Cee) ‘ “ % ORE MAT ; ata e “at r - 2 dome ees Pye yo PAPER EY i « Miritetecoctee Some Ona ete ee day ‘ . + 5 +e eae tarde’ AOD yee eae ® a ee « ! ; . aww yr oy beta ‘ Tee wens t pene M . ata peermie et Pry re) F Ck 4, ‘. Ms 88-4 08 Os sane ts i “ i i vee sab % i RDS: iy 3 ‘ * ‘ or heh ee ih neha, tee Pat tammanionsheats 3 ct berss sity ther ate fF fag reo are eee Wyre ait. soctcaiecg-orne #33 ta tstatienecapemtcnt tt Lied agape ton yer SES aia wert Savectessreses ni te erty wea ps ty Reape wyh ee Wie bieens Splaksarapaptobeecwres Pope meaner et it thesl say bene ea esthe stitaesit arewtr ‘ Syareen ‘ : p re ¥ » ‘) wee ¥ ? rey RS Oy ee me bat we # 2 ri ‘ er) ; ‘ yy ; oe Bae Reena paNtatt haat gH “anki gsearene wats x y soca 8 B24 ae ati sere . ee, mn x Y i ales “ye4el BY a winner ter erie sited i A ee : rarest SiS ammabgeerstrr etre eeeU stvaweete® pabehcay sea Oe ps Bronte Bs PaPEe MGT SET TS ; peat eentriee: . epee arnaemdet ere ‘ eeu rock. nidaatiteresesbrbcerseeees tsi ae ROL y Nore aa +", Cenk vole) bey yer REMIT) DUR Ae ‘ if x i ye} : 4 *Teta® ; : ae: 203 } Y meee esihest ere! “ey este mapas ered teat ie sitar nnee aa Nae 5 ( AP KR A shee ae ie nies NOOO ‘ +4 y : b a tere ed 4 414 a8 r chee ‘ arc toes ‘ Rett f ; : poasGatete4-O.u sess essere) : r ; ass betn010-0-9 oor Te y wie 84 tee Leg Nee ee SY Jyateate aes bee Lorareseie Sortaceptirod wep epee aE NT Pe AnG C7090) vented tadetem etree rier OPI ee ED gicbetr tee, Creer e en | Cece eee at on ieee te aes + 4 aete Stet Oe ee v4 7 Ey ‘ + “ed oy s pihereueaierads teense ohn on 4 Pe Oe Pisictatiatiatsate at Ry Least ‘ ati ama tues doch lett eis r “ ees : ; Sree asl « DO a ae ee ee deere oe Or riet se) ‘ae Tee idee ek 14 ee etm he ve pu, v ba herbed ; ii tathted heuer ie x eit ULE Huon ines Hiab ee eect e 0 Sah a ae Hatentad . . ating i SONG “e its te or geet ms aialares Fett Saat at ior F , ‘ mK : + Sep iy rt a eyes 3 ataitiest LO ‘ . ahreria tres) sate nates CRUSE S HHO gia eres eccrine ieeat thet Oalitetnsb eee sisal sea tone feat ; Aly shane eae ' ABE i WY vi Hi ae i “a t fbeeeh Y turk Oto ey eR) 5 ries peer te ts atees? rob oadmee etre recess Sesa53 so ‘ yee det 435? Paes: Pe oe a re fr ee ars) 4 PPPS SM AKG aa RP I RL UOK ere ae Vig k eer tee k WME ON Me PWT GLE re a4 Graded 4 A RAY hoe ‘ alee > SOOO: ale oe) See i a mtcertiad ROOTES) 744 A abea’ DOM B ARIS ABM clea ‘ Was Satie Hie asseactasie ; é i ‘ ey ie G Patties aire Ata : ¢e a ‘) f LOR 6 ae * q jean + 4 ry re * ‘ i brs bese a & € fosnines rat aed Sh) ; y : IK is re Rte pik th i tag EM areca ace i ‘ : igs Cer iat ox } MIO eh t y ary wag fest) Teves 4 ever * 4 Pees ys ‘ AA) “J 42 boas ‘ a wera ait Ht pias? Fae as 4 eM) 4 i pieadat hae aot « ie che tes Fad baled ‘ Orr ier (Crear i vemeds ON ORO: OA tea Ciena ad < Va tas dens y Peoria ere ree ss Ie) Pore be terre PCIE» Pree eh i enn acct aaa tala d suas J 4 ae Tek e886 trate , 4 a4 8 3 < rs i f of . ange amyiee SS Sess epbopeer ort pare emepat at 9a ten 92%; . Depecnpiees eratertentannte aad ‘ Vigan 69a vee ‘ ieee ON ie ea at ele eat ever euat 4 ‘ rts ‘ hat 4 atdeean® 4 eK WR ton) Wasa @ tect * 4 ‘ Tes en ane M M , ene. ate poeeng tee rte a6 e-tiest ote HD soa -4-0— a Pe reed | ee ern) Oe EMT + Cech Utes Was 8, 4 Fa 8 r CO aah ey Ware eae a4 ore , SMITHSONIAN faite x , poten ; “. . rete shearers + On iakie: + Se cree eI ais . Arn . hey pee aS, x Be or i “etter : : aerertes tesa tere #9 *! biases tse ate fa se ay 6 J es : ve POST-LARVAL FISHES. BY — LONDON : ” Ben % Co., teen: aT, Soho Square, We; : AND AT THE 1916. [au right teeta ‘s Sirs LARVAL AND POST-LARVAL FISHES. BY C. TATE REGAN, M.A. (Assistant in the Department of Zoology, British Museum (Natural History)). WITH FIVE TEXT-FIGURES AND TEN PLATES. ].— Antarctic and Subantarctic Fishes. PAGE 1. Notolepis and Myctophumn . : : : : ; : 125 2. A new Paraliparis . : : ; ; : ; 129 3. The Nototheniform Fishes ; : ; ; ' r 129 Il.—Fishes from the Tropical and South Temperate Zones : 134 II].—Notes and Conclusions. 1. Pelagic Larvae in Relation to the Distribution of Species 148 2. The Recapitulation Theory : : : p 149 3. Development of the Fins . ; : : : : ; 150 4. Characteristic Features of Pelagic Larvae. F : : 151 5. Systematic Importance of Larval Characters : : ; 152 I—ANTARCTIC AND SUBANTARCTIC FISHES. 1. NOTOLEPIS AND MYCTOPHUM. THE pelagic fishes of the order Iniomi are represented by a very perfect post-larval example of the strictly Antaretic Notolepis coats and by a series illustrating the development of the more widely distributed Myctophum antarcticum. This ranges. when adult, throughout the Antarctic, Subantarctic, and South Temperate Zones, but the larval and post-larval specimens were all taken in the Subantarctic Zone, some to the south of New Zealand, others to the west of Tierra del Fuego, Notolepis coatsii, Dollo. (P1. I, figs. 4, 5). A post-larval example, 70 mm. long, is very similar to one figured by Roule (Deux. Expéd. Antarct. Francaise, Poissons, pl. IID). It was taken on Dec. 28th, 1912, at Station 269, 68° 37'S., 166° 14’ W., surface. The adipose fin is contmuous with VOL. I. T 126 “TERRA NOVA” EXPEDITION. the procurrent part of the caudal, and extends forward to the dorsal. The small pelvic fins are a little in advance of the vertical from the origin of the dorsal, and the anus is a short distance in front of them; from the anus a membranous fringe runs backwards to the anal fin. Fig. 1.