> ))) #5 ■)?' > , u) - >' >' ■>'>/ hfli >^»- : ::)>>i. ■> -^) n>M 'v^y »■ ,».r i!>)) ^ li))'- r»)> i> 'A» : "V» >jte ^>y> ^ j V#’’ ' J* r- ■ MEMOIRS OF THE CARNEGIE MUSEUM VOL. X W. J. HOLLAND, Editor PITTSBURGH Published by authority of Board of Trustees of Carnegie Institute DECEMBER 1922— JUNE 1925 CARNEGIE INSTITUTE PRESS PITTSBURGH, PA. / PREFATORY NOTE The Tenth Volume of the Memoirs of the Carnegie Museum contains four papers. The first is a list of the Fishes of Hawaii by David Starr Jordan and Eric Knight Jordan, his son; the second is a paper upon The Fishes collected in Ja])an by David S. Jordan, in the year 1922. In the preparation of this catalog Vlr. Carl Leavitt Hubbs collaborated with Dr. Jordan as did also Messrs. Ernest A. McGregor and M. Kasawa, the former aiding Dr. Jordan in the study of the Salmonidce, the latter in preparing the account of Netuma osakce. The third paper is from the pen of Mr. Charles W. Gilmore, giving a preliminary description of a singularly perfect specimen of a sauropod dinosaur, which Mr. Gilmore has provisionally identified as Camarasaurus lentus Marsh. The fourth paper is also from the pen of Mr. Gilmore, and contains an account of the skeletal remains of three Ornithopodous dinosaurs collected at the National Dinosaur Monument in Utah by Mr. Earl Douglass and his assistants. These specimens add materially to our knowledge of the osteology of the group of reptiles which they represent. It is a matter for congratulation that we have been able to bring these papers through the press with reasonable promptness. A number of other important papers, which relate to the great paleontological and zoological collections in the Carnegie Museum are in hand or under way, and we anticipate with pleasure their appearance, as they contain important contri- butions to our knowledge. Progress in science is gradually made. In spite of the somewhat slow nature of the process, we flatter ourselves that what has been accomplished by the Carnegie Museum, since the publication of its scientific papers was initiated more than twenty years ago, has abundantly justified this phase of our activities. It is interesting to one, who is familiar with the scientific literature of the present time, to see how constantly reference is being made in text-books, as well as in scientific journals, to the papers which have from time to time been published by this Museum. In fact few students of paleontology and zoology write today without making reference to the results of the researches, which we have been permitted to carry on. W. J. Holland, Editor. Carnegie Museum, May 29, 1925. iii ■^**■*151!, ftj ¥-S’ ^5^ M- 1)/-, : ^W‘ '' '''i t^%v;^';’\ ’, ■■ »■' v.-^.'r ■ \ ■ ■ , \' - r. ‘/.:'^ '-J ,.e:^.,4 '- ■■:> ■■ ,■, - - ■’■— .4-', • • r-'>m -2 r. ■' . ..h< ' ' ' : »• ■; ,:' . J- ’iViM-'-” . >> . • ■ - ' .’> ■’ . »-,' Vi . .v-?«i:,,:: • o*. ;■ .,-. . .■ ' , '■■' I*'-' ' ', ■, • , ' r.’ ■ ■ ■4 4'4;:;:/4,:, '> 4 .• . ^ ' ' ' * ' -'^■ ■'5: V-m ',•4; Si^.. ; ■'■ <> IVjft ' ' j, -■ii;*- ■• ' ' ;:;4i4iyS ■ M^ki. ■■ 'S Q .4 4-^ji I yrinSK2BK9[fi^> , V-T''"' ' , f-*' -r., '. ..1? -." ^, ' 'i. ‘V ■;• i'' "> L* ' ; '■' j «'. ; i ' ?-k* ^ «!' : ’ ’ r /i ’ ^ • ' ' « .' - .A-h . .’ill •',at i-v.« • ■ ■ J^f 1 •• . ,'J' «j "s ,0 , '*- , V' ' ' ' I ■■ ,'■ . -!< '■ /'A- M JJg '* •riZ 3 >\|J • V*.* ■■.V’ '.•“ll', . ■■..•’id ■•; 4 \ * 4- TABLE OF CONTENTS PAGES I. Title pages i-ii II. Prefatory Note iii III. Table of Contents v IV. List of Figures in Text vii-viii V. List of Plates ix-x VI. Genera, Species, and Varieties described in this volume as NEW TO SCIENCE, OR REDESCRIBED AND FIGURED xi-XV VII. Errata, Corrigenda, and Supplementary Notes xvi-xvii Memoir No. 1. A list of the Fishes of Hawaii, with Notes and Descrip- tions of New Species. By David Starr Jordan and Eric Knight Jordan. (Plates I-IV) 1-92 Memoir No. 2. Record of Fishes Obtained by David Starr Jordan in Japan, 1922. By David S. Jordan and Carl Leavitt Hubbs (The Salmonidae by David Starr Jordan and Ernest Alexander McGregor) . (Plates V-XII) 93-346 Memoir No. 3. A Nearly Complete Articulated Skeleton of Camara- saiirus, a Saurischian Dinosaur from the Dinosaur National Monument, Utah. By Charles W. Gilmore. (Plates XIII-XVII) 347-384 Memoir No. 4. Osteology of Ornithopodous Dinosaurs from the Dino- saur National Monument, Utah. By Charles W. Gil- more. (Plate XVIII) 385-410 Index 411-449 V LIST OF FIGURES IN TEXT. Memoir No. 1. FIGURE PAGE 1. Vesposus egregius Jordan 23 2. Istiophorus gladius (Broiissonet) 31 3. Rhyacanthias carlsmithi Jordan 46 4. Peristedion engyceros Gunther 56 5. Peristedion gilberti Jordan. . 57 6. Roa {Loa) excelsa Jordan . 60 7. Masturus lanceolatus (Lienard) 88 Memoir No. 2. 1. Sebastodes thompsoni Jordan and Hubbs 266 Memoir No. 3. 1. Skull and lower jaws of Camarasaurus lentus Marsh. Viewed from left side 354 2. Skull of Do. Superior view 356 3. Skull and lower jaws of Do. Posterior view 359 4. Skull and neck of Do. Viewed from left side 361 5. Right sternal plate of Do. Superior view 377 Memoir No. 4. 1. Left scapula and coracoid of Camptosaurus 7nedius Marsh 390 2. Left fore limb and foot of Do 391 3. Skull of Dryosaurus altus Marsh 395 4. Anterior dorsal vertebrae with articulated ribs of Do 399 5. Posterior dorsal vertebrae, sacrum, and pelvis with articulated ribs of Do. . 399 6. Right ilium, vertebrae, and ossified tendons of Do 401 7. Reconstructed skull of Laosaurus gracilis Marsh 405 8. Pectoral arch and humeri of Do 407 vii LIST OF PLATES PLATE 1. Alhula virgata Jordan and Jordan, tijpe. Myctophum hollandi Jordan and Jordan, type; Physiculus grinnelli Jordan and Jordan, type. IL Eumegistus illustris Jordan and Jordan, type; Scapterias fragilis Jordan and Jordan, type. III. Centropyge potteri (Jordan and Metz) adult male; Centropyge tutuilce Jordan and Jordan, type; Cheilinus himaculatus Cuvier and Valen- ciennes. IV. Calliurichthys astrinius Jordan and Jordan, type; Calliurichthys zanectes Jordan and Jordan; Canthider7nis angulosus (Quoy and Gaimard). V. Psychichthys eidolon Jordan and Hubbs, type; Oncorhynchus adonis Jordan and McGregor, type; Oncorhynchus kawamurce Jordan and McGregor, type. VI. Oncorhynchus ishikawce Jordan and McGregor, type; Oncorhynchus macrostomus (Gunther); Oncorhynchus macrostomus (Gunther). VIL Oncorhynchus rhodurus Jordan and McGregor, type; Salvelinus pluvius (Hilgendorf) ; Salvelinus imbrius Jordan and McGregor, type. VIII. Scales of Japanese Salmonidae. IX. Netuma osakce Jordan and Kasawa, type; Gnathopogon majimce Jordan and Hubbs, type; Belligobio eristigma Jordan and Hubbs, type; Ocycrius japonicus (Doderlein) . X. Liopempheris sasakii Jordan and Hubbs, type; Malakichthys wakiyce Jordan and Hubbs, type; Brachirus bellus Jordan and Hubbs, type. XI. Iburiella kasawce Jordan and Hubbs, type; Encceura evides Jordan and Hubbs, type; Zalescopus tosce Jordan and Hubbs, type. XH. Zestichthys tanakce Jordan and Hubbs, type; Allolepis hollandi Jordan and Hubbs, type; Monomitopus kumce Jordan and Hubbs, type. XHI. Photograph of skeleton of Camarasaurus lentus (Marsh) in position as found. XIV. Drawing of skeleton of Camarasaurus lentus (Marsh) designed as a key to Plate XIH. IX X MEMOIRS OF THE CARNEGIE MUSEUM. XV. XVI. XVII. XVIII. Photograph of skeleton of Carnarasaurus lentus (Marsh) as mounted and displayed in the Carnegie Museum. Skull of Carnarasaurus lentus (Marsh). Two-thirds natural size. Restoration of Skeleton of Carnarasaurus lentus (Marsh). Drawn by Sidney Prentice under the direction of C. W. Gilmore. Skeleton of Carnptosaurus medius Marsh. GENERA, SPECIES, AND VARIETIES DESCRIBED AS NEW TO SCIENCE OR REDESCRIBED AND FIGURED IN THIS VOLUME. REPTILIA (Fossilia). Order DINOSAURIA. Suborder SAUROPODA. PAGE Camarasaurus lentus (Marsh) description of nearly complete articulated skeleton 347-384 Suborder ORNITHOPODA. Ca7nptosauriis medius Marsh 385-393 Dryosaurus altus Marsh 394-402 Laosaurus gracilis Marsh 403-409 Family DASYATID^: Family CHIMHIRID^: Family ALBULIDHl: Family SALMONIDdE: PISCES (Viventes). PAGE Dasyatis ushiei Jordan and Hubbs, sp. nov. . 114 Psychichthys eidolon, Jordan and IIul)bs, sp. nov. PL V, fig. 1 117 Phasmichthys Jordan and Hubbs, gen. nov., (type Chimocra mitsukurii Jordan and Snyder) 119 Albula virgata Jordan and Jordan, sp. nov. PL I, fig. 1 G Oncorhijnchus adonis Jordan and McGregor, sp. nov. PL V, fig. 2; VIII, fig. 4. . . . 127 Oncorhynchus kaivmnurce Jordan and McGregor, sp. nov. PL V, fig. 3; VIII, fig. 5 128 Oncorhijnchus ishikawcc Jordan and Mc- Gregor, sp. nov. PL VI, fig. 1; VIH, fig. 6 132 XI Xll MEMOIRS OF THE CARNEGIE MUSEUM. Oncorhynchus rhodurus Jordan and Mc- Gregor, sp. nov. PI. VII, fig. 1 ; VIII, figs. 1-2 137 Salvelinus pluvius (Hilgendorf). PL VII, fig. 1 141 Salvelinus hnhrius Jordan and McGregor, sp. nov. PI. VII, fig. 3; VIII, fig. 15. 142 Family CONGRID^: Anago Jordan and Hubbs, gen. nov., (Type Coviger anago Temminck and Schlegel. 193 Congriscus Jordan and Hubbs, gen. nov. (Type Congermurcena megastoma Glin- ther) 193 Astroconger Jordan and Hubbs, gen. nov., (Type Anguilla myriaster Brevoort) ... 194 Alloconger Jordan and Hubbs, gen. nov. (Type Leptocephalus flavirostris Snyder) . . . 195 Rhynchocymba Jordan and Hubbs, gen. nov. (Type Leptocephalus nystromi Jordan and Snyder) 195 Rhynchoconger Jordan and Hubbs, gen. nov. (Type Leptocephalus ectenurus Jordan and R. E. Richardson) 196 Congrina Jordan and Hubbs, gen. nov. (Type Co^igermurcena cequorea Gilbert and Cramer) 196 Family CYPRINIDTl; Gnathopogon suwce Jordan and Hubbs, sp. nov 166 Gnathopogon majimcB Jordan and Hubbs, sp. nov. PI. IX, fig. 2 167 Gnathopogon longifilis Jordan and Hubbs, sp. nov 169 Gnathopogon tsuchigce Jordan and Hubbs, sp. nov 170 Belligobio Jordan and Hubbs, gen. nov. (Type Belligobio eristicjma Jordan and Hubbs, sp. nov. PI. IX, fig. 3) . . 172-173 MEMOIRS OF THE CARNEGIE MUSEUM. Xlll Family ARIID^: Family MYCTOPHID^: Family GADIDiE; Family SCOMBRID.E; Family CYBRID^; Family GEMPYLIDiE: Family BRAMIDiE: Family CARANGID^: Family THUNNIDtE: Family CENTROLOPHID.E: Sarcocheilichthys morii Jordan and Hubbs, sp. nov 175 Acahara Jordan and Hubbs, gen. nov. (Type Richardsonius semotilus Jordan and Starks) 177 Moroco Jordan and Hubbs, gen. nov. (Type Pseudaspius hergi Jordan and Metz) 180 Moroco yamamotis Jordan and Hubbs, sp. nov 182 Netuma osakce Jordan and Kasawa, sp. nov. PL IX, fig. 1 157 Myctophum hollandi Jordan and Jordan, sp. nov. PI. I, fig. 2 11 Lamprossa Jordan and Hubbs, gen. nov. (Type Diaphus anteorbitalis Gilbert) . . 156 Pantophos Jordan and Hubbs, gen. nov. (Type Diaphus glandulifer Gilbert) 156-157 Physiculus grinnelli Jordan and Jordan, sp. nov. PI. I, fig. 3 22 Pneumatophorus peruanus Jordan and Hubbs, sp. nov 211 Sawara Jordan and Hubbs, gen. nov. (Type Cyhium niphonium Cuvier and Valenciennes 214 Ruvettus pacificus Jordan and Jordan, sp. nov 24 Eumegistus Jordan and Jordan, gen. nov. (Type Eumegistus illustris Jordan and Jordan, sp. nov. PI. II, fig. 7 35-36 Atule Jordan and Jordan, gen. nov. (Type Caranx affinis Rlippell) 38 Caranx jordani sp. nov., (Nichols MS). ... 40 Kishinoella Jordan and Hubbs, gen. nov. (Type Thunnus rarus Kishinouye) .... 219 Ocycrius Jordan and Hubbs, gen. nov. (Type Centrolophus japonicus Doder- lein) 229 XIV MEMOIRS OF THE CARNEGIE MUSEUM. Family APOGOGONID.E: Family OLIGORID^: Family PEMPHERIDRi:: Family CH.ETODONTIDyE: Family SCORPR^NID^E: Family PLATYCEPHALIDyE : Family AGON ID/E: Scepterias Jordan and Jordan, gen. nov. (Type Scapterias fragilis Jordan and Jordan, sp. nov. PL II, fig. 2) 44-45 Malakichthys ivakiyce Jordan and Hubbs, sp. nov. PL X, fig. 2 233 Liopempheris sasakii Jordan and Hubbs, sp. nov. PL X, fig. 1 228 Tifia Jordan, gen. nov. (Type Chmtodon corallicola Snyder) GO Loa Jordan, gen. nov., since changed to Roa Jordan (Copeia, May 20, 1923, p. 63) 61 and 252 C entropy ge tutuilce Jordan and Jordan, sp. nov., PL III, fig. 2 62 CeMropyge potteri (Jordan and Metz), fig’d PL III, fig. 1 62 Sebastocles Jordan and LIubbs, subgen. nov. (Type Sebastes elegans Stein- dachner) 260 Sebastocles {Sebastosomus) thoinpsoni Jordan and Hubbs, sp. nov 265 Brachirus bellus Jordan and Hubbs, s}!. nov. PL X, fig. 3 274 Wakiyus Jordan and Hulibs, gen. nov. (Type Flatycephalus crocodilus Tilesius) 286 Ratabulus Jordan and Hubbs, gen. nov. (Type Thysanophrys megacephalus Tanaka ' . . 286 Occella Jordan and Hubbs, gen. nov. (Type Agonus dodeca'Alron Tilesius. ... 291 Iburiella Jordan and Hubbs, gen. nov. (Type Iburiella kasavjcB Jordan and Hubbs, sp. nov. PL XI, fig. 1) . . 290-291 Iburina Jordan and Hubbs, gen. nov. (Type Occa iburia Jordan and Starks) 290-291 MEMOIRS OF THE CARNEGIE MUSEUM. XV Family ELEOTRIDvE: Family GOBIOIDID.E; Family LABRID^: Family CALLIONYMIDyE: Family URANOSCOPID.E: Family BLENNIDAH: Family OSTRACIIDA5: Family SOARCID.E: Family BROTULIDvE; Family CORYPHAHNOIDIDvE: Encaeura Jordan and Hnbbs, gen. nov. (Type Encceura evides Jordan and Hnbbs, sp. nov. PL XI, fig. 4) 303 Tcenioides snyderi, Jordan and Hnbbs, sp. nov 310 Hinalea Jordan and Jordan, gen. nov. (Type JuUs axillaris Quoy and Gaimard) 69 Calliurichthys astrinius Jordan and Jordan, sp. nov. PI. IV, fig. 1 80 CalliuricJdhys zanectes Jordan and Jordan, sp. nov. PI. IV, fig. 2 81 Zalescopus Jordan and Hubbs, gen. nov. (Type Zalescopus toscx Jordan and Hnbbs, sp. nov. PL XI, fig. 3) 312 Zalescopus satsurnce Jordan and Hubbs, sp. nov 313 Dasson Jordan and Hubbs, gen. nov. (Type Aspidontus trossulus Jordan and Snyder) 318 Oncesthes Jordan and Hubbs, gen. nov. (Type Fetroscirtes fluctuans Weber) ... 319 Triorus Jordan and Hubbs, gen. nov. (Type Lactophrys tritropis Snyder = Ostracion stellifer Bloch).- 256 Zestichthys Jordan and Hubbs, gen. nov. (Type Zestichthys tanakce Jordan and Hubbs, sp. nov. PL XH, fig. 1) 321 Allolepis Jordan and Hubbs, gen. nov. (l^pe Allolepis hollandi sp. nov. PL XII, fig. 2) 322-323 Alonomitopys kwmcB Jordan and Hubbs, sp. nov. PL XH, fig. 3) 324 Ccelorhynchus gilberti Jordan and Hubbs, sp. nov. 327 ERRATA, CORRIGENDA, AND SUPPLEMENTARY NOTES. P. 30, for “TETRAPTERUS Agassiz” read “TETRAPTURUS Rafinesque.” P. 39, NOTE: “The Ulua should perhaps stand as Caranx bixanthopterus Riip- pell until Caranx melampygus is certainly identified.” D. S. Jordan. P. 40, NOTE: “The species Caranx cguara and C. cheilio belong to an unnamed subgenus, the name Selenia being preoccupied. In Uraspis the strong spines on the tail are turned forward.” D. S. Jordan. P. 46, NOTE: “The genus Rhyacanthias is apparently not distinct from Symphy- sanodon Bleeker, well figured but wrongly placed. S. carlsinithi is apparently distinct from S. typus.” D. S. Jordan. P. 107, Footnote 13: Delete the period after “Natal”. P. Ill, Lines 18 and 19 from top: for “paratype” and “paratypes” read specimen and specimens. P. 114, Line 3 from top: Delete the “2” near end of line. P. 150, Line 25 from top: for “4-3” read 4+3. P. 157, Line 21 from top: insert the word “oriental” between the words only and species. P. 161, Through an unfortunate oversight Hemigrarnmocypris rasborella appears twice, once on page 161 and again on page 189. P. 168, 18th line from top: for “2.7” read 2,7. 19th line from top: for “2.6” read 2,6. P. 169, Line 25 from top: for “cayiiga” read heterolepis. Bottom line: for “line” read limb. P. 170, 5th line from top: for “2.7” read 2,7; for “2.6” read 2,6. P. 17+ 3d line from top: for “2.7” read 2,7. P. 179, First line: for “Awaya” read Iwate (Awai). P. 182, nth line from top: for “three” read four. P. 191, 2d line from bottom: for “pigmented” read pigment. P. 192, 1st line: for “pigmented” read pigment. P. 192, Under ee 5th line: for “its posterior” read posteriorly. P. 200, 21st line from top: place comma after “land” and transpose “G. loricatus,” to fall between the words “land” and the word “and.” XVI MEMOIRS OF THE CARNEGIE MUSEUM. XVll P. 201, At line 22: Alter the sentence to read as follows: The fully armed form of the Western Atlantic (the ‘‘biaculeatus” of most authors) is scarcely different from typical aculeatus, though it shows on the average an approach toward its partially plated fresh- water derivative {cuvieri), which has a relatively slender body and long spines. P. 202, 8th line from bottom: for “ Pungitius” read pungitius, P. 208, 17th line from top: for “37 or 41” read 37 to-41. P. 213, 3d line from top: for ^‘Thyrsion” read Thyrsio. P. 226, 13th line from top: for “12-15” read 12+15. P. 227, 11th line from bottom: for “VI, II” read VI, 11. P. 231, 8th line from bottom: for “J. bodps” read S. bobps. P. 243, 12th line from top: for “Shisuoka” read Shizuoka. P. 253, 15th line from top: insert “spots,” after “five” at end of line. P. 256, 19th line from top: for “ Tetrasornus” read Tetrosomus. P. 258, 3d line from bottom: insert “by” before “Abbott.” P. 259, 15th line from top: insert vermicularis after “Sphoeroides.” P. 284, 5th line from bottom: for “raker” read rakers. P. 285, 10th line from bottom: transpose “ (On+ocwncc, subfam. nov.)” to 9th line, after “c.” P. 286, 18th line from top: for “infra” read infraorbital. P. 290, 11th line from bottom: for “Ocella” read Occella. P. 298, 7th line from top: for Cleisthenes pinetorum Jordan and Starks” read Cleisthenes herzensteini (Schmidt). P. 307, 5th line from bottom: for “castanaea” read castanea. Publications of the Carnegie Museum, Serial No. 115 MEMOIKS OF THE OAENEGIE MUSEUM VOL. X NO. 1 W. J. HOLLAND, Editor A LIST OF THE FISHES OF HAWAII, WITH NOTES AND DESCEIPTIONS OF NEW SPECIES BY DAVID STARR JORDAN and ERIC KNIGHT JORDAN PITTSBURGH Published by the Authority of the Board of Trustees op the CAENEGIE INSTITUTE ■ FEB " B ]-23 ■ ^4^/; December, 1922 PRICE LIST OF PUBLICATIONS OF THE CARNEGIE MUSEUM ANNUAL REPORTS OF THE DIRECTOR 189.8', 30e. (scarce); 1899, 25e.; 1900, 30e. (scarce); 1901-22, 25c. each. REPORTS OF PROCEEDINGS OF FOUNDER’S DAY 1898-1922, 35e. each. ANNALS OF THE CARNEGIE MUSEUM The Annals are supplied to those who subscribe in advance in parts (paper-bound), as published, @ $3.50 per volume; Vols. I-XIV, 1901-1922, bound in green cloth @ $4.00; bound in i Morocco @ $4.50. MEMOIRS OF THE CARNEGIE MUSEUM The Memoirs are supplied to those who subscribe in advan ce in parts (paper-bound), as published, at $10 per volume; bound in green buckram at $10.75, and in half -morocco at $12.00. VOL. I. (1901-3) No. 1. Diplodocus, Its Osteology, Taxonomy, etc. No. 3. Osteology of the Steganopodes. Shufeldt $2.76 Hatcher ... $3.00 “ 4. Classification of Chalcid Flies. Ashmbad.. 6.50 ‘ ‘ 2. Oligocene Canidse. Hatcher 1.60 VOL. II. (1904^6) No. 1. Osteology of Haplocanthosaurus, etc. Hatcher $2.00 “ 6. Osteology of Diplodocus. Holland 1.60 “ 2. Osteology of Baptanodon. Gilmore 1.50 “ 7. Osteology of Frotostega. Wieland ....... .76 “ 3. FossilAvian Remains from Armissan. East- “ 8. New Suilline Remains from the Miocene of MAN 1.00 Nebraska. Peterson 76 “ 4. New Rodents and Discussion of Origin of 9. Notes on the Osteology of Baptanodon, etc. Dsemonelix. Peterson 1.76 Gilmore 1.0§ No. 5. Tertiary of Montana. Douglass $1.00 “ 10. The Crawfishes of Pennsylvania. Ortmann 4.00 VOL. m. No. 1. Archaeological Investigations in Costa Rica. No. 2. Osteology of Moropus. Holland and Peter- Hartman $6.00 son $6.00 VOL. IV. No. 1. Early Chinese Writing. Chalfant $3.00 No. 5. New Carnivores from the Miocene of West- ‘ ‘ 2. Formosan Fishes. Jordan 26 ern Nebraska. Peterson $1.60 “ 3. Entelodontidae. Peterson 2.60 “ 6. Monograph of the Najades of Pennsylvania. ‘ ‘ 4. Fishes of Formosa. Jordan and Richard- Pts. I and II. Ortmann 2.60 son 1.26 “ 7. Catalog Eocene Fishes from Monte Bolca in Carnegie Museum. Eastman 3,00 VOL. V. The Fresh Water Fishes of British Guiana. Eigenmann. $10 unbound; $10.75 cloth; $12.00 i morocco, VOL. VI. No. 1. Fishes from Korea. Jordan and Metz $1.60 No. 4. Fishes obtained in Japan. 1911. Jordan “ 2. Lantern-fishes of Japan. Gilbert l.CO and Thompson $3.60 “ 3. Gymnotid Eels of Tropical America. Max Nos. 5-7. Fossil Fishes in the Carnegie Museum. Mapes Ellis 1.50 Parts II-IV, Eastman 6.00 vn. No. 1. The Cheirodontinm. Eigenmann $3.60 No. 4. PimelodeUa and Tsiphlobagrus. Eigenmann. 2,00 ‘ ‘ 2. Turtles of Uinta Formation. Gilmore .... 2.00 No. 5, The Pygidildae. Eigenmann 2.50 No. 3, Ophidia from S. America. Griffin 2.00 No. 6. American Diceratheres. Peterson 2.25 VIII. No. 1. Monograph of the Naiades of Pennsylvania. Part III. Ortmann $5.00 No. 1. The Fishes of Western South America. Part I, Eigenmann $15.00 IX No. 2. Fossil Plants from Montana.. Jennings... $3.50 No. 3. South American Naiades. Ortmann 5.00 % No. 2. Neotropical Tingitidse. Drake... 2.00 No. 1. A List of the Fishes of Hawaii, with De- scription of New Species 4.00 REPRINTS FROM THE ANNALS OF THE CARNEGIE MUSEUM Nos, 1-59. See preceding lists. {Mostly out of print.) MEMOIKS OF THE CAENEGIE MUSEUM VoL. X. No. ]. A LIST OF THE FISHES OF HAWAII, WITH NOTES AND DESCRIPTIONS OF NEW SPECIES. By David Stake Jordan and Eric Knight Jordan. (Plates I-IV). The senior author of this paper spent most of the month of August, 1921, at Honolulu in attendance upon the Pan-Pacific Educational Conference. While there, he gave all available time to making collections of fishes, having the efficient assistance of Mr. Fordyce Grinnell, Jr., a former student of his, who visited the markets daily. The collections made have been distributed among a number of museums. The CarangidcB have been sent to the American Museum of Natural History in New York to be used in a monograph of that group by Mr. John Tread- well Nichols; the types of new species have been sent to the Carnegie Museum in Pittsburgh; and series of other species, more or less complete, to the American Museum of Natural History, the Carnegie Museum, the Southwestern Museum at Los Angeles, and to the Universities of Michigan and of Iowa. In addition to notes on new and rare forms we have given a complete list of the species thus far known from Hawaii, the whole serving as a revision of the two memoirs on the Aquatic Resources of the Hawaiian Islands, published bj^ the United States Fish Commission in 1905, the first by David Starr Jordan and Barton Warren Evermann on the survey of the shore-fishes made in 1901.; * the other by Charles Henry Gilbert on the deep-sea forms taken by the “Albatross” in 1902. t *Bull. U. S. Fish Comm., 1903 (1905), Vol. XXIII, Pt. I, pp. 1-574. \ Ibidem, Pt. II, pp. 575-713. 1 2 MEMOIRS OF THE CARNEGIE MUSEUM. In the memoir by Jordan and Evermann will be found an account of the earlier explorations of the islands, as well as a detailed statement of the character of the fish-fauna and its relation to that of the South Seas. Descriptions of species, notes on habits, and references to synonymy, where accurately given by Jordan and Evermann or by Gilbert, are in general not repeated in the following paper. The student who is using the present list is presumed to have the other two lists at hand, and references to their pages are given throughout this list. The principal articles upon the Hawaiian fish-fauna, published since the two above-named papers appeared, are- the following: 1. Jordan (David Starr) and Seale (Alvin) — “The Fishes of Samoa, with a Check- list of the Fishes of Oceania published by the United States Bureau of Fish- eries, 1906. 2. Bryan (William Alanson) — “Three New Hawaiian Fishes N Occasional Papers « of the Bernice Pauahi Bishop Museum, II, 1906. 3. Jordan and Snyder (John Otterbein) — “Notes- on Fishes of Hawaii, with De- scriptions of New Species.” Bulletin of the United States Bureau of Fisheries for 1906 (1907). 4. Gilbert (Charles Henry) — “The Lantern-fishes.” Memoirs Museum Compara- tive Zoology, XXVI, 1908. 5. Jordan and Dickerson (Alary Cynthia) — “On a Collection of Fishes from Fiji, with Notes on Certain Hawaiian Fishes.” Proceedings U. S. National Mu- seum, XXXIV, 1908. 6. Jordan and Aletz .(Charles William) — “Descriptions of Two New Species of Fishes from Honolulu, Hawaii.” Proceedings U. S. National Museum, XLII, 1912. 7. Jordan (David Starr) — “ Descriptio7i of Deep-sea Fishes from the Coast of Hawaii Killed by a Lava-flow fro7n Mauna Loa.” Proceedings U. S. National Mu- seum, FIX, 1921. JORDAN AND JORDAN: FISHES OF HAWAII. 3 THE PISHES OF HAWAII. Class LEPTOCARDII. Order CIRROSTOMI. Family I. BRANCHIOSTOMID^ (The Lancelets). Amphioxides Gill. 1. Amphioxides pelagicus (Gunther). (J. & E., p. 33.) Pelagic. This diminutive lancelet, supposed to be distinguished by the absence of buccal cirri and by its pelagic habit, is now regarded as a larval form. The supposed genus is nearer Branchiostoma than Epigonichthys (Asymmetron) , the only other genus of this family as yet found in the open Pacific. Class ELASMOBRANCHII. ‘ Order ASTEROSPONDYLI. Family II. SCYLLIORHINIDvE. Apristurus Garman. 2. Apristurus spongiceps (Gilbert). (Gilbert, p. 579.) Deep seas. This species is referred by Garman to Pristiurus, which genus is characterized by a row of prickly scutes along the upper side of the tail. These are not present in Gilbert’s type, though perhaps they may have been lost in the dredge. A cast in the Bishop Museum may belong to this species. Color plain light brown; dorsal fins small, subequal, the first slightly in advance of ventrals. The name Catulus is preoccupied in the Insecta. Family III. GALEORHINID^. {Carcliarhinidce of authors.) Galeorhinus Blainville. {Eugaleus Gill.) Amid the uncertainties regarding the application of the generic names Galeus Rafinesque and Carcharias Cuvier, we here follow the decision of the Interna- tional Commission of Nomenclature. 3. Galeorhinus japonicus (Muller and Henle). (J. & E., p. 36.) Recorded from Laysan by Steindachner. Not rare in Japan. 4 MEMOIRS OF THE CARNEGIE MUSEUM. Galeocerdo Mliller and Henle. 4. Galeocerdo tigrinus Muller and Henle. (J. & E., p. 36.) Rare. Taken once in Honolulu. Prionace Cantor. {Prionodon Muller and Henle, preoccupied; Cynocephalus (Klein) Gill.) 5. Prionace glauca (Linna3us). (J. & E., p. 37.) Occasional in Japan. A, cast in the Bishop Museum shows the pectoral fin rather longer than in the Atlantic P. glauca. A specimen taken by the “Albatross” agrees with this, the pectoral being 4.5 in total length, instead of 6.25. Carcharinus Blainville. {Eulamia Gill.) 6. Carcharinus melanopterus (Quoy and Gaimard). Mano. (J. & E., p. 38.) A fine cast of this common species is in the Bishop Museum. 7. Carcharinus phorcys (Jordan and Evermann). (J. & E., p. 39.) Occasional about Hawaii. 8. Carcharinus insularum (Snyder). (J. & E., p. 40.) Rather rare. 9. Carcharinus nesiotes (Snyder). (J. & E., p. 40.) Common about Hawaii. Family lY. SPHYRNIDHl (Hammer-head Sharks). Sphyrna Rafinesque. 10. Sphyrna zygaena (Linnaeus). Mano kihikihi. (J. & E., p. 41.) The common “Hammer-head” needs comparison with its fellows in, the Atlantic. Family V. ALOPIIDHl (Thresher-sharks). Alopias Rafinesque. 11. Alopias vulpes (Gmelin). (J. & E., p. 42.) Not common. Family VI. LAMNIDHl (Mackerel-sharks). IsuROPSis Gill. 12. Isuropsis glauca (Muller and Henle). (J. & E., p. 43.) Not rare. This shark, with others, needs comparison with Atlantic repre- sentatives. JORDAN AND JORDAN! FISHES OF HAWAII. 5 Carcharodon Muller and Henle (Man-eaters, or Great White Sharks). 13. Carcharodon carcharias (Linnaeus). Niuhi. Probably not rare . Family VII. SQUALID^ (Dog-fishes). Squalus Linnaeus. {Acanthias Risso.) 14. Squalus mitsikurii Jordan and Snyder. (J. & E., p. 45; G., p. 580.) Not rare. A common Japanese species. Etmopterus Rafinesque. (Spinax Cuvier.) 15. Etmopterus villosus Gilbert. (G., p. 580.) Deep seas. Taken off Molokai by the “Albatross.” Centroscyllium Aliiller and Henle. 16. Centroscyllium ruscosum Gilbert. (G., p. 580.) Deep seas. Taken off Kauai by the “Albatross”; identified by Carman with C. nigrum Garman from off the Galapagos. Order BATOIDEI. Family VIII. DASYATID^ (Sting-rays). Dasyatis Rafinesque. {Trygon Adamson; Dasibatus Garman, corrected spelling.) 17. Dasyatis sciera Jenkins. (J. & E., p. 47.) Rather common at Honolulu. 18. Dasyatis lata (Garman). (J. & E., p. 47.) One specimen known. 19. Dasyatis hawaiiensis Jenkins. (J. & E., p. 48.) Only the type known. Family IX. MYLIOBATIDiE. Aetobatus Blainville, as revised by Muller and Henle. {Stoasodon Cantor; Goniohatis Agassiz.) 20. Aetobatus narinari (Euphrasen) Hihimdnu. (J. & E., p. 49.) This species, rather common in Hawaii, seems indistinguishable from the Atlantic form. 6 MEMOIRS OF THE CARNEGIE MUSEUM. Famil}^ X. M0BULIDJ5 (Devil-rays). Morula Rafinesque. (Cephalopterus Diimeril, name preoccupied.) 21. Mobula japonica (Muller and*Henle). (J. & E., p. 50.) Class HOLOCEPHALI. Order CHIM^ROIDEI. Family XL CHIM^RID^. Chimera Linnaeus. 22. Chimsera purpurescens Gilbert. (G., p. 582.) Deep seas. Dredged off Kauai. Class PISCES. Order ISOSPONDYLI. Family XII. ELOPIDJll (Ten-pounders). Elops Linnaeus. 23. Elops hawaiiensis Regan. Awa. (J. & E., p. 53.) Regan has shown that Flops saufus Linnaeus, the common “Tenpounder’’ of the western Atlantic, is not really cosmopolitan, as supposed, but must be separated into several closely related species, of which the abundant Hawaiian form is one. Family XIII. ALBULIDiE (Lady-fishes). Albula (Gronow) Scopoli. {Butyrinus Lacepede.) 24. Albula virgata sp. nov. Jordan and Jordan. Oio. (J. & E., p. 55.) (PI. I, fig. 1.) Type No. 3896, Carnegie Aluseum, from Honolulu. 15.75 inches long. The common Oio of the markets of Hawaii differs markedly in color from Albula vulpes of the American coasts, as well as from all of the nominal species of the genus hitherto described. All of these are brilliantly silvery, with only vague dark lines or stripes. The Hawaiian fish is dusky, marked with distinct stripes much like the markings on a Striped Mullet {Mugil Cephahis). Head 3.33 in length; depth 4.33; dorsal rays 16; anal rays 8; scales 9-72-7; body elongate, moderately compressed; upper lobe of caudal somewhat the longer; JORDAN AND JORDAN! FISHES OF HAWAII. 7 a broad band of elongate, inembranaGeous scales along middle line of back; acces- sory ventral scale large. Color dusky olive, silvery below; a series of dark stripes extending lengthwise of the body, these mainly between the rows of scales, those below the lateral line fainter; dark lines above lateral line; below the lateral line the stripes composed * of stipplings of black dots; tip of snout black in color, forming a broken ring; a little black around nostrils; some faint dark blotches on head; all the fins finely dotted;, dorsal and caudal narrowly rimmed with black. Very common about Honolulu and Hilo, mostly inside the reefs. The genus Albula is widely distributed in most warm seas, only the Mediter- ranean being excepted. Valenciennes recognizes several distinct species, but ah recent writers have regarded all the forms as belonging to one species, no tangible differences in form, scales, or fins being evident. However, specimens from both coasts of America are brilliantly silvery without dark spots, and all the nominal species from the Red Sea, the East Indies, and the South Seas are also described as bright silvery. On the contrary all Hawaiian examples are dusky, with strong stripes along the sides. Family XIV. CHANIDHl. Chanos Lacepede. 25. Chanos chanos (Forskal). Aiva-awa, Awakalamoku, Puawa. (J. &E.,p. 56.) Valenciennes has indicated this common Hawaiian species under the name Chanos cyprinella, but we know of no characters to separate it from C. chanos of the Red Sea. Family XV. DUSSUMIERIIDHl (Round Herrings). Etrumeus Bleeker. 26. Etrumeus micropus (Temminck and Schlegel). Makiawa. (J. & E., p. 58.) We have "been unable to separate this species, which is not very common in Hawaii, from its fellow in Japan. The Californian species, Etrumeus othonops (R. S. Eigenmann), taken but once, and referred to a different genus, Perkmsia, may be different. It is a singular fact that none of the true herrings, Clupeidce, occur about Hawaii. 8 MEMOIRS OF THE CARNEGIE MUSEUM. Family XVI. ENGRAULID.E (Anchovies). Stolephorus Lacepede. {Anchoviella Fowler.) I have given elsewhere (“Genera of Fishes/’ p. 169) my reason for following Bleeker in the application of the name Stolephorus to an Anchovy {Anchoviella) rather than to a Round Herring {Spratelloides) . The genus Anchovia Jordan and Evermann is distinct from Stolephorus, which includes most of the tropical anchovies. 27. Stolephorus purpureus Fowler. Nehu. (J. & E., p. 60.) A common little fish used as bait. Family XVII. STOMATIDHl. Leptostomias Gilbert. 28. Leptostomias macronema Gilbert. (G., p. 607.) Deep sea, off Niihau. Family XVIII. ASTRONESTFIIDiE. Astronesthes Richardson. 29. Astronesthes lucifer Gilbert. (G., p. 605.) Deep sea off Kauai. Family XIX. GONOSTOMID^. Cyclothone Goode and Bean. 30. Cyclothone rhodadenia Gilbert. (G., p. 602.) Deep sea, Kaiwi Channel. 31. Cyclothone canina Gilbert. (G.,, p. 604.) Deep sea off Kauai. 32. Cyclothone atraria Gilbert. (G., p. 605.) Deep sea off Kauai. Family XX. MAUROLICIDHl. Argyripnus Gilbert and Cramer. 33. Argyripnus ephippiatus Gilbert and Cramer. (G., p. 601.) JORDAN AND JORDAN! FISHES OF HAWAII. 9 ViNCiGUERRiA Jordan and Evermann. {Zalarges Jordan and Williams, Proc. Cal. Ac. Sci., 1895, p. 793.) 34. Vinciguerria nimbaria (Jordan and Williams). Pelagic. Northeast of Hawaii. Family XXL STERNOPTYCHIDiE. Sternoptix^ Hermann. 35. Sternoptix diaphana Hermann. (G., p. 609.) Deep seas. Widely distributed. PoLYiPNUS Gunther. 36. Polyipnus nuttingi Gilbert. (G., p. 609.) Deep sea. Argyropelecus Cocco. 37. Argyropelecus heathi Gilbert. (G., p. 601.) Deep sea. Kauai Channel. Diplophos Glinther. 38. Diplophos pacificus Gunther. Deep sea, mid Pacific. Family XXII. HALOSAURID^. Aldrovandia^ Goode and Bean. (1895.) (Halosauropsis Collett, 1896.) 39. Aldrovandia kauaiensis Gilbert. (G., p. 611.) Deep sea off Kauai. 40. Aldrovandia verticalis Gilbert. (G., p. 611.) Deep sea off Kauai. 41. Aldrovandia proboscidea Gilbert. (G., p. 612.) Oahu and Molokai. Family XXIII. SYNODONTIDAE (Lizard-fishes). Trachinocephalus Gill. 42. Trachinocephalus limbatus Eydoux and Souleyet. Kawelea, Welea. (J. & E., p. 62.) This fish, generally common in the Pacific, requires to be compared with Trachinocephalus my ops of the Atlantic. ^ Usually corrected to Sternoptyx. ^ The name Aldrovandia apparently has priority over Halosauropsis. 10 MEMOIRS OF THE CARNEGIE MUSEUM. Synodus (Gronow) Scopoli. (Saurus Cuvier.) 43. Synodus varius (Lacepede). Ulae. (J. & E., p. 63.) Very common in shallow water. The color is very variable. 44. Synodus kaianus (Gunther). (O;, p. 588.) Deep sea. Taken by the “Albatross” off Maui. Saurida Cuvier and Valenciennes. 45. Saurida gracilis (Quoy and Gaimard). Ulae. (J. & E., p. 65; G., p. 589.) Common over coral sand. Family XXIV, CHLOROPHTHALMIDA5. Chlorophthalmus Bonaparte. 46. Chlorophthalmus proridens Gilbert and Cramer. (J. & E., p. 66; G., p. 589.) Deep sea. Common. Fmnily XXV. BATHYPTEROIDA5. Bathypterois Gunther. 47. Bathypterois antennatus Gilbert. (G., p. 590.) Taken by the “Albatross” off Kauai. Family XXVI. PARALEPIDID^. Lestidium Gilbert. 48. Lestidium nudum Gilbert. (G., p. 607.) Deep sea, off Molokai. Family XXVII. MYCTOPHIDiE. Neoscopelus Johnson. 49. Neoscopelus macrolepidotus Johnson. (G., p. 601.) Neoscopelus alcocki Jordan and Starks. Pelagic, widely distributed. According to Gilbert Japanese and Hawaiian specimens are wholly identical with the original Atlantic form, Neoscopelus macrd- lepidotus Johnson, from Madeira. Dasyscopelus Gunther. 50. Dasyscopelus pristilepis Gilbert and Cramer. (G., p. 600.) Pelagic, Hawaii to Marquesas. JORDAN AND JORDAN! FISHES OP HAWAII. 11 51. Dasyscopelus spinosus (Steindachner). (G., p. 599.) Pelagic, Hawaii and southeast. * Rhinoscopelus Liitken. 52. Rhinoscopelus tenuiculus Garman. Pelagic, open seas, southeast of Hawaii. Myctophum Rafinesque. 53. Myctophum fibulatum Gilbert and Cramer. (G., p. 596.) Pelagic, Pailolo Channel between Maui and Molokai. 54. Myctophum affine (Liitken). (G., p. 596.) Myctophum nitidulum Garman. Myctophum margaritatum Gilbert. Rhinoscopelus oceanicus Jordan and Evermann. Pelagic, widely diffused. 55. Myctophum evermanni Gilbert. (G., p. 597.) Pelagic, Hawaii to Marquesas. 56. Myctophum reinhardti Brauer. (G., p. 598.) Myctophum hraueri Gilbert, non Lonnberg. Myctophum luetkeni Gilbert (on plate). Pelagic, widely diffused throughout the tropics. 57. Myctophum hollandi sp. nov. Jordan and Jordan. (PL I, fig. 2.) T}q)e No. 3897, Carnegie Museum. From Honolulu. Head 3.33 in length; depth 4.25; eye 3 in head; snout 6; maxillary 1.5; dorsal rays 1.12; anal rays 1.17; scales 3-35-5; thirty-four photophores on each side. Body moderately elongate, deepest at the occiput, as usual in this group; eye very large; snout very short; mouth large, oblique; jaws even; maxillary rather broad, extending beyond eye nearly to margin of preopercle. Scales rather large; lateral line well developed. Photophores not divided by cross-line; using the nomenclature of Brauer’s Tiefseefische, p. 155, they are arranged as follows: Pectoral photophores (maculce peciorales PO) five, four in a continuous series, the last one higher; Suprapectorales fPLO) one, close to gill-opening and to lateral line; Subpectorales (PVO) two, one near lower axil of pectoral, the other a little lower, near gill-opening; Ventrales (VO) three, in a right line between ventrals and vent; Anales (AO) six, six in a right line with a vacant space equal to one spot above last rays of anal; Posterolaterales (Pol) one, just below lateral line and over 12 MEMOIRS OF THE CARNEGIE MUSEUM. space in anal series; Precaudales (Prc) one, close to lateral line on level of postero- lateral spot; Supra-anales (SAO) three, the upper close to lateral line, the two below out of line, a very obtuse angle at the middle one; Supraventral (VLO) wanting; Opercular (OP) two, close on edge of preopercle, both below upper base of pectoral; Mandibular (Bit) three, in a right line; Antorbital (Antorb.) none, no suborbital or postorbital spots. Dorsal fin high, its first ray equal to depth of body below it; adipose fin small; caudal deeply forked, its lobes 1.4 in head; anal fin rather long, falcate, its edge concave, its longest ray five-sixths height of dorsal, 1.8 in head; pectorals very long, reaching anal, as long as head; ventrals inserted just before dorsal, 2.4 in head. Color blackish, paler below the luminous spots ringed with black. A single example, 4.25 inches in length, was found in good condition by Mr. Grinnell in the market at Honolulu, perhaps a spewing from some large fish. The species is related to Myctophum hraueri as described by Gilbert {Myc- tophurn reinhardti Llitken) but has the anal shorter and the anal photophores fewer. According to Gilbert (7 he Lantern-fishes, Mem. Mus. Comp. Zook, XXVI, 1908, p. 219), Myctophmn reinhardti Liitken is based on two examples. The one figured by Llitken with fourteen dorsal rays and twenty-four anal rays is regarded as the type. This is from the tropical Atlantic. Gilbert observes: “Liitken’s fin-counts were taken from the second specimen, which belongs to a species which remains undescribed.” It is very likely identical with M. hollandi. Centrobranchus Fowler. 58. Centrobranchus choerocephalus Fowler. (G., p. 594.) Pelagic, widely distributed. 59. Centrobranchus gracilicaudus Gilbert. (G., p. 595.) Pelagic, off Niihau. Diaphus Eigenmann and Eigenmann. 60. Diaphus urolampus Gilbert and Cramer. (G., p. 591.) Pelagic, off Kauai. 61. Diaphus chrysorhynchus Gilbert and Cramer. (G., p. 592.) Pelagic, off Oahu and Molokai. 62. Diaphus adenomus Gilbert. (G., p. 592.) Pelagic, Kaiwi Channel. Lampanyctus Bonaparte. 63. Lampanyctus omostigma Gilbert. Pelagic, southeast of Hawaii. JORDAN AND JORDAN! FISHES OF HAWAII. 13 Nannobrachium Gunther. This genus is closely allied to Lampanyctus Bonaparte, Nyctimaster being distinguished by not having enlarged scales along the lateral line. It is distin- guished from Nannobrachium by the very small pectorals of the latter. 64. Nannobrachium nigrum Gunther. (G., p. 591.) Pelagic, south to the Philippines. 65. Nyctimaster reinhardti Jordan. {Cf. Proc. U. S. N. M., LIX, 1921, p. 645, fig. 2.) The three known specimens of this species were killed in a lava-flow from Mauna Loa into deep water off the southwestern coast of Hawaii. Order APODES (Eels). Family XXVIII. SYNAPHOBRANCHIDTl. Synaphobranchus Johnson. 66. Synaphobranchus brachysomus Gilbert. (G., p. 583.) Deep sea. Family XXIX. LEPTOCEPHALIDA5. ( Congridm.) Leptocephalus (Gronow) Scopoli. {Conger Cuvier, adult form.) 67. Leptocephalus marginatus (Valenciennes). Puhi uha. (J. & E., p. 76.) Common in crevices of lava-rock. 68. Leptocephalus bowersi (Jenkins). (J. & E., p. 77.) Rather common. This species belongs to the subgenus Ariosoina Swainson ( Congrellus Ogilby) characterized by the feebler organization and the rather more advanced dorsal fin inserted over the gill-opening. As in Leptocephalus {sens, sir.), the teeth are all sharp. 69. Leptocephalus sequoreus (Gilbert and Cramer). (G., p. 589; J. & E., p. 77.) Deep sea. Veternio Snyder. 70. Veternio verrens Snyder. (J. & E., p. 79.) One large example from Honolulu. Promyllantor Alcock. 71. Promyllantor alcocki Gilbert and Cramer. (G., p. 584.) Deep sea. 14 MEMOIRS OF THE CARNEGIE MUSEUM. Family XXX. MURJ5NES0CID.^. Rhechias Jordan. 72. Rhechias armiger Jordan. (Cf. Jordan, Proc. U. S. N. M., LIX, 1921, p. 644, fig. 1.) Off the southwestern coast of Hawaii, the type killed in deep water by a lava- flow from Maima Loa. Family XXXI. NETTASTOMIDffH (Sorcerers). Metopomycter Gilbert. 73. Metopomycter denticulatus Gilbert. (G., p. 585.) Deep sea, off Kauai. Family XXXII. NEMICHTHYIDHl (Snipe-eels). Nematoprora Gilbert. 74. Nematoprora polygonifera Gilbert. (G., p. 587.) Deep sea, off Bird Island. Serrivomer Gill and Ryder. 75. Serrivomer beani Gill and Ryder. (G., p. 586.) Deep sea. Stemonidium Gilbert. 76. Stemonidium hypomelas Gilbert. (G., p. 586.) Deep sea, off Niihau. Family XXXIII. OPHICHTHYIDHl (Snake-eels). Sphagebranchus Bloch. 77. Sphagebranchus flavicaudus Snyder. (J. & E., p. 80; G., p. 588.) Occasionall}'^ taken. Leiuranus Bleeker. {Stethoptenis Bleeker has line-priority, but later Leiuranus was preferred by the author.) 78. Leiuranus semicinctus (Lay and Bennett). (J. & E., p. 81.) Warm parts of the Pacific. Rare about Hawaii. Microdonophis Kaup. 79. Microdonophis fowleri Jordan and Evermann. (J. & E., p. 82.) Rare, but three specimens known. JORDAN AND JORDAN! FISHES OF HAWAII. 15 Jenkinsiella Jordan and Evermann. 80. Jenkinsiella macgregori (Jenkins). (J. & E., p. 82.) One specimen from Maui. Brachysomophis Kaiip. 81. Brachysomophis henshawi Jordan and Snyder. (J. & E., p. 83.) One large specimen from Honolulu. Myrichthys Girard. 82. Myrichthys stypurus (Smith and Swam). (J. & E., p. 84.) Johnston Island, one example known. 83. Myrichthys magnificus (Abbott). (J. & E., p.- 84.) Not seen since the original description was written. Callechelys Kaup. 84. Callechelys luteus Snyder. (J. & E., p. 86.) One large example from Molokai. Family XXXIV. MORINGUIDHl. - Moringua Gray. ’ (Raitaboura Gray has line-priority, but Moringua has been preferred by revisers) . 85. Moringua hawaiiensis Snyder. (J. & E., p. 86.) One example from Honolulu. Family XXXV. JMUR.ENIDHl (MorayG- Mur^na Linnaeus. 86. Muraena kailuae Jordan and Evermann. Pulii kauila; Puhi oa. (J. & E., p. 88.) The two nominal species, Murcena lampra Jenkins and Murcena kauila Jenk- ins, seem to be color variations of this highly variable species, the body of which is brown, marked by white spots, often dark-ringed and of various sizes and forms, usually largest on the tail. Enchelynassa Kaup. 87. Enchelynassa canina (Quoy and Gaimard). (J. & E., pp. 90, 91.) Enchelynassa hleekeri Kaup. . Gymnothorax vinolentus Jordan and Evermann. A very large Moray, found occasionally about Hawaii and Samoa. 16 MEMOIRS OF THE CARNEGIE MUSEUM. Gymnothorax Bloch. (Lycodontis McClelland.) 88. Gymnothorax eurostus (Abbott). (J. & E., p. 92.) Hawaii, not seen since the original description. 89. Gymnothorax laysanus (Steindachner). (J. & E., p. 93.) Not rare about Honolulu. 90. Gymnothorax meleagris (Shaw). (J. & E., p. 94.) South Seas, rare about Honolulu. 91. Gymnothorax steindachneri Jordan and Evermann. (J. & E., p. 101.) Not rare about Honolulu. 92. Gymnothorax gracilicauda Jenkins. (J. & E., p. 94.) Rare; possibly the young of G. steindachneri. 93. Gymnothorax ercodes Jenkins. (J. & E., p. 95.) One known from Honolulu. 94. Gymnothorax berndti Snyder. (J. & E., p. 98.) Rare about Honolulu. 95. Gymnothorax undulatus (Lacepede). Puhi laumili. (J. & E., p. 98.) The commonest Moray about Hawaii and especially ferocious. 96. Gymnothorax flavomarginatus (Riippell) . (J. & E., p. 99.) Rather common. 97. Gymnothorax thalassopterus Jenkins. (J. & E., p. 99.) Rare. Perhaps a variant of G. flavomarginatus. 98. Gymnothorax goldsboroughi Jordan and Evermann. (J. & E., p. 100.) One specimen known. 99. Gymnothorax petelli (Bleeker). (J. & E., p. 100.) {Gymnothorax leucacme Jenkins.) Rather common and widely diffused. 100. Gymnothorax mucifer Snyder. (J. & E., p. 97.) Honolulu, one example. 101. Gymnothorax leucostictus Jenkins. (J. & E., p. 96.) Two examples from Honolulu. 102. Gymnothorax waialuae Snyder. (J. & E., p. 97.) One specimen from Waialua Bay, Oahu. 103. Gymnothorax hilonis Jordan and Evermann. (J. & E., p. 102.) One example from Hilo. 104. Gymnothorax nuttingi Snyder. (J. & E., p. 103.) Only one example known. JORDAN AND JORDAN! FISHES OP HAWAII. 17 105. Gymnothorax pictus (Ahl). Puhi kapa’a. (J. & E., p. 103.) Common and variable, widely diffused. 106. Gymnothorax xanthostomus Snyder. (J. & E., p. 104.) Honolulu, rare. Eurymyctera Kaup. 107. Eurymyctera acutirostris (Abbott) . (J. & E., p. 105.) Not seen since the original discover}^; the species has been redescribed and figured by Fowler. Echidna Forster. 108. Echidna zebra (Shaw). (J. & E., p. 106.) Scarce about Hawaii; common in the South Seas. 109. Echidna tritor (Vaillant and Sauvage). (J. & E., pp. 106, 107, 108, 109.) {Echidna obscura Jenkins.) Abundant and excessively variable in color. It is believed that the nominal species E. leihala Jenkins, E. psalion Jenkins, E. zonata Fowler, E. vincta Jenkins, and E. zonophwa Jordan and Evermann are all variants of E. tritor, which is plain in color with a black spot at the angle of the mouth. These are variously marked with dark cross-bands, scarcely any two specimens being colored alike. The alleged differences in dentition need verification. 110. Echidna nebulosa (Ahl). Puhi kdpa. (J. & E., p. 110.) Common and widely distributed. Uropterygius Rlippell. (Ichthyophis Kaup, preoccupied.) 111. Uropterygius marmoratus (Lacepede). (J. & E., p. 111.) South Seas, scarce about Hawaii. 112. Uropterygius leucurus Snyder. (J. & E., p. 112.) Only one specimen known. ScuTiCARiA Jordan and Snyder. 113. Scuticaria tigrina (Lesson). (J. & E., p. 112.) South Seas, occasional about Hawaii. 18 MEMOIRS OF THE CARNEGIE MUSEUM. Order SYNENTOGNATHI. Family XXXVI. BELONID^ (Needle-fishes). Platybelone Fowler. {Eurycaulus Ogilby, Proc. Royal Soc. Queensland, XXI, 1908, p. 91, type Belone platyura Bennett, is preoccupied, and Platybelone ‘Fowler, Jan., 1919, is substituted. The gill-rakers are present as in Belone, the tail is broad, depressed, and keeled.) 114. Platybelone platyura (Bennett). (J. & E., p. 122.) South Seas. Not rare about Hawaii. Tylosurus Cocco. 115. Tylosurus giganteus (Temminck and Schlegel) Aha aha; Auau. (J. & E., p. 124.) This large Hawaiian fish requires to be compared with the original species from Japan. Not rare in the open sea. Ablennes Jordan and Fordice. (Originally written in error Athlennes.) 116. Ablennes hians (Cuvier and Valenciennes). (J. & E., p. 125.) It is very doubtful whether the rare Hawaiian form is identical with A. hians of the West Indies. Family XXXVII. HEMIRHAMPHID^E (Half-beaks). Hyporhamphus Gill. 117. Hyporhamphus pacificus (Steindachner). (J. & E., p. 126.) Common at times. Hemirhamphus Cuvier. 118. Hemirhamphus depauperatus Lay and Bennett. Me’eme’e; Iheihe. Locally abundant. Euleptorhamphus Gill. 119. Euleptorhamphus longirostris (Cuvier). Iheihe. (J. & E., p. 128.) Not rare in the open sea. JORDAN AND JORDAN: FISHES OF HAWAII. 19 Family XXXVIII. EXOCGETIDvE (Flying-fishes). Fodiator Jordan and Meek. 120. Fodiator rostratus (Gunther). (J. & E., p. 131.) ‘One example taken in Hawaii. The species seems to differ from Fodiator acutus of the Panama region in the sub vertical mouth and the shorter lower jaw It is nearer Fodiator than Parexocoetus. Evolantia Snodgrass and Heller. 121. Evolantia microptera (Cuvier and Valenciennes). (J. & E., p. 130.) Scarce about Hawaii. Parexoccetus Bleeker. 122. Parexocoetus brachypterus Solander. Pukiku. (J. & E., p. 131.) Very common, not exceeding seven inches. Exoccetus Linnaeus. (Ventral fin short, median.) Exocoetus Linnaeus, Syst. Nat., Ed. X, 1758, p. 316. Type Exoccetus volitans, lately shown to be based on an example of the species called Halocypselus evolans (Linnaeus). Halocypselus Weinland, Proc. Post. Soc. Nat. Hist., VI, 1858, p. 385 {meso- gaster = evolans = volitans). 123. Exoccetus volitans Linnaeus. (J. & E., p. 132.) By a confusion incident to correction of syonymy the plate on page 133, Jordan and Evermann, named ‘‘Exoccetus volitans,’' represents the species sometimes called by that name, = Exoccetus rubescens Rafinesque, not the true E. volitans, which has short ventral fins. Exonautes Jordan and Evermann. (Anal fin not shorter than dorsal.) 124. Exonautes gilberti Snyder. (J. & E., p. 134.) Rare. The species from near Samoa, identified by Jordan and Seale as Exoccetus unicolor Cuvier and Valenciennes, figured on page 209 of the “Fishes of Samoa,” is very close to Exonautes gilberti and perhaps the same. In the speci- mens of both, as figured, is the parasitic copepod Penella, to which a parasitic barnacle {Conchoderma) is attached. 20 MEMOIRS OF THE CARNEGIE MUSEUM. Cypselurus* Swainson. (Anal fin much shorter than dorsal; young (always?) with barbel at the chin.) 125. Cypselurus simus (Cuvier and Valenciennes). Malolo. (J. & E., p. 134.) The commonest large flying-fish about Hawaii, reaching a length of fourteen inches. The pectoral fins are usually, but not always, spotted with black. 126. Cypselurus spilonotopterus (Bleeker). Malolo. (J. & E., p. 136.) Cypselurus bahiensis Jordan and Evermann, p. 136; probably not Exoccetus bahiensis Ranzani. Usually common about Hawaii. A very large species, reaching twenty inches in length. It is known in life by its dark reddish-brown pectorals, which become blackish in spirits. The dorsal fin is largely black. The species is most likely distinct from the Atlantic form called C. bahiensis. 127. Cypselurus atrisignis Jenkins. (J. & E., p. 136.) Rare. Dorsal fin with a large black spot. Family XXXIX. MACROURIDaE (Grenadiers). ( Coryphsenoididse. ) Gadomus Regan. 128. Gadomus melanopterus Gilbert. (G., p. 658.) Deep water off Kauai. 129. Gadomus bowersi Gilbert. (G., p. 659.) Deep water off Bird Island. Melanobranchus Regan. 130. Melanobranchus micronemus Gilbert. (G., p. 661.) Deep water, Pailolo Channel. Chalinura Goode and Bean. 131. Chalinura ctenomelas Gilbert and Cramer. (G., p. 662.) Deep sea, very abundant. Optonurus Giinther. 132. Optonurus atherodon Gilbert and Cramer. (G., p. 663.) Deep sea; the most abundant member of the group. * The International Commission of Nomenclature has decided that the spelling Cypsilurus of Swainson is to be I’egarded as a misprint. JORDAN AND JORDAN! FISHES OF HAWAII. 21 Hymenocephalus Gigiioli. 133. Hymenocephalus striatulus Gilbert. (G., p. 665.) Deep sea off Oahu. 134. Hymenocephalus aterrimus Gilbert. (G., p. 666.) Kanai, in very deep water. 135. Hymenocephalus antraeus Gilbert and Cramer. (G., p. 663.) Deep sea, extremely abundant. A valid species, not to be confounded with H. aterrimus Gilbert. Macrourus Bloch. (This genus, distinguished by the subinferior mouth, is merged into CorijphcE- noides by Hubbs.) 136. Macrourus ectenes Gilbert and Cramer. (G., p. 667.) Deep sea. One specimen known. 137. Macrourus propinquus Gilbert and Cramer. (G., p. 667.) Deep sea off Kauai. 138. Macrourus holocentrus Gilbert and Cramer. (G., p. 668.) Deep sea off Oahu. One specimen known. 139. Macrourus gibber Gilbert and Cramer. (G., p. 668.) Deep sea; frequent. 140. Macrourus burragei Gilbert. (G., p. 668.) Deep sea off Oahu; one specimen known. 141. Macrourus obliquatus Gilbert. (G., p. 670.) Deep sea off Kauai. Only one specimen known. 142. Macrourus hebetatus Gilbert. (G., p. 671.) Deep sea off Oahu, one siiecimen known. 143. Macrourus longicirrhus Gilbert. (G., p. 672.) Deep sea off Kauai. Only the type known. CcELORHYNCHUS Gioma. 144. Coelorhynchus gladius Gilbert and Cramer. (G., p. 673.) Deep sea. 145. Coelorhynchus aratrum Gilbert. (G., p. 674.) Deep sea. Rather scarce. 146. Coelorhynchus doryssus Gilbert. (G., p. 675.) Deep sea. Occasional. 22 MEMOIRS OF THE CARNEGIE MUSEUM. Mat^ocephalus Berg. {Coelocephalus Gilbert and Cramer; preoccupied.) 147. Mataeocephalus acipenserinus (Gilbert and Cramer). (G., p. 676.) Deep sea. Common. Malacocephalus Giintlier. 148. Malacocephalus hawaiiensis Gilbert. (G., p. 677.) Deep sea off Oahu. Trachonurus Gunther. 149. Trachonurus sentipellis Gilbert and Cramer. (G., p. 679.) Deep sea, frequent. Family XL. GADIDffE. Antimora Giinther. 150. Antimora microlepis Bean. (G., p. 656.) Deep sea off Kauai. An Alaskan species. L.emonema Giinther. 151. Laemonema rhodochir Gilbert. (G., p. 657.) Deep sea off Oahu. But one specimen known. Physiculus Kaup. 152. Physiculus grinnelli sp. nov. Jordan and Jordan. (PI. I, fig. 3.) Type: No. 3898 Carnegie Museum. Twelve and one-half inches long. Found in the market at Honolulu. Head 4 in length to base of caudal; depth 4.8; eye 4.66 in head; snout 4.66; maxillary 2.16; barbel 4.5; height of first dorsal '2.5; length of ventral 1.16; liectoral 1.33; caudal 2; dorsal rays 7-73; anal rays 65; ventral rays 6; scales 6-127-26. Body moderately elongate, deepest under the first dorsal, the tail rather slender; head somewhat flattened, the jirofile depressed above the eye; mouth moderate ; the lower jaw included ; the narrow maxillary reaching about to posterior margin of eye; gill-rakers very short, blunt; eye moderate. First dorsal rather low, one and four- fifths times as high as long; second dorsal moderate, co- terminous with anal; caudal rounded. Ventrals reaching well past front of anal. Scales small, smaller posteriorly and below; snout and lower jaw scaleless. Soft fins with small scales; lateral line well developed. Color plain dusky, paler below, edges of fins darker. JORDAN AND JORDAN! FISHES OF HAWAII. 23 This species requires to be compared with Physiculus japonicus Hilgendorf from Tokyo. The following is the scanty description (Gesellsch. Naturforsch. Freunde Berlin, 1879, p. 80): “Von der Gattimg Physiculus sind bisher 3 Arten bekannt geworden; die erste, Ph. dalwigkii KP. wieder von Madeira, ist von imserer durch folgende Merkmale zii unterscheiden. Bei Ph. japonicus ist die Kopfliinge in der Kor|:erl. (ohne Gaud.) 5mal enthalten (bei P. Dalwigkii 4mal). Interorbitalraum gleich dem vertikalen Augendurchmesser (statt kleiner), D. I. ist 1 1/3 mal so hoch als lang (2 mal), imd die Hohe unter halber Kopflange (gleich der halben), die Fiiden der V. erreichen die A. (nicht). B. 7, D. 9/66, A. 73, V. 7. Die anderen beiden Arten, von Cuba und Sudaustralien, sind durch die Flossenformal hinreichend getrennt. Mus. Ber. No. 10624.” Order ZEOIDEA. Family XLI. ZEID.D (John Dories). Stethopristes Gilbert. 153. Stethopristes eos Gilbert. (G., p. 622.) Deep sea, Pailolo Channel. Cyttomimus Gilbert. (The presence of six soft rays in the ventral fins indicates that this genus belongs to the Zeidce rather than to the Caproidce.) 154. Cyttomimus stelgis Gilbert. (G., p. 624.) Deep sea off Oahu; but one specimen known. Fig. 1. Vesposus egregius Jordan. (Reproduced from Proc. U. S. N. M., Vol. 59, 1921, p. 650.) 24 MEMOIRS OF THE CARNEGIE MUSEUM. Family XLII. GRAMMICOLEPIDvE. Vesposus Jordan. 155. Vesposus egregius Jordan. (Proc. U. S. Nat. Mus., LIX, 1921, p. 650.) Deep sea off Hawaii; the type killed in overflow of lava from Maima Loa. Order CHONDRICHTHYES. Family XLIII. ATELEOPIDHi. Ateleopus Temminck and Schlegel. (Podateles Boiilenger, there being already a genus Atelopus.) 156. Ateleopus plicatellus Gilbert. (G., p. 653.) Deep sea, Pailolo Channel. Order HETEROSOMATA. Family XLIV. PLEURONECTIDHl. PcEciLOPSETTA Glinther. 157. Poecilopsetta hawaiiensis Gilbert. (G., p. 679.) Deep sea, Pailolo Channel. ' T^niopsetta Gilbert. 158. Tseniopsetta radula Gilbert. (G., p. 680.) Deep sea, Pailolo Channel. Platophrys Swainson. 159. Platophrys mancus (Broussonet). (J. & E., p. 513; G., p. 684.) {Rhoinboidichthys pavo Gunther.) Occasionally taken. 160. Platophrys pantherinus (Ruppell). (J. & E., p. 512.) Generally common about Hawaii. 161. Platophrys chlorospilus Gilbert. (G., p. 684.) Off jMaui in deep water. 162. Platophrys inermis Gilbert. (G., p. 685.) Deep sea, Pailolo Channel. 163. Platophrys coarctatus Gilbert. (G., p. 686.) Deep sea. JORDAN AND JORDAN! FISHES OF HAWAII. 25 This species and the preceding, with the interorbital very narrow, diverge considerably from the type of Platophrys. Scoops Jordan and Starks. {Platophrys Giinther, non Swainson.) 164. Sceeops hawaiiensis (Jordan and Evermann). (J. & E., p. 514; G., p. 687.) 165. Scaeops xenandrus (Gilbert). (G., p. 687.) Common in rather deep water. 166. Scseops arenicola (Jordan and Evermann). (J. & E., p. 515.) Among the large-scaled flounders known by the very narrow interorbital, thus approaching Engyprosopon Gunther, but the gill-rakers are very short, as in Scceops. Anticitharus Gunther. 167. Anticitharus debilis Gilbert. (G., p. 683.) Deep sea, Pailolo Channel. Chascanopsetta Gilbert. 168. Chascanopsetta prorigera Gilbert. (G., p. 689.) Deep sea, off Maui. Pelecanichthys Gilbert and Cramer. 169. Pelecanichthys crumenalis Gilbert and Cramer. (G., p. 690.) Deep sea. Samariscus Gilbert. 170. Samariscus corallinus Gilbert. (G., p. 682.) Deep sea, off Alolokai. Eamily XLV. CYNOGLOSSIDA5 (Soles). Symphurus Rafinesque. 171. Symphurus undatus Gilbert. (G., p. 690.) Deep sea, off Oahu. 172. Symphurus strictus Gilbert. (G., p. 691.) Deep sea, off Oahu. 26 MEMOIRS OF THE CARNEGIE MUSEUM. Order XENOBERYCES. Family XLVI. MELAMPHAIDJ5. ' MelamphaEs Gunther. 173. Melamphaes unicornis Gilbert. (G., p. 615.) Deep sea, off Kauai. Caulolepis Gill. 174. Caulolepis longidens Gill. (G., p. 616.) Dee]) sea, perhaps distinct from the Atlantic form. Order BERYCOIDEI. Family XLVIL POLY^MIXIIDA^. POLYMIXIA Lowe. 175. Polymixia berndti Gilbert. (G., p. 616.) Deep sea, off Oahu. Family XLVIII. HOLOCENTRID^ (Squirrel-fishes). Holotrachys Gimther. 176. Holotrachys lima (Cuvier and Valenciennes). (J. & E., p. 147.) Common in Hawaii and throughout the South Seas. OsTiCHTHYS (Langsdorf) Jordan and Evermann. 177. Ostichthys pillwaxi (Steindachner). (J. & E., p. 147.) Wry rare. Two specimens known from Honolulu. Myripristis Cuvier. (Freres Jacques.) 178. Myripristis multiradiatus Gunther. U'u. (J. & E., p. 149.) Abundant about Hawaii. 179. Myripristis chryseres Jordan and Evermann. Pauu. Not rare about Hawaii. 180. Myripristis symmetricus Jordan and Evermann. (J. & E., p. 151 Rather scarce. 181. Myripristis sealei Jenkins. (J. &E.,p. 151.) Not rare. 182. Myripristis murdjan (Forskal). U’u. (J. & E., p. 152.) JOKDAN AND JORDAN: FISHES OF HAWAII. 27 The commonest species of the genus, widely dispersed throughout the Pacific. Myripristis berndti Jordan and Evermann, p. 153, is probably not distinct from M. murdjan. 183. Myripristis argyromus Jordan and Evermann. (J. & E., p. 154.) One example known. Holocentrus. § Holocentrus. 184. Holocentrus diadema Lacepede. Alaihi kalaloa. (J. & E., p. 159.) Very common; one of the small species. 185. Holocentrus microstomus Gunther. (J. & E., p. 160.) Rather scarce. 186. Holocentrus spinifer (ForskM). (J. & E., p. 161.) Rare about Hawaii. 187. Holocentrus erythraeus Gunther. (J. & E., p. 161.) Scarce. 188. Holocentrus punctatissimus Cuvier and Valenciennes. (J. & E., p. 162.) A small fish generally common about Hawaii. 189. Holocentrus xantherythrus Jordan and Evermann. (J. & E., p. 164.) Common. The specific name of this species was rather unfortunately chosen, as its pale stripes are white, not yellow. The yellow streaks are characteristic of H. ensifer, for which the name was originally framed. 190. Holocentrus ensifer Jordan and Evermann. (J. & E., p. 165.) Rather common. § Flammeo Jordan and Evermann. 191. Holocentrus sammara Forskal. (J. & E., p. 155.) Common, widel}^ diffused. This species and the next belong to the subgenus Flammeo, distinguished by the larger mouth and projecting chin, characters of minor importance. 192. Holocentrus scythrops Jordan and Evermami. (J. & E., p. 157.) Abundant about Hawaii. Order AULOSTOMI. Family XLIX. AULOSTOVIIDiE (Trumpet-fishes). Aulostomus Lacepede. 193. Aulostomus chinensis (Linnseus). Nunu. (J. & E., p. 114.) Common. The original description of “ Fistularia chinensis” Linnseus in- cluded two Asiatic references and the species is said to inhabit the East Indies. 28 MEMOIRS OF THE CARNEGIE MUSEUM. The specific name chinensis should therefore remain with the Asiatic form, known as Aulostomus valentini by some later authors. Family L. FISTULARIIDdE (Cornet-fishes). Fistularia Linmeus. 194. Fistularia petimba Lacepede. (J. & E., p. 116.) Abundant. 195. Fistularia serrata Cuvier. (J. &E., p. 116.) Scarce about Hawaii. Family LI. MACRORHAMPHOSID.E. AIacrorhamphosus Lacepede. 196. Macrorhamphosus hawaiiensis Gilbert. (G., p. 613.) Off Laysan Island. Order LOPHOBRANCHII.^ Family LII. SYNGNATHIDdE. Microphis Kaup. 197. Microphis pleurotaenia (Gunther). (J. & E., p. 121.) Rare. Off Honolulu. ICHTHYOCAMPUS Kaup. 198. Ichthyocampus erythrseus Gilbert. (G., p. 613,) Off Molokai. Family LIII. HIPPOCAMPIDvE (Sea-horses). IIippocAMPUS Rafinesque. 199. Hippocampus hilonis Jordan and Evermann. (J. & E., p. 119.) One example from Hilo. 200. Hippocampus fisheri Jordan and Evermann. (J. & E., p. 119.) Scarce. Order HYPOSTOMIDES. Family LIV. PEGASIDAi; (Sea-moths). Pegasus Linnseus. 201. Pegasus papilio Gilbert. (G., p. 614.) Bird Island, and off Hawaii. ^ Solenosto'inus cyanopterus Bleeker has been reported from Hawaii in error. JORDAN AND JORDAN: FISHES OF HAWAII. 29 Order SELENICHTHYES. Family LV. LAMPRIDJE (Moon-fishes). Lampris Retzius. 202. Lampris regius (Bonnaterre). (J. & E., p. 166.) An example, six feet long, was once taken at Honolulu. It weighed 217 lbs. The Honolulu “Star-Bulletin ” in an issue early in 1922 re]iorts the capture at a depth of 1200 ft. of a second specimen, weighing much less. It was taken thirteen miles west of Oahu. Order PERCOMORPHI. Suborder PERCESOCES. Family LVI. ATHERINIDHl (Silversides). Hepsetia Bonaparte. 203. Hepsetia insularum (Jordan and Evermann). (J. & E., p. 138.) This little fish, common inside of the reefs, has the lower mandible straight, not abruptly elevated behind. It belongs, therefore, with most of the Pacific “Silversides” to the genus Hepsetia. Family EVIL MUGILIDHl (Mullets). Mugil Linnaeus. 204. Mugil cephalus Linnaeus. Ama-ama. (J. & E., p. 139.) The commonest food-fish in Honolulu, and one of the best, being largely- reared in salt-water ponds. We have been unable to distinguish the Hawaiian form from the Striped Mullet of Europe, and therefore let it stand under the same name. Ch^nomugil Gill. (This genus differs from Chelon Rose of the Mediterranean by having both jaws provided with papilliform teeth.) 205. Chsenomugil chaptalii (Eydoux and Souleyet). Uouoa. (J. & E., pp. 140- 141.) Myxus pacificus Steindachner seems to be the young of this species. Family LVUI. SPHYRmHD.E (Barracudas). Sphyr^na Lacepede. § Sphyrcena. 206. Sphyraena helleri Jenkins. Kawalea. (J. & E., p. 143.) A small species, not exceeding two feet in length. Generally common. 30 MEMOIRS OF THE CARNEGIE MUSEUM. § Agriosphyroena Fowler. (Giant barracudas with large scales, less than ninety.) 207. Sphyraena snodgrassi Jenkins. Kdku. (J. & E., p. 143.) This large and fierce Barracuda is common in the markets, and reaches a length of six feet. The species requires to be compared with other large Barracudas of the South Seas. Suborder RHEGNOPTERI. Family LIX. POLYNEMIl)^ (Thread-fishes). PoLYNEMUs Linna3us. ( Polydactylus Lacepede.) 208. Polynemus sexfilis Cuvier and Valenciennes. Moi; Moi-lii. (J. & E., p. 144.) Not rare at Honolulu. Suborder PERCIFORMES. Family LX. XIPHIIDYl. XiPHiAS Linmeus. 209. Xiphias gladius Linnaeus. (J. & E., p. 168.) The common swordfish is occasionally taken at Honolulu. Family LXI. ISTIOPHORID.E (Spear-fishes). Tetrapterus Agassiz, 210. Tetrapterus mitsukurii Jordan and Snyder. A’u. This large spear-fish, originally described from Japan, but since found to be abundant at Santa Catalina, may be seen every day in the Honolulu markets. It is taken in the open sea to the southwestward by Japanese fishermen. We have had no opportunity to compare Hawaiian specimens with those taken elsewhere. Pectoral longer than dorsal lobe. IsTioPHORUs Lacepede (Sail-fishes). ( Histiophorus of most recent authors.) 211. Istiophorus gladius (Broussonet). A cast of an example six feet long is in the Bishop Museum. It is not certain that the Atlantic form is really distinct from this. A photograph of the cast is given in fig. 2. JORDAN AND JORDAN! FISHES OF HAWAII. 31 Family LXII. SCOMBRIDtE (Mackerels). Pneumatophorus Jordan and Gilbert. 212. Pneumatophorus japonicus (Houttuyn). Opelu palahu. (J. & E., p. 169.) This small mackerel is rather rare about Hawaii. It needs comparison with the abundant geminate forms, P. japonicus of Japan and P. diego from California. P. colias of Europe and P. grex of our Atlantic coast also differ slightly, though all are very much alike. The “Chub-mackerels,” Pneumatophorus, differ from the mackerel of commerce. Scomber, in the development of the air-bladder. Fig. 2. Istiophorus gladius (BrousSonet). From a cast in the Bernice Pauahi Bisliop Museum, Honolulu. Auxis Cuvier (Frigate-mackerels). 213. Auxis thazard (Lacepede). (J. & E., p. 171.) This pelagic fish requires to be compared with A. rochei of the Atlantic and A. tapeinosoma of Japan. Euthynnus Ltitken (Oceanic Bonitos). (We let this genus stand until it can be compared directly with Gymnosarda unicolor, the type of the allied genus Gymnosarda.) 