to 25 copies of their communications gratuitously, and may obtain another 75 by payment, as shown on the printed slip which accompanies the proof. If more than 100 copies are wanted, application must be made to the Council. Abstracts of the proceedings at each General Meeting and Agenda for the next are supplied to all Fellows. B. DAYDON JACKSON, General Secretary. Frpruary 16, 1922. Price 103. THE JOURNAL THE LINNEAN SOCIETY, CONTENTS. Page ~ I. The Vertebrate Fauna of Houtman’s Abrolhos (Abrolhos Islands), Western Australia. By W. B. Aunxanpmr, M.A. (late Keeper of Biology in the Western Australian Museum). (Communicated by Dr. W. J. Danin, F.L.S8., Professor of Zoology in the Withiversity rol Wivenpools) ie. acs ncaeuyememenee ise. oethees cee tases os 457 il. Annélides Polychétes de ’ Archipel Houtman Abrolhos (Australie Occidentale) recuiellies par M. le Prof. W. J. Dakin, F.L.8. Par Prerre Favuver, Professeur 4 Université catholique d’ Angers. (Communiqué par M. le Professeur W. J. Dakin, D.Se., F.L.8.) (Avec 2 figures dans le texte.)..0.............-.+5- 487 ILI. Sherbornina: A new Genus of the Foraminifera from Table Cape, Tasmania. By Frepertck Cuapmay, A.L.S., F.R.M.S., Palee- ontologist to the National Museum, Melbourne. (Plate 32.)... 501 ~ IV. On a new Type of Teleostean Cartilaginous Pectoral Girdle YI g found in young Clupeids. By E.S.Goopricu, F.R.S., Sec. L.S., Professor of Comparative Embryology in the University of Se Saas. (OPSITOIUES so tn tae Co toa ghee Bem saee carceer Haine Me te pate Af lsu wlan Ins “, Index, Titlepage, Contents, ete. LONDON: SOLD AT THE SOCIETY’S APARTMENTS, BURLINGTON HOUSE, oN icine PICCADILLY, W.1, AND BY LONGMANS, GREEN, AND CO., AND WILLIAMS AND NORGATE. 1922. LINNEAN SOCIETY OF LONDON. LIST OF THE OFFICERS AND COUNCIL. Elected 27th May, 1921. PRESIDENT. Dr. Arthur Smith Woodward, F.R.S. VICE-PRESIDENTS. Horace W. Monckton, F E. T. Browne, M.A. : The Lord Rothschild, F. GS. C. C. Lacaita, M.A. RS. TREASURER. Horace W. Monckton, F.G.S. SECRETARIES. E. 8. Goodrich, F.R.S. | Dr. A. B. Rendle, £.R.S. GENERAL SECRETARY. Dr. B. Daydon Jackson. COUNCIL. Prof. Margaret Benson, D.Sc. | Gerald W. B. Loder, M.A. Prof. V. H. Blackman, F.R.8. | E. T. Browne, M.A. Henry Bury, M.A. Stanley Edwards, F'.Z.S. Prof. E. 8. Goodrich, F'.R.8. R. I. Pocock, F.R.S. Capt. John Ramsbottom, M.A. Dr. A. B. Rendle, F.R.S. Prof. Dame Helen Gwynne-Vaughan. Sir Sidney F. Harmer, K.B.E., F.R.S. Dr. B. Daydon Jackson. C. C. Lacaita, M.A. | Horace W. Monckton, F.G.S. | | The Lord Rothschild, F.R.S8. Dr. KK. J. Salisbury. C. I. Salmon, Esq. Thomas A. Sprague, B.Sc. Dr. A. Smith Woodward, F.B.S. LIBRARY COMMITTER. The Officers ex officio, with the following in addition :— James Britten, Esq. | Dr. W. T. Calman. Ii. J. Collins, B.A., B.Se. L. Y. Lester-Garland, M.A. Dr. R. R. Gates. Dr. A. D. Imms. Dy. E. J. Salisbury. Miss E. M. Wakefield. B. B. Woodward, Esq. THE VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 45 1 The Vertebrate Fauna of Houtman’s Abrolhos (Abrolhos Islands), Western Australia. By W.B. Atuxanper, M.A., Late Keeper of Biology in the Western Australian Museum. (Communicated by Dr. W. J. Dakty, Professor of Zoology in the University of Liverpool.) {Percy Sladen Trust Expedition to the Abrolhos Islands under the leadership of Prof. W. J. Dax1. | {Read 17th March, 1921.] CONTENTS. Page TLE OCILC ELON yer eeet ce. Beta dy reso ciSS aT N inte Here ete ices hed 457 SECTION |: —ThevandiVertebrates. ..... sacs gnc te vaca tireene 459 Naini ald Ape eee Ie eee tafe te. oe cee ee pep Ne forse 459 Vey til tatieerncie see a thers ar kerna a cre cee nce aE Mee cia ae 5 tea 460 AMM Phiblanprmnria et a) wise siecle bh cieidnie deletes aise accncie . 462 aN DILAS peewee gta here fe esse ccs sicko OR o oee beagles 462 RSM AT Versa treed tari inhsseeg cae Uehipo eiticaacs,ahl troker i ebametcs wise shit ste versl4 466 SECTION: 2:— LhesMarinesVertebratese sn. secs c wuss sic ote certo sl tipece ls 468 SETA t ys Ruiter ert een Bs sbepiiced Gud Secale RCE ocean 468 ANB na aV VLEs Ty ey tree Nok es ep it e PaPte aee 478 ROD LILA gene Mere nolzo sh nontoon res er a Eh PEER eRe ares Sle 479 IPAS CES 2 rast? wag eee ee a OTe el co Sea aan nas a RTE Cora eh oliobhe aueenseniiecs 479 SMA Sa oboe ancopas choos pro piclin cecum heed ROE R OTC 484 Weiberabureycrteds pvcets costae ees, c oe eke Mite eee hike oreecten eek eet 486 Inrropuctrion, THE first reference to the fauna of the Abrolhos is contained in the journal of Francis Pelsart, 1629 (1). He states thatin the two or three large islands (the Wallaby Islands) there were large numbers of a “species of cats,” of which he gives a very good description, the first proper account of a species of kangaroo written, though Tam not sure whether it was published till 1899, as it was omitted from the pupular account of his voyage published at the time. He adds that in these two islands they found a number of grey turtle- doves, but no other animals. Capt. Wickham, in command of H.M.S. ‘ Beagle,’ surveyed the islands in April 1840, and the officers under him made considerable collections, which are now in the British Museum —unfortunately, in most cases with nothing to indicate on which islands they were obtained. In his account of the LINN. JOURN.— ZOOLOGY, VOL. XXXIV. X Wy. : onal Muses e) 45% MR. W. B. ALEXANDER ON THE voyage (2) Stokes mentions hair-seals seen at Pelsart Island and Rat Island, the latter obtaining its name from “ the quantity of that vermin with which it was infested.” The Wallaby Islands derived their name likewise from the number of those animals found on them, and Stokes remarks specially that not a single wallaby was found on North Island. On Rat Island they obtained numbers of a lizard, named Silubosawrus stokesii by J. E. Gray in the appendix to the volume. One of these Lieut. Emery brought alive to England. The Pigeon Islands were so named because “ the common Bronze- winged Pigeon” was found there in great numbers. The burrows of the Sooty Petrel or Mutton-bird are mentioned as abundant on Rat Island and the south-west side of West Wallaby Jsland. Stokes remarks that the birds met with on Houtman’s Abrolhos, with the exception of one resembling in shape and colour a small quail (Hemipodius scintillans, Gld.) numerous on North Island, were known and common on the mainland. In 1842 Gould’s collector, John Gilbert, visited the islands, and wrote a vivid account of the nesting-habits of some of thesea-birds. He also seems to have made large collections of the reptiles etc., which are now in the British Museum. In 1889 the group was visited by Mr. A. J. Campbell, who wrote an account of the fauna (8), partly from information given him by Mr. Broad- hurst, whose firm had commenced working the guano, and by Mr. Beddoes, the firm’s manager on the islands. In 1894 Mr. O. Lipfert spent three months at the Abrolhos at the in- vitation of Mr. Broadhurst, collecting for the Western Australian Museum. T have to thank Mr. Lipfert for lending me a manuscript list of the birds that he found nesting. In 1897 Mr. R. Helms paid a short visit to the islands accompanied by Mr. Lipfert, and in 1902 an account of his visit was published (4). He added somewhat to the list of birds given by Campbell. In 1899 the islands were visited by Mr. R. Hall, who published a list of the birds (5), adding a few not recorded by Helms. In 1907 the Abrolhos Islands were visited by a party, amongst whom were Messrs. Milligan, Conigrave, and Gibson. The last wrote an account of the birds met with (6). Specimens obtained by Milligan and Conigrave are in the W. A. Museum. I visited the islands with Prof. Dakin in November 1913, with funds provided by the Percy Sladen Trust, and, where not otherwise mentioned, the observations hereafter recorded were made on that visit*. Mr. J. MeMillan, who accompanied us, gave valuable assistance on this expedition. * [This was the first Percy Sladen Trust expedition to the Abrolhos Islands, the second was made in 1915,—W. J. D.] VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 459 Srecrion 1—The Land Vertebrates. MAMMALIA. Rodentia. Muripa. Eiprmys Fuscires (Waterhouse) (not Gould). Dusky-footed Rat. Two examples of this species, obtained in November 1907, are in the Western Australian Museum. They were caught on the sand-hills on East Wallaby Island, the only locality in which we met with the species. The Abrolhos specimens are decidedly smaller than those in the W.A. Museum from other localities, and their skulls are narrower in proportion to their length. As the teeth are only slightly worn, it is probable that both specimens are young, and these differences may be due to age. When a longer series is available, it may be necessary to create a subspecies for them, Marsupialiaa Macropopip 4. Macropus EUGENIL HOUTMANNI (Gould). Dama Wallaby or Tammer. As already mentioned, this wallaby was first met with by Pelsart in 1629 on ‘two or three of the larger islands.” The types of Macropus houtmanni, Gould, were obtained by the naturalists of the ‘ Beagle’ in 1840 on Hast and West Wallaby Islands, the only islands on which they are found, and are in the British Museum, together with specimens obtained by Gilbert. In 1888 Oldfield Thomas wrote (7): “I have come to the conelusion that it is impossible to admit more than a single western species, notwithstanding the very striking differences that exist between the individuals Jong isolated in the islands of the Houtman’s Abrolhos and those living on the mainland. The differences fade away on the examination of a large series, and specimens from the small islets close to the coast are as a rule more or less inter- mediate.” Since 1888 the general use of trinomials to designate differences of this nature leads me to suppose that Mr. Thomas would now adopt the nomen- elature I am using, especially when it is borne in mind that the small islands to which he refers as inhabited by intermediate forms are not geographically intermediate, but lie off the south coast of Australia at the opposite extremity of the range of the species on the mainland. The animals are very plentiful on the two islands which they inhabit, chiefly amongst the coastal sand-hills and on the portions of the islands where the limestone rock outcrops. These are the only regions where the bushes are large enough to afford them cover during the daytime, which they appear to spend in the shade, only coming out at night. 34* 460 MR. W. B. ALEXANDER ON THE Introduced Mammals. Stokes found great numbers of rats on Rat Island, but it is not clear whether these were Rattus (Epimys) /uscipes or R. rattus or R. norvegicus. If they were from the latter they were presumably the result of one of the numerous shipwrecks which occurred on the islands from 1629 onwards. - Rats are stated also to have occurred on Pelsart Island, but we did not meet with them on either of these islands. Rabbits also were formerly found on Pelsart Island, though there is no evidence as to how they got there. Unfortunately, the domestic cat has been introduced. We saw an individual on Rat Island and the tracks of one on Pelsart Island, and this probably accounts for the disappearance of the rodents, as well as the larger lizards, on these groups. It is to be hoped that they will not reach the Wallaby Group. REPTILIA. Ophidia. Borp#. PYTHON SPILOTES VARIEGATUS (Gray). Carpet-Snake. Plentiful on West Wallaby Island. ‘he largest individual we met with was about 7 ft. in length. i CoLUBRID. Dentgsonta coronata (Schleg.). Crowned Whip-Snake. ‘The British Museum has a specimen obtained by Gilbert. A small grey snake seen by me on Hast Wallaby Island, but unfortunately not captured, was probably of this species. RHYNCHELAPS BERTHOLDI (Jan). Ringed Snake. Mr. Lipfert obtained a specimen of this species on West Wallaby Island in 1895. A large example obtained by the guano-workers was forwarded to Prof. Dakin in 1915, and is now in the W.A. Museum. Lacertilia. GECKONIDS. GyMNoDACTYLUS MItiusi (Bory). Flat-tailed Gecko. Frequent on West Wallaby Island ; also found on Pigeon Island. Herrronora BYNOEI (Gray). The types of this species, in the British Museum, were obtained on Houtman’s Abrolhos. PHYLLODACTYLUS MARMORATUS (Gray). The types, in the British Museum, are from the Abrolhos. PHYLLODACTYLUS OCELLATUS (Gray). The types of P. bil/neatus, Gray, regarded by Boulenger as a synonym of this species, were obtained on the Abrolhos. VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 461 DipLopactTyLts spinicErus, Gray. The types, in the British Museum, are from the Abrolhos. We obtained two specimens on West Wallaby Island. DrpLopactyLus virrarus, Gray. ‘The types of D. ornatus, Gray, regarded by Boulenger as a synonym of this species, were collected on Houtman’s Abrolhos. Prropus vartecatus (Dum. § Bibr.). A specimen in the British Museum is from Houtman’s Abrolhos. PYGOPODID&. DELMA FRASERL, Gray. Fraser's Slow-worm. We met with two indi- viduals of this species on West Wallaby Island. LIALIS BURTONIS, Gray. Burton’s Slow-worm. ‘The British Museum has a specimen from Houtman’s Abrolhos. We met with one example on West - Wallaby Island. AGAMID&. AMPHIBOLURUS BARBATUS (Cuv.).° Jew Lizard. Plentiful on Hast and West Wallaby Islands and North Island, especially amongst the sand-hills. Often observed in the bushes a couple of feet from the ground. SCINCIDA., Eaernta wairet (Lacep.). The British Museum has a specimen from the Abrolhos. EGERNIA KINGI (Gray). Common on West Wallaby Island, chiefly in the region covered by flat slabs of limestone, under which it lurks during the daytime. Also met with on East Wallaby Island and the Pigeon Islands. EGERNIA sTOKESI (Gray). Very plentiful wherever there are loose rocks on the Kast and West Wallaby Islands and Pigeon Islands. Sometimes as many as four or five individuals will be found by turning over a single stone. At the time of Stokes’s visit it was evidently common on Rat Island, and Campbell also met with it there in 1889. I think that it has probably been destroyed on that island by the introduction of cats. TRACHYSAURUS RUGOSUS, Gray. Stump-tailed Lizard. The British ) y I Museum has a specimen from Houtman’s Abrolhos, LyGosoma LesunuRI, Yum. § Bibr. The British Museum has a specimen obtained on the Abrolhos. 462 MR. W. B. ALEXANDER ON THE LyGosOMA RICHARDSONI (Gray). The type, in the British Museum, is from the Abrolhos. LyGOsOMA QUADRILINEATUM (Dum. § Bibr.). The British Museum has a specimen from the Abrolkos. Lyagosoma sp. A small species with an orange head and pink throat was very common in sandy localities on the Wallaby Islands. A specimen which J obtained is in the W.A. Museum, and appears to belong to a new species, but I hope to obtain further specimens. The small lizards of this genus are so active that they are very difficult to obtain, and several other species probably occur. A smal] form which lives among the scrub was seen on Long Island and Sandy Island in the Wallaby Group, as well as on Rat Island and Pelsart Island. A rather larger species was also observed on Rat Island. Lycosoma pRa&PEDITUM, Bingr., is recorded by Campbell. AMPHIBIA. Anura. CYSTIGNATHID &. LYMNODYNASTES DORSALIS (Gray). Specimens from Houtman’s Abrolhos are in the British Museum. “ BUFONIDA. MyopatRACHUS GOULDI (Gray). Specimens from the Abrolhos are in the British Museum. We did not see any Amphibia, but on Pigeon Island and Long Island we heard sounds after dark which sounded like those made by some species of frog. Considerable search with a lantern on Pigeon Island failed to reveal the animals from which these noises emanated. LAND-BIRDS. Turniciformes. TURNICIDS. ORTYGODES VARIUS SCINTILLANS (Gould). Painted Quail. Common on Hast and West Wallaby Islands, one of the Pigeon Islands, and North Island. At the time of our visit in November 1913 the breeding- season seemed to be practically over, as young birds were met with which could fly almost as strongly as the adults. VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 463 The Abrolhos birds were described by Gould (P. Z. 8. 1845, p. 62) as a distinet species, “ Hemipodius scintillans,” described as “ very nearly allied to but much smaller than ZZ. varius.” In the ‘ Birds of Australia’ Gould writes that the species ‘much resembles H. varius, but is little more than half the size of that species ; independently of which the colouring is much lighter, more varied and sparkling, the white margins of the back feathers more numerous and conspicuous, and the markings of the throat and breast of a crescentic instead of an elongated form.” In 1911 Mathews (8) separated the form of Zurnx varia found in South- West Australia as 7. varia stirlingi, writing that it “agrees with 7\ varia scintillans in its darker upper coloration but differs in its much paler under- surface and more white on the ear-coverts.” In 1913 the same writer (9) placed his own 7. varia stirlingt as a synonym of 7. varia scintillans, Gould. As Mr. Mathews informed me that he had not examined any birds from the Abrolhos, I presume he was relying on Gould’s description in taking this action. There are two Abrolhos birds in the Western Australian Museum obtained in 1894, and in my opinion they are quite sufficiently different from birds from the mainland as to need a subspecific name. They are much lighter in colour than all the birds from the mainland, for which I will use Mathews’s name of 7°. varia stirlingi. Now, though Gould’s figure shows scintillans as “ darker” than varius, in the description he says it is “lighter.” In the second place, the white patch on the throat extends further down, and the white markings on the breast are fairly narrow, only widening a little towards the tip instead of being crescentic as in stirlingi. I cannot help thinking that Gould meant to write that they were “of an elongated instead of a crescentic form,” instead of the exact opposite. In addition, both the specimens of scintillans have far more white spots on the cheeks and head than any of the specimens of stirlingi, in none of which do they extend right across the back of the head ; moreover, these spots and the streaks on the breast are white instead of buffish, which I presume is what Gould referred to when he said the colouring was “more varied and sparkling.” Other slight differences noticed are that the black patches on the back are distinctly more broken up by cross-bars in scintillans, the reddish colour on the sides of the breast extends much further down, and, as noted by Gould, “the white margins of the back feathers are more numerous and conspicuous.” In making these comparisons I have had before me the two specimens of Ortygodes varius scintillans from the Abrolhos, seven specimens of O. varius stirling’ from localities from Perth to Denmark, W.A., and one specimen of O. varius varius from Queensland. 464 MR. W. B. ALEXANDER ON THE Columbiformes. CoLUMBID &. COSMOPELIA ELEGANS NEGLECTA, Mathews. Brush Bronze-wing Pigeon. Plentiful on East and West Wallaby Islands and North Island, though, curiously enough, we did not meet with them on the Pigeon Islands, which ‘okes named from their abundance there. nest containing one fresh egg Stol df th bund there. A nest cont 2 fresh eg was found by me on Hast Wallaby Island in November 1913, and Hall captured a young bird that had just left the nest on Pigeon Island in November 1899. ‘Grey Turtle-Doves” were seen by Pelsart on the Wallaby Islands in 1629. Stokes recorded them as the “Common Bronze- winged Pigeon” and Campbell as Phaps chalcoptera. Owing to their rapid flight and their habit of flying just over the bushes and then ducking down again, it is by no means easy to get a good view of them, and perhaps Campbell did not secure a specimen. There are two specimens from the Abrolhos in the W.A. Museum, obtained in 1894, and I cannot discover any differences between them and specimens from the mainland, though they are both distinctly below the average in size. Ralliformes. RALLID &. Hypor@NipIA PHILIPPENSIS (Linn.). Buff-banded Rail. : Campbell saw this species on Rat and Pelsart Islands, and states that it is known to breed on the latter. We did not meet with the species. Birds from this locality would presumably be H. p. mellori, Mathews. PoRZANOIDEA PLUMBEA ROBERTI, Mathews. Spotless Crake. Campbell states that this species occurs on Pelsart Island, “about the mangrove swamp” (there are numerous mangrove swamps on the island). Gibson saw a pair in November 1907, ‘‘ on a rocky islet forming part of Rat I ’ ) $1 Island.” “This record for these birds,” he very justly adds, “is, I think ’ J y ? ) ’ somewhat unique.” Judging from the localities it frequents on the mainland, this is one of the most unlikely birds one could think of to be found on these J dry islands. Falconiformes. FaLcONIDS. CERCHNEIS CENCHROIDEs UNICOLOR (Milligan). Nankeen Kestrel. A bird of this species was seen on West Wallaby Island, and two days later another, or more probably the same individual, on North Island. Doubtless it was only a visitor from the mainland. VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 46 Or Coraciiformes. ALCEDINIDS. SAUROPATIS SANCTA WESTRALASIANA (Campbell). Sacred Kingfisher. Hall found a pair frequenting an abandoned jetty at Pelsart Island, going in and out among the plankiug as if nesting. He shot the female. Passeriformes. HIRUNDINIDZ. HIRUNDO NEOXENA CARTERI, Mathews. Welcome Swallow. Met with on almost every island. The birds appear to nest under the overhanging shelves or low cliffs found round most of the islands. On Pigeon Island a swallow was seen to carry nesting-material into a situation of this kind. MUSCICAPID A. WAHITEORNIS GOODENOVIL RUFICAPILLUS, Mathews. Red-capped Robin. Hall shot a young bird of this species on Pelsart Island, which he considered, doubtless correctly, as a stray visitor from the mainland. TIMELIID &. PreN@DUS MATHEWSI MATHEWS! (/redale). Rufous Song-Lark. Hall met with three birds of this species on Pelsart Island in the man- groves and shot one of them, a young male. He thought they might have nested on the island, but probably they were only visitors from the mainland. SYLVIIDS. SERICORNIS MACULATUS FUSCIPES, subsp. noy. Spotted Scrub-Wren. Common amongst the bushes on Hast and West Wallaby Islands. In the ‘ Birds of Australia’ Gould, after referring to the variability of S. macu- latus, states that specimens from Houtman’s Abrolhos differ from examples from the mainland in their rather smaller size, much greyer tint on the back, and much darker-coloured legs. In the Western Australian Museum are four rather poor specimens from the Abrolhos obtained by Milligan in November 1907. I have compared these with 16 specimens from various localities in 8.W. Australia, including two from Albany, the type-locality of S. maculatus. There are also available four specimens from the islands in Sharks Bay, one of them from Bernier Island, the type-locality of S. balstoni, Grant, and the other three, which agree closely with it, from Dorré Island. None of the specimens agrees with S. mathews? warreni, Mathews, which is said to have a‘ greenish (not greyish) olive back,” though there are examples from localities on both sides of the Warren River. 