—Notolepis coatsii. Type, actual size. Dr. W. 8. Bruce has very kindly lent me the type in order that the accompanying figure might be drawn, but owing to the condition of the specimen (cf. Tr. R. Soc. Edinburgh, XLIX, 1913, p. 233) this figure is largely a restoration. The “ Terra Nova” example agrees with the type in the number of myotomes (82) and fin-rays (8 dorsal and Fig. 2.—Distribution of Notolepis coatsii. All recorded captures are from the area bounded by the extreme limit of pack-ice, ~ ~ ~ ~~ ~—. Specimens taken by the “Terra Nova,” @ ; . . 9» oe Spy Es, *F)) ‘! py “Pourquoi Pas?” sf ; “Scotia,” 4 ; “ Challenger,” ©. 28 anal); but in the type, which is 105 mm. long, the anus is further back, below the anterior part of the dorsal fin, and there is a separate adipose fin. On the other hand, in a specimen of 50 mm., the anus is further forward, only a short distance behind LARVAL AND POST-LARVAL FISHES—REGAN. 1: b NI the head, so that the migration of the anus backwards during the development of this species is established. Moreover, it is evident that Prymnothonus is not a valid genus, but merely a larval form of Notolepis, Paralepis, ete. The capture of this example in the Ross Sea completes the evidence that Nofolepis coatsit 1s eircumpolar, for it had previously been taken near Peter Island, at the South Orkneys and in the Weddell Sea, and near Wilkes Land. Myctophum, sp. Some specimens, very much damaged, taken on March 27th, 1912, at Station Dac p2. lily Se 67> 25) EK, 30) metres: The largest, 15 mm. long, is very similar to the somewhat larger example of M. punctatum, Rafin. figured by Holt and Byrne (Fisheries Ireland Sci. Invest., 1910, VI, pl. I, fig. 1), but differs in that the anal papilla is separated by an interspace from the anal fin. I count 40 myotomes and 20 anal rays; the dorsal fin appears to have been bitten off. Myctophum antarcticum, Ginth. (Pl. I, figs. 1-3). A number of examples of this circumpolar species, 10 to 18 mm. in length, were taken in the Subantarctic Zone at Stations— 235) 52° 41'S., 168° 15' E., 10 metres, March 26th, 1912. 238 BP? GS, WBS Say IB. SO a March 27th, 1912. 240 HIS Bil" Shy WO Be Wy, 5p March 28th, 1912. 20a lea Seelane O Weysuttace Dees 20thel ilo: PRD 5A° 33'S., 176° 55! W., x4 Dec. 21st, 1912. 308 55° 29'S., 78°54’ W., 4 metres, April 9th, 1913. In adult examples, 60 to 100 mm. in length, I count 13 to 15 dorsal and 18 to 22 anal rays, and 40 to 42 myotomes. I find the same numbers in the larger specimens (14 to 18 mm.) of the ‘Terra Nova” collection, which have the fin-rays developed, but differ from the adult fish in the followme points :— (1) There are no scales or photophores, except a single photophore on each side above the base of the pelvic fin. (2) The dorsal and anal fins are lower, the caudal fin is less emarginate, the adipose fin is longer, and the paired fins are shorter. (3) The fish is more elongate, and the snout is proportionately longer, the eye smaller and the maxillary shorter, not reaching the vertical from the posterior edge of the eye. (4) There is a prominent anal papilla, and from it a membranous fringe runs forward to the base of the pelvie fins. (5) There is a small sinus, subdivided by septa, persistent above the occipital p 2 128 «TERRA NOVA” EXPEDITION: region of the head; this is an expansion of the anterior part of the dorsal fin-fold. (6) Pigment is wanting, except for a median black spot on the parietal region and some pigmented areas on the tail. In specimens of 14 mm. (and less) there is a dark area above and another below on the caudal peduncle ; in larger ones the dorsal area spreads forward on each side of the adipose fin, and the ventral one may disappear, or may he replaced by a median series of Fic. 3.—Distribution of Myctophum antarcticum. Recorded captures of the adult are marked + ; larval and post-larval stages taken by the Terra Nova,” JJ. The map shows the extreme limit of pack-ice = => ~—~-—, and the mean annual surface isotherms of 6 C.———-—-— and 12° CG. - dark spots, probably the precursors of the infracaudal plates of the adult female. Examples of 10 to 12 mm. are often more elongate, and have no photophores and no pelvic fins ; the adipose fin extends forward to the dorsal; the dorsal fin is low, and its rays are undeveloped or just evident. The maxillary reaches only to below the anterior part or middle of the eye, which is relatively larger than in the specimens of 14 to 18 mm., whilst the snout is correspondingly shorter. LARVAL AND POST-LARVAL FISHES—REGAN. 