214. Euthynnus pelamis ( Linnaeus) . Aku. (J. & E., p. 172.) This fish of the open sea is now very abundant in the markets of Honolulu and Hilo. It is extensively canned for commerce; more than any other species. The flesh is red, rather coarse, and oily. The better species of this group are not put up in tins, their use as fresh fish being more profitable. The best of them sell at present at fifty cents a pound in Honolulu. The various forms of striped “Oce- anic Bonitos” found in the warm parts of the Atlantic and Pacific need comparison one with another. 215. Euthynnus alleteratus (Rafinesque). Kdwakdwa. (J. & E., p. 173.) Very common in the markets. The young are taken in nets in the shallow waters of Hilo Bay. The flesh is paler than that of the Aku, and brings a higher 32 MEMOIRS OF THE CARNEGIE MUSEUM. price; hence it is less frequently tinned. From two to six round black spots appear in the adult fish along the sides of the breast. These are not shown in the figure (No. 65) given b}^ Jordan and Evermann. The Pacific form should be compared with true E. alleteraius of the Mediterranean. Sarda Cuvier (Bonitos). 216. Sarda chilensis (Cuvier and Valenciennes). (J. & E., p. 175.) Occasionally taken at Honolulu and canned with the Aku, packers making no fine distinctions. This species is quite different from the Atlantic Bonito, Sarda sarda, having the spinous dorsal always shorter. It -is not quite certain that Sarda lineolata from California and Sarda orientalis from Japan are identical with Sarda chilensis. Thunnus South (Tunnies). ( Thynnus Cuvier; preoccupied.) 217. Thunnus thynnus Linmeus. The great Tuna, regarded as identical with the European, and which is abun- dant about Santa Catalina Island, California, is not yet definitely known from Hawaii. 218. Thunnus orientalis (Temminck and Schlegel). A specimen seen in the market at Honolulu seemed distinct from the Cali- fornian Tuna, having the finlets dull yellow instead of blue. According to our notes the dorsal and anal lobes are high, the pectoral rather short, reaching two- thirds distance to anal. Finlets all dull soiled yellowish. Belly with twelve obscure pale cross-bars of grayish silvery, narrower than the interspaces, replaced by round spots above and below ; smaller spots alternating with the bars ; no clear yellow on fins. The silvery markings are characteristic of the young of several species of this group. Germo Jordan (Albacores). This groiq:) or subgenus differs from Thunnus only in the great length of the ribbon-like pectoral fins, which reach at least to the front of the anal, two and one half to three times in length of body. It should perhaps be merged in Thunnus. The species of this genus are much in need of careful revision. 219. Germo macropterus Temminck and Schlegel. Ahi. (J. & E., p. 174.) {Germo germo Jordan and Evermann.) This species, found both in California and Japan, is now rather abundant in the Honolulu markets. It reaches a weight of three hundred pounds. Dorsal and finlets all bright lemon-yellow without dark l)orders. The sides have faint JOEDAN AND JOEDANI FISHES OF HAWAII. 33 elongate dull silvery sjiots, not cross-bands. The dorsal and anal are very high and falcate. The flesh is coarse and red, like that of the AkUj with which it is often canned. This species was recorded by Jordan and Evermann in 1901 as Germo germo. It is brought in from deep water by the Japanese fishermen. 220. Germo sibi (Temminck and Schlegel). Soft dorsal moderately elevated, its lobe shorter than snout. Pectoral long, falcate, reaching to the second dorsal finlet. Finlets above bright yellow bordered by dark, the narrow margin white, the produced tips white; anal finlets all pale with no yellow. Flesh dark. Sides without distinct silvery markings. A large fish, frequently seen in the markets, and evidently distinct from G. macropterus and G. alalunga. It seems to be very near G. sibi of Jajian, but its identity cannot be positively decided without actual comparison of specimens. 221. Germo alalunga (Gmelin). (? Scomber germo Lacepede = Thynnus pacificus Cuvier and Valenciennes.) Another long-fin is occasionally taken with the others. Upon superficial examination it seems to be the same as the Californian Albacore, supposed to be Germo alalunga. Finlets all blue with no trace of yellow. Pectoral very long, reaching middle of dorsal lobe. Flesh pale. Weight twelve to fifteen pounds. This may be Scomber germo of Lacepede {pacificus C. & V.), but the long descripr tions of that author reveal no points of difference and the color of the finlets is not mentioned. In Jordan and Evermann, “Fishes of North and Middle America,” pp. 870-871, in the account of Thunnus thyjinus and Germo alalunga, the references to the flesh of the two are accidentally transposed. The flesh of the Tuna {Thunnus) is “coarse and oily”; that of the Albacore {Germo) is “excellent, that even of very large individuals being of fine flavor.” 222. Germo argentivittatus (Cuvier and Valenciennes). Dr. Nichols tells me that a specimen sent by Dr. Evermann in 1920 to the American Museum of Natural History corresponds to this species from “the Indian seas.” The color of the body, as stated by Cuvier and Valenciennes, corresponds to that of Thunnus orientalis, but the long pectorals arn said to be three and one-half in the length of body, not seven, as in Schlegel’s account of orientalis. 34 MEMOIRS OF THE CARNEGIE MUSEUM. Acanthocybium Gill (Petos). 223. Acanthocybium solandri (Cuvier and Valenciennes). Ono. This large fish is now common in the market of Honolulu, being taken with the hook in deep water thirty miles or more from the harbor by the Japanese. The flesh is excellent, being too costly to be used for canning. Jordan and Thomp- son have noticed that the Japanese form, Acanthocyhimn sara, is very distinct from A. solandri. The Cuban Peto, A. petus Poey, is also different. Brown, with narrow faint silvery cross-bars on sides. Teeth 75/60 on each side, compressed, smaller inwards; pectorals a little shorter than maxillary. The account of this species, given by Jordan and Evermann, is drawn from a Cuban example of Acanthocyhimn petus. The Japanese fish, Acanthocybium sara, called in Japan Okisawara, or “off-shore Sawara,” has the teeth much larger, 18/20 on either side, the snout blunter, the body less slender. ' (See Jordan and Metz, Memoirs Carnegie Museum, VI, p. 27.) The description copied by Cuvier and Valenciennes from Solander is not distinctive, and no locality is assigned to the species. As Solander collected prin- cipally about Tahiti, it is iiresumable that his species is the present. The huge size of these fishes debars them from collections. Family LXIII. GEMPYLIDtE (Snake-mackerels). Ruvettus Cocco. 224. Ruvettus pacificus sp. nov. Jordan and Jordan. Walu. (J. & E., p. 177.) Type: No. 04314, U. S. N. M. A single specimen, four and one half feet long, weighing forty pounds, was obtained by Jordan and Evermann from Honolulu. This is the only record, so far as we know, from the Pacific. This example we may take as the type of a new species. It is well described and figured by Jordan and Evermann under the name of Ruvettus pretiosus Cocco, but it differs from the Atlantic species in the number of fin-rays (D. XII, 15, II; A. 16, II, instead of D. XV, 18, II; A. 17-11) and in the deeper body, the depth being 5.4 instead of 6. It has been recorded from Japan. Prometichthys Gill. {Prometheus Lowe, preoccupied.) 225. Prometichthys prometheus (Cuvier and Valenciennes). (J. & E., p. 178.) Not rare in the open sea, occasionally brought into the markets. Our speci- mens seem identical with others from Japan. The Pacific form, Pro7netichthys solandri Cuvier and Valenciennes needs comparison with material from the Atlantic. JOED AN AND JORDAN.’ FISHES OF HAWAII. 35 Gempylus Cuvier. (Lemnisoma Lesson (1830). Gempylus Cuvier (1829) has priority.) 226. Gempylus serpens Cuvier and Valenciennes. Hauliuli puM. (J. & E., p. 179.) This rare fish is known from a painting at Hilo by Andrew Garrett and one at Honolulu by Mrs. J. B. Dillingham. Whether the Pacific form, G. thyrsitoides Lesson, differs from G. serpens of the Atlantic we cannot tell. Family LXIV. CORYPHA5NIDH: (Dolphins). CoRYPH^NA Linnseus. 227. Coryphaena hippurus Linnseus. Mahihi; Mdhimdhi. (J. & E., p. 204.) Now very common in the markets. Dorsal rays 54 to 58. 228. Coryphaena equisetis Linnseus. (J. & E., p. 205.) Recorded by Bennett and by Glinther. Not seen by us. » Family LXV. NOMEID^. Ariomma Jordan and Snyder. (It is not evident that this genus differs from Cubiceps Lowe of the Atlantic.) 229. Ariomma lurida Jordan and Snyder. ( J. & E., p. 217.) Pelagic. Two specimens from the markets in Honolulu. Two casts of this rare species of the open seas are in the Bishop Museum, from examples in much better condition than the original types. 230. Ariomma evermanni Jordan and Snyder. (Jordan and Snyder, Bull. U. S. Fish Comm., XXVI, 1906, p. 209.) Open sea. Only the type, from off Honolulu, is known. Family LXVL BRAMIDHl (Sea-breams). CoLLYBHS Snyder. 231. Collybus drachme Snyder. (J. & E., p. 203.) Open sea, scarce. Originally known from several young examples, some of them from the stomach of a dolphin ( Coryphcena) . A cast of a large example is in the Bishop Museum. Eumegistus gen. nov. Jordan and Jordan. Type: Eumegistus illustris Jordan and Jordan. This genus is nearly allied to Brama, differing in its much larger scales, which, at least in the adult, are smooth, entirely without vertical ridge, or emargination. 36 MEMOIRS OF THE CARNEGIE MUSEUM. Teeth small, sharp, even, in broad bands, none on vomer or palatines. Lateral line well developed. Each ray of dorsal and anal with a series of scales; these fins falcate, the front lobe acute. Caudal deeply forked, the lobes acute. Pec- torals long, falcate. Maxillary scaly. Snout and lower jaw naked. Gill-rakers of moderate length, stiff and strong, not numerous, the number about X -|- 12, the longest about half of eye. 232. Eumegistus illustris sp. nov. Jordan and Jordan. (PI. II, fig. 1.) Type: No. 3899, C. M., Honolulu. Collector D. S. Jordan. Head 3.4 in length; depth 2; dorsal rays III, 28; anal rays II, 20; ventrals I, 5; scales 9-58-22; eye 3.5 in head; snout 4.5; maxillary 1.75. Body broadly ovate, its outlines regular; an even curve from tip of snout to dorsal, a similar curve below; caudal peduncle rather slender. Head mode ate, high above eye; preorbital narrow; maxillar}" broad, its diameter at tip two-fifths of eye, extending to below middle of the large eye; mouth very oblique, the lower jaw heavy and projecting, its tip entering the profile. Preopercle entire, evenly rounded; opercle without spine or angle; scales on head small, smaller about the eye, lower jaw and forehead scaleless or nearly so. Scales on body thick, smooth, without emargination or vertical ridge, those on sides much larger than those along bases of dorsal and anal; each ray of dorsal and anal with a series of scales, each scale broader than high; lateral line well developed, concurrent with the back; a long scaly appendage at base of ventrals, the soft rays of which fin are also scaly. Lobe of dorsal acute, 1.1 in head, 2.1 in depth of body; anal lobe 3 in depth; upper caudal lobe slightly the longer, 1.9 in depth in fin, deeply lunate, with produced tips; pectoral reaching seventeenth dorsal ray, 1.6 in depth of body; ventrals short, 3.5 in depth. Color lustrous brownish black; the edge of dorsal and anal black above the paler scales; posterior edge of caudal aliruptly white; outer edges of pectorals and ventrals also white. The type of this sjiecies is a single specimen found in the market of Honolulu. It was about two feet in length, weighing nearly nine pounds. It was regarded as one of the best food-fishes, selling at fifty cents per pound, but no one seems to have ever seen it before. On account of its great bulk the senior author was unable to take the fish as a whole, but only those parts which upon the plate are delineated in detail. The white parts of this figure were left behind to be sold by the dealer. JORDAN AND JORDAN! FISHES OF HAWAII. 37 Family LXVII. CARANGIDvE (Cavallas). The tropical species of this family are widely" spread and very closely related among themselves. Our collections from Hawaii have been sent to Mr. John T. Nichols of the American Museum of Natural History to be used in a proposed monograph of the group. The present list is therefore tentative, based mainly on the account given by Jordan and Evermann, and liable after revision to undergo considerable change. A few indications given in a letter from Mr. Nichols are here accepted, as also the identification of Caranx bixcmthopterus made in an un- published paper by Yosiro Wakiya. ScoMBEROiDES liacepede (Leather-jackets). 233. Scomberoides tolooparah (Riippell). Lae. (J. & E., p. 180.) The species doubtfully listed. under this name is common at Honolulu. 234. Scomberoides sancti-petri (Cuvier and Valenciennes). (J. & E., p. 181.) Not common; the identification uncertain. Naucrates Rafinesque. 235. Naucrates ductor (Lacepede). (J. & E., p. 182.) Very rare. The Pacific form, Naucrates indicus (Lesson), needs comparison with the pelagic form from the Atlantic. Seriola Cuvier (Amber-fishes). 236. Seriola purpurascens Temminck and Schlegel. Kahdla; Pdakahdla. (J. & E., p. 183.) Supposed to be identical with the Japanese species. 237. Seriola sparna Jenkins. Kahdla opio. (J. & E., p. 184.) Rare. One large specimen was seen in the market. It may be the same as Seriola quinqueradiata of Japan. Color plain, without lateral stripes, and the fins rather low. Elagatis Bennett (Runners). {I rex Valenciennes.) 238. Elagatis bipinnatulus (Quoy and Gaimard). (J. & E., p. 185.) One fine specimen taken by us in Honolulu. I fail to find that the Atlantic species, E. pinnatulus (Poey), differs from E. hipinyiatulus of the Pacific. Decapterus Bleeker. 239. Decapterus pinnatulus (Eydoux and Souleyet). Opelu. (J. & E., p. 186.) Very abundant in the Honolulu market in August. It is sometimes canned as ^‘Sardines.” 38 MEMOIRS OF THE CARNEGIE MUSEUM. 240. Decapterus maruadsi (Temminck & Schlegel). A large species of Decapterus is represented by two examples in • our collection from Honolulu. It has been sent to the American Museum of ‘Natural History to be studied by Mr. Nichols, who regards it as identical with D. maruadsi of Japan. The lower jaw with verj' weak teeth, mouth otherwise toothless. Length eighteen inches. Selar Bleeker. ( Trachurops Gill.) The genus Selar was based upon various slender species belonging to Trachurus, Trachurops, and Atule of other writers. The first logotype, chosen by Jordan and Evermann, was Caranx hoops Bleeker. According to Fowler this is a species of Trachurops. Selar must therefore replace the latter name. 241. Selar mauritianus (Quoy and Gaimard). Akule; Halalalu. Trachurops crumenophthalma of authors; probably not the same as the latter, which is an Atlantic species. Atule gen. nov. Jordan and Jordan. Tj'pe: Caranx affinis Rtippell. This genus has the form of Selar (Trachurops). Elongate, the back low, without the peculiar notching of the shoulder-girdle distinctive of that genus, and with the last ray of the dorsal and of the anal semi-detached, joined by a low membrane to the rest of the fin. Like Selar and Caranx it has bony plates only on the straight posterior part of the lateral line. Teeth in jaws slender, small; vomer, palatines, and tongue with minute teeth. Atule (Akule in Hawaii) is the common name of fishes of this type in Polynesia. 242. Atule lundini (Jordan & Seale). Amuka; Puakahdla. (J. & E., p. 195.) Caranx affinis Rtippell, Neue Wirbelthiere, 1838, p. 49, pi. XIV, fig. 1. Red Sea. f Selar hasselti Bleeker, Verh. Batav. Genootsch., XXIV, 1852, p. 53. Moluccas. Decapterus lundini Jordan and Seale, 'Tishes of Samoa,” 1906, p. 229. Apia. Very common at Honolulu. Mr. Nichols finds tangible differences between the form in Hawaii and Samoa and the African affinis. He regards A. lundini as a subspecies of A. affinis, of which hasselti is a synonym. 243. Atuie polita (Jenkins). Maka. (J. & E., p. 194.) A rare species at Honolulu, probably referable to this genus, though deeper in body than the type. JORDAN AND JORDAN! FISHES OF HAWAII. 39 Caranx Lacepede. (Tricropterus Rafinesque; Carangws Girard.) Under this name we include the Carangoid fishes with the teeth in the jaws not in villiform bands, teeth on vomer and palatines; back more or less elevated, but not excessively so, and none of the dorsal spines filamentous. The group has been further subdivided by authors, but not very successfully. The proper logo- type of Caranx is yet to be determined. 244. Caranx ignobilis (Forskal). Pauu’u. (J. & E., p. 188.) Carangus hippoides Jenkins. This common and widely diffused species corresponds to Caranx hippos of the Atlantic. It is known from related species by the presence of a small patch of scales on the otherwise naked breast. 245. Caranx rhabdotus (Jenkins). (J. & E., p. 193.) Carangus rhabdotus Jenkins. A small deep-bodied species, marked by dark cross-bars. Anal fin yellow. It ascends into fresh waters. It has hitherto, perhaps correctly, been identified as Caranx sexfasciatus Quoy and Gaimard. 246. Caranx melampygus Cuvier and Valenciennes. Ulua. (J. & E., p.l91.) Caranx bixanthopterus Riippell. Caranx j or steri Jordan and Evermann, non Cuvier and Valenciennes. This species, distinguished from C. ignobilis by the scaly breast, is one of the most abundant and valued food-fishes of Hawaii. There, as elsewhere throughout the South Seas, it is known as Ulua. It corresponds to Caranx latus of the Atlantic. Pectoral fin bright yellow in life, anal dusky. It has Ijeen wrongly identified with C.forsteri C. & V., a species with fewer fin-rays. Wakiya regards C. bixanthop- ierus as the same species. C. heberi has fewer fin-rays. In the original description of this species it is said: Ce poisson parait d’ailleurs avoir ete argente, et teint vers le dos d’un plombe verdatre; Les deux pointes de ses nageoires sont noirMres, mais celle de Tanale plus que I’autre.” All this applies perfectly to the Ulua, but the dusky ‘^Omilu” with the sides sprinkled with small black points, could never have been described in this way. Both the Ulua and the O^nilu have dorsal rays in increased number — D. I. 23 or 24; A. I. 19 or 20. The Ulua is known in life by its dusky anal (hence melam- pygus) and its bright yellow pectoral. 247. Caranx marginatus Gill. (J. & E., p. 191.) This species is very close to Caranx for steri, but apparently distinct. 40 MEMOIRS OF THE CARNEGIE MUSEUM. if' 248. Caranx elacate (Jordan and Evermann). (J. & E., p. 190.) Only the type is as yet known. 249. Caranx stellatus Qnoy & Gaimard. Ornilu; Omilimilu. (J. & E.,p. 192.) Caranx melampygus Gunther and ’recent authors generally (not C. melam- pygus of Cuvier and Valenciennes). Caranx punctatus Cuvier and Valenciennes (name preoccupied). Caranx cceruleopinnatus Cuvier and Valenciennes (not of Riippell). A staple food-fish, not inferior to the Ulua and reaching a much larger size Specimens seen in the market at Hilo were five feet long. It is known by its dusky coloration, the back and sides usually with scattered small black spots. This species is rather common at Honolulu, and is readily known by the traits men- tioned above. 250. Caranx thompsoni Seale. (Jordan and Evermann, “Fishes of the Hawaiian Islands,” Addenda, p. 535.) Honolulu. Only the t^qie known. 251. Caranx dasson Jordan and Snyder. Only the type known. Uraspis Bleeker. {Sele7iia Bonaparte, Cat. Method., 1843, p. 75. Type Caranx luna St. Hilaire = Scomber guara Bonnaterre; Uraspis Bleeker, Amboyna, V, 1855, p. 418 {caran- goides). The name Selenia is preoccupied. Teeth in the jaws very small, in one or two series, none on vomer or palatines. 252. Uraspis helvolus (Forster). (J. & E., p. 196.) A very rare species, taken only once at Honolulu. 253. Uraspis cheilio (Snyder). (J. & E., p. 196.) A peculiar species with depressed head, elevated back, and thick lips. De- scribed from a single large specimen. A second was obtained by us in the Hono- lulu market. C’arangoides Bleeker. We retain this name for species with small teeth in villiform bands in the jaws and on vomer and palatines. 254. Carangoides jordani sp. nov. Nichols (MS). Oinilu. We adopt the name proposed by Nichols for this common Hawaiian species, hitherto, but certainly wrongly, identified with C. ferdau of the Red Sea. 255. Carangoides gymnostethoides Bleeker. (J. & E., p. 199.) Not seen by us. JORDAN AND JORDAN: FISHES OF HAWAII. 41 256. Carangoides evermanni Nichols. One specimen placed in the hands of Mr. Nichols, who regards the Hawaiian form as a subspecies of the preceding. 257. Carangoides ajax Snyder. (J. & E., p. 200.) A huge fish of peculiar form, notable for the small number of its fin-rays. Taken but once in Honolulu. Alectis Rafinesque. 258. Alectis ciliaris (Bloch). Ulua kihikihi. (J. & E., p. 200.) Not rare. 259. Alectis indicus (Riippell). A huge example, over two feet long, looking different from the small ones called ciliaris, was taken in the market. The relation of these two forms is yet to be established. Gnathanodon Bleeker. Jaws toothless; small teeth on tongue. 260. Gnathanodon speciosus (ForskM). Pdopdo; Ulua pauu. (J. & E., p. 197.) Common in the markets of Honolulu, as well as throughout the South Seas. Family LXVIIL KUHLIID.E (Seseles). Kuhlia Gill. {Moronopsis Gill; Boulengerina Fowler, Proc. Ac. Nat. Sci. Phila., 1906, p. 572. Type Dales mato Lesson = Dules ynalo Guv. & Val.) § Kuhlia. 261. Kuhlia malo (Cuvier and Valenciennes). Aholehole. (J. & E., p. 207.) Common in all running streams and descending to estuaries. The Hawaiian fish, called sandvicensis by Steindachner, needs further comparison with the original malo from Tahiti. Although the name Dales mato of Lesson, 1830, has apparent priority over Dules malo of Cuvier and Valenciennes, nevertheless the fact that Lesson quotes the latter in sjmonymy with the correct page shows that his report on the Voyage de la Coquille is later in date than Vol. VH of the Histoire des Poissons. The reference to Boulengerina on p. 507 of Jordan’s ‘‘Genera of Fishes” is erroneous, and should be cancelled. §§ Safole Jordan. (Proc. U. S. N. AI., 1912, p. 655. Type Dales tceniurus Cuv. & Val.) 42 MEMOIRS OF THE CARNEGIE MUSEUM. 262. Kuhlia tseniura (Cuvier and Valenciennes). (J. & E., p. 208.) Known from Johnston Island, south of Hawaii. Common about lava-rocks in the South Seas; strictly marine. Kuhlia arge Jordan and Bollman from the Galapagos is probably the same. Family LXIX. APOGONID.E^ (Cardinal-fishes). Pristiapogon Klunzinger. Both limbs of preopercle serrate; gill-rakers numerous; dorsal spines usually seven; scales large; caudal fin lunate. 263. Pristiapogon menesemus (Jenkins). Upapdlu. (J. & E., p. 215.) Common about the reefs. 264. Pristiapogon snyderi (Jordan and Evermann). (J. & E., p. 214.) Apogon Jrenatus Gunther, no7i Cuvier and Valenciennes. Common about Hawaii. 265. Pristiapogon erythrinus Snyder. (J. & E., p. 217.) Rare about the reefs. Apogon Lacepede. {Amia Gronow, 1763, not binomial. Not Amia Linnaeus, 1766.) § OsTORHYNCHUs Lacepede. Like Apogo7i proper, but with seven or eight dorsal spines, instead of six; preopercle serrate on the posterior limb only; lateral line complete; scales large (about twenty-five) ; teeth on palatines; gill-rakers numerous ; caudal fin more or less lunate, not convex. 266. Apogon maculiferus Garrett. (J. & E., p. 212.) A handsome little fish, common behind the reefs. Lepidamia Gill. 267. Lepidamia evermanni (Jordan & Snyder). (J. & E., p. 213.) (Proc. U. S. N. M., XXVIII, 1905, p. 123.) One specimen known from Honolulu. Foa Jordan and Evermann. 268. Foa brachy gramma (Jenkins). (J. &E., p. 211.) Scarce, on the reefs. We ina}^ retain the name Apogon until the question of the adoption of Gronow’s non-binomial names, not validated by Scopoli in 1777, is finally settled. JORDAN AND JORDAN! FISHES OF HAWAII. •43 Apogonichthys Bleeker. Preopercle entire; lateral lin^ complete; teeth on palatines; gill-rakers numerous; caudal lunate. 269. Apogonichthys waikiki (Jordan and Evermann). (J. & E., p. 210.) A rare little fish, found on the reefs. Synagrops Gunther. {Melanostoma Steindachner and Doderlein. Preoccupied.) 270. Synagrops argyrea (Gilbert and Cramer). (J. & E., p. 218; G., p. 618.) Deep sea. Rare. Note. The genera or subgenera allied to Apogon, some of them of questionable value, are provisionally diagnosed in the following key: a. Jaws without distinct canine teeth. b. Anal fin long, its rays about II, 16; preopercle entire; dorsal spines six; scales large, about twenty-five; caudal fin forked (bleekeri) Archamia Gill. bb. Anal fin short, its rays usually II, 8. c. Preopercle distinctly serrate on one or both limbs. d. Caudal fin lunate or forked. e. Preopercle distinctly serrate on both limbs; dorsal spines seven (frenatus). Pristiapogon. ee. Preopercle serrate on posterior limb onl3^ /. V omer and palatines with teeth. g. Scales large, about twenty-five. h. Dorsal spines six {ruber)-, (Awu'o Gronow; Monopn'on Poey). Apogon. hh. Dorsal spines seven or eight {jleurieui) . . . Ostorhynchus Lacepede. gg. Scales small, thirty-five to fifty; dorsal spines six (kalosoma). Lepidamia Gill. //. Vomer and palatines toothless {-parvula) Brephamia^ Jordan. dd. Caudal fin convex, its peduncle rather long; scales large; dorsal spines six (fusca). Nectamia Jordan. cc. Preopercle rigidly entire on both limbs. i. Caudal fin rounded. j. Dorsal fins not connected at base. k. Palatines with teeth. 1. Lateral line complete. ni. Gill-rakers few and small, about six; profile before dorsal S-shaped, concave above eye; dorsal spines six. n. Scales small, about forty; tongue with small teeth (aprion). Glossamia Gill. ® Brephamia gen. non., Jordan. Type Amia parvula Radcliffe. Differing from Apogon in having no teeth on vomer or palatines. 44 MEMOIRS OF THE CARNEGIE MUSEUM. nn. Scales large, twenty-five to thirty-one (lunatus). Mionorus Krefft. mm. Gill-rakers numerous, twelve to fourteen; profile even; dorsal spines seven. 0. Scales small, about forty-five {pandionis) . Xystramia Jordan. 00. Scales large, about twenty-five (perdix). Apogonichthys Bleeker. ll. Lateral line incomplete, imperfect or wanting on caudal peduncle. Gill- rakers numerous; dorsal spines seven {hrachygramma) . Foa Jordan & Evermann. kk. Palatines without teeth; lateral line incomplete; gill-rakers few, short; a large black ocellus on opercle (aurita) .... Fowleri.-v Jordan & Evermann. jj. Dorsal fins joined at base; dorsal spines eight {octospina). Neamia Smith & Radcliffe. ii. Caudal fin lunate or forked; scales large; gill-rakers long and slender. p. Lower teeth not enlarged; bodj^ much compressed, the back elevated; dorsal spines produced (grceffi) Zoramia Jordan. pp. Lower teeth enlarged; body not greatl}^ compressed; dorsal spines six, not pro- duced (clupeoides) Rhabdamia Weber. ua. Canine teeth present; teeth on palatines; anal fin short, its rays II, 8; lateral line complete; scales large, about twenty-five; caudal lunate. q. Preopercle entire; dorsal spines six; body rather elongate {lineatus) . . Cheilodipterus Lacepede. qq. Preopercle more or less serrate; dorsal spines more than six. r. Scales cycloid; dorsal spines about nine; gill-rakers numerous, about twelve. s. Dorsal spines smooth; body more or less compressed (japonica) . .Synagrops Giinther. ss. Dorsal spines anteriorly serrate (serratospinosa) Maccullochina® Jordan. rr. Scales ctenoid; dorsal spines seven. t. Lateral line anteriorl}^ with a conspicuous row of enlarged tubules; gill-rakers few and short (tubifera) Siphamia Weber. tt. Lateral line without enlarged tubules; gill-rakers numerous (grossidens) . Amioides Smith & Radcliffe. Hynnodus Gilbert. 271. Hynnodus atherinoides Gilbert. (G., p. 618.) Deep sea. Two specimens from Pailolo Channel. ScEPTERiAS^ gen. nov. Jordan and Jordan. Type Scepterias fragilis Jordan and Jordan. {Vide infra.) Allied to E pig onus Rafinesqiie and Hynnodus Gilbert. l^ody elongate, fragile, not so slender as in Hynnodus, but more so than in ® Maccullochina gemis novum. Type Synagrops serratospinosa Radcliffe, distinguished from Syna- grops by the serrated dorsal spines. The name is proposed in honor of Mr. Allan Riverston McCulloch of the Australian Museum, one of the most accurate workers in systematic ichthyology now living. ^ From (TKeirrepLas = open-eyed, sceptical. JORDAN AND JORDAN: FISHES OF HAWAII. 45 Epigonus; the mouth larger and the fins higher; teeth small, subeqiial; pre- orbital narrow; maxillary narrow, naked, not slipping under jireorbital; pores of lateral line simple; a weak spine on opercle, head otherwise unarmed; dorsal fins well separated, the first of seven slender spines, the second short, rather high, nearly opposite anal; anal with two feeble spines; caudal deeply forked; ven- trals below pectorals; both fins rather long ; ventral rays I, 5. Scales moderate, caducous. This genus differs from Hynnodus in the deeper body, smaller scales, and higher fins. Both genera are plainly allied to Epigonus Rafinesque of the Mediter- ranean, and should constitute a subfamily, Epigoninm, within the Apogonidce. 272. Scepterias fragilis sp. nov. Jordan and Jordan. (PI. II, fig. 2.) Type No. 3900 Carnegie Museum. Honolulu. Coll. D. S. Jordan. Head 3.33 in length; depth 4.75; eye 2.5 in head; snout 5.33; maxillary 2; dorsal rays VII, 1, 10; anal rays H, 9; scales 3-54-10. Body elongate, the outlines relatively straight and parallel; head rather broad above, the profile even; mouth rather large, terminal, oblique; jaws equal; maxillary narrow, naked, reaching nearly to middle of pupil, the tip not slipping under the narrow preorbital; a row of small subequal teeth in each jaw, a patch on vomer, no teeth on palatines; preopercle entire, the rounded angle somewhat produced; cheeks scaly; opercle scaly, ending in a short weak spine ; gill-rakers j;* + 14, rather long and very slender, about half diameter of eye; pseudo- branchiae large. Scales moderate, thin, readily falling; lateral line well developed, with large pores, concurrent with back, extending on caudal fin. Dorsal spines slender, the third rather the longest, a little more than half the head, first spine moderate, one-third length of the longest. Interspace between dorsals about one- third head; second dorsal higher than long, its first ray two-thirds head; caudal deeply forked, its lobes equal, pointed, two-thirds head; anal high, similar to soft dorsal, but inserted a little farther back; pectoral pointed, reaching front of soft dorsal, 1.33 in head; ventrals inserted just below pectorals, 2 in head. Substance soft and fragile. Color plain dusky, paler below, without markings; scales with fine punctula- tions, inside of gill-cavity black. Length of type 4.6 inches. Four specimens were found in the Honolulu market, ap]iarently spewings of some large fish, perhaps Epinephelus or Etelis. *The letter x indicates that the number is uncertain and not easily counted, as they dwindle in- to rudiments above. 46 MEMOIRS OF THE CARNEGIE MUSEUM. Family LXX. SERRANIDiE (Sea-bass). PiKEA Steindachner. 273. Pikea aurora Jordan and Evermann. (J. & E., p. 220.) A rare and very handsome species. Cephalopholis Bloch and Schneider. 274. Cephalopholis argus Bloch and Schneider. (J. & E., p. 221.) A common fish in the South Seas, recorded but once from Hawaii by Quoy and Gaimard, and therefore perhaps doubtfully. Epinephelus Bloch. This genus, abundant in both the East and West Indies, is very scantily represented in Hawaii. 275. Epinephelus quernus Seale. Hapu’u pu’u. (J. & E., p. 223.) This large fish is now rather common in the markets of Hawaii. Odontanthias Bleeker. Of the Serranince none at all are found in the waters of Hawaii. The AntMine forms are, however, well represented. 276. Odontanthias fuscipinnis (Jenkins). (J. & E., p. 225.) Rather common at moderate depths. PsEUDANTHiAS Bleeker. 277. Pseudanthias kelloggi (Jordan and Evermann). (J. & E., p. 226.) Rare; found in rather deep water. Rhyacanthias Jordan. (Proc. U. S. N. M., 1921, p. 647.) 278. Rhyacanthias carlsmithi Jordan. Fig. 3. Rhyacanthias carlsmithi Jordan. (Reproduced from Proc. U. S. N. M., Vol. 59, 1921, p. 647.) JORDAN AND JORDAN: FISHES OF HAWAII. 47 From deep water, off the southwestern coast of Hawaii. The type killed by a lava-flow from Mauna Loa. Grammatonotus Gilbert. 279. Grammatonotus laysanus Gilbert. (G., p. 619.) Deep water, off Laysan. Family LXXI. PRIACANTHIDHl (Catalufas). Priacanthus Cuvier. 280. Priacanthus alalaua Jordan and Evermann. Alalaua. (J. & E., p. 228.) Rather scarce. 281. Priacanthus cruentatus (LacepMe). Aweoweo. (J. & E., p. 229.) Very abundant. The Pacific form, Priacanthus carolinus Lesson, needs further comparison with the West Indian P. cruentatus with which we have hitherto identified it. 282. Priacanthus meeki Jordan and Evermann. Ulalauau. (J. & E., p. 231.) Abundant. A food-fish of some importance. Near Priacanthus hamruhr Forskal of the Red Sea. Family LXXII. EMMELICHTHYIDHl. Erythrocles Jordan. {Erythrichthys Temminck and Schlegel, name preoccupied.) 283. Erythrocles scintillans Jordan and Thompson. (Proc. U. S. N. M., XLI, 1912, p. 599.) (J. & E., p. 245.) Rather scarce. This beautiful fish differs somewhat from its Japanese con- gener, Erythrocles schlegeli. The genus Erythrocles is close to Emmelichthys Richard- son, but probably distinct. Boaxodon cyanescens from Chile, having a broad scaly maxillary, is closely related, but Inermia vittata from the West Indies and Diptery- gonotus leucogrammicus from the East Indies cannot be placed in the same family, having the maxillary narrow and naked. Family LXXIII. HISTIOPTERIDHl. Histiopterus Temminck and Schlegel. 284. Histiopterus typus Bleeker. A cast of a fine specimen of this large fish, otherwise only known from Japan, is in the Bishop Museum. 48 MEMOIRS OP THE CARNEGIE MUSEUM. Family LXXIV. LUTIANID.411 (Snappers). All the Hawaiian species of this familj^ belong to the aberrant group of Etelince, distinguished in several ways from the typical members of the family, but es- pecially by the scaleless dorsal and anal fins; and most of them by the broad flattish cranium. Rooseveltia Jordan and Evermann. In this genus the body is relatively deep, the canines strong, the tongue tooth- less, and the pectoral falcate. The typical species was at first referred to Serranus Cuvier, to which genus it bears little resemblance, and afterwards to Apsilus Cuvier, to which it is closely related. 285. Rooseveltia brighami (Seale). Ukikiki; Kalikali. (J. & E., p. 233.) This beautiful fish, one of the handsomest found in Hawaii, light crimson in color, marked with three broad golden cross-bands, is now common in the markets, as the Japanese fishermen operate in deeper water than the Hawaiians, whom as fishermen they have now succeeded. 286. Rooseveltia aloha Jordan and Snyder. Known only from the original type. Pristipomoides Bleeker. ( Platyinius Gill {vorax = macrophthalmus) ; Bower sia Jordan & Evermann.) We are unable to separate the Hawaiian species, called Bowersia, from the East Indian genus Pristipomoides. The only difference of any importance is in the slenderer body of the Hawaiian species. The West Indian form called Platyinius is equally close, the body being a little deeper than in either of the others. Pristi- pomoides sparus and P. microlepis seem to be genuine members of this genus. Sparopsis Kner, referred to the synonymy of Pristipofnoides bj^ Bleeker, belongs to the Denticinoe and to the genus or subgenus Synagris {Anemura Fowler) AXied io N emipterus. Pn'sb'pomofdcs has canines in both jaws; no filamentous spines ; no teeth on tongue; last ray of dorsal and of anal elongate; pectoral long, falcate; scales relatively large, about sixty. This genus and the next are offshoots from Aprion, to which both are closely related. The account of the teeth of Bowersia violescens by Jordan and Evermann is not correct, as the tongue is toothless. 287. Pristipomoides violescens (Jordan and Evermann). Opakapaka. (J. & E., pp. 234, 236.) Apsilus microdon, as described by Jordan and Evermann, is the young of this species. Steindachner’s fish was, however, Ulaula sieboldi. JORDAN AND JORDAN! FISHES OF HAWAII. 49 Ulaula Jordan and Thompson. (Jordan and Thompson, Proc. U. S. N. M., XXXIX, 1911, p. 439. Type Bower sia ulaula Jordan and Evermann = Chcetopterus sieboldi Bleeker, the name Chcetopterus preoccupied.) In this group, or subgenus, there are no canines; tongue with small teeth; pectoral falcate; mouth small. The name Ulaula, meaning “very red,” belongs properly to Etelis evurus. 288. Ulaula sieboldi (Bleeker). Kod’e. (J. & E., p. 237.) {Aprion microdon Steindachner.) This, like the preceding and the next two species, is a common food-fish of Hawaii, and, having the same olive-gray color with purplish reflections, they are often confused in the markets. We are not able to distinguish the Hawaiian form U. microdon (Steindachner) from Japanese specimens of U. sieboldi = Chcetopterus dubius Gunther. Aprion Cuvier and Valenciennes. Canines present; no teeth on tongue; pectorals very short; body elongate; scales large. The synonymy of this genus, as given by Jordan and Evermann, contains several errors. 289. Aprion virescens Cuvier and Valenciennes. Uku. (J. & E., p. 239.) This species, one of the most abundant and highly valued of the Hawaiian food-fishes, reaches a much larger size than the three just mentioned, attaining a length of three feet or more. Etelinus Jordan and Thompson. (Jordan and Thompson, Proc. U. S. N. M., XXXIX, 1911. Type Etelis niarshi Jenkins.) This genus has the notched dorsal and crimson colors of Etelis with the general form and dentition of Pristipomoides. The resemblance of the genus to the Japan- ese Ddderleinia is extremely close, although the latter, having a broad scaly maxil- lary, not slipping under the preorbital, must be placed in a different family in or near the Serranidce. 290. Etelinus marshi (Jenkins.) Ulaula. (J. & E., p. 240.) A common and valued food-fish. Etelis Cuvier and Valenciennes. Body elongate; dorsal deeply notched; caudal broadly forked; pectoral rather short; canines present; no teeth on tongue. Color deep crimson. 50 MEMOIRS OF THE CARNEGIE MUSEUM. 291. Etelis evurus Jordan and Evermann. Ulaula. (J. & E., p. 242.) This superb species, reaching a length of three feet, is now common in the markets, being taken in rather deep water. It is close to the West Indian Etelis oculatus and needs further comparison with Etelis carhunculus of the lie de France. The genera of the Etelince. have been much confused and misunderstood. They may be defined as follows : a. Etelin.e. Cranium solid; skeleton firm; dorsals connected; soft dorsal and anal scaleless; last ray of dorsal and anal more or less produced; scales above lateral line in rows parallel with the lateral line. b. Dorsal fin continuous, not deeply notched or divided. c. Cranium not flat above, much as in Lutianus; the interorbital area not separated from the occipital region, the median and lateral crests procurrent on it; frontal narrowed anteriorly; body rather deep. d. Canines none. Tongue with small teeth. e. Pectoral fins very short, shorter than ventrals; color dull olivaceous (fuscus). Apsilus. ee. Pectoral fins rather long, falcate (macrophthalmus) Tropidinius. dd. Canine teeth well developed; no teeth on tongue; skull thick, with three blunt ridges separated by narrow grooves; color red and golden (hrighami) Rooseveltia. cc. Cranium flat above, much as in Etelis; the interorbital area separated from the occipital region by a transverse line of demarcation, the median and lateral crests not procurrent on it; frontal broad anteriorly. /. Pectoral fin long, falcate. g. Canine teeth present; no teeth on tongue (iypus) Pristipomoides. gg. Canine teeth obsolete; tongue with a patch of very small teeth (sieboldi) . . Ulaula . //. Pectoral fin short, not falcate, formed as in Apsilus; body elongate; preorbital very broad {virescens) Aprion. bb. Dorsal fin divided or deeply notched; cranium broad, flattish, the median and lateral crests not procurrent on it; color red. h. Maxillary scaly; body elongate; canines strong. i. Caudal fin moderately forked; gill-rakers rather few (marshi) Etelinus. ii. Caudal fin deeply forked, the lobes produced; gill-rakers slender, numerous {car- buncuhis) Etelis. hh. Maxillary naked; body compressed; canines none'; gill-rakers slender (aquilionaris) . Etelides. aa. VERILINA5. Cranium cavernous; skeleton soft; form not elongate; dorsal divided to its base; second dorsal scaly at base; color black; deep-sea forms (sordidus) Verilus. The Japanese genus Ddderleinia {Eteliscus Jordan and Snyder) must stand very near to the Anthiince. The genus Verilus Poey, a deep-sea form, black in color, with cavernous skull and soft skeleton, should constitute a distinct sub- family, Verilince, allied to the Etelince. The dorsal fin is divided into two, and the second dorsal is scaly at the base. JOED AN AND JOEDAN! FISHES OF HAWAII. 51 Family LXXV. APHAREID.^. This family, allied to the Lutianidce and especially to the Etelince, differs in having no teeth on vomer or palatines; those of the jaws are very small. Aphaeeus Cuvier and Valenciennes. 292. Aphareus furcatus Cuvier and Valenciennes. (J. & E., p. 235.) Aphareus flavivultus Jenkins. This species seems to be widely distributed, but nowhere common. The type of A. flavivultus had the top of the head and forehead bright yellow, the fish being otherwise dull brownish purple. Our specimens do not show the yellow, which fades in spirits. Family LXXVL SPARID.E (Porgies). Monotaxis Bennett. (Sphcerodon Gunther.) 293. Monotaxis grandoculis (Forskal). Mu; Mamdmu. (J. & E., p. 243.) Rather common. Family LXXVII. KYPHOSID.T] (Rudder-fishes). Kyphosus Lacepede. 294. Kyphosus elegans (Peters). Nenue paiii. (J. & E., p. 247.) Kyphosus sandvicensis (Sauvage). Not rare about Honolulu. It seems to be identical with Kyphosus elegans (Peters) from Mazatlan. 295. Kyphosus fuscus (Lacepede). Manaloa; Nenue. (J. & E., p. 248.) Not rare at Honolulu. Sectatoe Jordan and Fesler. 296. Sectator azureus Jordan and Evermann. (J. & E., p. 248.) A beautiful fish, of which but one specimen is as yet known. Unknown to the fishermen. Family LXXVHI. A1ULLID.T: (Surmullets). Mulloides Bleeker. 297. Mulloides auriflamma (Forskal). Wekeula. (J. & E., p. 250.) Rather common. 298. Mulloides erythrinus Klunzinger. (J, & E., p. 251.) Recorded from Laysan Island. 52 MEMOIRS OF THE CARNEGIE MUSEUM. 299. Mulloides pflugeri Steindachner. Weke ula ula. (J. & E., p. 251.) Now rather common in the markets. Mulloides flammeus Jordan and Ever- mann is probably the young of this species. 300. Mulloides samoensis Giinther. Weke; Weke a’a. (J. & E., p. 253.) Not rare about Honolulu 301. Mulloides preorbitalis (Smith and Swain). (J. & E., p. 264.) Johnston Island. Occasional at Honolulu. 302. Mulloides vanicolensis (Cuvier and Valenciennes). (J. & E., p. 254.) South Seas. Recorded from Johnston Island. Upeneus Cuvier. ( Pseudupeneus Bleeker.) 303. Upeneus porphyreus (Jenkins). Kumu. (J. & E., pp. 261-262.) The largest species of the genus, constantly in the markets, and justly highly valued as food. Teeth very small; barbel short; characters which give the appear- ance of Mulloides. The specimen from Honolulu, recorded by Steindachner as U. fraterculus, is i)robably U. porphyreus. 304. Upeneus chryserydros (Lacepede). Moana kea. (J. & E., p. 255.) Not rare at Honolulu. 305. Upeneus multifasciatus (Quoy and Gaimard). Moana. (J & E., p. 256.) Very common in the markets, but apparently limited to the Hawaiian Islands; replaced in Polynesia by U. moana Jordan and Seale, a very similar species. 306. Upeneus bifasciatus Lacepede. Munu. (J. & E., p. 258.) Rather common about Honolulu. 307. Upeneus chrysonemus (Jordan and Evermann). (J. & E., p. 258.) Common. Known by the yellow barbels. 308. Upeneus crassilabris (Cuvier and Valenciennes). (J. & E., p. 259.) South Seas. Found at Johnston Island. 309. Upeneus pleurostigma (Bennett). (J. & E., p. 260.) Common about Honolulu. Upeneoides Bleeker. ( Upeneus Bleeker, not of Cuvier, as restricted by the first reviser.) 310. Upeneoides arge^ (Jordan and Evermann). Weke pueo; Weke pahula. (J. & E., p. 264.) Very abundant; close to Upeneus vittatus of the South Seas. ® Upeneoides tceniopterus (Cuvier and Valenciennes), an Indian species, was recorded from Honolulu by Steindachner, who mistook for it the young of U. arge. JORDAN AND JORDAN! FISHES OF HAWAII. 53 Family LXXIX. MALACANTHID^. Malacanthus Cuvier. 311. Malacanthus parvipinnis Vaillant and Sauvage. Makd’a. (J. & E., p. 275.) Common about Honolulu. Suborder CIRRHITIFORMES. Family LXXX. CHEILODACTYLIDHl. Goniistius Gill. This genus differs from Cheilodactylus mainly in the number of fin-rays (D. XVII, 27-32; A. Ill, 8: instead of D. XVIII, 23; A. Ill, 11). The outline of the dorsal is much more strongly angulated. 312. Goniistius vittatus (Garrett). Kikakapu. (J. & E., p. 447.) Two fine specimens of this very rare species were found by us in the Honolulu market. The name vittatus is ill-suited to the broad, oblique, black cross-bands, which are characteristic of this species. Family LXXXI. CIRRHITIDA5. CiRRHiTOiDEA Jenkins. 313. Cirrhitoidea bimacula Jenkins. (J. & E., p. 448.) Rare. Paracirrhites Bleeker. 314. Paracirrhites cinctus (Gunther). Piliko’a; Puopa’a; Oopuka-hai-hai. (J. & E., p. 449.) Very common. 315. Paracirrhites forsteri (Bloch and Schneider). Hilupilikoa. (J. &E.,p. 450.) Abundant. 316. Paracirrhites arcatus (Cuvier and Valenciennes). Piliko’a. (J. &E.,p.450.) Very abundant. We have no explanation of the two patterns of coloration; about half of the specimens having a broad, well-defined white stripe along the back posteriorly, while in others, similarly colored, this is absent. CiRRHiTUS Lacepede. 317. Cirrhitus marmoratus (LacepMe). Po^opda; Oopukdi. (J. & E., p. 452.) Abundant; large enough to acquire importance as a food-fish. 54 MEMOIRS OF THE CARNEGIE MUSEUM. Suborder PAREIOPLITM. ( Loricati.) Family LXXXII. CARACANTHIDiE. Caracanthus Kroyer. 318. Caracanthus maculatus (Gray). (J. & E., p. 453.) Scarce; about the reefs. Amphiprionichthys Bleeker. This genus differs from Caracanthus in having the dorsal fins hilly united. 319. Amphiprionichthys unipinna (Gray). (J. & E., p. 454.) A rare fish of the reefs. Family LXXXIII. SCORPaENIDAH (Scorpion-fishes; Rock-cod). Sebastapistes Gill. 320. Sebastapistes ballieui (Sauvage). Poopa’a. (J. & E., p. 455.) Rather common. 321. Sebastapistes corallicola Jenkins. (J. & E., p. 455.) Three specimens known. 322. Sebastapistes asperella (Bennett). (J. & E., p. 458.) Not recognized since recorded by Bennett. 323. Sebastapistes coniorta Jenkins. (J. & E., p. 458.) Common on the reefs. 324. Sebastapistes galactacma Jenkins. (J. & E., p. 459.) Common on the reefs. 325. Sebastapistes coloratus Gilbert. (G., p. 627.) Off Molokai, in deeper water. ScoRP^NODES Bleeker. (Sebastopsis Gill, and likewise Sauvage.) 326. Scorpaenodes kelloggi (Jenkins). (J. & E., p. 462.) Common on the reefs. 327. Scorpsenodes parvipinnis (Garrett). (J. & E., p. 463.) Very rare. Helicolenus Goode and Bean. 328. Helicolenus rufescens Gilbert. (G., p. 631.) Off Kauai, in deep water. JORDAN AND JORDAN! FISHES OF HAWAII. PoNTiNUS Poey. 329. Pontinus spilistius Gilbert. (G., p. 633.) Off Maui. Merinthe Snyder. 330. Merinthe macrocephala (Sauvage). Oopu kai Nohu. (J. & E., p. 461.) A beautiful fish, reaching a weight of about six pounds, now common in the markets, being taken in rather deep water. Setarches Johnson. 331. Setarches remiger Gilbert and Cramer. (G., p. 634.) Common in deep water. Plectrogenium Gilbert. 332. Plectrogenium nanum Gilbert. (G., p. 634.) ScoRP^NOPSis Heckel. 333. Scorpaenopsis gibbosa (Bloch and Schneider). Nohu; Omakaha. (J. & E., p. 468.) Scorpcenopsis catocala Jordan and Evermann. Abundant; known by the variegated breast. 334. Scorpaenopsis cacopsis Jenkins. (J. & E., p. 467.) Not rare. 335. Scorpaenopsis altirostris Gilbert. (G., p. 628.) Off Alolokai. Perhaps type of a distinct genus, the head not being depressed as in Scorpcenopsis, and the general appearance more like that of Sebastapistes. Peloropsis Gilbert. 336. Peloropsis xenops Gilbert. (G., p. 630.) Avan Channel between Alaui and Lanai. Iracundus Jordan and Evermann. 337. Iracundus signifer Jordan and Evermann. (J. & E., p. 470.) A rare fish of the coral-reefs. But two specimens are known. T^nianotus Lacepede. 338. Tsenianotus garretti Gunther. (J. & E., p. 471.) Known only from a drawing. 339. Tsenianotus citrinellus Gilbert. (G., p. 636.) Off Molokai. 56 MEMOIRS OF THE CARNEGIE MUSEUM. Brachirus Swainson, Dendrochirus Swainson. Unfortunately Brachirus has priority. Later Swain- son transferred the name Brachirus to a genus of Soles. Pectorals with the upper rays branched. 340. Brachirus barberi (Steindachner). (J. & E., p. 465.) Dendrochirus hudsoni Jordan and Evermann. 341. Brachirus chloreus Jenkins. (J. & E., p. 465.) Occasional about the coral-reefs. Pterois Cuvier. 342. Pterois sphex Jordan and Evermann. (J. & E., p. 464.) Taken but once at Honolulu. Family LXXXIV. BEMBRADID^. Bembradium Gilbert. 343. Bembradium roseum Gilbert. (G., p. 637.) Deep water; Pailolo Channel. Family LXXXV. PERISTEDIIDaE. Peristedion Lacepede. 344. Peristedion engyceros Gunther. Fig. 4. Peristedion engyceros Glinther. (Reproduced from Jordan, Proc. U. S. N. M., Vol. 59, 1921, p. 654.) Rare, in deep water. Besides the original type, found half-dried on the beach, we have found one specimen from the deeji sea and one killed in a lava-flow from Mauna Loa. There is also a cast in the Bishop Museum. The species may be distinguished from the next by the divergence of the long proboscideal horns and by the presence of dark cross-bands. JORDAN AND JORDAN: FISHES OF HAWAII, 57 345. Peristedion gilberti Jordan. (G., p. 639.) (Proc. U. S. N. M., LIX, 1921, p. 655, Peristedion engyceros Gilbert, not of Gtinther.) / Horns rigidly parallel; color red, often with small round olive spots. Found in the deep sea, abundantly in places. 346. Peristedion hians Gilbert and Cramer, (G., p. 638.) Frequent in deep water. Family LXXXVI. HOPLICHTHYIDHl. Hoplichthys Cuvier and Valenciennes. 347. Hoplichthys citrinus Gilbert. (G., p. 640.) Deep sea; abundant. 348. Hoplichthys platophrys Gilbert. (G., p. 642.) Deep sea off Laysan; only one specimen known. Family LXXXVII. CEPHALACANTHID/E (Flying Gurnards). ( Dactylopteridce. ) Dactyloptena Jordan and Richardson. (Proc. U. S. N. M., XXXHI, 1909, p. 665.) 349. Dactyloptena orientalis (Cuvier and Valenciennes). Lolo-oau. (J. & E., p. 473.) Rather scarce. We are thus far unable to separate the Hawaiian Fljdng Gurnard from the common Japanese species. The description of this species by Jordan and Richardson, 1. c., is from examples from Hilo. 58 MEMOIRS OF THE CARNEGIE MUSEUM. Suborder SQU AMI PE N N ES. Family LXXXVIII. CAPROIDiE. Antigonia Lowe. 350. Antigonia steindachneri Jordan and Evermann. (J. & E., p. 361 ; G., p. 621.) A Japanese fish, rarely seen about Hawaii in deep water. 351. Antigonia eos Gilbert. (G., p. 621.) Deep sea; Pailolo Channel. Family LXXXIX. CHAHTODONTIDvE (Butterfly-fishes). Forcipiger Jordan and McGregor. 352. Forcipiger longirostris (Broussonet). (J. & E,, p. 363.) Forcipiger flavissimus Jordan and McGregor. Rather common in the markets. In the plate published by Jordan and Evermann the brilliant yellow of this curious fish is not well represented. Ch^todon Linnaeus. Kihi kihi. {Tetragonoptrus Bleeker.) This genus covers a great variety of species agreeing in general form and in bright coloration, mainly yellow with black cross-bands or markings. It may perhaps be divisible into several genera, the division being based on the direction of the lines of scales, the size of the scales, and the form of the head. At present the subgenera, as proposed by Kaup and by Bleeker, are too ill-defined to permit of their recognition as genera. In the typical section of Chcetodon (type Choetodon capistratus Linnaeus), which is not represented in Hawaii, the lines of scales above the lateral line extend upward and backward, those below downward and back- ward, and none of the dorsal rays are prolonged. § Linophora Kaup. (One of the dorsal rays prolonged, whip-like, otherwise essentially as in Choetodon proper.) 353. Chsetodon setifer Bloch. Kihi-kihi; Kikakdpu. (J. & E., p. 364.) Common in Hawaii. § OxYCH^TODON Bleeker. (Scales of sides enlarged; snout sharp; a projection before eye.) 354. Chsetodon lineolatus Cuvier and Valenciennes. (J. & E., p. 365.) Rather rare. JOED AN AND JOED AN I FISHES OF HAWAII. 59 § Ch^todontops Bleeker. (Scales of sides moderately enlarged; snout moderate; profile even, no convexity before eye.) 355. Chsetodon lunula (Lacepede). Kikakdpu. (J. & E., p. 366.) Very common. The young have higher fins and a black dorsal ocellus. § Lepidoch^todon Bleeker. (Scales of sides anteriorly much enlarged, the rows nearly horizontal; teeth large; snout short; profile steep.) 356. Chaetodon unimactilatus Bloch. Kikikdpu. (J. & E., p. 368.) Chcetodon sphenospilus Jenkins. Not rare about the reefs. § ClTHAEGEDUS Kaup. (Snout short and blunt; scales moderate, the rows nearly horizontal.) 357. Chaetodon ornatissimus Solander. Kikikdpu. (J. & E., p. 373.) Rather rare; about the reefs. 358. Chaetodon punctatofasciatus Cuvier and Valenciennes. (J. & E., p. 369.) Not rare. § Rabdophoeus Kaup. (Scales subequal, moderate, arranged in series mostly horizontal, those above lateral line much reduced in size; snout rather short.) 359. Chaetodon ephippium Cuvier and Valenciennes. A common and showy species of the South Seas, once found in the Honolulu market. 360. Chaetodon fremblii Bennett. (J. & E., p. 375.) A handsome, but rather rare species. 361. Chaetodon trifasciatus Mungo Park. (J. & E., p. 372.) A Polynesian species, rather rare at Honolulu. The colored plate of Jordan and Evermann is from a Samoan example. This species and the next have rather large scales (forty, instead of fifty to sixty), thus approaching the next genus. 362. Chaetodon miliaris Quoy and Gaimard. (J. & E., p. 371.) Chcetodon mantelliger Jenkins. Generally common. The smallest species. 363. Chaetodon quadrimaculatus Gray. (J. & E., p. 373.) Rather common about Honolulu. 60 MEMOIRS OF THE CARNEGIE MUSEUM. Tifia gen. nov. Jordan. Type Choetodon corallicola Snyder. This group is distinguished from Rabdophorus by the very large scales above as well as below the lateral line, about thirty in a lengthwise series, and arranged in nearly horizontal rows. Teeth very small; snout moderately acute; the profile straight. The name Tifi-tifi {Kihi-kihi in Hawaiian) is applied to all species of Choetodon throughout the South Seas. The verb tifi is to adorn, and the name is given to the horns of the moon, the tips of the wing of a bird, and the like. 364. Tifia corallicola (Snyder). (J. & E., p. 374.) Rare. Found in rather deep water. Fig. 6. Loa excelsa Jordan. (Reproduced from Proc. U. S. N. M., Vol. 59, 1921, p. 653.) JORDAN AND JORDAN! FISHES OF HAWAII. 61 Loa Jordan. Scales small, even, the rows nearly horizontal; the first dorsal nearly scaleless, its first three spines thickened, the third and fourth very high. 365. Loa excelsa Jordan. (Jordan, Proc. U. S. N. M., LIX, 1921, p. 652, fig. 6.) Known from a single small specimen killed in a lava-flow from Maima Loa in rather deep water. This young fish bears some resemblance to the young of Cha’todon lunula figured bj^ Jordan and Evermann, but the dorsal spines are much longer and larger and the black markings are different. Micr ACANTHUS® SwainsoH. Dorsal and anal fins with few rays (D. XI, 17; A. Ill, 14). Scales small, about sixty. 366. Micracanthus strigatus (Cuvier and Valenciennes). (J. & E., p. 376.) A Japanese fish, occasionally taken at Honolulu. Heniochus Cuvier and Valenciennes. {Diphreutes Cantor, there being an earlier genus Henioche.) 367. Heniochus macrolepidotus (Linnaeus). (J. & E., p. 376.) Chcetodon acuminatus Linnaeus, this name having two pages priority, but the later and most frequently employed name is preferred by the International Com- mission of Nomenclature. A common fish of the Pacific, but rather rare at Honolulu. Holacanthus Lacepede. None of the Hawaiian species are at all closely related to the t^qie of this genus, Holacanthus tricolor, of the West Indies. In the typical group the scales are of moderate size; the lobes of the lunate caudal fin produced in long streamers. The numerous species of the South Seas need to be critically compared before the several subgenera proposed by Bleeker can be fully defined. Cha^todontoplus Bleeker. (Scales small; caudal fin rounded; suborbital entire; scales above lateral line small.) 368. Chaetodontoplus bicolor (Bloch). (J. & E., p. 380.) Common in Polynesia; recorded by Gimther from Hawaii. ® Misprinted Microcanthus by Swainson. 62 MEMOIRS OF THE CARNEGIE MUSEUM. 369. Chsetodontoplus arcuatus (Gray). (J. & E., p. 378.) Described in 1831 from Honolulu, but only the type is as yet knoMTi. Centropyge Kaup. Caudal rounded; scales large above as well as below lateral line, thirty to forty in a series; suborbital more or less serrate behind, but without strong spine. The original type of the genus, Holacanthus tibicen, was originally described as having four anal spines. Species of small size and variegated coloration. 370. Centropyge potteri (Jordan and Metz). (PI. Ill, fig. 1.) Holacanthus potteri Jordan and Metz, Proc. U. S. N. M., XLII, 1912, p. 525. Only the type of this dainty species of the reefs was known, until the senior author secured four additional examples from the reef at Honolulu. The largest of these, differing in certain minor respects from the original type, is figured on Plate HI, fig. 1. The serrations on the preorbital and preopercle are larger than in the type; the cross-streakings somewhat different. The black blotches above the pectoral have disappeared, and the posterior part of the body is suffused with deep inky purple, almost black, a color which obscures the vermiculations, although these can be traced. In life the dark streaks were dark bluish purple, the paler colors a rich light orange. 371. Centropyge tutuilse sp. nov. Jordan and Jordan. (J. &E., p. 378, PI. LVI. not description.) (PI. Ill, fig. 2.) Holacanthus bispinosus Gunther, Fische der Siidsee, H, 1874, p. 51, PL LVI, fig. C. Description in part. Not Holacanthus bispinosus Gunther, Cat. Fishes, II, 1860, p. 48, which is based on Bleeker’s description of Hola- canthus diacanthus from Amboyna. Holacanthus bispinosus Jordan and Evermann, Fishes of Hawaii, PI. LVI, not description, which was taken from two specimens of Centropyge diacantha. Type: No. 3902 Carnegie Museum. Three inches long. Tutuila, Samoa; paratype. No. 8750, Stanford University. Plead 4; depth 1.75; eye 3.5 in head; snout 4.25; maxillary 4.25; dorsal rays XIV, 17; anal rays III, 15; scales 35, 18; the number of fin-rays can not be exactly counted, nor the number of scales, as on the caudal peduncle these grow very small. Body broadly ovate, evenly rounded, the steep profile somewhat gibbous over the eye. Preorbital narrow, with two rather strong, blunt teeth; vertical line of JORDAN AND JORDAN: FISHES OF HAWAII. 63 preopercle with small serrse; angle with a stout curved spine reaching nearly to base of pectoral; a sharp spine about one fifth as long just below it. Scales on sides large, those above lateral line much smaller, merging into the scaly sheath of the dorsal fin; scales on belly much smaller; lateral line ending below last ray of dorsal; dorsal and anal scaly almost to their tips. Dorsal and anal rather high ; the tips angular, reaching beyond base of the rounded caudal ; ventrals filamentous, almost reaching front of anal, as long as head; pectorals a little shorter. The colors in life, as correctly stated by Jordan and Seale, “Fishes of Samoa,” p. 348, are as follows: Ground-color deep orange, or copper-red, clearer below; the head, back, and vertical fins blue-black, the color forming about eighteen narrow cross-streaks on side as wide as the ground-color; breast and belly orange; the lips and spines violet; lower lip very bright blue; anal and dorsal edged with blue; caudal with a broader blue stripe inside the margin; pectoral yellow, dusky at base; ventral orange, edged with blackish in spines, coppery red, more or less faded, with about twenty vertical black cross-bars a shade wider than the interspaces, which are about equal to the pupil; breast plain light orange, bars growing irregular below; without distinct markings; dorsal, anal, and caudal black, unmarked (dorsal and anal figured by Gunther with small blue spots); pectorals and ventrals pale, the ventral filament edged with black. This handsome little fish has had a rather unfortunate fate in the synonymy. It was first noticed by Gunther (“Fische der Siidsee,” II, p. 51, Tafel LVI), under the name of Holacanthus bispinosus, a name originally given by Gunther to a speci- men described by Bleeker from Amboyna under the erroneous name of Holacanthus diacanthus. But Bleeker’s fish and the present one, as shown by Bleeker’s figure, differ in color and in the armature of the preopercle. The true bispinosus, as figured, has much stronger spines on the lower limb of the preopercle, and the suborbital is very strongly serrate. The color is also very different, being yellow, with regular brown cross-bands, the caudal pale. Gunther claims to have had specimens in the British Museum from the New Hebrides, and one, in bad condition, from Hawaii, collected by Garrett. Presumably his figure, which represents, but none too well, o\xv C entropy ge tutuilce, was drawn from a New Hebrides example, colored after a sketch made by Parkinson in Tahiti. The only Hawaiian record is that of Gunther, mentioned above. Two speci- mens were obtained by Jordan and Kellogg at Pago Pago, Tutuila. One of these is the type of Centropyge tutuilce, and served also for the colored plate drawn by 64 MEMOIRS OF THE CARNEGIE MUSEUM. Morita and published by Jordan and Evermann under the name of Holacanthus bispinosus. But the description published by Jordan and Evermann (“ Fishes of Hawaii,” p. 378) was not taken from this species, but through some error, for which I cannot at present account, from two Samoan examples of Centropyge diacantha (Bloch). It is not clear that Gunther’s description (“ Fische der Siidsee”) belongs to the fish figured by him. Centropyge tutuilce is, therefore, until now known only from two colored plates, the first that of Gunther, indifferent in quality, the other that of Morita, which is excellent. In Samoa this species is known as Tuu’u pulepule mumu broad fish, red- striped. XiPHYPOPs’^® gen. nov. Jordan. Type: Holacanthus fisheri Snyder. Distinguished by the presence of two strong spines besides smaller serrse on the suborbital bone. The preopercle is also strongly armed. Scales large, those above the lateral line scarcely reduced; caudal rounded; profile convex; fourteen dorsal spines. 372. Xiphypops fisheri (Snyder). (J. & E., p. 379.) A handsome fish, taken a few times in rather deep water. Family XC. ZANCLIDA5 (Moorish Idols). Zanclus Cuvier and Valenciennes. 373. Zanclus cornutus (Linnaeus). Kihikihi. (J. & E., p. 382.) Very common about the reefs. Zanclus canescens Linnaeus is thought by Bleeker to be a distinct species, having a spine on the preorbital and no black markings before the eye. It may be, as the writer has supposed, the young of the common Za7iclus cornutus. The name canescens has one page priority over cornutus. 374. Zanclus ruthise Bryan. (Bryan, Report Bernice Pauahi Bishop Museum, II, 1905, p. 22, fig. 2 (1906).) A single young specimen taken at Honolulu, two and three-quarters of an inch long, remarkable for the great height of the first dorsal rays. The color is quite unlike that of Z. cornutus, young or old, there being only a faint dark bar across the interorbital and a broad obscure dark shade across body from dorsal to anal, and another on caudal peduncle. Caudal mostly black, as are the long rays of dorsal and the front of the anal and ventrals; lips black; tip of caudal pale; profile very steep; depth nearly equal to length. D. VII, 38; A. HI, p. 33. ^tS/w6ere-ef = Numb-ray. Misaki (Aoki). Family PLATYRHINID^. Genus Platyrhina Muller and Henle. If the name Platyrhina (1838) is regarded as distinct from Platyrhinus, the substitute names Analithis Gistel (1848) and Discobatus Garman (1880) are not needed. 