466 MR. W. B. ALEXANDER ON THE Specimens from the Abrolhos appear to be nearer to S. mathewsi balstont than to typical S. mathewsi maculatus ; they differ from both, however, in their very dark legs, as noticed by Gould. I am therefore using a new name for birds from these islands. They have much less of the rufous tinge on the rump than in S. mathewsi maculatus, and rather less than in S, mathewsi balstont. The dark streaks on the throat are narrower than in maculatus, but rather larger than in balstoni. The size is almost the same as balstoni, distinctly smaller than maculatus. Mr. Lipfert found a nest of this bird on West Wallaby Island on December 2, 1894. ZOSTEROPIDA. ZOSTEROPS GOULDI, Bonap. Green-backed White-eye. Noticed in small flocks or family-parties on almost every island or islet visited. They seem to have a special liking for mangroves. On several occasions they were seen flying from one islet to another. Mr. Lipfert found a nest containing one egg on Rat Island on November 23, 1894. MorTraciLLipa, ANTHUs AUSTRALIS, Vieillot. Australian Pipit. Helms obtained one of these birds on Gun Island. We met with a pair on North Island. No doubt they are only visitors from the mainland, and probably they belong to the subspecies A. a. bilbali, Mathews. SUMMARY. Mammats.—The two indigenous species of Mammals are only found in the Wallaby Group, the Wallaby on both the large islands, the Rat, as far as is known, only on Hast Wallaby Island. The latter is probably, and the former certainly, subspecifically distinct from the mainland species. Snakes.—The three species of snakes do not appear to differ from the mainland forms. At present two have been found only on West Wallaby Island and the third on last Wallaby Island. Lizarps.—Four families are represented by some 19 species. The seven species of Geckonidee are only known from the Wallaby Group, the two species of Pygopodide only from West Wallaby Island, and the single species of Agamide from the Wallaby Islands and North Island. The Scincidx are represented throughout the group, though we only met with the smaller species of Lygosoma outside the Wallaby Group; one of the larger forms, Lgernia stokesi, was, however, formerly plentiful on Rat Island. Froas.—Nothing is known as to the habitat of the two species recorded from the group. VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 467 31rDs.—Of the 12 land-birds recorded from the Abrolhos, four (Kestrel, Robin, Song-Lark, and Pipit) are certainly only casual visitors, five others (Rail, Crake, Kingfisher, Swallow, and White-eye) are probably visitors from the mainland, though all but the Crake and Kingfisher have been known to breed ; the remaining three are almost certainly residents. Of these, the Pigeon and the Quail are found in the Wallaby Group and on North Island, the Serub-Wren only on the Wallaby Islands. The two latter are subspecifically distinct from the mainland forms. Looking now at the different groups of islands we find that West Wallaby Island has two snakes and two slow-worms confined to it, Mast Wallaby Island has one snake and one rat confined to it, whilst the two Wallaby Islands possess subspecies of a wallaby and a bird peculiar to them. The Wallaby Islands, with North Island, possess also one lizard and two birds not found in the other groups, one of the birds being a subspecies peculiar to the group. This distribution strongly suggests that the whole land-fauna of the group has been derived from the Wallaby Islands *. One of the chief features of the weather on the Abrolhos is the prevalence during the summer of “ southerly busters,” extremely strong southerly winds. The presence of the two birds on North Island is thus easily accounted for, and it is noteworthy that the one lizard which habitually climbs bushes is the one which has managed to reach that island. Doubtless at times the bushes on the sand-hills which it frequents are blown into the sea, and occasionally reach North Island. The fact that the smaller skinks are found on many of the small sandy islets seems to show that the sea is no great barrier to their distribution. Probably their eggs are not damaged by floating in sea-water for some time. The larger Spiny-tailed Skink, Lgernia stokesi, is a favourite food of the Sea- Eagles, and it is possible that individuals may be captured by them on the Wallaby Group and carried to Rat Island, and if one occasionally escaped this would account for their presence on that group. A striking feature of the Abrolhos land-fauna is its southern character. Our knowledge of the distribution of animals in Western Australia is perhaps not sutticient to allow us to be dogmatic on the point, but there is every indication that when the islands were peopled from the mainland the fauna of the Geraldton district must have approximated much more than it does now to that of the extreme south-west. The range of Macropus eugenii does not now extend much north of Perth, whilst Lpimys fuscipes is only known from the south coast and the islands of the Recherche Archipelago. Oriygodes varius and Cosmopelia elegans are not recorded from farther north than the Moore River, though it is not improbable that they may occur ; Porzanoidea plumbea is not recorded from north of Perth. * {A discussion on the origin of the fauna of the Abrolhos [slauds will be given in the concluding paper of the series—W. J. Daxin. | 468 MR. W. B. ALEXANDER ON THE The most northerly record that I know for Denisonia coronata and for Egernia whiter, “ Perth,” is less than 30 miles north of Perth. Lygosoma richardsoni is known only from the Abrolhos. The remainder of the species are found farther north on the mainland, but there is not one of them which is not found in the south-west. Srcrion 2.—The Marine Vertebrates. SEA-BIRDS. Sphenisciformes. SeHENISCIDS. KubyPruLa MINOR woopwarbtI, Mathews. Little Penguin. tecorded by Hall without further particulars. I know of no other record of its occurrence narth of Fremantle, and consider the record needs confir- mation. Procellariiformes. HYDROBATID &. PELAGODROMA MARINA DULCIA, Mathews. White-faced Storm-Petrel. Gilbert met with this species on a small island about three miles south of East Wallaby Island. The young birds were almost ready to leave their holes in January. Campbell found them nesting on Beacon Island; the burrows contained young about 10 days old on December 15th. Lipfert found eggs on a sand-patch off Wooded Island in November 1894; and Hall obtained eggs and nestlings on West Wallaby Island and South Island, Pelsart Group, in November 1899. PROCELLARIIDA. PUFFINUS ASSIMILIS TUNNEYI, Mathews. Allied Shearwater, Campbell obtained a specimen of this bird at Rat Island, and Gibson noted them at several islands, principally on Wooded Island, where he obtained uumerous almost full-grown young in the burrows (November 1907). Hall states that the eggs have been found on Pelsart Group in July. We saw them several times about West Wallaby Island. One individual was found sitting on the water in the daytime, others flying about at night, when, like the Mutton-birds, they seemed to be attracted by the lights of the ship. They are not nearly so abundant as the next species. THYELLODROMA PACIFICA CHLORORHYNCHA (Lesson). Wedge-tailed Shear- water or Mutton-bird. All those who have written about the Abrolhos have referred to the great VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 469 numbers of these birds. During the daytime they are to be seen flying over the sea in the neighbourhood of the islands in small or large flocks. We saw them on one occasion in pursuit of a shoal of fish, on which Roseate Terns, Dolphins, and Bonetas were also levying toll. Unlike the Terns, which dive straight down, the Mutton-birds first settle on the water and then plunge under, apparently with the assistance of their wings. When rising from the water the wings are half-spread, but not flapped, and the bird gets up the velocity required to start it in flight by paddling along the surface with its feet. On the islands they appear from their burrows in great numbers as soon as it is dark, uttering the most weird and mournful cries. They are unable to stand on the ground, or to walk, so that they are obliged to use their wings, by whose aid they shuffle about among the bushes in a most awkward manner and are readily captured. They seem to be attracted by a light, as they often flew round the ship in the evening and once-or twice came on board. On many of the islands, especially West Wallaby Island and the southern end of Pelsart Island, their burrows are so numerous that when walking over the areas inhabited by them one sinks in, almost up to the knees, at every step. Their burrows were also seen on Rat Island, Long Island, and one of the Pigeon Islands. Fresh eggs were found in them on Pelsart Island, and Mr. Lipfert obtained them on Gun Island. During one moonlight night I spent some time watching a Mutton-bird excavating a burrow in the sand. It used its feet alternately, throwing out an almost continuous stream of sand behind it to a distance of about a yard. The burrows when completed extend to a distance of two or three feet into the ground at an angle. It strikes one as very remarkable that though the legs of these birds are not strong enough to support the weight of their body, yet they can be used for shovelling away sand continuously, apparently for hours. Moreover, they can use them as paddles on the water sufficiently rapidly to raise them- selves from the surface when about to fly. MACRONECTES GIGANTEUS ALBUS (Potts). Giant Petrel. A specimen in the W.A. Museum was obtained at the Abrolhos in 1894. Campbell states (10), on the authority of Beddoes, that they visit the islands every winter, which is not improbable, as they occur off Fremantle every year during that season. The subspecific name which Mathews uses for the New Zealand and Australian form is singularly unfortunate, as, so far as Tam aware, every specimen obtained in Western Australia has been com- pletely dark in plumage without a single white feather. There are six specimens in the W.A. Museum, and I have seen remains of several others washed up on the beach. 470 MR. W. B. ALEXANDER ON THE DIOMEDEID®. NEALBATRUS CHLORORHYNCHUS CARTERI (Rothschild). Yellow-nosed Mollymawk. A skull obtained on Pelsart Jsland in 1894 is in the W.A. Museun. When on the unfortunate Federal trawling ship ‘ Endeavour,’ in June 1912, T found these birds numerous in the neighbourhood of the Abrolhos. Lariformes. LaRripa, HypROPROGNE TSCHEGRAVA STRENUA (Gould). Caspian Tern. These birds occur in pairs or in small colonies on nearly all the islands, often nesting in company with other species. We found eggs or young birds on West Wallaby Island, one of the Pigeon Islands, Long Island, Wooded Island, and Pelsart Island. In most cases young birds were more numerous than eggs, and some of the young were almost fledged. Mr. Mathews, in the ‘Birds of Australia, states that the Australian form of Caspian Tern never nests in colonies, but on Wooded Island I counted eight nests close together, whilst, judging by the number of old birds, there was a larger colony at one spot on Pelsart Island. Hall reported a colony of some 13 pairs nesting on West Wallaby Island in 1899. THALASSEUS BERGIE (Lichtenstein). Crested Tern. Distributed throughout the islands. Breeding colonies were met with on West Wallaby Island, Wooded Island, and Pelsart Island. The young birds were much more numerous than the eggs, and many of them were almost fully fledged. There are two specimens in the W.A. Museam from the Abrolhos, but I have been unable to decide whether they should be referred to T. bergii pelecanoides (King) or 7. bergit gwendolene, Mathews. They do not appear to differ in size from birds from Barrow Island and Bedout Island, which would be the former according to Mathews, but they are also similar to birds from the neighbourhood of Perth, the type-locality of the latter. The only specimen in the W.A. Museum, which is decidedly larger than any of the others, is one from Hsperance on the south coast. SrERNA DOUGALLI GRAGILIS (Gould). Roseate Tern. These birds were met with in large flocks sitting on the reefs on North Island, East Wallaby Island, Long Island, Rat Island, Wooded Island, and Pelsart Island. We did not find them nesting until the last day of our visit, when we found a considerable colony on the north end of Pelsart Island which had just begun to lay on the piles of coral fragments. Campbell also found them nesting on Pelsart Island. VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 471 “ [STERNULA ALBIFRONS TORMENTI (Mathews). White-shafted Ternlet. Campbell states that he saw a pair of these birds near Rat Island in company with Little and Caspian Terns. He obtained a skin, but appears to be doubtful whether it may not be a young S. nereis. The record needs confirmation, as there is no other record of this species from the southern half of the western coast of Australia. | STERNULA NEREIS HORNI (J/athews). White-faced Ternlet. These little birds were seen on practically every island visited, and from their behaviour it seems probable that isolated pairs, or a few pairs together, nest on nearly all the sandy beaches or heaps of dead coral. We only found nests on the east side of West Wallaby Island, where there was a colony of several hundred individuals. The eggs were fresh or very slightly incubated, one, two, or three being the numbers found in a nest. Campbell, Lipfert, and Gibson all found colonies nesting on Pelsart Island. ONYCHOPRION FUSCATA SERRATA (Wagler). Sooty Tern. There can be no doubt that this is the most plentiful bird found on the Abrolhos at the present time during the nesting season. On Rat Island their numbers are prodigious, they nest under almost every bush and in many places also amongst the herbage ; there is also a very large colony on -the south end of Pelsart Island and another on Wooded Island. In the Wallaby Group they are not found on the Wallaby Islands, nor on North Island, but there are great numbers on all the smaller islands, the Pigeon Islands, Long Island, and Pelican Island. On Rat Island, Wooded Island, and Pelsart Island most of them had eggs at the time of our visit, though a fair number of young birds had already been hatched, but, curiously enough, those in the Wallaby Group had scarcely begun nesting, as only one ege was found—on Long Island. MELANOSTERNA AN.ETHETUS NOVEHOLLANDIA (Stephens). Bridled Tern. We only met with a single pair of this species on a small islet off Rat Island, and did not find a nest. Gilbert and Campbell both found them breeding in small numbers, but do not state on which island or islands. Lipfert and Gibson found a few nests on Pelsart Island. There is no doubt that this is the rarest of the Laride which nest in the Archipelago. t=} ANOUS STOLIDUS GILBERTI, Mathews. Noddy Tern. This species nests in very large numbers on Rat Island and the south end of Pelsart Island, whilst there is a smaller colony on Wooded Island. In each case their colonies occur amongst those of the Sooty Tern, most of the nests being built on the bushes, though in many cases they are flat on the ground ; these latter appear to be those of birds which have failed to obtain 472 MR. W. B. ALEXANDER ON THE a site on the top of the bushes within the limits of the colony, and, rather than utilise bushes only a few yards away from the rest of their species, they are content to take up a position on the ground. At the time of our visit the majority of the nests contained eges, though many young birds had already been hatched. All previous writers on the group have referred to the absurd tameness of the Noddies : they have to be lifted from the nest in order to see what it contains, though this has to be done with caution as they give vicious pecks with their beaks. Many of them undoubtedly fall victims to the cats, and were it not that these latter are kept down by the difficulty of finding food at other seasons of the year, when they appear to feed chiefly on crabs, it would doubtless not be long before the Noddies were exterminated. Ag it is, their numbers are only exceeded by those of the Sooty Terns. MErGALOPTERUS TENUIROSTRIS MELANOPS ((rowld). Lesser Noddy. These birds were discovered on Pelsart Island by Gilbert in 1842, though Stokes had previously observed their curious nests. Gilbert’s remarkably vivid account of their numbers and nesting-habits, first read by Gould at the meeting of the Zoological Society on February 27, 1844, and published in the P.Z.S. for that year, has been quoted by every writer who has since dealt with this subspecies, for the colony on the Abrolhos is still its only known breeding-place. The following quotations give an idea of the number of birds at the time of Gilbert’s visit. He wrote :—‘I have seen many vast flocks of birds, but I must confess [ was not at all prepared for the surprise I experienced in witnessing the amazing clouds (literally speaking) which these birds present when congregating in the evening .... even Audubon, who has been so accustomed to see such vast flocks of the passenger pigeon, could hardly avoid expressing surprise if he had an opportunity of seeing these birds at sunset, moving in one immense mass over and around their roosting-place ; while the noise of the old birds’ quack and the piping whistle of the young ones is almost deafening.” I regret to lave to record that these great flights, like those of the passenger pigeon, are now a thing of the past. Campbell wrote in 1890 :—‘* Now that a successful guano depot has been established upon Pelsart Island, no doubt in time the limited supply of mangrove trees will be used for fuel. What then will become of the extraordinary flights of the Lesser Noddies as they go to and from their fishing grounds? TI trust the photographs I took will not soon be the ‘ light of other days.” Mr. Lipfert tells me that when he visited the Abrolhos five years later, in 1894, the birds were still nesting on Pelsart Island, as they were also at the time of his visit with Helms in 1897, Hall also found them there in 1899. On Gibson’s visit in 1907 they were only found on Wooded Island, so that some time during the intervening eight years the whole colony moved VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 473, from one group to another. It is a great pity that it is not known how this exodus took place, nor for certain what was its cause. The cutting down of the mangroves, suggested by Campbell, has not taken place, as I believe they only furnish very inferior timber for burning. Prof. Dakin tells me that some years ago the soil of the mangrove swamp was dug out from the roots of the trees, being almost pure guano, and probably this disturbed the birds so much that they removed to their new home. Wooded Island, in spite of its name, has not so many mangroves as Pelsart Island, and almost every tree is covered by their nests. Their numbers at the present time, however, cannot be anything like what they were 3) years ago, as they are certainly far fewer than either the Sooty Terns or the Noddies. Moreover, there is every reason to fear that they are still decreasing, for every occupied nest on the trees there are several old ones, and Prof. Dakin tells me that on his last visit, in 1915, he thought there were fewer nesis occupied than when I was with him in 1913. I can suggest no reason for this decline; there are no obvious enemies of these birds on Wooded Island, but it appears that the guano accumulated below their nests on Wooded Island is likely to be worked before long, when presumably they will have to make another move. This could not have been long delayed in any case, as the accumulation of the guano round the roots of the mangroves is rapidly killing the trees, and it is unlikely that the birds would continue to nest on the boughs of the dead trees. It is to be hoped that a cbange to another island may lead to an increase in their numbers, otherwise I fear they must be regarded as the last remnant of a dwindling race. BRUCHIGAVIA NOVA HOLLANDIA LONGIROSTRIS (Masters). Silver Gull. These birds occur throughout the group. Small colonies were found nesiing on West Wallaby Island, an islet off Rat Island, Wooded Island, and Pelsart Island. A few eggs were still not hatched, and young birds were met with in every stage from newly hatched to fully fledged. GABIANUS PACIFICUS GEORGI (Aing). Pacifie Gull. Not so plentiful as the Silver Gull, but distributed over all the islands. No eggs were found, but young birds half-grown or fully fledged were noted on West Wallaby Island, Long Island, Wooded Island, and Pelsart Island. Mr. Lipfert found fresh eggs on an island off Rat Island on October 7, 1894. Charadriiformes. ARENARIIDE. ARENARIA INTERPRES OAHUENSIS (Blowham). Turnstone. A very common summer visitor, found round the shore on all the is'ands, LINN, JOURN.