129 Myctophum antarcticum is very similar in form, proportions, position of the fins and number of rays, number of myotomes, ete., to the northern JZ glaciale, Reinh., the principal differences between them residing in the number and arrangement of the photophores. Holt (P.Z.8. 1898, pp. 552-560, pls. XLVI, XLVIT) has described young stages of the northern species, and it is interesting to note the resemblances and differences between these and corresponding stages of JL antarcticum. There is a general resemblance in the development of the two species, but J/ glaciale at 8 mm. long corresponds to M. antarcticum of 10 to 12mm. and at 11°5 mm. to IL antarcticum of 16 to 18 mm., whilst at 14°5 mm. JL glaciale has the general characters of the adult fish, except that the eye, although considerably larger than in smaller examples, is not yet proportionately as large as in the adult. Post-larval examples of J/. glaciale differ from corresponding stages of J. antarcticum in that they are not noticeably elongate in form as compared with the adult, the abdomen is prominent, the dorsal sinus extends from head to dorsal fin, and the latter is much higher than in M. antarcticum. Even if post-larval J. glaciale appear less elongate than the same stages of M. antarcticum to a great extent because the depth is increased by the large dorsal sinus and the protuberant abdomen, yet if these features be neglected, the southern species 1s at this early age noticeably more slender than the northern one. Jn both, the eyes at first grow at a slower rate than the fish as a whole, as is the rule in fishes generally, but later on grow faster than the rest of the fish, so that they are propor- tionately much larger in the adult than in the young. Holt’s explanation—that this is due to the fact that the adults live at greater depths than the youne—may be oD unreservedly accepted. 2. A NEW PARALIPARIS. Paraliparis terrac-novae, sp. nu. (Pl. I, fig. 6). A young fish, 35 mm. long, was taken in McMurdo Sound on Jan. 16th, 1912, at Station 332, 77° 15'S., 166° 0' E., 0-550 metres. It is very similar in most respects to P. antarcticus, Regan, but differs notably in the fewer fin-rays (dorsal 55; anal 43) and in the form of the pectoral fin, which has no elongate lower rays. This is the second Antarctic species of this genus. 3. THE NOTOTHENIIFORM FISHES. The “ Terra Nova” collection includes young stages of Pleuragramma, Pagetopsis and Chionodraco, from the Ross Sea and MeMurdo Sound. As I have already made a systematic revision of the Nototheniiformes and have given a general account of their distribution, it seemed worth while to try and complete this work by describing the egos and young of the whole group, so far as possible, irom the material in the British 130 “TERRA NOVA” EXPEDITION. Museum (Natural History), including some undescribed larvae from the ‘* Discovery ” collection. NOTOTHENIIDAE. Notothenia (Pl. II, fig. 2). DoD 1910, at Roy Cove, Falkland Islands, by Mr. Rupert Vallentin, evidently belong A mass of eggs and a number of newly hatched larvae taken on Sept. 21st, to a species of Notothenia. The eges have a diameter of about 1°5 mm., and the capsules adhere by facets to form a loose mass, just as in our northern Coftus scorpius. The newly hatched larvae have a length of about 6 mm., and in their general structure are very similar to Coftus larvae, for they have a short abdomen and a long tail, the mouth is well-developed, and the median fin extends from the head round the tail to the yolk-sac ; the anus is placed a short distance behind the yolk-sac, instead of at its posterior edge, as in Coffus, nor can I see an oil-globule in the yolk. Lénunbere has noted that in August .V. tessellata and .V. sima had the ovaries well developed, with eggs measuring 1 mm. or less; it is quite likely therefore that the eggs and larvae described above belong to one of these species or to another species of the tessellata group. A second mass, also presented by Mr. Vallentin, is very similar to the first, but has the eggs somewhat larger (diameter 1°7 mm.); it was found under a stone at the Falklands on May 10th, 1910. Of the Antarctic species of Notothenia Lonnberg has noted that iV. larseni and N. nudifrons were nearly ripe on April 19, and that in V. rossi and N. gibberifrons the ovaries were very small in May, and in .V. coriiceps in August. Late larval and early post-larval stages of Notothenia have not yet been found. Two young examples of V. macrocephala, 40 mm. long, were taken by the “ Challenger ” in the tow-net on Jan. 8th, 1874, off Kerguelen. These are very different in appearance from the adult fish, as they are bright silvery, with the back bluish. From this coloration and from their method of capture it may be concluded that the young of this species swim at the surface and that its wide distribution may be connected with this. In WV. eyaneobrancha, which is restricted to Kerguelen, young examples of 40 mm. have the mottled coloration of the adult. Trematomus. In examples of 7. bernacchii and T. hansoni taken by the ‘Southern Cross” and “ Discovery” I find that the genital glands are much larger in April than in October, Females taken in April have well-developed ovaries with eges 1 to 15 mm. in diameter. All the specimens that were preserved of those captured in traps during the winter are males, with testes nearly ripe ; it may be that the females cease feeding at the approach of the breeding season. It is probable that spawning takes place about July, and that the eggs are not much, if any, larger than in Notothenia. oD LARVAL AND POST-LARVAL FISHES—REGAN. 