35. [72] Platyrhina sinensis (Bloch and Schneider). Uchiwa-zame = Fsiii-sha,rk. We have a specimen from Mikawa Bay (M. Ishikawa). Another from Sagami Bay was seen in the Yamada Museum. Family RAJIDtE. Genus Raja Linnaeus. Key to the Japanese Species of Raja. a. Upper parts wholly covered with spinules Raja isotrachysP aa. Upper parts mostly, but not wholly, beset with spinules Raja kujiensisP aaa. Upper parts largely smooth. h. Snout little or moderately produced, its length from eye much less than the distance from spiracle to inner angle of pectoral fin; front mai’gin of disc little concave. ^®Giinther, Challenger Reports, Deep Sea Fishes, 1887, p. 7, pi. 3. ^Allied to Raja isotrachys is Raja kujiensis Tanaka, Dobutsu-gaku Zasshi, XXVIII, No. 331, May 15, 1916, p. 173. An English translation of the Japanese description has been kindly supplied by Dr. Tanaka: “Disk rhomboid, wider than long (including ventral fins); snout slightly pointed, its tip making an angle of one hundred and forty degrees; the disk weakly crenulate in anterior margins, rounded at outer angle, making an angle of about one hundred degrees; posterior angles of disk also rounded, making an angle of about ninety degrees; longer diameter of eye shorter than interorbital width; length of snout (measured to front of eye) about two times interorbital width; spiracle shorter than diameter of eye; teeth in each jaw in twenty-five rows; two dorsals well apart. Body uniformly scattered with small spines, the spines very few in a small area behind spiracle, none on the border of eye only; large spines arranged in a single row from behind eye to second dorsal on the middle line of dorsal surface; near centre of disk a pair of rather large spines on either side of the spines of mid-dorsal line; lower surface entirely smooth. Color in formalin purplish brown, with a few black spots, which are scattered irregularly and unsymmetrically in relation to the mid-dorsal line; under surface dead-white, with but few dusky spots; margin dusky gray. Locality: Kuji, in Hitachi, northeast of Tokyo. Length: 825 mm.” no MEMOIRS OF THE CARNEGIE MUSEUM. c. Snout scarcely produced, its length equal to distance between outer edges of spiracles; line joining extreme angle of pectorals dividing the disc into two equal parts; body covered with black spots. Rostral cartilages united for more than half their length; caudal fin developed; lateral folds on tail conspicuous, extending to middle of caudal fin Raja fusca. cc. Snout moderately produced, its length much greater than distance between outer edges of spiracles; line joining extreme angles of pectorals dividing the disc into a larger more attenuate anterior part and a smaller posterior portion; body without black spots. d. Rostral cartilages united for only one-third their length; caudal fin barely evident. Lateral folds of tail very wide, extending almost to extreme tip of tail; tail depressed throughout; first dorsal separated from the second by more than half its basal length; a single row of spines on top of tail in specimens as long as 385 mm.; body unmarked, except for the two pectoral ocelli of the young, and faint lighter spots Raja smirnovi. dd. Rostral cartilages united for half their length; caudal fin rather high. e. Lateral folds of tail broad, extending to middle of caudal fin; tail de- pressed throughout; first dorsal separated from second by less than half its basal length; series of spines on top of tail multiple in speci- mens as small as 25 cm.; no spines developed on lateral edge of tail; body rather plainly colored, the pale blotches indefinite or absent, even in the young Raja toboe. ee. Lateral folds of tail obsolete; tail becoming compressed toward tip, the caudal fin unusually well developed; first dorsal separated from the second by more than half its basal length; series of spines on tail single in specimens as long as 35 cm. (a partial series of spines on each lateral edge of tail in a specimen 58 cm. long); body coarsely and conspicuously blotched with pale, except in large adults. Raja kenojei. bh. Snout greatly produced, its length from eye about equal to distance from spiracle to inner angle of pectoral fin; front margin of disc deeply concave. Rostral cartilages united for about half their length. Caudal fin rather high; lateral folds of tail weak anteriorly, well developed posteriorly, extending to middle of caudal fin; tail depressed throughout; first dorsal separated from second by more than half its basal length; series of spines single on top of tail; adult with a single additional series of spines on each lateral edge of tail; coloration plain Raja tengu}^ 36. [76] Raja fusca Garman. = Black Skate. Raia fusca Garman, Proc. U. S. N. M., VIII, 1885, p. 42. — Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 649. — Pietschmann, Sitzungsb. Akad. Wiss. Wien., CXVII, 1908, p. 645. — Garman, Mem. Mus. Comp. Zook, XXXVI, 1913, p. 349, pi. 24, figs. 4-5. Raja kenojei Jordan and Snyder, Proc. U. S. N. M., XXIII, 1901, p. 337. — Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 652; and of other American writers. — Pietschmann, Sitzungsb. Akad. Wiss. Wien, CXVII, 1908, p. 647. (Not Raia kenojei Muller and Henle, nor of other writers prior to 1900) . ’Vordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 654, fig. 8. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. Ill This species is obviously not the one called R. kenojei by Muller and Henle. We feel fairly certain that the form identified by recent workers as R. kenojei is the adult of Raja fusca, as we have mature individuals, others half-grown, a young one, and an embryo, well matching Carman’s type, all apparently representing stages of growth in one species. The black spots first appear on the pectoral ocellus. The embryo was collected by S. Yoshizawa at Toyama on the Sea of Japan; the young specimen by Aoki at Misaki; a male 39 cm. long, with small claspers,' by Jordan in the Yokohama market. 37. [78] Raja smirnovi Soldatov and Pavlenko. Megane-kasuhe. Raja binoculata Schmidt, Pisces Marium Orientalium, 1904, p. 291 (not of Girard, a Californian species). Raja meerdervoorti Snyder, Proc. U. S. N. M., XLII, 1912, p. 401 (in part at least; not Raia meerdervoortii Bleeker) . Raja smirnovi Soldatov and Pavlenko, Ann. Musee Zook, Acad. Imp. Nat. Sci. Petrograd, XX, 1915, p. 162, pL V, Peter the Great Bay. Of what seems to be this species we have a young male 385 mm. in total length (C. M. Cat. Fishes No. 7776), taken with a slightly smaller paratype off Fukui on the Japan Sea (Nonaka coll.). Six smaller paratypes were collected by S. Takayasu '^As the original account by the Russian author is not generally accessible, we append the sub- stance of it: Description op a New Species of the Family Rajid.e from Peter the Great Bay and from the Okhotsk Sea. By V. Soldatov and M. Pavlenko. Raja smirnovi sp. nov. Type, a male specimen 1077 mm. long from Peter the Great Bay; cotype, a female specimen 516 mm. long from Okhotsk Sea 58° 38' N, 152° 45' E., ol)tained by Dr. Derbek at depth of 69 fathoms. Disk much broader than long; the snout broad, not produced at tip. The anterior margin greatly arched; the rostral angle being about 100°. Interorbital area broad, concave. Both upper and lower surfaces are naked; only a few minute spines or prickles present along the anterior and posterior border of pectoral fins, on tip of snout, on anterior and posterior portion of orbital rim. Three strong spines on middle of back in male, and only two in female specimen. Two strong scapular spines in male, as in female. No spines on middle of disk. After interruption the spines reappear in a series of 22-26 on back of tail; a wide band of coarser minute prickles on each side of tail. Under parts without spines and prickles. A wide lateral fold along either side of tail. Pectoral hooks very well developed; they are usually radial in position, in 22-23 series, and have at most 6-7 hooks in a series. Dorsals very high and very near together. Between dorsals there is no spine in the males, and only one in the females. Teeth 20-24. Color in spirits: light brown above, whitish below, no spots or blotches. Length of body 1077 mm.; width of disk 785 mm.; length of disk 544 mm.; length of snout 136 mm. Named for Mr. Binirnov, Inspector of Fisheries. Collected in the Okhotsk Sea. 112 MEMOIRS OF THE CARNEGIE MUSEUM. at Takashima, near Otaru. It is doubtless northern in distribution, and is ap- parently the same as Raja smirnovi, lately described from the coast of Siberia. Our material is contrasted with the other Japanese species in the key given above. It is closely related to the Shore-rays of western North America, most closely perhaps to R. hinoculata and R. inornata. From R. hinoculata it differs in the sharply indented margin of the ventral fin and in coloration, from R. inornata in the less widely connected rostral cartilages, in having the second dorsal and caudal fins entirely separated, and in the spination of the back. Disk broader than long, the greatest width, when measured backward from tip of snout, extending to a point midway between the end of the ventral fin and the insertion of the first dorsal; its posterior margin weakly convex, while the anterior margin is slightly flexuous, a gentle convexity opposite the eyes not quite reaching to the chord across the general convex curve. A straight line between the extreme angles of the disk approximately coincides with the end of the second third of the length of the disk. The front angle of the whole disk is about ninety degrees in the larger specimens, about one hundred degrees in the smaller ones; the angle of the snout, fifty-five degrees in the larger, ninety degrees in the smaller specimens; the snout, as the measurements indicate, becomes produced with age, but the extreme tip is never sharply produced beyond the general contour; the length of the snout from the eye is nearly four times the width of the concave interorbital, about 1.7 times the distance between the outer edges of the spiracles; the length of the snout before the mouth is about 2.4 times the width of the mouth, or the least distance between the nostrils, and is contained 1.6 times in the distance from the mouth to the front of the anal slit. Outer edge of ventral fin sharply indented or deeply concave. Tail short, its length when measured from middle of anal slit only equal to the distance from that point forward to the middle of the pre-oral length of snout; it is much compressed throughout, becoming excessively flat toward its tip. Lateral keel, which originates near the extreme base of the tail, extending backward to within less than 2 mm. of end of tail, leaving a keelless tip scarcely longer than broad; the keel increases in size posteriorly, its dorsal width opposite the second dorsal being equal to the depth of the tail immediately behind that fin. The first dorsal fin is inserted nearer the end of the tail than the anus by a distance equal to its own base, and is separated from the second by an interval nearly equal to its base. Supracaudal extremely low, its height being only one-tenth its length; only separated by a short interspace from the second dorsal. Body wholly smooth, except for a few strong spines, arranged as follows: one at outer edge of front orbital rim, directed outward and backward; one at JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 113 inner edge of front orbital rim, directed straight backward; one on orbital rim opposite back of eye, directed outward and backward; only one at the shoulder; a row of eight to eleven strictly median spines on the tail, and one or two more be- tween the dorsals. The strong rostral cartilages converge in a slightly convex curve toward the tip of the snout, and are free for two-thirds of their length. The brown color of the upper surfaces is more or less indefinitely broken by paler blotches. The pectoral ocelli are vividly distinct in our smaller specimens, being reddish, with or without a gray center, within a narrow black ring, but in the larger specimens are rather indefinite. In the young the tail may be trans- versely banded. The lateral keels of the tail and the extreme front margin of the disk are pale. As our largest specimen has the claspers barely developed, this ray must attain a large size. Our specimens differ from the original account of this species in the entire absence of small prickles, and in the smaller number of spines on the middle line of the back, both features probably matters of age. 38. [73] Raja tob« Tanaka. Kasube = Skate . Raja meerdervoorti Jordan and Snyder, Proc. U. S. N. M., XXXIII, 1901, p. 337. — Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 650, fig. 7, and of subsequent writers, except Pietschmann, 1908, and Snyder, 1912. Probably not Raia meerdervoortii Bleeker. Raja tohm Tanaka, Zool. Mag.*, XXVIII, 1916, p. 313 (in Japanese); Fig. Desc. Fish. Japan, XXV, 1917, pp. 453, 455, pis. 124-127. Of this species we have a male, very like the type of R. tobcB, from Miyazu (Kyoto Fishery Institution); a female, like Tanaka’s allotype, from the Tokyo market (Jordan); and six smaller specimens, one from the Toba market (Jordan and Yamamoto), four from Kagoshima Bay (Wakiya), and one from Mikawa Bay (M. Ishikawa). It seems to us probable that R. meerdervoortii of the older authors (prior to 1900) is the same as the original R. kenojei. We therefore, adopt Tanaka’s name. It is a rather small ray, structurally well distinguished from the others of Japan. 39. [74] Raja kenojei Muller and Henle. = Rasp-skate. Raia kenojei Muller and Henle, Plagiostomen, 1838, p. 149, pi. 48, and of other authors prior to 1900. *Frequent references made in the synonomy in this paper are to the “Zool. Mag.” by which is intended the Dobutzu-Gaku Zasshi, a Japanese publication issued in Tokyo, of the title of which “Zoological Magazine” is a translation. W. J. Holland. 114 MEMOIRS OF THE CARNEGIE MUSEUM. (?) Raia meerdervoortii Bleeker, Act. Soc. Sci. Ind. NeerL, VIII, 1860, p. 66. (?) Raiajaponica Nystrom, Handl. Svensk. Vet. Akad., XIII, 1887, p. 52 (Nagasaki). Misaki (Aoki); Toyama (Yoshizawa); Miyazu (Kyoto Fish. Instit.). 2 9 9. The name kenojei is in current use for a very different species, Raja fusca Garman. The name meerdervoorti, which is probably synonymous with kenojei, has been used for the species lately named tob(B by Tanaka. Muller and Henle’s figure shows most of the trenchant characters of the present form, thus diverging widely from Raja fusca. The species, so far as we can ascertain, has not pre- viously been recorded by American ichthyologists. Family DASYATIDTl. 40. [79] Dasyatis akajei (Muller and Henle). Akaei = ^ed Ray. Miyazu, Mikawa Bay. This species is common in the markets of Japan, especially southward. 41. Dasyatis ushiei Jordan and Hubbs, sp. nov. C/s/wef = Cow-ray. The type, and only known specimen, (C. M. Cat. Fishes, No. 7778), is a young male, 988 mm. long to tip of tail, collected by Masashi Ishikawa in Mikawa Bay, a branch of the Gulf of Ise. This Sting-ray is obviously one of very large size, for our type has the claspers still rudimentary. It may be that the adult is the huge form known along the Japanese coasts as “ushiei,” or “Cow-ray.” An example of this Ushiei, from the coast of Mutsu, near Aomori, was seen by Dr. Jordan in the Yamada Museum. It was six feet long, had a great sting, and a single row of strong bucklers along the back and tail, but was otherwise apparently smooth; the front margin was rounded, not angular. The chief characters of the type follow. Disk nearly one-fourth broader than long (the greatest width 383 mm.; the length to opposite end of pectoral 310 mm.). Tip of the snout but slightly produced; the front angle of the snout is 150 degrees; the front angle formed between the nearly straight front margin of the disk is 105 degrees. The line across the greatest width of the disk traverses the anterior part of the disk at a median distance behind the tip of the snout, which is 2.4 times the length of the snout. The ocular and branchial regions are considerably elevated, so that the physiognomy. approaches that of the myliobatid rays. Eyes somewhat elevated ; orbit about as large as the spiracle, which is of rhombic outline and faces about equally outward, upward, and forward; interorbital flattish, two- thirds as long as the snout. Width of mouth about one-fourth less than least internarial distance, a little less than half pre-oral length of snout, which in turn JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 115 is one-third distance from mouth to inner posterior angle of ventral fins. Only twenty-three oblique rows of teeth in upper jaw. Oral papillae seven, in three groups, of which the median series comprises three papillae. Outer margin of ventral fins longer than either the anterior or posterior sides, slightly concave, and forming an angle of about eighty-five degrees with the posterior margin, which is subtruncate. Tail little more than twice as long as disk; slightly compressed an- teriorly, but terete and whip-like behind base of spine. Upper edge of the tail bearing a rudimentary keel shorter than the orbit, located not far behind the base of the spine; lower edge bearing a very low fold, extending from its origin opposite base of spine nearly one-third of distance to tip of tail, where it grades into a low keel, covered with shagreen, which extends almost to the extreme tip of tail. Except for the caudal spine, the body is without armature of any kind, and smooth shagreen is only developed on the posterior three-fourths of tail. Color of the upper surface gray, with some blackish margins of irregular form and disposition. Tail mostly blackish, white mottled with darker on the lower surface of the thickened basal portion. Under surface of disk white, with darker clouds toward the margin posteriorly. This species is characterized by its large size, smooth body, the form and keeling of the tail, the numerous (seven) buccal papillae, the blunt snout, etc. 42. Dasyatis species (?). Another very large sting-ray inhabits the shores of Japan. It has the tail armed with coarse strong tubercles, of very large size on the mid-line before the spine. The lower edge has a strong fold, armed, except anteriorly, with small tubercles, and extending from opposite the base of the caudal spine to, or nearly to, the end of the tail; the upper edge has no keel or fold. The tail of this species is often sold in the markets of Japan, being made into a cane. The species apparently is not yet described. 43. [83] Pteroplatea japonica Temminck and Schlegel. Tsuba-kuro-ei = Sword-guard Black Ray. A specimen from Miyazu was presented by Dr. Ishikawa. Others were seen, but not taken, in the markets of Tokyo, Osaka, etc. It is generally rather common. 44. [84] Urolophus fuscus Garman. /^Rro-ef = Black Sting-ray. Yokohama market (Jordan) ; Toba market (Jordan and Yamamoto) ; Misaki (Aoki); Kagoshima Bay (Wakiya). 116 MEMOIRS OF THE CARNEGIE MUSEUM. Family AETOBATIDiE. 45. [85] Aetobatus tobijei (Bleeker). T'o6e-e^ = Hawk-ray. Tokyo market (Otaki). Not rare southward. 46. Stoasodon narinari (Euphrasen). Tokyo market; a specimen sent by Professor Otaki. This is the first record from Japan of this widely diffused species. Examples from different regions need comparison. The name Aetobatus was proposed by Blainville in 1816 for the “Raies aigles,” of which Raja aquila Linnaeus was the commonest European and best known species. In his “Faune Frangaise,” edited by Serville in 1820, Blainville changed the particle hatus in each case to hatis and mentions under Aetohatis but one species, the “Rate aigle,” which he calls Aetohatis aquila. This species later became the type of Myliohatis (Dumeril) Cuvier, 1817. In 1838, Muller and Henle adopted Alyliobatis for the Eagle-rays, quoting the earlier Aetohatis Blainville as a synonym, while they established a new genus Aetohatis N for the exotic Raja narinari Euiihrasen. In 1849, Cantor reverted to the original arrangement, making Mylio- hatis a synonym of Aetohatis Blainville, and giving a new name, Stoasodon, to Raja narinari. It seems to us that this arrangement must stand in accordance with current rules. Family CHIMtERID^. In our opinion the genus Chimcera must be subdivided, for the species are distinguished by trenchant structural features. The seven species known to inhabit the waters of Japan may be divided among four genera. These may briefly be contrasted as follows: a. Claspers trifid. b. Anal distinct from the subcaudal; second dorsal fin not notched CJmncera. bb. Anal fin absent. c. Dorsal fin not notched Psiychichthys. cc. Dorsal fin deeply notched Bathyalo'pex. aa. Claspers bifid; anal fin absent. d. Dorsal fin not notched ; caudal filament excessively prolonged .... Phasmichthys. Genus Chimera Linnaeus. Three Japanese species, C. phantasma, owstoni, and jordani maybe retained in the genus Chimcera, which otherwise will include only C. monstrosa of the Atlantic Ocean. 47. [87] Chimaera phantasma Jordan and Snyder. = Silver shark. An adult female from the Osaka market (Jordan) corresponds well with the descriptions of this species given by Jordan and Snyder, by Tanaka, and by Garman. A mounted skin from Sagami Bay was seen in the Yamada Museum. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 117 Genus Psychichthys Fowler.^” We provisionally adopt this name for Chim^rids having trifid claspers, anal fin absent, and the dorsal fin not notched. In addition to the new species described below Psychichthys ogilbyi has been listed from Japan, a record which needs verification. 48. [89] Psychichthys eidolon Jordan and Hubbs, sp. nov. (Plate V, fig. 1.) Kachi-zame = Priceless shark. ChimcBra purpurascens (Gilbert MS.) Jordan and Snyder, Smiths. Misc. Coll., XLV, 1904, p. 235. — Tanaka, Jour. Coll. Sci., Tokyo, XX, Dec. 1905, p. 14. Not C. purpurascens Gilbert, Bull. U. S. Fish. Comm., XXIII, 1903 (Aug. 1905), p. 582, fig. 231, which is a Hawaiian species. Type, a specimen 128.5 cm. long, from “off Mishima, Izu, in Sagami Bay”; Cat. No. 12902, Stanford University Fish Collection. (C. M. Cat. Fishes, No. 7779). Jordan and Snyder recorded the same specimen in 1904, referring it to a Hawaiian species described in a manuscript by Dr. Gilbert, which did not appear until some time later. As the account given by Jordan and Snyder was not in- tended as a description, but merely as an indication that a black chimaera occurred also in Japan, the name purpurascens, dating from 1904, must be regarded as nomen nudum, until the appearance of the paper by Gilbert. Tanaka gave a brief account of the species as Chimcera purpurascens in his descriptions of Chimcera jordani and C. owstoni. Garman^^ for no apparent reason uses this name to replace C. jordani, a Japanese species unknown to Jordan, Snyder, or Gilbert, and then renames the true C. purpurascens, as Chimcera gilberti. Psychichthys eidolon is most like P. purpurascens, the Hawaiian species, with which it has been confused, but differs in the much higher first dorsal fin and much shorter pectoral fin, so that the spine is contained much less, instead of much more, than two times in the length of the upper pectoral margin. The soft dorsal rays are higher than in purpurascens, 4.5 in head. Body moderately deep (depth a little less than one-sixth the total length), and compressed throughout. Head massive, almost as deep (although only half as wide) as long. Snout produced as a conic projection about as long as eye, and with its base separated from the nostrils by a like distance; tip of snout on level with lower border of eye. Interorbital space convex, a little narrower than the eye is long. Anterior dental laminae of the upper jaw contain five or seven enamel Fowler, Proc. Acad. Nat. Sci. Phila., 1907, p. 419. Garman, Mem. Mus. Comp. Zool., XL, 1911, p. 86. 118 MEMOIRS OF THE CARNEGIE MUSEUM. rods; anterior laminae of lower jaw notched at symphysis. Eye, 4.2 in head; orbit, 3.35; interorbital, 4.4; snout, 1.8; pre-oral length of snout, 2.4; distance from snout to isthmus, 1.95; least distance from eye to gill-slit, 2.35, barely greater than least distance from eye to suborbital fold; eye to insertion of dorsal spine, 1.9. Dorsal spine comparatively slender, without anterior keel, posterolateral serrations, or posterior groove, and contained about 1.3 times in head. Soft portion of fin extending as a rounded tip adjacent to the spine ; slightly farther outward it has a margin, which is straight, except toward the base, where it is bent sharply backward to merge into the low keel, which, enclosed in a fleshy groove, connects the two dorsals. First dorsal inserted nearly the length of an eye behind vertical from end of head; second dorsal farther backward, a distance eipial to length of head minus eye. Second dorsal rising gradually to attain its full height (4.5 in head) above middle of the depressed ventral fin, then main- taining this height almost to its end. The short interval between second dorsal and caudal is filled by a fleshy ridge, which merges into the latter fin. Supracaudal rising to its greatest height near end of first third of length of its base, which equals the interval between origins of the two dorsals; it is separated slightly by a notch from a very low fleshy keel, which extends farther backward about half length of eye. Subcaudal similar in shape and height to supracaudal, but its rayed portion ex- tends farther in each direction, its base being about one-fourth longer; it is con- tinued without notches backward as an inconspicuous dermal fold well toward the end of the caudal filament, and forward indefinitely as a thick fold. Pectoral fin less than length of eye longer than head, and, when depressed, does not reach the ventral; its edge is very slightly falcate dorsally, but continued around in a wide circle to a distinct notch at the lower posterior end of the base. Caudal filament at least as long as the snout, as measured from the end of the rayed portion of dorsal. The tip of the tail is imperfect, and may have been broken off, but it was probably little, if any, longer. Color a uniform deep purplish black, as dark on the belly as on the back. The tone of color, and particularly the absence of counter-shading, indicates that the species inhabits waters of great depth. Its Hawaiian representative, Psychichthys purpurascens, was taken at a depth of between 957 and 1067 fathoms. Genus Bathyalopex Collett. 49. [93] Bathyalopex barbouri (Garman). H oshi-ginzame ^ Chimsera. Chimcera spilota Tanaka, Jour. Coll. Sci. Imp. Univ., Tokyo, XXIII, 1908, p. 15. We have not seen this species. JORDAN AND HUBBSt JAPANESE FISHES COLLECTED 1922. 119 Genus Phasmichthys Jordan and Hubbs, gen. nov. Genotype: Chimcera mitsukurii Jordan and Snyder. This genus differs from Hydrolagus (colliei) of the California coast (with which it agrees in having the claspers bifid and having no distinct anal fin) in the lack of a notch in the second dorsal and in the great prolongation of the caudal filament. Apparently Chimcera novce-zealandice Fowler should be referred to this genus. 50. [88] Phasmichthys mitsukurii (Jordan and Snyder). Chimcera 'phantasma Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 669 (not of Jordan and Snyder). Chimcera mitsukurii (Dean MS.) Jordan and Snyder, Proc. U. S. N. M., XXVII, January 23, 1904, p. 224, fig. 2, Sagami Bay. — Dean, Jour. Coll. Sci. Tokyo, XIX, February 15, 1904, p. 6, pi. 1, figs. 1, 2, Sagami Bay. — Garman, Mem. Mus. Comp. Zook, XL, 1911, p. 87. A male from Misaki (Aoki). Garman erroneously describes the claspers as trifid in this species. Jordan and Snyder’s description has slight priority over that of Dean, but the name being taken from Dean’s manuscript we must attribute the species to Dean. Family PTEROTHRISSIDtE. 51. [99] Pterothrissus gissu Hilgendorf. Gisu. Shizuoka and Yokohama markets (Jordan). This peculiar fish is common in rather deep water from the vicinity of Tokyo southward. At Shizuoka great numbers are used in the manufacture of kamoboku, producing a grade little inferior to that made from the scisenoid fishes. Family ELOPIDtE. 52. [101] Elops machnata ForskM. Xam-m>as/w = Whole Sardine. Misaki (Aoki); Kobe market (Jordan). The Elops of Japan and Formosa has been referred by Jordan and Richardson, Tanaka,^'* and Jordan, Tanaka, and Snyder^^ to the Hawaiian species, E. ha- waiensis, rather than to the East Indian species, E. machnata, with which Regan“ included it in his review of the genus. We have compared our three Japanese specimens with two from Hawaii, the latter representing E. hawaiensis. The two forms were found to be extremely similar, the character of the projection Mem. Car. Mus., IV, 1909, p. 165, pi. 66, upper fig. Fig. Desc. Fishes Japan, X, 1912, p. 184, pi. 50. Jour. Coll. Sci., Tokyo, XXXIII, 1913, p. 35. Ann. Mag. Nat. Hist., (8) III, 1909, pp. 37-40. 120 MEMOIRS OF THE CARNEGIE MUSEUM. of the lower jaw, emphasized by Regan in his key, being identical. This character, however, seems to have no specific value. Slight differences appear in the number of fin-rays: the Japanese specimens having twenty-three to twenty-five dorsal rays (sixteen to nineteen branched), and fifteen to seventeen anal rays (eleven or twelve branched) ; while the Hawaiian representatives have twenty-five to twenty- seven dorsal rays (twenty to twenty-one branched), and fifteen to seventeen anal rays (eleven to thirteen branched). A greater difference seems to exist in the number of vertebrae, sixty-four in one from Japan, sixty-eight in each from Hawaii. The Japanese specimens furthermore have the head slightly shorter, the eye smaller and the interorbital a little broader than in those from Hawaii. As Regan describes like differences in his accounts of the two species, we return to his view that the Elops of Japan and neighboring regions is referable to the East Indian E. machnata, rather than to the Hawaiian species. A comparison of larger series, however, is still to be desired. Family DOROSOMIDiE. Genus Clupanodon Lacepede. In accordance with the rules of the International Commission the name Clupanodon Lacepede is restricted to Clupea thrissa Osbeck, thus replacing Konosirus. 53. [103] Clupanodon punctatus (Temminck and Schlegel). Konoshiro, name of a dungeon-castle, in allusion to the black cross-streaks. Misaki (Aoki); Mikawa Bay (M. Ishikawa); Kagoshima Bay (Wakiya); Fukui (K. Nonaka); Miyazu, Tokyo, and Kyoto markets (Jordan). (C. M. Cat. Fishes, No. 7780, a-b.) This species is very common along shore, and is to be found in every market. It is subject to some variation in coloration; the lines of black spots and the black scapular blotch, although usually very distinct, being sometimes faint or obsolete. Family DUSSUMIERIIDTl. 54. [105] Etrumeus micropus (Temminck and Schlegel). Urume-iwashi = Market Sardine. Misaki (Aoki); Tokyo market (Jordan). Toba (Jordan and Yamamoto); Osaka, Kobe, Choshi. This species is inordinately abundant along the whole coast of eastern Japan, the commonest of the various fishes known as “iwashi” (sardine). It is dried by the ton. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 121 55. [106] Spratelloides japonicus (Houttuyn). Kibuna-iwashi. Tatoku Island, off Toba (K. Mikimoto). This little herring, known by its brilliantly silvery lateral stripe, swarms off the coast of Kyusyti. It does not seem necessary, for reasons given elsewhere, to regard Atherina japonica Houttuyn as the type of the genus StolepJiorus Lacepede. Family CLUPEIDTl. 56. [107] Sardinia melanosticta (Temminck and Schlegel). /’rt;as/u = Sardine; Ma-iwashi = True Sardine. Misaki (Aoki); Mikawa Bay (M. Ishikawa); Nagoya and Tokyo markets (Jordan); Toba market (Jordan and Yamamoto); Kushiro (Tanaka); Takashima, western Hokkaido (Takayasu). This species is a true sardine, or Sardinia, much like Sardinia pilchardus of Europe. 57. [109] Ilisha elongata (Bennett). Him = Broad. Osaka market (Jordan); Fukuoka (Hamada). The little fish from Korea, known as Zunasia chinensis (Basilewsky), seems to be the young of this species. 58. [110] Clupea pallasii Cuvier and Valenciennes. Nishin. Kushiro (Tanaka). Hakodate. The Pacific Herring is excessively common along the shores of the Hokkaido, forming the object of large fisheries. 59. [Ill] Harengula zunasi (Bleeker). Zunashi. Tokyo and Shizuoka markets (Jordan); Kagoshima Bay (Wakiya); Mikawa Bay (M. Ishikawa); Fukuoka (Hamada); Fukui (Nonaka); Misaki (Aoki). Generally abundant southward. Family ENGRAULIDTl. 60. [112] Engraulis japonicus (Temminck and Schlegel). Katakuchi ^Wide-mouth. Misaki (Aoki) ; Tokyo market (Jordan) ; Toba market (Jordan and Yamamoto) ; Tatoku Island (K. Mikimoto); Fukui (Nonaka). Specimens from Soo-chow, China have also been examined by us, the first to be recorded from that country. 122 MEMOIRS OF THE CARNEGIE MUSEUM. 61. [113A] Coilia ectenes Jordan and Seale. Etsu — Jolly. An adult specimen from near Fukuoka on the west coast of Kyusyu seems to be referable to this Chinese species, rather than to Coilia nasus^ of the eastern shores of Japan. From the same locality we have received the Chinese Nibea albiflora replacing the Japanese N. mitsukurii. This general region no doubt con- tains a considerable intrusion from the Chinese fauna. Pectoral filaments, 6; dorsal rays, 1-3-10; anal rays, 95; transverse scale- rows, about 80; scale-rows before dorsal, about 18. Eye and snout together a little less than half the postorbital length of the head. We have also examined a paratype of the species from Shanghai, several from Soo-chow, and a series from Port Arthur. We count the anal rays as 95 to 108, Jordan and Herre counted 100 to 113, while Jordan and Seale gave 123 for the type. Family SALMONID^ By David Starr Jordan and Ernest Alexander McGregor. Genus Oncorhynchus Suckley. In the north of Japan, especially around the island of Hokkaido, the Pacific salmon occur in great abundance, entering the rivers from the sea to spawn, their habits being identical with those of the species on the American coast from Alaska to California. So far as known, the species of Northern Japan are identical with those of the Northern and Eastern Pacific. In Southern Japan, where the sea-water is much warmer, the salmon and charr do not appear to enter the sea, the several species being land-locked in mountain-lakes or confined to mountain-streams. As this condition has existed undoubtedly for a long time, certain forms have come to be distinguishable from their presumable ancestry as recognizable species. These forms are more or less dwarfed and the maturing ages of individuals, as revealed through study of the scales, are always greater than those of their parent species of similar size. The younger individuals exhibit in the brooks the habits of trout and are known by the Japanese as Yamanie, the adults as Masu. It is of interest to note that these *Note: According to Rendahl the Japanese species, Coilia riasus Temminck & Schlegel, is not separal)le from Coilia clupeoides Lacepede of the Chinese coasts. According to Dr. Einar Lonnberg (Svensk. Akad. Vid. XXII), who has examined the Linnaean types, the original Clupea mystus L. (which became Mystus clupeoides of Lacepede), is the Chinese species. Coilia grayi Richardson. The common Japanese and Chinese species retains the name Coilia nasus (Schlegel). Coilia clupeoides Giinther (not Lacepede) is prol)ably the same. The name Coilia clupeoides shonld give place to Coilia myskis. D. S. Jordan. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 123 land-locked forms have undergone certain modifications in their anatomical characters analogous to those of the true trout of the subgenus Trutta. The increased number of vertebrae (63 to 70) is retained, however. None of the trout have more than sixty-two, so far as known. A final test is seen in the development of the hooked nose in the ripe males (at least in the case of 0. adonis and 0. rhodurus), and in the production of but one generation of ova and spermatozoa. The latter characteristic obviously implies the death of the individuals after once spawning, as in the marine species of Oncorhynchus. This condition has not been demon- strated in Oncorhynchus ishikawce nor in 0. macrostomus, the southernmost form of Japan, both probably derivatives of Oncorhynchus Kisutch. Key to the Species of Oncorhynchus. Subgenus Hypsifario. a. Gill-rakers comparatively long and numerous, 30 to 40 in number; scales relatively large, 126 to 133, of which the structural characters are as follows: circuli typically terminating abruptly in reticu- lations, which form a zone usually extending broadly along radial border of exposed area to near margin of scale; exposed surface comparatively free of concentric or other markings. h. Marine form. Vertebrae 64; pyloric cceca 75-95; branchiostegals 13-15; anal rays 14-16; gill-rakers about 37; scales about 130; vertebrae 64. Color bright clear blue above, silvery below; lower fins pale; upper fins dusky; young with obscure round black spots above, which fade with age, often distinct on caudal fin; spawning males blood-red, the shading irregular nerka. bb. Dwarf land-locked forms; vertebrae 66 to 69; pyloric cceca 51 to 69; anal rays 13 or 14; branchiostegals 12; gill-rakers about 32. c. Form lanceolate, symmetrical dorsal fin set posteriorly; caudal peduncle slender, 3 in head; scales 134; maxillary 1.66 in head. Color silvery; male with a narrow bright red stripe along side; no black or red spots; dorsal and pectoral fins pale, sharply and narrowly edged with black adonis. cc. Moderately elongate; dorsal relatively anterior; gill-rakers 39 to 41; pyloric coeca 51 to 59. Sooty blackish, no stripes or spots on adult males, no red on sides. kawamuroc. Subgenus Oncorhynchus. aa. Gill-rakers comparatively short and few, 19 to 30 in number; scales often smaller, 130 to 215, the circuli not terminating abruptly in the reticulations, when present. d. Scales very small, 185 to 215, circuli continuing through the reticulated zone; nuclear rings coarse; rarely with more than one annulus; branchiostegals 11 to 12; gill-rakers 20 to 30; pyloric cceca 180 to 217. Color bluish silvery below; hind part of back and adipose fin with many black spots; caudal fin with conspicuous black, more or less oblong, spots; males blotched with red; becoming excessively distorted, the shoulder conspicuously humped at maturity; size small; flesh pale, with little flavor gorbuscha. dd. Scales medium, about 130 to 145. e. Anal rays 13 to 15; branchiostegals 13 or 14; gill-rakers 20 to 25; pyloric cceca 111 to 150; circuli of scales typically traceable part way through the reticulated zone, which is narrow and rarely extends beyond second year’s growth; radiating scallopings commonly on exposed area; scale often broadest on transverse axis; nuclear circuli 124 MEMOIRS OF THE CARNEGIE MUSEUM. coarse; mature scale with two or three annuli. Color dull silvery; black spots small or obsolete; fins dusky; the adult male blotched or barred with darker, often brick-red; flesh pale and rather soft. .keta. ee. Anal rays 14 to 17; branchiostegals 13 to 18; gill-rakers 22 to 29; pyloric coeca 90 to 214; vertebrae 62 to 68; circuli of scales typically not continuing on exposed area (certain races may have a very few circuli present on exposed area); annuli often traceable on exposed surface. Back and upper fins with many small roundish black spots; silvery becoming dusky at spawning time; largest in size of the salmon; usually with red flesh tschawytscha. ddd. Scales rather large, 120 to 139; circuli not breaking up into reticulations at border of exposed surface, and at least seven or eight (usually many) of them continuing around on exposed area, even in older individuals; outline of scale typically longest on antero-posterior axis; pyloric coeca relatively few, 50 to 81; gill-rakers 19 to 25; anal rays 13 to 15; branchiostegals 12 to 15. /. Marine forms. g. Color silvery with few or no dark spots above; no red spots; parr marks not persistent; breeding males dull red; no red on caudal fin; dorsal fin more or less dusky above; caudal deeply forked, the lobes acute; length often twenty-five inches or more kisutch. ff. River and lake-forms. h. More or less dwarfish; color various; the caudal in life more or less edged above and below with red. i. Scales with few (some to ten) circuli con- tinuous around the exposed area; sides with more or less persistent round dark parr marks; black spots on sides often intermingled with crimson spots; sides below lateral line variously spotted; one to usually fifteen larger spots in a series along side; dorsal and caudal with few or no spots. j. Dorsal fin above jet-black, this color rarely fading in spirits, a pale area below and one behind the black; adipose fin mostly pale; caudal fin deeply forked, lobes acute, its edge dusky; anal rays 12 or 13 (rarely 11); gill-rakers 18 to 20 (rarely 21) ; pyloric coeca 40 to 60; scales 125 to 135, approximately the type of those of the Silver Salmon, 0. kisutch]circu\i invading the exposed area, revealing much greater age than in equal-sized scales of 0. kisutch] no spawning marks. ishikawoe. JORDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 125 jj. Dorsal fin without black above, more or less abruptly pale at tip, im- » maculate, or with a few small round inky spots along basal line, usually four in number (often six or eight, or only two, or wanting); no black spot below first ray; tips of anal and ventrals creamy white; anal relatively low and small; adi- pose fin dark-edged, but without distinct spots, blacker in front at base; sides of body more or less spotted, usually with a row of rather large spots along side of belly, often duplicated; caudal lunate, with bluntish lobes; anal rays 12 or 13; branchiostegals, 11 to 14; scales 120 to 140, structurally resembling those of the Silver Salmon, 0. kisutch, but of greater age than the scales of equal size in 0. kisutch. Dwarfish, not migra- tory, so far as known . . macrostomus. ii. Scales with many (21 to 31) continuous circuli on the exposed area; dorsal fin not black above, with inky spots at base, those on the fin rather numerous, mostly oblong, narrow, and set obliquely; a dark spot at base of first ray; caudal with no, or numerous spots, which are elliptical and parallel with the rays; the upper and esi)ecially the lower ray of caudal bright red in life; back and sides with numerous round or oblong spots, no red spots; tips of dorsal, ventrals, and anal pale; adipose fin with a dark spot. Spawning male with the head black, the jaws much distorted, and the maxillary very long. Anal rays 12 or 13; branchiostegals 12 or 13; verte- braj 63 (rarely 64); scales 127 to 133, ap- proaching the type of those of the Silver Salmon, 0. kisutch, but with traces of reticulations (in a four-year old specimen, twenty inches long) ; gill-rakers very short, 18 to 20; pyloric coeca 42 to 57; size relatively large rhodurus. 126 MEMOIRS OF THE CARNEGIE MUSEUM. Siibgeniis Hypsifario Gill. 62. [118] Oncorhynchus nerka (Walbaum). , Beni-masu = Jied Salmon. The Red Salmon of the North Pacific (Sock-eye, Blue-back, i^rasnai/a Ryba) is not yet definitely known from Japan by typical examples, although abundant in the Commander Islands and apparently in Kamchatka. It only enters rivers having lakes at the head-waters, spawning in the streams above the lake, and spending from one to three years in the lake. In certain lakes of Japan, however. Lake Akan in Kushiro, Lake Tozama in Ugo, and Lake Hakone in Sagami, specialized derivatives of this species occur, two of which may be regarded as distinct species. Oncorhynchus nerka is subject to considerable variation, the races of the different rivers of the North possessing distinctive marks recognizable by experts, though not available for specific distinction. In general the gill-rakers vary in number from 30 to 40; the pyloric coeca from 75 to 95; the branchiostegals from 13 to 15; the anal rays from 14 to 16. The scales are about 130. The structure of the individual scale of each species of salmon appears to be sufficiently characteristic to make it worth while to incorporate a description of each into the specific diagnosis. Scale of typical 0. nerka: Circuli nearly always terminating abruptly in re- ticulations along the radial border of exposed surface, which forms a zone usually extending broadly to or near to margin of scale. Typically only three to five unbroken circuli invade the exposed area. Rudiments of annuli are at times traceable around exposed area, but more often the latter presents a blank record. The dwarfed land-locked form in certain lakes of the State of Washington, known as subspecies ‘^kennerlyi,” differ from the ordinary 0. nerka in their small size (rarely over a foot in length), the body perhaps more compressed and the black spots on the back usually more distinct, extending upon the upper rays of the caudal and continuing over the whole caudal fin. The “kennerlyi” or dwarf form is probably ontogenetic, not to be noted in taxonomy, although, according to Dr. Gilbert, individuals artificially confined, become mature and spawn when very small. Two examples from Lake Akan in Kushiro, noted by Jordan and Snyder (See Proc. U. S. N. M., XXIV, 1902 (1904) p. 576) and preserved in the Stanford Museum, have anal rays 15; branchiostegals 14; gill-rakers 194-22 = 41; scales 130. Color pale, with few dark spots sparsely set on back, base of dorsal and upper rays of caudal (entirely wanting in one specimen), pectorals, and ventrals very dark above, paler on lower side. JORDAN AND HUBBS; JAPANESE FISHES COLLECTED 1922. 127 A specimen, No. 1928, from Kushiro (Tanaka), more resembles the type of 0. adonis, differing in the much more diffuse dark edging of the pectoral fins, the much larger number of pyloric coeca, slightly more numerous gill-rakers, and it also shows the following characters: length 11.31 inches; anal rays 17; branchio- stegals 12 to 13; gill-rakers 144-20 = 34; pyloric coeca 91; vertebrae 60-|-6 = 66-hu; scales 133, which structurally show three circuli invading the exposed area; reticu- lations present; annuli three (just completed third year). Body rather deep, compressed, the back elevated. Color rather dark, sides silvery, with no trace of red stripe; small black spots along the back and on the upper rays of caudal. Dorsal lighter toward base and tip, four or five distinct black spots at base. Pectoral with a broad dusky margin. Adipose fin pale, with a dark spot; caudal pale, with a few spots on upper rays. Peritoneum pale slaty-gray, with darker rib-stripes, but no stippling. This specimen must be referred to 0. nerka, having a like numlrer of pyloric coeca. It may not be land-locked. 63. [118A] Oncorhynchus adonis Jordan and McGregor, sp. nov. (Plate V, fig. 2; Plate VIII, fig. 4, scale.) Of this species but two examples were taken, a ripening male, 12.94 inches long. Collector’s No. 2190 (type), Car. Mus. Cat. of Fishes, No. 7784, collected by Doctor Jordan in Lake Hakone in Sagami in late October, and a ripe male 8 inches long (1. W. No. 107) taken by Mr. T. Ota in Lake Kizaki at the head of the Hime River near Nagano in Shinshu. Form of body gracefully elliptical, slender in head and tail and more evenly symmetrical than in other salmon, thus contrasting strongly with the chunky 0. rhodurus of the same waters. Dorsal fin set posteriorly. Head fully four times in length to base of caudal; depth of body about four times in length; caudal peduncle three in head; eye six in head; snout 3.33 in head; maxillary 1.8; pectoral 1.2 in head; branchiostegals 12-12 to 11-13; anal rays 13; dorsal rays 10; gill- rakers 13 or 14, plus 18 or 19 = 31 or 33; pyloric coeca 67 to 69; vomerine teeth small, about five in a long and narrow series; two rows, each of four or five teeth, on tongue, which is wide and slaty; peritoneum whitish, unspotted; vertebrae 61 + 5 = 66-1-11.^® Scales 131 to 134, structurally somewhat intermediate between those of the Sock-eye and Silver Salmon. About five circuli invade the exposed area, where they are much more widely spaced than on the concealed area, making This formula for presenting the vertebral count will be followed in this treatment of the Salmonids and is explainable as follows; The larger (first) number refers to those vertebrre not involved in the caudal support; the smaller (second) number represents the number of vertebra? from which arise the widened caudal, neural, and hsemal plates; “LT” stands for the urostyle. 128 MEMOIRS OF THE CARNEGIE MUSEUM. the focus excentric ; three annuli present in the fourth year, a very limited amount of reticulation exhibited; scales of lateral line conspicuously acute; circuli of nucleus widening toward focus (possibly a five-year old fish, if it is a two-year nucleus) . The type (Coll. No. 2190, Car. Mus. Cat. Fishes, No. 7784) caught in October, 1922, is beginning to appear hook-nosed. It was in life everywhere silvery; the head light yellowish-green; a conspicuous straight narrow band of bright light crimson along sides; back greenish, with many small round black spots; base of dorsal with four or five spots; caudal much spotted, some of the spots oblong; anal with faint spots; sides unspotted; fins all abruptly and narrowly edged with jet- black, as though the color had been inked on; this marking especially distinct on the pale pectorals; no red spots; no red on fins; no black nor pale tip to anal or dorsal; a few small spots on bases of dorsal and caudal. In spirits the characteristic red stripe disappears, leaving a white space. An example, taken by T. Ota from Lake Kizaki in Shinshu (I. W. No. 407), belongs also to the group of allies or derivatives of H. nerka and probably to H. adonis. Its anatomical characteristics have been incorporated with those of No. 2190 in the foregoing analysis. Length eight inches. Body symmetrical, as in the type of H. adonis. Back with small dark spots, extending on upper rays of caudal. No red markings (in spirits). Fins dusky, not evidently edged with darker. The example is a ripe male, although dwarfish. 64. [118B| Oncorhynchus kawamurse Jordan and McGregor, sp. nov. Kunimasu = 'Lo(ia\ Salmon. (Plate V, fig. 3; Plate VIII, fig. 5, scale.) This species, another land-locked derivative of 0. nerka^^ based on three examples. No. 1836 (type C. M. Cat. Fishes, No. 7785), No. 455, and No. 466, all ripe males from Lake Toyama in the mountainous western part of Ugo in the northwestern part of Hondo, presented by Professor Tamiji Kawamura. It is reputed to live at a considerable depth, coming to shallow water to spawn. The specimen. Collector’s No. 466, shows the following characteristics: length 12 inches; head 4 times in length; depth barely over 4 in length; eye 7.5 in head; snout 2.9; maxillary 1.7; pectoral 1.33. Anal rays 14; dorsal rays 11; branchio- stegals 12-12; gill-rakers 174-24 = 41; pyloric coeca 51; vomerine teeth 4 to 10 in a zigzag series; scales to end of vertebrae 126, structurally showing four or five circuli invading exposed area, badly absorbed, but at least in fourth year. Color dark blue, not silvery, almost black; no dark spots on body or fins; fins dark, tipped obscurely with black and black-edged. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 129 The paratype, Collector’s No. 455, and the type, No. 1836, are entirely similar, except that in the former the branchiostegals are 11-12; the scales 130; gill-rakers 15 + 24 = 39. In the type specimen the branchiostegals are 12-11; the scales 134; gill-rakers 17+23 = 40; and pyloric coeca 59. Of the two specimens (Coll. Nos. 455 and 1836) the former had the scales deeply imbedded, which prevented their study; but the latter clearly was in its fifth year with a two-year nucleus. All the specimens are ripe males, but not hook-nosed. This form differs from the land-locked variants of Oncorhynchus nerka chiefly in the dark color, the lack of spots, the smaller number of branchiostegals, and especially of pyloric coeca, as well as in the scale-structure, which shows almost no reticulation. Subgenus Oncorhynchus. 65. [123] Oncorhynchus gorbuscha (Walbaum). Aiara/R^o-masu = Saghalin Salmon; Koon-masu. (Plate VIII, fig. 3, scale.) This species occurs in the northern Hokkaido, where it is commonly salted with the Sake, 0. keta. We have one specimen from the Hokkaido, found in the market of Shizuoka. It may be known by its very small scales (usually about 200) ; its small size ; and especially by the large, mostly oblong, spots, which mark the caudal fin. The ripe male is characterized by a conspicuous hump above the shoulders so that the depth of the body frequently exceeds one-fourth its total length. The scale- structure shows circuli breaking up into reticulations along the radial border of the exposed surface, and visibly continuing through the reticulations; nuclear circuli coarse; rarely with over one annulus. Our example (Collector’s No. 2184) is a mature male. Length: 18.125 indies. Color: slaty-blue above lateral line, not darker on back; dorsal fin dusky especially at apex; adipose fin large, pale; no spots, except those on caudal, which are large and elongate; sides below lateral line and belly pale; pectorals and ventrals dusky above, the former with a wide distal darkish border; anal fin with a similar border. Anal rays 14; gill-rakers 14+18 = 32; branchiostegals 11-12; pyloric coeca (missing); scales 185, structurally resembling the scales of the typical Humpback, 0. gor- buscha, except that the second annulus is present at the margin (much absoi+ed, hence at least in its second year). 66. [124] Oncorhynchus keta (Walbaum). Sake. (Plate VHI, fig. 11, scale). Salmo keta vel kayko Walbaum, Artedi, Piscium, 1792, 72. Rivers of Kamchatka. After the Keta ov Kayko of Pennant and Krascheninnikow. 130 MEMOIRS OP THE CARNEGIE MUSEUM. Oncorhynchus keta Jordan and Snyder, Proc. U. S. N. M., XXIV, 1902, p. 572. (With detailed synonymy.) Salmo masou Brevoort, Exp. Japan, 1856, pi. IX, fig. 2, (name on a very bad drawing, changed in the text to Salmo orientalis Pallas, p. 275: Hakodate. ? Oncorhynchus masou Jordan and Snyder, Proc. U. S. N. M., XXIV, p. 571. (Specimen from Aomori; scales 190; A. 15; B. 13; gill-rakers 12 + 17 = 29. Color: blackish, unspotted.) Oncorhynchus haberi Hilgendorf, Monatsber. Ges. Ost-Asien, XI, 1876, p. 25. A. 16 to 18; B. 13-15; scales 136 to 150. Hokkaido, common in markets. ? Oncorhyyichus yessoensis Hilgendorf, Monatsber. Ges. Ost-Asien, XI, 1876, p. 25, Hokkaido. A. 16 or 17 ;B. 13-14; scales 133 to 137;coeca 132 to 161; vertebrae 68. This is the common large salmon of Japan, everywhere known as “Sake” or “Shake,” exceedingly abundant in the Hokkaido, extending its range southward as far as the Tone River, north of Tokyo. From Aomori, Hakodate, and other ports great numbers are shipped in salt to every part of the Empire. The flesh is pale and rather soft, but palatable when freshly caught, and as a salted fish much appreciated. This species we have hitherto identified with the common Dog-salmon, Calico-salmon, or Chum of Bering Sea and the American Coast, Oncorhynchus keta (Walbaum). This identification is probably correct, as renewed comparison discloses no difference. The young fish is plain whitish, more or less dusted with dark points, but without black spots. With age the body becomes blotched or barred with dusky and dull red. The dorsal and pectoral fins are more or less blackish; the caudal edged with dusky. Vomer without teeth in the adult. The usual weight is from six to eight pounds. The young female fish, described and figured by Jordan and Snyder as ^^Oncorhynchus rnasou” from Aomori, probably belongs to this species, but 190 scales were counted, the gill-rakers were 12+17 = 29, branchiostegals 13, anal rays 15. The fish in question was dusky, without spots. The old male figured liy the same authors from Hakodate, must be the same species: scales 160, anal rays 14, branchiostegals 15+15. In these salted fishes the pyloric coeca, and usually the gills, have been removed. It is possible that these two individuals represent a species distinct from 0. keta, darker in color and with smaller scales. If so, figures 20 and 22 in Jordan, Tanaka, and Snyder should belong to it. As in 0. keta, the body and fins are unspotted. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 131 In the present collection is a single specimen (Collector’s No. 452) a ripe female 19.5 inches long, from Sapporo, Hokkaido: anal rays 14; branchiostegals 13-14; gill-rakers 11-f 14 = 25; pyloric coeca 170; scales 138, structurally agreeing with the Dog-salmon, A:eto-type; one or two circuli continuous around exposed area, transverse axis widest, at end of its third year. 67. [121] Oncorhynchus tschawytscha (Walbaum). Masunosuke = ljOTdL of the Salmon. In Japan the great King Salmon is said to occur only in the extreme north of the Hokkaido. A mounted example, probably from the Columbia River is in the Yamada Museum. This is not represented in the present collection. Owing to the fact that studies in California by the junior author (working under the California Fish and Game Commission) have brought to light the occur- rence of several remarkably well differentiated races of King Salmon, it will not surprise us at a subsequent date to learn of the existence in the Hokkaido, or regions to the north, of well-marked Japanese or Siberian derivatives of this species. 68. [122] Oncorhynchus kisutch (Walbaum). Cmmasu = Silver Salmon. This common species of Alaska and the Pacific Coast of the United States is not yet definitely known from the main island of Japan. The land-locked or river-form, here called Oncorhynchus ishikawce, has been several times identified as 0. kisutch, of which species it is apparently a dwarfed off-shoot. Counts of anatomical parts in a fairly large series of Silver Salmon in Cali- fornia have yielded the following results: Anal rays. CcEca. Gill-rakers. Branchiostegals. Scales. Range 12-15 52-81 19-25 12-15 120-139 Mean 13.5 68.6 22.5 13.4 130 Number examined 34 66 28 34 17 Color silvery, with dark points and a few rather faint dark spots on top of head, back, dorsal fin, adipose fin, and upper rays of caudal; caudal unspotted below ; dorsal fin more or less tipped with dusky ; pectorals and anal dusky ; breeding males mostly dull red, the head not black; no red on young examples. In the United States this species normally reaches a length of about twenty to twenty-five inches and a weight of from six to ten pounds. The specimens from Otaru, Ura River, and Osatsubo, mentioned by Jordan and Snyder, may belong to this species, as also possibly those taken at Aomori. 132 MEMOIRS OP THE CARNEGIE MUSEUM. 69. [122A] Oncorhynchus ishikawse Jordan and McGregor, sp. nov. Yaniame; /vmramasR = River Salmon. (Plate VI, fig. 1; Plate VIII, fig. 6, scale). Salmo perryi Hilgendorf, Monatsber. Ges. Ost-Asien, 1876, p. 25, Hokkaido (not of Brevoort, Exped. Japan, 1856, p. 273, pi. IX, fig. 1, from Hakodate, which is a species ofHucho). — Jordan and Snyder, Proc. U. S. N. M., XXIV, 1902, p. 578 (in part). Specimens from Daiya River. This species has much in common with Oncorhynchus kisutch, from which siiecies it was probably originally derived. Our type of the species is an example from Lake Biwa (Collector’s No. 1896, C'ar. AIus. Cat. Fishes No. 7786), obtained with four others by Dr. Wakiya. It is a young male, 7 inches long, with a young of Rhinogobius similis in its stomach. Body symmetrical; head small, 4.33 in length; maxillary short, reaching little beyond eye, which is small, 1.6 in snout; caudal peduncle rather slender. Color very dark olive, paler on sides, with a few dark spots on back, but with numerous vJiite or pink spots irregularly scattered among the others (these sometimes M'anting) ; usually two or three small dark spots at base line of dorsal ; apical third of dorsal jet-black, paler at base, a whitish spot on upper half of last five or six rays; no spots on the fins, except at base of dorsal, where two to six (usually four) black inky spots are usually evident. Anal and ventrals pale, the tips of both fins abruptly yellowish white ; anterior rays of anal fin more or less produced ; pectorals dusky above on first few rays; caudal with dusky margin behind; one or more (sometimes several) round blackish spots on side of belly in a series, these rarely entirely wanting, though growing obscure with age as the parr marks disappear, one below the dorsal most permanent. Anal rays 13; branchiostegals 12-13; gill-rakers 9+12 = 21; pyloric coeca 51; vertebrae 58 + 6 = 64 + U.; scales 133, structurally resembling the Silver Salmon {kisutch-iype) ; about eight circuli continuous all around focus, which is central. Our ripe specimens have the scales rather badly absorbed and are seemingly in the third year. A larger example, an immature male from Lake Hakone (Collector’s No. 438), 9.625 inches long, shows the following characteristics: the spots and other markings mostly wanting or indistinct; the sides silvery; anal rays 12; gill-rakers 9 + 11 = 20; branchiostegals 13-14; pyloric coeca 55; scales 135, showing eleven or twelve circuli continuous around on exposed area; focus nearer anterior end; outline wide, ellijitic, in general conformed to the type of the Silver Salmon in fourth year; caudal shallow-forked, with sharp angles; parr-marks mostly obliterated; color JORDAN AND HUBBS: JAPANESE FISRES COLLECTED 1922. 133 silvery with no red; the black spots few, small, and very faint on back, none on caudal, which is not edged with red. Dorsal fin deep black above, rather abruptly paler below, especially at end of last few rays, about four elongate dark spots along its base; pectorals pale slaty below, blackish above, especially on the anterior two-thirds of their width; anal with dusky band traversing it midway, apex lighter, the fin as a whole pale and small. The black area on the dorsal fin grows darker with age, but the spots on the sides tend to disappear, the sides becoming silvery. This specimen approaches maturity. We have a smaller one, much like it, also from Lake Hakone. Thirteen specimens of different sizes from four to six inches long, paler in color, and with the parr-marks very conspicuous, were taken by Dr. Ishikawa on a special trip to the Shibu River (Shibugawa), a mountain-stream near Ikao in Kotsuke, Central Japan. These are variously spotted with black, a series of larger spots along side of belly being almost always conspicuous (sometimes wanting, or at other times duplicated) ; often with a dark spot below last ray of dorsal, and smaller inky spots along its base. Vertebrae 64 in all X-rayed specimens. From various localities in Southern Japan come thirteen specimens 3.5 to 7 inches long, their colors more or less faded. One of them from Uwajima has 69 vertebrae, one 68, the others 63 to 66. The branchiostegals vary from 11 to 14, usually 12-13, the left side usually having one more than the right; gill-rakers usually 8 + 10=18; coeca 37 to 58; scales 125 to 140, the latter number in the specimen (Coll. Wakiya, No. 536) from Uwajima; scales with five to eight circuli continuous around exposed area, well separated; the scales of the Silver Salmon (0. kisutch) type, also of the scale type of a paratype of Oncorhynchus formosanus (Jordan and Oshima); mostly in second year. Other examples come from the Kitakami River at Sendai (vertebrae 65, 65, 65; coeca 46, 41, 45; gill-rakers 18, 18, 18) ; one from the Kiso River in Shinshu (vertebrae 68; coeca 37; gill-rakers 18); two from Hamada (vertebrae 66, 64; coeca 45, 58; gill-rakers 18, 18); one from Toyama (vertebrae 63; coeca 42; gill-rakers 19); two from a stream in Hokkaido (vertebrae 64, 64; coeca 48, 41; gill-rakers 18, 18); two from Kumamoto, (vertebrae 64, 64; coeca 55, 41; gill-rakers 18, 18); and one from Uwajima (vertebrae 69; coeca ?; gill-rakers 18; scales 140). This is certainly the common ‘Trout” or “Yamame” of central and northern Japan, and it is most probably a derivative of Oncorhynchus kisutch, differing in its dwarf size and darker colors. It may always be known by the black upper part of the dorsal, which scarcely fades in spirits. The dark spots on the lower part of the sides are also characteristic in the young. 134 MEMOIRS OF THE CARNEGIE MUSEUM. The anatomical characters of the foregoing series of individuals may be summarized as follows: Anal Rays. Coeca. Gill-rakers. Branchio- stegals. Scales. Vertebrae. Age in years. Range 11-13 37-58 18-19 11-14 123-140 63-69 1-4 Mean 12.2 43.6 18.1 12.5 129.4 65.1 2 In all of these examples the ground-color of the peritoneum ranges from deep straw-color to smoky rose, with paler rib stripes and scattered black stipplings. 