—ZOOLOGY, VOL, XXXIV. BD 474 MR. We. B. ALEXANDER ON THE H ©EMATOPODIDS. HMATOPUS OSTRALEGUS PICATUS (King). Pied Oyster-catcher. Small parties of this species, or in some cases only single pairs, were seen on North Island, East and West Wallaby Islands, the Pigeon Islands, and Pelsart Island. On the latter they were several times met with on the ridges of broken coral in the centre of the island, suggesting that they were nesting there, but no nests were found. ‘The two specimens from the Abrolhos in the W.A. Museum are referable to the northern race picatus. HXMATOPUS NIGER BERNIERI (Mathews). Black Oyster-catcher. These birds were seen on the reefs at North Island, West Wallaby Island, Wooded Island, and Pelsart Island. They were not quite so numerous as the preceding species, with which they often associated. Mr. Lipfert found a nest containing one fresh egg on November 24, 1894. CHARADRIIDS. SQUATAROLA SQUATAROLA HYPOMELAS (Pallas). Grey Plover. A specimen in the W.A. Museum was obtained on the Abrolhos in the summer of 1894. We saw a flock of birds, either of this species or Golden Plovers, on the shore of West Wallaby Island. LEUCOPOLIUS RUFICAPILLUS TORMENT! (Mathews). Red-capped Dotterel. Very abundant on the sandy shores of North Island and Hast and West Wallaby Islands, but not met with in the southern groups, where sandy beaches are unusual. Breeding was apparently over, or nearly so, as young birds were often seen with the adults on the beaches, but on West Wallaby Island pairs of birds were met with on the sandy flats in the interior of the island as if nesting. ScOLOPACIDE. NumeENtus cyanopus, Vieillot. Australian Curlew. Campbell met with this species in the Wallaby Group in the summer of 1889. PHAOPUS PHAZOPUS VARIEGATUS (Scopoli). Whimbrel. Campbell records that he saw a small flock at “the mangrove swamp” on Pelsart Island on December 23, 1889. VETOLA LAPPONICA BAUERI (Vaumann). Barred-rumped Godwit. Lipfert shot a specimen on Gun Island in the summer of 1894, VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 475 HETEROSCELUS INCANUS BREVIPES (Vie/llot). Grey-rumped Sandpiper. A specimen in the W.A. Museum was obtained on the Abrolhos by Lipfert in the summer of 1894. ActITIS HYPOLEUCUS AURITUS (Latham). Common Sandpiper. A specimen in the W.A. Museum was obtained on the Abrolhos by Lipfert in the summer of 1894. (G{LOTTIS NEBULARIUS GLOTTOTDES ( Vigors). Greenshank. A.specimen in the W.A. Museum was obtained on the Abrolhos by Lipfert in the summer of 1894. PISOBIA MINUTA RUFICOLLIS (Pallas). Little Stint. Frequent, especially about the Wallaby Islands. Recorded by Gilbert and Campbell. Four specimens shot by Lipfert in the summer of 1894 are in the W.A. Museum. EROLIA FERRUGINEA CHINENSIS (Gray). Curlew-Sandpiper. Seen on the shores of West Wallaby Island. Recorded by Campbell. A specimen from the Abrolhos is in the W.A. Museum. ANTELIOTRINGA TENUIROSTRIS (Horsjield). Great Knot. A specimen obtained by Lipfert on the Abrolhos in the summer of 1894 is in the W.A. Museum. No doubt this list of wading birds would be enlarged by further collecting, as several other species commonly found in South-West Australia during the summer must visit the Abrolhos at times. Ardeiformes. ARDEID&. DEMIGRETTA SACRA COOKTOWNI, Jdathews. Reef-Heron. Seen on the reefs at North Island, one of the Pigeon Islands, Rat Island, Wooded Island, and Pelsart Island. Two nests were found on Wooded Island, in both cases situated ona ledge of rock behind a bush under an overhanging cliff at the edge of a lagoon. One nest was empty, the other contained two fresh eggs. All the birds were blue, except one of a pair seen flying together at North Island which was white. ‘lo my mind the evidence points to these two forms being an example of dimorphism within the species, and I am therefore using the well-known name of sacra for the species. It was rejected by Mathews as having been applied to an intermediate form which he regards as a hybrid, but whether the two forms are distinct species or not, I think the eyidence is clear that a true Mendelian segregation occurs 30* 476 MR. W. B. ALEXANDER ON THE when they breed together. In that case the origin of the rare intermediate forms may be due to incomplete segregation or they may be simply variations of the white form. Itis a pure assumption that they are of hybrid origin, hence I think the name sacra should be retained. Anseriformes. ANATIDS. CuHEnopis ATRATA (Latham). Black Swan. I am informed by Mr. O. Lipfert that a specimen was shot on Gun Island in 1894. [tis distinctly surprising to find that this bird flies so far out to sea, especially as the islands do not afford lakes of the kind usually frequented by swans. [ ViraGo CASTANEA (Hyton). Green-headed Teal. Campbell states that this species occurs in the Wallaby Islands. In view of the confusion between this species and V. gibberifrons, and the absence of any specimens, the record requires confirmation. | Pelecaniformes, PHALACROCORACIDA., HypoLnucus VARIUS PERTHI, Mathews. Pied Cormorant. These birds are found everywhere round the islands. We only found one colony nesting on West Wallaby Island, where they were in company with Caspian and Crested Terns, and Silver and Pacific Gulls. The nests con- tained fresh eggs, and when the birds flew off at our approach the Silver Gulls seized the opportunity to feed upon the eggs of their more timid neighbours. On an islet off Rat Island, we found the nests of an old colony not being used that year. On a little island at the western side of the lagoon, south of the Wallaby Group, we found a number of eggs lying about among the bushes. Cormorants were sitting on the shore on this islet, but they were evidently not nesting there, as there were no nests and the eggs were on or under the bushes promiscuously. Mr. Lipfert obtained eggs on Middle Island in 1894. PHETHONTID&. ScHOPHMTHON RUBRICAUDA WESTRALIS, Mathews. Red-tailed Tropic Bird. A pair of these birds with their single egg, obtained by Mr. Lipfert on Rat Island in November 1894, are in the W.A. Museum. Subsequently Beddoes wrote to Campbell (10) that he had “ found Tropie Bird nesting on Pelsart Island, month February, two eggs, both hard-set. Following February two nests, same kind, were taken on Rat Island ; two VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 477 eggs in each.” Mathews, in commenting upon this record, remarks that one egg is the usual clutch, and it seems to me probable that the late Mr. Beddoes made some mistake in the matter. [| LEPrOPHETHON LEPTURUS DOROTHFa, Mathews. White-tailed Tropic Bird. Recorded by Campbell as an occasional visitor, but he does not state that he saw the species himself nor on what evidence he relied. As there is no other record of the species on the west coast of Australia it requires sub- stantiation. | PELECANIDS. JATOPTROPELECANUS CONSPICILLATUS (Zemm. § Laug.). Australian Pelican. Small parties of these birds were seen on West Wallaby Island, opposite Pelican Island, as well as on Rat Island. We found no evidence that they were or had been nesting, though I specially visited the small island known as Pelican Island, as its name and the presence of pelicans in the vicinity suggested that it might be their stronghold. Campbell states that they have been known to nest on Pigeon Island, whilst Gibson was told that they nested on West Wallaby Island, appro- priating the nest of the Pied Cormorant in which to lay their egos, Both state that the breeding-season is early (Sept., Oct.). Falconiformes. AQUILIDA. CuNCUMA LEUCOGASTER (Gmelin). White-bellied Sea-Eagle. These birds were met with on North Island and all of the southern groups. Their nests were found on several of the small islets in ithe Wallaby Group, but the young had already flown. Campbell states that they lay in Sep- tember. Judging from the remains met with round the nests, they feed chiefly on the larger lizards and on Mutton-birds and Terns. PANDIONID&. PANDION HALIAETUS CristaTus (ellot). Osprey. Much more numerous than the Sea-Eagle, especially on Pelsart Island. Their nests were found on all the groups, generally on small islets. On Pelsart Island, where they nest on the island itself, three nests were found, in each of which was a fully-Hedged young bird. Though these were apparently able to fly they did not attempt to do so, but either lay flat in the nest or adopted a threatening attitude with their wings raised and feathers spread. ‘The old birds, meanwhile, circled round high in the air uttering shrill crics. 478 Mr. W. B. ALEXANDER ON THE MAMMALIA. Carnivora Pinnipedia. Ovariips. EuMeEropias ALBICOLLIS (Péron). White-necked Hair-Seal. [ Otaria albicollis, Péron, Voyage de Découvertes aux Terres Australes, vol. ii. p. 118 (1816) ; J. W. Clark, P. Z. 8. 1875. Arctocephalus lobatus, Gray, Spicilegia Zool., part i. 1828; Gray, Cat. of Seals and Whales, 1866; Gould, Mammals of Australia, vol. iil. pl. xlix. Otaria australis, Quoy & Gaimard, Voyage de lAstrolabe, Zool. vol. 1. 1830, Arctocephalus australis, Gray, Cat. of Seals and Whales, 1866. Zalophus lobatus, Ogilby, Cat. of Australian Mammals, 1892; Lucas and Le Sonef, Animals of Australia, 1909. ] Hair-Seals were noted on the group by Stokes; and Gilbert secured specimens which were figured by Gould in his ‘Mammals of Australia,’ as well as writing an account of their habits. They were evidently very plentiful on the Abrolhos at that time—at any rate, in the breeding-season. Nowadays only a few individuals are to be seen there, as on most other parts of the West Australian coast. I have given a full synonymy of this species, as, though Gould expressly states that he was not sure of the proper name to use for his Abrolnos speci- mens, he was, nevertheless, followed by Ogilby and by Lucas and Le Souef, the only recent writers, so far as | am aware, who have mentioned the species. I had reached the ecnclusion that Péron’s name should be used before I read the paper by J. W. Clark, and was confirmed in my view by finding that he had reached the same conclusion in 1875. Cetacea. BALANIDA. Mercarrera LonGimana, Rud. Humpback Whale. Portions of the skeleton of a whale of this species were lying about on the shore of Pelsart Island. In Oct. 1914, when travelling down the coast on the ss. ‘Minderoo,’ IT saw a number of Humpbacks in the neighbourhood of the Abrolhos. Prof. Dakin tells me that some came into Whale Bay in the Pelsart Group during his visit in 1915—apparently, as we were told by the fishermen, to scrape themselves on the rocks. DELPHINIDS. [Soratia GaADamu (Owen). Gadamu Dolphin. This appears to be the common ‘ Porpoise” of the West Australian coast, and I presume that the specimens seen at the Abrolhos belong to this species, but there is at present no definite svecimen from the group to determine the point. | VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 479 REPTILIA. Chelonia. CHELONIIDA. CHELONIA MypAs, Zinn. Green Turtle. This species visits North Island and West Wallaby Island, but we were not fortunate enough to see any *. It is curious that no species of Sea-Snake has yet been recorded from the Abrolhos, as several species occur farther south on the mainland. PISCES. A considerable number of small collections of fish have been made at the Abrolhos from time to time. Some coloured drawings were made by Lieut. Emery, one of the officers of H.M.S. ‘ Beagle’ in 1840, and are referred to by Richardson (11, 12), and it is possible that specimens were also obtained. Richardson (12) also refers to specimens collected by Gilbert in 1842, and Ogilby in 1899 described (13) some fish collected at the Abrolhos by Mr. A. M. Lea. Most of the records in the following list, however, are based on specimens in the Western Australian Museum. Of these some were collected by Saville Kent, some were presented by Mr. F. C. Broadhurst, others were collected by Lipfert in 1894 and 1897, and Conigrave in 1907; whilst, finally, the specimens obtained on the two Percy Sladen Trust Expeditions have been added. Unless otherwise stated the reason for inclusion in the following list is the presence of a specimen in the W.A. Museum :— Pleurotremata. GALEORHINIDA. GALEORHINUS ANTARCTICUS (Giinth.). Gummy Shark. Two specimens caught off the north end of Pelsart Island in 1913. HETERODONTIDS. Hereropontus paiprl, Bloch § Schneider. Bull-head or Port-Jackson Shark. Several fine specimens caught at North Island in 1913. Hypotremata. RuINoBATID 2. Ruropatus BANKSI, Miller § Henle. Shovel-nosed Ray. Isospondyli, CLUPEID&. CLUPANODON NEOPILCHARDUS (Steind.). Australian Pilchard. * (They were seen by me on the occasion of the Wxpedition of 1915,—W, J. Daxiy.] 480 Mh. W. B. ALEXANDER ON THE GONORHYNCHID4. GoNoRHYNCHUS GREYI, Rich. Beaked Salmon or Rat-fish. Ostariophysi. ARIID 4. [GALEICHTHYS THALASSINUS, Riipp. Salmon Cat-fish. A poor specimen, probably of this species. | PLOTOSID. CNIDOGLANIS MEGASTOMUS (fich.). Estuary Cobbler. Young specimens obtained in 1915. Apodes. MUR#&NID&. LycopontTIs WoopWARDI (McCulloch). The type of this species was from the Abrolhos (see ‘Records of W.A. Museum,’ vol. i. p. 80), and to it McCulloch assigns the specimen from the Abrolhos referred by Richardson (12) to Murena nubila. A third, very young, specimen was obtained by Prof. Dakin in 1915. Synentognathl. HemrrHampuipa, HypoRHAMPHUS INTERMEDIUS (Cant.). Sea-Garfish. Percesoces). SPHYR#NIDZ. z SPHYRHNA oBTUSATA, Cuv. § Val. Sea-Pike. Berycomorphi. Burrycrp#. ‘TRACHICHTHODES AFFINIS (Giinth.). Nannygai or Red Snapper. Percoideas SERRANIDA. THERAPON HUMERALIS, Ogilby. The type came from Pelsart Island, and the species has not been obtained except in the Abrolhos group. ACANTHISTIUS SERRATUS, Cuv. § Val. Wirrah. EipINEPHELUS MERRA, Bloch. IPINEPHELUS FASCIATUS (forsk.). EXPINEPHELIDES LEAT, Ogilhy. The type was from Pelsart Island. It would appear rather doubtful whether the species is really distinct from E. armatus (Cast.). CoLPOGNATHUS DENTEX, Cuv. § Val. HYPoPLECrRODES NIGRORUBRUM, Cuv. 5 Val. DAMPIERIA LINEATA, Cast. VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 48] PLESIOPIDA. PARAPLESIOPS MELEAGRIS, Peters. A small specimen obtained in 1915. CHILODIPTERIDSE. AMIA RUEPPELLI (Giinth.). Gobble-guts. SILLAGINID2. SILLAGO BASSENSIS, Cuv. § Val. School Whiting. CARANGIDA, CARANX GEORGIANUS, Cuv. § Val. Skipjack. TRACHURUS DECLIVIS, Jenyns. Horse-Mackerel. CENTROPOMIDSE. GLAUCOSOMA HEBRAICUM, Rich. Jew-fish. The type of this species was from Abrolhos. Several were obtained by us. It is the best food-fish found in the group. Lares cALcARIFER, Bloch. Giant Perch. LUTIANIDA. LUTIANUS CHRYSOTENIA, Bleek. NEMIPTERIDS. ScoLopsis BIMACULATUS, Riipp. LETHRINIDS., LETHRINUS OPERCULARIS, Cuv. § Val. Pentapus virta, Cuv. § Val. Butter-fish. Obtained by us at North Island. SPARID&. Pagrosomus AuRATUS (Forst.). Schnapper. The chief food-fish of the group ; it is stated to be much less plentiful than was formerly the case, Sparus SARBA, Forsk. Silver Bream. ScoRPIDIDS. Neatypus osiiquus, Waite. The type was from the Abrolhos (see Records of Austral. Mus. vi. p. 64) ; a second specimen from off Geraldton was presented to the W.A. Museum by Prof. Dakin in 1918. 482 Mk. W. B. ALEXANDER ON THE KYPHOSID. KYPHOSUS SYDNEYANUS, Giinth. Buffalo-Bream. GIRELLID&. TePHR&OPS TEPHRAOPS (/tich.). Buffalo-Bream. ENOPLOSIDA, EnopLosus AkMAtTUS (Shaw). Old-Wife. CHHTODONTIDS. MicrocanrHus stricatus (Cuv. § Val.). Footballer. CHIRONEMIDA. THREPTERIUS MACULOSUS, Rich. Spotted Kelp-fish. CHILODACTYLIDS. Gonustius arpBosus, ich, Magpie-Perch. DacryLopHoRA NiGRICANS, Rich. Dusky Morwong. PoMACENTRID&. HypsiIpops MICROLEPIS, Giinth. LABRIDS. PsEUDOLABRUS PARILUS (2ich.). LEPIDAPLOIS VULPINUS (ich.). AcHmRopDUS GOULDI (Rich.). Blue Groper. We obtained several of these fish ; they are very good eating when fresh. Corts AURICULARIS, Cuv. § Val. Parrot-fish. Probably the commonest fish of the group. Extremely variable in colour, but the dark tip to the operculum is characteristic. Lieut. Emery made drawings of two colour- varieties in 1840. OPHTHALMOLEPIS LINEOLATUS, Cuv. § Val. Rainbow-fish. THALASSOMA LUNARE (Linn.). Crescent-tail. THALASSOMA ANEITENSE, Giinth. These two species of Vhalassoma seem always to occur together, and it seems to me probable that they are the two sexes of the same spccics. ODACIDA. ODAX RICHARDSON, Giinth. Weedie or Rock-Whiting. OLIOTHOPS CYANOMELAS, /tich. Herring-Kale. VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 483 ScARIDA. HETEROSCARUS FILAMENTOSUS, Cast. PSEUDOSCARUS GMYNOGNATHUS, Bleek. Recorded from Pelsart Island by Ogilby. Gobioidea,. GoBiips. CALLOGOBIUS MUCOSUS ((Giinth.). Obtained in 1915. Several other species of Goby have been obtained, but have not been identified. Blennioidea. BLUENNIID &. BLENNIUS TASMANIANUS, ftich. Obtained at Sandy Island in 1913. Other Blennies have been obtained, but not identified. CONGROGADID&. CONGROGADUS SUBDUCENS, Rich. Scorpenoidea, ScorpmNIDS. SYNANCEJA HORRIDA (Linn.). Devil-fish. NEOSEBASTES PANDA (Atich.). This species was founded on a drawing made by Lieut. Emery at the Abrolhos in 1840. It is also in the W.A. Museum collection, ScorPANA SUMPTUOSA, Cast. PLATYCEPHALIDA. PLATYCEPHALUS BASSENSIS, Cuv. § Val. Flathead. TREC LDA. CHELIDONICHTHYS KUMU, Less & Garn. Gurnard. Plectognathi. Bauisripa, MonacaNTHUS CHINENSIS, loch. Recorded from the Abrolhos by Richardson. A specimen was obtained by Prof. Dakin in 1915. MonAcCANTHUS MEGALURUS, Rich. CANTHERINES GRANULATUS (Shaw). CANTHERINES HIPPOCREPIS (Quoy §° (raim.). CHATODERMIS MACCULLOCHI, Waite. The type-specimen came from the Abrolhos (see Records of Austral. Museum, vi. p. 81). It has not since been met with. 484 MR. W. B. ALEXANDER ON THE OSTRACIONTID &. ANOPLOCAPROS LENTICULARIS, Jtch, TETRODONTID&. SPHEROIDES PLEUROGRAMMA, Regan. Blow-fish Pediculati. BaTRACHID2. ConyzIcHTHYs DIEMENSIS (Jtich.). Recorded from the Abrolhos by Richardson. ‘There is a specimen in the W.A. Museum. SUMMARY. The marine fauna of the Abrolhos Islands was stated by Saville Kent (14) to be a remarkable mixture of temperate and tropical forms. Apart from the corals, which are outside the scope of the present paper, he specially instanced the fish, many of which he stated were common to the Abrolhos and the Barrier Reef and Torres Straits. It will be worth while therefore to examine how far his statement is borne out by the distribution of the marine vertebrates recorded in the present paper. Sea-Brrps.—The 36 species of sea-birds which occur on the group may be divided into the following groups. Summer visitors—11: wading birds from the Northern Hemisphere, all of which travel farther down the coast than the Abrolhos. Winter visitors—2: the Yellow-nosed Albatross has been recorded from as far north as Point Cloates, the Giant Petrel has not been met with north of the Abrolhos. ‘The breeding-places of both are unknown, but are doubtless much farther south. Other southern sea-birds probably occur about the Abrolhos in winter, but almost all the collecting done on the group has been carried out in summer. Casual visitor—1: the Black Swan. The remaining 22 species al! breed on the group. Of these 12 are forms found breeding in other localities along the west coast, both north and south of the Abrolhos. Four are not known to breed farther north than the Abrolhos, viz., Pelagodroma marina duleie, Puffinus assimilis tunneyi, Sternula nereis horni, and Gabianus pacificus georgi. Five are not known to breed farther south than the Abrolhos, viz., Sterna dougalli gracilis, Onychoprion fuscata serrata, Anous stolidus gilberti, Hematopus ostralegus picatus, and Scwophaethon rubricauda westralis. ‘he Abrolhos form of the Lesser Noddy, Megalopterus tenuirostris melanops, is not known to breed anywhere else, but as its near ally AZ. tenuirostris tenuirostris breeds in the Seychelles it may be accounted a northern form, Judging from the sea-birds, therefore, we may conclude that there is a distinct admixture of northern and southern forms, and the three most plentiful species—the Svoty Tern, Noddy, and Lesser Noddy—are all forms which are specially characteristic of tropical islands. VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. 485 Marine Mammars.