131 Pleuragramma antarcticum, Bouleng. (Pl. IL, figs. 3-6). Larval and post-larval examples were taken by the “ Discovery ” at Ross Island, at a depth of 6 to 10 fathoms. Date of Capture. Length. December 14th, 1902... -.. 6-7 mm. December 25th, 1902... 2) 8=10)mm: February 8th, 1904 ~ . O=05)mm: February 21st, 1902 ... ... 15-19 mm. April, 1903 Bor ... 15-25 mm. May 23rd, 1902 ee sco saab Tannin. August 13th, 1903 Sess ... 15-25 mm. September 18th, 1903... .. 20-25 mm. The “Terra Nova” also secured specimens from the Ross Sea and the coast of Victoria Land : Station. Date. Locality. Length. Near 186 December 31st, 1910 Ross Sea, 190 fathoms 30-35 mm. 325 August 8th, 1911 Cape Evans, 10 metres 16-18 mm. 326 January 9th, 1912 Terra Nova Bay, 10 metres = 10 mm. Sov January 22nd, 1912 Cape Bird, 80 metres 10 mm. 325 April 30th, 1912 Cape Evans, 10 metres 22-25 mm. From these data it seems probable that Pleuragramma may breed in the early Antarctic summer, that the eges may hatch out about the beginning of December, that the newly hatched larvae may be 6 mm. long or a little less, that by the winter they may reach a leneth of 15 to 25 mm., and when a year old may be as much as 35 mm. long. Larval and post-larval examples have been taken not far from the coast and at depths not exceeding 80 metres. Young fish (30 to 35 mm.) in the Ross Sea at a depth of 190 fathoms, and adults from the Ross Sea, 158 fathoms, from near the Balleny Islands, 254 fathoms, in addition to examples found frozen on the ice barrier and others taken from seals’ stomachs. Larvae of 6-7 mm. are very similar to those of Notothenia from the Falklands. Figures (Pl. Il) are given of examples 6, 11°5,13°5 and 25 mm. long. — It will be noted that the permanent caudal rays are making their appearance in the larva of 11 mm., and that in the larger ones they have assumed their final position. The differentiation of the dorsal and anal rays proceeds from behind forwards, and the full number is not developed until a length of 30 mm. is reached. Pectoral fins are present from the first, but the pelvics are rudimentary in fish of 25 to 35 mm. and are absent in smaller ones. The pigmentation at the bases of the vertical fins and on the dorsal surface of the abdominal cavity appears to be characteristic. In these larval and post-larval Plewragramma the eye is proportionately smaller than in the adult fish, no doubt because the latter descend to greater depths. 132 “TERRA NOVA” EXPEDITION. Artedidraco (P\. I, fig. 1). Examples of A. loennbergit, Roule, and A. skottsbergi, Lonnberg, taken by the “Terra Nova” in MeMurdo Sound, at a depth of 207 fathoms, on January 23rd, 1912, include nearly ripe females with eggs 25 to 8 mm. in diameter. Lonnberg has recorded nearly ripe females of A. mirus, with eggs of 25 mm., from South Georgia in May. On January 28th, 1904, the “ Discovery” obtained two larvae which I identify as A. skottsbergii, at Hut Point, Ross Island, at a depth of 3 fathoms. These are 13 mm. lone, and judging by the development of the vertical fins and the size of the yolk-sac they have not been hatched very long; one of them is figured (PI. II, fig. 1). The short tail and large yolk-sac make this larva quite unlike that of Pleuragramma in appearance, GYMNODRACONTIDAE. Gymnodraco acuticeps, Bouleng. (Pl. III, fig. 4). A post-larval fish 24 mm. long was taken by the “ Discovery” on Jan. 28th, 1904, at Ross Island, at a depth of 3 fathoms. The caudal and pectoral fins are fully developed, the pelvics are rudimentary, and the dorsal and anal rays are all present but do not reach the edge of the fin. A large yolk-sac is still evident. The snout is short and blunt as compared with that of the adult fish, and the oval nostril is relatively large; the opercular spme is not yet developed. CHAENICHTHYIDAE, The type of Chaenichthys rugosus, Regan, from Kerguelen, taken between October and February, is a ripe female with eggs 3 to 4mm. in diameter. Lonnhberg has examined a ripe female of Champsocephalus guinari, taken in May at South Georgia ; this had eges 4 mm. in diameter. Pagetopsis macropterus, Bouleng. (Pl. HT, figs. 1-3). Two larvae from the “Terra Nova” collection, taken at Cape Evans, MeMurdo Sound, may be referred to this species. The smaller, 14 mm. long, was captured on May 13th; the larger, 15 mm. long, on June 28th, 1911, at a depth of 20 metres. Two somewhat larger specimens, 19 and 20 mm. long, were obtained by the “ Discovery ” at Ross Island. The wide mouth, cleft to below or beyond the posterior edge of the eye, at once distinguishes these larvae from those of the Nototheniidae and Bathydraconidae ; another distinctive feature is the precocious development of the pelvic fins. The relatively short tail, the small number of myotomes, and the black colour of the pelvie fin membrane are characteristic of this species. The series illustrates well the growth of the produced snout of the adult fish from the snub-nosed form of the larva. LARVAL AND POST-LARVAL FISHES—REGAN. 