70. [122 in part] Oncorhynchus macrostomus (Gunther) . Ameno-uwo (male) = Rain-fish ; Amago (female); Enoha. (Plate VI, figs. 2-3; Plate VIII, figs. 8-9, scales.) Salmo macrostomus Gunther, Shore-fishes, Challenger, Exped., 71, pi. XXXI, fig. 8 (Yokohama market). Salmo perryi Jordan and Snyder, Proc. U. S. N. M., XXIV, 1902, 578, (in part); examples from Lake Biwa and other localities in Southern Japan (not of Brevoort) . This species is widely diffused in the streams and lakes of Southwestern Japan. It is highly variable in color, according to the nature of the water, and unfortunately in museum specimens as to its condition of preservation. As in other fresh- water Salmonidce, the differences between individuals are much more impressive than the characters which distinguish species. The species is, however, distinguishable at all times from its nearest relative, Oncorhynchus ishikawce, by the total absence of a large black blotch on the upper part of the dorsal fin. From 0. rhodurus it is separated by more technical char- acters, especially of the scales, and by the marked difference in character and adjustment of the dark spots, which in 0. rhodurus are not present below the lateral line. Of this species the most typical examples are a series from Lake Hakone, seven to nine inches in length, both sexes being represented, the largest quite mature. These in life show the following colors: dorsal fin pale above, smoky below, never black above, usually with three to eight small prevalently round spots along its base, no black spot on base of first ray; adipose fin unspotted, usually dusky at base in front. Head very dark above. Back with small dark spots, these in the more sharply marked examples continuing upon the head in rather definite JORDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 135 arrangement; a bowed line of eight or ten along occipital border; a similar arched line over each eye; none reaching lateral line. Caudal fin with two or three spots, sometimes with none; with pale edges which are red in life, as in 0. rhodurus and 0. ishikawcB. Lower fins dusky, paler than in 0. rhodurus, the tips vaguely lighter. Anal anteriorly dusky, the tip rather sharply pale; no spots at base. Pectorals dark-edged above. Sides of body with seven to nine large round blackish blotches, parr-marks, which, however, do not seem to be wholly lost with age. Above these and alternating with them are large round obscure blotches, which fade with age. Mixed with the black spots and blotches are scattered crimson spots above the lateral line, usually larger than the black ones. Sides below lateral line with one to five series of often twenty round black spots, unequal in size, most of them very distinct, but reduced in some examples to a single large spot on side well below the dorsal fin. Usually there are several series of these spots, the smaller ones extending upon the belly. Of the two examples figured in this paper, the one is a young male, collected by Dr. Ishikawa in Nagara River near Gifu. This specimen was much faded and the original markings have been restored from a specimen of the same size from Lake Hakone. The other figure is from a very highly colored young male from Lake Biwa. The first of these differs from Gunther’s plate only in the much smaller mouth. From Gifu we also have a more mature male, in \\diich the maxil- lary is very much longer, a character apparently due to greater maturity. None of our examples show a hook-nose, nor any tendency in that direction. The example figured (Plate VI, fig. 3) from Lake Hakone (Coll. No. 450) (Car. Mus. Cat. Fishes, No. 7790) shows the following traits: Head 3.75 in length to base of caudal; depth 5; eye 5.75 in head; snout 3.15; maxillary 2 in head (1.5 in more mature examples). Anal rays 12; branchiostegals 13-14; gill-rakers 94-11 = 20; scales 116; pyloric coeca (lost), 57 or 58 in other examples; vomer with a long line of teeth in zigzag; caudal lunate, with bluntish lobes, not deeply forked; length 8.75 inches; scales with focus central, about six circuli continuous around exposed area, much more widely spaced posteriorly; wide transversely; in third year (other examples from Lake Hakone mostly in fourth year). Body in spirits much faded, but showing faint parr-marks, scarcely silvery; anal dusky at base, with conspicuous white tip; back with rather few spots; sides nearly plain; caudal nearly or quite unspotted; dorsal dusky, paler at tip, unspotted, except for three to five black spots present at its base; anal rather low, pale at its tip, or entirely pale. 136 MEMOIRS OF THE CARNEGIE MUSEUM. The other of the two specimens figured (Plate VI, fig. 2) is an immature male, seven inches long, from Lake Biwa, obtained from a private hatchery in Otsu by Jordan and Kawamura. In this specimen the very dark color obscures the black spots, leaving the crimson spots even more conspicuous; in life the dark spots on the side are very obscure and there are none on the head; the anal is broadly tipped with orange; the unspotted caudal is trimmed above and below with red. Ten other examples, similar to this and as brightly colored, ranging in length from four to seven inches, were obtained. The following characteristics are, ex- hibited by the specimen figured (Collector’s No. 454, Car. Mus. Cat. Fishes, No. 7791). Anal rays 12; branchiostegals 12-13; gill-rakers 6-}-ll = 17; pyloric coeca 49; scales 130, structurally of the type of the Silver Salmon, 0. kisutch; focus central; 11 circuli continuous around on exposed area; in second year. Of Oncorhynchus 7nacrostomus we also have three examples from the Hiki River at Wakayama, which are comparatively large and dark in color, the black spots almost obliterated. In addition we have one from Hamada, five from Himeji, three from the Kuma River at Kumamoto, five from Toyama, Fukui, one each from Lake Kizaki, Echizen, and Uwajima, besides those already noted from Lake Hakone, Lake Biwa, and the Nagara River. Of these only four show the maxillary as long as shown in Gunther’s figure (1.5 in head). This species seems to be the most abundant form in the streams south and west of Lake Biwa; its northern limit, so far as our collections show, being Lake Hakone. How many kinds of these 'Wamame” in this remarkable lake are really indigenous, and what may have been introduced in the long period of Japanese civilization, no one can tell. It is certain now that 0. rhodurus, 0. macrostomus , and 0. ishikawce all exist there in almost equal abundance, besides 0. adonis, a species of a very different type. 0. macrostomus is reputed to be non-migratory. It seems to reach only a small size and the males seen, while having the jaws prolonged backward, show no signs of a hook. The character of the scales, which, as in 0. ishikaivce, have but few (seven to ten) circuli continuous on the exposed area, sharply distinguishes these species from 0. rhodurus, in which the circuli range from 21 to 31. The black tip of the dorsal fin is the only constant character we have found sharply to set off 0. ishikawce from 0. macrosto7n,us. In both the dark spots normally extend below the lateral line, and are often mixed with red ones. The round parr-marks in both are very persistent, and in the young the additional alternating dorsal row is conspicuous. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 137 Tabular Summary of Characters of Oncorhynchus Macrostomus. Localities. 1 Coeca Gill-rakers. Anal Rays Branchio- stegals. Scales. Vertebrae^^ Head, per ct. in length. Maxillary per ct. in length Lake Biwa. . . . Average .... 41-61 51.2 17-21 19.2 12-13 12.5 11-14 12.5 126-133 64-66 64.8 24-28 9, 25-27 d', 15-17 9, 15 Lake Hakone. . Average .... 18 12-13 12.4 12-14 13.1 126-143 63-64 63.4 c^, 27-30 9,25 d, 16-19 9, 14 Fukui 40-45 17-20 12-13 11-14 124-142 63-65 cf , 25-26 c?, 13-15 Lake Kizaki . . . 49-55 17-19 12-13 11-14 122-132 64 c?, 26-27 15-16 Six other localities. . . . 43-57 18-21 12-13 12-14 120-130 62-64 25-31 13-20 Closely allied to Oncorhynchus mucrostomus is the species recently described from the mountains of Formosa, as Sabno fonnosanus Jordan and Oshima (Proc. Acad. Nat. Sci. Phila., 1919, 122). In this species, the anal is very high, its first rays reaching beyond tip of the last. A. Ill, 10; scales 130; branchiostegals 13; gill-rakers 7-t-9 = 16; distal parts of all fins dusky. 71. [124] Oncorhynchus rhodurus Jordan and McGregor, sp. nov. (Plate VII, fig. 1; Plate VIII, figs. 1-2, scales). Oncorhynchus masou Jordan and Thompson, Mem. Car. Mus., VI, 211, pi. XXIV, fig. 3, (not text). The type of this well marked species is a mature male 20.25 inches long (Col- lector’s No. 2218) (Car. Mus. Cat. Fishes, No. 7794) with very strongly hooked jaws and nuptial colors, taken in Lake Hakone, Sagami, November 20, 1923, presented by Mr. Henry Kanaya Yamaguchi, proprietor of the noted Fuji-ya Inn at Miyanoshita. This specimen shows the following characters: body robust, compressed; jaws much hooked, the long upper jaw overlapping the lower; anal fin of moderate height, first ray fully reaching to middle of last ray; tip of first few rays produced; caudal subtruncate or shallowly lunate; anal rays 12; branchiostegals 11-12; gill- rakers 8+10=18 (very short); pyloric coeca 42; vertebrae 63; scales 133, (in poor shape from absorption), structurally of a modified 0. kisutch type, showing a few reticulations, at least six circuli completely surrounding focus, these much more widely spaced posteriorly; at least in its fourth year 21 to 31 circuli in all on exposed area. ^ Determined by X-ray photographs. 138 MEMOIRS OF THE CARNEGIE MUSEUM. Color in life; head jet-black; sides bright greenish, with only a few black spots, these chiefly confined to the back. Parr-shades faint, the interspaces pink, contrasting with the green of the sides; no red stripe along sides; a bright red stripe along outer rays of caudal above and below, fading in spirits to whitish, that on the lower rays most distinct; black spots on dorsal numerous, small, more or less oblong and oblique; spots on base of fin few, oblong and obliquely set; tips of dorsal, anal, and ventrals broadly creamy white; pectorals with obscure dark edging distally; caudal lobes short, acute, a few black spots above and below: inside of mouth blackish, one vomerine tooth remaining. A microscopic study of the scales of No. 2136 (a paratype thirteen inches in length) reveals that the individual was seven years old and had never spawned before. This ecpials the maximum longevity hitherto recorded for Oncorhynchus. Three female specimens, ripe with eggs, about a foot long, from Lake Hakone (November 20) are much smaller and slenderer; dull sooty-silvery in color with traces of dark spots above, and faint pinkish shades in the interspaces, alternating with traces of dark bars. Upper and lower edge of caudal pale, but (in spirits) no longer red; numerous smaller oblong, oblique spots, arranged vaguely in rows along the rays, those at base of the fin oblong, oblique, ink-like, sometimes running together in a broken line; caudal with a few dark scattered spots; none on ventrals or pectorals; paired fins dark, with abrupt yellowish white tips. A younger male is like the old one, but less brightly colored and barely be- ginning to be hook-nosed. It has still a long row of five small vomerine teeth. Two spawning females, (from some locality in Shinshu) one about a foot in length, and one about 17.5 inches long, exuding ova, are dusky metallic-sooty, with traces of dark bars, these quite unlike the rounded parr-marks of other species; no red on fins, except on lower ray of caudal; spots mostly obliterated; anal with a broad white tip; adipose fin with a dark spot above; anal usually with a few dark spots at base; no series of dark spots along side below lateral line in either specimen. Parr-marks mostly early replaced by broad vertical dark cross-bars, which in the males tend to grow sharper with age; pectorals deep slaty, as are also the other lower fins, except for the pale tips; young without red spots on sides. The faint broad dark cross-shades are characteristic of this species. Of Oncorhynchus rhodurus we have six examples of various sizes and both sexes from Lake Hakone, and two from some lake in Shinshu, near Nagano. The species may be known at sight by the pale dorsal as well as by the absence of dark spots below the lateral line. A much more important character is found in the presence on each scale of many (21 to 31, the average 25.4) unbroken circuli on JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 139 the exposed part of the scale. In Oncorhynchus macrostomus and 0. ishikaum, the unbroken circuli on the posterior surface are few (7 to 10, averaging 8) and the spotting is different. In spite of resemblance to Salmo, this species is a true Oncorhynchus. Like Salmo, it has twelve anal rays, mostly eleven or twelve branchiostegals, seventeen to twenty gill-rakers, and one hundred and thirty-three scales in lateral series. The fact that the first anal ray reaches beyond the middle of the last ray marks most of our species of Salmo. On the other hand, the number of coeca (averaging 57), the number of vertebrae (about 63), but more particularly the well developed hooked nose in the ripe males, and the death of all spawning individuals, identify it as an Oncorhynchus. Oncorhynchus rhodurus is the species figured by Jordan and Thompson, Mem. Car. Mus., VI, 1913, p. 211, pi. XXIV, fig. 3, as Oncorhynchus masou. The speci- men figured. Car. Mus. Cat. Fishes, No. 6002a, was taken in Lake Chiusenji, a land-locked mountain lake, in which it was said to have been introduced from Akita in Ugo in the northwestern part of Japan. We were told that all individuals in the lake die after spawning. The figure in question accurately represents a young male, except that the black spots have faded. A specimen from Naoetsu, Echigo, is in the Stanford University Museum, collected by K. Otaki. The following table gives the record of our various examples. T.\bular Summary of Characters of Oncorhynchus rhodurus. Cceca. Gill- rakers. Anal Rays. Branchio- stegals. Scales. Vertebrae. Head pr.ct. in length. Maxillary pr.ct. in length. Range 42-76 17-20 12-13 11-13 124-133 62-63 c?, 29-33 9 , 24-25 c^, 19-22 9, 13-14 Mean 57.4 18.6 12.4 11.8 128 62.7 Subgenus Trutta Linnseus. 72. [Introd.] Salmo shasta Jordan. Shasta Rainbow-trout A specimen from Lake Biwa, presented by Dr. Ishikawa, belongs to this American species, which was introduced from the hatchery at Baird on the McCloud River near Mount Shasta, California. The body and upper fins are profusely spotted, much more so than in any Japanese species. 140 MEMOIRS OF THE CARNEGIE MUSEUM. Genus Salvelinus (Nilsson) Richardson. The species of Charr found in tributaries of the North Pacific are very far from final determination. The name Salvelinus malma evidently belongs to a northern form, known from Unalaska to Kamchatka. Close to this, but with the head constantly larger, is the “Dolly Varden,” or “Bull-trout” of Northern Cali- fornia and northward. This may stand as Salvelinus spectahilis‘^^ (parkei), and probably grades into the preceding. Both of these enter the sea, growing to the weight of eight to ten pounds, the red spots becoming silvery. In all the mountain- streams of northern and middle Japan the common trout, or Iwana, Salvelinus pluvius, agrees closely with S. malma, but its pyloric coeca and gill-rakers average fewer and the spots on the side and back are larger. Another group of Charr has the pale spots much larger, some as large as the eye, otherwise much like S. pluvius. Specimens from Kamchatka {S. leucomcenis) have the head short, 4.5 to 4.66 in length. A related form or species of this type, from Hamada in Iwami, has the head 4 in length and the coeca very few, only seventeen. This we call Salvelinus imbrius, granting it, pending study, the rank of a distinct species. Key to Pacific Species of Salvelinus. a. Lateral spots all smaller than eye, mostly smaller than pupil, bright crimson, becoming silvery in sea-run examples; no spots on head, and usually few or none below lateral line; lower fins with the first ray bright red. h. Head long, more than one-fourth length of body to end of vertebrie (3.66 in length) gill- rakers 18; pyloric coeca 18. Shasta region and northward spectabilis. bb. Head shorter, less than one-fourth length of body. c. Pyloric coeca 24 to 33; gill-rakers 20 to 24; branchiostegals 11-12; head 4.25 to 4.5 in length; North Pacific malma. cc. Pyloric coeca 23 to 25; gill-rakers 18 to 20; branchiostegals 12-13; head 3.875 to 4.4 pluvius. aa. Lateral spots pale yellowish, irregular in size, mostly larger than pupil, some about as large as eye, extending more or less below lateral line; lower fins pale; scales about 200. d. Head 4.5 to 4.66 in length; no spots on head; caudal well forked; coeca 20; gill-rakers 18; branchiostegals 13-13. Kamchatka and Hokkaido. leucomcenis. dd. Head 4 in length; pale spots covering top of head; caudal shallow-forked; coeca 17; gill-rakers 14; branchiostegals 12-13. Hamada in south western Japan imbrius. The foregoing analysis of these forms is provisional and all may prove to be variants of Salvelinus malma. The name spectabilis Girard must take precedence over parkei for the American “Dolly Varden.” Girard called the species Salmo spectabilis. The earlier name Salar spectabilis was applied by Cuvier and Valenciennes to a European species of Salmo, but the combination, Salmo spectabilis, was first used by Girard for the Charr of the Columbia. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 141 73. [127] Salvelinus pluvius (Hilgendorf ) . Iwana. (Plate VII, fig. 2, example from Shinshu; Plate VIII, fig. 14, scale). The common Red-spotted Charr, or Iwana, is found in all the mountain- rivers and brooks from the southern part of Hondo throughout the Hokkaido. Our specimens are from the Daiya River at Nikko; Lake Chiusenji (introduced); the Kawajiri River in Rikuchu; and Ohata, near Aomori, besides those obtained by Jordan and Snyder in 1900. The two from Ohata are very slender, apparently starvelings. Whether the Iwana is really separable from Salvelinus malma of the Kamchatka-Alaskan region is doubtful, as already indicated. Besides the ordinary Iwana, found in all mountain-streams of Japan, we have a fine specimen, 11.5 inches long, (I. W. No. 207) in the collection. It is a ripe male with the lower jaw somewhat hooked and slightly prolonged, and was taken in a stream or lake in Shinshu, near Nagano, by Soji Nakona. It differs from others in the subtruncate caudal and its rather larger spots. Body rather elongate and moderately compressed; head slightly under 4 in length to end of vertebrae; depth 5; depth of caudal peduncle 2.84 in head; eye 6.5; interorbital space 3.33; snout 3.6; maxillary 1.66, extending beyond eye a distance about equal to diameter of eye; D. 9; A. 8; scales in lateral series 185, in transverse series above lateral line 38. Vomer flush with roof of mouth, on its anterior portion teeth in a fan-shaped cluster, none on the depressed shaft; palatine teeth extending to anterior vomerine teeth, with a very short gap between; roof of mouth, within palatines and behind teeth of vomer, black; branchiostegals 13-13; gill-rakers 7-1-8 = 15; pyloric cmca 21; dorsal fin a little longer than anal, the former 1.75 in head; ventrals 2.1 in head; caudal almost squarely truncate, the lobes not acute, over 2 in head, the middle rays more than two-thirds the outer; ventral appendage 2.5 in fin; longest ray of dorsal 1.75 in head. The difference in the form of the caudal is a striking feature of this specimen and is probably due to age. Scales under microscope show that the outline is usually rectangular-elliptic, with the antero-posterior axis the longest; focus nearly central, sub-circular; most of the circuli continued around on the exposed area, which appears much like the con- cealed area; circuli widely spaced, inclined to interbranch, especially on the anterior radius; blank area restricted to two posterior marginal circuli; submarginal blank zone limited to axial portion of circuli nine, ten, eleven, and twelve posteriorly; radial spurs common. This specimen is in its fourth year. Color (in alcohol) smoky-brown above lateral line, paler below; back blackish-gray; many small round pale spots (red in life) scattered over sides and back, most of these about half the diameter of the pupil, not visibly arranged in rows; a few below larger; all 142 MEMOIRS OF THE CARNEGIE MUSEUM. fins unspotted; head, especially above, dark slaty; dorsal fin slaty, the first ray palest; caudal fin slaty, unspotted, margined above and especially below with whitish; pectorals, ventrals, and anal pale slaty, all bordered anteriorly with rather wide, pale edging; pectorals and ventrals darker above. 74. [126] Salvelinus leucomaenis (Pallas). (Plate VIII, fig. 13, scale.) Salvelinus kundscha Jordan and Gilbert and of other authors, probably not Salmo kundscha Pallas (fide Berg). This trout, long known from Kamchatka, is recorded by Jordan, Tanaka, and Snyder from Nemuro and the Shiribeshi River, in the northern Hokkaido. Also from Iturup Island of the Kuriles, and from Shinano. It is visibly known by its large pale spots, creamy in preserved examples, of unequal size, some about as large as the eye, none extending on the head. According to Dr. Leo S. Berg, the early specific name kundscha should not be used for this species, as the original of the name was probably Salvelinus alpinus, a European species entering the Arctic seas. We, therefore, take the earlier name of reasonably certain application. 75. [126A] Salvelinus imbrius Jordan and McGregor, sp. nov. (Plate VII, fig. 3; Plate VIII, fig. 15, scale.) This species is based on an immature female, 9.25 inches long (W. No. 510) Car. Mus. Cat. Fishes, No. 7797, obtained from a stream near Hamada in Iwami in the southwestern part of Japan. It is very close to Salvelinus leucomcenis, which subarctic species it may represent southward. Body moderately elongate, not much compressed, dorsal contour slightly elevated back of occipital region; caudal fin shallowly forked; head about 4 in length; depth 4.5; depth of caudal peduncle 2.6 in head; eye 5.4; interorbital space 3.125; snout 4.33; maxillary 1.875; extending beyond eye a distance about equal to diameter of pupil; dorsal rays 12; anal-rays 9; scales in lateral series (immediately above lateral line) 195. Vomer flush with roof of mouth, four teeth in a cluster on anterior portion; roof of mouth longitudinally striated with pale and blackish shades; distance from tip of teeth on vomer to tip of snout one-third interorbital space; about twelve palatine teeth on each side; tongue short, blunt, of a rusty color, with two rows of about five teeth on each side; teeth essentially as in Salvelinus pluvius and other species of Salvelinus. Branchiostegals 12-13; gill-rakers 5 4-9=14; pyloric cceca 17; vertebrae 59 (53 4- 6 -h urostyle) only the last two slightly up-turned; dorsal equaling anal, 1.66 in head; pectorals JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 143 1.5 in head; ventrals fully two in head; ventral appendage about four in the fin; caudal shallowly forked, the lobes obtusely angled; its length 2.25 in head; peri- toneum smoky, with pale stripes along ribs and a few scattered stipples. Scale under the microscope shows the outline usually wide rectangular-elliptic, with the antero-posterior axis longest; focus central, circular; circuli nearly all continuous entirely around, widely spaced, nowhere angled; blank area restricted to one or two peripheral circuli pos.teriorly, and occasionally also to axial portion of circuli four, five, and six; nucleus of about nine circuli, which are spaced increasingly wider toward focus; radial spurs present or absent (when present scarce). Our specimen is in its fourth year. Color in spirits pale slaty-brown above, paler below, but back hardly darker than sides; numerous round pale spots equally distributed over back, sides, and top of head, the largest larger than pupil and some almost as large as eye; top of head with about sixteen spots smaller than those on sides; spots extending to below lateral line. Dorsal fin rather pale, a horizontal slaty bar across middle of rays, a blackish border on first ray; caudal somewhat dusky; basal half of ventrals whitish; anal, ventrals, and pectorals pale, the first rays of each paler, pectorals very slightly dusky above; lateral line conspicuously whitish. Tabular Statement of Anatomical Characters of Pacific Species of Salvelinus. Species. Locality. No. of cceca. Gill-rakers. Branchiostegals. Head in length to end of vertebrse. S. spectabilis Rattlesnake Cr.j Mont. 18 18 12-13 3.66 S. malnia Unalaska 27 20 12-11 4.5 33 20 11-11 4.5 Pt. Hope, Alaska 24 24 12-12 4.5 23 10-11 4.25 S. pluvius Chiusenji 12- 13 1 -13 13- 14 14- 15 3.875 Kamajiri R. 25 18 12-13 4 Rikuchu 23 20 12-12 4 Ohata 23 18 11-11 4.4 19 12-12 4.4 Shinshu 21 15 13-13 3.8 S. leucomcBnis Nemuro 20 18 13-13 4.25 Petropavlovsk 22 20 13-13 4.125 ■ S. imbrius Hamada 17 14 12-13 4 144 MEMOIRS OF THE CARNEGIE MUSEUM. For an account of the pyloric coeca and measurements of head of examples of Salvelinus spectabilis from near Seattle we are indebted to Professor John N. Cobb of the University of Washington. Length to Base Caudal Ray Length of Head to Nape Length of head to Margin of Opercle Length of Head in Body Number of Pyloric Coeca Sex In. Cm. In. Cm. In. Cm. Head to Nape Head to Opercle 1. 25.5 64.77 4.0 10.16 6.5 16.51 6.37 3.92 25 Male 2. 23.0 58.42 3.5 8.89 5.75 14.60 6.57 4.00 ' 23 Male 3. 19.56 49 . 68 3.31 8.41 5.62 14.29 5.9 3.48 23 Female Characters of Scales of Salvelinus. (Best applicable to scales from between dorsal fin and lateral linel A. All circuli extending entirely around scale; focus central; no true blank area. B. Outline cuneate-ovate; radial spurs usually present (at times lacking); circuli interbranching considerably, especially anteriorly, numbers one to four commonly angled postero-axially. leucomcenis (Kamchatka). BB. Outline ovate; radial spurs normally absent; circuli relatively little interbranched, inner ones more apt to be angled anteriorly spectabilis (Seattle). (Plate VIII, fig. 17) A A. All but two or three circuli extending entirely around scale; marginal blank area restricted to one to three posterior peripheral circuli; a submarginal blank area present on exposed area in most species. C. Circuli usually abruptly bent on radii bounding exposed area; outline cuneate-ovate; submarginal blank area restricted to axial portion of circuli Nos. 6 to 8. mahna (Bering Sea). (Plate VIII, fig. 16) CC. Circuli not so bent. D. Focus sub-central, often somewhat nearer the posterior end; outline oval, widest posteriorly; circuli commonly angled on long axis; radial spurs present; sub- marginal blank area restricted to axial portion of circuli Nos. 5 to 7 pluvius (Chiusenji). DD. Focus central; circuli not angled, widely spaced; outline elliptic to rectangular- elliptic. E. Submarginal blank area not present, several posterior circuli usually some- what broken; radial spurs usually lacking; circuli relatively unbranched (starved examples) pluvius (Ohata). EE. Submarginal blank area present. F. Radial spurs conspicuous; submarginal blank area restricted to circuli Nos. 9 to 12; circuli inclined to interbranch on anterior radius. pluvitis (Shinshu). FF. Radial spurs few and inconspicuous; marginal blank area restricted to axial portion of circuli Nos. 4 to 6 imbrius (Hamada). AAA. Few to eight circuli extending all around scale. G. Focus central; circuli not angled. H. Focus usually well elongate; only two to four circuli extend- ing all around scale; blank area comparatively large. Jontinalis (Maine). JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 145 HH. Focus only slightly elongate; six to eight circuli extending entirely around; blank area relatively small, restricted to a rather narrow triangle at posterior end. . . . pluvius (Rikuchu). ' GG. Focus commonly somewhat nearer posterior end; first three or four circuli usually angled posteriorly. I. Focus sub-circular; six or seven circuli extending entirely around; blank area large, its anterior margin almost a vertical line; radial spurs present; circuli closely spaced; outline rectangular-ovate, widest at a point one-third distant from focus to anterior end..(S. malma (Unalaska). II. Focus elliptical; six to eight circuli extending all around; blank area restricted to outer four circuli (No. 6 usually interrupted postero-axially) ; no radial spurs seen ; outline , rectangular-elliptic. fontinalis. (Swannanoa R., North Carolina). From this table it will be seen that a close alliance is indicated between S. pluvius and S. imbrius. Our key, based on scale-characters, brings these two forms closely together. Salvelinus fontinalis from Maine shows a rather close relationship to S. malma from Unalaska, when the scales are the only features considered. It also indicates a substantial difference in the number of coeca and gill-rakers as between the Bering Sea and Unalaskan individuals thus far referred to S. malma. These two are widely separated in the scale-key. On the other hand the critical study of the scales tends to refute in two in- stances the integrity of our tabulated analysis (See p. 140). First, the three collec- tions of Salvelinus from Rikuchu, Ohata, and Chiuzenji (R. Daiya) (all listed in the Stanford Collection as S. pluvius), when subjected to the scale scrutiny fall, two into one primary structural division and one into another, indicating a lack of identity. Secondly, Salvelinus imbrius (from Hamada) has been placed, tenta- tively, near S. leucomcenis, chiefly on account of the possession in common of body-spots larger than the pupil. On the contrary, it will be noted that the critical scale characters are quite different in these two forms, which discrepancy is ac- companied by a difference of four in the number of gill-rakers. Finally, two individuals from Montana {S. spectabilis) when placed in our table, exhibit rela- tionship both with S. fontinalis from North Carolina and S. leucomcenis from Nemuro. However, when subjected to the scale-test, this form from Montana shows no kinship with the eastern Brook-trout, Salvelinus fontinalis, while the alliance with leucomcenis (Nemuro) is greatly emphasized. 76. [125j Hucho perryi (Brevoort) . 7to-R?TO = String-fish. (Plate VIII, fig. 7, scale). Salmo perryi Brevoort, Exped. Japan, 1856, 273, pi. IX, fig. 1, (Hakodate). Salmo blackistoni Hilgendorf, Monatsber. Ges. Ost-asien, 1876, 25, (Hokkaido). A specimen of this interesting trout-like fish in the Stanford collection was 146 MEMOIRS OF THE CARNEGIE MUSEUM. obtained by Professor Otaki at Naoetsu in Echigo. A larger example was examined in the Lakeside laboratory of the University of Kyoto. It is not so slender as the one figured by Jordan, Tanaka, and Snyder, and the spots, though similarly placed, are more diffused, and not lunate. This species is smaller and slenderer than its congener, Hucho hucho of the Danube, but otherwise the two species have very much in common. This “Huchen’' is said to reach a weight of sixty to one hundred pounds. Hucho stands out as very distinct from the other Salmonidce, especially in its dentition, its fewer vertebrae, and the great number of its pyloric coeca. It is not known to enter the sea. It should apparently form a separate subfamily, Huchonince. Tlie following characters are shown by the example from Naoetsu: head 3.33 in length to end of last vertebra, depth 5; eye 6.75 in head; maxillary 2; snout 3.75; depth of caudal peduncle 2; height of dorsal 1.66; pectoral fin 1.75; dorsal rays 11, anal rays 10, pectoral 14; gill-rakers rather short and stout, 8+12 = 20; branchio- stegals 12“12; pyloric coeca 157, short, densely imbricated, the largest only 8 mm. long by 1.5 thick. Scales 105; pores in lateral line 109. Vertebrse 57 (51 + 6 up- turned, besides the urostyle). Body long and low, somewhat pike-like, not much compressed, the mouth large. Dentition peculiar; only vomer with a short trans- verse series of teeth, none on the shaft; the teeth continuous with the palatine series of the two sides, the whole forming an unbroken and nearly uniform U-shaped series. In other SabnonidcB, the palatine teeth are interrupted by the presence of the vomerine element. Scales relatively large, the entire surface marked with concentric circuli, about one-seventh of those of the concealed surface not continued on exposed area, resulting in a slightly wider spacing on this part. First four or five (central) circuli angulate at the posterior axis. The type of scale is thus inter- mediate between Oncorhynchus and Salvelinus. Caudal rather deeply formed, with acute lobes; ventrals nearly under middle of dorsal; adipose fin rather large. Head and body rather profusely covered with small black spots, few of them extending below lateral line and most of them not larger than a scale; fins all unspotted. Comparison of Genera of Salmonid/e and Plecoglossid.®. Genera. Cceca. Gill-rakers. Anal Rays. Branch iostegals. Scales. Vertebrae. OncorJiynchus 40-230 18-40 12-16 11-17 120-215 62-69 Salrno 65. 9 11 120 60 Trutla 40-60 20 10 11 120-190 58-60 Salvelinus 17-21 14-20 8-12 12-13 185-250 59-65 Hue ho 157 20 10 12-12 105 57 Plecoqlossus 368-425 33-39 9-17 5-5 140-165 60-61 JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 147 Family PLECOGLOSSID^E. (By David Starr Jordan and Ernest Alexander McGregor.) 77. [128] Plecoglossus altivelis Temminck and Schlegel. Ayu;Koayu. (Plate VIII, figs. 10 and 12, scales). Common throughout Japan and Formosa, ascending clear streams from the sea, spawning in the early fall. Also found land-locked and much dwarfed in Lake Biwa. It is one of the very choicest of Japanese food-fishes. In this genus the vertebral column ends squarely at the base of the caudal fin, the terminal vertebrae being neither reduced in size, nor turned upward (hetero- cercal) as in the Salmonidce, a fact finely shown in X-ray photographs made by Mr. McGregor. Our specimens are from Noo, Toyama, Himeji, Mikawa, (four), and from Lake Biwa at Otsu (dwarf examples),^® locally known as “Koayu.” Specimens of Ayu from Formosa seem to be identical with the ordinary form. Gill-rakers 41; dorsal rays 11; anal rays 15. The scales are decidedly longer on the dorso-ventral axis, at which line the circuli bend sharply. Branchiostegals typically 5-5 (rarely 4-4); anal rays about 14; gill-rakers about 36; pyloric coeca about 388; vertebrae 60-61; scales in linear series just above lateral line about 161; scales comprising lateral line about 62. Scales structurally as follows: decidedly broadest on transverse axis; first four or five circuli with segments lying nearly straight and parallel behind the focus, which is compressed elliptical; balance of circuli tangentially deflected posteriad; major portion of exposed area (which is normally ampler than concealed area) not bearing circuli (juvenile individuals excepted); outline of scale concave laterad of focus. Mostly two and three years old. The breeding males have the exposed surface of each scale usually bearing- three cone-shaped warty processes (pearl-organs). The following table contains a summary of the more important anatomical characters shown by our material: Branchio- stegals. Anal Rays. Scales along lateral line. Gill-rakers. Pyloric ccEca. Vertebrae. Range 4:4-5:5 9-17 149-165 33-40 368-425 60-61 Mean 5:5 14.4 161.4 36 388.3 60.5 These have gill-rakers 35, anal rays 14, and scales in lateral series 152. In the lot are some ripe males between 2.75 and 3 inches long. Only three counts as follows: 368; 372; 425. 148 MEMOIRS OF THE CARNEGIE MUSEUM. The breeding males of Plecoglossus undergo a very marked nuptial transfor- mation involving chiefly the dermal armature and the fins. Warty growths, called “pearl-organs,” or “nuptial tubercles,” develop on the scales and also overlie the fin-rays, especially on the anal fin. This imparts to the fish a conspicuously roughened appearance. Accompanying this is a very noticeable enhancement of the size of the fins. To reveal this we append the measurements of a ripe female, of a ripe male, and of an unripe male of equal length. Length of Pectoral. Length of Dorsal. Length of Ventral. Length of Anal. Length of Anal Base. Length of Maxillary. Ripe Male 17.8 20.5 17.2 11.3 18.0 11.7 Um'ipe Male 14.0 16.6 11.2 12.7 12.4 12.6 Ripe Female 14.8 . 17.6 11.0 12.8 14.4 12.1 The length of the maxillary of the male, which with most of the Salmonid fishes is augmented at maturity, would appear to undergo a reverse process in Plecoglossus as our nuptial males invariably possess the shortest maxillaries. Another conspicuous nuptial development concerns the anal fin. The base of this fin in the spawning male becomes lengthened by about half its original basal length (although the radial length appears actually to shorten slightly). At the base of the first anal rays a double series of enlarged, imbricated scales partially conceals this portion of the fin. These peculiar scales in reality are free cuticular outgrowths on the exposed surface of the underlying true scales. This nuptial transformation may occur at different ages, as evidenced by a male two and three- quarters of an inch long from Lake Biwa (one year old) and our No. 43 from Mikawa Bay in its third year. In an article published in Japanese a few years ago liy Professor T. Kagiya the writer states that it is usual for the Ayu to mature, spawn, and die in its first year. We have made a careful study of scales from several of the above Ayu from different localities in Japan with reference to age. Our diagnosis was corroborated liy Mr. Hulibs. For nine typical fish the results were as follows: in the first year, one; in second year, two; in third year, five; in fourth year, one. Nothing resemiDling the spawning-mark of trout was seen in any of our scales. This sug- gests the occurrence of but one spawning for the Ayu, followed by the death of the individual as in Oncorhynchus. Following is a condensed translation by M. Kasawa, of Professor Kagiya’s account of the life-history of the Ayu, in Japan: Rcrcentase of total length of fish. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 149 “The Ayu is strictly a river-fish. The spawning, which covers a period of about two months, takes place in August and September. The actual locations chosen for spawning are clear pools between one and two feet deep, with sandy bottoms, near the bank of the stream. The females average about 50,000 ova, and these are deposited in the sand on the bottom, usually during late afternoon or evening. “The eggs hatch in about three weeks and the young fish migrate downstream to brackish water (about late September and October) where they remain about two months. Following the feeding period in the estuary, by which time they have grown to about two or three inches, they again migrate far up to the head- waters. Here in the cold clear tributaries they feed on diatoms, moss, and other aquatic plants through the spring and summer; and grow to maturity. The ascend- ing Ayu are called Nohori Ayu. “With the advent of late summer the development, both as to size and sexual maturity, is greatly hastened. Preparatory to spawning the mature fish undergo a downstream migration to their spawning grounds in the middle reaches of the river. These mature migrating fish are called the Kudadi Ayu or Descending Ayu. The ripe fish take on a rusty color just before spawning time. “Nearly all the adult fish die after spawning, but the writer believes that a small number survive and persist another year. To these alleged survivors the name Tomari Ayu {Remaining Ayu) is applied. “There is great mortality among the young Ayu, especially following hatch- ing, due to the activity of predatory species of catfish, carp, and other dace. “The Ayu occasionally reaches the length of a foot or more and a weight of one and one-quarter pound. It does not occur in Hokkaido. “In a recent season six provinces marketed Ayu with an aggregate value of 167,000 yen.” (S83,500). It ‘is a singular fact, that abundant as are the Coregonidoe and TJmjmallidce in Eastern Siberia, no specimen of either family has been found in Japan. Family OSMERIDA5. 78. [129] Osmerus dentex Steindachner. = Cucumber-fish. One specimen from Kushiro, Hokkaido. (Tanaka.) 79. [129A] Spirinchus lanceolatus (Hikida). Osmerus dentex Franz, Abh. Bayer. Akad. Wiss., Vol. I, Suppl. 4, 1910, p. 6, pi. 3, fig. 5 (not of Steindachner), young, from Misaki. Osmerus lanceolatus Hikida, Zook Mag., XXV, 1913, p. 127, pi. 150 MEMOIRS OF THE CARNEGIE MUSEUM. We have not seen Hikida’s description (which we quote from the Zoological Record), but have a specimen labelled “Osmerus lanceolatus,” a nuptial male, 134 mm. long to caudal fin, from Iburi, Hokkaido, presented to Dr. Jordan by Dr. Tamiji Kawamura. It undoubtedly belongs to the genus Spirinchus Jordan and Evermann. Five other specimens, representing both sexes in the breeding condition, were taken in Kushiro by Tanaka. The characters of Spirinchus lanceolatus are as follows: Dorsal rays, 2, 8 (2, 9 in one specimen); anal rays, 4, 16 (3 or 4, 15 or 16); pectorals, 11 (11 or 12); ventrals, 8 (constant); scales 7-62 to 65-8 to 10; length of head to end of opercular membrane, 4.35 (4.4 to 4.6) in length to caudal base; depth of body, 5.3 (to 5.4 in males; 4.65 to 5.0 in ripe females); least depth of caudal peduncle, 3.0 in head (3.0 to 3.2); snout, 4.5 (4.4 to 4.6); upper jaw, 2.1 (2.0 to 2.15) not longer in male than in female; mandible, 1.95 (1.8 to 2.0); orbit, 4.5 (4.2 to 4.6); eye 5.6 in breeding male, larger in females; fleshy interorbital, 3.35 (male); bony interorbital, 4.6 (4.2 to 4.7). Body of breeding male oblong, the greatest depth being almost evenly maintained from the shoulders to the anal fin, behind which point the ventral contour rises rather sharply to the fairly slender caudal peduncle; females more trimly formed. In both sexes the head is rather massive, and blunt anteriorly; ventral contour curved rather sharply upward to tip of the projecting mandible; dorsal contour much flatter, weakly concave behind eyes, but a little convex on snout. Tip of mandible on level of middle of eye, tip of premaxillaries somewhat higher. Interorbital space weakly convex; suborbital narrower than the upper jaw, which has a concave upper border and a tip more rounded dorsally than ventrally; maxillary extending to below posterior border of eye. Branchiostegals 4-3; gill-rakers, 11 + 26 = 37, the longest, 1.6 in orbit. Pyloric coeca four, of very uneciual length. Teeth weak for a Smelt; tongue covered with strong, but rather short, teeth, of which four or five around the tip are somewhat enlarged, scarcely canine-like. Other teeth arranged in single file, those of premaxillary and mandible of moderate size, rather close-set; those of maxillary very fine; those of vomer, palatines, and pterygoids very small canines; each side of the vomer and each palatine bone with four or five of these teeth. Breeding males are all very dark in color, dusky above the blackish lateral stripe, which extends along and just above the lateral line; scale-margins ventrally dark; top of head dark, becoming blackish on the snout, rest of head heavily punctate with black; dorsal fin dusky, blackish on the front edge; adipose pale; caudal blackish throughout; anal black on basal half; ventrals dusky, with dark specks medially; pectorals dark, blackish on upper border. Females pale in color, showing the same markings. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 151 In the breeding male the entire head and body, and both sides of all the fins except the caudal and adipose are thickly covered with small warty pearl-organs, not represented in the females, even by rudiments. The skin is everywhere turgid, the scales being particularly soft and thick. The lower fins are greatly expanded and lengthened in the breeding male; the pectorals reaching to within less than the length of the eye from the insertion of the ventral fins, which in turn extend to (or nearly to) the origin of the anal, each fin being nearly as long as the head; the anal, when depressed, reaches to the root of the lower caudal rays, and its longest rays are contained 1.6 to 1.7 times in the head. The membranes of these lower fins are widened and the rays thickened, so that the fins are spread in fan-like fashion. In the females each of the lower fins falls short of the next fin by a dis- tance about equal to the length of the eye, and the highest anal ray is only two- fifths as long as the head. Our breeding males are 134 to 148 mm. long to caudal fin; ripe females 129 to 136 mm. Genus Hypomesus Gill. The generic name i^ypomcsws Gill under the present rules should be adopted in preference to Mesopus Gill, given to the same group on the preceding page, but left as a typographical error (failure to completely correct proof). 80. [130] Hypomesus olidus (Pallas). ITa/casa(/f = Pond-smelt. Lake Kawaguchi, (Masashi Ishikawa); Lake Yamanaka (near Fuji-san); Nagano (Nakano); Lake Suwa (Ota); Lake Suwa (Jordan); Lake Mikata; Fukui, Akita, Aomori, (Beppu) ; Iwate (Awaya) ; Lake Kasumigaura (Hattori) . It is not entirely clear that Salmo inghaghitsch Walbaum can be identified with this species, for which Pallas’ name olidus may apparently be retained. Hypomesus olidus differs sharply homHtjpomesus preliosus of California in the smaller size of the scales (there being fifty-four to sixty-two rather than about seventy in the course of the lateral line); in the more anterior position of the ventral fins in reference to the dorsal, the ventral insertion being usually in advance of, instead of a little behind, the vertical from the origin of the dorsal; in the much larger size of all the fins ; in the darker color ; and in the habit of spawning in fresh- water ponds, rather than in the surf of the sea. The other species known from Japan, Hypomesus japonicus {Osmerus oligodon Kner), differs from H. olidus and resembles H. preliosus in all of the characters listed above, with the exception of the size of the scales, which is intermediate, and possibly of the breeding habits, which have not been recorded foriJ. japonicus. Hypomesus olidus and H. preliosus agree with one another and with the species 152 MEMOIRS OF THE CARNEGIE MUSEUM. of Osmerus and Plecoglossus in having the head, body, and fins in the breeding male almost completely covered with minute warty pearl-organs, and the fins thickened and enlarged. H. olidus ranges from the lakes of central Japan northward through Kam- chatka and southern Alaska to the Arctic shores of North America. We have compared series from the different lakes of Japan, but fail to find any consistent differences. The dorsal rays vary from 8 to 10, of which 2 or 3 are simple; the anal rays from 15 to 18, 3 or 4 unbranched; the scales along lateral line to caudal base, from 54 to 62. Our largest male is 126 mm. long to the caudal fin; the largest female, 113 mm. long. The specimens from the southernmost locality. Lake Suwa, are dwarfed, but not otherwise different. We may note here that the records of Osmerus thaleichthys from the Nushagak River, Alaska, which have several times appeared in reports^^ refer to the present species, as we have determined by a re-examination of the material involved. Spirinchus thaleichthys is therefore to be eliminated from the lists of Alaskan fishes. 81. [131] Hypomesus japonicus (Brevoort). We identify with this species two specimens taken by Tanaka at Kushiro, Hokkaido. Dorsal rays, 10 (two simple); anal, 14 or 15 (three unbranched); the last ray counted as branched from base. Scales to caudal base, 66 to 68. Ventral fin inserted a little behind the dorsal origin. Hypomesus japonicus seems to be most closely related to H. pretiosus, with which it is compared in the account of the preceding species. It is probably the Western representative of that form ranging from Japan northward to Kamchatka. Material from the latter region (collected at Petropavlovsk) was erroneously iden- tified by Jordan and Gilbert'’® with Mesopus olidus. Family ARGENTINIDTl. 82. [133] Argentina semifasciata Kishinouye. = Second Gisu. Toyama; Noo, (Niigata). Our seven specimens are all from the Sea of Japan. Gilbert, Kept. U. S. Fish Comm., 1893, (1895), p. 400. — Jordan and Evermann, Bull. U. S. N. M., XLVII, pt. 1, 1896, p. 522. — Jordan and Gilbert, Kept. Fur Seal Inves., Ill, 1898, j). 440. — Evermann and Goldsborough, Bull. Bur. Fish., XXVI, 1906 (1907), p. 40 (not figure). Kept. Fur Seal Invest., Ill, 1898, p. 440. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 153 Family SALANGIDvE. 83. [135] Salangichthys microdon (Bleeker).” = White-fish. Aomori (Beppu); Lake Kasumigaura (Hattori) ; Yamada (Tsuchiga). 84. [136] Parasalanx ariakensis (Kishinouye). By re-examining the paratype of this species we are able to confirm Oshima’s'*® reference of this species to Parasalanx. 85. [Extraterr.] Hemisalanx prognathos Regan. A breeding pair received from Soo-chow, China (Gee). Family GONORHYNCHIDiE. 86. [137] Gonorhynchus abbreviatus Temminck and Schlegel. Nezumi-gisu = Rat-Gisu. One specimen, Mikawa Bay (M. Ishikawa). Head, 4.25 in length to caudal; depth, 8.7; pectoral fin, 5.7. Eye, 4.65 in head; snout, 2.45; interorbital, 4.25. Dorsal rays, 2.8; anal, 2.6; scales, 21-166-16. Family AULOPIDtE. 87. [138] Hime japonica (Gunther). Hime = Princess. Ten specimens from Misaki (Aoki). The species, hitherto called Aulopus japonicus, differs from Aulopus in im- portant respects, and is the type of the genus Hime Starks. (Copeia, March 29, 1924, p. 30.) Family ASTRONESTHID^. 88. [144] Astronesthes iijimai Tanaka. T'o/ca(7e-/RRiafca = Naked Lizard. We have fourteen specimens of this deep-sea fish, collected by Owston in the Sagami Sea, and eleven taken by Aoki at Misaki. The species needs to be compared with the Indian A. martensi and the Hawaiian A. lucifer. The photophores in our material of Astronesthes iijimai vary as follows: operculars 2; branchiostegals 18 to 21; jugulars 8; upper thoracics (to ventral fin) 18 or 19; lower thoracics 22, the third and fourth opposite the pectoral bases; upper ventrals, 21 or 22, a few overlapping the front of anal base; the lower ventrals 21 to 23, 4 (rarely 3) opposite or before ventral bases, the last 2 opposite anal base (counted by Gilbert and Tanaka as the first two of the anal series) ; anals 7 or 8, the last one to four variously elevated; caudals 4 or 5. We have not compared our material with the descriptions (in Japanese) of Salangichthys kishi- nouyei and S. ishikawce Wakiya and Takahashi (Zool. Mag., XXV, 1913, pp. 551-555). Ann. Car. Mus., XII, 1919, p. 174. 154 MEMOIRS OF THE CARNEGIE MUSEUM. Family MAUROLICIDtE. 89. [145] Maurolicus pennanti (Walbaum).^® = Cucumber. Five specimens from Toyama, Sea of Japan, (Yoshizawa). We find the following photophores present in these examples: one before and one behind middle of eye; one before and one behind preopercular angle; six in branchiostegal series ; one on chin (on each side) ; six on isthmus ; nine of irregular shape in the series from pectoral base to above anal origin; twelve from isthmus to ventral fin; two behind ventral fin, and then a series of four from between these to anal origin; one above anal origin over the gap between the ventral and anal series; fourteen to sixteen anals, the last narrowly separated from first of the eight or nine paired postanals; last postanal median. Two specimens from the Mediterranean Sea at Messina agree exactly in the number and arrangement of the photophores. Family SYNODONTID.E. 90. [151A] Synodus fuscus Tanaka. Tanaka has lately described three new species of Synodus from Japan. From a translation of his paper and a study of one specimen of S. fuscus we have prepared the following: Key to the Japanese Species of Synodus. a. Scales relatively small, about 65 in lateral line japonicus. aa. Scales larger, 45 to 55 in lateral line. b. Eye much shorter than the sharp snout, about 6 to 7 in head; dorsal, 11 or 12; anal 9 or 10. fuscus. bb. Eye about as long as snout, 4 to 4.5 in head. c. Dorsal rays, 12; anal, 11; scales, 52 macrops. cc. Dorsal rays, 13; anal, 9; scales, 55 hoshinonis. The following is the substance of the original description of Synodus fuscus : Head, 3.5; depth, 7; eye, 6.25; interorbital width, 6; snout, 3.8; depth of caudal peduncle, 6. Dorsal rays, 11; anal rays, 10; pectoral, 12; ventral, 8. Scales 3.5- 53-5.5. Snout about half longer than the eye. Ventral fin inserted a little in ad- vance of dorsal origin; tip of pectoral reaching ventral insertion; caudal emarginate. Color greenish, the lower side whitish, marked on the sides with nine indistinct blotches; dorsal also marked; pectoral dark, ventral and anal white; margin of caudal dark. Tokyo market. We do not have at hand the description of Maurolicus japonicus Ishikawa (Jour. Coll. Agric., VI, 1915, pp. 183-191, 2 pis.). Zool. Mag., XXIX, No. 340, 1916, pp. 37-38. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 155 Our single specimen of Synodus fuscus (collected by Masashi Ishikawa in the Bay of Mikawa) shows the following characters: Head, 3.8; depth, 7.4; eye, 6.6; interorbital, 8.5; snout, 4.0; depth of caudal peduncle, 5.6. Dorsal, 12; anal, 9; scales 4-52-6. 91. [148A] Saurida grandisquamis Gunther. Saurida grandisquamis Gunther, Cat. Fishes Brit. Mus., V, 1864, p. 400, (Louisiades, Gunther) Fische der Slidsee, III, 1909, p. 377. — Weber and DE Beaufort, Fishes Indo-Austral. Arch., II, 1913, p. 141. — Fowler and Bean, Proc. U. S. N. M., LXII, 1922, 3 (Sept.). Saurida macrolepis Tanaka, Zook Mag., XXIX, 1916, 39. Tanaka’s account of Saurida macrolepis agrees so well with the descriptions of Saurida grandisquamis noted above, that we accept the two as probably identical. The specimens described by Fowler and Bean came from Formosa. 92. [149] Saurida argyrophanes (Richardson). Eso. Misaki (Aoki); Kagoshima Bay (Wakiya); Tokyo and Kobe markets (Jordan). Scales 48 to 52 between gill-opening and end of last vertebra. Pectoral fin reaching to above insertion of ventral. Adipose eyelid not very wide or thick. 93. [150] Saurida eso Jordan and Herre. T'oA;aj/e-eso = Lizard-Eso. Tokyo and Kobe markets (Jordan); Mikawa Bay (M. Ishikawa); Toyama (Yoshizawa); Miyazu, Noo. We have also examined a specimen of this species collected by the late Professor Walter Fong at Hong Kong, China. Scales 5 or 6 — 61 to 63 (to caudal base) — 7. Weber and de Beaufort^^ refer both S. argyrophanes and S. eso to the synonymy of Saurida tumhil (Bloch). We are not prepared to identify either of these Chinese- Japanese species with S. tumbil, and are certain that they differ from one another. Saurida eso has the scales constantly smaller; the pectoral fin shorter, not reaching the ventral insertion; and the adipose eyelid wider and thicker. It also attains a larger size. This important species is valued in the preparation of “kamoboku.” 94. [152] Trachinocephalus myops (Forster). = Off-shore Eso. Toba market (Jordan and Yamamoto); Tokyo and Kobe markets (Jordan); Mikawa Bay (M. Ishikawa); Miyazu, Misaki (Aoki). The Fishes of the Indo-Australiaii Archipelago, II, 1913, p. 142. 156 MEMOIRS OF THE CARNEGIE MUSEUM. Family MYCTOPHIDtE. 95. Diaphus latus Gilbert. Diaphus latus Gilbert, Mem. Car. Mus., VI, 1913, 95, pi. XIII, fig. 1. This species may provisionally be retained in the genus Diaphus, pending a comparative study of the known species. Three specimens were obtained at Misaki (Aoki). The photophores are like those of the type in all essential respects, the only significant deviations affecting the anals. First anal as near as, or not much farther from, the second than is the third, the interspace being decidedly less than that between third and fifth organs of the series ; last antero-anal but little elevated on one side of one specimen. Genus Lamprossa Jordan and Hubbs gen. nov. This genus agrees with Diaphus, as now restricted, in all essential respects, except the development of light organs about the eye. These are well described by Gilbert as follows: “A well-developed supra-orbital luminous body, .... in the form of a narrow streak along the upper border of the orbit, not extending behind the pupil; a small superior pre-orbital, in its usual position above the nostril; a third luminous body occupying the position of both suborbital and inferior preorbital, extending on the inferior border of the eye to, or beyond, the middle of the orbit, becoming abruptly narrowed beneath front of eye and sending a narrow upward extension nearly to level of the superior pre-orbital.” Type: Diaphus anteorbitalis Gilbert. A Hawaiian species, Diaphus adenomus Gilbert, approaches Lamprossa in the development of orbital photophores, and may for the present be referred to this genus. 96. [156B] Lamprossa anteorbitalis (Gilbert). Diaphus anteorbitalis Gilbert, Mem. Car. Mus., VI, 1913, p. 92, pi. XII, fig. 1. Six specimens were taken by Aoki at Misaki. They agree well with the types in the disposition and number of the photophores, the only point of deviation being the height of the first antero-anal. This photophore may occupy any posi- tion between the horizontals from the lower and the middle supero-anals. Genus Pantophos Jordan and Hubbs, gen. nov. We separate this genus from Diaphus, as the type species has glandular bodies connected not only with the suprapectoral photophore, but also with nearly all JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 157 of the other light-organs on the side of the body. The antorbital photophore is confined to a cavity above the nostril, and is directed forward, as in several species of Diaphus, but there is also characteristically developed a large suborbital photo- phore, located on the lower orbital margin. The type species, Diaphus glanduUfer Gilbert, alone is known. 97. ’[156C] Pantophos glandulifer (Gilbert.) Diaphus glandulifer Gilbert, Mem. Car. Mus., VI, 1913, p. 90, pi. XI, fig. 2. One specimen from Misaki, (Aoki). Dorsal rays, 15; anal, 16; pectoral, 11 or 12; ventral, 8; scales, 36; head .29 and eye .08 of total length. Photophores arranged exactly as indicated by Gilbert, and of the same number, with the exception that there are six postero-anals on one side. 98. [159A] Nyctimaster jordani (Gilbert). Type from the east coast of the Hokkaido. Family ARIIDA5. 99. [162A] Netuma osakse Jordan and Kasawa, sp. nov. (Plate IX. Fig. 1.) Type, a specimen 37.5 cm. long to caudal fin, found by Jordan and Yama- moto in the fish-market at Osaka (Car. Mus. Cat. Fishes, No. 7808.) This species differs widely from the common East Indian form, Netuma thalassina, the only species hitherto recognized in the genus, in having the top of the head covered with smooth skin, the granular bony area being chiefly restricted to the median portion of the occipital process. The snout is more bluntly rounded, when viewed from above, in N. osakce, and the body is a little deeper, although the caudal peduncle is slenderer; the fins are lower. The differences in proportion are indicated by the following comparison of the type of N . osakce with a Philippine specimen of A. thalassina of comparable size. Netuma thalassina Netuma osakce Width of granular area on occipital process in length of process 1.4 6.0 Length of snout in head 2.5 2.7 Depth of body in length 4.6 4.0 Depth of caudal peduncle 4.0 4.6 Length of dorsal spine in head 1.6 1.8 Length of longest dorsal ray in head 1.35 1.6 Length of longest anal ray 2.4 2.8 Length of ventral fin in head 1.5 2.1 This is the first Sea-catfish to be definitely recorded from Japan, Thunberg’s type of Silurus maculatus {Arius maculatus) having perhaps come from China. 158 MEMOIRS OF THE CARNEGIE MUSEUM. Body rather robust, highest at front of dorsal fin, wedge-shaped anteriorly to the snout, which is sharply pointed, when viewed from the side, but very broadly rounded, when viewed from above; behind the deepest point the contours are gently convex to the caudal peduncle, which is narrow medially, but abruptly expanded at base of caudal fin. Head 3.2 times in length to caudal; its width about two-thirds its length and one-fourth greater than its extreme depth, and much narrower than the body is deep (in thalassina width of head and depth of body about equal). Orbit 7 in head; greatest interorbital width 2.1. Top of head covered with smooth skin; weak granulations appearing only in a small patch on each side of occiput; occipital process granular medially, but its sides wholly smooth; the granulations become weak posteriorly toward the very small, smooth, basal plate at dorsal origin; interorbital space weakly arched, without the median depression found in N. thalassina. Maxillary barbels unequal, the left, when laid straight back, not reaching far beyond the vertical from posterior border of eye, right maxillary barbel extending a little beyond bony margin of opercles; outer mental barbels a little longer than inner ones, about as long as maxillary; posterior nostril with a pair of flaps along its hinder border, which are just covered by the large semi-circular flap arising between the two nostrils. Teeth all villiform, forming wide bands; premaxillary band arcuate and continuous; mandibular band more curved, and interrupted at symphysis. Vomer with a broad band made of two roundish subequal patches on each side; palatines with a longer posterior patch fitting into a broad notch between the two members of the vomerine set, its form triangular, prolonged backwards. Gill-membranes forming a wide fleshy fold; gill-rakers rather stiff, cylindrical, 6-b6 in number, the longest one-third as long as eye. Dorsal spine rather obscurely granulate anteriorly (not granulate- serrate as in thalassina) ; adipose fin rather large, free ; located above middle of anal base; anal fin falcate; caudal deeply forked, the upper lobe the longer and the more sharply pointed; pectoral spine about as long as dorsal spine, obscurely granulate on front margin, and weakly serrulate posteriorly. Color rich brownish above, grayish white below; dorsal and anal fins pale brown; caudal fin brown, with a greenish tinge on the lobes; paired fins brown on upper surface, gray below. Family PLOTOSIDTl. 100. [163] Plotosus anguillaris (Forskal). Umi-gigi — Sea, Gigi. Misaki (Aoki); Mikawa Bay (M. Ishikawa); Wakanoura (Yamamoto); Toba market (Jordan and Yamamoto). This species is annoyingly common along the shore of Kytisyu. JORDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 159 Family SILURID^. 101. [164] Parasilurus asotus (Linnaeus). Namazu = Catfish. Fresh water at Kumamoto (Wakiya); Ozu (Wakiya); Himeji (Abe); Nagano (Nakano) ; Lake Suwa (Ota) ; Lakes Suwa and Hakone (Jordan) ; Fukuoka (Hamada) ; Hino River, Lake Kasumigaura (Hattori). This species is generally abundant in the quiet parts of streams in the southern half of Japan. It reaches a larger size than the other N amazu (Cat-fishes) of the rivers. Dorsal rays 4 to 6; anal rays 72 to 83. The coloration varies from plain to brightly mottled, the sharpest contrasts appearing in a specimen from Himeji. The lateral lines are connected across the back by vertical branches, as indicated by Schlegel in his figure of Silurus japonicus. Family BAGRIDiE. 102. [165] Pelteobagrus* nudiceps (Sauvage). H age-gigi = Bald Gigi. Lake Biwa (Wakiya) ; Lake Biwa (Jordan) ; Himeji (Abe) ; Okayama(Mikamo). Anal rays, including rudiments, from 18 to 23 in number. The young are very coarsely blotched or vertically barred with pale areas, while the adults are uniformly marked, or streaked with two lengthwise light areas. In well preserved specimens the bones of the head are entirely covered with skin. 103. [Extraterr.] Pelteobagrus fulvidraco (Richardson) . A specimen collected by Gee at Soo-chow, China, has been sent to the Museum of Zoology, University of Michigan. 104. [166] Pseudobagrus aurantiacus (Temminck and Schlegel). Gigi. Fukuoka (Hamada). Like Pelteobagrus nudiceps, this species may be either coarsely blotched, or striped with light areas. 105. [167] Liobagrus reini Hilgendorf. A fca2:a = Red Cat-fish. Himeji (Abe) ; Toyama (Yoshizawa) ; Lake Biwa (Jordan) ; Lake Kitagata, Fukui. The variations in proportions in this little Cat-fish are very wide, but are not correlated with one another. We thus agree with Tanaka, that Liobagrus sugubrii Regan^^ cannot be recognized. Interocular space, 2.45 to 3.0 in head; dorsal spine, 3.0 to 5.5; depth of body, 5.0 to 7.3 in length to caudal; pectoral spine varying from much less to much more than half the length of that fin. *The group called Fluvidraco by Jordan and Fowler apparently cannot be separated from Pelteo- bagrus Bleeker, based on Silurus calvarius Basilewsky. Ann. Mag. Nat. Hist. (8) I, 1908, p. 152; Proc. Zool. Soc. London, 1908, p. 62. 160 MEMOIRS OF THE CARNEGIE MUSEUM. Family COBITID^E. 106. [168] Misgurnus anguillicaudatus (Cantor). Z)ojo = Loach. Nagano (Nakano); fresh water at Kumamoto (Wakiya); Himeji (Abe); Toyama (S. Yoshizawa); Noo, (Awaya); Lake Kasumigaura (Hattori); Lake Biwa at Otsu (Jordan and Kawamura); Lake Suwa (Jordan); Soo-chow (Gee). Extremely common in streams and lakes. One of the Lake Biwa specimens has only nine barbels, there being but one on one side of the mandible. 107. [170] Hymenophysa curta (Temminck and Schlegel). = Sea-loach. Three specimens of this peculiar loach were collected in Lake Biwa at Otsu by Jordan and Kawamura. 108. [172] Lefua echigonia Jordan and Richardson. Hotoke-dojo = Buddha-loach. Lefua nikkonis Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 768, fig. 1 (specimens from Shimotsuke, near Nikko, but not the type from Chitose, Hokkaido) . Lefua echigonia Jordan and Richardson, Proc. U. S. N. M., XXXIII, 1907, p. 264, fig. 1. — Tanaka, Annot. Zook Jap., VII, 1909, 129. With some doubt we refer to this species two specimens from the Hiki River in Kishu, collected by H. Kuroiwa, and bearing the local name “Hotoke-dojo.” They differ from the type in their plain coloration, the spots on the body and fins being obsolescent, and in the slenderer build. The coloration is very variable, however, as Tanaka has indicated, and as we find in examining the specimens from Shimotsuke which Jordan and Fowler confused with their L. nikkonis. The difference in form seems to be largely due to the lesser deposition of fat, for the adipose folds on the caudal peduncle are not developed, and the whole texture is firmer; furthermore this variation is paralleled in other loaches, as in the common Misgurnus. 109. [174] Cobitis biwae Jordan and Snyder. AS/ima-dojo = Striped Loach. Kumamoto (Wakiya); Himeji (Abe); Okayama (Mikama); Lake Kasumigaura (Hattori); Noo, Lake Biwa at Otsu (Jordan and Kawamura). We have not reconsidered the validity of this abundant species, which scarcely differs from Cobitis tcenia of Europe. Genus Barbatula Linck. The generic name Barbatula Linck, 1790, antedates and replaces Oreias Sauvage, 1874, and Orthrias Jordan and Fowler, 1903. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 161 110. [173] Barbatula oreas (Jordan and Fowler). = Secondary Loach. This species is not represented in the collections at hand. Family CYPRINIDiE. 111. [175] Pseudoperilampus typus Bleeker. Zem-tonagfo = Genuine Chub. Lake Kasumigaura (Hattori); Lake Biwa (Jordan). Genus Hemigrammocypris Fowler. H emigrammocypris Fowler, Proc. Acad. Nat. Sci. Phila., LXII, 1910, p. 483. Type: Hemigrammocypris rasborella Fowler. Brevigobio Tanaka, Dobutsu-gaku Zasshi, Tokyo, XXVIII, 1916, p. 102. Type: Brevigobio kawabatce Tanaka, = H. rasborella Fowler. “No barbels about mouth; pharyngeal teeth 3-rowed, 4, 4, 2-2, 4, 4; dorsal fin inserted nearer to base of caudal than to tip of snout; a sharp ventral keel between anus and origin of ventral; lateral line decurved, incomplete, running along lower part of body, ending near last ray of anal.” {Tanaka.) This genus is in many ways similar to Rasbora, a genus now comprising many species of southern Asia and the East Indies, but differs from that genus in having the abdomen sharply keeled, and in the normal structure of the lower jaw. 112. [176A] Hemigrammocypris rasborella Fowler. Proc. Acad. Nat. Sci., Phila., LXII, 1910, p. 483. (Japan.) Brevigobio kawabatce Tanaka, Dobutsu-gaku Zasshi, Tokyo, XXVIII, 1916, p. 102 (Lake Biwa); Fig. Desc. Fishes Japan, XXIV, 1916, p. 420, pi. 115, figs. 339, 340 (Lake Biwa, pond near Tsu in Ise). Genus Acheilognathus Bleeker. We follow Jordan and Thompson"*^ in our treatment of the fishes of this group. 113. [177] Acheilognathus rhombea (Temminck and Schlegel). Kanehira = Flat-money. Kumamoto (Wakiya); Fukuoka (Hamada). The depth of the body in this minnow is highly variable, being contained from 2.2 to 2.8 times in the length of body to caudal base, increasing very irregu- larly with age. The branched rays in the dorsal fin vary from twelve to fourteen in number. Mem. Car. Mus., VI, 1914, 217-227. 162 MEMOIRS OF THE CARNEGIE MUSEUM. 114. [179] Acheilognathus limbata (Temminck and Schlegel). Bote. Kumamoto (Wakiya); Hamada (Wakiya); Himeji (Abe). We identify our half-grown specimens (43 to 53 mm. long to caudal) with this species. The body is less elevated than in adults, being contained 2.6 to 2.7 times in the standard length. Some of the specimens of similar size from Funayado (reported by Jordan and Fowler) are as slender as these, while others are deeper. The increase in depth with age is obviously irregular, as in A. rhombea. In some of the specimens the lateral line appears to be variously incomplete. Teeth 5-5, smooth, or with distinct traces of plications on one of the teeth. Scales about 33. Anal fin sometimes black-margined. 