—No conclusions can be drawn from the presence of the Seal, as it is, or was, found all round the western half of the continent. Reprites.—The Green Turtle, plentiful in the north-west, has not been recorded from farther south than.the Abrolhos. Fisu.—The fish-fauna of the north-west coast of Australia is still too little known to allow of a very complete comparison of the northern and southern forms found at the Abrolhos. Sixty-seven species are recorded from the group, of which 5 are not at present known from any other locality, and 4 more do not appear to have been found on the west coast of the mainland. Of the remaining 58, 14 are found generally along the coast-line, 34 do not appear to have been found farther north than the Abrolhos, and 10 are not known from farther south. Of these 10 species, two, Hpinephelus merra and Lethrinus opercularis, have been obtained as far south as New South Wales on the east coast of Australia, so that their most southerly record on the west coast is not remarkable. The remaining 8 northern forms are as follows :—Lates calcarifer (N. Australia to India), Lutianus chrysotenia (N.W. Australia through Malay Archipelago to Nicobar Islands), Scolopsis bimaculatus (N. Australia to China, India, and Red Sea), Thalassoma lunare, and 7. aneitense (N. Hebrides, Norfolk I., and Lord Howe I. to N.W. Australia), Congrogadus subducens (N.W. and N. Australia), Synanceja horrida (N. Australia to India), Coryzichthys diemensis (N.W. and N. Australia). These 8 species certainly appear to range further south on the west coast of Australia than they do on the east, but there is no good evidence that they occur much further south on the Abrolhos than they do on the mainland. From the next area northward of whos> fish anything is known, Shark’s Bay, at least 4 of these species have been obtained, viz., Lutianus chrysotwnia, Thalassoma lunare and 7. aneitense, and Synanceja horrida. It will be worth while to quote here Saville Kent’s summary for contrast with the foregoing facts. He states (14)—‘ The fish fauna of Houtman’s Abrolhos was found, as might be anticipated in virtue of its essentially migratory constituents and its proximity to areas of relatively cool water, an interesting admixture of both tropical and temperate species. Conspicuous among the fishes indigenous to the temperate Australian sea-board may be mentioned such species as the Schnapper (Pagrus major), the Sergeant Baker (Aulopus purpurissatus), Australian Whiting (Sillago ciliata), Yellow-tail (Seriola gigas), and a species of what in the Sydney market would be designated a Morwong (Chilodactylus). Characteristic tropical fish were, on the other hand, specially represented by innumerable varieties of Parrot- fishes, Labride and Scaridee. Many of these, it is interesting to observe, such as species of Julis and Pseudoscarus, had not been met with by the writer farther north on the Western Australian coast, but were familiar to 486 THE VERTEBRATE FAUNA OF HOUTMAN’S ABROLHOS. him, as in the case of the Holothuride, as inhabitants of Torres Straits and the Queensland Great Barrier region. Such species, again, as Platax orbicularis and Mesoprion Johni, the Golden Schnapper of Thursday Island, Torres Straits, may be mentioned among the essentially tropical forms that were found frequenting the Abrolhos reefs.” We may, perhaps, suggest that the “innumerable varieties of Parrot-fishes” were chiefly forms of Corts auricularis, which, as already mentioned, is the commonest fish and is very variable in colour : the record of Platax orbicu- laris may refer to P. teira, which has been found on the west coast as far south as Fremantle, whilst Lutianus chrysotenia may have been mistaken by Saville Kent for L. (Mesoprion) johni. Otherwise, it would be strange if the few species mentioned by him in support of his statement had none of them been obtained by anyone else. To the writer it would seem more remarkable that ‘he fish-Ffauna of the Houtman’s Abrolhos coral-reefs should consist mainly of species characteristic of the southern coast of Australia, than that a few of the tropical species associated with coral-reefs further north should be found there. Literature cited. (1) Hurres. The Part borne by the Dutch in the Discovery of Australia, 1606-1765. Leiden, 1899. (2) Svoxes. Discoveries in Australia by H.M.S. ‘Beagle, 1837-1843. London, 1846. (3) CamppetL. Proc. Austral. Assoc. Adv. Science, 1890, p. 492. (4) Hetus. Journ. Dep. Agric. W.A. vol. v. p. 33. Perth, 1902. (5) Hatu. Ibis, 8th series, vol. ii. 1902, pp. 121-143 and 180-206. (6) Gipson. Emu, vol. viii. 1908. (7) OtprieLD THomas. Cat. Marsupials and Monotremes in Brit. Mus. London, 1888. (8) Maruews. List of the Birds of Australia. Novitates Zool, xviii. 1911. (9) Maraews. List of the Birds of Australia, London, 1913. (10) Camppety. Nests and Eggs of Australian Birds. Sheffield, 1900. (11) Ric#arpson. Contributions to the Ichthyology of Australia: Ann. & Mag. Nat. Hist. ii. (1843). (12) Ricrarpsox. Ichythology of the Voyages of H.M.S. ‘Erebus’ and ‘Terror. London, 1848. (13) Ociupy. Contribution to Australian Ichthyology. Proc. Linn. Soc. N.S. Wales, xxiv. 1899. (14) Savitte Kewr. The Naturalist in Australia. London, 1897. ANNELIDES POLYCHNTES DE L’ARCHIPEL HOUTMAN ABROLHOS. 487 Annélides Polychetes de ? Archipel Houtman Abrolhos (Australie Occidentale) recueillies par M. le Prof. W. J. Dakin, F.L.S.* Par Prerre Fauve, Professeur a l'Université catholique d’Angers. (Communiqué par le Professeur W. J. Daxin, D.Sc., F.L.S8.) (Avec 2 figures dans le texte.) [Lu le 17 mars, 1921.] L’arcuiper diles coralliennes désigné sous le nom de Houtman Abrolhos est situé entre 28° 15’ et 29° de latitude sud. Bien qu’il ne soit éloigné que d’une quarantaine de milles dela céte occidentale de Australie, il semble jouir d’un climat notablement plus chaud que le continent voisin 4 Ja méme latitude. Tout au moins, la température de la mer y est-elle de quelques degrés plus élevée. Le phénoméne semble avoir pour cause une pointe méridionale formée par un courant chaud de ’Océan Indien. Quoi qu’il en soit, les conditions de la vie marine doivent s’y rapprocher singuli¢rement de celles que l’on rencontre dans Ja zone des tropiques, aussi les coraux y prospérent et ce petit archipel est probablement le groupe d’attolls le plus méridional connu. Dans une note préliminaire fort intéressante, M. le Prof. Dakin (1917) a exposé et discuté en détail cette situation exceptionnelle qui l’a engagé a entreprendre l’exploration rationnelle des Houtman Abrolhos, principalement au point de vue faunique. Il était, en effet, intéressant de se rendre compte dans quelles proportions les espéces tropicales et les especes méridionales pouvaient sy trouver mélangées. M. le Prof. Dakin ayant eu Pamabilité de me confier la détermination des Annélides Polychétes recueillies dans cette expédition, je pourrai apporter une modeste contribution a cette étude de zoogéographie. A Vexception dune Avrenicola cristata et de quelques Armandia lanceolata et Hunice Siciliensis, trouvées dans lestomac de poissons péchés dans l’estuaire de Swan River, prés de Perth, toutes les especes étudiées ici proviennent de Varchipel des Abrolhos. * (The Percy Sladen Trust Expeditions to the Abrolhos Islands. ] 488 PROF. P. FAUVEL: ANNELIDES POLYCHETES Huit familles seulement sont représentées par 15 genres et les 21 espéces suivantes, dont 19 des Abrolhos :— APHRODITIENS. | Eunicrens. Lepidonotus oculatus, Baird. (?) Eunice tentaculata, Quatrefages. Lepidonotus cristatus, Grube. Dunice siciliensis, Grube. Lepidonotus carinulatus, Grube. Eunice antennata, Savigny. Lepidonotus acantholepis, Grube. Bunice australis, Quatrefages. Euthalenessa djiboutiensis, Gravier. Lysidice collaris, Grube. AMPHINOMIENS, Chloeia flava (Pallas). OPHESLIENS. Notopygos hispidus, Potts. Armandia lanceolata, Willey. Eurythoé complanata (Pallas). Polyophthalmus pictus, Dujardin. FLESIONIENS, Hesione pantherina, Risso. ARENICOLIENS. Néréwiens. Arenicola cristata, Stimpson. Nereis denhamensis, Augener. Pseudonereis anomala, Gravier. SERPULIENS. Platynereis dumerilii, Aud, & Tidw. Protula bispiralis (Savigny). La premiére remarque qui s’impose est l’absence totale des Polychétes du sable et de la vase (I Arenicola cristata et les Armandia lanceolata ne venant pas des Abrolhos). Les Sédentaires ne sont représentées que par le Polyo- phthalmus pictus et un individu de la rare Protula bispiralis qui semble n’avoir été observée qu’une seule fois, par Ehlers, depuis Savigny. Ce sont les Aphroditiens et les Huniciens qui dominent de beaucoup, avec les Amphinomiens et quelques Néréidiens. Nous avons done 14 une faunule d’Annélides rampant sur les pierres et les Polypiers. La faune des fentes vaseuses des rochers et des cavités des Madrépores n’est guére représentée que par quelques Hunices. Il serait imprudent d’étayer de vastes considérations théoriques sur-un aussi petit nombre d’espéces correspondant seulement & un habitat restreint de la faune des Polychetes. L’examen de cette petite collection peut suggérer néanmoins quelques reflexions. Nous remarquerons d’abord que trois de ces especes—Lepidonotus oculatus, Nereis denhamensis et Eunice tentaculata—n’ont encore été signalées, jus- qwici, qu’en Australie, sur les cdtes sud et sud-ouest. Les 18 autres especes appartiennent a la zone intertropicale. Parmi elles, les 5 suivantes—Lepidonotus cristatus, L. carinulatus, L. acantholepis, Euthalenessa djiboutiensis et Pseudonereis anomala—habitent les régions les plus chaudes de Océan Indien, Mer Rouge, Golfe Persique, Ceylan, par exemple, et n’ont pas encore été signalées en Australie. Les 13 suivantes—Chloeiu flava, Notopygos hispidus, Eurythoé complanata, FHesione pantherina, Platynereis dumerilii, Hunice siciliensis, E. antennata, E. australis, Lysidice collaris, Armandia lanceolata, Polyophthalmus pictus, DE L’ARCHIPEL HOUTMAN ABROLHOS. 489 Arenicola cristata et Protula bispiralis—dépassent la zone tropicale au Nord ou au Sud et ont toutes été déja signalées en Australie et en Nouvelle-Zélande. Parmi elles, Platynereis Dumerilii, Eunice Siciliensis, Polyophthalmus pictus et Arenicola cristata, tout & fait cosmopolites, se rencontrent aussi dans la Méditerranée et dans l’Atlantique Nord. Ein résumé, sur 19 espéces des Abrolhos, nous en avons 5 spéciales aux régions chaudes de ?Océan Indien, 11 de la zone tropicale, mais la dépassant souvent au Sud et habitant aussi l’Australie méridionale, et 3 seulement propres a cette dernicre région. Nous pouvons en conclure, semble-t-il, que la faune des Polychetes des Abrolhos est la méme que celle de la plupart des récifs coralliens de Océan Indien avec, en outre, quelques espéces jusqu’ici spéciales a l Australie. Famille des APHRODITIENS, Savigny. Tribu des PoLYNOINKES, Grube. Genre Lrepiponotus, Leach. LEPIDONOTUS CRISTATUS, (rrube. Lepidonotus cristatus, Gravier (1901), p. 210, pl. 7. fig. 104-110; pl. 9. fig. 186, Fauvel (1919), p. 329. (Synonymie.) ” ” Localité.— Wooded Island. Cette espéce n’est représentée que par la partie antérieure d’un grand spécimen, large de 25 millimetres, soies comprises. Les trois antennes sont sensiblement égales, Ainsi que I’a fait remarquer Gravier, qui a donné de cette espéce une description trés détaillée, les soies ventrales du 2° sétigere different complétement des suivantes, elles sont longues, fines, presque capillaires et ornées de petites membranes laciniées disposées en spirale. Les élytres sont ornées chacune d’une grosse eréte charnue, plus ou moins nettement bilobée, couverte de petites épines chitineuses coniques. Les élytres sont bordées postérieurement d’un ¢troit liséré couleur de rouille, et portent une tache de méme couleur au-dessus de V’élytrophore. Les différentes papilles de Vélytre sont conformes aux figures de Gravier. Distribution géographique—Mer Rouge, Ceylan, Philippines, Iles de la Sonde, Amboine, Abrolhos, Ile Maurice, Zanzibar. Leriponorus ocuLatus, Baird. Lepidonotus oculatus, Baird (1865), p. 184. Haswell (1883), p. 281. " ; Fauvel (1917), p. 171, pl. 4. fig. 20-23, (?) Thormora argus var., Haswell (non Quatrefages) (1883), p. 280, pl. 8. fig. 9-11, ” ) Localités—Dragué entre Wallaby et Rat Island. Dragué A l’extérieur de Varchipel des Wallabys. LINN, JOURN,— ZOOLOGY, VOL. XXXIV. 36 490 PROF. P. FAUVEL: ANNELIDES POLYCHETES Le premier specimen mesure 37 mill. sur 15 mill., soies comprises ; le second atteint 47 mill. sur 15 et il est décoloré 4 l'exception d’une petite tache brun pale sur ’élytrophore des élytres postérieures. Ges spécimens sont bien semblables 4 ceux des Golfes Saint-Vincent et Spencer, sur la céte Sud de l’Australie. Distribution géographique.—Australie méridionale, Houtman Abrolhos. LEPIDONOTUS CARINULATUS, Grube. Lepidonotus carinulatus, Potts (1909), p.83 Fauvel (1911), p. 367, fig. 1. (Synonymie.) ” ” . 5 Horst (1917), p. 69, pl. 15. fig. 10. . . Fauvel (1919), p. 330. Localité.—Dragué entre les Iles Wallaby et Rat Island. Il n’a été recueilli qu’un seul spécimen de petite taille dont les élytres sont détachées. Il est tout & fait semblable & ceux de Madagascar que j’ai eu occasion d’étudier. La caducité de ses élytres, et ses soies ventrales hidentées, distinguent facilement cette espéce du Lepidonotus squamatus auquel elle ressemble beaucoup par ailleurs. : Distribution géographique.—Mer Rouge, Golfe Persique, Ceylan, Philip- pines, Amboine, Japon, Madagascar, Abrolhos. LEPIDONOTUS ACANTHOLEPIS, Grube. Tepidonotus acantholepis, Grube (1878), p. 24, pl. 2. fig. 1. 5 - . Michaelsen (1892), p. 5 ss Horst (1917), p. 67, pl. 15. fig. 3-4. Localité.—Récif de Long Island, Pelsart Group. Cette rare et curieuse espéce est représentée par un seul individu one de 30 mill. et large de 12, soies comprises. Jusqu’ici, elle semble n’avoir été recueillie qu’aux Philippines par Grube, a Ceylan par Michaelsen et aux Iles de la Sonde par Vexpédition du ‘ Siboga.’ Michaelsen a fait remarquer que les soies ventrales ne sont pas bifurquées, comme Vindiquait Grube, mais, en réalité, trifurquées. Horst a donné une bonne figure de ces soies, qui, vues de profil, rappellent des soies ventrales d’ Hermione, quand les deux dents inférieures se profilent Pune sur l’autre. Ces deux dents ressemblent aux deux moignons émoussés des soies du premier sétigére des Serpules (Fig. 1, d). Les soies dorsales, implantées au nombre de 2a 3 seulement sur an petit mamelon parapodial, sont minces, finement épineuses et terminées en pointe lisse, mousse ou légérement renflée. Le prostomium, du type Lepidonotus, est en grande partie caché par le bord antérieur du premier segment, qui forme un repli nucal tres accusé. insertion des antennes est terminale. L’impaire est portée par un DE L’ARCHIPEL HOUTMAN ABROLHOS, 491 volumineux cératophore cylindrique, brun foncé. Elle est notablement plus longue que les antennes latérales. Les cirres tentaculaires sont accompagnés d’un petit faisceau de soies fines. Tous ces appendices sont annelés de brun et présentent un renflement ovoide au-dessous de leur extrémité distale effilée. Les élytres de la premicre paire sont plus grandes que les suivantes, elles se croisent complétement et cachent la téte. Celles de la deuxiime paire se touchent seulement par le bord et celles de la 3° paire laissent déja une partie du dos & nu. Ces trois paires sont imbriquées d’avant en arriére, tandis que les suivantes, beaucoup plus petites, laissent entre elles un espace plus grand que leur diamétre, qui ne dépasse pas l’épaisseur du parapode. Les premiéres Fia. 1. (iB © ©@® ® “o G@ 6 @ 00 5% 98 L6 OM, ; 2 @9@° Lepidonotus acantholepis—a. Elytre, X 25. 6. Granulations superficielles d'un tubercule, x 800. ce. Papille filiforme du bord de l’élytre, x 140. d. Soie ventrale, x 100, élytres sont done plus grandes et les suivantes plus petites que sur le type de Grube. Ces élytres sont ovales, ou presque rondes (Fig. 1,a). Leur surface est couverte de gros tubercules ovoides, 4 surface finement granuleuse et a contenu brun foneé (Fig. 1,/). On remarque, en outre, au bord des élytres, quelques molles papilles cylindriques (Fig. 1, ¢). L’anus, dorsal, s’ouvre sur l’avant dernier sétigére. Le pygidium est bilobé avec. 2 longs urites. Cetie espéce présente des affinités avec Y Hermenia verruculosa, Grube, qui est un Lepidonotus. Distribution géographique.—lles de la Sonde, Philippines, Ceylan, Australie occidentale. 36" 492 PROF. P. FAUVEL: ANNELIDES POLYCHETES Tribu des SIGALIONINAES, Grube. Genre EutHaLenessa, Darbouw. EUTHALENESSA DJIBOUTIENSIS, G'ravier. Thalenessa djiboutiensis, Gravier (1901), p. 281, pl. 7. fig. 114-117. Euthalenessa djiboutiensis, Fauvel (1918), p. 331; (1919), p. 345, Localité.—A brolhos. L’unique spécimen est tronqué postérieurement et completement incolore. Il correspond bien a la description détaillée de Gravier et ne différe pas des spécimens du Golfe Persique que j’ai eu l’occasion d’étudier récemment. Cette espéce se rapproche de la Thalenessa oculata,M‘Intosh, du Détroit de Bass. Distribution geographique. — Mer Rouge, Golfe Persique, Houtman Abrolhos. Famille des AMPHINOMIENS, Savigny. Genre EurytHoii, Kinberg. EurytHoi COMPLANATA (Pallas). Eurythoé complanata, Augener (1913), p. 87. (Synonymie.) Bs 6) Fauvel (1919), p. 348. » aleyonia, Gravier (1901), p. 248, fig. 257- 258, Z 5, levisetis, Fauvel (1914), p. 116, pl. 8. fig. 29-30, 33-37. Localités.—Récif de Long Island, et ala céte, Pelsart Group. Wooded Island. Les quelques individus de cette espéce, si répandue dans tous les récifs coralliens, ne présentent rien de particulier a signaler. Distribution géographique.—Atlantique, Antilles, Guyane, Canaries, Golfe de Guinée, Océan Indien, Mer Rouge, Madagascar, Philippines, Pacifique, Australie, Chili, Iles Gambier. Genre CHLOEIA, Savigny. CHLOEIA FLAVA (Pallas). Chloeia flava, Fauvel (1917), p. 190. (Synonymie.) Localité.—King’s Sound, Cote Nord-Ouest. Cette magnifique espéce, dont Milne-Edwards a donné une belle figure en couleurs dans ‘Le Régne Animal Illustré,’ sous le nom de Ch. capillata, a les soies dorsales d’un jaune bronzé a reflets un peu verdatres, tandis que les ventrales sont d’un blane albitre tirant sur le gris. A la face dorsale, chaque segment est orné d’une grosse tache arrondie, pourpre foncé, DE L’ARCHIPEL HOUTMAN ABROLHOS. 493, Sur ce spécimen, les denticulations de soies en harpon ont complétement disparu, rongées par la formol. MNistribution géographique.— Océan Pacifique, Australie meéridionale, Houtman Abrolhos, Océan Indien, Japon. Genre Notopyaos, Grube. Noropyaos Hisprpus, Polts, var. sbkRATUS, Fauvel. Notopygos hispidus, Potts (1909), p. 359, pl. 45. fig. 6-7; pl. 46. fig. 3-5. mi F Bae (1911), p. 248. 4 Fauvel (1917), p. 192; (1919), p. 350. (?) Nota ds labiatus, Benham (1915), p. 205. Localité—Abrolhos. Les deux individus de cette espéce, un petit et un autre de taille moyenne, ont les soies dorsales et ventrales également lisses, & Pexception de celles des premiers sétigéres qui sont légcrement dentelées aux deue rames. Aux derniers sétigéres, les soies ventrales sont aussi légerement dentelées et les dorsales le sont indistinctement. Ils rentrent done dans la variété serratus. Reste a savoir quelle valeur on doit attribuer 4 ce caractére et si cette espece nest pas une simple yariété du JV. labiatus, M‘Intosh. La 1° branchie apparait au 6° sétigére et l’anus s’ouvre au 21°. Distribution géographique.—Mer Rouge, Océan Indien, Australie méri- donale (Houtman Abrolhos). ‘ Famille des HESIONIENS, Grube. Genre HEsiong, Savigny. HESIONE PANTHERINA, Risso. Hesione pantherina, Saint Joseph (1898), p. 829. (Bibliographie.) . a Gravier (1900), p. 179, pl. 9. fig. 16. y Fr Fauvel (1911), p. 874, fig. 4; (1919), p. 370. Hesione Ehlers’, Gravier (1900), p. 175, pl. 9. fig. 14-15. Hesione splendida, Augener (1913), p. 187. (Synonymie.) Localité.—Dragués a lextérieur de ?Archipel Wallaby. Sauf quils sont enti¢rement décolorés, les deux spécimens des Wallabys ne différent en rien de ceux de la Méditerranée. Mais, ainsi que je l’ai déja fait remarquer, cette décoloration, dont Gravier faisait un trait distinctif de VP Hesione Ehlersi, n’a rien de caractéristique. Distribution géographique.-—Atlantique, Méditerranée, Mer Rouge, Golfe Persique, Ceylan, Philippines, Malaisie, Japon, Australie, Afrique méri- dionale. 494 PROF. P. FAUVEL: ANNELIDES POLYCHETES i 1 Famille des NEREIDIENS, Quatrefages. Genre NEREIS, Cuvier. NEREIS DENHAMENSIS, Augener. Nereis denhamensis, Augener (19138), p. 156, pl. 3. fig. 51. Fauvel (1917), p. 204, pl. 6. fig, 45-46. ” ” Localité.—Pelsart Group, a Vintérieur du lagon. Bien que trés macéré, ce spécimen est bien reconnaissable a la disposition de ses paragnathes et a ses soies en serpe homogomphe a la rame dorsale des parapodes postérieurs. La réduction des groupes vii.-viii. & un seul rang de gros paragnathes, avec quelques petits intercalés, est & peu prés le seul caractére distinguant cette espece, qui semble spéciale a l’Australie, de notre Nereis pelagica. Distribution géographique.—Australie, cdtes sud, sud-ouest et nord-ouest (Houtman Abrolhos). Genre PsEuDONEREIS, St. Joseph (non Kinberg). PSEUDONEREIS ANOMALA, Gravier. Pseudonereis anomala, Gravier (1901), p. 191, pl. 11. fig. 50-52. Willey (1904), p. 262. Fauvel (1911), p. 895; (1919), p. 421. ” ” ” ” Localité.—A mer basse, Pelsart Group. Ce spécimen est malheureusement macéré. La trompe porte des rangées serrées, pectinées, de petits paragnathes aux groupes de Vanneau maxillaire. Les groupes vil.