133 Chionodraco kathleenae, Regan (PI. IV, figs. 2, 3). A larva of 21 mm. was taken by the “Terra Nova” off Cape Evans at a depth of 10 metres on October 19th, 1911. This is not quite so advanced as the young Pagetopsis of 19-20 mm. From these it is readily distinguished by the longer tail and the more numerous myotomes (about 60), whilst the pelvic fins are scarcely pigmented except between the spine and the first soft ray. A fish of 32 mm. from the “ Discovery ” collection is much more advanced, but the dorsal and anal rays are still undeveloped. Cryodraco (PI. Vee hie: 1). Three specimens, 16, 17 and 21 mm. in total length, from Ross Island ( Discovery ” collection) may belong to an undescribed species of this genus. The myotomes number 55 or 56, fewer than in C. atkinsoni, in which I count 62. The body is crossed by two dark bands, the anterior on myotomes 27 to 29, the posterior on myotomes 40 to 42; there are indications of a third band at the base of the caudal fin. The slender form, produced snout and elongate pelvic fins further distinguish these fishes from larval or post-larval Pagetopsis and Chionodraco. The data Nototheniiformes may be summarized as follows :-— D viven above with regard to the breeding and development of the Of the Bovichthyidae nothing is known, and examination of the material in the elves no results. Natural History Museum @ The breeding season varies considerably ; approximate dates are May and September for species of Notothenia, July for Trematomus hansoni and T. bernaechii, November for Pleuragramma antarcticum, January and May for species of Artedidraco, May for Champsocephalus qunnari, etc., ete. The eggs vary in diameter from 1 to 4 mm. when taken from ripe females preserved in spirit ; the diameter would no doubt be greater if eggs freshly taken from the sea were examined. The Nototheniinae seem to have the smallest eges, Artedidraco and the Chaenichthyidae the largest. In all probability the eges are demersal throughout the group, and in some species of Notothenia they are known to adhere together im masses. In some of the Nototheniinae the newly hatched larvae are about 6 mm. long, but in Artedidraco and the Chaenichthyidae they are probably twice as long or even more. The newly hatched larvae have a distinct mouth, the membranous median fin extends forward above to the head and below to the yolk-sac; the anus is situated at or a little behind the posterior end of the yolk-sac, and at the edge of the fin- membrane ; pectoral fins are present, but the pelvics do not appear until much later, except in the Chaenichthyidae, During growth the caudal is formed and assumes its VOL. I. U 134 “TERRA NOVA” EXPEDITION. terminal position before the dorsal and anal rays develop ; of these the posterior rays appear first, at any rate in Pleuragramma. Of the forms identified the larvae of the Nototheniinae differ from the rest in the small size of the yolk-sac. Of those with a large yolk-sac Artedidraco is distinguished from Gymnodraco by the short tail, whilst the Chaenichthyidae differ from both in the wide mouth and the early development of the pelvic fins.* Except Pleuragramma artarcticum larval and post-larval Notothenioids have only been taken quite near the coast. There is some evidence that the young of Notothenia macrocephala, a widely distributed species, may be pelagic. Il—FISHES FROM THE TROPICAL AND SOUTH TEMPERATE ZONES. Larval and post-larval fishes were taken, by means of plankton nets and the young fish trawl, at or near the surface in the following areas :— (1) North of New Zealand and round the Three Kings Islands. Stations 85-142. July to September, 1911. (2) Melbourne Harbour. Station 161. October, 1910. (3) Temperate South Atlantic, about 200 miles from the coast of Uruguay. Station 311. April, 1913. ) Off Rio de Janeiro. Stations 39-40, April, 1913. (5) Western Tropical Atlantic, south of the Equator. Stations 43-57. May, 1913. (6) Atlantic, south of the Canaries. Stations 16-17. June, 1910. (7) Atlantic, south of the Azores and west of the Canaries. Stations 68-69. May, L913. The nature of the collections made in these areas is shown by the following summary —— New ZeanaAnp ANd Turee Kinas Isnanps. Sardina neopilchardus Limnichthys fasciatus Prymnothonus, sp. Cubiceps caeruleus Myctophum coccot Centrolophus maoricus Diaphus, sp. Thyrsites atun Lampanyctus macropterus Lepidopus caudatus