115. [180] Acheilognathus lanceolata (Temminck and Schlegel). Yari-tanago = Spear-chub. Eleven specimens from Kumamoto, collected by Wakiya, fully agree with Jordan and Thompson’s account of this species (L c., p. 224). In fin-formula the only variant has ten-branched anal rays. Thirteen others from Noo near Niigata are essentially similar. All have nine-branched anal rays, but three have nine- branched dorsal rays. Teeth, 5-5, with rather wide grinding surfaces, but with at most bare traces of plications. Scales 34-37. 116. [180A] Acheilognathus intermedia (Temminck and Schlegel). Bay of Mikawa (M. Ishikawa); Himeji (Abe); Lake Mikata, Lake Kasumi- guara (Hattori); Lake Biwa, at Otsu (Jordan and Kawamura); tributary of the Sumida River near Tokyo (Jordan). We follow Jordan and Thompson (pp. 223-227) in the identification of this species. The barbel varies considerably in length, being usually more than two-thirds as long as the snout; it is contained 1.0 to 1.7 times in length of snout. The branched fin-rays are: dorsal, 9 (8 to 10) ; anal, 10 (9 to 11). Scales to caudal base, 35 to 36. Pharyngeal teeth, 5-5, with wide grinding surfaces, but no plications. 117. [181] Acheilognathus cyanostigma (Jordan and Fowler). Lake Mikata near Fukui; Lake Biwa at Otsu (Jordan and Kawamura). Branched rays of dorsal, 8; of anal, 7 to 9. 118. [181A] Acheilognathus tabira (Jordan and Thompson). = Rice-chub; Acheilognathus limbata Tanaka, Annot. Zool. Jap., VII, 1909, p. 133 (not of Temminck and Schlegel). JORDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 163 Acheilognathus tabira Jordan and Thompson, Mem. Car. Mus., VI, 1914, p. 220, pi. XXV, fig. 1. Nagano (Nakano); Fukuoka (Hamada); Lake Mikata, Lake Biwa at Otsu (Jordan and Kawamura). The branched dorsal rays vary from 10 to 12; anal, 9 to 10. 119. [183A] Rhodeus kurumeus Jordan and Thompson. Rhodeus kurumeus Jordan and Thompson, Mem. Car. Mus., VI, 1914, p. 229, pL XXVI, fig. 2. Two small specimens of this species were collected by Wakiya at Kumamoto, not far from the type-locality, Kurume. Half-grown specimens show two dark blotches on the front of the dorsal fin, one near the base, the other near its tip. The depth of the type was given as 2._75, which appears to be a misprint for 2.25. This species needs to be compared with the Chinese Rhodeus ocellatus. Genus Gnathopogon Bleeker. Jordan and Thompson^^ have given a very useful revision of the species of this genus, of which Leucogohio Gtinther, Squalidus Dybowsky, and Otakia Jordan and Snyder are synonyms. We have a large amount of new material from Japan and Korea, and find it necessary to go over the group again. Key to the Known Species of Gnathopogon. a. Anus not in advance of the normal position just before the anal fin. Eye small, usually less than one-fourth length of head. Pectoral fin usually reaching less than two-thirds the distance to ventral insertion. b. Snout narrower; maxillary reaching nearly or quite to below front of orbit. Pharyngeals with 2 or 3 teeth in inner row. c. Pharyngeal teeth smooth or obsolescently serrate on the grinding surface (character not definitely described for the Asiatic species, group dd); barbel about as long as, or longer than pupil. d. Barbel much longer than pupil (1.2 to 1.7 in eye). Origin of dorsal usually nearer base of caudal than tip of snout. Scales in four to five rows above, and 34 to 39 rows along the lateral line. e. Body rather robust, its depth 3.4 to 4.3 in standard length; caudal peduncle less than twice as long as deep, its depth 7.4 to 8.7 in body; width of body about 1.75 in its depth. Length of pectoral fin 1.4 to 1.7 in distance between insertions of paired fins. Color lighter; less spotted. Lake Biwa to Kyusyu elongatus. ee. Body slenderer, its depth 4.0 to 4.7 in standard length; caudal peduncle about twice as long as deep, its depth 8.6 to 10 in body length; width of body about 1.5 in its depth. Length of pectoral fin 1.6 to 1.9 in distance between origin of paired fins. Color darker; more spotted. Lake Suwa and adjacent waters suwce. Mem. Car. Mus., VI, 1914, pp. 214-217. 164 MEMOIRS OF THE CARNEGIE MUSEUM. dd. Barbel scarcely as long as pupil. Origin of dorsal equidistant between tip of snout and caudal base. Caudal peduncle less than twice as long as deep. Asiatic mainland, f. Scales in 6 rows above and 40 along lateral line. Headwaters of the Yangtsekiang. tceniatus}^ //. Scales in 6 rows above and 36 along lateral line. Amur basin toeniatus Berg. (? not of Gunther. ///. Scales in 4 or 5 rows above and 36 to 38 along lateral line. Korea strigatus.*^ cc. Pharyngeal teeth strongly dentate on grinding surface; barbel much shorter than pupil; body slender; origin of dorsal nearer tip of snout than base of caudal. Scales in 42 rows to caudal base. Lake Biwa and Yodo River, Japan coerulescens. bh. Snout broader and shorter; the maxillary only reaching to below middle of snout. Pharyngeals with only one tooth in the inner row. Origin of dorsal nearer tip of snout than base of caudal. China , herzensteini^^ aa. Anus in advance of normal position just before anal fin. Eye usually large, more than one-fourth length of head, e.xcept in large specimens of G. japonicus. Pectoral fin usually reaching more than two-thirds distance to ventral insertion. Origin of dorsal nearer tip of snout than base of caudal. g. Distance from anus to anal fin less than length of eye. h. Scales above lateral line in 2.5 to 3.5 rows; scales more or less enlarged on mid-dorsal line. Body with conspicuous dark spots; top of head spotted. i. Barbel minute, about one-sixth as long as pupil. Scales above lateral line in 3 or 3.5 rows; scales of mid-dorsal line little enlarged. Ping-yang River, Korea. rnajimce. a. Barbel about two-thirds as long as eye. Scales above lateral line in 3.5 rows; scales of mid-dorsal line moderately enlarged. KyusyO gracilis. Hi. Barbel longer than eye, but not reaching to below hind border of eye. Scales above lateral line in 2.5 rows; scales of mid-dorsal line much enlarged. Head 3.45, eye, 3.4, scarcely shorter than snout, longer than interorbital, 1.4 in post- orbital. Ping-yang River, northwestern Korea longifilis. hh. Scales above lateral line in 4 rows; (enlargement of scales on mid-dorsal line not described). Body with dark spots; top of head spotted. Barbel at least as long as eye, reaching to below hind margin of eye. Head, 4.0; eye, 3.5 to 3.6, shorter than snout, shorter than interorbital, about 1.5 in postorbital. River Sambu, southern Korea coreanus. hhh. Scales above lateral line in 4.5 or 5 rows; scales not enlarged along mid-dorsal line. Body with dark spots, inconspicuous; top of head not spotted. j. Barbel nearly as long as eye, but scarcely reaching to below hind border of pupil. Head, 3.7. Eye large, 3.15 in head, as long as snout, longer than inter- orbital, nearly as long as postorbital. River Ping-yang tsukhigce. jj. Barbel short, not reaching to below middle of pupil, 0.3 to 0.6 as long as eye. k. Eye 3.5 longer than snout or interorbital, 1.4 in postorbital. Scales 34 to 36. Chanka Lake, Amur basin chankanensis.^^ kk. Eye about 4.0 shorter than snout, little more than half postorbital. Scales 35 to 37. Streams of southern Japan japonicus. Gunther, Ann. Mus. Zook, Acad. Imp. Sci. Petersb., 1896, p. 214, pi. 2, fig. A. Berg, Ichth. Amur., 1907, p. 84. Regan, Proc. Zool. Soc. London, 1908, p. 59, Chong-ju, Chung-Chong Province, Korea. Giinther, 1. c., p. 213, pi. 2, fig. B. Berg, Ann. Mag. Nat. Hist., (7) XVIII, 1906, p. 394. Dybowsky, Verb. Zool. Bot. Ges. Wien. XXII, 1872, p. 215; Berg, Ichth. Amur., 1909, p. 83. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 165 gg. Distance from anus to anal fin at least as long, usually longer, than diameter of the large eye. Depth of body 4.5 to 6.0 in length. 1. Scales above lateral line in 4 rows, along lateral line in 33 rows. Head more than one-fourth the standard length. Formosa ll. Scales above lateral line in 5 rows, along lateral line in 39 rows. Head less than one-fourth standard length. Lake Biwa. Japan biwoe}^ 120. [184 and 188] Gnathopogon elongatus (Temminck and Schlegel). Moroko = Minnow. Capoeta elongata Temminck and Schlegel, Fauna Jap., Pisces, 1846, p. 200, pi. 100, fig. 1. Gnathopogon elongatus Bleeker, Ned. Tijdschr. Dierk., I, 1863, p. 202 (as type of Gnathopogon) ; Verb. Akad. Amsterdam, XVIII, 1879, p. 23. — Jordan and Snyder, Proc. U. S. N. M., XXIII, 1901, p. 343 (Lake Biwa); Annot. Zool. Jap., Ill, 1901, p. 47 (Lake Biwa). — Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 822 (Lake Biwa). — Jordan, Tanaka, and Snyder, Jour. Coll. Sci., Tokyo, XXXIII, 1913, p. 66. — Jordan and Thompson, Mem. Car. Mus., VI, 1914, pp. 215, 217 (Lake Biwa; Nagoya; Yodo River). Barbus homogenes Gunther, Cat. Fishes Brit. Mus., VII, 1868, p. 136 (after Schlegel; the name elongatus pre-occupied in Barbus). Leucogobio guentheri Ishikawa, Annot. Zool. Jap., Ill, 1901, p. 161, pi. 3, fig. 1 (Matsubara on Lake Biwa). — Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 826 (Lake Biwa; Nagoya; Hatata; Chikugo River). — Berg, Ann. Mag. Nat. Hist., (7) XVIII, 1906, p. 395 (in key). — Tanaka, Annot. Zool. Jap., VII, 1908, p. 6 (Lake Biwa); Zool. Mag., No. 237, 1908, p. 235 (Lake Biwa) . \ Kachi River, at Nagoya (Jordan); Lake Biwa (Wakiya); Lake Kawaguchi, Mikawa Bay (M. Ishikawa); Himeji (Abe); Kumamoto (Wakiya); Ozu (Wakiya); Nagasaki. Counts and measurements of a considerable series from various parts of Japan follow; head, 3.4 to 4.0 in standard length; depth of body, 3.4 to 4.3; depth of caudal peduncle, 7.4 to 8.7; eye, 4.2 to 5.0 in head, 1.2 to 1.7 times as long as barbel; pectoral fin contained 1.4 to 1.7 times in interval between insertions of paired fins; scales in lateral line to caudal base, 34 to 39; gill-rakers, x plus 6; teeth 5, 5-3, 5. Oshima, Ann. Car. Mus., XII, 1919, 219, pi. LI, fig. 2, p. 128 (Tozeh River, Formosa); Proc. Acad. Nat. Sci. Phila., 1920, p. 125 (Rihikuton, Formosa). Berg, 1. c., 1909, p. 84, footnote, indicates the existence in Lake Chanka of a second and unnamed species, which he regards as related to G. biwoe. 166 MEMOIRS OF THE CARNEGIE MUSEUM. 121. Gnathopogon suwae Jordan and Hubbs, sp. nov. Type, 72 mm. long to caudal base, collected by Jordan in Lake Suwa, at Kamisuwa in Shinshu (Car. Mus. Cat. of Fishes No. 7814). Eight paratypes were taken with the type; five other paratypes were taken in Lake Kisaki by T. Ota. This species seems to be the local representative of Gnathopogon elongatus, which is widely spread to the southward. It also bears a fairly close resemblance to several species of the Asiatic mainland. It is compared with these and other species in the key, which has been given. Form moderately trim, heaviest forward; the nape slightly elevated; width of body about two-thirds the depth, which is contained 4.2 (4.0 to 4.7) times in standard length; length of the rather slender caudal peduncle about twice its least depth, which measures 9.4 (8.6 to 10.0) times in length to caudal. Head symmetrical, rounded anteriorly; its length, with opercular membrane, contained 3.8 (3.7 to 4.0) times in length to caudal; eye round and rather small, being contained 4.7 (4.4 to 5.0) times in head; decidedly shorter than either the snout or the convex interorbital, less than half the postorbital. Barbel well de- veloped, but not long, reaching to below middle of eye; its length contained 1.4 (1.3 to 1.5) times in eye; upper jaw as long as snout, but being somewhat oblique, does not quite reach to below front of orbit; tip of premaxillaries on a level with lower part of pupil. Six gill-rakers on lower limb of arch, not counting extreme rudiments. Pharyngeal teeth more or less hooked, with grinding surfaces, which are at most obsoletely serrate; 4 or 5 teeth in the outer and 2 or 3 in the inner row. Peritoneum silvery, with clusters of black dots. Intestine shorter than body. Anus not advanced in position. Scales 4.5 to 5.5 between origin of dorsal fin and lateral line, 35 to 37 along lateral line to caudal base, 3 or 3.5 between lateral line and anal origin, and 3 to 4 between lateral line and ventral insertion. Dorsal rays, 2, 7 (rarely 6); anal, 2, 6; all fins short; pectoral extending much less than two-thirds distance to ventral, its length 1.75 (1.6 to 1.9) times in interval between origins of the two fins; ventral not reaching anus, extending only two- thirds to origin of anal; anal, when depressed, falling short of lower caudal rays by a distance longer than eye. Length of depressed dorsal contained 1.8 (1.6 to 1.8) times in distance from its origin to occiput. Origin of dorsal over ventral base, a little nearer base of caudal than tip of snout (sometimes almost equidistant) . Color grayish brown above, pale below. A dark stripe extends from the origin of the lateral line to the caudal base, becoming more distinct posteriorly, and running largely above the slightly decurved lateral line anteriorly. Dark JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 167 spots of varying form and distinctness developed on the upper half of the body; most distinct in a paratype from Lake Kisaki. 122. [189, 198, and 203] Gnathopogon ceerulescens (Sauvage). Y anagi-moroko — Willow-minnow. Squalius ccerulescens Sauvage, Bull. Soc. Philom. Paris, 1883, p.3 (Lake Biwa). Leuciscus ccerulescens Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 845 (after Sauvage). Otakia rasborina Jordan and Snyder, Proc. U. S. N. M., XXIII, 1900, p. 345, pi. 9, fig. 3 (Lake Biwa); Annot. Zool. Jap. Ill, 1901, p. 46 (Lake Biwa). — Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 842, fig. 6. — Tanaka, Zool. Mag., No. 237, 1908, p. 235. — Jordan, Tanaka, and Snyder, Annot. Zool. Jap., XXXIII, 1913, p. 70, fig. 44. Leucogohio jordani Ishikawa, Annot. Zool. Jap., Ill, 1901, p. 163, pi. 3, fig. 2 (Lake Biwa). — Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 827 (after Ishikawa). — Berg, Ann. Mag. Nat. Hist., (7) XVIII, 1906, p. 395 (in key). — Tanaka, Zool. Mag., No. 237, 1908, p. 235, (Lake Biwa); ibid., VII, 1908, p. 5 (Lake Biwa). — Jordan, Tanaka, and Snyder, Jour. Coll. Sci., Tokyo, XXXIII, 1913, p. 67 (after Ishikawa). Gnathopogon jordani Jordan and Thompson, Mem. Car. Mus., VI, 1914, p. 217 (in key). Lake Biwa (Wakiya, Jordan and Kawamura) ; Yodo River, outlet of Lake Biwa, at Kyoto (Jordan). We also have a series collected in Lake Biwa by Jordan and Snyder in 1900. The species currently known as Gnathopogon (or Leucogobio) jordani fits the description of Squalius ccerulescens Sauvage in all respects. Sauvage makes no mention, however, of the minute barbel, but he obviously overlooked it. Otakia rasborina seems also to be the same, though the barbel is described as absent. Jordan and Seale in 1906*^^ recorded some minnow from Kawatana, near Nagasaki, under the name of Leuciscus ccerulescens, but we do not locate this material. 123. Gnathopogon majimae Jordan and Hubbs, sp. nov. (Plate IX; fig. 2.) Type 43 mm. long to caudal fin, collected by Yojiro Wakiya in the Ping-yang River, northwestern Korea; Car. Mus. Cat. of Fishes, No. 7816. The type is unique. Proc. U. S. N. M., XXX, 1906, p. 144. 168 MEMOIRS OF THE CARNEGIE MUSEUM. Body rather heavy forward, attenuate posteriorly; dorsal contour rising from tip of premaxillaries, which are on the level of lower border of pupil, in a convex curve to above front of orbit, thence nearly straight to occiput, beyond which it is considerably elevated. Greatest depth of body below the dorsal origin, more than twice the least depth, and contained 4.0 times in standard length. Head rather long and narrow, its dorsal contour more curved and more elevated than the ventral; length of head 3.35 in standard length; eye slightly oval and large, longer than snout or interorbital width, contained 1.3 times in postorbital, 3.0 times in head; interorbital flat; barbel slender and very short, only about one-sixth as long as the pupil, so short as to be hidden in the groove at angle of mouth, not reaching to below front of orbit; length of upper jaw equal to distance from tip of snout to the fold between nostrils; about seven rudimentary gill-rakers developed on lower limb of outer gill-arch; pharyngeal teeth 5-3, obscurely hooked, with entire grinding surfaces; peritoneum silvery, with black dots; intestine short; anus in advance of anal fin a distance nearly equal to length of eye. Scales large, 3 or 3.5 from origin of dorsal fin to lateral line; 2.5 from lateral line to insertion of ventral or origin of anal; eleven scales somewhat, but not greatly, enlarged, from dorsal fin to occiput; only thirty along lateral line to caudal base. Dorsal rays, 2.7; anal, 2.6. Fins all long; pectoral extending just to insertion of ventral, slightly more than two-thirds as long as head; ventral extending slightly beyond anus; the depressed anal reaching within scarcely more than half length of eye from lower caudal rays; dorsal especially elevated, its height being equal to its distance from occiput. Origin of dorsal slightly in advance of ventral insertion, a little nearer tip of snout than base of caudal. Color pale, but with a deep ‘brown band from front of snout to base of caudal, extending everywhere along the nearly straight lateral line, most intense on the snout, more or less interrupted at each scale margin. Top of head, and the back, but not the sides of the body, with numerous brown spots somewhat smaller than the pupil; the spots less distinct, as though more blurred, than in G. longifilis. Named for Toyoji Majima of the Imperial University of the Hokkaido. 124. [185] Gnathopogon gracilis (Temminck and Schlegel). I to-moroko = Slim. Minnow CapoMa gracilis Temminck and Schlegel, Fauna Japonica, Pisces, 1846, p. 201, pi. 100, fig. 2 (Nagasaki). Gnathopogon gracilis Bleeker, Verh. Akad. Amsterdam, XVHI, 1879, p. 23 (after Schlegel). — Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 823 (after Schlegel). JORDAN AND HUBBS”. JAPANESE FISHES COLLECTED 1922. 169 Barbus homozonus Gunther, Cat. Fishes Brit. Mus., VII, 1868, p. 137 (after Schlegel; the name gracilis pre-occupied in Barbus). Leucogobio mayedce Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 828 (specimens from Chikugo River, and part of material from Lake Biwa; not Gobio mayedce Jordan and Snyder). — Jordan, Tanaka, and Snyder, Jour. Coll. Sci., Tokyo, XXXIII, 1913, p. 67 (in part; after Jordan and Fowler). Gnathopogon ishikawce Jordan and Thompson, Mem. Car. Mus., VI, 1914, p. 215, pi. XXIV, fig. 4, (Chikugo River). It appears almost certain to us that Gnathopogon ishikawce Jordan and Thomp- son, rather than Gobio biwce Jordan and Snyder, is identical with Capoeta gracilis Temminck and Schlegel. The figure of C. gracilis indicates a fish differing from G. biwcE in the lesser distance between anus and origin of anal fin; the deeper body, more elevated anteriorly, in the larger scales, etc. In these and other re- spects, except the height of the anal fin, which we find variable, the figure agrees with that of G. ishikawce. Moreover the type-localities of C. gracilis and G. ishikawce are in close proximity, while G. biwce has only been taken in Lake Biwa, far to the northward. The present collection contains no specimens of G. gracilis, but we have examined the type material of G. ishikawce. 125. [Extraterr.] Gnathopogon longifilis Jordan and Hubbs, sp. nov. Type, a fine specimen 49 mm. long to the caudal fin, collected by Yojiro Wakiya in the Ping-yang River, northern Korea; Car. Mus. Cat. of Fishes, No. 7819. We have no paratypes. The distinctive features of this form are indicated in our key. This species and G. majimce bear a strong superficial resemblance respectively to Notropis cayuga and Notropis heterodon, which live together in weedy waters in the eastern United States. The similarity probably has an ecological basis. The body is even more attenuate posteriorly and less elevated anteriorly, than in Gnathopogon majimce, the whole form being trimmer. Greatest depth, considerably more than twice least depth, contained 4.45 times in standard length; nuchal elevation slight; dorsal profile of head rather gently decurved to tip of premaxillaries, which are scarcely higher than lower margin of eye. Head, 3.45; eye, 3.4; scarcely shorter than snout, a little longer than width of interorbital, 1.4 in postorbital; interorbital almost flat. Barbel long and slender, a little longer than eye, but not quite reaching to below posterior margin of orbit; maxillary reaching to below posterior nostril. Only five rudimentary gill-rakers developed on lower line. Pharyngeal teeth 5-3, obscurely hooked; almost smooth on grinding 170 MEMOIRS OF THE CARNEGIE MUSEUM. surfaces. Peritoneum silvery, but rather densely punctate. Intestine short. Anus in advance of anal fin a distance contained 1.4 times in the orbit. Scales large; 2.5 from origin of dorsal to lateral line; 2.5 from lateral line to insertion of ventral or origin of anal; 11 greatly widened from origin of dorsal to occiput; 34 along lateral line to caudal base. Dorsal rays, 2.7; anal, 2.6; fins all long; pectoral reaching almost to ventral, contained 1.3 times in head; ventrals slightly over- lapping the anus; the anal when depressed, extends to within diameter of pupil from lower caudal rays. Height of the pointed dorsal fin a little greater than distance from dorsal origin to occiput; dorsal beginning considerably in advance of ventral insertion, diameter of one orbit nearer tip of snout than base of caudal. Back light brown, with numerous rich deep brown spots, which are also traceable on top of head. The spotted back is abruptly marked off from the sides by a pale streak, which overlies the lateral stripe. The latter begins at the eye, not being developed on the snout, as in G. majhnce, and extends along the straight lateral line to the caudal base, being more or less interrupted at the scale margins. 126. [Extraterr.] Gnathopogon tsuchig^ Jordan and Hubbs, sp. nov. Gnatho'pogon tsuchigce is one of the three new species of the genus collected by Yojiro Wakiya in the Ping-yang River, northern Korea. It is known only from the type, a specimen 69 mm. long to the caudal fin (Car. Mus. Cat. Fishes, No. 7818). Body deepest slightly before the middle of its length, little elevated at either the nape or the dorsal origin; dorsal contour gently curved from dorsal fin to snout, where it is decurved almost to the horizontal from lower margin of orbit; greatest depth of body below dorsal origin a little more than twice the least depth, and contained 4.8 times in standard length; caudal peduncle twice as long as deep. Head rather heavy and deep, its length 3.7 in body; eye large and roundish, its upper edge almost flush with the flat surface of the interorbital, its length equal to that of snout, exceeding the interorbital width, as long as the postorbital exclusive of the opercular membrane and contained 3.15 times in head. Barbel slender, of moderate length, contained 1.3 times in the eye, but scarcely reaching to below posterior margin of pupil; upper jaw somewhat projecting and extending backward slightly behind vertical from posterior nostril. Gill-rakers less rudi- mentary than in related species, only five, however, below angle. Three teeth in lesser row (those of the outer row appear abnormal in the type, two being very broad, the third very slender and separated by a gap from one of the wider ones). Peritoneum finely punctate; intestine short; anus in advance of anal fin a distance nearly equal to length of the large eye. Scales moderate, 4.5 from origin JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 171 of dorsal to lateral line, 2.5 from lateral line to ventral, 3 from lateral line to anal origin; about 13, not at all enlarged, fronl occiput to dorsal fin, 36 along lateral line to caudal base. Dorsal rays, 2.7; anal, 2, 6, as in related forms; pectoral fin one diameter of pupil shorter than interval between insertions of paired fins, which is equal to length of head; ventrals extending beyond anus, but not to anal fin, which, when depressed, reaches to within half of the diameter of orbit from caudal; length of the depressed dorsal scarcely shorter than occiput to dorsal; third dorsal ray over insertion of ventral, which is midway between tip of snout and base of caudal. Color pale, with clusters of dots forming faint roundish spots in a series before dorsal fin, and in a series along the axial septum of the longitudinal muscles, the latter forming a chord across the curve made by the considerably decurved lateral line. Head without spots. Named for Yasukei Tsuchiga, science-teacher at Yamada. 127. [190] Gnathopogon japonicus (Sauvage). Deme-moro/co = Pop-eye Minnow. Squalius japonicus Sauvage, Bull. Soc. Philom., 1883, p. 4, (Lake Biwa). Leuciscus japonicus Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 845 (after Sauvage). Gohio mayedce Jordan and Snyder, Proc. U. S. N. M., XXIII, 1900, p. 342, pi. 9, fig. 2 (Lake Biwa); Annot. Zool. Jap. Ill, 1901, p. 46 (Lake Biwa). Leucogohio mayedce Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 828, fig. 3, (in part; see Jordan and Thompson, 1911). — Berg, Ann. Mag. Nat. Hist., (7), XVHI, 1906, p. 395 (in key). — Tanaka, Annot. Zool. Jap., VII, 1908, p. 5 (Lake Biwa); Zool. Mag., No. 237, 1908, p. 235 (Lake Biwa). — Snyder, Proc. U. S. N M., XLH, 1912, p. 404 (Yamaguchi). — Jordan, Tanaka, and Snyder, Jour. Coll. Sci., Tokyo, XXXHI, 1913, p. 67, fig. 41 (in part). Gnathopogon mayedce Jordan and Thompson, Mem. Car. Mus., VI, 1914, pp. 215, 217 (Lake Biwa). Lake Biwa (Wakiya). Kachi River at Nagoya (Jordan). The species hitherto called Gnathopogon mayedce is doubtless the one described by Sauvage as Squalius japonicus. As in the case of G. ccerulescens, the small barbel was overlooked by Sauvage, and the species has been lately listed as a Leuciscus. 128. [191] Gnathopogon biwae (Jordan and Snyder). Gohio biwce Jordan and Snyder, Proc. U. S. N. M., XXIII, 1900, p. 340, pi. 9, fig. 1 (Lake Biwa); Annot. Zool. Jap. HI, 1901, p. 46 (Lake Biwa). 172 MEMOIRS OF THE CARNEGIE MUSEUM. Leucogohio biwcB Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 829, fig. 4 (Lake Biwa). — Berg, Ann. 'Mag. Nat. Hist., (7) XVIII, 1906, p. 395 (in key). — Tanaka, Zool. Mag., No. 237, 1908, p. 235 (Lake Biwa). — Jordan, Tanaka, and Snyder, Jour. Coll. Sci., Tokyo, XXXIII, 1913, p. 68, fig. 42 (after Jordan and Snyder). Gnathopogon gracilis Jordan and Thompson, Mem. Car. Mus., VI, 1914, pp. 214, 217, fig. 3 (Lake Biwa), not Capoeta gracilis Temminck and Schlegel. This species is still known only from Lake Biwa. Jordan and Thompson made an unsuccessful effort to identify it with Capoeta gracilis Temminck and Schlegel, and in the same paper redescribed what appears to be the true gracilis as Gnathopogon ishikawce. 129. [186] Hemibarbus barbus (Temminck and Schlegel). Afzj/of = Skin-carp. Osaka and Tokyo markets (Jordan); Lake Suwa (Jordan); Nagano (Nakano); Hino River, Fukui; Lake Kasumigaura (Hattori). Generally common southward in lakes and quiet streams. One young specimen from Nagano, otherwise typical, has the body and head much darker than usual, and the spots on the body more numerous and darker. The normal coloration at different ages has been well described by Tanaka. We have not sufficient material from outside Japan to test the alleged identity of this species with H emibarbus labeo of the Asiatic mainland. Genus Belligobio Jordan and Hubbs, gen. nov. Type: Belligobio eristigma Jordan and Hubbs. Belligobio is related to Hemibarbus, but differs in having the last simple ray of the dorsal fin unossified, flexible, articulated, not a coarse bony spine, and in having the snout more produced. Both genera are related to Gobio, having scales of the same type. Body elongate; head long, the snout sharply produced, longer than postorbital; a rather conspicuous flap between the nostrils; mouth very low, horizontal, chiefly lateral; upper lip normal, largely covered by the slightly projecting rostral fold; lower lip produced backward as a triangular median flap in contact with the gular groove; lower jaw entirely included; posterior maxillary barbels present, rather short, arising from lower edge of maxillary near its end ; gill-rakers short but thick, four on lower limb of outer arch; pharyngeal teeth 5, 3, 1, the larger ones with a broad and deeply excavated grinding surface; bones of snout expanded, sub- orbitals, preopercle, and mandible with highly developed mucous cavities; dorsal Annot. Zool. Jap., VII, 1909, p. 131. JORDAN AND HUBBS*. JAPANESE FISHES COLLECTED 1922. 173 fin with 7, anal with 6 branched rays; dorsal beginning a little before ventrals; anal far behind the dorsal; lateral line scarcely decurved, running on caudal peduncle a little below middle of depth. 130. [186A] Belligobio eristigma Jordan and Hubbs, sp. nov. (Plate IX; fig. 3.) Type, an adult female, 118 mm. long to caudal, collected by Kumachichi Mikamo near Okayama (C. M. Cat. Fishes No. 7820.) One paratype was taken with the type; another was collected by Abe at Himeji. Both of these localities are in the Inland Sea drainage of Hondo. Head, 3.65 (3.45 in paratypes) in standard length; greatest depth of body, 4.9 (4.6 to 5.2). Least depth of caudal peduncle about 2.0 in its length, 3.15 (to 3.2) in head; snout, 2.3 (2.2); interorbital, 4.0 (to 3.8); postorbital, 2.55 (to 2.65) ; eye to end of fold of mouth, 4.0 (to 4.4) ; length of upper jaw, 4.3 (4.0 to 4.4). Form rather slender, not particularly trim, the outlines not being even curves; body weakly compressed; snout much produced, considerably longer than postorbital; interorbital flat, bony, with three longitudinal ridges, of which the outer two diverge outwardly toward the nostrils. In preserved specimens a hori- zontal groove extends forward horizontally from below each nostril, joining its fellow in a wide curve, as viewed from above; anterior nostril with a raised rim, the posterior nostril with a fimbriate border; between the two there is a wide rounded flap, large enough to cover either nostril when depressed; extreme tip of snout abruptly decurved to form the rostral fold, which medially is on a level a pupil’s length below eye, and which partially overhangs upper jaw. The moder- ately thick upper lip slightly overlaps the base of the barbel, which is inserted on the lower edge of maxillary near its end; barbel about two-thirds eye, extending slightly beyond vertical from hinder rim of posterior nostril; upper jaw a very little longer than eye, but not reaching to below the anterior nostril; gape largely lateral, very little oblique; lower lip moderately thick laterally, deflected backward from the symphyseal region to the gular groove as a flap one-fifth as long as the orbit, and having the shape of an isosceles triangle; gill-rakers rather large, but short and fleshy, angular in form, 24-4 in number; pharyngeal teeth triserial, 5, 3, 1, the larger ones with broad, deeply excavated grinding surfaces and little hooked tips. Peritoneum silvery with brown spots; intestine shorter than body; air- bladder very large. Lateral line complete, barely decurved anteriorly, running a little below middle of depth on caudal peduncle. Scales moderate, 5.5 or 6 from origin of dorsal to lateral line, 13 between dorsal fin and occiput, 40 (39 to 41) 174 MEMOIRS OF THE CARNEGIE MUSEUM. along lateral line to caudal base, 3 (3 to 4) between lateral line and ventral inser- tion, 4.5 between lateral line and anal origin, 15 around narrowest part of caudal peduncle. The individual scales are broadly semioval in outline, with basilateral angles narrowly rounded; the basal margin truncate, with a broad median con- vexity; the focus within the basal fourth of the length of the scale; the basal circuli close-set and parallel with the basal margin, while the lateral circuli are moder- ately spaced, about as in most American CyprinidcB, and flare outwardly to meet the scale margin at an angle; the anterior field traversed only by rudimentary circuli, but with the radii rather strong and numerous, extending from very near the focus and from intermediate origins to points on the margin, between which the scale is produced in narrow scallops; no lateral or basal radii. Dorsal rays, 2, 7; anal, 2, 6; dorsal fin inserted well forward, its fourth ray lying over ventral insertion, its origin nearer by length of eye from tip of snout than caudal base; origin of anal but little nearer end of dorsal base than caudal base; length of the depressed dorsal contained 1.2 times in the dorso-occipital interval; pectoral reaching to within less than one diameter of eye from ventral, which fails to reach the anus by a somewhat greater distance; the depressed anal reaches about to middle of caudal base; second dorsal ray a very little stiffened basally, but still quite flexible; it remains thin and preserves the articulations, being essentially like the corresponding anal ray, and barely showing an approach toward the bony spine of Hemibarhus. The most striking feature of the species is the coloration, correctly indicated on Plate IX, Fig. 3. 131. [192] Pseudogobio esocinus (Temminck and Schlegel). Kamatsuka = Iliver Dodger. Lake Biwa, Kachi River at Nagoya (Jordan); Kumamoto (Wakiya); Lake Suwa (Jordan); Nagano (Nakano); Himeji (Abe); Okayama (Mikamo); Chikuma River (Ota); River at Yamada (Tsuchiga). A gudgeon of remarkable appearance, looking like the American genus Hypentelium. Genus Sarcocheilichthys Bleeker. In Sarcocheilichthys the barbels are very variable. In S. variegatus, though constantly short and slender in the young, they become shorter and thicker with age, often becoming reduced to papillae, or are even entirely lost. The lower lip is thick and fleshy on each side, but obsolete toward the hard and narrow symphy- seal projection; the gape as viewed from below is strongly sigmoid on each side. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 175 The anus is advanced in position, being almost as near the ventral insertion as the anal origin. The oviduct is extended as a long white tube, which is retractile within a large pocket, and which carries the rectum outward for some distance along its front border. In Pungtungia and Pseudorasbora the lower lip is also practically restricted to the lateral lobes; in Pungtungia the mouth is also inferior, but the median ex- tension of the mandible is broad and truncate, and the barbel is longer; in Pseudorasbora the mouth is dorso-terminal and transverse, and the barbel wholly lacking. In Biwia the lower lip is continuous, but everywhere thin, the mouth small and inferior, the gape a half-oval; in Abbottina the lower lip on each side is heavy and divided into a lateral lobe and an anterior almost barbel-like process; in Pseudogobio the greatly expanded lips, as well as the fleshy structure surrounded by the lower lips, are all strongly papillate. 132. [193] Sarcocheilichthys variegatus (Temminck and Schlegel). Higai = 'Red Dace; Sakura-bae = Cherry -Dsice, Lake Biwa (Jordan and Kawamura); Kachi River at Nagoya (Jordan); Kumamoto (Wakiya); Himeji (Abe); Lake Suwa (Ota); Lake Mikata, Lake Kasumigaura (Hattori). A common species in Lake Biwa. 133. [Extraterr.] Sarcocheilichthys morii Jordan and Hubbs, sp. nov. Type, a female 100 mm. long to caudal fin, collected by Dr. Ydjiro Wakiya in the River Ping-yang, Korea, (C. M. Cat. Fishes, No. 7824). Dorsal rays, 2, 7; anal, 2, 6; scale rows 4.5 or 5-39-5 (to anal origin) or 3.5 (to ventral insertion); pharyngeal teeth 5, 1-1, 5, weakly and bluntly hooked, with developed grinding surfaces; no barbels. Head, 4.2; depth of body, 3.8; depth of caudal peduncle two-thirds its length, 1,75 in head; eye 4.5; snout, 2.65; interorbital, 2.95. Mouth small, the upper jaw reaching to below nostril only; lower jaw pointed, included; lower lip thick and pendant on sides, interrupted in the middle. Fourth dorsal ray over insertion of ventral, and midway between tip of snout and base of caudal; top of dorsal straight; the first branched ray longest, 1.15 in head; caudal forked; pectoral rounded, its length 1.3 in head, or 1.3 in pectoral-ventral interspace; ventral fins reaching almost to anus, 1.45 in head; anal fin with straight margin, its longest ray, 1.6 in head; anus distant from anal fin, but nearer anal than ventral; rectum carried out for some distance along front of produced oviduct; lateral line straight, complete. 176 MEMOIRS OF THE CARNEGIE MUSEUM. General color pale, deepening on back, and marked by irregularly disposed, vertically elongate, deep brown spots, and by dark scale-edgings. Dorsal with some red medially, with a subterminal blackish bar on the rays only, black streaks distally on first two interradial membranes, and medially in last two membranes; base of rays black, the first branched ray for over one-third its length, those follow- ing for a progressively lesser distance; caudal lobes, especially the lower one, darkened mesially. Prof. Mori in his recently published “List of the Fresh -wgter Fishes of Korea,” simply records the names of this species. Two other species have been described from the Yangtse-Kiang and one from the Amur Basin. “ The Korean form appears to be related to Sarcocheilichthys maculatus (Gunther),®® but differs in various counts and proportions. We name this minnow for Professor Tamezo Mori, mammalogist of the Heijo High School, Seoul, Korea, who is making a study of Korean vertebrates. Pro- fessor Mori, at Stanford University, while these lines are being written, has a specimen of the same species (which we designate as a paratype), from the Han River, Korea. Genus Pungtungia Herzenstein. Fungtungia Herzenstein, Bull. Ac. Imp. Sci. St. Pet., XHI, 1892, p. 231 (P. herzi) . Zezera Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 837, (Z. hil- gendorfi Ishikawa). We have been able to examine specimens of Pungtungia herzi (the typical species of the genus) through the courtesy of Prof. Tamezo Mori, who collected them in the Han River, Korea. This Korean species usually has the teeth 5-5, as in P. herzensteini, and in fact differs but slightly from that species. The snout is somewhat longer, a little longer than the postorbital. The Japanese genus Zezera from Lake Biwa seems inseparable from Pungtungia. 134. [195] Pungtungia hilgendorfi (Ishikawa). Kumamoto, Hamada (Wakiya); Himeji (Abe); Yamaguchi (Ishikawa). Jordan and Fowler described this species on the basis of their own material, although Ishikawa had already named it in manuscript. The specimen described by Jordan and Fowler (Cat. No. 7722, Stanford Fish Collection) becomes the type of the species, and Funayado on Kyusyu Island the type-locality. Leo Berg, Mem. Acad. Imp. Sci. St. Petersb., (8), XXIV, 1909, 91. ®® Pseudogobio maculatxis Gunther, Ann. Mag. Nat. Hist., (6), I, 1888, pi. 432. — Steindachner; Denksch. Akad. Wiss. Wien, LIX, 1892, p. 370, fig. 4 (figure of oviduct). JORDAN AND HUBBS; JAPANESE FISHES COLLECTED 1922. 177 135. [197] Pseudorasbora parva (Temminck and Schlegel). Moroko = Minnow, Kachi River, at Nagoya (Jordan); Kumamoto (Wakiya); Lake Mikata, Fukiu, Fukuoka, Lake Biwa at Otsu (Jordan and Kawamura). Nuptial males have the pearl-organs restricted to a single series along the sides of the cheeks, and to a series on the lower side of the head, inside the curve of the mouth. Mr. George S. Myers has lately shown that the nominal species, Fundulichthys virescens Schlegel is based on a bad figure of Pseudorasbora parva, Acahara Jordan and Hubbs, gen. nov. Type: Richardsonius semotilus Jordan and Starks. Mori in his “List of Fresh-water Fishes of Korea” makes incidental reference to the as yet unpublished generic name Acahara, Jordan and Hubbs, referring to it Richardsonius semotilus and Richardsonius brandti. As this was the first appear- ance of this generic name in print, R. semotilus may be designated as the type of the genus. Of the Japanese minnows currently referred to Leuciscus two, having large scales, belong in Gnathopogon (ccerulescens and japonicus); three, hakonensis, phalacrocorax, and taczanowskii, together with Phoxinus septentrionalis, comprise a group- which we here name Acahara; while other species jouyi and dorobce, to- gether with Phoxinus steindachneri and Pseudaspius atrilatus, form another natural series, which we call Moroco. Both Acahara and Moroco are also repre- sented by species on the Asiatic mainland. Neither of these groups seem to be congeneric with Cyprinus leuciscus Lin- naeus, which through tautonymy is the type of the genus Leuciscus Cuvier (1817). That genus corresponds to Dobula Rafinesque (1820) to Squalius Bonaparte (1837) and to Cephalus Bonaparte (1845). At present Acahara cannot be sharply dif- ferentiated from Telestes {Telestes multicellus) , another European group, nor from some of the American divisions of Leuciscus, such as Siboma and Tigoma. But while awaiting a critical study of all the leuciscine types, we think it safer and more natural to define and name the group Acahara than to refer it to any of the genera mentioned. The genus Moroco is still more readily separated. The provisional genus Acahara comprises large dace having the following characters: body elongate, subterete; head bluntly subconic, with the mouth slightly overhung by the rostral fold; gape slightly oblique, arched; lower jaw included; lips normal, the lower with a broad frenum; gill-rakers short and slender, 9 to 12 in number on lower limb of outer arch; pharyngeal teeth biserial, with four or five hooked teeth in the outer row, and two teeth in the lesser series ; dorsal 178 MEMOIRS OF THE CARNEGIE MUSEUM. fin inserted over ventral base a little behind middle of length of body; anal fin inserted far behind dorsal; dorsal with seven^ anal with seven or eight branched rays; scales rather small, 60 to 93 along lateral line; each scale oval in outline, with the focus well basad of middle; lateral and apical radii strongly developed; circuli subcontinuous, present on all fields, nearly parallel to scale-margin, and moderately well spaced, as in American minnows in general; intestine short; peri- toneum pale; color uniform, without specialized darkened scales. In breeding males the pearl-organs are thickly scattered over the top of the head and body; of these a few (about one to a scale on the back and a corresponding number on the top of the head) may be somewhat enlarged, and white in color; minute hooks are ranged in single file along the rays on both sides of the dorsal and anal fins, and on the inner side of the paired fins. Three species of Acahara may be recognized in Japanese waters, being distin- guished by the size of the scales. In most of the rivers and adjacent coasts the scales are about seventy-five in number, the extreme known range being from sixty-five to eighty-three, the average varying with the locality from about seventy to about eighty, without definite geographical correlation. The names hakonensis (by error hakuensis), taczanowskii, and septentrionalis belong with this seemingly inseparable complex. In the streams tributary to Tokyo Bay and Lake Kasumi- gaura there is a better defined race, with only 60 to 71 scales (average 66.2) in the row; this form, phalacrocorax, may be retained as a valid species. In Lake Jusan, near Aomori, there is an equally distinct form, which we call jusanensis, known by a single specimen having more than 90 scales along the lateral line. The vernacular name, Akahara (Red-belly), currently applied to the species of this genus, refers to a narrow stripe of bright red running straight from the head to the tail in the adult male. 136. [201 and 202] Acahara hakonensis (Gunther). Ugui = Dace ; Aka-hara = Red-belly. Leuciscus hakuensis Gunther, Challenger Reports, Shore-Fishes, 1880, p. 72, pi. 31, fig. B, (adjudged by the International Commission on Zoological Nomenclature to be a slip from hakonensis, Lake Hakone being misread “Hakoue^’ on the label). Phoxinus septentrionalis Jordan and Seale, Proc. U. S. N. M. XXX, 1906, p. 143, fig. 1. Tokyo market (Jordan); Hamada (Wakiya); Lake Kawaguchi (Wakiya and Ishikawa); Lake Yamanaka (Wakiya and Ishikawa); Nagano (Nakano); Himeji (Abe); Okayama (Mikamo); Akita, Hiki River, Kishu, (Kuroiwa); Lake Kisaki JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922, 179 (Ota); Aomori (Beppu); Toyama (Yoshizawa); Noo (Morioka), Awaya, Lake Togo, Tottori (Inomata); Kamishibi near Fukui, Lake Mikata, Fukui, Lake Suwa (Jordan); Lake Biwa (Jordan); Lake Hakone (Jordan). A very narrow straight bright orange stripe along lower part of body in males ; some orange on lower part of head. The minnows of Japan referred by Jordan and Fowler®^ to Leuciscus hakuensis, L. phalacrocorax, and L. taczanowskii appear to represent a group of local races, between none of which can any trenchant differences be discovered. Their L. taczanowskii is composed of two forms, one with about 80 the other with about 90 scales in the lateral line; the former represents L. taczanowskii proper, while the latter lies without the known limits of variation of the other Japanese forms, and is here regarded as a new species, L. jusanensis. Our material from tributaries of the Japanese Sea shows fewer scales on the average, not more, as currently indicated, than is characteristic of true hakonensis (of which we have topotypic material). We are in fact unable to distinguish nomen- claturally between the majority of the local races. One race, however, that named L. phalacrocorax by Jordan and Fowler, seems well enough differentiated to warrant its retention as a valid species. Scale-variation in Acahara hakonensis. Scales along Lateral Line to Caudal Base. 6s 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 ■ Lake Kawaguchi ..... 1 2 2 1 70 3 Aomori, Morioka, and Otaru 1 2 4 2 1 2 1 1 1 72 3 Lake Yamanaka 1 2 1 2 1 73 4 Drainage of Sea of Japan 1 2 2 1 2 3 2 1 1 1 74 7 Ebisu, Sado Island . . . 1 ■ 2 1 74 5 Lake Biwa 2 4 4 1 74 8 Nagano 1 1 2 1 1 1 75 7 Drainage of Inland Sea 1 1 2 1 76 0 Lake Hakone 1 1 1 78 0 Tokyo Market 1 2 79 0 Totals 1 2 2 2 9 8 6 4 14 12 7 4 4 1 . . 1 74.5 Phoxinus septentrionalis Jordan and Seale, of which we have paratypes at hand, is based on young specimens of Acahara hakonensis. The supposed incom- pleteness of the lateral line is largely due to the accidental loss of scales. Proc. U. S. N. M. ,XXVI, 1903, pp. 844-848. I 180 MEMOIRS OF THE CARNEGIE MUSEUM. 137. [205] Acahara phalacrocorax (Jordan and Fowler). Aka-hara = 'Red-he\\j. Tama River, just below the original type-locality (Jordan); Lake Kasumi- gaura (Hattori) ; Fukui. Males with an orange lateral streak, lower fins bright red. This species is merely the local form of the wide-spread Leuciscus hakonensis, representing that form in the waters tributary to Tokyo Bay and Lake Kasumi- gaura just to the northward. It is distinguished by the larger size of its scales, which number 13 to 16, most usually 14, from the origin of the dorsal fin to the lateral line, and 60 to 71 (average, 66.2) along the lateral line to the caudal base. In Leuciscus hakonensis there are 13 to 18, most frequently 15 scales from the dorsal fin to the lateral line; and 65 to 83 along the lateral line to the caudal fin, the average ranging in different races from about 70 to about 80, and being for the entire series as counted by us, 74.5. Further work, however, will probably show still further intergradation between the two forms, and it may become impossible to recognize L. phalacrocorax as distinct. 138. [206] Acahara jusanensis Jordan and Hubbs, sp. nov. Leuciscus taczanowskii Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 848 (description, and Lake Jusan record only); (not of Steindachner). The type is an adult specimen from Lake Jusan, near Aomori, Province of Mutsu, presented to Dr. Jordan by Director Sotaro Saito of the Aomori Museum in 1900; Cat. No. 7352, Stanford University collection (Car. Mus. Cat. Fishes, No. 7828).' This species is sufficiently well described by Jordan and Fowler. The scales in the lateral line, 93 on one side, 90 on the other, are smaller than in related species.^* According to Dr. Berg Leuciscus taczanowskii is identical with Leuciscus brandti (Dybowsky). This species is recorded by Jordan and Metz from Chin- nampo and Gensan. A very distinct species of Acahara {A. semotilus Jordan and Starks) has been described from Fusan in Korea. In A. hrandti from Lake Chanka, Siberia, the scales are 83, the back blackish. Genus Moroco Jordan and Hubbs, gen. nov. Type: Pseudaspius hergi Jordan and Metz. Mori in his “List of Fresh-water Fishes of Korea” has listed Moroco hergi Jordan and Hubbs without explanation, having evidently taken the name from a labelled specimen in the Stanford University Collection. This publication induces us to make P. hergi the genotype of Moroco, as it is the first species to be published under the new generic name, which we are here proposing. See Berg, Ichth. Amurensis, 1909, pp. 105-108. JOEDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 181 Our reasons for separating this group from Leuciscus appear :n the discussions under the head of the genus Acahara and in the following description. Body approaching the form of a Top-minnow (Gambusia), the anterodorsal and posteroventral outlines being subhorizontal and parallel; snout pointed, some- what overhanging the upper lip; lower jaw included, with its lip rather thick, with a broad frenum; gill-rakers more or less rudimentary, only 3 to 6 on lower limb; pharyngeal teeth biserial, with four or five teeth in the outer row, and two in the lesser row; dorsal fin posteriorly located, its origin being behind the end of ventral base; origin of anal under its last ray; dorsal and anal each with seven branched rays; scales small or minute; individual scales oval in outline, with the focus well basad of middle; the radii strongly developed on all fields of scale; the circuli continuous, well-spaced and parallel with margin of scale; intestine short; peri- toneum dark. Body marked with specialized darkened scales, as in certain American genera (Rhinichthys, etc.). Nuptial males with the upper parts of the head and body covered with minute pearl-organs. Moroko is a vernacular name for small dace of this type. In addition to the type-species we refer three other species to Moroco. They may be distinguished by means of the following key. c. Caudal peduncle excessively deep in the adult, its least depth nearly two-thirds length of head. Scales in lateral line to caudal base fewer than 70. Tsushima jouyi. cc. Caudal peduncle moderately deep in the adult, its least depth about half length of head. Dark scales more conspicuous than the lateral stripe. d. Scales in the lateral line 70 to 80. Eye in adult only half length of snout. Southern Japan. steindachneri. dd. Scales in the lateral line about 100. Korea bergiR ccc. Caudal peduncle slender in the adult, its least depth about one-third length of head. Dark lateral stripe more conspicuous than the darkened scales. Scales in lateral line about 80. Lake Yamanaka yamamotis . 139. [200] Moroco steindachneri (Sauvage). Abura-hai — Fat minnow; Doro5ae = Mud-minnow. Phoxinus steindachneri Sauvage, Bull. Soc. Philom., Paris, 1883, p. 5. — Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 850 (after Sauvage). Hemitremia steindachneri Jordan, Tanaka, and Snyder, Jour. Coll. Sci., Tokyo, XXXIII, 1913, p. 71 (after Sauvage). Pseiidaspius bergi Jordan and Metz, Mem. Car. Mus., VI, 1913, p. 22, pi. Ill, fig. 2 (scale-count apparently too high). Pseudaspius modestus Jordan and Metz, 1. c., p. 23, pi. Ill, fig. 3 (scale-count too low). We have examined the paratypes of these nominal species, and find the scale-counts in essential agreement. 182 MEMOIRS OF THE CARNEGIE MUSEUM. Leuciscus jouyi Jordan and Fowler, 1. c., p. 849 (Kaminutani River, record only); — Tanaka, Annot. Zool. Jap., VII, 1909, p. 134. Not Leuciscus jouyi Jordan and Snyder. Leuciscus dorohoe Ishikawa, Proc. Dept. Nat. Hist., Tokyo Imp. Univ., I, 1904, p. 6, pi. 3, fig. 2. Pseudaspius atrilatus Jordan and Thompson, Mem. Car. Mus., VI, 1914, 231, pi. XXVI, fig. 3 (paratypes seen). Kumamoto, Hamada (Wakiya); Toyama (Yoshizawa); Kinano River, Kishu (Kiiroiwa); Kamishibi, Fukui, Nagano (Nakano); Yamanashi (Imperial Museum 1900). It has not heretofore been suspected that the three names listed in the synonymy really refer to the same fish, but we are quite sure that this is the case. The species is extremely close to Moroco jouyi, which appears to be a local form confined to the island of Tsushima, and differing from both the Japanese M. steindachneri and the Korean species, M. hergi. These forms are contrasted in the key given above. Nuptial males have the upper parts of the head and body covered with minute pearl-organs. 140. [200AJ Moroco yamamotis Jordan and Hubbs, sp. nov. Type 110 mm. long to caudal base, collected by Masashi Ishikawa in Lake Yamanaka on the East side of Fuji-San in Koshu, Car. Mus. Cat. Fishes, No. 7829. Two paratypes, 96 and 63 mm. long, were obtained with the type. This species bears a fairly close resemblance to Leuciscus hakonensis (of which we have specimens taken in the same lake), but it differs trenchantly in having the gill-rakers reduced to three to five fleshy projections, in place of the nine to twelve well developed slender rakers of A. hakonensis; in having the dorsal fin inserted more posteriorly, nearer base of caudal than the middle of eye (rather than the reverse); in the thicker, heavier lips; in the reduction in the size of the scales on the back, there being about 50 instead of about 35 rows before the dorsal fin; in coloration, etc. In most of these respects it agrees with Moroco jouyi and M. steindachneri, of which it may well be the local representative. In general appearance it is quite unlike M. jouyi, the body being slenderer, the caudal peduncle attenuate, rather than greatly deepened, in the adult about one-third, instead of two-thirds, as long as the head. In general the coloration is much paler, the JORDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 183 darkened scales being only indistinctly developed, although the lateral band is more conspicuous; moreover the scales are smaller than in M. jouyi. In the form of the caudal peduncle it is less sharply separated from M. steindachneri, but the same color differences hold; the eye is larger than in M. steindachneri. These differences are given in tabular form in our key. Body formed as in Moroco jouyi, the anterodorsal outline being horizontally flattened, the tail bent downward, but much slenderer and especially attenuate posteriorly. Greatest depth of body, 4.65 (to 4.8) in standard length; least depth of caudal peduncle from anal base to center of caudal base, 3.1 (2.9 to 3.2) in length of head. Top of head flattened, not very convex either transversely or longi- tudinally only gently decurved anteriorly to the edge of the rostral fold, which is well developed, and partially covers and overhangs the upper lip, and lies on level of lower border of pupil; lips full laterally, the lower with a broad median frenum; the mouth slightly curved and somewhat oblique; gape largely lateral; upper jaw as long as snout, extending to below front of eye. Head with membrane, 3.5 in standard length. Eye 1.6 (to 1.0 in young) in snout, 4.8 (to 3.8) in head; snout, 3.1 (to 3.6); fleshy interorbital, 3.6 (to 3.65). Gill-rakers 3 to 5 below angle on outer arch, all soft and very weak. Pharyngeal teeth 4, 2-2, 5, hooked, but without developed grinding surfaces. Peritoneum dark. Intestine short; diet carnivorous. Scales smaller than in other species of Moroco except M. hergi, 21 or 22 from origin of dorsal to lateral line, 78 (77 to 83) in lateral line; 53 from origin of dorsal to occiput ; 8 to 1 1 from lateral line to ventral ; as seen under microscope oval, with focus well basad of middle; circuli rather coarse, well separated, running parallel with margins of scales, radii strongly developed on all fields, including the basal. Lateral line very weakly decurved anteriorly; running a little below middle of depth posteriorly. The dorsal fin, when depressed, not quite half as long (except in young) as distance from its origin to occiput; pectoral contained 1.7 (to 1.5) times in interval between bases of paired fins; ventral fin reaching anus; anal not reaching much more than half-way to caudal. Color pale, becoming darker above, particularly on snout; specialized darkened scales present, but not conspicuous; a dark lateral band follows the axial septum of the body muscles, indistinct anteriorly, where it curves upward well above the lateral line, but blackish posteriorly, where it follows the lateral line; all fins with some pigment, but the anal and ventrals quite pale. The species is named for Dr. Senzi Yamamoto of the Imperial University of Kyoto. 184 MEMOIRS OF THE CARNEGIE MUSEUM. Genus Zacco Jordan and Fowler. The species of this genus have been considerably confused by various authors. We offer a brief synopsis of them, as far as they are known. a. Scales relatively large,' 8 (usually) or 9 from origin of dorsal to lateral line, 39 to 45 along lateral line to caudal base. Mouth of moderate size, the upper jaw much less than half length of head; body marked with vertical dark bars; anal fin of breeding male extended far beyond caudal base. h. Pectoral fins of moderate length, in females and immature males not nearly reaching vertical from ventral fin, in breeding males barely reaching that point; nuptial tubercles on cheek fused at base. Japan and Korea platypus. bb. Pectoral fins elongate, in females and immature males about reaching vertical from ventral fin, in breeding males much longer; nuptial tubercles on cheek separate. Formosa evolans. aa. Scales smaller, 9 (rarely) to 15 from origin of dorsal fin to lateral line, 46 to 62 along lateral line to caudal base. c. Mouth of moderate size, the upper jaw scarcely more than one-third length of head; rostral fold large, nearly concealing the premaxillaries. d. Nuptial tubercles of head smaller and separate; side with a dark longitudinal streak, but without cross-bars. Japan and Korea ternmincki. dd. Nuptial tubercles on side of snout and on cheek hugely developed, and arising in each case from a broad horny common base; male at least with vertical cross-bars. China . acanthogenys.^° cc. Mouth very large, the upper jaw almost half as long as the head; rostral fold not expanded, and not concealing the premaxillaries; nuptial tubercles of head separate; vertical cross-bars often developed. Formosa pachycephalus. 141. [211] Zacco platypus (Temminck and Schlegel). Hmja ^Minnow. Leuciscus platypus (Temminck and Schlegel) Richardson, Kept. Brit. Assoc. Adv. Sci., for 1845 (1846), p. 300 (on proof-sheets of Schlegel’s later account. — Temminck and Schlegel, Fauna Japonica, Pisces, 1846, p. 210, pi. 101, fig. 3. (Nagasaki). Opsariichthys platypus Gunther, Cat. Fishes Brit. Mus., VII, 1868, p. 296 (Formosa records excepted). — Sauvage, Bull. Soc. Philom., 1883, p. 5. Barilius platypus Bleeker, Verh. Akad. Amsterdam, XVIII, 1879, p. 23. — Jordan and Snyder, Proc. U. S. N. M., XXIII, 1901, p. 344; Annot. Zool. Jap., Ill, 1901, p. 47. — Regan, Proc. Zool. Soc. London, 1908, p. 59 (Korean record). Zacco platypus Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 851. — Smith and Pope, ibid., XXXI, 1906, p. 462. — Tanaka, Zool. Mag., No. 237, 1908, p. 235; Annot. Zool. Jap., VII, 1908, p. 7, ibid., VII, 1909, p. 133. — Cockerell, Zool. Anz., XXXVIII, 1911, p.85. — Tanaka, Fig. Desc. Fishes Jap., IV, 1911, pi. 200, figs. 72-74; ibid., V, 1912, p. 83. — Snyder, Proc. U. S. N. M., XLII, 1912, p. 404. — Jordan, Tanaka, and Snyder, Jour. Coll. Sci., Tokyo, XXXIII, 1913, p. 75. — Jordan and Thompson, Mem. Car. Mus., VI, 1914, p. 232. — OsHiMA, Ann. Car. Mus., X, 1919, p. 237 (Japanese specimens only). Opsariichthys acanthogenys Boulenger, Proc. Zool. Soc. London, 1901, p. 269, pi. 24, fig. 1. The University of Michigan has material of this species from Foo-chow, China. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 185 Leuciscus minor (Temminck and Schlegel) Richardson, Kept. Brit. Assoc. Ad. Sci., 1845 (1846), p. 300. — Temminck and Schlegel, Fauna Japonica, Pisces, 1846, p. 210, pi. 101, fig. 3. Barilius minor Bleeker, Verb. Akad. Amsterdam, (2) III, 1867, p. 248; Verb. Akad. Amsterdam, XVIII, 1879, p. 23. Leuciscus macropus Temminck and Schlegel, Fauna Japonica, Pisces, 1846, p. 209, pi. 101, fig. 2. Barilius macropus Bleeker, Verb. Akad. Amsterdam, XVIII, 1879, p. 23. Lake Biwa at Otsu (Jordan and Kawamura); Kumamoto (Wakiya); Himeji (Abe); Okayama (Mikamo); Lake Suwa (Jordan); Fukuoka (Hamada); Lake Mikata, Fukui, Kachi River at Nagoya (Jordan); Lake Kasumigaura (Hattori). Generally common southwards. Barilius acutipinnis Bleeker, from the Yang-tse-Kiang of China, is perhaps identical with Zacco platypus. The relation of Zacco to Barilius is in need of definition. Zacco platypus has erroneously been identified with a Formosan species, Z. evolans, by Oshima, and with a Chinese species, Opsariichthys bidens, by Boulenger. The pharyngeal teeth may be 4 or 5 in the outer row. The scales do not vary widely in number, there being usually 8, occasionally 9 from the origin of the dorsal to the lateral line, and from 39 to 45, usually 41 to 43, along the lateral line to the caudal base. In fully developed males, the anal fin is extended (by the growth of the rays, not by adipose extentions, as has been suggested) far beyond the caudal base, (much farther than in Z. temmincki)] the nuptial tubercles on the cheeks are united basally to form a plate approaching that of Z. acanthogenys (a condition not noted in Z. temmincki), but the pearl-organs on the lower side of the caudal peduncle are very small, several to a scale (in Z. temmincki these organs are enlarged and only one is developed on each scale). 142. [Extraterr.] Zacco evolans Jordan and Evermann. Opsariichthys platypus Gunther, Cat. Fishes Brit. Mus., VII, 1868, p. 296 (For- mosan record only). Zacco platypus Oshima, Ann. Car. Mus., X, 1919, p. 235 (most of synonymy and note on “cotype” of Z. evolans excepted); Proc. Acad. Nat. Sci. Phila., 1920, p. 130 (not of Richardson). Zacco evolans Jordan and Evermann, Proc. U. S. N. M., XXV, 1902, p. 322, fig. 5 (type, but not “cotype”). — Jordan and Richardson, Mem. Car. Mus., IV, 1909, p. 170, fig. 6. 186 MEMOIRS OF THE CARNEGIE MUSEUM. Zacco temmincki Oshima, Ann. Car. Mus., X, 1919, p. 240 (specimens from Daito River only). We have examined the type of this Formosan species (the paratype is referable rather to Zacco pachycephalus) , part of the material called Z. platypus by Oshima, and two specimens from Daito River, wrongly referred by Oshima to Zacco temmincki. 143. [212] Zacco temmincki (Temminck and Schlegel). = River-mutsu. Leuciscus temmincki (Temminck and Schlegel) Richardson, Rept. Brit. Assoc. Adv. Sci. for 1845 (1846), p. 300 (scales erroneously counted, but identifica- tion fixed by note on color; account prior to that by Schlegel, who is quoted as unpublished). — Temminck and Schlegel, Fauna Japonica, Pisces, 1846, p. 210, pi. 101, fig. 4. Opsariichthys temminckii Gunther, Cat. Fishes Brit. Mus., VII, 1868, p. 295. — Matsubara, Jap. Intern. Fisch.-Ausst., Berlin, 1880, p. 17. — Namiye, Class. Cat. Spec. Vert., 1881, p. 107. — Sauvage, Bull. Soc. Philom., 1883, p. 5. — IsHiKAWA, Zool. Mag., VII, 1895, p. 121. — Ishikawa and Matsumura, Prel. Cat., 1897, p. 11. Barilius temmincki Bleeker, Verb. Akad. Amsterdam, XVIII, 1879, p. 23. — Jordan and Snyder, Annot. Zool. Jap., Ill, 1901, p. 47. Zacco temmincki Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 852. — Tanaka, Zool. Mag., No. 237, 1908, p. 235. — Snyder, Proc. U. S. N. M., XLII, 1912, p. 404. — (?) Jordan and Metz, Mem. Car. Mus., VI, 1913, p. 21 (Korean record). — Jordan and Thompson, ibid., VI, 1914, p. 232. — Oshima, Ann. Car, Mus., X, 1919, p. 238 (most of synonymy only). — Tanaka, Annot. Zool. Jap., VII, 1909, p. 134. — Jordan, Tanaka, and Snyder, Jour. Coll. Sci. Tokyo, XXXIII, 1913, p. 75. — Tanaka, Fig. Desc. Fishes Jap., XVII, 1914, pi. 81, fig. 275; ibid., XVIII, 1914, p. 296. Leuciscus sieboldi Temminck and Schlegel, Fauna Japonica, Pisces, 1846, p. 211, pi. 101, fig. 5. Opsariichthys sieboldi Gunther, Cat. Fishes Brit. Mus., VII, 1868, p. 295. — - Sauvage, Bull. Soc. Philom., 1883, p. 5. Barilius sieboldi Bleeker, Verh. Akad. Amsterdam, XVIII, 1879, p. 23. — Jordan and Snyder, Annot. Zool. Jap., Ill, 1901, p. 47. Zacco sieboldi Jordan and Fowler, Proc. U. S. N. M., XXVI, 1903, p. 854. — Tanaka, Annot. Zool. Jap., VII, 1908, p. 6; Zool. Mag., No. 237, 1908, p. 235. — Jordan and Thompson, Mem. Car. Mus., VI, 1914, p. 232. JORDAN AND HUBBS: JAPANESE PISHES COLLECTED 1922. 187 Zacco mitsukurii Ishikawa, Proc. Dept. Nat. Hist., Tokyo Univ., I, 1904, p. 4, pi. 1, fig. 1. Zacco mitsukurii, var. a, Ishikawa, ibid., p. 5, pL 2, fig. 2. Hamada, Kumamoto, Ozu, Island of Shikoku, Lake Biwa (Wakiya); Mikawa Bay (Ishikawa); Himeji (Abe); Okayama (Mikamo); Fukuoka (Hamada); Yamaguchi. The number of scale-rows in this species varies widely, but we are unable to find evidence indicating the existence of two species, Z. temmincki and Z. sieboldi, differing in the number of scales, and ranging side by side through southwestern Japan. The variations seem rather to be of a purely local character. Our counts may be tabulated as follows: ZACCO TEMMINCKI. Number of Transverse Rows of Scales Localities 46 47 48 49 50 SI 52 53 54 55 56 57 58 59 60 Averages Shikoku Island, Drainage of Inland Sea 1 46 Mikawa Bay (East of Inland Sea) . Hondo Drainage of Inland Sea. . . . Hamada, Drainage of Sea of Japan Kyusyu (South of Inland Sea) .... 1 48 1 2 7 1 2 3 3 1 1 3 52 1 2 4 2 1 52 1 1 2 1 54 Lake Biwa {mitsukurii) 1 2 4 3 6 3 1 57 Scales from Dorsal Fin to Lateral Line Localities 9 10 II 12 13 14 IS Averages Shikoku Island, Drainage of Inland Sea 1 11.0 Mikawa Bay (East of Inland Sea) 1 11.0 Hondo, (Drainage of Inland Sea) 1 12 4 2 3 3 11.1 Hamada (Drainage of Sea of Japan) 8 4 11.3 Kyusyu, (South of Inland Sea) 1 1 3 13.4 Lake Biwa {mitsukurii) 5 12 3 13.9 The variations in the number of scales seem to show no very definite geo- graphical correlation. The most noteworthy feature is the high average number in the specimens from Lake Biwa. 144. [Extraterr.j Zacco pachycephalus (Gunther). Opsariichthys pachycephalus Gunther, Cat. Fishes Brit. Mus., VH, 1868, p. 296 (Formosa). 