—viii. forment un rang de paragnathes aplatis alternativement dans le sens transversal et dans le sens longitudinal. Les groupes vi. sont formés d’une rangée de paragnathes coniques, tendant a se dédoubler en. deux lignes transversales. Ils sont done, sous ce rapport, intermédiaires entre la Ps. anomala typique et la P. masalacensis, Grube, qui est une espece trés voisine. Distribution géographique.—Mer Rouge, Golfe Persique, Mer d’Oman, Houtman Abrolhos. Genre PLATYNEREIS, Kinberg. PuaTYNEREIS DuMERILII, Audouin et M.-Edwards. Platynereis Dumerilit, Fauvel (1911), p. 397; (1914), p. 1938; (1919), p. 421, (Synonymie.) es 3 Izuka (1912), p. 158, pl. 17. fig. 7-8. insolita, Grayier (1901), p. 197. pl. 12. fig. 53. ”? Localité.— Houtman Abrolhos. Aucune indication de loealité particulitre n’accompagnait les deux petits spécimens de cette espece cosmopolite, si répandue. DE L’ARCHIPEL HOUTMAN ABROLHOS. 495 Distribution géographique-—Mers Arctiques, Manche, Atlantique, Médi- terranée, Océan Indien, Mer Rouge, Golfe Persique, Madagascar, Mer du Japon, Pacifique, [les Gambier, Australie (Houtman Abrolhos). Famille des EUNICIENS, Grube. Genre Eunicr, Cuvier. (?) HUNICE TENTACULATA, Quatrefages. Eunice tentaculata, Fauvel (1917), p. 209, fig. 17. (Synonymie.) Eunice pyenobranchiata, M‘Intosh (1885), p. 294. Eunice Elsy?, Baird (1870), p. 344. Localité.—Abrolhos. Un fragment postérieur d’une grande Hunice aplatie dont les branchies persistent jusqu’aux avant-derniers sétigéres me parait appartenir a cette espece, trés répandue en Australie. Les acicules et soies aciculaires sont noirs, ces derniéres sont bidentées. Distribution géographique.—Australie, Nouvelle-Zélande (Ceylan ?). EUNICE ANTENNATA (Savigny). Eunice antennata, Crossland (1904), p. 312, pl. 22. fig. 1-7. A oi Fauvel (1917), p. 225. (Synonymie.) OE Fauvel (1919), p. 877. Tipnie bassensis, Benham (1915), p. 219, pl. 41. fig. 67-74. Localite.—W ooded Island. Sur un individu entier, la 1** branchie se montre au 5° sétigére, comme c'est d’ailleurs le cas le plus fréquent. Les antennes sont moniliformes. Les soies aciculaires sont jaunes et tridentées. Dans la région postérieure, les branchies redeviennent plus longues et plus fournies que dans la région moyenne. Ovest d’ailleurs un des traits caractéristiques de cette espece, bien qwil ne soit pas toujours aussi marqué. Distribution géographique—Mer Rouge, Golfe Persique, Ceylan, Mada- gasear, Zanzibar, Moluques, Philippines, Détroit de Torres, Détroit de Bass, Australie méridionale, Hes Gambier. EUNICE AUSTRALIS, Quatrefages. Eunice australis, Fauvel (1917), p. 228. (Bibliographie.) » Murrayi, M‘Intosh (1885), p. 288 » leuconuchalis, Benham (1900), p. 21 (fide Ehlers). Localité.—Abrolhos. Cette espéce, représentée par un seul individu, ressemble beaucoup a VE. antennata, mais elle s’en distingue principalement par ses branchies a nombreux filaments qui disparaissent brusquement vers le 38°-50° sétigére. 496 PROF. P. FAUVEL: ANNELIDES POLYCHETES Le soies aciculaires sont jaunes et tridentées. Elles sont souvent au nombre de 3 ou de 4 dans les parapodes postérieurs. Distribution géographique——Nouvelle-Zélande, Australie occidentale et meéridionale, Zanzibar, Ceylan, Maldives, Cap de Bonne-Espérance. IUUNICE SICILIENSIS, Grube. Eunice siciliensis, Fauvel (1917), p. 251 et (1919) p. 879. (Bibliographie.) » vahda, Gravier (1900), p. 264, pl. 12. fig. 80-82. » leucodon, Khlers (1901), p. 128, pl. 16. tig. 1-10, Localités — Wooded Island, Uagoon flat; Pelsart Group, a lintérieur du Lagon ; Freshwater Bay, 9 Mai, 1914. L’ Eunice siciliensis est trés répandue dans tous les récifs des mers inter- tropicales, tout autour du globe. Elle est représentée des Abrolhos par de nombreux fragments souvent dépourvus de téte, mais bien reconnaissables a leurs longues branchies simples et & leurs parapodes a un seul acicule et dépourvus de soies pectinées et de soies aciculaires. Distribution géographique—Meéditerranée, Atlantique, Antilles, Golfe de Guinée, Détroit de Magellan, Océan Pacifique, Hes Gambier, Hawai, Australie, Océan Indien, Ceylan, Philippines, Mer Rouge, Golfe Persique. Genre Lysip1cE, Savigny. LysIDICE COLLARIS, Grube. Lysidice collaris, Fauvel (1917), p. 286. (Synonymie.) » fallax, Khlers (1898), p. 15. Localités.—Pelsart Group, sur le rivage et a Vintérieur du lagon. Cette espece, trés voisine de la Lysidice ninetta, dont elle n’est probablement qu’une simple variété, ne s’en distingue que par la forme de ses yeux qui sont réniformes ou semi-lunaires, au lieu d’étre ovales. Distribution géographique-—Mer Rouge, Golfe Persique, Ceylan, Philip- pines, Japon, Seychelles, Madagascar, Zanzibar, Australie méridionale et occidentale, Samoa, Iles Gambier, Guyane. Famille des OPHELIENS, Grube. Genre ARMANDIA, Filippi. ARMANDIA LANCEOLATA, Willey. Armandia lanceolata, Willey (1905), p. 228, pl. 5. fig. 120. 5 5 Fauvel (1917), p. 259; (1919) p. 435. Localité—F reshwater Bay (cdte occidentale d’ Australie). 3 Six spécimens, comptant de 27 & 30 sétigcres. Beaucoup de branchies DE L’ARCHIPEL HOUTMAN ABROLHOS. 497 sont tombées et les yeux latéraux ont en grande partie disparu. Le tube anal est caractéristique. Distribution géographique.—Golte Persique, Ceylan, Australie. Genre PoLYOPHTHALMUS, Quatrefages PoLyoPHTHALMUS PICTUS, Dujardin, Polyophthalmus pictus, Fauvel (1919), p. 487. (Synonymie.) Loealité.——Abrolhos. Un seul spécimen, encore bien pigmenté, représente cette espece si cosmopolite et de coloration si variable. Distribution géoyraphique. —Atlantique, Méditerranée, Mer Rouge, Mada- gascar, Ceylan, Philippines, Chine, Australie occidentale, Iles Gambier. i . Famille des ARENICOLIENS, Awdouin et M.- Edwards. Genre ARENICOLA, Lamarck. ARENICOLA CRISTATA, Stimpson. Arenicola cristata, Ashworth (1912), p. 105, pl. 5. fig. 12, 13; pl. 8. fig. 17; pl. 10. fig. 80; pl. 13. fig. 41, 42. (Bibliographie et Synonymie.) Arenicola antillensis, Ehlers (1892), passim. Localité.—Canning River, trouvée flottant avec le courant. 18 Feévrier, 1914 (Australie occidentale). Cet unique représentant de cette belle espeéce est d’assez grande taille, mais en trés mauvais état. La partie moyenne du corps est tellement macérée qu il est difficile de compter les segments. II semble bien exister 17 sétigeres, les branchies, qui commencent au 7°, sont du nombre de I1 paires. Elles sont trés développées, pennées et réunies & la base par une forte membrane. Le prostomium, malgré son mauyais état, est encore bien reconnaissable et conforme & la figure d’Ashworth (1912, p. 106, fig. 45). Les otocystes renferment chacun un gros otolithe sphérique. Les deux sacs diaphragmatiques sont gros et allongés. Les deux ceecums digestifs sont courts, épais, renflés et terminés en pointe mucronée. Les premiers segments de la région caudale ne portent pas les petits processus digitiformes qui gion spéciaux a la variété américaine. Ashworth a déja signalé cette espéce au nord-ouest de |’ Australie (Barrow Island). Distribution géographique.— Atlantique Nord, cdtes de PAmérique, Médi- terranée, Océan Indien, Mer Rouge, Australie occidentale, Pacifique, Japon, Californie. 498 PROF. P. FAUVEL: ANNISLIDES POLYCHETES Famille des SERPULIENS, Burmeister. Genre Protuta, Risso. PROTULA BISPIRALIS (Savigny). Serpula bispiralis, Savigny (1820), p. 75. Protula bispiralis, Quatrefages (1865), tome il. p. 467. 53 y Ehlers (1907), p. 31. Lovalité.—Abrolhos. La Protula bispiralis a d’abord été décrite par Savigny d’aprés un individu provenant de la mer des Indes. De Quatrefages en a complété la description Waprés le méme spécimen. Depuis, elle n’a été mentionnée que par Ehlers Fie, 2. Protula bispiralis.—a, Plaque onciale abdominale, x 300. 0. Dent inférieure vue de face, x 800. ¢. Soie abdominale, x 140. d. Longue soie capillaire abdominale, x 140. e, Soie thoracique courte, aplatie, x 140. f. Extrémité d’une longue soie thoracique unilimbée, x 140. qui en a étudié un exemplaire de Nouvelle-Zélande communiqué par Benham. Ehlers ne pense pas que cette espéce puisse étre identifiée & ?Urtica marina americana de Seba. Le spécimen des Abrolhos n’est malheureusement pas accompagné de son tube, qu’aucun auteur n’a encore décrit, &ma connaissance. Il mesure 60 mill. DE L’ARCHIPEL HOUTMAN ABROLHOS. 499 de longueur, y compris le panache branchial de 12 mill., et la région thora- cique est large de 10 mill., pieds compris. I] répond tout & fait a la deseription de Savigny et de Quatrefages, sauf que les deux volumineux lobes branchiaux ne font que 6 tours de spire, au lieu de 8-9 ; il faut Vailleurs remarquer qu’il est de taille un peu moindre, le type mesurant 80 mill. La collerette et la membrane thoracique sont malheureusement déchirées. Le nombre des sétigéres tloraciques est de 7. Pas plus qu’Ehlers, je n’y observe d’uncini. Les soics thoraciques sont trés nombreuses et de deux sortes: les unes Jongues, minces, & extrémité trés fine lége¢rement courbée, n’ont qwun limbe étroit d’un seul cdté (Fig. 2,7); les autres, plus courtes, sont aplaties a V’extrémité et terminées en pointe plus obtuse, arquée (Fig. 2, e). Ces soies aplaties sont homologues aux soies d’ Apomatus de certaines Protules. Le soies de la région abdominale-sont faiblement géniculées, aplaties et finement découpées, sur leur bord conyexe, en petites dents coupées carrément (Fig. 2,¢). D’aprés de Saint-Joseph, cet aspect serait di & un plissement du bord mince de la soie. A Vextrémité postérieure, il s’y joint de longues et fines soies capillaires (Fig. 2, d). Les uncini abdominaux sont des uncini typiques de Protula & nombreuses et fines dents inclinées, & grosse dent inférieure légerement renflée A Pextrémité qui, vue de face, est légerement bilobée (Fig. 2, a, 5). Toutes ces soies sont identiques a celles de la Protula intestinum, dont la P. bispiralis ne semble se distinguer que par le grand développement de ses lobes branchiaux a tours de spire beaucoup plus nombreux. Distribution géographique.—Nouyelle-Zélande, mer des Indes, Abrolhos. INDEX BipnioGRAPHIQUE. 1912. AsHwortH (J. H.).—Catalogue of the Chetopoda in the British Museum. Part I. Avenicolidze. (London.) 1918-1914. AuGEneEr (H.).—Die Fauna Siidwest Australiens. Bd. iv. Polycheta Errantia, 1918, and Bd. v. Polycheta Sedentaria, 1914. (G. Fischer, Jéna.) 1865. Bairp (W.).—Contributions towards a Monograph of the Species of Annelids belonging to the Aphroditacea. Jowrn. Proc. Linn. Soc. London, vol. viii. 1870. Barrp (W.).—Contribution towards a Monograph of the Species of Annelids belonging to the Amphinomacea. Journ. Linn. Suc. London, Zool., vol. x. 1990. Brnuam (W. B.).—Report on Wixperime:tal Trawling. (New Zealand.) 1915-1916. Benram (W. B.).—Report on the Polychieta obtained by the I’.1.8. ‘ Endeavour’ on the Coasts of New South Wales, Victoria, Tasmania, and South Australia. Pt. 1.,1915; Pt. IL., 1916. (Sydney: W. E. Smith.) 1903-1904. Crossnanp (C.).—On the Marine Fauna of Zanzibar and British East Africa from Collections made by Cyril Crossland in the Years 1901-1902. Parts I, IL, III. Proc. Zool. Soc. London, vols. i.-ii. 1908; vol. i. 1904. 1917. Dakin (W. J.).—The Percy Sladen Trust Ixpeditions to the Abrolhos Islands (Indian Ocean). Report I. Jowrn. Linn, Soe, London, Zool., vol. xxxiy. 1892. Wuurrs (I.).—Die Gehdrorgane der Arenicolen. Zettschr. fiir Wiss., Zool. Ld. liii., Suppl. 500 1898. 1901. 1907. 1911. 1914. 1914. 1917. 1918. 1919. 1919. ANNELIDES POLYCHETES DE L’ARCHIPEL HOUTMAN ABROLHOS. Euurrs (E.)—Ueber Palolo (Zunice viridis). Nachr. der K. Ges. der Wiss. zu Gottingen, Math.-Phys. Kl. Enters (E.).—Die Polycheten der Magellanischen und Chilenischen Strandes. (Berlin.) Euters (I.).—Neuseelindische Anneliden. JI. Nachr. der K. Ges. Wiss. xu Gottingen, N.F., Ba. v. Facvet (P.).—Annélides Polychétes du Golfe Persique. Arch. de Zool. Expér. et Gén. 5e série, vol. vi. (Paris.) Fauve. (P.).—Annélides Polychétes de San Thomé, Golfe de Guinée. Arch, de Zool. Expér. vol. liv. fase. 5. (Paris.) Favuvet (P.).—Annélides Polychétes non pélagiques, provenant des Campagnes de I‘ Hirondelle’ et de la ‘Princesse-Alice,’ 1885-1910. és. Camp. Scient. du Prince de Monaco, fase. xlviii.. (Monaco.) Fauve. (P.).—Annélides Polychétes de l’Australie Méridionale. Arch. de Zool. Expér. vol. lvi. fase. 3. ( Paris.) Favuven (P.).—Annélides Polychétes des Cétes d’Arabie. Bull. Mus. Hist. Nat. Paris, No. 5, 1918. Fauve (P.).—Annélides Polychétes des Iles Gambier. Bull. Mus. Hist. Nat. Faris, No. 5, 1919. ~ Fauvet (P.).—Annélides Polychétes de Madagascar, de Djibouti et du Golfe Persique. Arch. de Zool. Exvpér. vol. lviii. fase. 8. (Paris.) 1900-1908. Gravire (Cu.).—Contribution 4 étude des Annélides Polychétes de la Mer 1878. 1888, 1911. Rouge. Nouvelles Archives du Muséum du Paris, 4° série, t. ii. fase. 2, 1900; t. iii. fase. 2, 1901; t. vili., 1906; t. x., 1908. GruseE (Ep.),—Annulata Semperiana. Beitrige zur Kenntniss der Anneliden Fauna der Philippinen. Mém. de l'Acad. Imp. des Sc. de Saint-Pétersbourg, vol. xxv. No. 8. = Hasweu (W. A.).—A Monograph of the Australian Aphrodite. Proc. Linn. Soc. New South Wales, vol. vii. Horst (R.).—On the Genus Notopygos, with some new Species from the Malay Archipelago. Notes from the Leyden Museum, vol. xxxiii. 1912-1917. Horsr (R.).—Polycheta Errantia of the ‘Siboga* Expedition.—Part I. 1912. 1885. 1892. Amphinomidee (‘ Stboga’ Expeditie, Monographie xxiv. 1912), Part II. Aphro- ditidee and Chrysopetalide. did. xxiv b. (Leyden.) - Izuxa (A.).—The Enrrantiate Polycheta of Japan. Journ. Coll. Se. Tokyo, vol. xxx. no. 2. M‘Inrosu (W. C.).—Report on the Annelida Polycheta collected by H.M.S. ‘Challenger.’ ‘ Challenger’ Reports, Zoology, vol. xii. (London.) MicHarLsen (W.).—Polycheten von Ceylan. Jahr, der Hamb. Wiss. Anst. Bd. ix. (Hambourg.) 1909-1910. Porrs (F. A.).—Polycheeta of the Indian Ocean. Trans. Linn. Soc. London, 1865. 1898. 1820. 1905. vol. xii. et vol. xiii. QuaTREFAGES (A. DE).—Histoire Naturelle des Annelés marins et d’eau douce, 8 vols. (Paris: Roret.) Sarnt-JOSEPH (DE).—Annélides Polychétes des Cotes de France (Mancbe et Océan). Ann. Se. Nat. Zool, 8° série, vol. v. (Paris.) Savieny (J. C.).—Systéme des Annélides. Description de Egypte, Hist. Nat. vol. xxi. (Paris.) Wittry (A.).—Report on the Polychaeta. Ceylon Pearl-Oyster Fisheries, Suppl. Report, xxx. (London.) ON A NEW GENUS OF FORAMINIFERA. 501 Sherbornina: A new Genus of the Foraminifera from Table Cape, Tasmania. By Freperick CHapman, A.L.S., F.R.M.S., Paleontologist to the National Museum, Melbourne ; Hon. Pal. Geol. Sury. Vict. (PLatE 32.) {Read 17th March, 1921.] History or THE Specimens.—Ilarly in 1912 I received from the late Mr. R. N. Atkinson *, of Sulphur Creek, N.W. Tasmania, some foraminifera obtained from the Tertiary beds at Table Cape, Wynyard, Tasmania. Recognising their novel character, I put them aside for further investigation, and their description has been still further delayed owing to their requiring to be sliced, a work which had to be postponed on account of the lack of specimens. Mr. Atkinson’s father, Mr. E. D. Atkinson, has lately obtained additional specimens, and to him I express my indebtedness. GeEoLocicAL Horizon or THE Foraminirera.—The tests of this type, of which a fair number have been found, have all occurred in the lower zone of the Table Cape fossil deposits—the Crassatellites Bed. The presence of Crassatellites oblonga, T. Woods, sp., and the various volutes determine the age of this bed as Janjukian, which is equivalent to the Middle Tertiary stage of the southern Australian Tertiaries and to the Miocene elsewhere. It lies below the Turritella Bed containing Turritella warburtonensis, Tate, and the interesting marsupial Wynyardia bassiana, Spencer t. That upper bed may. be regarded as of Upper Miocene age, since it appears to represent the upper beds at Torquay. DESCRIPTION :— Fam. ROTALITD/. Subfam. ROTALIINE, Genus SHERBORNINA f, gen. noy. Generic Characters.—Test discoidal, moderately thin ; median arch con- eave. Shell built up of a median annular series of chamberlets with a discorbine commencement ; the loculi of the annuli widely spaced. External * Mr, R. N. Atkinson, after whom I have named the species here described, was a keen and observant collector of the Tertiary fossils of Table Cape. He unfortunately met with an untimely death in 1915, whilst cleaning a military rifle, + Proc. Zool. Soc. Lond. 1902, pp. 776-794, pls. xlix. & 1. t{ Named in honour of my friend Charles Davies Sherborn, Hon.F.Z.S., A.L.S., who has contributed so much to the literature of this group. 502 MR. F. CHAPMAN ON SHERBORNINA: A NEW GENUS layer formed of small overlapping spatulate chamberlets. The primordial series of about 7 globular to reniform segments, lying in the median system, is discorbine—that is, depressed rotaline.. Shell-wall perforated with coarse tubuli. SHERBORNINA ATKINSONI, gen. et sp. nov. (Plate 82.) Description of Species.—Test discoidal, thin, complanate. Median area depressed. Texture subtranslucent. Surface faintly showing the annular series of the median layer with the crenulations of the superimposed spatulate chambered layer on either side. A thin median horizontal section shows the commencing series as a discorbine shell of about 7 chambers, followed by 4 imperfectly annulate series, the last two almost completely ring-like and embracing the earlier series. Primordial chamber spherical, measuring 58 in diameter ; chambers 2, 3, and 4 subglobular, 5-9 sublunate; 10 and 11 subannulate and of equal width throughout *. Chamberlets of superficial layers spatulate and squamosely arranged. Diameter of test in holotype 1:9 mm. Thickness of test on periphery *25 mm. Thickness in median area ‘09 mm. Diameter of another specimen 2°3 mm. Relationships.—This very remarkable form represents for the Rotaliidee the cycloclypeine annulate multichambered character combined with the lepidocycline spatulate-chambered type found in the Nummulinide. But instead of the lepidocycline character being found in the median layer, it is here, strangely enough, found evenineccd on the cycloclypeine_ median series, to form a complex unrecognized in the Orbitoidine. This structure of both median and external layers is merely 7somorphous with the Cyclo- elypeus and Lepidocyclina groups, as: proved by the commencing series, which is distinctly rotaline, and especially discorbine. The true relationship of this generic type lies with the planorbuline group, of which a remarkable and allied generic type, Cyeloloculina, has been described by Messrs. Heron-Allen and Earlandt. The present specimens show the same stages of development for the median series of chambers—the discorbine, the pavonine, and the annulate. As in Cycloloculina, the tubuli in Sherbornina are fairly coarse perforations, and the “deposit of shell- substance between the tubuli” noted by Heron-Allen and Earland in Cycloloculina appear to have developed in the present form into small scale- like chamberlets forming the external layers, besides which there is also the crenulate and warty appearance of the surface seen also in Cycloloculina. As in Cycloloculina, Sherbornina has no distinct apertural orifice beyond the openings on the face and periphery leading through the coarse tubuli to the chamberiets. * In this character Sherbornina atkinsoni resernbles Cycloloculina annulata rather than C. polygyra of Heron-Allen and Harland. See postea and Journ. Roy. Micr, Soc. 1908, pp. 536 and 5388. + Journ. Roy. Micr. Soe. 1908, pp. 529-548, pl. xii, Chapman. JOURN. LINN. Soc., ZOOL. Sard BLN \ — Lat Tr Leb Soros cart cor OE IP EOC F.C. phot. et del. ad nat. SHERBORNINA ATKINSONI, F, Chapm. x VoL. XXXIV. PL. 32. 34 OF FORAMINIFERA FROM TABLE CAPE, TASMANIA, 503 Age-relations of Cycloloculina and Sherbornina.—Messrs. Heron-Allen and Earland obtained their specimens of Cycloloculina from amongst derived fossils from Tertiary strata at Selsey. The majority of the fossils found there are, however, of Middle Hocene age, so that the greater chances point to that rock-series as their source of origin. The present Miocene Sherbornina seems to be a distinct modification of Cycloloculina through the accession of secondary shell-growth and the development of external chamberlets super- imposed on the median series. Occurrence.-—Moderately common. In the Crassatellites Bed of Wynyardi, Table Cape, Tasmania. Age.—Miocene or Janjukian. EXPLANATION OF PLATE 32. Fie. 1. Sherbornina atkinsoni, gen. et sp. nov. Holotype. Miocene (Janjukian). ‘Table Cape, Tasmania, External surface of test showing annulate plan of growth over- lain with tegulous layer of secondary chamberlets. 56 Pale Fig. 2. S. atkinsont. Horizontal section through median plane, showing the discorbine stage passing into the annulate series. The widely-spaced partitions of the annu- late series of chamberlets are seen near the centre and also towards the periphery. The tegzulate or overlapping chamberlets of the secondary or external layer are seen halfway to the centre of the test. x 24. Vig. 3. 