L. longipinnis, sp. n. Tripterygium varium Anguilla australis Monacanthus scaber Scombresox forsteri Diodon, sp. Scorpis violaceus Haplophryne mollis * Onos, Brosmius, Molva, etc., resemble the Chaenichthyidae in the early development of the pelvic fins, in this respect differing from Gadus (cf. Ehrenbaum, Nordisches Plankton, Eier und Larven von Fischen, 1905-1909). LARVAL AND POST-LARVAL FISHES—REGAN. ite Mrteourne Harsour. Odax balteatus Platycephalus, sp. Pentaroge marmorata TremMPERATE SoutH ATLANTIC. Stylophthalmus paradoaus M. laternatum Cyclothone microdon Ceratias, sp. Myctophum benoiti 10 DE JANEIRO. Sardinella pseudohispanica Ancylopsetta quadrocellata Cyclothone microdon Ancylopsetta, sp. Glyphidodon, sp. Symphurus plagqusia Gobiosoma molestum Tropica Sourm ATLANTIC. Stylophthalmus macrenteron, sp. 0. L. acuticeps, sp. n. Cyclothone microdon L. hexastiqgma, sp. n. Vineiquerria lucetia Hemirhamphus unifasciatus Synodus synodus Cryptotomus ustus Prymnothonus, spp. Scorpaena, sp. Lampanyctus maderensis Bothus ocellatus Leptocephalus muraenae unicoloris ATLANTIC, S. OF THE CANARIES. Lampadena chavesi ATLANTIC, S. OF THE AZORES. Vinciquerria lucetia Scombresox saurus Paralepis speciosus The majority of these were captured well out at sea, and are young stages of oceanic fishes, or oceanic larvae (Leptocephalus, Bothus ocellatus) of coast fishes. Exceptions to this are the larvae and young fishes taken in Melbourne Harbour and others captured a few miles off Rio de Janeiro (Sardinella, Glyphidodon, Gobiosoma, Ancylopsetta, Symphurus) ; also some of the New Zealand species were taken near the coast and may never be oceanic. Of some importance in its bearings on the geographical distribution of coast-fishes is the capture of young Lemirhamphus unifasciatus and Cryptotomus ustus, Brazilian species, far out in the Atlantic. Of greater interest is the case of Limnichthys fasciatus, known previously as a little fish of the rock-pools of New South Wales and Lord Howe Island ; now its range is extended to New Zealand, and its occurrence in localities so wide apart is explained by the capture of the young fish at or near the surface between the Three Kings Islands and New Zealand. ~ 136 «TERRA NOVA” EXPEDITION. ISOSPONDYLI. CLUPEIDAE. Sardine lla pseudohispanica, Poey. (El: We fic. Ny Typical Clupeoid larvae, 7 to 8 mm. long, with the anus far back, anal fin undeveloped, dorsal fin posterior, and caudal rayed and terminal. There are stellate chromatophores on the head, below the heart, and above and below the gut; they are most distinct on the dorsal border of the hinder two-thirds of the gut. Vertebrae 46. The number of vertebrae and the resemblance to the larval Pilchard (Sardina pilchardus) lead me to identify these larvae as Sardinella pseudohispanica, which is the Western Atlantic representative of the Mediterranean Allache (Sardinella aurita, Cuy. and Val.); the genera Sardina and Sardinella ave very closely related. Stations 39, 40. Six miles off the mouth of Rio de Janeiro Harbour. 2 metres. April 27th, 1913. Sardina neopilchardus, Steind. (Pl. V, figs. 3, 4). Three larval and post-larval fishes may be referred to this species. I count 16 dorsal and 16 anal rays and 52 or 53 myotomes. The smallest example, 12 mm. long, is more advanced than the 11°5 mm. larva of Sardina pilchardus figured hy Cunningham, as the caudal fin is fully formed and terminal im position, and the anal rays are appearing. The largest, 18 mm. long, appears to differ from Cunningham's 24mm. S. pilchardus chiefly m details of pigmentation, but after examination of Pilchard larvae from Plymouth, kindly lent by Dr. E. J. Allen, Tam doubtful whether there are any constant differences between the larvae of the Kuropean species and its representative in the seas of Australia and New Zealand. Station 135. Spirits Bay, near North Cape, New Zealand. 3 metres. Sept. Ist, 1911. STOMIATIDAE. Stylophthalmus macrenteron, sp. n. (Pl. V, fig. 1). ) A post-larval fish, 33 mm. long, shows several resemblances to Stylophthalmus paradoavus, Brauer, and may therefore be described as a Stylophthalinus, although it is unlikely that it is congeneric with any of the species associated under that name. In all probability it belongs to the family Stomiatidae ; the only other family that seems possible is the Alepocephalidae. If this be a Stomiatid it may represent a young stage of Lustomias obscurus, Vaillant, described from a single specimen taken near the Azores. Form elongate; head one-fifth of the length of the fish. Snout produced and depressed ; lower jaw prominent ; maxillary toothed, not nearly reaching eye; inter- orbital region broad and flat. Myotomes about 70. Dorsal 22, placed posteriorly. Anal 40, extending forward in advance of dorsal. Protruding terminal portion of LARVAL AND POST-LARVAL FISHES—REGAN. 137 intestine very long and supported behind by a cartilaginous rod. A series of seven dark spots along the back. Tropical Atlantic. Station 49. 