188 MEMOIRS OF THE CARNEGIE MUSEUM. Zacco pachycephalus Jordan and Evermann, Proc. U. S. N. M., XXV, 1902, p. 322. — Jordan and Richardson, Mem. Car. Mus., IV, 1909, p. 170. — Cockerell, Zool. Anz., XXXVIII, 1911, p. 87. — Oshima, Ann. Car. Mus., XII, 1919, p. 240; Proc. Acad. Nat. Sci., 1920, p. 130. Zacco platycephalus Fowler and Bean, Proc. U. S. N. M., LXIV, 1922, Art. 2, p. 7 (lapsus for pachycephalus). Zacco evolans Jordan and Evermann, Proc. U. S. N. M., XXV, 1902, p. 322 (“cotype,” not type). Zacco platypus Oshima, Ann. Car. Mus., XII, 1919, p. 236 (note on “cotype” of Z. evolans only). Zacco temmincki Oshima, 1. c., 1919, p. 238 (most of synonymy and specimens from Daito River excepted). — Fowler and Bean, Proc. U. S. N. M., LXIV, 1922, Art. 2, p. 6. Zacco temminckii Oshima, Proc. Acad. Nat. Sci. Phila., 1920, p. 130. (Not Leuciscus temmincki Richardson) . Oshima and Fowler and Bean have both referred the finer-scaled Formosan species of Zacco to two species, pachycephalus and temmincki. But temmincki is a trenchantly different form, as the key we have prepared indicates. We have examined most of the material discussed by Oshima, but fail to find more than the one species represented. One of Oshima’s lots, that from the Daito River, should have been referred to Zacco evolans. The “cotype” of Zacco evolans is a male specimen of Z. pachycephalus. 145. [213] Opsariichthys uncirostris (Temminck and Schlegel). Hasu. Lake Biwa, at Otsu (Jordan and Kawamura); Hachi, Fukui. It is abundant in Lake Biwa, reaching a length of a foot or more, and is much valued as food, its flesh being rich and delicate. Nuptial males have coarse, broadly conic pearl-organs thickly set on the mandibles, the suborbital region, and the preopercle, occurring also at the tip of the lower jaw, and even on the lower and upper lips laterally, also between the nostrils, and on the cheeks, and interopercle. Other pearl-organs are scattered over the sides of the tail, usually several on a scale; they are strongest near the anal fin, becoming smaller dorsally and obsolete toward the back and toward the caudal fin; on the lower surface of the caudal, as also along the anal rays, they are in contrast much strengthened, and only one is located on each scale. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922, 189 146. [Extraterr.] Opsariichthys bidens Gunther. Opsariichthys bidens Gunther, Ann. Mag. Nat. Hist., (4) XII., 1873, p. 249. Two specimens of this species were collected by Dr. Wakiya in the Ping-yang River, Korea. The Opsariichthys of the Asiatic mainland (China, Korea, and the Amur drainage) differs from the Japanese species in having somewhat larger scales: 44 to 48 along the lateral line, rather than 47 to 53; 9 from the dorsal fin to the lateral line, not 10 to 12; 3.5 rather than 4.5 or 5 from the lateral line to the anal origin. In our two specimens the scales are 9-45 or 46-3.5. Berg, who gives a good synonymy in his Ichthyologia Amurensis (1909), fails to separate these two species. To Berg’s synonymy for bidens should be added: Opsariichthys platypus Boulenger, Proc. Zool. Soc. London, 1901, Part I, p. 24, pi. 24, fig. 2 (not of Richardson) . 147. [176] Ishikauia steenackeri (Sauvage). W adaka ; Unia-uwo = Ilorse-iish. Lake Biwa, at Otsu (Jordan and Kawamura). It is abundant in the lake, reaching a length of nearly two feet and is a food-fish of importance. The flesh is, however, far inferior to that of the excellent Hasu, with which it is associated. Genus Hemigrammocypris Fowler. {Brevigobio Tanaka.) “No barbels about mouth; pharyngeal teeth 3-rowed, 4, 4, 2-2, 4, 4; dorsal fin inserted nearer to base of caudal than to tip of snout; a sharp ventral keel between anus and origin of ventral; lateral line decurved, incomplete, running along lower part of body, ending near last ray of anal” (Tanaka). This genus in many ways is similar to Rasbora, a genus comprising many species from southern Asia and the East Indies, but it differs from that genus in having the abdomen sharply keeled, and in the normal structure of the lower jaw. 148. [176A] Hemigrammocypris rasborella Fowler. Hemigrammocypris rasborella Fowler, Proc. Acad. Nat. Sci. Phila., LXII, 1910, 483, Lake Biwa. Brevigobio kawabatce Tanaka, Zool. Mag., XXVIII, 1916, p. 102 (Lake Biwa); Fig. Desc. Fishes Japan, XXIV, 1916, p. 420, pi. 115, figs. 339, 340 (Lake Biwa; pond near Tsu in Ise). Kachi River at Nagoya (Jordan) ; Lake Biwa at Otsu (Jordan and Kawamura) ; Lake Biwa (Jordan). 190 MEMOIRS OF THE CARNEGIE MUSEUM. Our specimens from the Kachi River are pale in color (like other fishes from the same place), almost lacking the longitudinal band. In this species the fine black streak which follows the axial septum is characteristically interrupted to form a series of dashes. 149. [214] Cyprinus carpio Linnaeus. Koi. Lake Kawaguchi, Mikawa Bay (Ishikawa); Ozu, Shikoku Id. (Wakiya); Nagano (Nakano) ; Himeji (Abe) ; Lakes Hakone and Suwa (Jordan) ; Lake Kasumi- gaura (Hattori). The specimen from Lake Hakone is strikingly blotched with orange and black, as are manyKof domesticated in ponds. 150. [215] Carassius auratus (Linnaeus). Funa;Hiwara = 'Red-he\ly. Very abundant as- a native fish, the Funa is domesticated everywhere and greatly modified in ponds, where it receives various names. Lake Kawaguchi; Lake Yamanaka; Mikawa Bay; Kachi River; Kumamoto; Nagano; Himeji; Okayama; Akita, Sakurai; Yamaguchi; Toyama; Lake Hakone; Lake Suwa; Fukuoka; Noo, Morioka; Lake Kozan; Lake Togo (Inomata); Lake Kasumigaura (Hattori). We also have material from Soo-chow, China (Gee). Most of the specimens are of the ordinary “wild type” of Gold-fish, only a few of these obtained showing the increased depth, elongated or fantastic fins, or orange color characteristic of many of the “domestic races.” The dorsal rays are fewer than indicated in most descriptions, being H, 12 to II, 16 in the specimens counted. Family FLUTIDiE. Genus Fluta Bloch and Schneider. The name Fluta Bloch and Schneider (1801) is prior to Monopterus (Lacepede), which until 1806 appeared only as “les Monopteres.” Monopteros Volta (1796) is a genus of fossil fishes {= Platinx Agassiz) and is prior to Monopterus Lacepede. 151. [216] Fluta alba (Zuieuw). T'awna^f = Rice-field Eel. Ponds near Kyoto (Kawamura); Formosa (Kawamura). Family ANGUILLIDiE. 152. [217] Anguilla japonica Temminck and Schlegel. Unagi = Eel; Ounagi = Great Eel. Mikawa Bay (Ishikawa); Himeji (Abe); Lake Suwa (Ota); Lakes Suwa and Hakone (Jordan); Fukuoka (Hamada); Lake Togo (Inomata); Lakes Mikata JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 191 and Kasumigaura (Hattori). We also have a specimen from Soo-chow, China. Everywhere abundant and highly valued as food. One of the specimens from Lake Kasumigaura belongs to the black-speckled type, which Jordan and Snyder and Ishikawa have regarded as a variant of the common species of Japanese eel. Family SYNAPHOBRANCHID.E. 153. [220] Synaphobranchus affinis Gunther. /Jora-ana(/o = Cave-eel. Synaphobranchus taketce Tanaka, Zool. Mag., XXVIII, 1917, p. 257 (text in Japanese). Kushiro (Tanaka); Shizuoka (Jordan); Misaki (Aoki). The position of the dorsal origin varies considerably, in our material being- over, or rarely (as in the type of S. taketce) a little before the anus, or as much as one-fourth length of head behind vertical from anus. 154. [222] Synaphobranchus jenkinsi Jordan and Snyder. One of the eels taken by Aoki at Misaki agrees with the description of the type of this species in all respects, except that the eye is contained 2.5 times in the snout, the pectoral fin two times in the head. Family CONGRIDAil. In previous descriptions of the Congers of Japan little or no attention has been paid to the dentition or the structure of the snout. A more careful examina- tion of the several species has brought out characters apparently of generic signifi- cance, as the following key will show. Key to the Genera of Congers of Japan and Formosa. a. Premaxillary teeth entirely within the closed mouth (rarely slightly exposed in Anago) ; teeth hiserial on jaws, those of the inner series very much the smaller. Tail blunt. Snout not notably produced beyond the mouth. Anterior nostrils low. Teeth all small, scarcely canine-like. b. Teeth of jaws conic, not in contact basally, and not forming a common cutting edge. Mouth small, the gape reaching only to below middle of eye. Tip of snout smooth. Dorsal fin beginning over, or very slightly behind, pectoral base Anago. bb. Teeth of jaws incisor-like, compressed, and in contact basally, forming a common cutting edge (the tips more or less separate in Congriscus). Mouth larger, the gape subtending all or most of eye. Tip of snout more or less distinctly tricarinate. Dorsal fin beginning well behind i^ectoral base, c. Teeth of jaws less truncate, with the tips largely free, not forming a very well defined cutting edge. Dorsal fin beginning a little in advance of middle of pectoral Congriscus. cc. Teeth of jaws evenly and abruptly truncate, the tips forming a well defined common cutting edge. Dorsal fin beginning over or behind middle of pectoral. d. Pores not surrounded by pigmented free areas, confined on bodj^ to lateral line and sparsely developed on head Conger. 192 MEMOIRS OF THE CARNEGIE MUSEUM. (Id. Pores surrounded by conspicuous pigmented free areas, forming a series below the dorsal fin, as well as along the lateral line, and densely developed on top of head anteriorly. . Astroconger. aa. Premaxillary teeth largely or entirely in front of the mouth on the lower surface of the projecting snout. Teeth in jaws in a band or in two series of similar size. Tail more attenuate. e. Anterior nostrils on lower surface of snout, beside premaxillary band of teeth, just within anterior end of upper lip, far below the anterior rostral pits. Upper lip separated from maxillary teeth by a wide flat ridge. Teeth all small, none canine-like, in a patch on the vomer; premaxillary teeth not entirely in advance of mouth, the posterior edge being covered when the mouth is closed. Gape short, extending only to below middle of eye. Gill-openings directed downward and backward. f. Snout short, barely projecting beyond premaxillary teeth (and without pocket or keel on midline). Teeth fewer and larger, those on jaws mostly in rows, those on vomer bluntly conic, forming an elongate-triangular band (which separates the maxillary rows). Alloconger. //. Snout long, its fleshy tip projecting sharply beyond the premaxillary teeth. Teeth fine and close-set, those of jaws forming narrow bands, those on the vomer largely molar- like, forming a broad patch. g. Premaxillary patch of teeth much smaller than the vomerine patch, and separated from it by the widely confluent anterior ends of the maxillary bands. Anteroventral line of snout occupied by a deep pocket. No enlarged pores between the nostrils. Posterior nostril a horizontal slit, with entire rim Rhynchocymba. gg. Premaxillary patch of teeth larger than the vomerine patch, and in full contact with it, the anterior ends of the maxillary bands thus being widely separated. Antero- ventral line of snout occupied by a fleshy keel ending posteriorly in a small free process. A pair of enlarged pores between the nostrils. Posterior nostril widely open, with fimbriate border Rhynchoconger. ee. Anterior nostrils on lateral face of snout, well above premaxillary teeth, far in advance of, and above, upper lip just behind the anterior rostral pits. Upper lip separated from maxillary teeth by a ridge, which is not flattened. Teeth largely canines, none molar- like; those on vomer not forming a patch; the premaxillary band entirely in advance of mouth, its posterior edge composed of large canines shutting outside the lower jaw. Gape relatively wide, extending almost to below hinder border of eye. Gill-openings directed downward and forward. A pair of enlarged pores just before premaxillary teeth. Posterior nostrils more or less slit-like, with entire rims. h. Premaxillary patch of teeth not separated from the vomerine teeth by the maxillary series, which are widely separated anteriorly; teeth on sides of jaws in two even rows; vomerine teeth numerous, in a very long even file behind the canine. Ridge between maxillary teeth and upper lip with entire edge. Snout shorter, its fleshy tip scarcely projecting beyond the premaxillary teeth, and without ridge or keel on mid-ventral line Uroconger. hh. Premaxillary patch of teeth separated from the vomerine teeth by the widely con- fluent maxillary bands; teeth on sides of jaws in narrow bands; vomerine teeth rather few and arranged as an A-shaped figure behind the large canine. Ridge between maxillary teeth and upper lip with the border finely fimbriate. Snout long, its fleshy tip projecting sharply beyond the premaxillary teeth, and with a ventro-anterior keel Congrina. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 193 Anago Jordan and Hubbs, gen. nov. Type: Conger anago Temminck and Schlegel. This genus is very close to Conger, with which it is connected through Con- griscus, but the teeth, though similarly disposed, are very different in shape, and the mouth is smaller, the dorsal fin inserted farther forward. From Ariosoma (balearica) and Alloconger {flavirostris, etc.), it differs in the arrangement of the teeth, and from Gnathophis in having all the teeth sharp. 155. [233] Anago anago (Temminck and Schlegel). Anago = Conger. Misaki (Aoki); Mikawa Bay (M. Ishikawa); Toba (Jordan and Yamamoto); Tokyo market (Jordan). Generally common. The three characteristic dark dashes on the side of the head are evident in a young specimen 15 cm. long, but not in one 12 cm. long. CoNGRiscus Jordan and Hubbs. gen. nov. Type: Congromurcena megastoma Gunther. This genus, which we define and compare with related types in the key, seems to stand directly between Anago and Conger in its technical characters. 156. [234] Congriscus megastomus (Gunther). 0/cf-ana^o = Off-shore Conger. Misaki. Genus Conger (Cuvier) Oken. {Leptocephalus (Gronow) Scopoli; name assigned to larval forms.) Two species of the Indo-Asiatic fauna may be retained in the typical genus Conger. One of these, C. japonicus^ is interpreted by us as the representative of the Atlantic species, C. conger, while the other, C. cinereus, differs considerably in the point of origin of the dorsal fin. 157. Conger cinereus Rlippell. Conger cinereus Weber and de Beaufort, Fishes Indo-Austral. Arch., Ill, 1916, p. 258, figs. 107, 108 (with entire synonymy excepting Leptocephalus nystromi Jordan and Snyder). Leptocephalus riukiuanus Jordan and Snyder, Proc. U. S. N. M., XXIII, 1901, p. 852, fig. 4. (Okinawa). This species, which has the dorsal fin inserted farther forward than in Conger japonicus or Conger conger, has been recorded from Japan proper by Gunther and by Nystrom as Conger marginatus, and a young specimen has been described as 194 MEMOIRS OF THE CARNEGIE MUSEUM. Leptocephalus riukiuanus, from the Riu Kin Islands. A comparison of this speci- men with a large one from Samoa (called marginatus) reveals no evident differences. 158. [224, 228, and 231] Conger japonicus Bleeker. Hama-anago = 'Beach-conger. Conger vulgaris Temminck and Schlegel, Fauna Jap., Pisces, 1846, p. 259. — Bleeker, Atl. Ichth., IV, 1864, p. 26, pi. 149, fig. 2 — Gunther, Cat. Fishes Brit. Mus., VIII, 1870, p. 70 (in part). (Not of European authors.) Conger japonicus Bleeker, Verh. Akad. Amsterdam, XVIII, 1879, p. 32, pi. 2, fig. 2. Leptocephalus japonicus Jordan and Snyder, Proc. U. S. N. M., XXIII, 1901, p. 851. Leptocephalus erehennus Jordan and Snyder, ibid., p._849, fig. 3. (Misaki). Leptocephalus kiusiuanus Jordan and Snyder, ibid., p. 851. Nagasaki. — Snyder, Proc. U. S. N. M., XLII, 1912, p. 406. Conger conger Weber and de Beaufort, Fishes Indo-Austral. Arch., Ill, 1916, p. 259. Three specimens from Misaki (Aoki). Conger japonicus Bleeker is doubtless based on the young of the conger later described as L. erebennus and L. kiusiuanus by Jordan and Snyder. On re- examining the types of the two nominal species we fail to find any wide difference in the length of the trunk. This is apparently also the species referred by Schlegel, Bleeker, and Weber to the Atlantic Conger conger. Indeed on comparison of material we find no constant differences in proportions, coloration, or dentition. There appears, however, to be a difference in the number of pores in the lateral line on the trunk (and the pores correspond in number to the body segments) . Between the verticals from the origin of the pectoral and the anus we count 33 to 35 pores in six Japanese specimens: 36 to 37 in one from Beaufort, North Carolina, and 40 in another from the same locality; 39 to 41 in two from the Canary Islands, and 39 to 40 in three from Naples. Astroconger Jordan and Hubbs, gen .nov. Type: Anguilla myriaster Brevoort. This common Japanese conger, known currently as Leptocephalus myriaster, must be taken as the type of a genus distinct from Conger, on account of the very extensive development of sensory pores. The series of pores along the back is, so far as we know, not developed in any other eel. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 195 159. [227] Astroconger myriaster (Brevoort). ilfa-ana{/o = TiTie-conger. Misaki (Aoki); Fukuoka (Hamada); Mikawa Bay (Ishikawa); Tokyo, Yoko- hama, and Osaka markets (Jordan); Fukui (Nonaka). Generally common every- where in the markets. Alloconger Jordan and Hubbs, gen. nov. Type: Leptocephalus flavirostris Snyder. This genus seems similar to Ariosoma Swainson®^, but it at least differs in having the premaxillary teeth largely exposed. It is compared with the other Japanese genera of conger-eels in the preceding key, and the remarks there given will serve as its definition. 160. [230] Alloconger flavirostris (Snyder). Leptocephalus flavirostris Snyder, Proc. U. S. N. M., XXXV, 1908, p. 93; ibid., XLII, 1912, p. 405, pi. 51, fig. 1 (Misaki). Alloconger flavirostris is very close to the East Indian species, A. anagoides (Bleeker), of which we have examined the Formosan material recorded by Jordan and Richardson in 1909®^ and to the Hawaiian species, A. howersi (Jenkins), which we also have at hand. It seems to differ from both species in having the teeth smaller and more numerous. In the paratype of A. flavirostris there are about fifty pores before the anus. Rhynchocymba Jordan and Hubbs, gen. nov. Type: Leptocephalus nystromi Jordan and Snyder. The definition and comparisons of Rhynchocijmba may be found in our generic analysis given above. 161. [226] Rhynchocymba nystromi (Jordan and Snyder). In working out the generic diagnosis of Rhynchocymba we have used the typical material of this species. In describing Leptocephalus nystromi Jordan and Snyder compared it only with the very distantly related Leptocephalus marginatus = Conger cinereus, and referred Gunther’s and Nystrom’s Japanese records of The group of conger-eels typified by Murosna halearica De la Roche of the Mediterranean Sea has successively received the names Ariosoma Swainson, Ophisoma Swainson, Conger murcena Kaup, and Congrellus Ogilby. The first restriction of Ophisoma by Bleeker, 1864, to 0. acuta Swainson ( = balearica) carries its synonym Ariosoma with it. To the Congers with blunt teeth of the type of Conger hahenata, the name Gnathophis Kaup, based on Myrophis heterognathus Bleeker from Nagasaki is apparently applicable. Mem. Car. Mus., IV, 1919, p. 171 62 196 MEMOIRS OF THE CARNEGIE MUSEUM. Conger rnarginatus to their new species, making no reference to the dentition. This circumstance has led Weber and de' Beaufort (1916) to refer nystromi improperly to the synonymy of Conger cinereus. The species is doubtless different from Myrophis heterognathos Bleeker, type of Gnathophis Bleeker, whatever the latter may be, though the last version gives it the blunt teeth of Conger habenatus. Rhynchoconger Jordan and Hubbs, gen. nov. Type: Leptocephalus ectenurus Jordan and R. E. Richardson. This genus with Rhynchocymba and Congrina have many characters in common with Congerniurcena nasica Alcock, the type of Bathycongrus Ogilby, but all seem to differ generically. Unfortunately we have no specimens of Alcock’s species. 162. [Extraterr.] Rhynchoconger ectenurus (Jordan and Richardson). Leptocephalus ectenurus Jordan and R. E. Richardson, Mem. Car. Mus., IV, 1909, p. 171, p. LVI, lower figure. We have studied the type of this Formosan species. Genus Uroconger Kaiip. This genus has long been separated from Conger or Leptocephalus, but it is no more distinct than most of the genera we have just characterized. 163. [226A] Uroconger lepturus (Richardson). Leptocephalus retrotinctus Snyder, Proc. U. S. N. M., XLII, 1912, p. 405 (Kago- shima). (Not of Jordan and Snyder.) Failure to consider the dentition has led Snyder to refer specimens of Uroconger lepturus, a well known Chinese and East Indian eel from Kagoshima, to Leptoce- phalus retrotinctus Jordan and Snyder {^Congrina retrotinctd) . We have re- examined this material. One specimen has the tail unusually short (only 1.4 times the head and trunk), and the caudal fin is extremely large and base broad. These facts would seem to indicate that the extremely attenuate tail normally developed in this species is subject to injury, but that after loss it regenerates a pseudo-caudal fin, as in the macrouroid fishes. Congrina Jordan and Hubbs, gen. nov. Type; Congerniurcena cequorea Gilbert and Cramer. This genus is defined in our generic analysis given above. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 197 164. [229] Congrina retrotincta (Jordan and Snyder). Leptocephalus retrotinctus Jordan and Snyder, Proc. U. S. N. M., XXIII, 1901, p. 853, fig. 6 (Tokyo Market). A specimen 461 mm. long to caudal was taken by Aoki at Misaki. Snyder’s record of Leptocephalus retrotinctus from Kagoshima refers to Uro- conger Upturns . Congrina retrotincta is extremely close to Congrina cequorea, the type-species of the new genus Congrina. On comparison it seems to differ in having the head shorter and the teeth perhaps a little stronger and fewer. In all the characters given in the key the two species are alike. The type of L. retrotinctus was described as having the snout blunt, which is true of that specimen, but the bluntness is due, we now find, to the fact that the protruding fleshy tip had been broken off or torn away. As the type was not only injured, but young, we have prepared the following description of our adult specimen: Body fairly robust, about as wide as deep anteriorly, but posteriorly becoming compressed, tapering to a very slender tail. Postorbital region tumid. The bluntly conic snout projects well beyond mouth, its preoral length being nearly equal to orbital length. Greatest depth of body 2.45 in head, 17.7 in total length to caudal. Head, 7.75 in total length, 1.7 in trunk, 4.4 in tail. Distance from dorsal fin to occiput 2.8 in head; highest dorsal ray, 3.15; length of caudal, 7.4; length of pectoral, 3.65; snout, 3.55; eye, 6.65; interorbital width (fleshy), 6.15; gape, 3.2; width of head, 2.7. Head and trunk together 1.85 in tail. Origin of dorsal fin a little behind base of pectoral; pectoral fin rounded. Teeth of jaws coarse and irregular, in narrow bands; vomerine teeth few, one greatly enlarged, conic, sharp, somewhat curved backward, preceded by two in line, much smaller, the first scarcely canine-like; followed by a group of three teeth forming an isosceles triangle with the apex pointing backward; premaxillary teeth about thirteen in number, enlarged, canine-like posteriorly along front of gape. Tip of snout with a large deep pit on each side just in front of the anterior nostrils, which open in a short tube; a series of five longitudinal slits runs from just behind the anterior nostril to below middle of eye; the posterior nostril is a horizontal slit with scarcely elevated rims, located just before and a little above the horizontal through middle of pupil. Other pores occur near tip of chin and a pair on the lower surface of the snout just in front of the premaxillary cluster of teeth and on each side of a fleshy keel; lips rather full laterally; the upper narrow, separated by a fringed fold from maxillary band of teeth ; lower lip thicker and more pendant ; gill-openings extending 198 MEMOIRS OF THE CARNEGIE MUSEUM. downward and forward from the middle of the pectoral base to within their own length apart. Lateral line a rather wide ridge, originating abruptly at side of nape, running a little above midline of sides on anterior half of body, a little below the mid-line posteriorly; seventh pore above pectoral base; eighth and ninth on each side of the vertical from dorsal origin, thirty-ninth above the anus. Color dusky above, rather abruptly pale below; pectoral fin whitish, with a dusky blotch above its center; vertical fins pale anteriorly with a dusky base, which soon widens, especially on the dorsal fin to exclude the pale color from all but the margin. Family MUR^NESOCID^. 165. [235] Muraenesox cinereus (Forskal). Hamo. Tokyo, Kyoto, and Kobe markets (Jordan). Not rare in the markets and reaching a considerable size. Outer mandibular teeth not directed outward; vomerine teeth broad, with strong basal lobes. Family OPHICHTHYIDJE. 166. [243] Pisoodonophis zophistius Jordan and Snyder. One adult and two young, Mikawa Bay (Ishikawa); two, Misaki (Aoki). On comparison of material we find that P. zophistius differs from the East Indian P. cancrivorus in having the dorsal much higher and more darkly colored anteriorly, in the much smaller size of the two barbels on the upper jaw, and in the smaller eye, which is two-fifths instead of half as long as the snout. The pores of the head, however, are alike in the two species. 167. [254] Ophisurus macrorhynchus Bleeker. = Off-shore Snake. Mikawa Bay (Ishikawa). Family MURtENIDtE. 168. [258] Gymnothorax reticularis Bloch. t/ien6o = Moray. Misaki (Aoki). None of the other Morays, numerous about Kytisyu, were obtained in 1922. Family CYPRINODONTID^. Genus Oryzias Jordan and Snyder. Regan®^ and Weber have identified Oryzias Jordan and Snyder with Aplocheilus (usually, but not at first, written H aplochilus) . Gunther, in referring the types of Aplocheilus and Panchax to the same species, was certainly in error. Ann. Mag. Nat. Hist., (8) VII, 1911, p. 324. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 199 The group, called Oryzias Jordan and Snyder, is the Indian genus Aplocheilus McClelland, as first restricted by Bleeker to Aplocheilus melastigmus McClelland. In a recent paper upon these fishes Dr. Ernst AhP^^ recognizes Oryzias as a distinct genus, as it lacks enlarged teeth on the sides of the premaxillary, these being characteristic of Aplocheilus. The teeth in Oryzias are very small and slender. 169. [268] Oryzias latipes (Temminck and Schlegel). Med(2/i;;a = High-eyes. Kachi River at Nagoya (Jordan) ; brook at Yamawa near Kagoshima (Wakiya) ; Aomori (Beppu); Lake Biwa (Jordan); Lake Kasumigaura (Hattori). A little inhabitant of ditches in the rice-fields. 170. [269] Pseudorasbora parva (Temminck and Schlegel). The genus Fundulichthys Bleeker has been shown by Mr. George S. Myers to have no real existence, its type-species Fundulus virescens Temminck and Schlegel, being founded on a bad figure of Pseudorasbora parva. Family NOTACANTHID^. 171. [270] Polyacanthonotus challengeri (Vaillant). The names Polyacanthonotus Bleeker, Versl. Akad. Amst., (2) VIII, 1874, p. 368 (type Notacanthus rissoanus De Filippi and Verany) and Zanotocanthus Gill, Johnson’s Cyclopedia, III, 1876, p. 883 (the same type) have priority over Mac- donaldia Goode and Bean. Family SYNGNATHIDiE. 172. [274] .Syngnathus schlegeli Kaup. = Tooth-pick fish. Bay of Mikawa (C. Ishikawa); Misaki (Aoki). 173. [287] Hippocampus japonicus Kaup. = Northern Sea-horse. Enoshima (Jordan). 174. [288] Hippocampus coronatus Temminck and Schlegel. Uma-umi = Sea-horse ; Tatsu = Dragon. Enoshima (Jordan). Family AULORHYNCHIDTl. 175. [291] Aulichthys japonicus Brevoort. Kwda-ya^ara^Pipe-Yagara. Misaki (Aoki). Ahl, Zool. Anz., May 1924, p. 50. 200 MEMOIRS OP THE CARNEGIE MUSEUM. Family FISTULARIIDtE. 176. [292] Fistularia petimba LacepMe. AA;(i-^agrara = Red Yagara. Shizuoka market (Jordan) ; Toba market (Jordan and Yamamoto) ; Kagoshima Bay (Wakiya). 177. [293] Fistularia serrata Cuvier. Ao-^a^ara = Green Yagara. Osaka market (Jordan); Toba market (Jordan and Yamamoto); Misaki (Aoki); Mikawa Bay (M. Ishikawa). Family MACRORHAMPHOSID^. 178. [295] Macrorhamphosus sagifue Jordan and Starks. = Egret-piper. Ten specimens from Misaki (Aoki). Depth, 3.75 to 4.1. Dorsal spine serrated, inserted well before anus, when depressed reaching beyond base of caudal. Regan®^ has lately well reviewed the species of this genus. Family GASTEROSTEIDtE. Genus Gasterosteus Linnaeus. The three-spined Sticklebacks have long been known to be among the most variable of all fishes, and they have been referred to a large number of nominal species. We have examined many series representing localities in all the northern continents. The marine form of northern Europe, G. aculeatus = G. trachurus of Green- land and the northern Pacific on both sides, G. loricatus, G. cataphractus = obolarius = insculptus, seem not to be separable into local species or subspecies. This form is characterized by the large size attained, by the complete development of lateral plates and caudal keel, the long, strong pubic plate, the long falcate pectoral fin, the serrate ventral spine, and the high average number of dorsal and anal rays. Up the streams and southward this circumarctic form, G. aculeatus, varies through a most complex and irregular, though complete, intergradation toward and into a very different type, both in Europe and on both sides of the North Pacific. The change involves a reduction in the adult size, a loss of lateral arma- ture, a shortening of the pubic plate, a shortening and rounding off of the pectoral fin, and a reduction in the number of dorsal and anal fin-rays. The change, how- ever, is not fully identical in the two oceans, for in Europe the loss of plates on the average is brought about more abruptly and from a more posterior point, so Ann. Mag. Nat. Hist., (8) XIII, 1914, p. 17. JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 201 that intermediate types frequently have the plates irregularly absent in advance of the caudal keel which still persists; while in the Pacific the posterior plates become gradually shortened and eliminated together, so that the caudal keel is rarely evident, except in fully plated individuals. Moreover, the European stream-type, G. leiurus, has on the average stronger serrulations on the fin-spines than does the analogous Pacific form G. microcephalus. Toward the southern end of the range of Gasterosteus in both Europe and California a few especially distinct, almost wholly unarmed, races have been pro- duced, as Regan has indicated.®’ Of these the race inhabiting the Santa Ana system of streams of Southern California, G. williamsoni Girard = santo-annee Regan®® is particularly extreme, and has the fin-spines extremely short and posteriorly inserted. It is connected with G. microcephalus by a wholly intermediate type, occurring in neighboring waters both to the northward (Santa Clara River) and southward. It is closely analogous to algeriensis of northern Africa. The occurrence of extensive and complete intergradation seems to call for the trinomial distinction of these forms. In the Pacific region we may recognize there- fore three subspecies: Gasterosteus aculeatus aculeatus Linnaeus Gasterosteus aculeatus microcephalus (Girard) Gasterosteus aculeatus williamsoni (Girard). The fully armed form of the western Atlantic (biaculeatus) and its partially plated fresh-water derivative {cuvieri) have not been thoroughly studied in the present connection, but the examples we have seen seem quite distinct from any of the forms of G. aculeatus. The junior author is accumulating material for a more exhaustive study of all the Sticklebacks. Our Japanese specimens should be referred to two subspecies. 179. [297] Gasterosteus aculeatus aculeatus (Linnaeus). Hari-uwo = 'Needle-fish. Nineteen specimens from Kushiro, Hokkaido (Tanaka) represent a race with the body slenderer and the body-plates smoother than usual. Largest specimen 58 mm. long to caudal; plates and keel complete, but the posterior plates abruptly shortened; pubic plate long; pectoral long, narrow, and falcate; dorsal soft rays ®® Ann. Mag. Nat. Hist., (8) IV, 1909, pp. 435-437. ®® The exact type-locality of Gasterosteus williamsoni is perhaps still in doubt, although Lieutenant Williamson is known to have crossed the Sierra Madre range at the head of the Santa Ana River, near Banning, California. Nevertheless the description leaves little doubt in regard to its identity with the form lately named santa-annce by Regan. The number of vertebrae used by Regan is not constant, but may prove to be an average character of some value. 202 MEMOIRS OF THE CARNEGIE MUSEUM, 12 to 14, anal soft rays 9 to 11 (counting last ray as only partly divided). Depth of body, 4.2 to 4.8 in length to caudal. 180. [297A] Gasterosteus aculeatus microcephalus (Girard), Harniko = Little Needle. Fifteen specimens from Lake Biwa represent the partly plated Japanese form of Gasterosteus aculeatus. They seem wholly inseparable from specimens referred to G. microcephalus from California and the Aleutian Islands. Dorsal soft rays, 11 or 12; anal soft rays, 8 or 9; plates, 4 to 7. Franz has named and figured Japanese specimens of this subspecies as Gas- terosteus williamsoni japonicus. If the Japanese race should prove sufficiently distinct to be separated nomenclaturally, it will, however, require a new name, as both Houttuyn and Steindachner have used the combination Gasterosteus japonicus. There can be little doubt that these naked Sticklebacks of the rivers, however similar, have diverged independently from the marine form. As “ontogenetic species” they need not enter systematic lists. Genus Pungitius Coste. Pungitius Coste, 1846, replaces Pygosteus Gill, 1861.'^^ In the present connection we follow Berg's review®* of the species of Pungitius in all respects but one. We refer the partially naked species of northern Japan to P. brevispinosus rather than to P. pungitius. 181. [298] Pungitius sinensis (Guichenot). i7ansa5a = Needle-mackerel. Three specimens from Noo, Niigata-Ken, on the western side of Japan, have small plates along the entire course of the lateral line, and for this reason alone are referred to P. sinensis. As used by Berg, P. sinensis seems to us to be a complex of more or less unrelated races, which have developed plates along the lateral line anteriorly. Our specimens differ from those described by Jordan and Starks as Pygosteus steindachneri, and by Tanaka as Pygosteus kaiharce, in having only one instead of two soft rays in the ventral fin (except on one side of one specimen). In P. brevispinosus and in specimens from Kamchatka referred to Pungitius, two rays are only occasionally evident in the ventral.®® ®’^ See Jordan, Stanford Univ. Publ. (Biol.), Ill, 1923, p. 174. ®* Proc. U. S. N. M. XXXII, 1907, pp. 451-454. ®® We have examined large series of the Pacific forms of Gasterosteus aculeatus, but fail to find a single variant in the number of ventral rays (I, 1). In three paratypes of G. gladiunculus Kendall from Maine, however, we find two soft ventral rays on both sides of two specimens and on one side of the third. This well marked species is now regarded by Kendall as the original G. bispinosus Walbaum. JORDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 203 The six types of Pygosteus undecimalis ( = Pungitius tymensis) at hand have no soft ventral rays at all, and one specimen lacks even the ventral spines. As Berg has remarked, Day has described and figured a local race of Pungitius from Ireland, which is closely comparable with P. tymensis. The specimens from Noo have 8 or 9 dorsal spines. 182. [300 and 301] Pungitius brevispinosus (Otaki). Ira-tomiyo. We refer to this species the specimens without anterior armature referred by Jordan and Starks to Pygosteus steindachneri, the Hokkaido specimens referred by Berg to Pygosteus pungitius and the material in the present collection, consisting of three specimens from Aomori (Beppu) and a series from Sapporo (T. Kawamura). This form, which will probably prove to grade into P. pungitius, differs in having fewer dorsal spines, as the following table shows. In P. pungitius proper the number varies from 7 (abnormally as few as 2) to 12, 10 least frequently occurring. Number of dorsal spines VII VIII • IX X Number of specimens 2 . 15 22 1 Specimens from Petropavlovsk, Kamchatka, seem to have the spines in about the usual number for P. pungitius. Number of spines IX X XI XII Number of specimens 4 28 9 1 Family EXOCCETIDiE. Genus Exoccetus Linnseus. The name Exoccetus of Linnseus, included his species E. volitans (1758) and E. evolans (1766) both of which belong to the Halocypselus-iype, having short ventrals not used for flight. 183. [303] Exoccetus volitans Linnseus. I daten-tori = Swift bird. One adult specimen, 6.5 inches long, was taken by Dr. Jordan at sea, about three hundred and fifty miles east of Yokahama. It is evident that Exoccetus volitans of the Tenth Edition of the Systema Naturce and Exoccetus evolans of the Twelfth Edition are both based on species with the ventral fins short, the group called Halocypselus by Weinland. Probably both names belong to the same species. But to what extent the species is cosmopolitan, and whether all forms of this type the world over belong to one species is prob- lematical. The oldest name assigned to a Pacific member of this group seems to be Exoccetus splendens Abel, ‘Hourn. China, 1818, 4.” 204 MEMOIRS OP THE CARNEGIE MUSEUM. The specimen noted above shows the following characters: Mouth unusually small; jaws and palate toothless; snout short, shorter than eye; scales 44; D. 13; A. 14; pectoral fin extending to last ray of dorsal; ventral twice in head, reaching half-way to anal; pectoral with the first ray long, not branched, reaching nearly to tip of fin; second ray slenderer, forked; dorsal fin low, anal a shade higher; upper caudal lobe unusually long, more than two-thirds length of lower. Color in life, plain dark blue above, center of scales a little darker; pectoral plain dusky, its basal half translucent, its edge narrowly white; dorsal plain olive; caudal plain dusky, and pure white; ventral white, its edge slightly dusky; length 6.5 inches. This description discloses no differences on which an Exocoetus splendens could be differentiated as the Pacific representative of E. volitans. 184. [304, 306] Cypselurus agoo (Temminck and Schlegel). Tobino-uwo = Bird-fish.; Agu. Cypselurus hirundo Jordan and Starks, Proc. U. S. N. M., XXVI, 1903, p. 542 (not Exocoetus hirundo Steindachner) . We provisionally identify as the young of this common Japanese Flying-fish eight specimens, 53 to 65 mm. long to the caudal fin, collected by Aoki at Misaki. They differ from the adult, as do the young of other species of the genus, in having the pectoral fins somewhat shorter, and in having a barbel at the tip of the chin. This barbel is short, constricted at its base but broadly expanded into a sub- triangular black flap, having the lower edge uneven; the flap is about half as wide as the eye. The coloration is also wholly unlike that of the adult, and corresponds with the description given by Jordan and Starks in the account quoted above. Our specimens agree entirely with the young identified by Jordan and Starks as Cypselurus hirundo. It is certain, however, that they are specifically different from Steindachner’s fish, also described from a young specimen. The young of Cypselurus hirundo differs widely in coloration and in the character of the barbels: “Die Unterlippe ist verdickt und endigt jederseits in einige zarte Tentakeln von geringer Liinge.” 185. [307] Exonautes brachycephalus (Gunther). A young flying fish, 74 mm. long to the caudal, obtained by Aoki at Misaki, Japan, probably belongs to this species. At least it appears to be identical with the Japanese specimen 107 mm. long referred by Jordan and Starks^” to Cypselurus brachycephalus. Proc. U. S. N. M., XXVI, 1903, p. 539, fig. 2. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 205 Head, 4.3; depth, 6.3; dorsal, 10; anal, 11; the two fins commencing at the same vertical (the type has two more rays in each fin, but the specimen described by Jordan and Starks is intermediate). Snout short and not produced, only two- thirds as long as the orbit, which is contained nearly three times in head; inter- orbital gently elevated toward orbital margins. Scales about 50 in lateral line, about 33 from occiput to dorsal fin, 7 between dorsal fin and lateral line; ridge of \ lateral line not more than usually conspicuous. Pectoral fin extending to opposite tips of last dorsal and anal rays, when depressed; ventral fin extending a little farther, but not quite to caudal base; the structure of the rays of the paired fins is as described by Jordan and Starks. There is no trace of a barbel. Body clear light brown above, darkest on snout and caudal base, light below. Dorsal with a large and conspicuous black spot; anal fin clear; lower margin of upper caudal lobe and greater portion of lower lobe darkened, as also in the larger Japanese specimen; ventrals mostly black, pale around the margin; pectoral light in *the rays, but deep brown on the membranes, except along the lower edge of the fin, on a large oblong area near middle of fin and on a large triangle between the oblong and the tip of the fin; these specified areas without pigment. Family HEMIRAMPHID^. 186. [310] Hyporhamphus sajori (Temminck and Schlegel). Sayori. Sapporo market (Majima); Otaru market (Takayasu); Tokyo, Osaka, and Nagoya markets (Jordan); Mikawa Bay (M. Ishikawa); Toyama, Miyazu, Misaki (Aoki). Generally common throughout Japan. This species reaches a length of at least 26 cm. (measured from tip of pre- maxillaries to caudal base). 187. [312] Hyporhamphus kurumeus (Jordan and Starks). Kachi River at Nagoya (Jordan); Lake Kasumigaura, north of Tokyo (Hattori). These are the first records of the species from Hondo, the main island of Japan, the types having been obtained in Chikugo River, Kyusyu, and no specimens having been subsequently reported. Dorsal rays, 15; anal, 17; scales, about 80 to 90 (deciduous, very difficult to count, given as about 70 in the description of the type, but we may count more than 80 in the types) ; origin of anal a little in advance of that of dorsal. 206 MEMOIRS OF THE CARNEGIE MUSEUM. Family SCOMBRESOCID.E. 188. [315] Cololabis saira (Brevoort). Samma; Saira. Tokyo market, Nagoya market (Jordan); Kushiro (Takayasu); Misaki (Aoki). Very abundant everywhere and much valued as food. Mr. Hubbs” has lately indicated that the Californian form, C. hrevirostris, is not tangibly different from the^ Japanese saira, which was first named. Family BELONIDTl. 189. [316] Tylosurus anastomella (Cuvier and Valenciennes). = Gar-fish. Miyazu; Fukuoka (Hamada). 190. [317] Tylosurus schismatorhynchus (Bleeker). H ama-datsu = ^hore Gar-iish. Kobe market (Jordan); Toyama, Sea of Japan (S. Yoshizawa). Not rare. Family SPHYRiENIDiE. 191. [320] Sphyraena japonica Cuvier and Valenciennes. Kamasu = Barracuda. We have one adult specimen of the large Barracuda from the Tokyo market (Jordan) and seven young from Misaki (Aoki). Jordan and Snyder in 1900 took the species at Nagasaki. Head, 2.8 to 3.3; depth, 7.2 to 8.2; depth of caudal peduncle, 4.6 to 5.7 in head; eye, 4.9 to 5.9; snout, 2.2 to 2.4; interorbital, 5.1 to 5.7; upper jaw, 2.2 to 2.4. Dorsal, V-I, 9; anal, II, 8; scales, 12-111 to 125-14. Ventral inserted under origin of first dorsal, a little behind tip of pectoral, or at a distance from anal origin contained 3.6 times in the length to caudal. Preopercular margin squarish. Color dark; the mouth parts largely black. 192. [322] Sphyraena pinguis (Gunther). Kamasu = Good Salmon (Barracuda). Kyoto and Kobe markets (Jordan); Mikawa Bay (M. Ishikawa); Toyama (Yoshizawa); Miyazu, Fukuoka (Hamada). This small Barracuda is generally common in the markets and much valued as a pan-fish. Records of Sphyrcena ohtusata from Japan probably refer to this species, which seems different from true S. obtusata. Head, 3.1 to 3.4; depth, 6.2 to 7.5; depth of caudal peduncle, 4.0 to 4.5 in head; eye, 4.6 to 5.6; snout, 2.2 to 2.4; interorbital, 5.1 to 6.0; upper jaw, 2.4 to 2.6. Dorsal, V-I, 9; anal, II, 8; scales 8 to 10 — 80 to 90 — 11 to 12. Ventral inserted Publ. Zool. Univ. Cal., XVI, 1916, p. 157. JOED AN AND HUBBS: JAPANESE FISHES COLLECTED 1922, 207 under middle of pectoral; pectoral not quite reaching vertical from dorsal origin; distance between insertion of ventral and origin of anal one-third the length to caudal. Preopercular margin squarish, somewhat produced backward at angle. Family OSPHRONEMID.E. 193. [Extraterr.] Macropodus opercularis (Linnaeus). Three specimens obtained from Professor Yosiro Manabe, taken at Okinawa, Ryukyu Islands. Dorsal rays XIII or XIV, 8; anal XX, 14 or 15. Soft dorsal and anal lobes considerably produced and sharp; caudal lobes not at all, or only slightly, produced; soft dorsal rays branched, anal rays unbranched. Vertical fins ensheathed by scales at their bases. Material of this species from Soo-chow, China, has been sent to the Museum of Zoology, University of Michigan, by Dr. Cora B. Reeves. Family OPHICEPHALID.E. 194. [Extraterr.] Ophicephalus argus Cantor. Specimens of this common Chinese fish were obtained by Professor Gee at Soo-chow, China. Family ATHERINID^. 195. [323] Atherina bleekeri Gunther. Bay of Kagoshima (Wakiya). 196. [325] Atherina tsurugae Jordan and Starks. Gin-iso-iwashi = Silvery Surf-sardine. Two silver-sides referable to Atherina tsurugce were given to Dr. Jordan by Mr. Mikimoto, who obtained them at his pearl-plantation on Tatoku Island. Family MUGILID.E. 197. [328] Mugil cephalus Linnaeus. Bora = Mullet. Tokyo, Nagoya, and Osaka markets (Jordan); Kagoshima Bay (Wakiya); Mikawa Bay (Ishikawa) ; Fukuoka (Hamada) ; Lake Kasumigaura (Hattori) ; Misaki (Aoki). Everywhere excessively common in the markets. The length of the gape is contained 1.15 to 1.3 times in the width of the mouth. Young, as small as 51 mm. to caudal, have three anal spines like the adult. 208 MEMOIRS OF THE CARNEGIE MUSEUM. We are still unable to follow Gunther and Oshima^^ in separating the common Striped Mullets into more than one species the world over. We find no specific differences in the form of the mouth in different regions, and such differences as may exist in the number of scale-rows are slight and overlapping. In our material we count the transverse rows as follows: Naples, 39 to 43; Florida, 39; Texas, 38 to 40; Peru, 38 to 41; mouth of Colorado River, 38 to 41; Hawaii, 38 or 39; China, 38 to 41; Japan 36 (rarely) to 41. In Formosan material, which he referred to two different species, Oshima counted 38 to 42 scale-rows. Genus Liza Jordan and Swain. Three species of Liza occur in Japan. They have lately been differentiated by Tanaka in a paper published in Japanese. From a translation of this paper made by Mr. Kasawa, and an examination of two of the species (no specimens of Liza akame are in the collection), we have prepared the following key. Key to Japanese Species of Liza. a. Tip of pectoral not reaching vertical from origin of first dorsal. b. Back broad, depressed, not carinate; snout and eye only two-thirds as long as rest of head. Scales smaller, 37 or 41 in lateral series; depth of body about one-fifth of length to caudal menada. hh. Back narrower, compressed to a weak keel along mid-line; snout and eye nearly as long as rest of head. Scales larger, 35 or 36 in lateral series; depth of body, 3.85 to 4.5 hcematocheila. aa. Tip of pectoral reaching vertical from front of first dorsal. Scales larger, 34 in lateral series; body very deep (the greatest depth 3.5 in length to caudal (after Tanaka) akame . 198. Liza menada Tanaka. Liza hcematochila Tanaka, Fig. Desc. Fishes Japan, VIII, 1912, p. 137, pi. 37-39 (not Mugil hcematocheilus Temminck and Schlegel). Liza menada Tanaka, Zook Mag., No. 336, Oct. 15, 1917. One specimen of this species, well figured by Tanaka as “Liza hsematochila,” is in the present collection. It was secured in the Nagoya market. Jordan and Snyder in 1900 took the species at Osaka, Wakanoura, Hiroshima, Hakodate, and the Ishikari River. 199. [329] Liza haematocheila (Temminck and Schlegel). Menada. Three specimens are in the present collection, all from the Bay of Mikawa (C. Ishikawa). Seen in the markets of Yokohama and Osaka. Jordan and Snyder took specimens in 1900 at Katase, Enoshima, and Wakanoura. Not rare in markets, reaching a large size. Ann. Car. Mus., XII, 1919, p. 268; XIII, 1922, p. 240. Zool. Mag., No. 336, Oct. 15, 1917. JORDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 209 Family BERYCID.E. 200. [332] Beryx splendens Lowe, = Golden-eye Porgy. Tokyo and Yokohama markets (Jordan). Dorsal rays, IV, 13 to 15; anal, IV, 26 to 28; depth of body, 2.8 to 2.9 in length to caudal base; head, 2.7; scales with spinules arranged in quincunx order on each side of a central spineless groove, 9-67 or 68-19 in number; mouth bright red within. We have no Atlantic material at hand for comparison. 201. [333] Hoplostethus mediterraneus Cuvier and Valenciennes. Kagoshima Bay (Wakiya). Family HOLOCENTRIDtE. 202. [336] Holocentrus spinosissimus Temminck and Schlegel. = Head Porgy. Toba market (Jordan and Yamamoto). 203. [340] Ostichthys japonicus (Cuvier and Valenciennes). Ehisu-dai = Ebisu-porgy ; Fish-god. Tokyo market (Jordan); Kochi, Miyazu (Wakiya). Family POLYMIXIID^. 204. [343] Polymixia japonica Gunther. = Silver-eye. Polymixia japonica Gunther, Ann. Mag. Nat. Hist. (4) XX, 1877, p. 436. — Steindachner and Doderlein, Denksch. Akad. Wiss. Wein, XLVII, 1883, p. 261, pi. 4, fig. 2. — Jordan and Fowler, Proc. U. S. N. M., XXVI, 1902, p. 18. Shizuoka market (Jordan). This species by oversight is currently credited to “Steindachner.” Family MONOCENTRID^. 205. [344] Monocentris japonicus (Houttuyn). Matsukasa-uwo = Pme-cone-hsh.. Misaki (Aoki), Commoner southward. Family SCOMBRIDTl. This family, as now restricted, contains species of rather small size, with the spinous dorsal short, remote from the soft dorsal; the corselet indistinct; the vertebrae 31; the posterior without the “trellis-like” structure, produced by the 210 MEMOIRS OF THE CARNEGIE MUSEUM. union of haemal processess; interspinal bones weak and slender; mouth large, with minute teeth. We recognize three genera, only one of them found in Japan proper, a. Scombrinw. Gill-rakers moderate, about 26; body fusiform; vomer and palatines toothed; vertebrse 31. b. Air-bladder wanting. Atlantic Scomber. bb. Air-bladder well-developed Pneumatophorus. aa. Rastrelligerince. Gill-rakers very long and numerous, feathery, about 57, filling the mouth; body compressed; no teeth on vomer or palatines. South Seas Rastrelliger. In arranging the species of Mackerels and Tunnies of our collections in Japan, we have rather closely followed the determinations given in the elaborate and painstaking monograph of Dr. Kamakichi Kishinouye, entitled ‘'Contributions to the Comparative Study of the So-called Scombroid Fishes,” Journ. Agri., Imp. Univ. Tokyo, March, 1923. Dr. Kishinouye divides the old family ScombridcB, on the basis of skeletal and muscular characters into four, which seem to be well-defined and are certainly natural groups. He also adds certain new genera, as given below. Genus Pneumatophorus Jordan and Gilbert. Pneumatophorus Jordan and Gilbert, Proc. U. S. N. M., V, 1883, p. 593 (as subgenus). — Starks, Science, LIV, 1921, p. 222 (elevated to generic rank). The Chub-mackerels have until lately been usually regarded as comprising a single cosmopolitan species, called Scomber colias or Scomber japonicus, the latter name being the earliest. Evermann and KendalP'* found certain differences in their material and consequently recognized two species, colias and japonicus. In 1922 Starks^^ further compared specimens from California and Massachusetts. In the meantime, however, Tanaka^® had recognized the existence of two distinct species in Japan, distinguishing Scomber tapeinocephalus Bleeker from the commoner species S. japonicus. A more extended comparison of these mackerel from various parts of the world has obviously been needed. We have with the aid of Masunosuke Kasawa, a student from Sapporo, examined all of the specimens at our disposal. This material seems to comprise seven species, of which two are represented in Japan, and so far as we know, confined to the waters of that country. They are P. tapeinocephalus and P. japonicus. Besides these we have specimens from Aus- tralia, Hawaii, and Socorro Island, off the Coast of Mexico, which seem to belong to a Polynesian species, which may be provisionally known by the old name Proc. U. S. N. M., XXXVIII, 1910, p. 327. Copeia, No. 103, 1922, p. 9. ^® Zool. Mag., XXIX, 1917, p. 347. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 211 australasicus of Cuvier and Valenciennes. The other species are nearer japonicus. Each of them occupies a special faunal area: P. colias European; P. grex Western Atlantic; P. peruanus, sp. nov., Peru; and P. diego California. Seven species are contrasted in the following key. Key to the Species op Pneumatophorus. a. Dorsal spines 11 or 12; scales fewer than 200 in the lateral line. Gill-rakers 25 to 27 on lower limb of outer arch. h. Scales between occiput and first dorsal fin, 35 to 45; scales between origin of second dorsal fin and lateral line, 19 to 23; length of head .285 to .295 of length to caudal base; distance from tip of snout to first dorsal, .365 to .38; distance from tip of mandible to ventral fin, .345 to .355. australasicus P hh. Scales between occiput and first dorsal fin, 24 to 32; between second-dorsal and lateral line, 15 to 19; length of head, .275 to .285 of length to caudal; snout to dorsal, .355 to .36; mandible to ventral, .33 tapeinocephalus. aa. Dorsal spines 9 or 10; scales more than 200 in the lateral line (186 to 209 in P. grex). c. Gill-rakers 29 to 32 on lower limb of outer arch. Scales along lateral line, 200 to 229; scales from occiput to dorsal fin, 39 to 50; scales from origin of second dorsal fin to lateral line, 20. Measurements in hundredths of length from tip of snout to base of caudal: length of head, .28 to .295; snout, .09 to .095; upper jaw, .105 to .115; snout to dorsal, .37 to .385; mandible to ventral, .335 to .38 colias.’’^ cc. Gill-rakers 23 to 28 on lower limb of outer arch. d. Scales in lateral line, 186 to 209; before first dorsal fin, 34 to 41; from origin of second dorsal to lateral line, 19 to 21. Gill-rakers, 23 to 27. Measurements in hundredths of length to caudal fin: length of head, .27 to .28; snout, .085 to .09; upper jaw, .10 to .105; snout to dorsal, .36 to .37; mandible to ventral, .325 to .34 grex^^ dd. Scales in lateral line, 205 to 231; before first dorsal fin, 40 to 60. Gill-rakers, 26 to 29. e. Measurements in hundredths of length to caudal base: head, .275 to .29; snout, .09 to .095; upper jaw, .10 to 115; snout to dorsal ,.355 to .375; mandible to ventral, .325 to .35. /. Scales between the second dorsal fin and lateral line 19 to 26, usually fewer than 23. japonicus. ff. Scales between the second dorsal fin and lateral line 22 to 27, usually more than 23. diego.^^ ee. Measurements in hundredths of length to caudal base: head, .29 to .32; snout, .095 to .105; upper jaw, .115 to .135; snout to dorsal, .38 to .405; mandible to ventral, .365 to .375. peruanus.^^ Specimens from off Moreton Bay, Queensland; Lord Howe Islands (near Australia); Hawaii; Socorro Island, off the west coast of Mexico. Diagnosis based on a specimen from Naples, and one of three from the Canary Islands. Diagnosis based on a series of adults from Woods Hole. Diagnosis based on series of adults from California. Pneumatophorus peruanus Jordan and Hubbs, sp. nov. Type-specimen 201 mm. long to the caudal fin, from the Bay of Callao, Peru; collected by the late Admiral L. A. Beardslee; Cat. No. 6218, Stanford University Collection, C. M. Cat. Fishes No. 7847. Paratypes are from the same locality and from the Galapagos Islands. 212 MEMOIRS OF THE CARNEGIE MUSEUM. 206. [345, in part] Pneumatophorus tapeinocephalus (Bleeker). Marusaba^ Hound Mackerel. Scomber tapeinocephalus Bleeker, Nat. Tydsk. Ned. Ind., VI, 1854, 407 (Nagasaki), Tatoku Island (Mikimoto). Jordan and Snyder (in 1900) took this species at Hakodate, Tateyama, Kobe, and Matsushima, though not distinguishing it at the time from the common mackerel or Saba. It is nowhere abundant. 207. [345] Pneumatophorus japonicus (Houttuyn). = Mackerel ; Hirasaba = Broad Mackerel; Go7nasaba — Oil Mackerel. Sapporo, Takashima, Tokyo, and Shizuoka markets (Jordan); Kushiro; Misaki; Miyazu; Yokohoma; Kobe; Yamada; Hakodate; Nagasaki; Same; Mat- sushima. This is the common mackerel of Japan, found daily in abundance at every port. In the markets, according to Kishinouye, there are three distinct forms of mackerel, which he regards as not definable as subspecies. The usual shore-form, Gomasaba = Oil Mackerel, Scomber pneumatophorus minor Schlegel, has gray spots below the lateral line, and the dark markings of the back do not cross the line which is marked by a row of round spots. The Hirasaba, or Flat Mackerel (variety major) of Schlegel has but nine dorsal spines, while the Marusaba, or Round Mackerel {Scomber tapeinocephalus Bleeker) has eleven or twelve. In the Hirasaba j the dark streaks on the back cross the lateral line, and the caudal is yellowish. The ‘‘Marusaba’’ is the species named Scomber tapeinocephalus by Bleeker. The others correspond respectively to the Scomber pneumatophorus major and minor of Schlegel, Scomber saba, and Scomber janesaba of Bleeker. The marusaba is no doubt a distinct species, and the others may prove to be so. The scanty descriptions of Scomber japonicus and Scomber auratus in the original paper of Houttuyn are identifiable only on the supposition that but one species of mackerel with eight or nine spines occurs at Nagasaki. Family CYBIIDiE. This family comprises the allies of the Spanish Mackerel of America, the Seer-fishes, or Sawara of Japan. It is visibly characterized by the many-rayed first dorsal, which extends nearly to the front of the second. Body compressed; mouth large; teeth strong, in one row, usually more or less compressed; lateral line sinuous, often branched or duplicated; vertebrae 31 to 64, usually more than 40; interhaemels well developed, usually long and numerous; gill-rakers small and JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 213 few, sometimes wanting, not more than 13 in number. Besides the living genera, numerous others are characteristic of the Miocene of California. The following are recognized: Ocijstias, Thyrsocles, Turio, Thyrsion, Zaphleges, Auxides, Zestias. This group is probably the most primitive of the Mackerels. Key to the Japanese Genera of Cybiid.e. a. Acanthocybiince. Lamellse of the gills reticulated, as in the Sword-fish; gill-rakers none; intermuscular bones inserted on the ribs; body elongate; teeth very strong, trenchant; vomer with teeth. Size very large AcantJionjhium. aa. Lamellte of gills not reticulated; gill-rakers present; intermuscular bones inserted on vertebr®. b. Cybiince. Body elongate; teeth in jaws trenchant; teeth on vomer, c. Lateral line single; teeth compressed. d. Air-bladder present; lateral line simple. e. Gill-rakers three only Cybium ee. Gill-rakers 11 to 13 Scomber omorus. dd. Air-bladder wanting; lateral line with numerous short branches at right angles Sawara. bb. Sardince. Body plump; teeth not trenchant, their edges rounded, no teeth on vomer; lateral line simple; corselet at shoulder more or less distinct. /. Body scaly; tongue toothless; dorsal spines 18 to 22 Sarda. //. Body naked outside of corselet; tongue and palatines with villiform teeth; dorsal spines 14 Gymnosarda. Acanthocybium Gill. This genus contains giant mackerels, differing considerably from the smaller forms, especially in the peculiar netted structure of the gills, as in the Sword-fish, Xiphias. There are no gill-rakers and the broad triangular serrated teeth resemble those of a shark. It might well be made type of a distinct family. 208. [355] Acanthocybium sara (Lay and Bennett). Osawara, great Sawara or Seer-fish ; Okisawara = Off-shore Sawara. A valued food-fish, but from its huge size, eight to twelve feet in length, seldom preserved in collections. Teeth ^ to on each side. The Hawaiian species, Acanthocybium solandri (Cuvier and Valenciennes) has much smaller teeth ^ to ^ on each side. Cybium Cuvier. This genus, of which the type is Scomber commersoni Lacepede, may perhaps be retained as distinct from Scomberomorus Lacepede, on account of the reduction in the number of its gill-rakers (1 + 2 = 3). The teeth are triangular and minutely serrated. 214 MEMOIRS OF THE CARNEGIE MUSEUM. 209. Cybium commersoni (LacepMe). f/s/izsairam = Cow-sawara. This species, abundant in Formosa and southward is recorded from Japan proper by Kishinouye from a specimen taken at Yamaguchi. It was not seen by us. Cyhium commersoni may be separable from Scomberomorus by the few gill- rakers and serrulate teeth, as already indicated. Sierra Fowler (cavalla) has but eight gill-rakers, very strong teeth, and fifteen, instead of seventeen, dorsal spines; Apodoniis Bennett (immunis) and Chriometra Lockington {concolor) have the teeth subconical and more numerous. These may represent one or two distinct genera. Grammaiorycnus Gill {bilineatus) has two lateral lines and Lepidocybium Gill {flavobrunneum) has the lower teeth much enlarged, the dorsal spines but 12, and the finlets reduced to 4 or 5. Scomberomorus Lacepede. The name Scoynberomorus of Lacepede cannot be set aside to be replaced by Cijbium for the reasons assigned by Kishinouye. There seems to be no question that Lacepede’s Scomberomorus plumieri, based on a rather poor copy of the painting of Plumier, which had previously been the basis of Scomber regalis Bloch, belongs to the species now called Scomberomorus regalis. 210. [354] Scomberomorus sinensis (Lacepede). iJasa-saieam = Toothed Sawara; Inusawara = Dog-sawara ; Ushi-sawara = Cow-sawara. This species is not rare in southern Japan, though less abundant than the common “Sawara.” The original form of the specific name, sinensis, must replace chinensis. Genus Sawara Jordan and Hubbs, gen. nov. Type: Cybium niphonium Cuvier and Valenciennes. This genus differs from Scomberomorus in lacking the air-bladder, and in the presence of sharp branch canals placed at right angles along the course of the lateral line. 211. [353] Sawara niphonia (Cuvier and Valenciennes). Sawara. Kobe market (Jordan). The “Sawara” is an excellent food-fish, common everywhere southward, and valued for its rich and delicate flesh. Sarda Cuvier. Pelamys Cuvier and Valenciennes, name preoccupied. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922, 215 212. [352] Sarda orientalis (Temminck and Schlegel). Hagatsuwo ^TooihQd Albicore. Not uncommon in the markets. The species needs further comparison with Sarda chilensis of Chile and Sarda lineolata of California. Genus Gymnosarda Gill. 213. [347] Gymnosarda nuda (Gunther). 7somagrRro = Surf-tunny. Tokyo market (Jordan), one specimen, 29 cm. in length. It agrees well with the descriptions of Gunther and Khmzinger, and has never before been noticed in Japan proper, although found in Formosa. Dorsal XIV-12+VI; anal III-IO+VI; eye 5.1 in head; pectoral 6.8. Scales present only on the corselet and along the lateral line, very thick and deeply imbedded. The last dorsal spine, which is almost imbedded in the skin, lies about midway between the preceding spine and the front of the second dorsal fin. Teeth uniserial in jaws, stronger below than above; lacking on the vomer, in a fine band on the palatines. Hsemal processes normal, showing no trace of the specialized structure seen in Euthynnus, Katsuwonus, and Auxis. Body bluish black above, shading into silvery below, and showing no trace of dark stripes; cheeks silvery; opercles blackish; dorsal fins blackish, soft fin and finlets entirely white; pectoral dark; ventral black, except on the proximal half of the soft rays, which are white. Family THUNNIDTl. This family, separated from the Scombridce through the researches of Dr. Kishinouye, is thus characterized by him: “Cutaneous vascular system connected with a vascular plexus developed as sheets in the lateral line. Portions of the lateral muscle surrounding these sheets, situated on both sides of the vertebral column, dark red almost black in color. Another peculiar vascular plexus developed on the inner side of the liver or in the haemal canal. Circulation of blood in the liver especially well developed.” For the Tunnies and Albacores, Kishinouye proposes a separate order, Plecostei, defined as “having a cutaneous vascular system, connected with the vascular plexus developed as sheets in the lateral muscle. Another peculiar vascular plexus is developed in the inner side of the liver or in the haemal canal.” These derivatives of the Cyhiidce, are regarded by Kishinouye as among the most specialized of fishes. / 216 MEMOIRS OF THE CARNEGIE MUSEUM. The Plecostei, as thus defined, contain two families, Thunnidce and Katsu- wonidce. The first of these is thus described: ‘‘Body wholly covered with scales; second dorsal and anal with the anterior rays elevated; vertebrae 18+21 = 39; transverse process present; first vertebra short, anchylosed to the skull; ali- sphenoids meeting on the ventral median line; air-bladder present (except in one genus).” Key to Japanese Genera of Thunnid^. a. Cutaneous blood-vessels passing through the myotome of the fifth vertebra; surface of liver striated with fine venules. b. Pectoral fin short, about half head, not reaching front of second dorsal; dorsal and anal lobes low. Thunnus. bb. Pectoral fin very long, ribbon-like, reaching to the anterior dorsal and anal finlets, dorsal and anterior lobes moderate Germo. aa. Cutaneous blood-vessels passing through the myotome of the seventh vertebra; surface of liver not striated with venules. c. Posterior cardinal vein not contiguous with the Cuvierian ducts. Vascular plexus on the inner side of the liver; pectorals long, reaching to near end of second dorsal, dorsal and anal lobes moderate, about half head Parathunnus. cc. Posterior cardinal vein contiguous with the Cuvierian ducts; vascular plexus in the hsemal canal. d. Air-bladder well developed, long and narrow; pectorals long, extending beyond middle of second dorsal; dorsal and anal lobes much elevated, as long as head; gill-rakers about 30. Neothunnus. dd. Air-bladder wanting; gill-rakers about 26; pectoral fin shorter, reaching to near end of spinous dorsal; head shorter; mouth smaller Kishinoella. Thunnus South. {Thynnus Cuvier, preoccupied; Orcynus Cuvier, preoccupied; Albacora Jordan.) Pectoral fins very short, about two-thirds length of head; dorsal and anal lobes low. 214. [351] Thunnus orientalis Temminck and Schlegel. Meguro = Tuiiny; Yoko. Orcynus schlegeli Steindachner, Fische Japans, 1885, 178, taf. fig. 1 (Tokyo). Kobe market (Jordan and Yamamoto); Misaki, young (Aoki). This species, the young of which is locally known as Yoko, is rather common in the markets southward. It is said to reach a very large size. It needs comparison with its Californian cognate. Dorsal finlets bluish; those of the anal dull yellowish; no yellow on fins; body plump, with narrow bars of silvery, the bars more or less broken into lines of dots, and especially conspicuous in the young. Dorsal XIII to XV-14, VIII or IX. Anal 13 to 15, VII to VIII. Gill-rakers 12+13 = 25. JOEDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 217 This tunny is certainly T. orientalis of Schlegel, and we fail to note any char- acters by which T. schlegeli can be separated from it. Kishinouye takes the same view. A stuffed specimen about six feet long, with the dorsal and anal lobes about half of head, is in the Fisheries Institute in Tokyo. The species may be identical with the Tunny of the Atlantic or the Tuna of the islands of California, but only comparison of material can decide. This and other species of Maguro (Tunnies) and Shibi (Albacores) are so large as to forbid preservation by ordinary collectors. They can best be studied in the field, in Japan and in Hawaii, where hundreds of specimens, ranging from five to ten feet in length are now daily brought into the markets. This condition has been well utilized in the recent studies of Dr. Kish- inouye. Germo Jordan. This genus, or subgenus, differs from Thunnus in the elongate, more or less ribbon-like pectoral fin, which is much longer than head, reaching at least to the second anal finlet. Internally Germo differs little from Thunnus. Kishinouye does not accept the genus Germo, remarking: “many systematists put too much weight on the length of the pectorals, but it has little value in the classification.” The species of Germo are imperfectly known, the size of market examples making preservation difficult. 