8. atkinsoni. Vertical median section through test, showing the relation of the median to the external series of chamberlets. x 34. S. atkinsoni. Structural sketch drawn from micro-section. ., loculi of median series connected by coarse tubules ; ez., loculi of external series. x 68. = 2 ee Fig. 5. S. atkinsoni, Structural sketch drawn from micro-section. m., loculi of median series; ev., loculi of external series ; s., septa of median series: ¢., shell-wall between median and outer chamberlets, showing coarse perforations. x 72. Aine: Witenes h it oe Says Ee RE A NEW TYPE OF TELEOSTEAN CARTILAGINOUS PECTORAL GIRDLE, 505 On a new Type of Teleostean Cartilaginous Pectoral Girdle found in young Clupeids. By E.S8. Goopricn, F.R.S., Sec.L.8., Prof. of Comparative Embryology in the University of Oxford. (With 6 Text-figures.) [Read 18th November, 1920. ] WHILE examining some thionin preparations of the cartilaginous skeleton of a young sprat, Clupea sprattus, caught at Plymouth *, I found that the pectoral girdle presented a quite unexpected structure, which seems to have hitherto escaped the notice of anatomists. On investigating the young of Clupea harengus and C. pilchardus the same peculiar form of girdle was found, and it may be considered as typical of the genus Clupea, and possibly may be found to occur in related genera. In all living Teleostei the endoskeletal girdle, preformed in cartilage in the young, is in a reduced condition relatively to that of the large dermal bones which support it, and consists of right and left halves usually clearly separated in the middle line. Hach half is firmly fixed to the cleithrum, and is formed of a short dorsal scapular region and of a generally much larger ventral coracoid region. These are separately ossified, and in many lower Teleostei, including the Clupeiformes, there is another ossification in a dorsal mesocoracoid arch, forming an inner buttress to strengthen the girdle near the articulation of the fin. Although in many Teleosts the coracoids may be quite large, and may extend towards the middle line and meet, ventrally, yet, so far as I know, the left and right halves are always said to remain separate (3). Neither in the well-known monographs of Parker (5) and Gegenbaur (2), nor in the works of Stannius (6) and other authors, is there any mention of a median fusion of either cartilage or bone (the only possible exception I have met is that of two little ventral epicoracoid cartilages, widely separated from the coracoids, described by Parker as partially fused (5)). On the other hand, it isa familiar fact that in the Elasmobranchii and Dipnoi the two halves of the cartilaginous pectoral girdle, originally separate in the embryo, usually become firmly fused below the pericardium (3). In the young Sprat, Herring, and Pilchard about 20 to 30 mm. in length, the endoskeletal pectoral girdle is in the form of a solid transverse bar of * T am indebted to Miss Lebour for all the material on which this study is founded, and have to thank her for placing it at my disposal. The specimens were preserved in formol. LINN, JOURN.—ZOOLOGY, VOL. XXXIV, 37 506 PROF. E. S. GOODRICH ON A-NEW TYPE OF cartilage extending ventrally from side to side below the pericardium (figs. 1, 2,3, & 5). At this stage the whole girdle consists of two slender dermal cleithra passing dorso-ventrally to meet below, and the cartilaginous Ite. 1. ~ pfn Clupea sprattus. Cartilaginous skeleton of the pectoral fins and girdle. Clupea sprattus, 28 mm. e. Fig. 2, transverse section showing the pectoral girdle below the pericardium, Fig. 3, similar section further forward. TELEOSTEAN CARTILAGINOUS PECTORAL GIRDLE, 507 ventral bar expanding on either side into a triangular plate perforated by three foramina (fig. 4). These plates and the ventral bar represent the coracoid regions fused. On each side is a small latero-dorsal seapular process projecting in front of the fin-base ; while a longer postero-ventral process extends backwards in the body-wall. At a little later stage, shown Iie. 4. Clupea harengus, 30 mm. Pectoral girdle and fins with muscles attached. Fie. 5. Clupea pilchardus. Cartilaginous skeleton of the pectoral fins and ‘girdle. in fig. 6, the cross-bar becomes bent in the middle to form an acute angle pointing forwards, and at the bend the cartilage becomes absorbed so that the right and left halves soon are separated—a condition which, so far as I have been able to make out, persists in the adult. The development of the pectoral girdle in Teleosts has been: studied by Swirski (8) and Wiedersheim (10), and has lately been very completely Bf! 5308 PROF. E. S. GOODRICH ON A NEW TYPE OF worked out by Swinnerton in Salmo and Grasterosteus (7), by Derjugin in Ewocetus (1), and by Haller and by Vogel in Salmo (4 & 9). These authors have ascertained that the endoskeletal girdle and fin-skeleton arise from paired rudiments widely separated. From each rudiment develops a procartilaginous plate in which become differentiated the fin radials and the girdle. The coracoid process of the latter grows downwards and forwards, and may even meet its fellow of the opposite side, without, however, fusing with it. A posterior process grows out behind, and subsequently dwindles and disappears in the adult. In Clupea,I find that the girdle arises in just the same way from paired rudiments, and that its subsequent history differs only in the formation of the transverse bar by fusion across the mid-ventral line, As to what significance should be attached to the peculiar structure of the Fra. 6. Clupea sprattus, 36 mm. Cartilaginous skeleton of the pectoral fins and girdle. cartilaginous girdle in Clupeids described above, it would be rash to come to any definite conclusion before the corresponding stages in the development of allied forms have been studied. Two points may, however, be noticed :— (1) that the condition when the two halves form a continuous bar is transitory, the girdle being of paired origin and the halves separating again in later life when the dermal bones become well-developed ; (2) that in no other Teleostomes, however primitive, do we find the paired elements of the girdle fused in the middle line. Therefore we can hardly suppose that the cross-bar represents an ancestral condition ; rather would we suggest that it may be a specialisation due to the necessity of affording a firm base of support for the large pectoral fins in young Clupeids, which are active swimmers, TELEOSTEAN CARTILAGINOUS PECTORAL GIRDLE, 509 Summary :—In the young of Clupea sprattus, C. harengus, and C. pilchardus about 20 to 30 mm. in length, the right and left coracoid regions fuse’ to a solid cartilaginous ventral bar, which becomes bent and again subdivided in later stages. This fusion is probably a specialisation to strengthen the support of the pectoral fins before the complete development of the dermal bones of the pectoral girdle. REFERENCE-LETTERS FOR Figures 1-6. at, atrium; ba, bulbus arteriosus; 5, branchial arch; c, coracoid region; cf, coracoid foramen: e/, cleithrum; 2m, hypoglossal muscles; @, cesophagus; p, pericardium ; pf, posteoracoid foramen; pfm, pectoral fin muscles; pfn, pectoral fin skeleton ; pr, posterior process; 7, radial; rd, distal vadial cartilage; s, scapular process; sf, scapular foramen; ve, ventricle; vm, ventral body-wall muscle. References.” 1. Dersuern, K.—Die Entw. der Brustflosse und des Schultergiirtels bei Evwocetus volitans. Zeit. £. Wiss. Zool. xci., 1908. 2. Graenpaur, CU.—Unters. zur vere]. Anat. der Wirbeltiere, ii., 1865. 3. Goopricu, EH. 8.—Vertebrata Craniata: Part 9 of a ‘Treatise on Zoology,’ 1909. 4. Hatter, B.—Ueber den Schultergiirtel der Teleosteer. Arch. f. Mikr. Anat. Ixvi., 1905. 5. Parker, W. K.—A Monograph of the Shoulder-girdle and Sternum., Ray Soe., 1868. 6. Srannrus, H.—Handb. d. Anat. d. Wirbelthiere, 2te Aufl., 1854-6. 7. Swryvertron, H, H.— A Contribution to the Morphology and Development of the Pectoral Skeleton of Teleosteans. Quart. Journ. Micros. Sci. xlix., 1906. 8. Swirskr, G.—Unters. u. d. Entw. des Schultergiirtels des Hechts. Inaug.-Diss., Dorpat, 1880. 9. Vocen, R.—Entw. des Schultergiirtels der Forelle. Jen. Zeitschr. f. Naturw., xlv., 1909. 10. WireprersHEIM, R.—Das Gliedmassenskelett der Wirbeltiere. 1892. Oey ; ae peer ais ve Pa ck ran ‘ ie 4 me ae air bay Put an ef Neirs paint 6 , AA ons eet | feat ¢ ates Tariig At hen meER res seh yee [Synonyms and native names are printed in 7falies. A star * denotes the first publication of a name. | Abrolhos Islands, Perey Sladen Trust Ex- peditions to, by W. J. Dakin, 127-180; general description, 135-180; Verte- brate Fauna, by W. B, Alexander, 457-486, Abundance of Marine animals (Spolia Runiana IV.), by W. A. Herdman, 247-259. Acanthistius serratus, Cuv. § Val., 480. Acartia clausi, Giéesbr., distrib., 98, 110, 112, 119, 124. Achverodus egouldi (Rich.), 149, 174, 482. Achamarchis dentata, Lamx., 5. Acrocecrops, Henrich § De Giryse, 44%. Acrocephalus schcenobeenus, Nevwtom, mis- takes artificial for real insect, 58. Acroloxus lacustris (Zinn.), 210. Actinia equina (Zinn.), abundance, 254. Actinopora, d Orb., mentd., 37. — regularis, VOrb., 38. Actitis hypoleucus auritus (Latham), 475. Adeona Ifeckeli, Reuss, 4. — sp., 24. Adeonella, Busk, mentd., 16. — contorta, Waters, 8, 5, 28, 24. — insidiosa, Jull., 4. — lichenoides, Bush, 24. — platalea, Busk, 24. — polymorpha, Bush, 24. — polystomella, Reuss, 24. Adeonellopsis distoma, Bush, 24. — Crosslandi, Waters, 24. geon pennatus, Sp. Bate, 346. Aetea recta, Hincks, 8, 5. Ageniaspis, Thoms., mentd., 332. Agriolimax agrestis (Zinn.), 206. Albatross, Yellow-nosed, 484, Alexander, W. B., Vertebrate Fauna of the Abrolhos Islands, 457-486. Algee, occurr.in the Abrolhos Islands, 171; —, green, mentd., ftnote 321; —, red, mentd., 307 ; —, mentd., 219. Alkalinity of the sea in its relation to phytoplankton, 124. Alpheus aeuletpes, Cout., 379. —alcyone, De Man, 350, 379. — audouinii, Cout., 351, 381. —bouvieri, A. M.-£d., var. bululensis, Cout., 851, 381. — bucephaloides, Nobili, 845, 350, 879. — bucephalus, Cowt., 380. — consobrinus, De Man, 345, 350, 380. — djeddensis, Cout., 346. — djiboutensis, Cowt., 346. — gracilis, Heller, 350, 378. — — var. alluaudi, Cout., 378. — — var. luciparensis, De Man, 378. — insignis, Heller, 351, 380. — lottini, Guér., mentd., ftnote 378. — macrodactylus, Ortm., 346. — malleodigitus (Sp. Bate), 346. — microstylus (Sp. B.), 350, 379. — pachychirus, Stimps., 351, 580. — paracrinitus, Wiens, 351, 380. — — var. bengalensis, Cout., 381. — parvirostris, Dana, 351, 381. — spp., 161, 345, 350, 351, 379. — strenuus, Dana, 351, 381. \ 7 512 Alpheus ventrosus, H. M.-Ed., 350, 378. Amathia, Rathke, mentd., 20. — connexa, Busk (sp. dub.), 40. — distans, Busk, 40, 41. — lendigera, Zamu., 40. — obliqua, MacG., 40. — tortuosa, 7’. Woods, 3, 89-41. — Vidovici, Waters, 3, 40, 41. — Wilsoni, Kirkp., 41. Amia, Zinn., mentd., 186. — ruppelli (Giinth.), 481. Amphibolurus barbatus (Cuv.), 461. Anabrus, Pack., 484. Anatidee, mentd., 52. Anatis ocellata (Linn.), 444, 445. Anchistus inermis, Miers, 345, 346, 351, 391. — miersii (De Man), 351, 391. Ancylus fluviatilis, Miil/., 204, 210. Andricus, Hart, mentd., 329-332. Anemones, oceurr. in Abrolhos Islands, 161. Anemonia sulcata, Penn., 257. Annélides Polychétes de l’Archipel Hout- man Abrolhos, par P. Fauvel, 487-500. Anodonta anatina (Linn.), 215. — cygnea (Linn.), 215. Anomalocera, Templ., sp., distrib., 115. Anoplocapros lenticularis, Rich., 484. - Anous stolidus gilberti, Mathews, 471, 484. Anser, Briss., sp., mistakes artificial for real insect, 58. Ant, oecurr. on the Abrolhos Islands, 160. Anteliotringa tenuirostris (Horsfield), 475. Anthea cereus, 1. § Sol., abundance, 257. Anthus australis, Viedlot, 466. — — bilbali, Mathews, 466. — pratensis, Bechst., mentd., 52; mistakes artificial for real insect, 58. Anurida, Lab., 486. — maritima, Guérin, 433. Apanteles, F6v'st., mentd., 283, 331. Aplysia, Zinn., sp., 170-171 ; ftnote 259, Aplysilla, #. £. Sehulze, 322. Apsendesia, Lamw., 37. — cristata, Lame., 37, 38 Apterygota maxillule in the, 430-433. Aretecephalus australis, Gray, 478. — lobatus, Gray, 478. Arenaria interpres cahuensis (Blovham), 473. mentd., INDEX. Arenicola antillensis, Whlers, 497. — cristata, St/mps., 487-489, 497. — marina (Linn.), abundance, 255, Avianta arbustorum (Zinn.j, 207. Arion ater (Zinn.), 206. — empiricorum, Fér., 206. Armandia lanceolata, Willey, 487, 488, 496. Arthropoma Cecilii, Lev., 3, 20. Ascaltis cerebrum, Haeck., 311, ftnote 512 316, 317; mentd., 301. — reticulum, Haeck., 316, 317; mentd., 299. Ascetta primordialis, aeck., ftnote 307. Ascidians, occurr. in Abrolhos Islands, 161 : mentd., ftnote 259. Ascopodaria discreta, Busk, 42. Ascosia, Jullien, 408. Asellus aquaticus, de Fuure., reprod., 335- 343. Astacillinea, Collinge *, new group of Val- vifera, 71. Astrea, Gmel., sp., 158. Astralium, Zink, 161. Astrophyton, Link, sp., 162. Athanas crosslandi, Tuttersall *, 345, 350, 372. — dimorphus, Ortmann, 350, 371, 373. — djiboutensis, Cout., 350, 368. — haswelli, Cowt., 373. — minikoensis, Cout., 373. — naifaroensis, Cowt., mentd., 373. — nitescens, Aud. & Sav., 846, 375. — orientalis, Pearson, 373. — parvus, De Man, 345, 350, 372. — polymorphus, Kemp, 345, 369, 370, 378. — sp., 345, 350. Aulopus purpurissatus, Rich., 485. Auricularia stagnorum, Klein, 211. Axiopsis ethiopica, Mobili, 351, 894. 3 Baétis binoculatus, Leach, food of Trout, 50, 57. — pumilus, Burm., food of Trout, 50. —vernus, Curtis, sub-imago, the food of Trout, 48. Balanus balanoides (Linn.), 250-252, 255. Balea fragilis, Drap., 210. — perversa (Lrnn.), 210. Barentsia discreta, Avrkp., 3, 42. —misakiensis, Asajiro Oka, 42. abundance, INDEX. Rat mistakes artificial for real insect, 58. Batopora, Reuss, 407, 408, 421; mentd., 401. — conica, Seguenza, 424, — conica, Hantken, 424. — multiradiata, Ress, 404. Beania elongata, Hinchs, 8. — hirtissima, Heller, 3, 8. — — var. robusta, Hincks, 8. — magellanica, Bush, 4, 8. — spinigera, MacG., 8. Bibio, Geoffr., 451, 453. — johannis, Mezy., 452; food of Swallow, 52; imago, the food of Trout, 49. — pomone (Fudr.), 451, 452. Bicupularia, Reuss, 408. Bidder, G. P., Fragrance of Calcinean sponges and the spermatozoa of Guancha and Sycon, 299-3804; Syn- crypta spongiarum, Bidder *, 305-13; Notes on the physiology of sponges, 315-326. Biddulphia, Gray, distrib., 99, 100. — mobiliensis, Bail., distrib., 99, 101, 124. — regia, O. Schultze, distrib., 101. — sinensis, Grev., distrib., 98, 101, 124. Bipora, Whitelegge, 408. — cancellata, Whitelegge, 419. — (?) elegans, Whitelegge, 421. — philippinensis, Whitelegve, 420. Birds, feeding-habits, 47-60. Biselenaria, Gregory, 409. Bithynia tentaculata (Zinn.), 215. Blatta, Linn., 455. Blennius tasmanianus, Pich., 483. Bonellia, Desh., sp., 175. Boneta, mentd., 469. Bowerbankia, Marre, 41. — gracillima, AMincks, 42. — imbricata, Adams, 41. — pustulosa, Hincks, var. Waters*, 3, 41. Brachycentrus subnubilus, Curtzs, food of fishes, 48, 51; food of birds, 62. Bream, Buffalo, 149. British Guiana, new genus of Isopoda from (W. E. Collinge), 61-63. Bruchigavia noveehollandize (Master's), 473. Buccinum acicula, Miill., 209. Bugula dentata, Busk, 3, 5. alternata, longirostris 513 Bulimus acutus, Miill., 206. Bulla fontinalis, Linn., 213. — hypnorum, Linn., 204, 213. Caberea Boryii, dud., 6. — bursifera, Ort., 6, — Darwinii, Busk, 6. — HEllisii, Flem., 6. — glabra, MacG.,, 6. — grandis, Mincks, 6. — Hookeri, Busk, 6. — lata, Busk, 6. — ligata, Calv., 4. — minima, Bush, 6. — rostrata, Bush, 6. Cacospongia, Schmidt, 313. Calanus finmarchicus (Gunner), distrib., 98, 110-117, 122, 124, 125. — sp., food-value, 259. Calcinean sponges, their fragrance( Bidder), 299-304. Callogohius mucosus (Giinth.), 483. Campodeee lubbocki (Silvestr?), 431. Campylus linearis (Zinn.), 444. Cancer, Linn., sp., mentd., ftnote 259. Cantherines granulatus (Shaw), 483. — hippocrepis (Quoy § Gaim.), 485. Cape Crawfish, early larval stages, 189. Cape Verde Islands, marine fauna, by A. W. Waters, 1-45. Caprimulgus europeus, Linn., mistakes artificial for real insect, 58. Carabus violaceus, Linn., 444. Caranx georgianus, Cuv. § Val., 481. — platessa, Cuv. § Val., 149. Carteriospongia, Hyatt, 310, 318, 322. Carychium minimum, Miill., 204. Catenaria Lafontii, Aad., 405. Catoptropelecanus conspicillatus (Temm. § Laug.), 477. Cats introduced to keep down rats in the Abrolhos Islands, 158, 457; mentd., 460, 461, 472. Cellaria, Ellis, 405, 417 ; mentd., 17. — biseriata, Mapl., 4. Cellepora, Gmel., 16, 17, 407. — americana, Osburn, 19. — avicularis, Hincks, 19. — globularis, Reuss, ftnote 402. — intricata, Calv., 4. j14 Cellepora serspinosa, Waters, 22. — spongites, Pall., 17. Celleporelia hyalina, Norm., 20. Centropages, A7éy., sp., distrib., 120. Cerchneis cenchroides unicolor (Milligan), 464, “Cercids,” proposed new term for “minute wandering cells,” 315. Cheetoceras, Kiitz., distrib., 99, 100. — boreale, Baii., distrib., 101, 104, 124. — contortum, Schiitt, distrib., 104, 124. — criophilum, Castr., distrib., 104, 124. — debile, Cleve, distrib., 101, 104, 106, 124. — decipiens, Cleve, distrib., 104, 106, 124. — densum, Cleve, distrib., 101, 104, 124. — diadema, Cleve, distrib., 104, 124. — sociale, Laud., distrib., 100, 101, 104, 124, — spp., distrib., 98, 100, 101; food-value, 259. — subtile, Cleve, distrib., 101. -— teres, Cleve, distrib., 101, 104, 106, 124. Cheetoceros, Ehrenb.=Cheetoceras, Kiitz. Cheetodermis maccullochi, Waite, 483. Chapman, F., Skerbornina*, a new genus of Foraminifera from Table Cape, Tasmania, 501-503. Chaunosia hirtissima, Busk, 8. Chelidon urbica, Bote, its food, 51. Chelidonichthys kumu, Less. § Garn., 488. Chelonia mydas, Lenn, 479. Chenopis atrata (Latham), 476. Chilodactylus, Lac., 485. Chiridotea, Harger, 69-72. — ceca, Harger, 73. — sabini, Stebbing, 74. — tuftsiil, Harger, 74. Chiriscus, Richardson, 70, 71. — australis, Richardson, maxillipedes, 69. Chironomus, Woll., 451-453. — sp., 452. Chitin, new method of making micro- scopical preparations of, 67-68. Chloeia flava (Pall.), 488, 492. —- capillata, Sav., 492. Chloéon, Leach, 458. — dipterum (ZLinn.), 485-437. Chloride of lime used as fish-poison in rock-pools, 180, 160. Chloroperla, Newm., 436. INDEX. Chorizopora Brongniarti, Hincks, 3, 21. Chrysopa, Leach, 440, 455. — flava (Scop.), 441. Cidarids, 174. Clathrina, Gray, 301; mentd., 307. — clathrus, O, Schmidt, 307, 311, 317. — contorta, Haeck., 311. — coriacea (Dendy), 801, 307, 516, 317. — sp., 3ll. Clausilia bedentatu, Strom, 205. — rugosa, Drap., 210. — spp., 204, 210. Cleantiella, Richardson, 69-71. Cliona, Grant, mentd., ftnote 524. Clupanodon neopilchardus (Steind.), 479. Clupea harengus, Zinn., pectoral girdle, 506-509. —pilchardus, Wadlb., pectoral girdle, 505— 509. — sprattus, Linn., pectoral girdle, 505- 509. Clupeids, new type of Teleostean Carti- laginous Pectoral Girdle in, by E. 8. Goodrich, 505-509. Cnidoglanis megastomus (Rich.), 480. Cochlicella acuta (Mu/Z.), 205. — barbara (Linn.), 206. Codium, Stack., spp., 160. Coleoptera, maxillule in the, 442-445. Colidotea, Richardson, 69-71. Collinge, W. E., Paracubaris spinosus*, a new genus and species of terrestrial Isopoda from British Guiana, 61-63 ; Oral appendages of certain species of Marine Isopoda, 65-93. Coloration, warning, in fish and birds, 47-60. Colpognathus dentex, Cuv. § Vail., 480. Conchodytes meleagrine, Peters, 351, 392. Conescharellina, d’Orb., 401, 403, 406-7. — angulopora, Hasw., 405, 411, 420-422, — angustata, d’Orb., 419. -— cancellata (Bush), 419-428. — cancedluta, Waters (pars), 420. — clithridiata, Gireg., 407. — conica, Hasw., 421-424. — conica, Hantken, 425, — depressa, Hasw., 422. — elegans, \Waters, 421. — eoccena, Vevian?, 404, 424. — flabellaris, Zev., 405, 420-423, INDEX. Conescharellina znc?sa, Waters (pars), 423. — philippinensis (Bush), 419-428, ef. 404- 406, 411. Congrogadus subducens, Rich., 483, 485. Conus magus, Linn., 154. Copepoda, distrib. in Ivish Sea, 95-126. Copidosoma, Ratz., mentd., 331, Coral Formations of the Abrolhos Islands, 176-180. Coralliocaris lamellirostris, De Man, 390. — lucina, Nobili, 351, 390. — superba (Dana), 351, 390. Coris auricularis, Cuv. § Val., 149, 482, 486, Cormorants, mentd., 476. Coryzichthys diemensis (Rieh.), 484. Coscinodiscus, Ehr., distrib., 99, 100. — concinnus, W. Sm., distrib., 101, 103. — Grani, Cleve, distrib., 103. — radiatus, Ehr., distrib., 101, 103. Cosmopelia elegans (Temm.), 467. — — neglecta, Mathews, 464. Cotile riparia, Bove, its food, 51. Crahs, mentd., 170. Crabyzos, Sp. Bate, 69-71. Crassatellites oblonga, 7’. Woods, mentd., 501. : Cribrilina acanthoceros, MacG., 12. — annulata, Fabr., 10. — Balzaci, Waters, 10, 12, 13. — figularis, Hincks, 12, 18. — Gattyxe, Hincks, 10, 11, 12, 13. — innominata, Norm., 12. — latimarginata, Busk, 12. — monoceros, Busk, 10, 12. — nitido-punctata, Smtt, 12. — patagonica, Waters, 12. — projecta, Waters, 10, — punctata, Hass., 10, 12. — radiata, Smith, 3, 10; mentd., 2. Crinoid, sp., occurring in the Abrolhos Islands, 170. Criserpia, H. M. Edw., 32, Crisia, Zam., 24-30; mentd., 2. -— acropora, Busk, 28. —aculeata, Hass., 25. — biciliata, MacG., 25. — conferta, Busk, 4, 25, 29. — Corbini, Canu, 25. — cribraria, Stimps., 25. — cuneata, Mapl., 25. | | | 515 Crisia denticulata, M. Edw., 25, 28. — — var. verdensis, Waters *, 3, 27, — denticulata, Waters, 29. — eburnea,'Hincks (pars), 25, 30. — — var. laxa, Busk, 30. — eburneo-denticulata, Smt, 25, 30. — Kdwardsiana, d’Ord., 25, — elongata, WM. Edw., 25, 28. — elongata, Harmer, 25, 28. — elongata, Norm., 4. — fistulosi, Heller, 25, 30. — geniculata, M. Edw., 25. — Holdsworthii, Busk, 25. — howensis, MacG., 25. — inflata, Waters, 25. — kerguelensis, Bush, 25, 80. — maxima, Jod., 25, 27. — operculata, Rob., 25. — pacifica, Rob., 27. — pugeti, Fob., 25, 27. —yramosa, Harmer, 4, 25, 30. — serrata, Waters, 28. — sertularoides, Z’Orb., 25. — sigmoidea, Waters, 3, 29. — sinclairensis, Busk, 28. — tenella, Calv., 4. — tenuis, MacG., 25. — tubulosa, Busk, 3, 25, 26. — vincentensis, Waters *, 3, 29. Crossland, Cyril, coll., 1, 345. Cryptoplax, Gray, sp., 170. Cubaris, Lrandt, mentd., 61. Cumulipora angulata, v. Maehr., mentd., 15, Cuncuma leucogaster (Gmel.), 477. Cupularia, Lame., 401-415; mentd., 6. — bidentata, Reuss, 404. — canariensis, Busk, 399-403, 406, 407, 410, 413, 421. — deformis, Busk, 413. — denticulata, Conrad, 412-415. — doma, Smitt, 413. — guineensis, Busk, 411. — Haidingeri, Reuss, 415. — Haidingert, Manzoni, 414. — Johnsoni, Busk, 411, 418; mentd., 899- 402. — Lowei, Busk, 400, 402, 412, 413, 415; mentd,, 399. — minima, Busk MS., 419. — Oweni, Gray, 4, 413. 