18° 51'S., 33° 40’ W. Surface. May 6th, 1913. Stylophthalmus paradoaus, Brauer. A larva, 6 mm. long, of the type figured on pl. V, fig. 5, of the Valdivia report. South Atlantic. Station 311. 35°29'S., 50° 26’ W. 2 metres. April 22nd, 1913. GONOSTOMATIDAR, Cyclothone microdon, Giinth. (Pl. V, fig. 5). Several examples 6 to 8 mm. long have the fins fully developed, but the head is intermediate between that of the Gonostomatid larva figured by Holt and Byrne (Fisheries Ireland Sci. Invest. 1912, I, pl. II, fig. 5), and that of the adult fish, the maxillary not extending back beyond the eye. I count 12 dorsal and 16 anal rays and 30 myotomes. Tropical Atlantic. Station 39. 6 miles off mouth of Rio de Janeiro Harbour. 2 metres. April 27th, 1913. South Atlantic. Station 311. 35° 29'8., 50° 26’ W. 2 metres. April 22nd, 1913. Vineiguerria lucetia, Garm. (PI. V, figs. 6, 7). Seven examples, 8 to 10 mm. in total length, differ from the adult fish in that they are more slender and the photophores are not developed; also the mouth is smaller. A specimen of 15 mm. has the lower series of photophores developed, but the upper series is incomplete. All are from the Atlantic, taken at the surface. Station 45. 21°S., 37°50’ W. May 4th, 1913. Station 50. 18°S., 31° 45’ W. May 7th, 1913. Station 68. 27° 22’ N., 33° 40’ W. May 28th, 1913. INIOMI. SYNODONTIDAE. Synodus synodus, Linn. (Pl. VU, fig. 4). A larva of 14 mm. differs from those described by Max Weber (Siboga Fishes, p. 82, fig. 28) in having 13 patches of dark pigment on each side of the gut instead of 11. Probably this is a specific difference distinguishing the Atlantic S. synodus from the Indo-Pacific S. varius. I count 60 myotomes and 9 anal rays. Station 46. 20° 30’ S., 36°50’ W. Surface. May 4th, 1913. SUDIDAE. Prymnothonus, Richards. The only larvae known to belong to this family are of the type described as Prymnothonus ” (Giinther, “ Challenger” Pelagic Fishes, p. 39, pl. V, 1889). These are 138 “TERRA NOVA” EXPEDITION. larvae of Paralepis and related genera with produced snout. It has been shown above (p. 126) for the Antarctic Noftolepis coatsi/ that there is an extended backward migration of the anus during the transition from the Prymnothonus stage to the adult fish. A larva of 12 mm. has the snout moderately produced. There are 80 myotomes and about 20 anal rays. The anus corresponds to the twenty-eighth myotome and the origin of the anal fin to the fifty-fourth (Pl. VII, fig. 1). Tropical Atlantic. Station 50. 18°8., 31° 45’ W. Surface. May 7th, 1913. A second larva of 16 mm. has the snout more produced than the preceding. There are 116 vertebrae (52 + 64) and 30 or more anal rays (Pl. VII, fig. 2). Tropical Atlantic. Station 47. 20° 30’ S., 36° 30’ W. Surface. May 4th, 1913. In the number of anal rays these examples agree well enough with known species of Paralepis, and in the number of myotomes the first agrees with the Mediterranean species that I have examined. Possibly P. borealis, a species that I have not seen, may have the larger number of myotomes found in the second specimen. A third * Prymnothonus” is probably generically distinct from these ; 1t is a post- larval fish, 22 mm. long, evidently related to Paralepis, which it resembles in the structure of the head, but it has only 60 myotomes and 11 anal rays. The adipose fin is above the posterior end of the anal. Dorsal and pelvic fins are undeveloped GEV EE fies): Station 85. 24 miles W.N.W. from Cape Maria van Diemen, New Zealand. 2 metres. July 24th, 1911. Paralepis speciosus, Bellotti. Omosudis elongatus, Brauer, Valdivia Tiefsee Fische, p. 140, fig. 68 (1906). This species is represented in the British Museum collection by two examples of 65 and 75 mm. from Messina. Bellotti’s specimens were 75 and 90 mm., Brauer’s from 8 to 30 mm., the larger full grown. There is therefore reason to suppose that this is a small species, and that it assumes the adult form at an early age. A larva of 8 mm. that I refer, with some doubt, to this species has the fin-rays not yet developed, but the form of the head is already as in the adult fish. The patches of pigment on each side of the gut number only five, instead of eight, but this may be a larval character. Station 69. 29° 10’ N., 33° 36’ W. Surface. May 29th, 1913. This species was described from the Mediterranean (Bellotti, Atti. Soc. Ital. XX, 1877, fase. 1, p. 2, fig.), and has been recorded by Brauer from the Gulf of Guinea and the Indian Ocean. MycrorHIDAr. Larval and post-larval stages of Myctophum and related genera were taken to the north of New Zealand and in the Atlantic. The species of this group are so numerous LARVAL AND POST-LARVAL FISHES—REGAN. 