215. [349] Germo germo (Lacepede). Tombo-shibi. A large Albacore was seen by Dr. Jordan at Shizuoka. It has the finlets black, edged with paler; dorsal lobe dusky, no yellow; dorsal black, its lobe one-third of head; body plump; snout rather long, longer than eye; pectorals very long, reaching second anal finlet. We follow Kishinouye in provisionally identi- fying the Japanese Albacore with Germo germo and Thynnus pacificus of the South Seas. But with the latter author we see no final certainty to be secured, except through comparison of material from Japan, Europe, California, and the South Seas. The Hawaiian Albacore, recorded provisionally by Jordan and Jordan as Germo alalunga is probably this species. According to Fowler, the form Scomber germon, was earlier used by Lacepede. Parathunnus Kishinouye. This genus is distinguished by internal characters, as already stated. The single recognized species has the pectoral fin elongate, though shorter than in 218 MEMOIRS OF THE CARNEGIE MUSEUM Germo] the lobes of the dorsal and anal moderate, about half length of head; free dorsal finlets eight only, the first being coalescent with the fin; body relatively robust. 216. [348 part] Parathunnus sibi (Temminck and Schlegel). Shibi; Mebachi = Wasp-eye; Daruma-Shibi = Chimkj-shihi. Thynnus sibi Temminck and Schlegel, Fauna Japonica, p. 97, pi. L. Common about Nagasaki. Germo sibi Jordan and Jordan, Fishes of Hawaii, Mem. Car. Mus., X, 1922, 33. Thunnus mebachi Kishinouye, Sui. Gak. I, 1915, 19; pi. I, f. III. Parathunnus mebachi Kishinouye, Study of Scombroid Fishes, 442, 1923, figs. Tokyo, Yokohama, Yamada, Kobe, Osaka; common in the markets. With Mr. Kitahara we identify this species with the Thynnus sibi of Tem- minck and Schlegel. Kishinouye rejects this decision, as only the deep body (the depth about equal to length of head) and long pectorals form tangible points of resemblance. But he leaves T. sibi unidentified, while this common species would seem to have been unnoticed by previous authors. Besides the two points, admitted by Kishinouye, we may add another noted by Schlegel. The first dorsal finlet is joined to the last ray, leaving but eight which are free. There is some discrepancy in the description of the color of the dorsal finlets. We found them dull yellow, each crossed by a rather broad, sharply defined, angular band of black, the edgings pale; dorsal lobes dull yellow; head with a marked brassy luster. Schlegel figures the finlets as dull yellow, describing them with the second dorsal and caudal as passing ‘Vers leur extremite au fond olivatre.” Kishinouye merely says “finlets yellow,” but figures them as having a faint dark margin. It may be that two species of this genus exist in Japan, distinguishable by the color of the finlets. The generic name Pelamys, resurrected from Klein and Walbaum by Fowler, cannot be substituted for Katsuwonus, as Walbaum simply quotes from Klein, without indicating acceptance of the pre-Linnsean name. (See Opinion No. 5, International Commission of Zoological Nomenclature.) Neothunnus Kishinouye. Neothunnus Kishinouye, Studies, etc. 1923, p. 445. {macropterus) . This genus is characterized by anatomical features, as already indicated. Its type species is distinguished by the very high lobes of the dorsal and anal, the height of each being two-thirds as long as head; pectoral very long, almost as ribbon-like as in Germo. In the known species the finlets are bright yellow. JORDAN AND HUBBSt JAPANESE FISHES COLLECTED 1922. 219 217. [350] Neothunnus macropterus (Temminck and Schlegel). Kiwada; Kihata. The Yellow-fin Albacore, found also in Hawaii and in California, is known at once by the clear lemon-yellow finlets, not edged with black, and by the very high lobes of the dorsal and anal, which are largely yellow. It is the most abundant of the Japanese Albacores; generally common in the larger markets. Kishinoella Jordan and Hubbs, gen. nov. Type: Thunnus rarus Kishinouye. This genus differs from Neothunnus in the absence of the air-bladder. In the type species, the pectoral fin is of moderate length, scarcely as long as the head; dorsal and anal lobes moderate; body fusiform; mouth small; finlets nine above; gill-rakers about 21. The single species is the smallest of the group of Albacores. 218. Kishinoella rara (Kishinouye). Koshinaga. Rarely taken off southern Japan, and not seen by us. Finlets yellowish, with a greyish margin; tip of second dorsal and anal washed with yellow. Family KATSUWONID^, Kishinouye. This family comprises small tunnies in which the haemal bones posteriorly are united below, forming what Liitken calls a ^hrellice-like” structure; air-bladder wanting; teeth weak, one-rowed; pectorals short; soft dorsal and anal low, elevated in front. Various internal characters are noted by Kishinouye, who divides the group into three genera. Key to Genera of Katsuwonid^. a. First dorsal long, reaching base of second; a pair of foramina on the dorsal surface of the skull; inferior foramen of vertebrte well developed, forming a well-marked trellice. b. Hypaxial as well as epaxial blood-vessels under the skin well developed; teeth in jaws only; verte- brse 41 Katsuwonus. bb. Hypaxial blood-vessels atrophied; teeth on jaws, palatines, and sometimes on vomer; vertebra? 49. Euthynnus. aa. First dorsal short, well separated from the second; hypaxial blood-vessels atrophied; lower foramen of vertebrae little developed; teeth in jaws only; vertebrae 49; epiha?mal spines well developed. Auxis. 219. [347] Katsuwonus vagans (Lesson). Katsuwo = Victor-fish ; Magatsuwo = True Katsuwo. Generally common southward. Kishinouye and authors generally apply to this species the name of the “Oceanic Bonito” of the Atlantic, which also occa- sionally is taken on the coast of southern California. It needs comparison with 220 MEMOIRS OF THE CARNEGIE MUSEUM. Mediterranean examples of Katsuwonus pelamis (Lesson). In our record of the fishes of Hawaii (p. 31) we have called this fish, the “Aku” of Hawaii, Euthynnus pelamis, but its generic separation from Euthynnus seems justified. Euthynnus Llitken. In addition to the superficial characters already named, Kishinouye finds internal features, which differentiate Katsuwonus from Euthynnus. Gymnosarda, with which genus Euthynnus has been confounded, proves to have quite a different anatomy and finds its place next to Sarda. 220. [347] Euthynnus yaito Kishinouye. Faffo = Moxa-scar. One specimen from Choshi (Ishikawa). This species is frequently taken along the shore in southern Japan. What seems to be the same species, locally known as ‘Tvawakawa,” is very abundant in Hawaii, whence it is recorded by Jordan and Jordan as Euthynnus alleteratus. But Kishinouye regards it as distinct from that Atlantic species, as well as from Euthynnus affinis (Cantor) of the Malayan region. It is the Thynnus tunnina of Schlegel, but that name belongs to the European alleteratus. The round dark spots, three to five in number, under the pectorals, are characteristic of the Yaito at all ages. In a Mexican species, Euthynnus lineatus, lately described by Kishi- nouye from Manzanillo, Mexico, these spots are replaced by short dusky bands. Auxis Cuvier. In this genus, the spinous dorsal is short, as in Scomber, but there is little agreement in other regards, and, as Kishinouye justly observes, the two genera belong at opposite ends of the series. All the species of ‘'Frigate Mackerels” have been provisionally referred to Auxis thazard from the South Seas. The Atlantic form is Auxis rochei (Risso) or Auxis bisus (Rafinesque) . As Kishinouye finds two well-marked species in Japan, we cannot identify either with Auxis thazard without comparison of specimens, 221. [346] Auxis tapeinosoma Bleeker. Medika, Maru-medika = Kound Medika. Auxis tapeinosoma Bleeker, Verh. Bat. Gen., XXVI, 98, tab. 7, fig. I, 1854. (Nagasaki). Auxis maru Kishinouye, Sui. Gak. Ho., I, 1915, 24, pi. I. JORDAN AND IIUBBS: JAPANESE FISHES COLLECTED 1922. 221 This species is common in the markets southward. It runs in schools. We have examples from Tokyo, Kyoto, Misaki, and Toyama. Kishinouye considers it a species distinct from A. tapeinosoma, which is described as otherwise similar, but with nine, instead of eight, dorsal finlets. Gill-rakers 36; D. IX or X-12-8; A. 13+Vri. But as Bleeker’s type came from Nagasaki it is probably the same. 222. Auxis hira Kishinouye. //fra-medzVra = Broad Medika. This species, not recognized by us, is said to be abundant in southern Japan. The fin-rays are the same as in the preceding species, the gill-rakers 39, the body deeper and more compressed, the depth about equal to length of head. Family GEMPYLIDtE. 223. [357] Promethichthys prometheus (Cuvier and Valenciennes). Kuro-shihi-kamasu — Black Tunny Barracuda. Tokyo market (Jordan). This needs comparison with its Atlantic cognate, originally described from St. Helena. 224. [358] Ruvettus tydemani Weber. Ruvettus pretiosus Jordan, Tanaka, and Snyder, Journ. Coll. Sci. Tokyo, XXXIII, 1913, p. 123, and other authors (not of Cocco). Ruvettus tydemani Weber, Fische Siboga-Expedition, 1913, p. 401. Ruvettus pacificus Jordan and Jordan, Mem. Car. Mus., X, 1922, p. 34. (Honolulu). Weber and Jordan and Jordan, independently noticing that in the Pacific Ocean Ruvettus has fewer fin-rays than in the Atlantic, have each given a new name to the Pacific form. Weber’s name has priority. This species has been recorded from Japan, but is not represented in the present collection. 225. [359] Nealotus tripes Johnson. A single specimen of this rare fish was collected by Aoki at Misaki. Dorsal, XX-3, 15, H; anal I-I, 1, 15-1- II; scales between origin of dorsal and lateral line, and nine scales between end of spinous dorsal and lateral line. We have, no Atlantic material for comparison. 222 MEMOIRS OF THE CARNEGIE MUSEUM. Family TRICHIURID^. 226. [363] Trichiurus japonicus (Temminck and Schlegel). Tachino-uwo = Sword-fish. Tokyo market, Toyama, Fukuoka, Noo. This excellent food-fish is very common in all markets. The key given by Klunzinger, (Fische d. Rothen Meeres, 1884, 120) is useful in distinguishing species of Trichiurus. Family ISTIOPHORIDTl. 227. [367] Tetrapturus mitsukurii Jordan and Snyder. Makajiki = TY\xQ Spear-fish. Common in the markets, reaching a length of twelve to fifteen feet, or more. Two other species of this genus, T. mazara Jordan and Snyder, and T. angusti- rostris Tanaka, have been described from Japan, but in the rush of the market they could not be discriminated. 228. [365] Istiophorus orientalis (Temminck and Schlegel). Bashokaijiki = Banana Spear-fish. Mr. Manabe presented Dr. Jordan with a young specimen of this species, taken by him at Kobe. It is also represented in the Museum at Yamada. The species is seldom seen in the markets. Family CARANGIDTl. 229. [369] Scomberoides orientalis (Temminck and Schlegel). I ke-katsuwo^ Bond Victor-fish. Kagoshima Bay (Wakiya). Dorsal rays, I-VI-I, 20; anal, II-I, 18 or 19. 230. [371] Seriola aureovittata Temminck and Schlegel. Buri; Fukuraji = B\ump Aji, or Caranx. Otaru market (Takayasu) ; Sapporo market (Majima) ; Osaka market (Jordan). 231. [371 A] Seriola quinqueradiata Temminck and Schlegel. With the preceding and quite as abundant. Dorsal rays V or VI (the sixth becoming buried in the skin with age) — I, 30 to 33; anal, II-I, 18 to 20; pectoral 17; caudal, 17, of which 15 branched; gill- rakers on lower limb of first arch, 21 to 23. Everywhere common southward in the markets. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922, 223 232. [372] Seriola purpurascens Temminck and Schlegel. i7^ramasR = Broad Salmon; Aka-buri = Amber-fish. Tokyo and Osaka markets (Jordan); Toyama (S. Yoshizawa); Noo. Rather common in the markets. Gill-rakers on lower limb, 13 or 14 and 2 rudiments. 233. [373] Seriolina intermedia (Temminck and Schlegel). Ai-huri = Blue Amber-fish. This small species, known generically by its rudimentary, tubercle-like gill- rakers, was very abundant one morning in the Nara market, (coming from Kobe) (Jordan) but no specimens were taken at the time, and none were seen afterwards. Body plump, shorter, and deeper than in Seriola. No yellow stripe on side. 234. [375] Decapterus maruadsi (Temminck and Schlegel. Ao-aji= Green Aji; Maru-aji = Round Aji. Toba market (Jordan and Yoshizawa); Fukui (Nonaka); Misaki (Aoki); Kobe market (Jordan); Miyazu. Common. 235. [376] Decapterus muroadsi (Temminck and Schlegel). Muro-aji — Bone-aji. Tokyo and Osaka markets (Jordan) ; Toba market (Jordan and Yamamoto) ; Kagoshima Bay (Wakiya); Toyama (Yoshizawa); Misaki (Aoki). Generally common. 236. [378] Trachurus japonicus (Temminck and Schlegel). Ma-aji^True Aji. Misaki (Aoki); Tokyo, Shizuoka, and Kobe markets (Jordan); Kushiro (Tanaka) ; Tatoku Island (Mikimoto) ; Kagoshima Bay (Wakiya) ; Mikawa Bay (M. Ishikawa); Toyama (S. Yoshizawa); Fukui (Nonaka); Miyazu, Noo. Very common in markets. 237. [379] Trachurops mauritiana (Quoy and Gaimard). Me-aji = Big-eye Aji. Tokyo and Shizuoka markets (Jordan); Wakanoura (Yoshigawa); Misaki (Aoki). We have compared Japanese specimens with others from various parts of the world, but are unable to appreciate any differences in proportions. In scale-counts slight average differences are suggested, but these are not sharp enough to warrant the division of the species. We count the scales along the lateral line to caudal 224 MEMOIRS OF THE CARNEGIE MUSEUM. base as 87 to 94 in Japan; 87 to 94, Philippine Islands; 86 and 90 in two from the Hawaiian Islands; 88 to 94 in Samoa; 83 to 93 in the region of Panama; 85 to 87 in Jamaica; 84 to 89 in Cuba; 90 in one from Woods Hole, Massachusetts. The gill-rakers vary from 9 to 11 + 27 to 30, without apparent geographical correlation. The two isolated anal spines seem to grow relatively longer with age. The name Selar must by our rules supersede Atule not Trachurops. 238. [380] Carangoides equula (Temminck and Schlegel). iJfra-ajf = Broad Agi. Shizuoka and Osaka markets (Jordan) ; Toba market (Jordan and Yamamoto) ; Kochi (Wakiya); Mikawa Bay (M. Ishikawa); Misaki (Aoki). Rather common. Body everywhere suffused with brassy; fins light yellow. 239. [381] Longirostrum delicatissimum (D6derlein)®h Shi7na-aji ^ Striped Aji. Osaka and Tokyo markets (Jordan). 240. [383] Caranx bixanthopterus RuppelL This common food-fish of the Middle and Western Pacific, widely known as Ulua, was not obtained in 1922, unless represented among the numerous young examples from Misaki, Tokyo, Toba, and Kagoshima. These specimens have been placed in the American Museum to be studied by Mr. John T. Nichols. The name Caranx for steri Cuvier and Valenciennes, cannot possibly belong to this species. Caranx melampygus Cuvier and Valenciennes, wrongly identified with Caranx stellatus Quoy and Gaimard, has the same long dorsal and anal fins (D. 23, A. 19), the anterior rays of both fins much elevated and black at tip. But Dr. Wakiya finds a species of this type in Ryukyu with these lobes proportionately still higher. This species he calls Caranx melampygus. The oldest name certainly applicable to the Ulua or Kasumi-aji is bixanthopterus, an appropriate term as the pectoral fins are bright yellow in life. 241. [387] Alectis ciliaris (Bloch). = Silk-thread Porgy. Tokyo market (Jordan); Mikawa Bay (M. Ishikawa); Misaki (Aoki). Young specimens with long filaments on the fins. In this account of the Carangidce we have closely followed the nomenclature of Dr. Wakiya in his paper entitled “The Carangoid Fishes of Japan,” recently issued in the Annals of the Carnegie Museum, Vol. XV, pp. 139-292. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 225 Family RACHYCENTRID^. 242. [389] Rachycentron canadum (Linn^us). Suji. Osaka market (Jordan). Rare. This form, which is R. 'pondicerrianum of Cuvier and Valenciennes, requires comparison with Atlantic specimens of R. canadum. Family LEIOGNATHIDtE. 243. [390] Leiognathus nuchalis (Temminck and Schlegel). Hiiragi = Holly-tree. Mikawa Bay (M. Ishikawa); Tokyo and Shizuoka markets (Jordan); Toba market (Jordan and Yamamoto); Kochi (Wakiya); Fukuoka (Hamada); Fukui (Nonaka). Generally common. 244, [391] Leiognathus rivulata (Temminck and Schlegel). Okihiiragi = Offshore Hiiragi, Wakanoura (Yamamoto); Kobe market (Jordan); Tatoku Island (Mikimoto); Toyama (Yoshizawa); Misaki (Aoki); Noo. The greatest depth of the body is contained 2.8 to 3.2 times in the length to caudal fin. Family CORYPHTINID^. 245. [394] Coryph^na hippurus Linnaeus. >S/iMra = Dolphin, Tokyo market (Jordan); Misaki (Aoki); Fukui. Generally common. The specimen from Misaki is a young individual, in all essential respects like the type of Ectenias brunneus Jordan and Thompson.®^ The other specimens are half-grown individuals, and serve to connect these young specimens definitely with the adult of Coryphwna. Family BRAMIDTl. 246. [401] Taractes steindachneri (Doderlein). Banzai-uwo = Hurrah-fish. Argo steindachneri Doderlein, Denksch. Akad. Wiss. Wein,'XLVII, 1883, p. 242 (named as “n. sp., n. gen.” and well figured, but not described); ibid., XLIX, 1884, p. 174 (description and excellent figures). A fine specimen was secured in the Shizuoka market by Jordan. The species must be very rare in Japan. 83 Mem. Car. Mus., VI, 1914, p. 241, pi. XXVII, fig. 3. 226 MEMOIRS OF THE CARNEGIE MUSEUM. Family PAMPID^E. 247. [404] Pampus argenteus (Euphrasen). Managatsuwo. Yokohama and Kyoto markets (Jordan); Soo-chow, China (Dr. Cora B. Reeves) . Generally common southward. A valued food-fish. Family STROMATEID^. 248. [406] Psenopsis anomala (Temminck and Schlegel). Ebodai. Kagoshima Bay, Kochi (Wakiya) ; Tokyo market (Jordan) ; Mikawa Bay (M. Ishikawa); Misaki (Aoki); Miyazu, Sea of Japan; and Yamada. Seen in Ozaka, Kobe, and Yokohama. This small species is generally common southward in the markets, and is valued as a table-fish. Dorsal spines 6, graduated; vertebrae strong, about 12-15 in number; flesh firm, and much as in Poronotus or Palometa. Ventral fins well developed. Family CENTROLOPHIDiE. Genus Ocycrius Jordan and Hubbs, gen. nov. Type: Centrolophus japonicus Doderlein. This genus is closely related to Palinurichthys Bleeker, (perciformis) of the eastern coast of the United States, differing in the more elongate body, the longer and much more pointed lobes of the deeply forked caudal, and the lanceolate pectoral. It is also related to the Australian Hyperoglyphe, in which genus the dorsal spines are higher, the median highest, the scales smaller and the fins also pointed. Regan unites all these, with others, under the preoccupied name of Lirus Lowe, which must be replaced by Mupus Cocco. 249. [407] Ocycrius japonicus (Doderlein). Medai = Big-eye Porgy. (Plate IX; fig. 4.) Two specimens from the Tokyo market (Jordan), 29 and 41.5 cm. long to caudal fin. These specimens agree with Doderlein’s account in all respects, except those which might have been produced by the drying out or poor preservation of his type. Such a condition would account for the much larger size of the orbit, the presence of a keel on top of head, the concave contour on front of interorbital space, the greater prominence of the serrations on the margins of the opercular bones, etc., of the type, as compared with our material. JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 227 Dorsal rays, VII or VIII, 23 to 26; anal, III, 18. Dorsal spines, very short, subequal, largely concealed in a scaly thick sheath, not graduated into the soft rays. Head evenly gibbous, without a keel above. Eye (not orbit) much shorter than interorbital width, a little shorter than snout, not quite one-fourth length of head. Preopercle, subopercle, and interopercle with weak denticulations. Supple- mental maxillary small and indistinct. Gill-rakers long and strong, about 6+16; lateral line becoming straight at tip of pectoral; roof of mouth rounded without evident groove. Family PEMPHERIDiE. Genus Catalufa Snyder. Snyder has distinguished a genus Catalufa, with the new species, umbra, from Japan as the type, and containing also Pempheris compressus of Australia, the scales being all strongly ctenoid in these two species, instead of largely smooth, as in most Pempherids. But Pempheris 'compressus is the original type-species of Pempheris.. The species retained by Snyder in Pempheris have been placed by Ogilby (1913) in a new genus, Liopempheris. Catalufa and Pempheris may, however, be regarded as generically distinct, as the lateral line in the type of Catalufa is almost straight, whereas in Pempheris, as in Liopempheris, it is strongly curved. 250. [409 and 412] Catalufa japonica (Ddderlein). Agonashi. Pempheris japonicus Steindachner and Doderlein, Beit. Fische Japans, II, 1884, p. 29. Catalufa umbra Snyder, Proc. U. S. N. M., XL, 1911, p. 528; XLII, 1912, p. 412, pi. 52, fig. 3. Two specimens, 77 and 80 mm. to caudal fin, from Toba (Jordan and Yamamoto) . Dorsal rays VI, II; anal rays, III, 35 or 37; scales 14-76-29. Depth of body, 2.2; head, 3.35 in total length to base of caudal. Depth of caudal peduncle, 3.0 or 3.2 in length of head; eye, 2.35; snout, 5.5; upper jaw,_1.9 or 2.0; interorbital space, 3.3 or 3.4; scales on front half of interorbital space imbedded in the fleshy skin, but not absent as described by Doderlein; the “5-6 mehr oder minder stark entwickelte zahnahnliche Dornen’’ described by Doderlein as developed along the outer mandibular edge near the symphysis are merely a row of upturned spinules on the strongly ctenoid scales of the mandible; other head characters are as de- scribed by Doderlein and by Snyder. Origin of dorsal about twice as far from tip of middle caudal rays as from tip of mandible. Highest dorsal spine 1.4 in head; first soft ray (not the fourth as obviously misprinted in Snyder’s description) 228 MEMOIRS OF THE CARNEGIE MUSEUM. longest, 3.4 or 3.6 in total length to caudal, or 1.1 in length of head; base of dorsal, 4.6 or 5.2 in total length, equal to length of head to posterior margin of eye. Origin of anal equidistant between tip of mandible and base of caudal, or a little nearer the former; the base of the fin exceeding length of head by a distance somewhat greater than orbital length. Ventral inserted barely in advance of vertical from end of pectoral base. Second and longest anal ray a trifle longer than eye; margin of the fin weakly concave anteriorly. Ventral spine about two-thirds the total length of fin, which equals the dorsal base, being contained 4.6 or 5.2 times in total length; pectoral pointed, its length 1.2 in head, or 3.8 in total length. Scales all evenly disposed, of regular size, closely imbricate and strongly ctenoid (except on top of head anteriorly). Lateral line very slightly curved anteriorly, running nearly straight from upper end of gill-opening to end of middle caudal rays, not parallel with the dorsal contour posteriorly; least distance from origin of dorsal to lateral line equal to length of eye. Color deep brown, with about eleven dark vertical bars about as wide as the lighter interspaces, and very indefinite, except near middle of body. Lobes of dorsal, caudal, and anal broadly marked with black; ventrals blackish posteriorly, puncticulate; pectorals nearly clear. The excellent description by Steindachner and Doderlein is obviously based on an example of the present species. They err, however, in referring SchlegeFs Pempheris molucca, which is rather the species here called sasakii to the synonymy of japonicus. It is this fact which probably led Professor Snyder to rename the present species, and to apply the name japonicus to a species of Liopempheris. The latter we here provisionally identify as Liopempheris vanicolensis. 251. Liopempheris sasakii Jordan and Hubbs, sp. nov. (Plate X; fig. 1.) Pempheris molucca Temminck and Schlegel, Fauna Japonica, Pisces, 1844, p. 85, pi. 44, fig. 3 (not of Cuvier and Valenciennes). Type, a specimen 97 mm. long to the caudal fin, from Toba (Jordan and Yamamoto) C. M. Cat. Fishes, No. 7861. This species is not, as Doderlein and others have indicated, synonymous with Pempheris ( = Catalufa) japonicus, being referable rather to the genus Liopempheris Ogilby (1913). From other species of that genus it differs in the smaller size of the scales, about as numerous as in Catalufa japonica. Dorsal rays, VI, 9; anal. III, 41; scales, 8-74-23. Depth, 2.45 in length; head, 3.6; eye, 2.6; interorbital width, 4.5; width of head, 1.9; snout, 4.35. Inter- orbital space, snout, and preorbital region entirely scaled, except immediately JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 229 about eye and nostrils. Lateral line rather strongly curved throughout its length, the highest point of the curve about as distant from the chord of the curve as the lateral line is from the origin of dorsal; lateral line everywhere completely parallel- ing the dorsal contour. Scales deciduous, cycloid, or very weakly ctenoid where remaining, their exposed faces deeper than long. Origin of dorsal about 2.45 times as distant from tip of caudal as from tip of mandible; sixth dorsal spine longer than snout and eye, contained 1.4 times in head; first and longest soft ray slightly longer than postorbital; base of dorsal a little shorter than distance from tip of mandible to posterior orbital margin; base of anal equal to distance from anal origin to front of head, exceeding length of head by a distance greater than twice diameter of eye; origin of anal about below middle of dorsal base; second and longest anal ray slightly longer than eye; insertion of ventral fin slightly in advance of lower posterior end of pectoral base; ventral spine as long as eye, the soft rays 1.4 times longer, just reaching anal origin; length of pectoral, 1.1 in head. Color olive-brown, becoming darker on tip of snout, in front of preopercular ridge, on opercle along subopercular margin, on abdomen behind line joining pectoral and ventral bases, just within the anal base, on internal base of pectoral, and about the lateral line near its origin; silvery with black specks below the eye and below and before the pectoral and ventral bases; lateral line an unpigmented streak. Dorsal broadly blackish along front margin and at tip, otherwise pale; caudal light dusky, blackening toward free edge; anal black near front, but through- out most of its length whitish proximally, abruptly blackish on distal half; paired fins clear, except for black punctations at their bases. This species is named for Madoka Sasaki, Professor of Marine Zoology at the Imperial University of the Hokkaido at Sapporo. Liopempheris nyctereutes (Jordan and Evermann) from Formosa seems closely allied to L. sasakii. 252. [Extraterr.] Liopempheris vanicolensis Cuvier and Valenciennes. Pempheris japonicus Snyder, Proc. U. S. N. M., XLII, 1922, p. 497 (Okinawa) (not of Doderlein). The specimens of Pempheris recorded as P. japonicus from Okinawa (Ryukyu) by Snyder, were certainly incorrectly determined. The original P. japonicus is the species subsequently described by Snyder as Catalufa umbra. The Okinawa specimens, which we have re-examined, are close to P. vanicolensis, and are probably identical with that species, of which we have three specimens from the Philippines and Samoa. These have 43 to 45 anal soft rays (Gunther counted 38 to 40, Bleeker 39 to 45); the two from Okinawa have only 36 rays. They have about 9 or 11-49 230 MEMOIRS OF THE CARNEGIE MUSEUM. or 51-12 scales, but the scale-count is here of little value, except along the lateral line, because the normally very large scales are replaced in irregular and inconstant patches by very much smaller scales; thus one specimen has three, another eight rows between the lateral line and the median dorsal row on the caudal peduncle, and the number varies on the two sides of the same fish; sometimes as many as six or eight small scales will replace a single large one on the lower side. There is no spot on the base of the pectoral fin, but the dorsal and anal fins were probably originally black; the fin-membrane- is here lost, however, and the rays more or less broken, so that the fins appear nearly clear. In general appearance these specimens from Okinawa are like those of L. vanicolensis. Family APOGONID^. 253. [413] Apogonichthys carinatus (Cuvier and Valenciennes). Mato-ishimochi = Target Rice-fish. Wakanoura (Yamamoto); Toba (Jordan and Yamamoto); Kochi (Wakiya); Misaki (Aoki). Young, 5 to 6 cm. long, have the body crossed mesially by about eight dark shades of varying width, somewhat expanded toward their truncated upper and lower ends. Specimens 7 to 8 cm. long still show these bars, but less distinctly. In the adults they have vanished completely. 254. [414] Apogon lineatus Temminck and Schlegel. Moc/w'Rico = Rice-fish. Wakanoura (Yamamoto); Toba (Jordan and Yamamoto); Kobe market (Jordan); Mikawa Bay (M. Ishikawa); Misaki (Aoki); Toyama, Fukui, and Noo, all on the Sea of Japan. Dr. Cora B. Reeves has sent specimens from Soo-chow, China, to the University of Michigan. Everywhere common southward. A little shore-fish. 255. [415] Apogon niger Doderlein. = Black-rock Rice-fish. We have one example of this species, collected by Aoki at Misaki. It shows no definite markings on the body. 256. [420] Apogon semilineatus Temminck and Schlegel. Wakanoura (Yamamoto); Misaki (Aoki). Coloration typical. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922, 231 257. [422] Apogon kiensis Jordan and Snyder. A specimen from Misaki (Aoki) agrees well with the types. The upper band is continued backward faintly to the caudal fin, and there is an indistinct dusky line both above and below the main black band, and finally another, somewhat wider, running straight backward from the upper margin of eye to below the interval between dorsal fins. Traces of these less distinct lines remain in the types. 258. [425] Synagrops japonica (Doderlein). Sumikui-uu)o = lvi]^-i\sh.. A single specimen was obtained by Aoki at Misaki. Family SCOMBROPIDA5. 259. [426 and 427] Scombrops boops (Houttuyn). Mutsu. ? Telescopias gilberti Jordan and Snyder, Proc. U. S. N. M., XXIII, 1901, p. 909, pi. 45. Yokohama, Shizuoka, Osaka, and Tokyo markets (Jordan) ; Toba (Jordan and Yamamoto); Misaki (Aoki). The small red form (boops) is generally common; the large black form {gilberti) is distinctly rare. The only sharply distinctive feature accredited to Telescopias gilberti is the number of anal spines, two, instead of three, as in Scombrops boops. But the type specimen has three anal spines, the first indeed very small, but evident. Differ- ences in proportions and in scale-counts do not seem to hold. The color is pale dull red in the ordinary Mutsu {boops) and perhaps changes to black with age {gilberti) . Depth of body 3.6 to 4.3; orbit, 3.0 to 3.5 (slightly decreasing in relative size with age); depth of caudal peduncle, 3.2 to 3.6. Scales 51 to 65 along lateral line, 17 to 21 in transverse series; developed gill-rakers, 1 to 4 and 11 to 14. We have a young example from Misaki; a series of specimens corresponding with J. boops from the same place and from Yokohama, Tokyo, and Toba markets; one intermediate in size and color between boops and ‘‘gilberti,” from the Shizuoka market, and a large adult, like the type of gilberti, from the Yokohama market, where several others were seen. In spite of this evidence, we are not yet fully convinced that the rare gilberti is the adult of the common boops. Family ACROPOMIDTl. 260. [430] Acropoma japonicum Gunther. Dabo-mutsu. Two specimens were obtained by Yamamoto at Wakanoura. 232 MEMOIRS OF THE CARNEGIE MUSEUM. Family KUHLIIDtE. 261. [Extraterr.] Kuhlia rupestris (Lacepede). Mikyu. We have a specimen of this species from Chatan, Ryukyu, collected by H. Kuroiwa. Dorsal, X, 11; anal. III, 10; scales, 5.5-39-8.5; gill-rakers, 7 and 18. Depth, 2.7 ; longest dorsal spine a little shorter than longest soft ray, and not quite one- half length of head. Family PRIACANTHID.E. 262. [433] Priacanthus macracanthus Cuvier and Valenciennes. Beni-mebaru = Red Pop-eye ; Kinme = Gold-eye. Osaka market (Jordan); Mikawa Bay (M. Ishikawa); Fukui (Nonaka); Miyazu. Six specimens in all. Dorsal rays, X, 12 (one specimen) or 13 (five specimens); anal rays. III, 14 (all cases) . Depth of body decreasing with age, but greater than length of head in all our specimens (depth 2.6 in a specimen 117 mm. long to caudal; 2.8 to 2.9 in four 130 to 142 mm. long; 3.3 in adult). Preopercular spine reaching to gill- opening in one young, farther in all other specimens. Both dorsals and ventral blotched with dark, especially in the young; anal largely clear. 263. [434] Priacanthus meeki Jenkins. Hoseki- Kinme = Gem Gold-eye. Tatoku Island (Mikimoto); Misaki (Aoki). Two young specimens, 94 and 99 mm. Dorsal rays X, 13 or 14; anal, III, 15 (both cases). Depth of body greater than length of head, 2.6 in standard length. Preopercular spine not reaching gill-opening. Vertical fins dusky, becoming black anteriorly; ventrals largely black with some lighter markings. This species seems identical with Priacanthus meeki, the common form in Hawaii. In P. hamruhr of the Red Sea and East Indies the body is said to be more elongate, the depth 3. to 3.5 in length. 264. [435] Priacanthus japonicus Cuvier and Valenciennes. Chikamo- Kinme = Near-sighted Gold-eye. Tokyo and Yokohama markets (Jordan); Kochi (Wakiya); Toyama (Yos- hizawa); Miyazu, Misaki (Aoki). JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 233 265. [436] Pseudopriacanthus niphonius (Cuvier and Valenciennes). Kuruma-dai = WheeVporgy. Tokyo market (Jordan); Kochi (Wakiya); Misaki (Aoki). One of the Misaki specimens is a young one, matching Schlegel’s figure (pi. 7a, fig. 2). Family OLIGORID^. 266. [437] Malakichthys griseus Doderlein. = Big-eyed Bass. We have nine specimens of this peculiar bass-like fish; eight collected by Mr. Ybjiro Wakiya at Kagoshima; and one by Dr. Jordan in the Shizuoka fish-market. There seems to be no reason to doubt the identity of Satsuma macrops Smith and Pope with this species. 267. [437A] Malakichthys wakiyae Jordan and Hubbs, sp. nov. (Plate X; fig. 2.) Type a specimen 119 mm. long to the caudal fin, collected by Wakiya in Kagoshima Bay; C. M. Cat. Fishes, No. 7863. Six paratypes were obtained at the same place. This sharply marked species differs from M. griseus, hitherto the only one known, in many ways. The anal fin has 9, or rarely 10, soft rays, instead of only 7, the base of this fin is about half longer, instead of shorter than, the length of the anal spine. All of the fins are smaller; the soft portion of both dorsal and anal fins are longer, instead of shorter than high; the pectoral reaches not quite so far as, or barely to, the vertical from front of anal (in one specimen to above second anal spine), instead of to above first soft ray of anal (in one specimen only to above second spine); the ventral does not nearly reach to the vent, only one- half to two-thirds the distance to the origin of the anal, instead of nearly to the vent, and more than two-thirds the distance to the anal. The head is shorter and much narrower; the eye smaller; head 2.6 to 2.75 (instead of 2.35 to 2.6), much less than, instead of about half as wide as long; eye shorter than, instead of as long as, postorbital, 2.8 to 3.0 (not 2.5 to 2.7) in head. The vomerine teeth form a distinctly Y-shaped figure, instead of a triangle with moderately concave sides. The scales are slightly smaller. Dorsal rays, IX, I, 10; anal. III, 9 (10 in one specimen); pectoral, 13; ventral, I, 5; caudal, 17 (15 branched); scales, 5 to 6-47 to 49-13 or 14 (5 or 6-42 to 45-12 to 14 in M. griseus) ; gill-rakers 22 to 24 below angle on first arch; branchiostegals, 7. Body rather angular, deepest below 234 MEMOIRS OF THE CARNEGIE MUSEUM. origin of dorsal (depth 2.6 to 2.7). Dorsal contour before this point gently sigmoid, being concave above eye behind front of dorsal, gently curved to caudal peduncle; ventral contour flattish medially, but ascending rather abruptly on mandible and along anal base. Head, as viewed from above, much thinner and less wedge- shaped than in M. griseus. Interorbital slightly convex, cavernous, about two- thirds as wide as eye (scarcely more than half eye in M. griseus)] border of orbit and preopercular ridge sharp (in preserved specimens) ; preopercular margin den- ticulate on lower border and about the produced rounded angle; opercle ending in two sharp points, the lower the longer; opercular membrane wide, extending well behind pectoral base. Mandible strongly projecting, with two sharp spines at tiji; maxillary with a long narrow supplementary bone above a strong sharp ridge, which fits against the edge of the narrow suborbital. Teeth minute, granular, except in outer row of jaw, in rather wide bands in jaws, in a Y-shaped group on the vomer (the base of the Y pointing forward), and in a very narrow band running into a single series posteriorly on the sharp elevated ridge of the palatine; none on the tongue. Scales rather firm, about twice as deep as long, the basal margin straight, scalloped between the ends of the strong radii; focus near apical margin; circuli only moderately angulated between dorsal and lateral fields, running into the lateral margin at an angle of about 45 degrees; no circuli, radii, or ctenii on the narrow apical field; the apical margin sharply denticulate. Head almost com- pletely scaled. Body light olive-brown above, becoming darker on the snout; the sides and lower surfaces bright silvery. Fins with some dusky color, the spinous dorsal becoming black at margin; a black spot on axil of pectoral, more distinct than in t M. griseus. 268. [439] Lateolabrax japonicus (Cuvier and Valenciennes). Suzuki = Sapporo market (Majima); Tokyo and Kobe markets (Jordan); Kagoshima Bay (Wakiya); Mikawa Bay (Ishikawa); Miyazu, Fukuoka (Hamada); Lake Kasumigaura (Hattori). A very common and valued food-fish throughout Japan. Family NIPHONIDTl. 269. [440] Niphon spinosus Cuvier and Valenciennes. Ana = Sea-bass. Tokyo market (Jordan) ; Toyama (Yoshizawa) ; Misaki (Aoki) ; Miyazu, Fukui. An uncommon species, reaching a considerable size. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 235 Family EPINEPHELIDtE. 270. [441] Bryttosus kawamebari (Temminck and Schlegel). Kawa-mebaru = 'R\YQV Pop-eye or “Rock-cod.’’ Tsuyama, near Okayama (Kawamura); Kumamoto (Wakiya); Himeji (Abe); Fukuoka (Hamada). The preopercle, subopercle, interopercle, and the large bony scapular scale, all have their margins denticulate, but covered with skin. We count only 37 scales in the lateral line to caudal base. It is a river-fish, bearing a strong resemblance to the American Centrarchidce or Sun-fishes. 271. [Extraterr.j Siniperca chua-tsi Basilewsky. A half-grown specimen of this common River-bass of China has been sent by Dr. Cora B. Reeves to the Museum of Zoology, University of Michigan. It agrees well with Boulenger’s description. 272. [Extraterr.j Coreoperca herzi Herzenstein. Wakiya has sent a specimen of this River-bass from the Ping-yang River in Korea. Jordan and Metz erroneously indicated Coreoperca whiteheadi Boulenger®^ as a synonym of C. herzi. Dorsal rays, XIV, 13; anal, III, 8. 273. [450] Epinephelus craspedurus Jordan and Richardson. A young specimen taken in the Toba market (Jordan and Yamamoto) agrees well with the description of the type, except that the caudal is truncate, not slightly emarginate. Otherwise known only from Kagoshima. 274. [455] Epinephelus epistictus (Temminck and Schlegel). Komon-hata = Backward Bass. One young: Wakanoura (Yamamoto). 275. [456] Epinephelus morrhua (Cuvier and Valenciennes). lyagobata. Toba (Jordan and Yamamoto). We follow Boulenger in our identification of this species, and consider E. doderleini Franz (1910) as probably not a distinct species; we have, however, no material of that type. Proc. Zool. Soc. London, 1889, p. 960, pi. 68. Boulenger here gives a good re-description of C. herzi. Herzenstein, Ann. Mus. Zool. Acad. Imp. Sci. St. Petersb., 1896, p. 11. 236 MEMOIRS OF THE CARNEGIE MUSEUM. In the adult the bands of the young become replaced by double rows of spots. The disruption of the bands takes place in a regular manner and sequence. The pigment becomes concentrated toward the edges of the bands, so as to form two streaks, which later break up into lines of spots. The lower posterior part of the body is first affected, the progress of disruption being toward the nape, where the bars still remain intact in our adult specimen. 276. [458] Epinephelus moara (Temminck and Schlegel). Moara = Sea- weed Bass. Toba market (Jordan and Yamamoto); Mikawa (M. Ishikawa). 277. [459] Epinephelus awoara (Temminck and Schlegel). d.oam = Green Bass. Mikawa Bay, Miyazu, Fukui. 278. [460] Epinephelus septemfasciatus (Thunberg) . Ma/iato = True Hata, or bass. Tokyo market (Jordan); Toba (Jordan and Yamamoto); Mikawa Bay (M. Ishikawa); Fukui (Nonaka); Misaki (Aoki); Miyazu. 279. [461] Epinephelus tsirimenara (Temminck and Schlegel). Afca/iato = Red Bass. Toba (Jordan and Yamamoto). 280. [476] Doderleinia berycoides (Hilgendorf). Aka-mutsu ^ Ked Mutsu. Shizuoka (Jordan); Toyama (S. Yoshizawa); Miyazu; Noo near Niigata. Not common, its color a brilliant brick-red. Family SERRANIDA5.«« 281. [464] Chelidoperca hirundinacea (Cuvier and Valenciennes). Hime-ko-dai = Little Princess Porgy. Kagoshima, Kochi (Wakiya); Misaki (Aoki). 282. [465] Sayonara satsumae Jordan and Seale. Sakura-dai = Cherry Porgy. Kochi (Wakiya). 283. [466] Caprodon schlegeli (Gunther). Aka-isagi = Jied Isagi, or Croaker. Kyoto market (Jordan); Toba (Jordan and Yamamoto). Eye golden-yellow; body mostly scarlet. ^®The SerranoicI genus, Rhyacanthias Jordan (carlsmithi) (Proc. U. S. N. M., 1925 lately described from Hawaii, seems inseparable from Symphyanodon Bleeker {typus), Arch. Neerl. XII, 61, 1878. The minute teeth on the palate were overlooked by Bleeker, who placed the genus near Pomadasys. JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 237 Family LUTIANIDJi:. 284. [481] Lutianus russelli (Bleeker) . = Black-star Porgy. Kochi, Shikoku (Wakiya). 285. [483] Lutianus vaigensis (Quoy and Gaimard). Okifuyedai = Off-shore Air-bladder Porgy. Kagoshima Bay (Wakiya). 286. [484] Lutianus vitta (Quoy and Gaimard). Kinseisaki — Symmetrical Isaki. Mikawa Bay (M. Ishikawa); Toyama (Yoshizawa); Fukui (Nonaka); Miyazu. Not rare. 287. [487] Ulaula sieboldii (Bleeker). Yoko-suji-fuyedai = Jled Air-bladder Porgy. A fine specimen 23 cm. long to the caudal fin, was taken at Toba. (Jordan and Yamamoto). Head, 3.2; depth, 3.5; eye, 5.8; dorsal, X, 11; anal, III, 8; scales, 8.5-70-16; gill-rakers, 2.4 in eye; fifth dorsal spine longest, 2.35 in head; third anal spine, 3.5; pectoral fin, 1.07. The presence of a patch of teeth on the tongue and the slight development of canines warrant the recognition of Ulaula^^ as of generic rank. Family THERAPONIDTl. 288. [490] Therapon oxyrhynchus Temminck and Schlegel. Shima-isagi = Striped Croaker. Kachi River at Nagoya, with fresh-water fishes (Jordan) ; Toba market (Jordan and Yamamoto); Mikawa Bay (Ishikawa); Miyazu, Lake Mikata near Fukui. Common. The young of this species live in the bays and estuaries, ascending streams. These, when about 5 or 6 cm. long, differ widely from the adult, described by Jordan and Thompson.®^ The head is longer (2.7 to 2.8 in length to caudal), and the depth of the body greater (2.7 to 2.9). The coloration is markedly different from that of the adult, which is well shown in SchlegeFs plate (pi. 6, fig. 3). The secondary bands are ®®Jordan and Thompson, Proc. U. S. N. M., XXXIX, 1911, p. 460; Jordan and Jordan, Mem. Car. Mus., X, 1922, pp. 49, 50. Proc. U. S. N. M., XLI, 1912, p. 538. 238 MEMOIRS OF THE CARNEGIE MUSEUM. not developed, only the primary four being evident. A tendency toward a vertical, as well as a longitudinal, pattern is brought out by a slight intensification of the ground-color in line with dilated and especially darkened portions of the stripes. The spinous dorsal is marked with blackish mottlings on ,base and with large distal blotches on the membranes. The soft dorsal and caudal fins are longitudinally streaked with black, but pale-bordered. Similar specimens were recorded from Swatow, China, by Rutter.** Other material from Japan, both in the present collection and in the Stanford Museum, fully connects these young with the adult. They show little approach toward the very small fish described and figured by Franz*® as the young of this species. 289. [491] Therapon servus (Bloch). Yagati-isagi. Kachi River at Nagoya, Kobe market (Jordan); Mikawa Bay (M. Ishikawa). The young of this fish, as those of the last, enter fresh water and associate with Brook-gobies and Minnows. In the young, 5 to 10 cm. long, the form and color are about as in the adult, but the upper band is not broken longitudinally, but, on the contrary, twice vertically by the lighter ground-color. Family BANJOSIDAH. 290. [492] Banjos banjos (Richardson). = Hurrah-porgy. Misaki (Aoki). A rare and curious fish. Family POMADASIDiE. 291. [493] Parapristipoma trilineatum (Thunberg). I sagi, or I saki. Mikawa Bay (M. Ishikawa); Toyama (S. Yoshizawa); Tokyo and Kobe markets (Jordan); Wakanoura (Yamamoto). Common southward. 292. [494] Plectorhynchus pictus (Thunberg). Korodaz = Elder Porgy. Wakanoura (Yamamoto); Kobe market (Jordan). 293. [495] Plectorhynchus cinctus (Temminck and Schlegel). Kosho-dai. Wakanoura (Yamamoto); Toba (Jordan and Yamamoto); Toyama (S. Yoshizawa); Fukui (Nonaka); Misaki (Aoki). Dorsal rays, XII, 16 to 18; anal. III, 8. ** Proc. Acad. Nat. Sci. Phila., 1897, p. 75. *® Abh. Bayer. Akad. Wiss., Vol. I, Suppl. 4, 1910, p. 46, pi. 5, fig. 31. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 239 294. [496] Hapalogenys mucronatus (Eydoux and Souleyet). Higedai = Bearded Porgy. Osaka market (Jordan). 295. [497] Hapalogenys nigripinnis (Temminck and Schlegel). Letodai — Channel porgy. Toba (Jordan and Yamamoto); Kochi (Wakiya); Soo-chow, China (sent by Dr. Cora B. Reeves to University of Michigan). The specimen from Kochi, corresponds with the descriptions of H. kishinouyei. which seems to be the young of H. nigripinnis. 296. [498] Scolopsis inermis (Temminck and Schlegel). Tamagashira. Toba (Jordan and Yamamoto). Family C^SIONIDTl. Genus C^sio Lacepede. 297. Caesio lunaris Ehrenberg. Three specimens of this species, taken by Wakiya at Kagoshima, constitute the first record of the species from Japan proper. 298. Caesio cserulaureus (Lacepede). Hist. Nat. Poiss. Ill, 1802, p. 85. A specimen from Kagoshima (Wakiya) is the first to be recorded from Japan. 299, Caesio chrysozonus Kuhl and Van Hasselt. Twelve examples of this species, which was originally described from the Moluccas, were taken by Wakiya at Kagoshima. The two largest specimens are especially deep, the greatest depth being contained thirty-five times in length to caudal fin. All three of these handsome tropical fishes correspond v^ell with current descriptions. Family LABRACOGLOSSIDvE. It is obvious that Labracoglossa has no affinity with the stromateid fishes, with which it has been aligned. It has lately been made the type of a distinct family, with which certain New Zealand genera have been associated. It seems to us that Labracoglossa is as near to Ccesio as to any other genus. 300. [408] Labracoglossa argentiventris Peters. Takabe. Ten specimens, Misaki (Aoki). Jordan, Stanford Univ. Publ. (Biol.) 3, 1923, p. 203. 240 MEMOIRS OF THE CARNEGIE MUSEUM. Family DENTICID^. 301. [499] Euthyopteroma virgatum (Houttuyn). /togforo = Thread; Adventurer. Osaka and Kobe markets (Jordan); Mikawa Bay, Fukui, Wakanoura, Miyazu. A beautifully colored fish, generally common southward. 302. [500] Euthyopteroma bathybium (Snyder). Soko-itoyori. Kagoshima Bay (Wakiya). Two specimens, one young, with a very large eye (3 in head). Family SPARIDtE. 303. [503] Sparus latus Houttuyn. Kriro-daf = Black Porgy. Bay of Mikawa (M. Ishikawa). Scales 5-45-11; depth of body, 2.55; length of longest spine 1.95 in head; second anal spine greatly strengthened and elongate, 1.6 in head; membranes of lower fins largely black medially; streaks on scale-rows obscure; vertical bars obsolete. We identify our specimens of this genus with considerable doubt. The char- acters given®^ to separate S. latus from S. macrocephalus (S. swinhonis) do not hold well. 304. [504] Sparus macrocephalus Basilewsky. Morokoshidai = Mmnowlovd-taii Chrysophrys swinhonis Gunther, Ann. Mag. Nat. Hist. (4) XIII, 1874, p. 155. Tokyo and Shizuoka markets (Jordan) ; Toba market (Jordan and Yamamoto) ; Mikawa Bay (M. Ishikawa); Toyama (S. Yoshizawa); Misaki (Aoki); Miyazu, Fukui. Generally common southward. As a food-fish inferior to the Red Tai ( Pagrosomus) . Scales 6 to 9-48 to 56-14 to 16. Depth of body, 2.4 to 2.6; highest dorsal spine (third or fourth) 2.1 to 2.5 in head; second anal spine, when depressed, reaching to about opposite the middle of the last anal ray, and contained 1.95 to 2.15 times in the head. A dark spot at origin of lateral line; well marked olive streaks along the rows of scales, including those on cheeks; body with several rather narrow cross-bars; lower fins blackish. Dorsal rays, XI, 10 or 11; anal, III, 8. 305. [505] Lethrinus haematopterus Temminck and Schlegel. Fuefuki-dai. Kagoshima Bay (Wakiya); Miyazu, Sea of Japan. See Jordan and Thompson, Proc. U. S. N. M., XLI, 1912, pp. 581-588. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922, 241 306. [506] Lethrinus nematacanthus Bleeker. Toba market (Jordan and Yamamoto). 307. [508] Pagrosomus major (Temminck and Schlegel). Tai, Akadai = ^ed Tai, or Porgy. Kobe market (Jordan); Mikawa Bay (W. Ishikawa); Misaki (Aoki); Miyazu. Common throughout Japan; the most valued food-fish of the Empire, favored by the Fish-God, Ebisu, In this species the four scale-rows adjoining the lateral line above run parallel with it, whereas the rows between these four and the dorsal fin extend obliquely upward and backward. In the other Japanese sparids all of the scale-rows above the lateral line are parallel with it. In life the Red Tai has over the eye a spot of the most intense blue, which fades at death. The round light blue spots on the body fade with age and with exposure. 308. [509] Taius tumifrons (Temminck and Schlegel). Golden Tai. Tokyo and Kobe markets (Jordan) ; Kagoshima Bay (Wakiya) ; Misaki (Aoki) ; Toyama (S. Yoshizawa); Miyazu, Noo. In life this species is pale red, largely washed with golden; whole front and snout yellow; upper lip mostly yellow; three large diffuse round yellow spots on side of back, fading in death. It is now taken in great numbers in trawl-nets in rather deep water. 309. [510] Evynnis cardinalis (Temminck and Schlegel). = Little Tai. Tokyo market (Jordan); Toba market (Jordan and Yamamoto); Mikawa Bay (M. Ishikawa); Toyama (S. Yoshizawa); Miyazu. Our specimens vary widely in the form of the body, the degree of elongation of the dorsal spines, and the development of vomerine teeth, but we are unable to refer them to more than one species. For this reason we doubt the validity of Parargyrops edita Tanaka.®^ The species is unique among Sparoid fishes in having a few large bluntish teeth on the vomer. It is common southward and valued as food. Family GIRELLID^E. 310. [511] Girella punctata Gray. Mezina. t Toba market (Jordan and Yamamoto); Misaki (Aoki). Dorsal, XV, 13 or 14; anal. III, 12; scales, 53; a dark spot on each scale, Fig» Desc. Fishes Jap.^ 24, 1916, p. 425, pL 116, fig. 342; pL 117, fig. 343. 92 242 MEMOIRS OF THE CARNEGIE MUSEUM. Family KYPHOSID^. 311. [514] Kyphosus cinerascens (Forskal). T'enj^/cR-^sa^/^ = Indian Croaker; Kuroisagi = Black Croaker. Family GERRID^. 312. [516] Gerres erythrourus (Bloch). Kagoshima Bay (Wakiya). As the original type of Gerres Cuvier, as published in advance by Quoy and Gaimard, was a Xijstoema, the name Gerres must be retained here. Family SCIAENID.E. Genus Nibea Jordan and Thompson. We now regard Scicena aquila (Lacepede) as the legitimate type of the genus Scicena Linnseus, as this species was the original form included in the confused synonymy of Scicena umbra” Linnaeus, as first shown by Cuvier. The name Scicena accordingly is restored to the Mediterranean species, called Pseudoscicena by Bleeker and earlier Argyrosomus by De la Pylaie. Scicena umbra (L.) Cuvier is characterized by the obsolescence of preopercular spines, and of the slits and pores above the mouth, conspicuous in most of the species of this group; teeth in bands, the outer enlarged in both jaws. The characters of the genera allied to Scicena are singularly elusive, each one being subject to degrees of development. The form of the mouth, the dentition, the armature of the preopercle, the size of the second anal spine, the scaling of the fins indicate groups, which are more or less natural, but which are as yet imper- fectly separated. For the present we may refer the Japanese species to the genus Nibea Jordan and Thompson (type, Scicena mitsukurii). Nibea has the teeth in narrow bands, the outer row above and the inner below somewhat enlarged; gill- rakers moderate; preopercle with some stiff serrse, usually soft; dorsal fin scaleless, with a sheath at base; dorsal fins somewhat connected. The related genus Bola Buchanan (type Scicena coitor) seems to differ mainly in having the soft dorsal covered with small scales, the preopercle usually without liony serrse; gill-rakers relatively short. Johnius Bloch (type, J . carutta) has the lower teeth in a villiform band, none of them enlarged; soft dorsal fin scaly; anal spine small. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 243 313. [518] Nib ea mitsukurii (Jordan and Snyder). = Big Mouth, Toba (Jordan and Yamamoto); Shizuoka (Jordan); Choshi (Ishikawa), Very common in Suruga Bay, where it is much used in making Kamohoku. Dorsal rays, IX or X-I, 29 to 31; anal, II, 7, (rarely III, 8); developed gill- rakers on lower limb, 12 to 15; rows of spots continuous above lateral line, becoming straight below middle of soft dorsal fin. 314. Nibea albiflora'*'* (Richardson). On comparison of material we find that the characters supposed to distinguish Nibea mitsukurii from the Chinese Nibea albiflora hold. Of the latter we have one specimen from Fukuoka (Hamada). This province is on the west coast of Kyusyu, while the known localities for mitsukurii are all on the east coast of Hondo, the species being especially abundant about Shisuoka. Dorsal rays, X-I, 28; anal, II, 7; only 10 well developed gill-rakers on lower limb; rows of spots not continuous above lateral line, becoming straight only behind dorsal fin and in a strip along lateral line extending forward to below middle of soft dorsal fin. 315. [520] Nibea schlegeli (Bleeker). Ishimochi = Rock Rice-fish. Shizuoka, Osaka, Tokyo and, Kobe markets (Jordan); Mikawa Bay (M. Ishikawa); Misaki (Aoki); Fukuoka (Hamada); Fukui, Shizuoka (Jordan); Miyazu. Common southward, largely used for Kamohoku. Dorsal rays, X-I, 24 to 27. Anal spine sometimes a little longer than the eye. Black blotch on opercle and shoulder usually distinct. 316. [521] Nibea nibe (Jordan and Thompson). Hama-omFe = Great Beach-Nibe. Osaka market (Jordan); Mikawa Bay (M. Ishikawa). The softness of the flesh and the protrusion of the stomach into the mouth in one of our Osaka specimens indicates that this species descends to considerable depths. Dorsal rays, X-I, 30 to 32; anal, II, 7 or 8. The preopercular spines are moderately enlarged at angle. The scale-rows are marked by dark streaks, less definite, but arranged much as in Nibea albiflora, the streaks becoming horizontal only behind the dorsal base. The second anal spine may be nearly as long as the eye. See Jordan and Thompson, Proc. U. S. N. M., XXXIX, 1911, pp. 246-252, figs. 1-2. 244 MEMOIRS OF THE CARNEGIE MUSEUM. 317. [522] Nibea japonica (Temminck and Schlegel). Nihe] 0m5e = Great Nibe. Tokyo and Shizuoka markets (Jordan), This species, described in detail by Jordan and Metz from Fusan in Korea, reaches a much larger size than the others, three to four feet in length. It is also distinguished by the loose scales and generally ragged appearance, as compared with the neatness of the others. Teeth irregular, mostly in one row, some in each jaw irregularly enlarged; no true canines. Small specimens are long and slender, suggesting Cynoscion in form and aspect, the scales more even than in the adult. Caudal long, on a long peduncle, irregularly lanceolate, the lower lobe much longer; pectoral fin long. General color gray; mouth orange within; dorsal, caudal, and pectoral black-edged. In this species the large air-bladder is used in making the jelly-isinglass, known as Nibe. Genus Othonias Jordan and Thompson. This group is technically close to Bola, having, as in Bola, the soft dorsal scaly, but it differs in several important characters. It definitely approaches Collichthys in the large size and obliquity of the mouth, the dentition, the develop- ment of sensory cavities, and in the width and convexity of the head. It also resembles that genus in the presence of silvery glandular organs in the skin of the ventral surface, one under each scale; these are possibly light-organs (photophores). The soft dorsal, anal, and caudal fins are closely covered with small scales. 318. [522] Othonias undovittatus (Jordan and Seale). Pseudoscicena undovittata Jordan and Seale, Proc. Davenport Acad. Sci., X, 1905, p. 11, pi. 6 (Hong Kong). Scicena manchurica Jordan and Thompson, Proc. U. S. N. M., XXXIX, 1911, p. 255, fig. 3 (Port Arthur).- — Jordan and Metz, Ann. Car. Mus., VI, 1913, p. 38, fig. 28 (after Jordan and Thompson). We have compared the type of Pseudoscicena undovittata, paratypes of Scicena manchurica, and a specimen obtained in the Osaka market (Jordan), and find tliem all specifically identical. The type of S. undovittata has nine soft rays, in the anal fin, not eight, as described. The species is new to Japan. The Osaka specimen has IX-I-33 dorsal rays; II, 10 anal rays; 9-|-16 gill-rakers. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 245 Family EMMELICHTHYID^. 319. [523] Erythrocles schlegeli (Richardson). = Red Coat. Toba market (Jordan and Yamamoto). This species should be accredited to Richardson, not to Bleeker. The name Erythrichthys is preoccupied in the Characinidce, Family OPLEGNATHID^. 320. [524] Oplegnathus fasciata (Temminck and Schlegel). Ishidai = 'Rock Tai. Toba (Jordan and Yamamoto); Tatoku Island (Mikimoto); Mikawa Bay (M. Ishikawa); Toyama (Yoshizawa); Miyazu, Misaki (Aoki). Our material ranges from young to adult, and illustrates the changes in growth figured by Schlegel; the dark bands, so conspicuous in the young, fading with age. 321. [525] Oplegnathus punctata (Temminck and Schlegel). I shigaki-dai — Hock-imp Tai. Kobe market (Jordan) ; Mikawa Bay (Ishikawa) ; Misaki (Aoki) ; Toyama (Yoshizawa) ; Miyazu. Generally common southward, more so than the preceding. Family HISTIOPTERID^. 322. [527] Histiopterus typus Temminck and Schlegel. Kawahisha. Misaki (Aoki). A rare species. Family MULLID^. 323. [529] Upeneoides bensasi (Temminck and Schlegel). Beni-sashi = Red Surmullet. Shizuoka and Kobe markets (Jordan); Wakanoura (Yamamoto); Toba market (Jordan and Yamamoto); Tatoku Island (Mikimoto); Kagoshima (Wakiya); Mikawa Bay (Ishikawa); Toyama (Yoshizawa); Misaki (Aoki); Fukui (Nonaka); Noo, Miyazu. Generally common. 324. [530A] Upeneoides vittatus (Forskal). Two examples of this East Indian Surmullet were obtained by Wakiya at Kagoshima Bay, the first to be recorded from Japan proper. 325. [537] Upeneoides tragula (Richardson). Fome-/umeje = Bride Surmullet. Toba (Jordan and Yamamoto); Kagoshima (Wakiya). 246 MEMOIRS OF THE CARNEGIE MUSEUM. 326. [531] Mulloides japonicus (Houttuyn). Aka-himeji = Surmullet. Kagoshima (Wakiya). The main dorsal spines are seven in number, as described by earlier authors, and in all the material we have available, not eight, as reported by Snyder. An additional spine in front of these is minute when present, but often absent. Our specimens preserve much of their original color. The general tone is a bright yellow, becoming pink along the middle of the sides, above the large blackish opercular blotch, along the anterior free margin of the subopercle (which other- wise, like the interopercle, is silvery) and on the postorbital region. 327. [535] Upeneus pleurotenia Playfair, i/orai-urro = Bamboo-fish. Mullus pleurotcenia Playfair, Fishes of Zanzibar, 1866, p. 41, pi. 5, fig. 3. Pseudupeneus ischyrus Snyder, Proc. U. S. N. M., XXXII, 1907, p. 91, fig. 2. (Tokyo Bay). Upeneus ischyrus Jordan, Tanaka, and Snyder, Jour. Coll. Sci., Tokyo, XXXIII, 1913, p. 183, fig. 133. Upeneus pleurotcenia Snyder, Proc. U.S.N.M., XLII, 1912, p.501. (Naha, Okinawa). One specimen from Kagoshima Bay (Wakiya) corresponds well with the type of Pseudupeneus ischyrus Snyder, and with the specimen later recorded by the same author as Upeneus pleurotcenia. The type of U. ischyrus still retains definite traces of the peculiar coloration of this species, and the figure shows these markings in part. Head, 3.15; depth, 3.1; scales in lateral line, 29; dorsal rays VIIK9; anal, 7. Color in formaldehyde: upper parts very dark, particularly on the scale- margins; a light stripe from near nostrils to near end of soft dorsal fin, curved upward slightly, and interrupted by the upper rim of the eye, becoming obscure posteriorly; a nearly parallel stripe from middle of maxillary, skirting the lower margin of the orbit, extending toward the light saddle across front of caudal peduncle, strongly tinged with red; a third stripe extends from angle of mouth backward and upward, parallel with the upper two, but scarcely evident on body; between the light stripes the dark color is intensified, this being particularly true in a spot at upper end of preopercle; there is a short golden bar below this spot, and another on the interior base of the pectoral fin and a flush of red about these spots, and near the anal fin; the light blotch on the caudal peduncle is bounded posteriorly by a rather indefinite vertical dark bar. We have also examined material from Hong Kong, China, collected by Walter Fong. IToc. U. S. N. M., XXXII, 1907, p. 96. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 247 328. Upeneus indicus (Shaw). This species is represented by a specimen from Kagoshima Bay (Wakiya). Franz (1910) records it from the same place and from Yokohama, but it is not listed by Jordan, Tanaka, and Snyder. Family APLODACTYLID^. '329. [538] Goniistius zonatus (Cuvier and Valenciennes). TaA^ano/ia-daf = Hawk-Tai. Toba market (Jordan and Yamamoto); Misaki (Aoki). 330. [539] Goniistius zebra (Doderlein). Migimaki. A specimen of this rare fish from Misaki (Aoki) corresponds with this species, as contrasted with Goniistius zebroides Tanaka. Family CIRRHITIDJE. 331. [540] Cirrhitichthys aureus (Temminck and Schlegel). Oki-gonbe = Off-shoie Gonbe, or Cirritid. One specimen from Misaki (Aoki). It closely corresponds with Jordan and Herre’s account,®^ differing chiefly in having groups of cirri at the tips of the dorsal spines and on the posterior rim of the anterior nasal flap. The cirri are present on the dorsal spines in the specimen noted by Jordan and Herre, but we find no remaining trace of nasal cirri. The agreement of our specimens with SchlegeFs figure, however, is not close, and it is highly possible that the species at hand is still without a name. Family POLYNEMIDiE. 