516 INDEX. Cupularia Reussiana, Manzoni, 413. — unbellata, Defr., 4, 400, 415, 414. Cycloloculina, ZHeron-Allen § arland, mentd., 502. —annulata, Heron-Allen § Larland, mentd., ftnote 502. —polyegyra, Heron-Allen § Harland, mentd., 502. Cyclostoma acutum, Drap., 209. Cymbalopora, Hag., 408. Cynthia, Gist/., mentd., 288. Cypselus apus, Z//., its food, 51. Cytherea reticulata, Zinn., 154. Dactylophora nigricans, Rich., 482. Dakin, W. J., Percy Sladen Trust Expedi- tions to the Abrolhos Islands, 127- 180, Dampieria lineata, Cast., 480. Darwin, Charles, his visit to the Abrolhos Islands, 133. Dascillus, Zatr., 443. Defrancia, Bronn, mentd., 37, 38. Delma fraseri, Gray, 461. Demigretta sacra cooktowni, Mathews, 475. Dendronotus, Ald. § Hance., mentd., 170. Denisonia coronata (Schley ), 460, 468. Dermaptera, maxillule in the, 438-434. Diachoris hirtissima, Heller, 8. Diastrophus, Hart., mentd., 329. Diatoms, distrib. in Irish Sea, 95-126. Diloptaxis, Reuss, 409. Diplodactylus ornatus, Gray, 461. — spinigerus, Gray, 461. — vittatus, Gray, 461. Diplopterus, Agass., mentd., 184, 187. Diplosoma, Farre, sp., abundance, 257. Diptera, oceurr. on the Abrollios Islands, 170; maxillule in the, 451, 452. Dipterus, Sedg. § Murch., mentd., 181. Discocavea, Marss., 37. Discoescharites, Roemer, 408. Discoflustra doma, VOrb., 414. Discoflustrella, VOrb., 408, 414. — doma, VOrb., 413. Discoflustrellaria, VOrb., 408, 414. Discopora albirostris f. pusilla, Smitt, 22. — coccinea, Srwvtt, mentd., 13. ~- stenostoma, Smztt, mentd., ]3. Discopora ventricosa, Smitt, mentd., 13. Discoporella, VOrb., 401, 408. — Beradana, dOrb., 414. — umbellata, V’Orb., 414. Discosparsa marginata, d’ Orb., 37. Dolphins, mentd., 469. Dorodotes levicarina, Sp. Bate, 346, Doryphora, Ji., mentd., 453. — decemlineata, Say, 445. Dry-rot, efflorescence, mentd., 299. Echinometra mathe, Blainv., 148, 170. Kehinus esculentus, Linn., 175. —sp., mentd., ftnote 259. Edotia, Guér.-MWén., 69-72. -— bicuspida, Miers, 88. — bicuspida, G. O. Sars, 88. — hirtipes, Miers, 86. — nodulosa, Miers, 89. Egernia kingi (Gray), 461. — stokesi (Gray), 461, 466, 467. — whitei (Lacep.), 461, 468. Emberiza schceniclus, Zinn., its food, 52. Engidotea, Barnard, 69-71. Ennalipora, Gibb §& Horn, 409. Enoplosus armatus (Shaw), 482. Entalophora raripora. d’Orb., 4. — reticulata, Calv., 4. — subverticillata, Cadlv., 4. Enteropneust, a new, mentd., 180. Ephemera, Zinn., 435, 486. — danica, Miill., food of fishes, 48, 51; food of birds, 52, 53. Ephemerella ignita, Poda, food of Trout, 57; as artificial insect taken by birds, 58. Ephemeridee, maxillulee in the, 485-436, Epimys fuscipes ( Waterhouse), 459, 467. Epinephelides armatus (Cust.), 420. — leai, Ogilby, 480. Ipinephelus fasciatus, Forsk., 480. — merra, Bloch, 480, 485, Erichsonella, Benedict, 69-72. — attenuata, Richardson, 35. Erichsonia attenuata, Harger, 85. Erolia ferruginea chinensis (Gray), 475. Eschara spongites, Pall. (pars), 16. Euchzeta norvegica, Boeck, distrib., 114, 122. Eudy ptula minor woodwardi, Mathews, 468. INDEX. Kuwidotea, Collinge *, 69-71, 84. — peronii, Collinge *, 84. Wumetopias albicollis (Péron), 478. Hunice antennata (Sav.), 488, 495. — australis, Quatrefages, 488, 495, — bassensis, Benham, 495. — LHleyi, Baird, 495. — leucodon, Ehlers, 496. — leuconuchalis, Benham, 495. — Murrayi, MeInt., 495. — pycnobranchiata, MeInt., 495. — Niciliensis, Grude, 487-489, 496. — tentaculata, Quatrefages, 488, 495. — valida, Gravier, 496. Eurythoé aleyonia, Gravier, 492. — complanata. (Pall.), 488, 492. — levisetis, Fauv., 492. Husicyonia carinata (Oliv.), 345, 350, 368. Eusymmerus, Richardson, 69-71. Tutbalenessa djiboutiensis, Gravier, 488, 492, Evans, A. M., Structure and oceurr. of Maxillule in the Orders of Insects, 429-456, Exoceetus, Linn., mentd., 508. Exogone clavigera, Claparéde, 234, 256 — fustifera, Hasw.*, 218, 222, ee tnae — gemmifera, Digoatien 218, 2 — heterocheta, McInt. (sphalm. 21 221. — heterosetosa, MeInt., 221, 239. — naiidina, Girsted, 219, 236. — verrugera, Claparede, 219, 221, 227, 239. Exogonee, 217-242. TDiagwack sot to the Abrolhos Islands (W. J Dakin), 127-180. 219, 289, Fauna of Houtman’s Abrolhos Islands (Alexander), 457-486. Fauvel, P., Annéliles Polychétes de V’Archipel Houtman Abrolhos, 487- 500. Feeding-habits of fish and birds, by J. C. Mottram, 47-60. Fenusa melanopoda, Cam., 449, 450. Filisparsa superba, Jui/., 4. — tubulosa, Busk, 35. Fish, feeding-habits, 47-60. Flabellopora elegans, d’Orb., 422 Flustrelia hispida, Gray, 5, 38, 402. 517 Forficula, Zinn., 452, 453. — auricularia, Tine 433, 484, 439-442. Fringillide, mentd., 59, Frogs, mentd., 159. Fucus, Zenn., spp., abundance, 256. Fungia (Zam.), its absence in the Abrolhos Islands noted, 158. Fusus, Lam., sp., mentd., ftnote 259, Gabianus pacificus georgii (Aing), 478, 484. Galeichthys thalassinus, Zéipp., 480. Galeorhinus antarcticus (Giinth.), 479. Gasterosteus, Linn., mentd., 508. Gatenby, J. B., Germ-cells, fertilization and early development of Grantia (Sycon) compressa, 261-297. Gelignite, used for capture of fish, 130. Gemellipora arbuscula, Calv., 23, D4, Geotrupes, Fabr., 443. Gephyrea, Koch, mentd., 160. Gerris thoracica, Fabr., not taken by Trout, 50. Gilbert, John, collector, 458. Gilchrist, J. D. F., Post-Puerulus stage of Jasus lalandii (Milne Kdw.), Ortmann, 189-201. Glaucosoma hebraicum, Wich., 481. Globigerina, d’Orb., sp., mentd., 153. Glottis nebularius Mage) ( Vigors), 475. Glyptidotea, Stebbing, 69-72. — lichtensteini, Stebding, 7. Glyptonotus, Lights, 71. — cecus, Miers, 73. — sabint, Miers, 74. — sp., mentd., 70. — tuftsit, Miers, 74. Goniistius gibbosus, Rich., 482. Gonodactylus brevisquamatus, Paz/s., 345, 351, 362-364. — chiragra (Labr.), 161, 359; mentd., 363. — — var. smithii, Pocock, 351. — demani, Henderson, 545, 351, 359. —— var. spinosus, Bigelow, 345, 351, 361. — fimbriatus, Lenz, 345, 362-364. — glaber, Brooks, em. Llenderson, 351, 362. — — var. rotundus, Borradaile, 362. -— — var. ternatensis, De Man, 362. . DAG. — pulchellus, Miers, 345, 346, 351, 364. — graphurus, Jer: 518 INDEX. Gonodactylus spinosus, Lenz, 861, 365. Gonorhynchus greyi, Rich., 480. Goodrich, KE. S., Restorations of the head of Osteolepis, 181-188; uew type of Teleostean Cartilaginous Pectoral Girdle found in young Clupeids, 505— 509. Grantia, Flem., 317, 322. — compressa (O. Fub.), 299, 800-812, 317, 818, 321, 824; germ-cells, fertilization and early development, by J. B. Gatenby, 261-297. Grubea clavata, Quatrefages, 236, 240. — fureelligera, dug., 218. — kerguelensis, McInt., 218, 228, 228, 281, 236, 240. — limbata, Claparéde, 228, 229. — quadrioculata, Aag., 218, 223, 228, 281, 236, 239, 240. — protandrica, dag., 235. — pusilla, Dujardin, 222, 255, 236. — pusilloides, Hasw.*, 222 2 241, — sp. undet., 251. -— tenuicirrata, Claparéde, 228. Guano-collecting, method used in the Abrolhos Islands, 138; its effect upon flora, 138. Guancha coriacea, (?), 315-317, ftnote 307; mentd., 299. — — tropus primordialis, 300. Guinardia, Perag., distrib., 99, 101. — flaccida, Perag., distrib., 108, 124. Gulls’ nests on Abrolhos Islands, 160; mentd., 476. Gymnodactylus miliusii (Bory), 460. Hematopus ostralegus picatus (Iving), 474, 484. — niger bernieri (Mathews), 474. Halichondria panicea, Johnst., 254. Haliporus steindachneri, Balss, 346. Haplarmadillo, Dodifus, mentd., 61. Harpiliopsis, Borrad., 388. — beaupresi, Borrad., 588, 389. —— depressus, Borrad., 388, 389. Harpilius, Dana, 388. — beaupresii (Avd.), 351, 589. — consobrinus, de Man, 389. — depressus (Séemps.), 845, 851, 389, Harpilins gerlachei, Nobili, 345, 351, 388, 389, 390. — lutescens, Dana, 388, 389. Haswell, W. A., The Exogonew, 217-242. Hlaswellia alternata, Calv., 4. Helicella itala (Zinn.), 206. — sp.?, 204. — virgata (Da (.), 205, Helicigona lapicida (Linn.), 205, 207. Helicodonta obvoluta (Miill.), 205. Heliodoma implicata, Calvet, 419. Helix acuta, Miill., 206. — albella, Linn., 204. — arbustorum, Linn., 204, 207, —aspersa, Miill,, 205, 207. -— auricularia, Linn., 210. — balthica, Zinn., 211. — barbara, Linn., 206. — Basteri, Linn., 213. — carinula (Gimel.), 206. — cellaria, Miill., 205. — complanatus, Linn., 204, 212. — concinna, Jeftreys, 207. — contecta, Millet, 214. — contortus, Linn., 212. — corneus, Linn., 211. — fasciata, Miill., 214. - — fragilis, Linn., 204. — gothica, Linn., 206. — vrisea, Linn., 205, 207. — hispida, Linn., 207. — hortensis, Miill., 208. — ttala, Linn., 206, — lapicida, Linn., 206, 207. — leucas, Linn., 208. -— limosa, Linn., 204, 211. — lubrica, Mill., 209. — nemoralis, Linn., 207, 208. — octona, Linn., 209. —pisana, Miill., 205, 208. — pisana, Ehrbg., 209. — planorbis, Linn., 211. — pomatia, Linn., 207, 208. — pulcheila, Wiill., 205. — putris, Linn., 210. — sptrorbis, Linn,, 204, 212. — stagnalis, Linn., 211, 213. — stagnorum, Gwel., 213. — subcylindrica, Linn., 209. — tentaculata, Linn., 211, 213. — vivipara, Linn,, 211, 218, 214. Helix vortex, Linn., 212. Helodes, Payk., 442. Hemerobius, Linn., sp., 441, 442. Hemimerus hanseni, Sharp, 443, 444. — talpoides, Hansen, 433. Hemipodius scintillans, Gould, 458, 463. — varius, Temm., 463. Hepialus humuli, Fadr., 447. Neptagenia, Vayssiére, 435. — sulfuria, Mill., not taken by ‘Trout, 60, 51; as bird food, 58. Herdman, W. A., Spolia Runiana.—III. Distrib. of certain Diatoms and Cope- poda.... in the Irish Sea, 95-126,— IV. Notes on the abundance of some marine animals, 247-259. Hermenia verruculosa, Grube, 491. Hevrings, their food, 122. Hesione Ehlers?, Gravier, 493. — pantherina, Fzsso, 488, 493. — splendida, Aug., 493. Heteractis, Cun. § Bassl., 407. Heterodontus philippi, Bloch § Schneider, 149, 479. Heteronota bynoei (Gray), 460. Heteroscarus filamentosus, Cast., 483. Heteroscelus incanus brevipes (Vierllot), 475. ; Heterostegina, d’Orb., absence on Abrolhos Islands, noted, 160. Hippolyte orientalis, Heller, 383. — proteus, Pauls., 351, 382, — varians, Leach, 383. — ventricosus, M, Edw., 588. Hippothoa Brongniartii, d’Ord., mentd., ftnote 11. — distans, MacG., 3, 20. — divaricata, Lamv., 3, 20. — hyalina, Waters, 3, 20. — spongites, Smitt, 16. Hircinia variabilis, Schulze, mentd., 309, 310, 312. Hirundo neoxena carteri, Mathews, 465. —rustica, Linn., its food, 51; mistakes artificial for real insect, 58. Histriobdella, Ben., mentd., 332. Hoghen, L. T., Nuclear Phenomena in the Oocytes of the Gall-tly Mewroterus, 327-333. Holoporella, Waters, 404, 407. — aperta, Waters, 3, 23. INDEX. 519 Holoporella pusilla, MVaters, 3, 22. Holothurians, occurr. in Abrolhos Islands, 161. Hornera brancoensis, Calv., 4. — eburnea, Caly., 4. — tubulosa, Busk, 33. Houtman’s Abrolhos Islands, see Abrolhos Islands. Hydroprogne tschegrava strenua (Gould), 470. Hydropysche, Pict., sp., 447. Hygromia hispida (Linn.), 204, 207. Hylotoma rose, Berlese, 448. Hymenoptera Tenthredines, maxillulee in the, 448-451, Hypoleucus varius hypoleucus, Brande, its nest, 160, — -—— perthi, Mathews, 476. Hypoplectrodes nigrorubrum, Cuv. § Val., 480. Hyporhamphus intermedius (Cant.), 480. Hypotenidia philippensis (Zinn.), 464. — — mellori, Mathews, 464. Hypsipops microlepsis, Giinth., 482. Ichthyaria avicularia, Calv., 4. Idmonea, Zam., mentd., 31. — atlantica, Forbes, 4. Idotea, Labr., 69-72. — affinis, M. Edw., 81. — baltica (Paill.), 79, 80. — bicuspida, Owen, 88. — bicuspida, Streets & Kingsley, 88. ceca, Say, 73. caudacuta, Haswell, 84. distincta, Guér.-Mén., 84. edwards, Guér.-Mén., s1. excavata, Hasw., 81. — festiva, Chilton, mentd., 70. — granulosa, Rathke, 79-81. — hirtipes, Dana, 76. — hirtipes, M, Edw., 86. — wrorata, Harger, ftnote 79. — lalandii, M. Edw., 81. lichtensteint, Krauss, 77. — marina var. phosphorea, Miers, 79. marmorata, Packard, 88. media, Dana, 76. nitida, Teller, 81. nodulosa, Kroyer, 89, 520 INDEX. Idotea oregonensis, Dana, 76. — peronti, M. Kdw., 84. — phosphorea, Haryer, 79. — pulchra, Lockington, 88. — rectilinea, Lockington, 78. — resecata, Stimps., 75. — sabini, Kroyer, 74. — stricta, Dana, 84. — tuftsi, Stimps., 74. — ungulata, Lam., 81. — whitet, Stimps., 76. — wosnesenskiti, Brandt, 76. Idoteinea, Collinge *, new group of Valvi- fera, 71. Llybius, Lrichs., 452. Trish Sea, distrib. of Diatoms and Cope- poda in, 95-126, Tsias, Boeck., sp., mentd., 122. Tsolepisma, Stlvestr?, 432. Isopoda, a new genus of, from British Guiana (W. E. Collinge), 61-63 ; Oral appendages of, by W. IE. Collinge, 65-93. : Tsotoma palustris, Mill., 435. Japyx, Hal, 431. Jassa falcata (Mont.), 870. Jasus lalandii (Milne Ldw.) Ortmann, a Post-Puerulus stage of, by J. D. F. Gilchrist, 189-201. Julis, Cuv., sp., 485. Kangaroo, early description of, 457. Kennard, A. S., and B. B. Woodward, Linnean species of Nou-Marine Mol- lusca in the Linnean Collection, 203- 215. Kent, Saville, his visit to the Abrolhos Islands, 138, 184; incorrect about plutonic rccks in the Abrothos Islands, 137. Kionidella, Hosch., 408. Kyphosus sy dneyanus, Giith., 482. Lacuna, Twt., spp., abundance, 257. Lagenipora, Hincks, ftuote 402. Lagoa crispata, De Giryse, 446. Laminaria, Gray, spp., abundance, 253, 256, Laminopora contorta, Mich., 28, Laridze, mentd., 52. Lates calearifer, Bloch, 481, 485. Lauderia, Cleve, distrib., 99, 100. — borealis, Gran, distrib., 101, 106, 124. Leander concinnus ( Dana), 345, 351, 393. — erraticus, Desm., 393. — longicarpus, Ortm., 393. — natator, M. Edw., 346, 393. — natator, Nobili, 392. — squilla (Linn.), 392. — tenuicornis, Say, 346, 351, 392. Lepidaplois vulpinus (Ztich.), 482. Lepidocampa fimbriatipes, Carpenter, 431. Lepidonotus acantholepis, Grube, 488- 491, — carinulatus, Grube, 488, 490. — cristatus, Grube, 488, 489. — oculatus, Baird, 488, 489. — squamatus (Linn.), 490. Lepidoptera, maxillule in the, 445-447, Lepisma, Linn., 452-5, — saccharina, Linn., 452. Lepralia brancoensis, Calv., 4. —- ciliata, Barrois, 23. — collaris, Jull., 22. — crassimarginata, Hinchs, 22. — cucullata, Busk, 3, 21. — fissa, Busi, 22. Magnevilla, Busk, 21, — peristomata, Waters, 3,21. — Poissonii, Aincks, 4; mentd., ftnote 11. — Waters’, Cal ., 22. Leptophethon lepturus dorothez, Mathews, Cia ® Leptophlebia marginata, Maton, not taken by Trout, 50. Lethrinus opercularis, Cuv. § Val., 481, 485. Leucandra aspera, O. Schmidt, 299, 316, 317. Leuciscus cephalus, ZVem., its food, 51, — leuciscus, White, food of, 48, 51. — rutilus, Flem., its food, 51. Leucopolius — ruficapillus (Mathews) 474. Leucosolenia, Bow., 801, 310, 315. — coriacea (Zvem.), mentd., 307. — gardineri, Dendy, 807, 312. — poterium, Haeckel, mentd., ftnute 307, — stolonifer, Dendy ftnote 302. tormenti INDEX. 521 Leuctra, Steph., 436. Lialis burtonis, Gray, 461. Lichenopora, Defr., 37, 38. — hboletiformis, @’Orb., 37. — ballata, MacG., 36. — californica, Waters, 35. — conjuncta, MWich., 37. — cumulata, Wich., 37. — echinata, MacG.., 35. — formosa, d’Orb., 36. — Francquana, d’Orb., 36. — gregaria, d’Orb., 37. — hispida, Flem., 36. — Holdsworth, Busk, 34, 35, 36. —irregularis, Norm., 8, 33, 36, 37, — — var. composita, Waters *, 3, 34, 38. — magnifica, MacG.,, 36. — meandrina, Peach, 36, 37. — micropora, Reuss, 37. — multistellata, @ Orb., 37. — neapolitana, Waters * (nom. tent.), 36. —- novee-zelandiz, Busk (sp. dub.), 34, 35, 36, — pristis, MacG., 86, 37. — pustulosa, Waters, 36. — radiata, Hincks, 36. - sp., mentd., 2. — suecica (Hennig), 37. — turbinata, Dej., 37. — verrucaria, Fabr., 37. — yictoriensis, Waters, 35, 36. Limax agrestis, Linn., 206. — albus, Linn., mentd., 203. — ater, Linn., 206. — — var. rufus, Linn., 206, — — yar. albus, Miill., 206. — flavus, Linn., 206. — maximus, Linn., 206. Limestone in the Abrolhos Islands, 137, 154. Limnea auricularia (Linn.), 210, — glabra (Dfiill.), 209. — palustris (Wiill.), 205. — pereger (Miill.), 210, 211. — stagnalis (ZLrnn.), 211. — truncatula (Miill.), 211. Linnean Collections, Non-Marine Mollusca in (Kennard & Woodward), 203-215. Liolphus planissimus (Herbst), occurr. in Abrolhos Islands, 161. LINN. JOURN.—ZOOLOGY, VOL. XXXIV. Lithodomus, Cuv., 160, Littorina, Fér., spp., abundance, 255, 257. Lizards, mentd., 133, 170, 174, 477. Lunularia, Busk, 417. Lunulites, Zam., 401-408, 418. — axgulopora, \WVoods, 422, — Beisseli, Warss., 403. — canariensis, Manzoni, 411. — canariensis, Busk, 400. — eancelluta, Busk (pars), 419. —- conica, Busk, 418. — eupulus, Bush, 403, 417. — gibbosa, Busk, 403, 417. — goldfussi, Hag., 402, 403, — Hagenowi, Bosc, 403. — Haidingeri, Reuss, 414, — ineisa, Hincks, 422. — levigata, Canu, 402. — witra, Marss., 403. — patelliformis, Mapl., 403, 417. — patelliformis, Marss., 417. — philippinensis, Busk, 420. — radiata, Zame., 418. — repandus, Mapi., 417. — salebros, Warss., 403. — sella, Marss., 402, 403. — semilunaris, Marss., 417. —- transiens, Gregory, 403. — umbellaia, Defy., 414. — urceolata, Cuv., 410. — urceolata, Blainv., 414. — urceolata, Lamk., 410. Lutianus chrysoteenia, Bleek., 481, 485, 486. — johni, (?), 486. Lycodontis woodwardi (McCulloch), 480. Lygosoma lesueuri, Dum. § Bibr., 461. — prepeditum, Blngr., 462. — quadrilineatum (Dum. § Bibr.), 462. —richardsoni (Gray), 462, 468. — sp., 462, 466. Lymnodynastes dorsalis (Gray), 462. Lysidice collaris, Grube, 488, 496. — fullax, Ehlers, 496. — Ninetta, dud. § Ldw., mentd., 496. Lysiosyuilla miultifasciata, JV0od-Mason, 351, 359. Lysmata trisetacea, Heller, 346. Machilis, Latr., 429, 432, 452, 455, Mackerel, its food, 122. ae 38 522 INDEX. Macrochiridothea, Ohlin, 69-71. Macronectes giganteus albus (Potts), 469. Macropus eugenii, Desm., 159, 467; Pel- sart’s description of, 152. — — houtmanni (Gould), 459. Macrura, occurr. in the Abrolhos Islands, 170. Macrurans, occurr. in A brolhos Islands, 161. Macrurous Decapoda, Red Sea, 3865-398. Madrepora corymbosa, P. Forsk., 154,161. — sp., 158, 161. Mallophaga, maxillule in the, 437. Mamestra persicarie (Linn.), 446. Mamillopora, Smtt, 408. — bidentata, Reuss, 401. — cupula, Smrtt, 404. — Edwardsi, Juil., 4. — simplex, Kosch., 4, 404. — Smuittit, Caly., 4. Mangroves, ocewr., on Abrolhos Islands, 35, 166, 170, 174. Maregravia diemensis (err.), 485. Margaritana margaritifera, Zinn. 215; mentd., 372, 376. Marine animals, Notes on the abundance of (Spolia Runiana IV.), by W. A. Herdman, 247-259, Maxillule of Insects, Structure and occurr., by A, M. Evans, 429-456, Meandrina, Lam., sp., 158. Megalichthys, Agass., mentd., 187. Megalopterus tenuirostris melanops (Gould), 472, 484. —— — tenuirostris (Temm. § Laug.), 484. Megaptera longimana, Rud., 478. Melanopus femoratus, Scudd., 434. Melanosterna anzethetus novee-hollandice (Stephens), 471. Meleagrina, Zinn., mentd., 392. Meliceritites, Roem., 16. — sp., mentd., 15, Melobesia, Zam., mentd., 18. Membranipora bellula, inels, 4. — canariensis, Simitt, 411. — cervicornis, Busk, 4. — crassimarginata, Hincks, 9. — eurvirostris, Wincks, 3, 9. — Dumerilii, Awd., 402. — (?) filum, Judl., 4. — Flemingii, Waters, 9. — eranulifera, Hincks, 4. Membranipora lineata (Linn.), 402. — nitida, Johnst., 402. — patellaria, Moll., 4, 18. — pilosa (Zinm.), 402. — quadricornuta, Waters *, 3, 9. — tenuirostris, Hincks, 3, 9, 402. Membraniporella, Simztt, 11. — lineata, Hincks, 10. — nitida, Sm7tt, 10-12. Mesembryanthemum, Zinn., sp., oceurr. on the Abrolhos Islands, 165. Mesidotea, Richardson, 69-72. — sabini, Richardson, 74. Mesoprion Johni, Renard, 486. Metapeneus curvirostris, Nobili, 367, — stebbingt, Nobili, 365, — vaillanti, Nobili, 366. Microcalanus pusillus, Sars, distrib., 98. Microcanthus strigatus (Cuv. § Vai.), 482. Micropora, Gray, 408; mentd., 6, 16. Microporella ciliata, Harmer, 3, 23, 402. — coronata, Aud., 408. — diadema, MacG.., 402. — impressus, dud., 402. — Malusii, Awd., 402. — vtolacea, Johnst., 4. Microscopical preparations, new method described, 67-68. : Mitiolina, Khy., sp., 153. Mimicry in fish and birds, 47-60. Minea, Pearse, mentd., 61, 63. Minnows, food of Trout, 48. Mollia hyalina, Barrois, 20. Mollusca, Linnean species of Non-Marine, in the Linnean Coll, by A. S. Kennard & Bb, B. Woodward, 203- 215. Monacanthus chinensis, Bloch, 483. — megalourus, Rich., 488. Montipora, Quoy S Gaim., sp., 168. Motacilla boarula, Zinn., its food, 52; mis- takes artificial for real insect, 58. -— lugubris, Zrnn., its food, 52. rayi, Bonap., its food, 52; mistakes artificial for real insect, 58. Mottram, J. C., Feeding-habits of fish and birds, with special reference to warning coloration and mimicry, 47-60. Miiller, O. F°., his work on Mollusca referred to, 203, INDEX. 