139 and have such a wide range that the definite assignment of larvae to their species is very difficult. This difficulty is increased by the fact that the head, owing to the relatively smaller size of the eye and mouth and the greater leneth of the snout, has a physiognomy quite unlike that of the adult fish. However, by counting the myotomes and fin-rays and taking into consideration the position of the fins, the size of the mouth, ete., it is possible to make determinations which may, in some cases, approximate to the truth. The general character of the development has already been described in dealing with Myctophum antarcticum. Myctophum benoiti, Cocco (Pl. VI, figs. 1, 2). Several examples, from 4 to 7 mm. in total length, may belong to this species. They have much in common with the larvae described and figured by Holt and Byrne (Fisheries Ireland Sci. Invest. 1910, VI, p. 29, pl. I, fig. 8), from the Irish Atlantic slope under the name ‘‘ Scopelid larva, R 2,” but seem to be specifically distinct. They are distinguished by their rather deep form, conical snout, strong teeth, and by the distribution of the stellate pigment spots ; four large ones are present on each side, respectively at the origin of the dorsal and anal fins, below the adipose fin and above the end of the anal fin; usually there is a fifth on the side between the two last-named. There is also a spot on the back behind the head and a mid-ventral series of small spots from the end of the lower jaw to the origin of the anal fin. I count 12 or 13 dorsal and 17 or 18 anal rays and 37 myotomes. South Atlantic. Station 311. 35° 29'8., 50° 26’ W. 2 metres. April 22nd, 1913. Myctophum laternatum, Garm. (Pl. VI, fig. 7). A specimen 8 mm. long is probably of this species. It is moderately elongate ; the snout is short and the mouth is small, the maxillary extending to below the middle of the eye. The anal fin has 14 rays and commences below the posterior part of the dorsal, which has 11. The myotomes number 35. South Atlantic. Station 311. 35° 29'S., 50° 26’ W. 2 metres. April 22nd, 1913. Myctophum coccoi, Cocco. An example 20 mm. long has the characters of the adult fish, except that the eye is relatively smaller, its diameter being less than the length of the snout. Station 86. Off Three Kings Islands. 3 metres. July 25th, 1911. Diaphus sp. (Pl. VI, figs. 3, 4). Form rather deep; snout obtuse and mouth oblique. Length 4 to 5 mm. Dorsal and anal fins each with about 14 rays; about 35 myotomes. The distribution of the pigment is shown in the figures. Station 135. Spirits Bay, near North Cape, New Zealand. 3 metres. September Ist, 19L#. 140 “TERRA NOVA” EXPEDITION. Lampadena chavesi, Collett (Pl. VI, fig. 8). Four specimens, 10 to 12 mm. long, may belong to this species. I count 12-13 dorsal, 13-14 anal, and 15 pectoral rays and 38 myotomes. The fins are placed as in the adult fish. The maxillary extends to below the middle of the eye. On each side there is a photophore in front of the eye, another at the base of the pectoral fin, and a third above the base of the pelvic fin. Two examples have 2 mid-dorsal stellate blackish spots behind the adipose fin and a mid-ventral series of 3 or 4 linear spots behind the anal fin. Station 17. South of the Canaries. 26° 17’ N., 20° 54’ W. 10 metres. June 30th, LO LO. Lampanyctus maderensis, Lowe (Pl. VI, fig. 6). An example of 9 mm. shows the supraorbital ridge ending in an antrorse spine ; the cleft of the mouth extends to below the posterior edge of the eye. The myotomes number 36 and the fins are as in the adult fishes figured by Goode and Bean, and Brauer. There is a bar of pigment at the base of the caudal fin. In specimens of 5 mm. the spine is not developed, the mouth is smaller, and the adipose fin is longer. Station 50. Tropical Atlantic. 18°8., 31° 45'N. Surface. May 7th, 1913. Lampanyctus macropterus, Brauer (Pl. VI, fig. 5). A specimen of 10 mm. may belong to this species; it is very similar to the example of LZ. crocodilus of the same size figured by Holt and Byrne (Fisheries Ireland Sci. Invest., 1910, VI, pl. I, fig. 3). There are 12 dorsal and 20 anal rays and 40 myotomes. There is a black spot at the base of the caudal fin, another above the anal papilla, and some smaller ones on the lower part of the head. Station 135. Spirits Bay, near North Cape, New Zealand. 3 metres. September Ist, 1911. Lampanyetus longipinnis, sp. n. (Pl. VI, fig. 9). Dorsal 22. Anal 25. Myotomes 38. Evidently related to L. procerus, Brauer, differing in the more numerous dorsal and anal rays. A young fish of 15 mm. Station 113 (N.E. of Three Kings Islands). 33° 12’! 8., 171° 05’ E. 3 metres. August 9th, 1911. APODES. Following the plan adopted by workers on this group, I use the name Leptocephalus as a generic term for larval Eels, and give new specific names to those that cannot be identified. Leptocephalus acuticeps, sp. n. (Pl. VII, fig. 5). Very similar to L. orycephalus, Pappenh. (Deutsche Siidpolar Exped. XV, Zool. VII, LARVAL AND POST-LARVAL FISHES—REGAN. 141 p- 190, pl. IX, figs. 3, 5), from the Indian Ocean, but with fewer myotomes, 207 (174 + 33) instead of 220-230 (180-190 + 40). Form elongate ; head small: snout rather produced and acute. Teeth in 1 + 6 pairs in the upper jaw and 1 + 5 in the lower. Gut with a swelling (liver) at about the thirtieth myomere.