523 Multitubigera gregaria, VOrb., 87. — micropora, Reuss, 37. Mureena nubila, Rch., mentd., 480. Musca domestica, Linn., focd of Wagtails, Muscicapa grisola, Zinn., mentd., 52; feeding-hahits, 57. Mutton-birds, presence on the Abrolhos Islands, 138, 155, 160, 174, 458, 477. Mya lutraria, nom, Linn, 204, — margaritifera, Linn., 204, 215. — pictorum, Linn., 214. — twmidus, Linn., 214. Mycetophila, Mezg., 451. Myobatrachus gouldi (Gray), 462. Myriozoum strangulatum, Calv., 4. Myrmecophila, Zatr., 454. Mytilus anatina, Linn., 215. — cygnea, Linn., 215. — edulis, Zinn., abundance, 252-254. Myxomycetes, mentd., ftnote 322. Nautilus crista, Linn., 212. Nealbatrus chlororhynchus carteri (2oth- schild), 470. Neatypus obliquus, Matte, 481. Nebria brevicollis, Fabr., 448, 444. Nematus Hrichsonii, Hart., 449, Nemura, Latr., 486, 453. — sp., 437. Neosebastes panda (Rich.), 485. Nereis denhamensig, dug., 488, 494. — pelagica, Zinn., mentd., 494. Nerita fluviatilis, Linn., 214. — lacustris, Linn., 204, 214. — littoralis, Linn., 204. Neuroptera, maxillule in the, 439-442. Neuroterus, Hart., nuclear phenomena in the oocytes, 527-333, — lenticularis, Oliv., 327, 329. — numismatis, Oliv., 328, 329. Nikoides danee, Pauls., 394. — maldivensis, Borrad., 394. — spp., 345, 351, 594. Nitzschia delicatissima, Cleve, distrib., 99. Notodoris citrina, Bergh, mentd., 161. — sp., oceurr. in Abrolhos Islands, 161. Notopygos hispidus, Potts, 488. — — var. serratus, Fauy., 495. —labiatus, M‘Znt., mentd., 495. Notopygos labiatus (?), Benham, 493. Nuclear phenomena in the oocytes of Neuroterus, by L. T. Hoghen, 327- 333. Nudibranch, occurr. of a new, in the Abrolhos Islands, 170. Numenius cyanopus, Verllot, 474. Nyctiphanes norvegica (JZ. Sars), distrib., 114. Odax richardsoni, Giinth., 482. Oithona helgolandica, Claus, diswib., 98, 110-112, 118, 124. — similis, Ald. & Hanc., 98, 118. Oliothops cyanomelas, Rich., 482. Oniscus ungulatus, Pall., 81. Onychocella angulosa, Reuss, 4, 417. Onychoprion fuscata serrata (Wagler), 471, 484. Ophthalmolepis lineolatus, Cuv. § Val., 482. Oral appendages of Isopoda, by W. E. Collinge, 65-93. Orbitoides stellatus, Giimb., 401. Orbitolites, Lam., sp., mentd., 153. Orbitulipora petiolus, Lonsd., 404. Orthophragma stellata, Giimbel, 401. Orthoptera, maxillule in the, 454-435, Ortygodes varius, Gould ?, 467. — — scintillans, Gould, 462. — — stirlingi, Mathews, 468. — — varius, Mathews, 465. Osearella, Vosmaer, 310, 322. Oscillatoria, Vauch., 322. Ospreys, 174. Osteolepis, Val., Restorations of the head of, by E. S. Goodrich, 181-188. — macrolepidotus, dy., 181-186. Osthimosia avicularis, Waters, 3, 19. Otaria albicollis, Péron, 478. — australis, Quoy & Gaim., 478. Otionella, Can. § Bassl., 407. Otiorhynchus, Germ., 450. — suleatus, Fadr., 445. Pachymatisma johnstoni, Bowerb., 254. Peedophylax claviger, Husw., 229. -- verrugera, Claparede, 219. —- veruger, Claparéde, 219. Pagrosomus auratus, Forst., 481. Pagrus major, Temm., 485. Pagurid, sp., 175. Palemon petitthouarsit, Aud., 385. — torensis, Pauls., 393. Palemonella gracilis, Pauls., 395. — tenuipes, Dana, 351, 383. — tridentata, Borrad., 384. Palm trees, absence on Abrolhos Islands noted; 135. Paludestrina stagnalis, d’Ord., 213. Paludina minuta, Totten, 213. — sp., 209. Pamphilius dentatus, MacGill., 448. Pandion haliaétus cristatus ( Viedllot), 477. Panulirus penicillatus (Oliv.), 174. Parabetzeus culliereti, Cowt., 346. Paracalanus parvus (Claus), distrib., 98, 110, 112, 118, 124. Paracubaris, W. E. Collinge *, 61. — spinosus, W. Z. Collinge *, 62. Parapandalus adensameri, Balss, 346. — pristis (Lisso), 346. Parapeneeus fissurus (Sp. Bate), 346, Paraplesiops meleagris, Peters, 481. Paratypton siebenrocki, Balss, 346. Paridotea, Stebbing, 69-72; mentd., 78. — fucicola, Barnard, 83. — peronit, Stebbing, 84. — reticulata, Barnard, 88. — rubra, Barnard, 838. — unguiata, Stebbing, 81. — — var. atrovirens, Collinge *, 82. Paschocaris paschalis, Nobili, 882. Patella compressa, Linn., 210. — lacustris, Linn., 205, 210, — lutea, Linn., 210. — pellucida, Zrnn., 210, — vulgata, Zinn., abundance, 255. Pedicellina australis, Jull., 42. — cernua, Smitt, 3, 45. Pelagia, Lam., 37. — clypeata, Lamx., 37, 38. Pelagodroma marina dulcie, Mathews, 468, 484. Pelsart, Francis (1629), and the Fauna of the Abrolhos Isl., 457; his disastrous voyage to the Abrolhos Island, 131- 152. Penzopsis stebhingi (Nobili), 345, 350, 365. 524 INDEX. Penzeopsis stridulans (Wood-Mason), 345, 350, 366. — yaillanti (Nodil2), 245, 350, 366. — velutinus, Dana, 346, 366. Penzeus ashiaka, Wishinouye, 367. — carinatus, Dana, mentd., 367. — japonicus, Sp. Bate, 850, 367. — semisuleatus, De Haan, 350, 367. Peneroplis, Montif., sp., mentd., 153. Pentapus vitta, Cuv. § Val., 481. Pentias, Richardson, 69-71. Pentidotea, Richardson, 69-72; mentd., 78. — resecata, Richardson, 75. -— whitei, Richardson, 76. — wosnesenskii, Richardson, 76. Periclimenes borradailei, Rathbun, 387. — calmani, Vattersall *, 345, 851, 885. — demani, Kemp, 386. — ensifrons, Dana, 386. — kolumadulensis, Borrad., 587. petitthouarsii (Awd.), 851, 385. — potina, Nobili, 388. — seychellensis, Borrad., 386. -— spiniferus, De Man, 386. — spp., 345, 551, 386-388. — tenuipes, Borrad., 387. Perla, Geoffr., 486, Ee Perlodes, Banks, 456, 453. — dispar, (?), 436. Peropus variegatus (Dum. § Bibr.), 461. Petrel, Sooty, mentd., 458; —, Giant, 484. Petrobius oudemansi, Carpenter, 480. Pheopus pheopus variggatus (Scopol), 474. S Phalacrocera, Schin., 451. Phups chalcoptera, Lath., 464. Philosamia cynthia, (?), 446. Phryganea, Linn., 447. Phyllodactylus bilineatus, Gray, 460. —marmoratus (Gray), 460. — ocellatus (Gray), 460. Phyllopertha, Aiby, 443. Physa fontinalis (Zinn.), 213. — hypnorum (Linn.), 213. Physalia, 7%/., sp., 170. Pieris brassicee, Linn., 446. Pigeon, Bronze-winged, mentd., 458, Pinna, Zinn., sp., mentd., 9. Pisobia minuta ruficollis (Pall.), 475. Plaice, abundance, 248; its food, 122. Planorbis albus, AZill/., 205. INDEX. 525 Planorbis carinatus, Miill., 205. — complanatus (Lenn.), 212, 213. — contortis (Linn.), 212. — corneus (G'mel.), 211. — crista (Linn.), 212. -- leucostoma, Mrllet, 212. — planorbis, Linn,, 205, 211. — spirorbis (Linn.), 212. — umbilicatus, Miill., 205, 211, — vortex (Linn.), 212. Planorbulina, @’Orb., sp., mentd., 153. Platax orbicularis (Forsk.), 486. — teira (Lursk.), 486. Platycephalus bassensis, Cuv. § Vai,, 483. Platynereis Duwerilii, dud, § WM. LEdw., 488, 489, 494. — imsolita, Gravier, 494. Plecoptera filipalpia, A/ap., 436. — setipalpia, Alup., 436. Plutonic rocks, mnon-occurrence in the Abrolhos Islands, 137, 154. Pocillopora, Lem., sp., 158, 168. Podiceps fluviatilis, Degl. § Gerbe, mis- takes artificial for real insect, 58. Polita cillaria (Miill.5, 204. — Draparnaldi (Leck), 204. — radiatula, Alder, 204.. Polyides rotundus, Girev., 40. Polyophthalmus pictus, Dujardin, 488, 489, 497. Polypterus, Geoffr., mentd., 181, 186. Polystomella, Zam., sp., mentd., 153. Pomatias elegans (Miill.), 205, 214. Pontonia pinne, Ortmann, sp. dub., 391. Porella leevis var, subcompressa, Bush, 4. Porina borealis, Busk, 4. Porpita, Lam., sp., 161. Porzanoidea plumbea, Mathews, 467. — — roberti, Mathews, 464. Potamogeton, Tovn., sp., mentd., 47. Prattia, @’ Archiac, 408. Proboscina, Aud., 32, — Boryt, Aud., 32. — merassata, Smitt, 32. — Lamourourti, Aud., 31. Proidotea, Rac. § Sev., 70, 71. Protospongia, Kent, 322. Protula bispiralis (Sav.), 488, 489, 498. Pseudanodonta grateloupiana (Gassies), 215. — normandi (Dupuy), 215. Pseudarmadillo, Saussure, mentd., 61. Pseudocalanus elongatus (Boeck), distrib., 98F LONI TZ. Pseudolabrus parilus (Rich.), 482. Pseudonereis anomala, Gravier, 488, 494. — masalacensis, Grube, mentd., 494. Pseudoscarus gymnognathus, Bleek., 483. — sp., 485. Pseudosquilla ciliata (Fabr.), 551, 357. — megalophthalma, Bigelow, 351, 357. Psocide, maxillulee in the, 437. Ptencedus mathewsi mathewsi (Iredale), 465. Pterocaris typica, Heller, 346. Pterostichus, Bon., 443, 444. Ptychodera flava, Eschscholtz, 175. — pelsarti, Dakin *, 175. Ptychoptera, Meig., 451. Puellina, Jull., 10. Puerulus, Ortm., 199. Puffinus assimilis tunneyi, Mathews, 468, 484, Pulvinulina, Carp., sp., mentd., 153. Pupilla muscorum (Linn.), 204, 209. Purpura lapillus, Zinz., abundance, 255. Pyramidula rotundata (Mzil/.), 206. — ruderata (Studer), 205, 206. Python spilotes (Lacép.), oceurr. on the Abrolhos Islands, 159. — — variegatus (Gray), 460. Rabbits, 460. Radiopora conjuncta, Mich., 37. — cumulata, Mich., 37. — formosa, VOrdb., 37. — Francquana, VOrb., 34, 36, — wregularis, J. Yates Johuns., 33. —- multistellata, VOrb., 37. — pustulosa, VOrb., 36. Raphidia, Linn., 453. — sp., 439, 441. Rat, its oecurr. on the Abrolhos Islands, 138, 159. Rattus fuscipes, Waterhouse, 460. — norvegicus, Erv., 460. — rattus (Linn.), 460. Reniera, Gir., 8321; mentd., ftnote 324. — simulans, (?), 320. Retepora Imperati, Bush, 4, 22. — ramulosa, Calv., 4. 526 Rhamphostomella, Waters, 18. Rhinobatus banksii, Miller §- Henle, 479. Rhizocephalan parasite, mentd., 395. Rhizodopsis, Young, mentd., 187. Rhizosolenia, Ehr., distrib., 99-101. — alata, Brightw., distrib., 107, 124. —semispina, Hensen, distrib., 101, 107, 124. — setigera, Brightw., distrib., 10], 107, 124. —Shrubsolei, Cleve, distrib., 101, 124. — spp., distrib., 101. — Stolterfothi, Perag., distrib., 107, 124. Rhodites, Hart., mentd., 327, 329, 332. Rhbyacophila, Pict., sp., 447. Rhynchelaps bertholdi, Jan, 460. 107, Sabellaria alveolata (Zznzn.), abundance, 249-250. Salmo, Zinn., mentd., 508. — fario, Linn., feeding-habits, 47-57. Salvatoria kerguelensis, McInt., 228. Sagartia miniata, Gosse, 254. Saron neglectus, De Man, 846, 351, 882. Sauropatis sancta westralasiana (Campbell), 465. Screophzethon rubricauda Mathews, 476, 484. Scatophaga stercoraria, Latr., food of Wag- tails, 53. Schizoporella argentea, Hincks, 18. — Cecilit, Robertson, 20. “— contorta, Caly., 23. — cucullata, Jull. & Caly., 21. — divisopora, Waters, 19. — fissa, Hincks, 22. — hyalina, Ley., 20. | — nivea, Busk, 16, | — oligopus, Robertson, 3, 18. — porrelliformis, Waters *, 15. — pulchra, Neviant, 19. — sanguinea, Norm., 17. — spp., mentd., 2. ’ — spongites, Smit, 3,15, 16. — spongites, Thornely, 16. — trichotoma, Waters *, 3, 19. | — unicornis, Pergens, 8, 14, 16, 17, 402. — viridis, Thornely, 15. westralis, Schizotheca fissa, Hineks, 8, 22. INDEX. Schizotheca (?) Talismani, Calv,, 4. Scolopsis bimaculatus, Zéipp., 481, 485. Scorpzena sumptuosa, Cast., 485. Scrupocellaria Bertholetti, Hincks, 3, 5, 45. — cervicornis, Bush, 7. — cornigera, Smztt, mentd., 7. — Macandrei, Bush, 3, 6. — pusilla, Simztt, mentd., 7. — reptans var. Bertholletii, Waters, 5. — tridentata, Waters *, 3, 7. Sea-birds, great number of species on Abrolhos Islands, 188, 138. Seal, 485. Segmentina, Flem., 212. — lacustris, Lightfoot, 212 — nitida (Miill.), 205, 213. Selandria sixii, Voll., 449, 450. Selenaria, Leach, 401-407. — concinna, Busk, 402, 404. — concinna, 7. Woods, 4165. — fenestrata, Hasw., 416. — maculata, Busk, 402, 415, 417. — magnipunctata, Mapl., 416. — nitida, Mapl., 407, 418. — petaloides, d’Orb., 402. —- punctata, 7. Woods, 416. - Selenariade, 399. Selenariopsis, Maplestone, 409. Septifer bilocuiaris, Linz., 154. Serialaria lendigera (Linn.), 41. — semiconvoluta, Lamu, (sp. dub.), 41. Sericornis maculatus fuscipes, IV. B. Alev- ander *, 465. — — warreni, Mathews, 465. — halstoni, Grant, 465, 466. Seriola gigas, Giinth., 485. Serpula bispiralis, Sav., 498. Setosellina Roulei, Cadv., 4. Sherbornina, 2. Chapman *, 501. — atkinsoni, 1. Chapman *, 502. Sialis, Latr., 453. — lutuaria (Linn.), 440, 442, 450. Sticyonia sculpta, H, M.-Edw., 368. Sillago bassensis, Cuv. § Vai., 481. — ciliata, Cuv. § Val., 485. Silubosaurus stokes, J. 2. Gray, 458. Simulium, Zati., 451; food of fishes, 47, 48, 50. — sp., food of Trout, 67. Skeletonema costatum (Girev.), distrib., 99. INDEX. 527 Smittina cervicornis, Pallas, 4. — jacobensis, Bush, 4. -— trispinosa var, protecta, Thomps., 8, 21. — tropica, Waters, 8, 21. Snails, food of Trout, 48. Sotalia gadamu (Owen), 478. Sparus sarba, Forsk., 481. Spherium corneum ian 215. Spheroniscus, Gerstaecker, mentd., 61. Spheeropora, Hasw., 408. — fossa, Hasw., 409. Spheerosyllis antarcticus, Gravier, 226. —- hirsuta, Hhlers, 218, 226. — hystrix, Claparéde, 223, 224, 227, 231 235, 236, 240. — fexuee MclInt., 226, — Macintoshi, Ehlers, 225. — perspicax, Lh/., 218, 226. — pirifera, Claparéede, 224, Spheroides pleurogramma, Regan, 484. Sphyrena obtusata, Cuv. § Val., 480. Spirogyra, Link, sp., mentd., 47. Spolia Runiana.—IIT. Distrib. of certain Diatoms and Copepoda in the Irish Sea, by W. A. Herdman, 95-126; IV. Notes on the abuudance of some common marine animals, by W. A. Herdman, 247-259. Spongelia, Mardo, 321, 322. Sponges, occurr. in Agate: islands, 161, 162; Notes on their physiology (Bidder), 315 526 ; their origin, 322 -324; mentd., 219, ftnote 259, 395. Spongilla, Zam., 317, ftnote 321 ; fertiliza- tion, 263, 279. — fluviatilis (Padl.}, fertilization, 286. — sp., 268, Squatarola squatarola hypomelas (Padl.), 474. Squilla massavensis, Koss., 351, 356, Starfish, abundance, 248. Steganoporella, Hincks, mentd., 6. — Rozieri var. indica, Hincks, mentd., 13, 14, — — var. falcifera, Hincks, 13. — — var. labiata, Lev., 18, 14. Stelletta collingsi, Schmidt, 254. Stenopus spinosus (2zsso), 346. Sterna dougalli gracilis (Gould), 470, 484. — sp., mistakes artificial for real insect, 58, Sternula albifrons tormenti (Mathews), 471. Sternula nereis, Gould, 471. — — horni (Mathews), 471, 484. Stichopora, Hagenow, 408. Stichoporina, Sto/., 408, Stomatopoda and Macrurous Decapoda coll, in the Sudanese Red Sea (Tattersall), 345-598. Stomatopora, Goldf., 32. —- granulata, M. Edw., 4. — incrassata, Hincks, 33. Sturnus vulgaris, Naewm., mentd,, 52 Styela grossularia, Van Bened., abundance, 254-6, Stylasterina, mentd., 4. Stylotella, Lendenf., mentd., ftnote 316, Suceinea pfeifferi, Rossm., 205, 210. — putris (Zinn.), 210,211. Swan, Black, 484, Syeandra, Hick. 32 — raphanus, O. Schmidt, fertilization, 263, 277, 284, 286. Syeon, Risso, 801, 317, 322. — ciliatum, Fabr., 306; fertilization, 262, 265, 277, 284. — compressum, Dendy, 261. i aes O. Schmidt, 310, 317, 318, 322, 524; fertilization in, 802-304. — — tropus aquariensis, mentd., 299. Sylviidee, mentd., 52. Symmius, Richa ih (Al ee: — caudatus, Lichardson, 73. Synalpheus bispinosus, De Man, mentd,. 377. — biunguiculatus, Stzmps., 350, 375, 377. — fossor (Pauls.), 350, 374. — -— var, propinqua, De Man, 374. — gravieri, Cout., 350, 373. — heroni, Cowt., 346, 850, 374. — hululensis, Cowt., 345, 350, eee — pachymeris, Cout., mentd., 876 — quadridens, De Man, mentd., 37 78, — quadrispinosus, De Man, mentd., 378. — quinquidens, Tuttersall*, 345, 350, 376. —— savignyl (Guér.), 845, 346, 350, 375. —- septemspinosus, De Man, mentd., 378. — streptodactylus, Cowt., 345, 350, 373. — triacanthus, De Man, mentd., 377. — tricuspidatus, Heller, mentd., 375. — trispinosus, De Man, mentd., 378. — triunguiculatus ( Pardls.), 350, 874. — tumidomanus, Cowt., 874, 528 INDEX. Synaneeja horrida, Zinn., 485, 485, Synapta, Eschsch., sp., 162. Synerypta, Ehr., 822, 325, — spongiarum, Bidder *, 305-3813, — volvox, 306, 313. Synidotea, Haryer, 69-72, 85-89; ley to species, 86. — angulata, Benedict, 87. — bicuspida, G. O. Sars, 88. — hirtipes, Benedict, 86. — levis, Benedict, 89. — marmorata, Benedict, 88. —nebulosa, Benedict, 87. — nodulosa, Harger, 89. — pallida, Benedict, 87. Synisoma, Collinge, 69-71. Tadpole, not taken by Trout, 51, cf. 53, Tanypus, Meig, 451. Tattersall, W. M., Stomatopoda and Macru- rous Decapoda coll. by Mr. Cyril Crossland in the Sudanese Red Sea, 346-398. Telephorus lituratus, Fall., 443. Tellina corneum, Linn., 215. Temora longicornis, G. Sars, distrib., 98, 110, 112, 120-125. Tenebrio mollitor, Lann., 445. Tephreops tephreeops (Rich.), 482. Terns, 174; mentd., 476, 477; enormous numbers on the Abrolhos Islands, 188, 160,166,169 ; —, Roseate, mentd., 469. Thalamita ? stimpsoni, A. M. Edw., 148. Thalamoporella Jervoisi, Hincks, 14. — mamillaris, Lamz., 14. — Rozieri, Hincks, 3, 13. — nove-hollandie, Lev., 18. Thalasseus bereii (Lichtenstein), 470. — — ewendolene, Mathews, 470. — — pelecanoides (King), 470. Thalassiosira, Cleve, distrib., 99, 101. — eravida, Cleve, distrib., 106, 124. — Nordenskioldi, Cleve, distrib., 106, 124. Thalassoma aneitense, Giinth., 482, 485. — lunare, Linn., 482, 485. Thalenessa djiboutiensis, Fauv., 492. — oculata, McInt., mentd., 492. Theba leucas, Beck, 209. Theodoxus fluviatilis, Montf., 214. — lacustris, Montf., 214. Therapon humeralis, Ogilby, 480. Thor paschalis (/Zel/e7'), 351, 382. Thormora argus yar., Hasw., 489. Threpterius maculosus, Rich., 482. Thursius, 7ray., mentd., 187. Thye >7roma pacifica chlororhyncha (Les- son), 468. Thymallus vulgaris, Nelss., its food, 51. Tineina, as food of Trout, 47. Tipula, Zinn, 451. Trachichthodes affinis (Giinth.), 480. Trachurus declivis, Jenyns, 481. Trachypenzeus anchoralis (Sp. Bate), 350, 367. — curvirostris, Stémps., 367. Trachysaurus rugosus, Gray, 461. Trichogramma, Westw., mentd., 332. Trichoptera, Maxillule in the, 447-448. Trichopterous larvee, food of Trout, 47. Tridacna, Da Costa, mentd., 392. Trochopora, d’Orb., 407, 418-419. Trochus, Linn., spp., abundance, 257. Trout, Brook, feeding-habits, 47-57. Truncatella Montagui, Lowe, 209. — truncata, Mont., 209. Tryphoena pronuba, Hiibn., food of Spar- rows, 53. 2 Trypostega venusta, Norman, 4. Tubucellaria, d’Ord., 405. Tubularia indivisa (Zinn.), 254. Tubulipora aperta, Zarmer, 31, 33. — fimbria var. pulchra, Waters, 30. — flabellaris, Fabr., 31, 37. —- inerassata, Waters, 3, 32. — Lamourouxi, Waters, 3, 31, 33; mentd., 2, — liliacea, Harmer, 31. — occidentalis, Roberfson, 31, — organizans, d’Orb., 31. — phalangea, Couch, 31. — plumosa, WW. Thomps., 31. — pulchra, MacG.., 3, 30. Turbo erista, Linn., 212. — muscorum, Linn., 204, 209. — perversa, Linn, 204; 210. — pulcher, Reeve, 154. — reflexus, Linn., 214. — stagnalis, Bast., 215. — ulve, Penn., 213. — ventrosus, Mont , 213. Turnix varia, Mathews, 463, INDEX. Turniz varia scintillans, Gould, 463. — — stirling’, Mathews, 463. Turritella warburtonensis, Zate, mentd., 501. Turtle, Green, 485. Turtle-doves, mentd., 457. Ulex europeus, Linn., mentd., 47. Ulva, Zinn., sp., 166. Unio, Retz., 215. — pictorum, Refz., 214. — tumidus, Peétz., 205. Unwin, E. If., Notes upon the reprod. of Asellus aquaticus, 535-545, Upogebia heterocheir, Kemp, 596. — pseudochelata, Tattersall *, 345, 351, 395, — savignyi (Strahl), 351, 895, Uroglenopsis, (?), 306. Urtica marina americana, Seba, mentd., 498. Vallonia excentrica, Sterh7, 205, Valvata piscinalis (Wiill.), 205. Valvifera, new groups of, 71. Vanessa urticee (Linn), 446, : Velia cwrrens, Latr., not taken by Trout, 50. Vermetus, Adais., sp., 148, 150, 153, 160, 161, 170. D2) vo Verongia, Bowerd., 310, 322. Vetola lapponica baueri (Naumann), 474. Vibracella trapezvidea, Zewss, 401, 403. Vinenularia nove-hollandie, Masw., 18. Virago castanea (Lyton), 476. — gibberifrons (iS. Miill.), 476. Viviparus vivipara, Montf., 214. Volvox,, Miill., mentd., ftnote 307. Wallaby, 177; early description, 132. Waters, A. W., Littoral Marine fauna of the Cape Verde Islands, 1-45; Obs. upon the relationships of the (Bryozoa) Selenariadze, Conescharellinide, etc., fossil and recent, 399-427, Whales, mentd., 175. Whiteornis goodenovil — ruficapillus, Murhews, 465, Woodward, B. B., see Kennard, A. 5. Wynyardia bassiana, Spencer, mentd., 501, Zalophus lobutus, Ogilby, 478. Zenobiana, Stebbing, 69-71. Zonitoides nitidus (Miil/.), 204. Zoobotryon pellucidum, Ehr., 3, 41. Zostera, Linn., mentd., 248. — marina, Linn., abundance, 257. — sp., 160. Zosterops gouldi, Bonap., 466. PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET. es Ae, Taek LINNEAN SOCIETY OF LONDON. RULES FOR BORROWING BOOKS FROM THE LIBRARY. 1. No more than Six volumes shall be lent to one person at the same time without the special leave of the Council or one of the Secretaries. 2. All books shall be returned before the expiration of Six weeks from the time of their being taken out, but if not required by any other Fellow, they may, on application, be kept for a further period of Six weeks. 3. All books lent shall be regularly entered by the Librarian in a book appropriated for that purpose. 4. No work forming part of Linneus’s own Library shall be lent out of the Library under any circumstances. Norn.— Certain other works are included in this prohibition, such as costly illustrated works, volumes belonging to sets which could not be replaced if lost, and unbound parts of serials. 5. All carriage charges on books to be defrayed by the Fellow borrowing the same. 6. All borrowed books when returned to be sufficiently wrapped and protected. Borrowers will be held liable for any damage incurred owing to improper packing. 7. Newly published books shall not ordinarily leave the Library until after the expiration of oue month from the date of their being received. 4A GENERAL INDEX to the first twenty Volumes of the Journal (Zoology) may be had on application, either in cloth or im sheets for binding. Price to Fellows, 15s.; to the Public, 20s. NiO TA CsEes. Tar attention of the Mellows, and of Librarians of other Societies, is requested to the fact that the parts of the Journal (Zoology) of the present volume have been issued as follows :— Vou. 34. Nos. 225, 226, 227, 228, 229, and 230 the present number. Authors are entitled to 25 copies of their communications gratuitously, and may obtain another 75 by payment, as shown on the printed slip which accompanies the proof. If more than 100 copies are wanted, application must be made to the Council. Abstracts of the proceedings at each General Meeting and Agenda for the next are supplied to all Fellows. B. DAYDON JACKSON, General Secretary. Ne ep? ts aS: " ‘ i BY Me iNT