332. [543] Polynemus plebeius (Broussonet) . Agonashi; Tsuhame-Konoshiro — Swallow Gizzard-shad. Shizuoka (Jordan); Misaki (Aoki); Mikawa Bay (Ishikawa). .Our specimens agree well with the excellent figures of this species given by Broussonet and Gunther, with the distinctive features of plebeius as pointed out by Jordan and McGregor,®® with specimens of plebeius from Samoa, and finally with “cotypes” of Polydactylus agonasi Jordan and McGregor in all respects save color. The types of P. agonasi seem entirely faded, however, and we see no reason for regarding the Japanese Polynemus agonasi as different from plebeius. We use the specific name in its original form, plebeius. Proc. U. S. N. M., XXXIII, 1907, p. 161, fig. 1. Proc. U. S. N. M., XXX, 1906, p. 814, fig. 248 MEMOIRS OF THE CARNEGIE MUSEUM. Family SILLAGINID.E. 333. [544] Sillago japonica (Temminck and Schlegel). Ao-gisu = Green Gisu. Osaka, Tokyo, and Kobe markets (Jordan); Toba market (Jordan and Yama- moto); Mikawa Bay (Ishikawa); Toyama (Yoshizawa); Misaki (Aoki); Miyazu, Noo, Fukui (Nonaka). Very common; a choice food-fish, as such much like the English whiting. Scales 3 to 4-64 to 74; cheek-scales weakly to strongly ctenoid. The specimen from Fukui (Sea of Japan) is young, only 42 mm. long, and is included in a large series of small anchovies, Engraulis japonicus. 334. [546] Sillago parvisquamis Gill. Tokyo market (Jordan). Scales 6.5 to 8-79 to 81-12 to 13. Dorsal, VII-I, 22; anal, I, 23. Family BRANCHIOSTEGID^. Genus Branchiostegus Rafinesque. {Latilus Cuvier.) The name Branchiostegus offered in 1814 by Rafinesque as a substitute for Coryphoenoides Lacepede, (not of Gunner), must unfortunately replace Latilus. 335. [548] Branchiostegus japonicus (Houttuyn). Amadai = Girl Ten. ? Coryphcena branchiostega Gmelin. Latilus argentatus Cuvier and Valenciennes. Tokyo market (Jordan); Mikawa Bay (Ishikawa); Toyama; Shizuoka; Yoko- hama; Kobe; Misaki; Miyazu. A common and valued, but tasteless, food-fish, the soft flesh extremely white. We find the coloration and proportions to vary considerably in this genus, and are not able at present to recognize more than one species. This is light red in life, with shadings of blue or purple especially on the caudal, but in spirits it becomes more or less yellow, as described by Houttuyn. If more than one species exists in Japan, Houttuyn’s scant description is unidentifiable. Family CEPOLIDtE. 336. [555] Acanthocepola krusensterni (Temminck and Schlegel). Akatachi-uwo = 'Red Sword-fish. Kobe market (Jordan); Mikawa Bay (Ishikawa); Fukui (Nonaka), Miyazu. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 249 337. [557] Cepola schlegeli Bleeker. = Red Sword-fish. Dr. Ishikawa sent us one specimen of this species from, the Sea of Japan taken at Noo. Family EMBIOTOCIDiE. 338. [558] Ditrema temmincki Bleeker. Umi-tanago = Sea Chub. Tokyo market (Jordan); Toba (Jordan and Yamamoto); Misaki (Aoki); Fukui (Nonaka). Rather common southward. Dorsal rays, IX to XI, 20 or 21 ; anal. III, 24 to 28. The scales are not enlarged in the region between the pectoral and the ventral fins. Ditrema thus agrees with Tceniotoca, but differs from Embiotoca. In life bluish, with more or less bronze-green or copper-red checks and streaks on the body, one row to each scale-row; the streaks more or less expanded into spots on each scale. 339. [570] Neoditrema ransonnetii Steindachner. = Off-shore Chub. One specimen from the Tokyo market (Jordan). Dorsal rays, VI, 20; anal rays. III, 26. Family POMACENTRID.E. 340. [563] Chromis notatus (Temminck and Schlegel). Suzumedai = Sparrow-porgy. Wakanoura (Yamamoto). 341. [564] Abudefduf saxatilis (Linnaeus). Oya-hitsuchiya. Wakanoura (Yamamoto); Kobe market (Jordan). The specific name, saxatilis, seems to have been intended for this Asiatic species in the Tenth Edition of the Systema Naturce. In the Twelfth Edition it was transferred to the Brazilian form, Abudefduf marginatus. 342. [577] Pomacentrus coelestis Jordan and Starks. Wakanoura (Yamamoto). Very rare. Family LABRID.K. 343. [578] Choerodon azurio (Jordan and Snyder). Xandaf == Korean Tai. Osaka and Tokyo markets (Jordan). Not rare. Dorsal, XIII, 7; anal. III, 10. 250 MEMOIRS OF THE CARNEGIE MUSEUM. 344. [579] Semicossyphus reticulatus (Cuvier and Valenciennes). Kobudai = Little Soldier-tai. Toba (Jordan and Yamamoto). 345. [582] Pseudolabrus japonicus (Houttuyn). Sasanoha-bera = Bamhoo-hera or Wrasse. Yokohama, Tokyo, and Shizuoka markets (Jordan); Toba market (Jordan and Yamamoto); Misaki (Aoki). Common southward. 346. [584] Duymseria flagellifera (Cuvier and Valenciennes). Misaki (Aoki). 347. [588] Stethojulis terina Jordan and Snyder. Kaminari-bera = Thunder- wrasse. Shizuoka (Jordan); Misaki (Aoki). Jordan, Tanaka, and' Snyder (1913, p. 200) in their Catalogue identify this species with the East Indian S. kalosoma. Our formalin specimens, which ap- parently have preserved their fresh colors, correspond, however, with Jordan and Snyder’s color account of terina, and not with the descriptions of S. kalosoma. We retain the name terina for the Japanese form, provisionally regarding it as distinct from S. kalosoma. The species has posterior canines and is therefore a true Stethojulis, not a Hinalea. 348. [592] Halichoeres poecilepterus (Temminck and Schlegel). Ao5era = Blue Bera. Toba market (Jordan and Yamamoto) ; Tokyo market (Jordan) ; Toyama, Misaki (Aoki) ; Miyazu. Common, the two sexes quite different, as recognized by Jordan and Snyder, H. pyrrhogrammus (Temminck and Schlegel) being the female. We preserve the original spelling poecilepterus, of Schlegel. Those who regard Halichoeres as preoccupied by Halichoerus may call this genus Hemiulis; Parajulis and Chcerojulis being synonyms. 349. [595] Halichoeres tenuispinis (Gunther). Tokyo and Shizuoka markets (Jordan); Misaki (Aoki); Toyama. Scales 25 in the lateral series, or 27 along the lateral line, not counting 3 on the caudal. Jordan and Snyder,®^ in recording this species from Japan under the name Proc. U. S. N. M., XXIV, 1902, p. 639. 97 JORDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 251 Halichceres bleekeri (Steindachner and Doderlein), state: “It is possible that Halichceres tenuispinis (Giinther), with the black on the dorsal restricted to a single spot and the ventrals not filamentous and shorter than the pectorals, may prove to be the same species.” In our specimens the black on the dorsal extends backward to the third spine in some (as described for tenuispinis), or as far as the sixth spine; in one half-grown specimen, otherwise similar, the black spot is little developed; in another entirely absent. The ventral fin may be non-filamentous and shorter than the pectoral on the same fish, in which the opposite fin is slightly filamentous and longer than the pectoral. There remains no basis for the distinc- tion of the Japanese bleekeri from the Chinese tenuispinis. Further comparison of actual material, however, is to be desired. One specimen in formalin seems to retain the life-colors, which are as follows: ground-color pink; the scales with green borders, especially on upper posterior fourth of body; three green streaks diverging back from the eye form the anterior ends of horizontal rows of green spots, one on each scale, on the fore part of the body; these head-streaks are green, bordered narrowly with blue and then with deep red; the first extends slightly upward, and is the least definite, the lower one extends slightly downward toward the triangular deep blue spot on the upper edge of the base of the pectoral fin, but before reaching the opercular margin turns upward at right angles. The front portion of the spinous dorsal is banded from the base outward successively by green, red, yellow, and blue-black, then by red and finally green at tip of rays, the black marking being most extensive; elsewhere the fin is pink, bordered by greenish, and then very narrowly by dusky; between each ray, behind the black marking, there is a large round spot of green, red- bordered, between each two rays. The anal fin is pink, pale-margined, and marked by a median red-bordered band of green. The caudal is pinkish, with a basal and a median band and some distal spots of green. The pectoral and ventral fins are pale pinkish, the former fin having the rays margined by pencilled lines of dark red. 350. [599] Thalassoma cupido (Temminck and Schlegel). Nishiki-uwo = Brocade-fish. Misaki (Aoki). 351. [605] Iniistius dea (Temminck and Schlegel). T ensuidai = Murked Tai. Yokohama and Tokyo markets (Jordan). The lateral black spot covers the greater portion of from one to three scales. Dorsal rays, II-VII, 12; anal. III, 12. 252 MEMOIRS OF THE CARNEGIE MUSEUM. Family SCARID^. Genus Leptoscarus Swainson. The name Leptoscarus replaces Calotomus Gilbert. 352. [607] Leptoscarus japonicus (Cuvier and Valenciennes). Budai = Soldier-Tai. Toba market (Jordan and Yamamoto). Family ZEIDTl. 353. [610] Zeus japonicus Cuvier and Valenciennes. Mato-dai = Target Tai. Tokyo and Osaka markets (Jordan) ; Kochi (Wakiya) ; Tatoku Id. (Mikimoto) ; Kagoshima Bay (Wakiya); Mikawa Bay (M. Ishikawa); Miyazu; Noo; Misaki (Aoki). Generally common. 354. [611] Zenopsis nebulosus (Temminck and Schlegel). Kagami-dai = Mirror Tai. Tokyo market (Jordan). Family CHTITODONTIDTI.^* 355. [619] Chsetodon setifer Bloch. To^/e-c/ioc/io-Rw^o^ Prickly Butterfly-fish. One young individual, 42 mm. long to the caudal fin, from Shizuoka. This specimen is essentially like the adult, except that the dorsal rays are not produced, and the posterior oblique streaks are rather faint. 356. [621] Chaetodon lunula (Lacepede) . Tsw/cf-c/ioc/io-Rw^o = Moon Butterfly-fish. One specimen, 45 mm. long to caudal, corresponds closely with a specimen of like size from Samoa, and with the young as figured by Gunther in ‘‘Fische der Sudsee” (plate 33). 357. [623] Coradion modestum (Temminck and Schlegel). Genroku-dai = Elder-Tai. Osaka market (Jordan); Toba market (Jordan and Yamamoto); Fukui (Nonaka). Genus Acanthoch^etodon Bleeker. This genus, characterized by the lunate caudal, small scales, and gill- membranes narrowly joined to the isthmus, is well distinguished from Holacanthus. The name Loa given by Jordan (Proc. U. S. N. M., 1921, 633) to a species of this family from Hawaii is preoccupied by Loa Stiles, 1905, a genus of worms. Roa has been substituted for it by Jordan, Copeia, May 20, 1923, p. 63. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 253 358. [624] Acanthochaetodon septentrionalis (Temminck and Schlegel). Kinchaku-dai Tai. Misaki (Aoki). 359. [626] Heniochus diphreutes Jordan. Hatatate-dai. The single young Heniochus in the collection was taken by Masashi Ishikawa at the Bay of Mikawa. The anal fin is pale anteriorly. Family ACANTHURIDTl. 360. [629] Naso unicornis (Forskal). = Long-nosed Scraper. One young specimen, 69 mm. long to caudal, from Shizuoka (Jordan). The frontal horns are yet undeveloped, but the two caudal plates are evident. Dorsal rays, V, 28; anal, II, 28 (the last counted as doubled); ventral, I, 3. 361. [630] Xesurus scalprum (Cuvier and Valenciennes). Sannoji-dai. Tokyo and Kobe markets (Jordan); Misaki (Aoki). Dorsal, IX, 23 or 24; anal. III, 22 or 23. Each of our specimens has only four spines on the caudal peduncle, although some have five. 362. [631] Acanthurus matoides (Cuvier and Valenciennes). Kurohagi = Black Hagi or Surgeon-fish. Wakanoura (Yamamoto). One young specimen. Family TEUTHIDTl. 363. [635] Teuthis fuscescens (Houttuyn). Aigo; Ginhagi = Silver Hagi. Tokyo and Shizuoka markets (Jordan) ; Toba (Jordan and Yamamoto) ; Mikawa Bay (M. Ishikawa); Fukui (Nonaka); Miyazu. A common fish often caught from the wharves by boys. Dorsal, I-XII or XIII, 9 or 10; anal, VI (rarely) or VII, 9. Depth of body 2.4 to 2.9 in standard length, decreasing very irregularly with age. Family TRIACANTHODIDiE. 364. [636] Triacanthodes anomalus (Temminck and Schlegel). Benikawamuki = 'Red Rough Skin (File-fish). A rare and most interesting fish. Kagoshima Bay (Wakiya). 254 MEMOIRS OF THE CARNEGIE MUSEUM. Family TRIACANTHID^. 365. [637] Triacanthus brevirostris Temminck and Schlegel. Gin-kawamuki = Silver File-fish. A rare species from Mikawa (Masashi Ishikawa). Family BALISTIDtE. 366. [639] Sufflamen niger (Park). M ongara- Kawahagi = Mongsirsi Kough-skin. Kochi (Wakiya); Misaki (Aoki). 367. [642] Canthidermis rotundatus (Proce). One specimen of this uncommon form was obtained at Wakanoura (Jordan). Family MONACANTHID^. 368. [643] Monacanthus cirrhifer (Temminck and Schlegel)^. Kawahagi = Brough Skin. Tokyo, Shizuoka, Yokohama, and Kobe markets (Jordan); Toba market (Jordan and Yamamoto) ; Kagoshima Bay (Wakiya) ; Mikawa Bay (M. Ishikawa) ; Misaki (Aoki); Toyama (Yoshizawa); Miyazu, Noo. This File-fish, or Leather- jacket is generally common southward. Dorsal rays, 11-31 to 34. 369. [647] Cantherines modestus (Gunther) . Umatsura-hagi = Horse-face File-fish. Osaka market (Jordan); Mikawa Bay (M. Ishikawa); Toyama (Yoshizawa); Misaki (Aoki). Generally common, even northward. Dorsal, 1-35 to 37; anal, 34 (last ray branched from base). The young are marked with several streaks of dark, which in some break into spots much like those of C. tessellatus. These become diffuse in the half-grown individuals and disappear in the adult. The group called Pseudomonacanthus , characterized by the retrorse spinules on the dorsal spine, shade by degrees into Cantherines, in which the spines are merely rough. 370. [647A] Cantherines tessellatus (Gunther). Challenger Repts., Shore-Fishes, 1880, p. 54, pi. 23, fig. 13. Cantherines nigromaculosus Tanaka, Fig. Desc. Fishes Japan, 8 and 9, 1912, pp. 144-196, pi. 38, fig. 145. JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 255 Pseudomonacanthus nigromaculaius Jordan and Thompson, Mem. Car. Mus., VI, 1914, p. 267, pi. XXXI, fig. 2 (Specific name misspelled). Tokyo market (Jordan); Misaki (Aoki). We refer our material to Tanaka’s species nigromaculosus, but that nominal form is apparently identical with Monacanthus tessellatus, described by Gunther from the Philippine Islands. Our specimens have only 32 or 33 dorsal soft rays (the last counted as double), fewer than in either Tanaka’s or Gunther’s types. 371. [647B] Cantherines howensis (Ogilby). One specimen was obtained at Wakanoura by Yamamoto. It agrees well with Tanaka’s description and figure®^ of Japanese material of this species, other- wise known only from Lord Howe Island. We provisionally adopt the synonymy as given by that author. Dorsal, 11-35; anal, 32. 372. [648] Rudarius ercodes Jordan and Fowler. = Net-work Hagi. Misaki (Aoki); Mikawa Bay (M. Ishikawa); Toyama (Yoshizawa). Dorsal rays, 11-23 to 26. 373. [651] Alutera monoceros (Osbeck). Ikkaku-hagi = \]mcorn Hsigi. Tokyo and Kobe markets (Jordan). Dorsal rays, 11-47 or 48. 374. [652] Alutera liturosa (Shaw). ASos/iA/iaf/t = Hoodlum. Wakanoura (Yamamoto) Perhaps identical with the Atlantic form Alutera {Osheckia) scripta (Osbeck). . . Genus Tetrosomus Swainson. We may regard Tetrosomus as a valid genus, related to Lactoria and to Triorus. It may be characterized as follows: carapace closed behind the dorsal and anal fins, pentagonal in cross-section, the dorsal ridge greatly elevated and surmounted by a very large spine; the upper lateral ridges rather sharp, but weak, closely approximated; the lower lateral ridges greatly expanded and bearing five nearly equally spaced spines; sides concave between the ridges; a single supraorbital spine; width of body equal to length to anus. Fig. Desc. Fishes Japan, 20, 1915, p. 354, pi. 105, fig. 300. (the figure in Vol. 19). 256 MEMOIRS OF THE CARNEGIE MUSEUM. 375. [654] Tetrosomus gibbosus (Linnseus). Known from Japan only by one record, made by Bleeker, and this possibly referring to Triorus stellifer. 376. [655] Ostracion immaculatum Temminck and Schlegel. Hakofugu = Box-puffer. Misaki (Aoki). 377. [656] Lactoria diaphana (Bloch and Schneider). [7mi-SR2uma = Sea-sparrow. Misaki (Aoki). 378. [657] Lactoria cornuta (Linnaeus). Kongofugu ^M.edlej-puffer. Mikawa Bay (M. Ishikawa). Triorus Jordan and Hubbs, gen. nov. Orthotype, Lactophrys tritropis Snyder = Ostracion stellifer Bloch. Carapace closed behind dorsal and anal fins, triangular in cross-section; the two sides of the triangle convex, but without a ridge behind the supra-orbital ridge, except for a trace in the young; the bottom of the triangle nearly flat; the body very wide; supra-orbital ridge strong, with two spines; dorsal ridge very high and sharp, with two spines; ventral ridges each with four spines, the last remote from the anterior three, the two median spines most closely approximated. We regard this genus as more closely related to Tetrasoinus than to the American species now classed together under the name Lactophrys. 379. [658A] Triorus stellifer (Bloch). Ostracion gihhosmn Franz, Abh. Bayer, Akad. Wiss., Vol. I, Suppl. 4, 1910, p. 91 (not of Linnseus). Lactophrys concatenatus Gilchrist and Thompson, Ann. Durban Museum, I, 1917, p. 423. One specimen from the Bay of Kagoshima (Wakiya). Twenty others from Misaki (Aoki). 380. [659] Kentrocapros aculeatus (Houttuyn).^“° Itornaki-fugu. One specimen of this little Trunk-fish was obtained by Wakiya at Kagoshima. Sixteen taken by Aoki at Misaki. In young, 3.0 to 3.5 cm. long to the caudal fin, the spines on the ridges are represented by mere tubercles. In one adult one of the pair of most prominent superolateral spines is tripartite on one side. See McCulloch and Waite, Trans. Roy. Soc. S. Austral., XXXIX, 1915, p. 492. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 257 Family TETRAODONTID^. 381. [662] Lagocephalus spadiceus (Richardson). Saha-fugu = Mvickerel puffer; Gin-fugu = Silver puffer. Wakanoura (Yamamoto); Toba (Jordan and Yamamoto); Toyama (S. Yoshizawa); Noo. This species is best characterized by the extent of the dorsal prickly area. The area is widest midway between verticals from eye and gill-opening, here covering the whole dorsal surface; from this area prickles extend forward in a narrow area, remaining conspicuous, to the nostrils, but obsolescent farther for- ward; backward from the widest point the area is of similar shape to that before that point, and extends about half-way to the dorsal fin. In some specimens prickles extend farther back in a rather narrow and irregular band in the dorsal line, but they do not even then closely approach the dorsal fin. In the East Indian species L. lunaris the dorsal prickly area in contrast extends backward in a but slightly narrowed band to the dorsal fin. In both species the sides of the trunk and the entire urosome are wholly devoid of prickles. In both the lateral folds of the two sides arise together at the angle of the chin and extend backward, strong throughout their course to the middle of the caudal base, where each is met at an acute angle by the poorly developed dorsal fold of the caudal peduncle. The two species are unquestionably closely related, but different. Further differences, also pointed out by Bleeker, are noticeable in proportions, L. spadiceus having a rather slenderer body and shorter head (a little less rather than a little more than one-third of the total length to caudal). In addition to the Japanese material of L. spadiceus reported on by Jordan and Snyder, we have examined one from Swatow, China, collected by Miss Adele M. Fielde and recorded as Lagocephalus lunaris by Rutter, two from Manila, recorded by Jordan and Seale, and by Jordan and Richardson, in each case together with true L. lunaris in the same lot as Sphoeroides lunaris. Of L. lunaris we have other Philippine material, two specimens from Hong Kong, collected by Captain Finch, and one from Moreton Bay, Queensland, collected by Ogilby. A definite color-pattern is sometimes developed in the young, rarely in the adult. It consists of a sharply defined double cross-bar across the back about midway between the pectoral and dorsal fins; a large blackish blotch below the Proc. U. S. N. M., XXIV, 1901, p. 235. Proc. Acad. Nat. Sci. Phila., 1897, p. 81. Jordan and Seale, Bull. Bur. Fish., XXVI, 1906 (1907), p. 36. Jordan and Richardson, Bull. Bur. Fish. XXVII, 1907 (1908), p. 273. 258 MEMOIRS OF THE CARNEGIE MUSEUM. dorsal base; irregular longitudinal blotches along the sides; indefinite cross-shades connecting the darkened supra-orbital ridge, and irregular dark spots on the back, especially distinct on the top of the caudal peduncle, or there united to form an indefinite saddle. 382. [664] Sphoeroides^°^ alboplumbeus (Richardson). Komon-/R^R = Belly-puffer. Mikawa Bay (M. Ishikawa); Osaka market (Jordan); Kagoshima Bay (Wakiya); Soo-chow, China (Dr. Cora B. Reeves). The synonymy of this species is greatly confused. We find no Japanese or Chinese specimens corresponding with S. oblongus Bloch of the Indian Ocean, with which alboplumheus has been identified, but we are not certain that the two are really different. The oldest name seems to be guttulatus described by Richard- son (1845) as a variety of ocellatus, and having page-priority over alboplumheus, but later (1846) doubtfully referred to the synonymy of alboplumheus by its describe!'. All Japanese and Formosan records of ocellatus probably refer rather to this species.^”® Spheroides stictonotus is probably the adult, as it differs from typical alboplumheus in much the same w^ay as the adult (abbotti) of vermicularis differs from the half-grown and young. But nearly all of these suggestions require confirmation. The prickly area extends from between the nostrils backward dorsally to the caudal fin, very closely approaching the dorsal fin, and from this dorsal area downward to include most of the lower surfaces, except on the caudal peduncle; the entire preorbital region to the lateral fold and the chin are without prickles. The prickles are sufficiently enlarged to make the skin rough to the touch on the back from between the eyes half-way or almost to the dorsal fin, on the belly, and sometimes also in lateral bands before and behind the pectoral fins. Where weak, the prickles may become obsolete. The course of the lateral lines on the snout is variable. 383. [666] Sphoeroides rubripes (Temminck and Schlegel). Tora-fugu = Tiger-puffer. Kobe market, Enoshima (Jordan); Onomichi (Jordan and Snyder). Young specimens of this species from Onomichi were recoi'ded by Jordan and Snyder'”’ as Spheroides alboplumheus. The name Sphoeroides (1798) antedates Spheroides 1806. Philippine records of S. ocellatus (Jordan and Seale, Proc. U. S. N. M., XXVIII, 1905, p. 791, and Bull. Bur. Fish. XXVI, 1906 (1907), p. 36), refer to Chelonodon patoca, as we have determined by re-e.xamining the material so recorded. A Chinese record of Lagocephalus ocellatus Abbott (not Osbeck), refers to S pheroides macclellandi Regan. Proc. U. S. N. M., XXIV, 1901, p. 244. JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 259 With age the pectoral spot becomes blacker and more strongly ocellated, rather than grading into the general color scheme. Moreover additional, or postpectoral, ocelli become developed. The prickles commence well behind the lips, not far in advance of the nostrils, and extend backward to within a short distance of the dorsal fin, where the patch is abruptly truncated and completely terminated, except for a few prickles, which extend along the lateral line below the dorsal base. The ventral patch extends from the angle of the chin almost to the anus. The two patches are either separated or very narrowly joined before and behind pectoral fins. Wide areas about the mouth, eyes, and gill-openings, and almost the entire tail are wholly devoid of prickles. Wherever the prickles occur they are large and strong, being readily visible to the unaided eye, and very sharp. Dorsal rays 16 or 17. The use of the inflated skins of this large species as ornamental lanterns is a specialty of Enoshima. 383a [668] Sphoeroides (Temminck and Schlegel). Yokohama, Tokyo, Toba (Jordan); Mikawa (M. Ishikawa); Misaki (Aoki). 383b [670] Sphoeroides pardalis (Temminck and Schlegel). Toba (Jordan); Fukui (Nonaka); Misaki (Aoki). Family SCORPvENID^. 383c [697] Sebastolobus macrochir (Gunther). Western Hokkaido (Majima); Kushiro, Nemuro (Tanaka); Misaki (Aoki). Genus Sebastodes Gill. {Mebaru; Soi). For the present we refer all the Japanese species allied to Sehastes, but having only thirteen or fourteen dorsal spines, to the single genus Sebastodes. We realize perfectly that not one of the Japanese forms is naturally congeneric with Sebastodes paucispinis, the Californian type of the genus, nor yet with nigrocinctus, the type of the genus Sebastichthys, a Californian species, which stands at the opposite extreme of the series of fifty or more known forms. It will be necessary to break up the group into from five to ten genera, but the divisions, thus far proposed, fail to satisfy, as the characters lack definiteness, or are subject to intergradation. It seems best to leave the arrangement of genera to some monographer. We may note, that, the more extensive the material in hand, the more difficult the problem, as appears in Frank Cramer’s elaborate paper on the “Cranial Characters of Sebastodes” (Cal. Ac. Sci., V, 1895). 260 MEMOIRS OF THE CARNEGIE MUSEUM. The evolution of the group from Miocene forms {Sebastoessus, Rixator, etc.) allied to Sebastodes revealing greater and greater divergence and modification of the cranium is evident. The Japanese species of this group may be provisionally referred to the follow- ing subgenera, already more or less fully defined, but by no means adequately covering the entire group. Emmelas Jordan and Evermann, glaucus. Primospina (Eigenmann and Beeson) owstoni, sasakii. Sebastosomus Gill, inermis, tokionis, joyneri, thompsoni, schlegeli, taczanov%^kii, Jkmimeus, itinus, steindachneri. Acutomentum Eigenmann, matsubarce, iracundus, scythropus. Rosicola Jordan and Evermann, fuscescens. Pteropodus Eigenmann, vulpes, nivosus, trivittatus, mitsukurii, pachycephalus. Sebastocles^°^ Jordan and Hubbs, elegans. Subgenus Primospina Eigenmann and Beeson. 384. [687A] Sebastodes (Primospina) owstoni Jordan and Thompson. Toyama (Yoshizawa); Noo. This species is represented in the present collection by five small specimens (97 to 130 mm. long), all from the Sea of Japan. All of the specimens have fourteen dorsal spines. The color in life was obviously red, marked with about five indefinite dark saddles along the back, of which the one below the tenth and eleventh dorsal spines is the most conspicuous. 385. [687B] Sebastodes (Primospina) sasakii Tanaka. This species seems to be known only from the original description published in Japanese.^®® We give here a translation of the original description, made for us by Mr. Kasawa: ‘Head 2.9 in length of body; depth, 3.0. Diameter of eye 4.5 in head; inter- orbital width, 3.5; snout 3.33; maxillary 2.1; depth of caudal peduncle, 4.0. D. XIII, 15; A. Ill, 7; pectoral, 19, all soft, the first and last unbranched, the first eight branched, the last nine unbranched, thickened; V. I, 5; C, about 12 (branched rays only counted). Scales about 115 in lateral line; 15 between D. and lat. 1.; 27 between lat. 1. and anal; pores about 59. Gill-rakers 28; longest gill-rakers 1.5 Sebastodes Jordan and Hubbs, subgen. nov. Dorsal spines low, normally fourteen; interorbital deeply concave, size small. Type Sebastes elegans Steindachner. Zool. Mag., XXVIII, No. 333, p. 257. JORDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 261 in eye. Body elongate, not elevated, evenly rounded. Maxillary not reaching to anterior margin of pupil. Pectoral fin reaching beyond ventral fin, but not to anus; distance from tip of pectoral to origin of anal equals length of V. Caudal fin emarginate. Scales ctenoid. Maxillary scaled. ‘In formahn color brownish red, pale ventrally ; irregular brownish black blotches on sides; about five wide bars above lateral line; many irregular spots between bars. Four brownish bands about eye, one band across the head in front of eyes; one band branched behind eye, one branch extending across head back of eye, the other branch running along side of body; next band below eye also branched, one branch running to opercle, one branch extending obliquely downward and back- ward; the last band extending downward and forward from eye. Another band on interorbital space, not touching eye. Very little black on lower part of opercles; belly pale. ‘Specimen taken in deep water (about 120 fathoms). ‘Named for Madoka Sasaki, professor of fisheries. Imperial University of Hokkaido. ‘Caught off the coast of Rikuzen (Matsushima).’ Subgenus Sebastosomus Gill. 386. [689] Sebastodes (Sebastosomus) inermis (Cuvier and Valenciennes). Kuro-soi = Black Rock-cod. Sebastes inermis Cuvier and Valenciennes, Hist. Nat. Poiss., IV, 1829, p. 346. — Hilgendorf, Sitzungsb. Ges. Nat. Freunde. Berlin, 1880, p.l72. Sebastodes inermis Jordan and Starks, Proc. U. S. N. M., XXVII, 1904, p. 103. — Franz, Abh. Bay. Akad. Wiss., Vol. I, Suppl. 4, 1910, p. 70. — Snyder, Proc. U. S. N. M., XLII, 1912, p. 425. — Jordan and Metz, Ann. Car. Mus., VI, 1913, p. 51. — Jordan and Thompson, ibid., VII, 1914, p. 271. Sebastodes ventricosus Temminck and Schlegel, Fauna Japonica, Pisces, 1843, p. 48, pi. 20, figs. 1, 2. — Bleeker, Verb. Bat. Gen., XXVI, 1859, p. 80. — Nystrom, Kong. Verh. Hand!., 1887, p. 20. Sebastodes fuscescens Jordan and Snyder, Proc. U. S. N. M., XXIII, 1900, p. 756 (Probably not of Houttuyn). Sebastodes guntheri Jordan and Starks, Proc. U. S. N. M., XXVII, 1904, p. 102, fig. 2. — Jordan and Metz, Ann. Car. Mus., VI, 1913, p. 49, fig. 43. Tokyo and Yokohama markets (Jordan); Toba (Jordan and Yamamoto); Misaki (Aoki); Mikawa Bay (M. Ishikawa). Snyder, and Jordan and Thompson were undoubtedly right in referring guntheri to the synonymy of inermis. 262 MEMOIRS OF THE CARNEGIE MUSEUM. Snyder expresses doubt as to whether Cuvier and Valenciennes had the species, now called inermis or tokionis in mind, and suggests that Sebastes ventricosus and Sehastodes tokionis are identical. In our opinion, however, Schlegel’s figure of ventricosus is assuredly not based on a specimen of tokionis, and agrees in detail with inermis, and not with any other species of the inermis group. Hilgendorf, furthermore, compared the type of inermis with a specimen identified by Schlegel as ventricosus and declares the two to be identical. Consequently, until contrary evidence is forthcoming, we propose to regard ventricosus as a synonym of inermis, and to continue to apply the name inermis to the species here discussed. It is probable, however, that the descriptions of Gunther (ventricosus) and of Stein- dachner (inermis) were based on specimens of tokionis. Sehastodes inermis is typical of a complex group of species, the genus Sebas- tosovnus of Gill, which differ widely from other species of Sehastodes, and agree among themselves in nearly all trenchant characters. In all of these forms the head is nearly smooth, the supraorbital and occipital ridges being weak; nasal, preorbital, postorbital, and occipital spines alone are developed on the top of the head, and they are small and rather depressed; the preorbital is armed by two sharp spines directed downward and backward; behind them the suborbital is extremely narrow; of the two opercular spines the upper is the larger; the five preopercular spines are all directed backward; of these the first is very small, the second much the largest; the interorbital is evenly and very slightly convex and rather broad; the gill-rakers are long and slender, twenty-four to twenty-seven in number on lower limb of first arch; the mandible projects as a rather sharp knob, and a double symphyseal knob is developed; the broad end of the maxillary ex- tends about to below the middle of orbit; the suborbital stay is complete; the head is covered with ctenoid scales to the lips, the snout, preorbital, suborbital, maxillary, mandible, exposed gular and branchiostegal regions, all being closely scaled; the body-scales are of moderate size, the pores thirty-four to fifty-three in number; there are few accessory scales; the dorsal fin is of moderate height, and composed of thirteen spines and twelve to fifteen soft rays; the anal rays are III, 5 to 8; the paired fins are pointed and long; the anus is well in advance of the anal fin; the peritoneum in all is white. In addition to S. inermis this group includes tokionis, joyneri, and thompsoni. These four species may be divided into two groups. In the first pair, inermis and tokionis, the caudal fin is strictly truncate, and the ventral fin is very long, reaching far beyond the anus, which is distant about two-thirds the orbital length from the anal fin. In the other two species, joyneri and thompsoni, the caudal JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 263 fin is distinctly emarginate and sharply lobed; the ventral fin is shorter, reaching but little, or not at all, beyond the anus; the anus is distant the full length of the orbit from the anal fin. Description of Sebastodes inermis: Dorsal rays, XIII, 12 to 15, most fre- quently 14; anal rays. III, 5 to 8, usually 7 or 8. Pores in lateral line to caudal base, 39 to 47 (the type of guntheri has 42, not 50 pores, as described); scales vertically below first dorsal spine, to lateral line, 11 to 17; scales in an oblique row from first anal spine to lateral line, 18 to 27. Scales somewhat rougher and firmer than in related species, and with more accessory scales at bases; each scale near middle of body, above lateral line, with 7 to 12, usually 9 to 11, radii. Back rather strongly elevated, depth of body greater than length of the head, and contained 2.4 to 2.7 times in standard length. Length of orbit 2.9 (rarely) to 3.4 in head to end of opercular flap, and 1.15 to 1.4 times the interorbital width (1.45 to 1.6 times interorbital in young); length of upper jaw, 2.2 to 2.4 in head; length of ventral spine 1.4 to 1.9 in total length of fin. Color brassy green to reddish brown or blackish, becoming dusky silvery to dull reddish below; the sides crossed by wide, indistinct, and much disrupted cross-bands, which are about as well developed below as above the lateral line; young boldly marked with rather large spots, which later merge with the bars. 387. [691] Sebastodes (Sebastosomus) tokionis Jordan and Starks. Aka-mebaru = ^ed Rock-cod. Sebastes ventricosus Gunther, Cat. Fishes Brit. Mus., II, 1860, p. 97 (not of Temminck and Schlegel). Sebastodes ventricosus Jordan and Evermann, Bull. U. S. N. M., XLVII, pt. 2, 1898, p. 1829 (after Gunther). Sebastes inermis Steindachner and Doderlein, Denksch. Akad. Wiss. Wien, 1884, p. 205 (not of Cuvier and Valenciennes). Sebastodes inermis Jordan and Evermann, 1. c., p. 1829 (after Steindachner and Doderlein) . Sebastodes tokionis Jordan and Starks, Proc. U. S. N. M., XXVII, 1904, p. 104 fig. 3. — Franz, Abh. Bay. Akad. Wiss., Vol. I, Suppl. 4, 1910, p. 70. — Snyder, Proc. U. S. N. M., XLII, 1912, p. 425. — Jordan and Thompson, Mem. Car. Mus., VI, 1914, p. 425, fig. 40. Osaka market (Jordan); Misaki (Aoki); Toba (Jordan and Yamamoto). The proportionate measurements of depth of body and of eye, as given by Gunther and by Steindachner, indicate that they had before them the present species, rather than S. inermis. This species is extremely close to Sebastodes inermis. 264 MEMOIRS OF THE CARNEGIE MUSEUM. but differs in having a lower average number of pores and scales, a higher average number of radii on the scales, the back not elevated, and the body slenderer, the eye usually much larger (at comparable sizes), and the color in life lighter and redder. Description of Sebastodes tokionis (based on the material listed above, on the type and four paratypes from Misaki, and one specimen from Tokyo, collected by Jordan and Snyder in 1900): Dorsal rays, XIII, 13 to 15; anal rays. III, 6 to 8. Pores in lateral line to caudal base, 35 to 45, usually fewer than 42 (in the type we count 39); scales below first dorsal spine, II to 14; obliquely above first anal spine, 17 to 22 (in each case to lateral line). Scales near middle of body, above lateral line, averaging somewhat longer and more widely exposed than in related species, each with 8 to 15, usually 10 or more, radii. Dorsal contour much less elevated than in S. inermis and more gently curved, the greatest depth conse- quently less, about equal to length of head, contained 2.7 to 2.9 times in length to caudal. Orbit very large, 2.7 to 3.2 in head; in the adult 1.4 to 1.65, and in the young 1.55 to 1.65 times the interorbital width; length of upper jaw 2.3 to 2.4 in head; length of ventral spine, 1.55 to 1.85 in total length of fin. Color of a fresh specimen brown above, becoming bright red on lower parts; the dorsal dusky, becoming blackish outwardly, but with the margins of the membranes from the fifth to the tenth spines red; all other fins red; nasal tubes bright red. In alcohol the back is dusky (lighter than in S. inermis), the lower parts silvery; dorsals, anal, and tip of ventral dusky; pectorals colorless. In the young the fins are darker, and the sides are marked with spots, which, though clearly evident, are smaller, more numerous, and less distinct than in the young of S. inermis. Adults are indefinitely marked with cross-bars, which are well shown in the figure of the type; these, as in S. inermis, are about as well developed on the lower side as above the lateral line, and are nowhere sharply defined. 388. [692] Sebastodes (Sebastosomus) joyneri (Gunther). T'aA:eno/co-me5arw = Bamboo Rock-cod. Sebastes joyneri Gunther, Ann. Mag. Nat. Hist., I, 1878, p. 485; Challenger Reports, Shore-Fishes, 1880, p. 64, pi. 29, fig. A. — Steindachner and Doder- LEiN, Denksch. Acad. Wiss. Wien, 1884, p. 206. Sebastodes joyneri Jordan and Evermann, Bull. U. S. N. M., XLVII, pt. 2, 1898, p. 1829 (after Steindachner and Doderlein). — Jordan and Snyder, Proc. U. S. N. M., XXIII, 1900, p. 757 (in part). — Jordan and Starks, Proc. U. S. N. M., XXIII, 1900, p. 105 (in part, not specimens from Miyako). — Snyder, Proc. U. S. N. M., XLII, 1912, p. 426. Shizuoka (Jordan). JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 265 We separate from S. joyneri, as representatives of a new species, which we name Sebastodes thompsoni, certain specimens, which hitherto have been con- founded with this species. Excluding these we have seven specimens (five pre- viously recorded) which correspond in full detail with the descriptions of Gunther and Steindachner. These specimens serve as the basis for the description which follows. Description of Sebastodes joyneri: Dorsal rays XIII, 14 or 15; anal rays, III, 7 or 8; pores in lateral line 43 to 51 (42 to 49 counted by Steindachner); 13 to 15 scales between lateral line and origin of dorsal, 24 to 29 scales to origin of anal; scales near middle of body and above lateral line with 8 to 13 radii. Both contours rather evenly curved, the body rather slender, its greatest depth 2.7 to 2.9 in length to caudal; orbit very large, its length 2.65 to 3.4 in head to end of flap, 1.2 to 1.7 times the interorbital width; length of upper jaw, 2.35 to 2.5 in head; length of ventral spine, 1.4 to 1.75 in total length of fin. Color red, a little darker above, with highly intensified and sharply margined oblique black bars of constant form, five partly below and partly on dorsal base; first composed of two separated rounded spots, one at base of fin and one on lateral line; the second bilobed, extending a little below lateral line; the third comma- shaped, barely extending below lateral line, the ventral end showing some variation, sometimes being partly or completely separated from the rest of the bar, the detached portion rarely divided vertically; fourth round, extending only half-way to lateral line; the fifth smaller and squarer in shape, a saddle across caudal peduncle at end of dorsal base, extending a little out on fin, but not to latest line; the sixth a small spot at upper edge of caudal base. Largest specimen 188 mm. long to caudal. 389. [692A] Sebastodes (Sebastosomus) thompsoni Jordan and Hubbs, sp. nov. Sebastodes joyneri Jordan and Snyder, Proc. U. S. N. M., XXIII, 1900, p. 757 (in part). — Jordan and Starks, Proc. U. S. N. M., XXVII, 1904, p. 105 (specimen from Miyako). — Jordan and Thompson, Mem. Car. Mus., VI, 1914, p. 271, pi. XXXII, fig. 2. (Specimen from Tsushima Straits). This species is nearest Sebastodes joyneri, with which it has heretofore been confused, but differs sharply in coloration, as is well shown in the figures of the twb species. The pores in the lateral line are more numerous (52 or 53) than in any other species of the mermfs-group. Type, 189 mm. long to caudal fin, collected by Jordan and Snyder (in 1900) at Miyako, Japan; Cat. No. 7167, Stanford University Collections. Paratypes of For certain other characters of S. joyneri see the comparisons made under S. inermis, p. 262. 266 MEMOIRS OF THE CARNEGIE MUSEUM. about the same size were collected by Doctor Jordan (in 1911) at Osaka; Cat. No. 22640, Stanford collections (No. 6037a Car. Mus. Cat. Fishes). Head, 2.85 to 3.05; depth, 2.7 to 2.9. Body rather slender, the back not especially elevated; the dorsal and ventral contours about equally curved. Head rather smooth, supraorbital and occipital ridges being low and partly scaled over; nasal, preorbital, postorbital, and occipital spines alone developed on top of head, all small and depressed; preorbital armed by two sharp spines (one bifid in type) directed downward and backward; suborbital extremely narrow; upper opercular spine the longer; five preopercular spines, the first very small, the second much the largest, those following progressively shorter, all directed backward. Inter- orbital nearly smooth, very slightly convex, its edges nowhere gibbous, its least Fig. 1. Sebastodes thornpsoni Jordan and Hubbs, sp. nov. Reproduction of the figure of S. joyneri Jordan and Thompson, Mem. C. M., VI, PL XXXII, fig. 2. width 1.35 to 1.6 in orbit; suborbital stay complete; orbit, 3.2 to 3.25 in head; maxillary, 2.2 to 2.4, its broad end reaching to below middle of orbit; mandible projecting as a rather sharp knob; a double symphyseal knob of teeth fitting into interspace between expanded anterior lobes of premaxillary teeth; sides of jaws and palatines with narrow bands of teeth; gill-rakers rather long and slender, twenty-seven on lower limb of outer arch. Head covered with ctenoid scales to the lips, the snout, preorbital, suborbital, maxillary, mandible, and exposed portions of gular and branchiostegal membranes all being closely scaled. Body- scales rather smaller than in related species, the pores of lateral line numbering 52 or 53 to caudal base; about 17 scales in a series from first dorsal spine vertically JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 267 downward to lateral line, 28 to 30 in an oblique row from first anal spine to lateral line. Accessory scales few, except on nape and along base of dorsal. Scales with about eleven radii. Dorsal rays, XIII, 14; anal. III, 7; pectorals, 16. Dorsal spines short and moderately robust, the fourth or fifth longest, a little longer than highest soft ray, contained 2.3 to 2.4 times in head; membranes between dorsal spines not deeply incised; the dorsals not very deeply emarginate. Caudal dis- tinctly emarginate; the lobes rather sharp. Third anal spine about as strong as second and a little longer, a little more than two-thirds the longest soft ray, 2.4 to 2.7 in head. Ventral pointed, scarcely extended beyond anus, the length of the fin 1.6 to 1.65 times the length of the spine, contained 1.55 to 1.65 times in head. Pectoral rather pointed, reaching to above anus, 1.2 in head. Unbranched pectoral rays not enlarged. Anus in advance of anal fin a distance about equal to length of orbit. Peritoneum white. Color doubtless red in life; in spirits dusky above, and silvery below the lateral line. Upper sides marked with dark brown bars, not black as in S. joyneri, and of different form. The form of the bars is better indicated by the figure than by description. Mem. Car. Mus., Vol. VI, pi. XXXII, fig. 2. 390. [690] Sebastodes (Sebastosomus) schlegeli (Hilgendorf). Kuro-mebaru = Black Rock-cod. Otaru market (S. Takayasu); Sapporo market (Majima). 391. [695] Sebastodes (Sebastosomus) steindachneri (Hilgendorf). Yanagi-mebaru ^Willow Rock-cod. A single specimen taken by Tanaka at Nemuro. Dorsal rays, XIII, 15; anal. III, 7; pectoral rays, 1, 9, 8=18; pores, 32; head, 2.7; depth, 2.7; orbit, 4; snout, 4. 392. [697] Sebastodes (Sebastosomus) taczanowskii (Steindachner) . Otaru market (Takayasu); Sapporo market (Majima). In this species there is wide variation in structural features. The postorbital spine may be either absent, or present; the posterior suborbital lobe may become divided into two parts; scales sometimes extend forward on the mandible over the articular bone, and along the upper edge, and sometimes even scatteringly over the posterior outer face of the dentary bone. The body, when fresh, shows more or less distinctly pearly spots on the centers of scales on the sides. 268 MEMOIES OF THE CARNEGIE MUSEUM. 393. [698] Sebastodes (Sebastosomus) flammeus Jordan and Starks. = Scarlet Rock-cod. One specimen, 345 mm. long to the caudal fin, taken at Kushiro by Tanaka, agrees well with Jordan and Thompson’s redescription and figure of this species. Measurements in hundredths of length to caudal: head (to end of opercular membrane) .40; upper jaw, .175; snout, .10; mandible, .225; pectoral, .28; ventral fin, .185; ventral spine, .102; fourth dorsal spine, .105; third anal spine, .098; bony interorbital, .083; depth of caudal peduncle, .10. Dorsal rays, XIII, 15; anal. III, 8; pores in the lateral line, 33; gill-rakers, 1-1-20. The teeth of the upper jaw are in a band anteriorly, the few next the anterior notch being canines, in a double series of fairly strong incurved teeth medially, and in a band of fine teeth near the angle of the gape. The mandibular teeth are developed as a group of small canines on the strong symphysial knob, but rapidly narrow to a single series, which extends along the entire sides of the jaws. The dark blotch mentioned by Jordan and Snyder is not on the opercular flap, but on the upper part of the branchiostegal membrane. The membrane above the anterior half of uppermost branchiostegal is also brownish black. These two marks, ordinarily concealed, are sharply diagnostic of the species. Subgenus Acutomentum Eigenmann. 394. [700] Sebastodes (Acutomentum) iracundus Jordan and Starks. Itten-ako = Spot Rock-fisk. Sebastes matsuharce var., Hilgendorf, Sitzungsber. Ges. Naturforsch. Freunde, Berlin, 1880, p. 170, and plate. Sapporo market (Majima); Kushiro (Tanaka). The four specimens fully agree with the fish recorded under the above name by Snyder, from Mororan. They differ from the type chiefly in coloration, which is probably largely due to changes in preservation. The black blotch on the side of the large type, as described, is very much smaller on one side than on the other; such irregular spots appear on various Californian species of the genus. In the following counts and measurements the items applying to the large type are put in parentheses, when different from the measurements taken from the five other specimens at hand. Mem. Car. Mus., VI, 1914, p. 270, pi. XXXII, fig. 1. Proc. U. S. N. M., XLII, 1912, p. 426. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 269 Dorsal rays, XIII, 14 (13); anal rays, III, 8; pectoral rays, distinguishing branched and unbranched rays, 1 or 2+8 or 9 (9 or 10) + 8 or 9 (9) = 18 or 19 (19 or 20); pores, 30 to 35 (30). Head, 2.55 to 2.6 (2.45), depth, 2.4 to 2.6 (2.8); depth of caudal peduncle, 3.7 to 4.1 (4.7); orbit, 3.3 to 3.5 (3.8); interorbital, 4.6 to 5.1 (5.1); snout, 3.8 to 4.1 (4.25); upper jaw, 1.9 to 2.1 (2.1). Scales covering all exposed surfaces of the head, excepting only the eyes and lips, everywhere ctenoid. Teeth in a moderate band in premaxillaries, somewhat enlarged anteriorly ; mandibular teeth coarse on the small symphysial knob, chiefly in one series of strong teeth, laterally flanked to well toward the end of the series by one or two irregular rows of smaller teeth. Spines of head subject to much variation; supra- ocular spine occasionally reduced to a sharp knob; coronal spines indifferently present or absent, directed outward or backward, and vatying in position; nuchal spines distinct, or more or less completely fused with the parietals. Lower border of orbit without a raised and broken crest, but sometimes showing a small spine at base of the upper edge of suborbital stay. Interorbital flat or weakly concave, with a pair of very weak to rather strong ridges. Color red, with dusky markings arranged in diagnostic fashion. Three narrow bars cross the top of the head at the front and back of orbit and at nape; rather faint cross-bars or wedges extend downward from about the front of the dorsal to the black opercular blotch, from the middle of the spinous across the lateral line, and as irregular fragments well toward the ventral fins, from near the end of the spinous dorsal not nearly to the lateral line, and from below the end of soft dorsal base almost to lateral line; a longitudinal row of spots, more or less blurred into blotches, opposite the dark bars, extends along the body just above and below the lateral line, the ventral series being the more conspicuous. A dark blotch on the branchiostegal membrane near opercular spot, and a variable amount of like color on the hidden portion of the membrane medially. Buccal and branchial cavities coarsely blotched with black; peritoneum black. 395. [693] Sebastodes (Acutomentum) scythropus Jordan and Snyder. Ukeguchi-mebaru = Lucky-mouth Rock-cod. A ripe female, 202 mm. long to the caudal fin, was obtained in the Yokohama market (Jordan). Dorsal, XIII, 12; anal. III, 5 (counting last rays as branched); second anal spine little longer than third. Ground-color pale pinkish, becoming silvery below, and marked dorsally by large orange-red blotches, broadly bordered with bluish gray, and arranged as follows: a streak along front of lateral line ending in a large blotch below middle 270 MEMOIRS OF THE CARNEGIE MUSEUM. of spinous dorsal; a similar streak between lateral line and dorsal; four blotches diagonally disposed about the one mentioned above, the foremost behind opercle, the upper two extended into dorsal base, the one under the four last spines of the first dorsal fin wedge-shaped; the fourth blotch indefinitely connected along lower sides with the lower end of a prominent bar below soft dorsal; another bar on caudal peduncle near base of caudal. Head pale, with some mottling of bluish dusky above, and with a large and conspicuous rich deep brown opercular blotch. Dorsal fins with extensions of the red body-markings, and with red dashes in the membranes distally. Caudal and ventrals red; pectorals and anal pinkish. Subgenus Pteropodus Eigenmann. 396. [702] Sebastodes (Pteropodus) vulpes (Steindachner and Doderlein). Ma-soi = True Rock-cod ; Kitsune-mebaru = Fox Rock-cod. Otaru market (Takayasu). 397. [703] Sebastodes (Pteropodus) pachycephalus (Temminck and Schlegel.) Hachigara = Cranium-belly; Mura-soi = Irregular Rock-cod. Otaru market (Takayasu); Yokohama market (Jordan); Toyama (S. Yo- shizawa). 398. [706] Sebastodes (Pteropodus) mitsukurii (Cramer). Toba market (Jordan and Yamamoto); Mikawa Bay (M. Ishikawa); Fukui (Nonaka). Young, about 4 cm. long to caudal, are very similar to the adult in form and color, differing chiefly in having the gill-rakers about one-third as long as the orbit. 399. [708] Sebastodes (Pteropodus) trivittatus (Hilgendorf) , »S/?ima2of = Striped Rock-cod. Otaru market (Takayasu); Nemuro (Tanaka). Dorsal rays, XIII, 13; anal, 6; pores, 38. Subgenus Sebastodes* Jordan and Hubbs. 400. [704] Sebastodes (Sebastodes) elegans (Steindachner and Doderlein). Y oroi-mebaru = M.a,iled Rock-cod. Misaki (Aoki). Ripe females are 13 to 15 cm. long to caudal. All four specimens have fourteen dorsal spines. This dainty little fish is especially characteristic of the Inland Sea. * See footnote No. 108, p. 260. JORDAN AND HUBBS: ApANESE FISHES COLLECTED 1922. 271 Genus Scorp2enodes Bleeker. This name replaces Sehastopsis Gill, and Sebastopsis Sauvage. Sebastella Tanaka (1918) appears to be another synonym. 401. [709] Scorpsenodes littoralis (Tanaka). 7so-A:asa^o = Surf Scorpion-fish. Sebastella littoralis Tanaka, Zool. Mag., XXIX, 1918, p. 10. We offer a translation of the original account of this species, which was in Japanese: ^Head 2.66 in length of body without caudal fin; depth 3. Diameter of eye 3.875; interorbital, 6.89; snout, 3.44; maxillary, 1.77; caudal peduncle, 3.875. Dorsal, XIII, 9; anal. III, 5; pectoral fin, 17, the lower nine rays unbranched; ventral, I, 5; caudal, 11 (branched rays). Scales in series upon lateral line, 46; 6 above, 11 below lateral line. Teeth on both jaws and on vomer, but not on palatines. Second anal spine stronger and longer than third. Tip of pectoral fin reaching a little beyond origin of anal ; ventral fin not reaching anal origin. Caudal rounded. Color in formalin brown. Six indistinct oblique bars on the -sides; three bands radiating downward from the eye; subopercle with a large blotch; many brown spots and streaks on all the fins. Length from front of head to tip of middle caudal rays 99 mm.’ This fish is abundant along the shore at Misaki, and does not attain a large size, a fact which accounts for its not being brought into the markets. It resembles some species called Sebastodes, but differs in having no teeth on the palatine bones. 402. [710] Sebasticus albofasciatus (Lacepede). Ayame-kasago = Bright-colored Kasago. Shizuoka and Osaka markets (Jordan); Misaki (Aoki). The subocular spine is absent in all our specimens but one, on which it occurs on but one side. Dorsal spines normally 12, but occasionally 11 or 13. In large adults the eye is less than one-fourth as long as the head, but in specimens of similar size, the eye is larger than in S. marmoratus. 403. [711] Sebastiscus marmoratus (Cuvier and Valenciennes). Kasago. Sebastiscus tsuraara Tanaka, Zool. Mag., Vol. XXIX, No. 339, 1917, p. 10 (Misaki). 272 MEMOIRS OF THE CARNEGIE MUSEUM. Shizuoka, Tokyo, and Kobe markets (Jordan); Misaki (Aoki); Mikawa Bay (M. Ishikawa). This species varies much in color and pattern, grading from blackish to reddish brown, but not approaching, so far as known, the red color (fading to white in alcohol) characteristic of S. albofasciatus. One of the darker color phases has lately been distinguished specifically by Tanaka as S. tsuraara, but we do not think the form so named to be separable. In this species the spine on the sub- orbital stay near the lower margin of the orbit is but rarely developed, but it is also frequently, perhaps usually, absent in albofasciatus. These two forms are very closely related, and indeed albofasciatus may be merely the deep-water race of marmoratus, and it is likely that Jordan and Thompson (1914) are right in uniting the two, under the older name albofasciatus. 404. [714] Helicolenus emblemarius Jordan and Starks. Yokohama market (Jordan). 405. [715] Scorpaena neglecta Temminck and Schlegel. Fusa-kasago = Fringe Kasago. Misaki (Aoki). 406. [716] Scorpaena izensis Jordan and Starks. Misaki (Aoki) ; Miyazu. 407. [717] Scorpaenopsis cirrhosa (Thunberg). Oni-kasago = F>evi\ Kasago. Kobe market (Jordan). Snyder’s record of this species from Kagoshima refers to S. gibbosa. The two species are, however, well distinguished. Head 2.25 to 2.3; depth, 3.1 to 3.4; orbit, 5.4 to 6.1; dorsal rays, XII, 8 or 9; anal. III, 5. 408. [719] Scorpaenopsis gibbosa (Bloch and Schneider). Scorpcena gibbosa Bloch and Schneider, Ichth., 1801, p. 192, pi. 44. — Gunther, Fische der Siidsee, I, 1874, p. 79, pi. 53; Cat. Fishes Brit. Mus., II, 1860, p. 119. — Bleeker, Verb. Akad. Amster. XVI, 1876, p. 38, pi. 2, fig. 1; Atl. Ichth., IX, 1877, pi. 416, fig. 4 and 4a. Scorpcena kagoshimana Doderlein, Denksch. Akad. Wiss. Wien, 49, 1884, p. 28; ibid., 53, 1887, pi. 3. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 273 Scorpcenopsis kagoshimana Jordan and Starks, Proc. U. S. N. M., XXVII, 1904, p. 137. — Smith and Pope, Proc. U. S. N. M., XXXI, 1906, p. 482. Scorpcenopsis cirrhosa Snyder, Proc. U. S, N. M., XLII, 1912, p. 428 (not of Thunberg) . One specimen obtained at Kagoshima Bay (Wakiya). In Japan this species has only been taken at Kagoshima. Smith and Pope and Snyder err in regarding this species as identical with the very different S. cirrhosa. We have compared this topotypic example of S. kagoshimana with a specimen of /S. gibhosa from Samoa, and with the figures of the species, and find no basis for specific distinction. Especially diagnostic is the coloration of the under surface of the pectoral fin, clear white, with black spots in base, gray mesially, then broadly and regularly black within the narrow white margin. The Hawaiian species Scorpcenopsis catocala has been regarded as identical with S. gibbosa, but probably in error. The spines are much sharper, the cavities of the head are deeper, and the coloration is different, especially on the under surface of the pectoral fin; the base highly mottled, followed dor sally first by a large black blotch, then by a white area, then by a row of black spots parallel with, but well removed from, the margin of the fin; the pale margin of the ventral fins is wider. In these respects S. catocala agrees better with S. diabolus, as de- scribed and figured by Bleeker. We hesitate, however, to make the identification. 409. [722] Pterois lunulata Temminck and Schlegel. Mino-kasago = Rain-coat Kasago. Mikawa Bay (M. Ishikawa); Misaki (Aoki). Genus Brachirus Swainson. The name Brachirus, occurring several pages earlier in Swainson’s work (p. 71) must replace Dendrochirus (p. 180). 410. [723 and 725] Brachirus jordani (Regan). Seto-mino-kasago'— Channel Mino-kasago. Pterois jordani Regan, Ann. Mag. Nat. Hist,, (7) XV, 1905, p. 20. Dendrochirus jordani Snyder, Proc. U. S. N. M., XLII, 1912, p. 428. Ebosia starksi Franz, Abh. Bayer, Akad. Wiss., Vol. I, Suppl. 4, 1910, p. 72, pi. 9, fig. 69. One specimen secured by Wakiya in Kagoshima Bay. Franz’s description of Ebosia starksi agrees with Brachirus jordani in all details, except the number of anal soft rays, given as nine; probably a misprint. 274 MEMOIRS OF THE CARNEGIE MUSEUM. 411. [723A] Brachirus bellus Jordan and Hubbs, sp. nov. (Plate X; fig. 3.) Type 64 mm. long to caudal fin, 90 mm. long to end of caudal; collected at Misaki, Japan, by Aoki; C. M. Cat. Fishes No. 7894. No other specimen has been seen. Dorsal rays, XII, I, 9; anal. III, 5; caudal, 12 (10 branched); pectorals. 1 + 7+9=17. Head, 2.4; depth, 2.5 in standard length. Least depth of caudal peduncle, 3.5 in head, just equal to length of orbit; interorbital, 7.6; snout, 3.9: least sub- orbital width, 8.0; length of upper jaw, 2.5. The ovate form of the body and the contours of the head are well shown by the artist. When viewed from in front, the head is widest at the preopercular angle; the interorbital deeply concave, with a fine groove between the two sub- median ridges, which do not end in spines; no nasal spines; only two blunt spines on upper posterior margin of orbit; infraorbital keel complete, but spineless, ex- tending to the preopercular margin opposite the uppermost and strongest (though weak) preopercular spine; the two additional preopercular spines blunt; opercles devoid of developed ridges or spines; a series of three short ridges ending in blunt spines from eye to below origin of lateral line; occipital ridges moderately elevated and sharp, widely divergent, their weak terminal spines being more than twice as far apart as their origins; a small spine on each side between occipital ridge and orbit. The two suborbital, one nasal, and one supraorbital leaf-like flaps are well drawn by the artist. The teeth are all small and blunt, arranged in moderate bands on jaws and vomer, absent on palatines. The scales of the head are cycloid, rather loosely covering the top of the head forward to middle of interorbital groove, but are well imbricate on opercle, subopercle, and the cheeks above and below the stay; sides and front of interorbital, snout, preorbital, suborbital, both jaws, interopercle, and gular and branchiostegal membranes devoid of scales. The body-scales are ctenoid, rather regularly arranged, without accessory scales, and of rather large size, there being only about twenty-eight along the lateral line to caudal base. The form of the fins is shown in accurate detail by the artist, so that a de- scription would add nothing not evident from an examination of the figure. The body and fins are prettily spotted with brown or black, as pictured. 412. [724] Ebosia bleekeri (Doderlein). Eboshi-kasago = lie\met-Ksisago. A female specimen, without the elevated occipital crest, which features the male of this species, was taken at Misaki by Aoki. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 275 413. [727, 728] Minous monodactylus (Bloch and Schneider). Hime-Okoze = Princess-Okoze, or Stinging Fish. Scorpcena monodactyla Bloch and Schneider, Ichth., 1801, p. 194. Minous monodactylus Cuvier and Valenciennes, Hist. Nat. Poiss., IV, 1829, p. 424, pi. 95, fig. 2. — Gunther, Cat. Fishes Brit. Mus., II, 1860, p. 148 (with synonymy). — Bleeker, Acad. Sci. Roy. Amst., 1876, p. 64 (with synonymy). — Jordan and Thompson, Ann. Car. Mus., VI, 1914, p. 276, fig. 47. Minous adamsii Richardson, Voy. Samarang, Fishes, 1850, p.7. — Jordan and Starks, Proc. U. S. N. M., XXVII, 1904, p. 151. Minous echigonius Jordan and Starks, 1. c., p. 153, fig. 14. Lysoderma satsumce Smith and Pope, Proc. U. S. N. M., XXXI, 1906, p. 484, fig. 7. Kobe market (Jordan); Toba market (Jordan and Yamamoto); Fukui (Nonaka). We find no characters, by which we can distinguish the Japanese and Chinese M. adamsii from the East Indian monodactylus. Regan states that he has found no difference between Japanese and East Indian material. Minous echigonius seems to be likewise identical, and Lysoderma satsumce is doubtless based on a specimen of Minous monodactylus, which had lost its three semi-free anterior dorsal spines, as the type and only known specimen of that nominal species, as described and figured, otherwise agrees to minute details with normal specimens. There is no difference in scales nor in the character of the anal spines, features imperfectly described by authors other than Smith and Pope. 414. [731] Erosa erosa (Langsdorf). Daruma-okoze ^DavumaStinger. Kagoshima Bay (Wakiya); Misaki (Aoki). Rare. 415. [732] Inimicus japonicus (Cuvier and Valenciennes). Oni-okoze = Devil-Stinger. Tokyo, Yokohama, and Osaka markets (Jordan); Mikawa Bay (Ishikawa); Misaki (Aoki); Fukui (Nonaka). Common in the markets southward. All of the specimens taken belong to the dark-colored inshore form {typical japonicus) . Genus Hypodytes Gistel. The name Hypodytes Gistel (1849) should replace the later Paracentropogon. 416. [735] Hypodytes rubripinnis (Temminck and Schlegel). Ha-okose = Chief-okose. Kochi (Wakiya); Mikawa Bay (M. Ishikawa); Misaki (Aoki); Toyama (Yoshizawa). Very common southward. 276 MEMOIRS OF THE CARNEGIE MUSEUM. Family HEXAGRAMMIDiE. 417. [740] Hexagrammos otakii Jordan and Starks. Ainame-, Ahura-koi = Y^ii^\8h.. H exagrammos aburaco Jordan and Starks, Proc. U. S. N. M., XXVI, 1903, p. 1008, fig. 1. Sapporo market (Majima) ; Otaru market (Takayasu) ; Tokyo market (Jordan) ; Toba (Jordan and Yamamoto); Mikawa Bay (M. Ishikawa). Generally common especially northward. We find that the characters supposed to distinguish H. aburaco from H. otakii do not hold; indeed, as Snyder has noted, the fifth lateral line of one side may not meet its fellow {aburaco), which in contrast is continued forward as the median line of the breast {otakii). The fourth line ends at any place between the base of the ventral fins and a point midway between the tip of the ventrals and the anus. 418. [744] Pleurogrammus azonus Jordan and Metz. Pleurogrammus monopterygius Berg, Pise. Mar. Orient., 1904, p. 71. — Tanaka, Annot. Zool. Jap., 6, 1908, p. 240. Pleurogrammus azonus Jordan and Metz, Ann. Car. Mus., VI, 1913, p. 47, pi. VIII, fig. 2 (Korea). Sapporo market (Majima); Otaru market (Takayasu). A ripe female is 365 mm. long; other specimens are larger. Dorsal rays XXI, 28; anal, 26 (last double); depth 4.35 to 4.65; body with indefinite cross-mottlings, as shown in the figure of the type, though denied in the description of the type. Records of Pleurogrammus monopterygius from the fauna of the Japan Sea doubtless refer to the species under consideration. This Alaskan species should hence be eliminated from lists of Japanese fishes. 419. [745] Agrammus agrammus (Temminck and Schlegel). Kujime. Misaki (Aoki); Toba market (Jordan and Yamamoto); Nagoya market (Jordan); Mikawa Bay (Ishikawa); Toyama (Yoshizawa). Not common. Family ERILEPID^E. 420. [746] Erilepis zonifer (Lockington). Aburabozu = Fat priest. Shizuoka (Jordan). Not common, reaches a weight of 40 pounds, or more. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 277 Family COTTID^. 421. [761] Dasycottus japonicus Tanaka. Dasycottus setiger Jordan, Tanaka, and Snyder, Jour. Coll. Sci., Tokyo, XXXIII, 1913, p. 263 (Niigata). (Not of Bean.) Dasycottus japonicus Tanaka, Fig. Desc. Fishes Japan, XVIII, 1914, p. 308, pi. 83, fig. 280 (Niigata). Noo; Fukui (Nonaka). Another specimen, from Nagaoka, Japan, was sent to Stanford University by Nakamina several years ago. Dorsal rays VIII (not XI as given by Tanaka)-12 to 14 (counting the last ray as a double one); anal, 12 or 13; pectoral 23 to 25. On comparison of material representing D. japonicus and D. setiger we find that the differences in the form of the body and the curvature of the upper pre- opercular spine, used by Tanaka, do not hold. The Japanese species differs, however, having the filaments on the head fewer and more scattered, and the maxillary longer, reaching to below the posterior margin of the orbit. 422. [762] Trachidermus fasciatus Heckel. = String-hook Sculpin. A specimen from Fukuoka (Hamada) has only seven simple rays in the pectoral fin. The University of Michigan has lately received this species from Soo-chow, China. 423. [763 and 769] Rheopresbe kazika (Jordan and Starks). Takitaro = Cascade-fish ; Kamakiri = Mantis-fish. Cottus kazika Jordan and Starks, Proc. U. S. N. M., XXVII, 1904, p. 265, fig. 15 (Niigata). Rheopresbe Jujiyamee Jordan and Starks, Proc. U. S. N. M., XXVII, 1904, p. 270, fig. 16 (Odawara). — Tanaka, Fig. Desc. Fishes Japan, VII, 1912, p. 112, pi. 28, figs. 110-111; pi. 29, figs. 115-116. ^One specimen collected at Hamada (Wakiya). We find that Cottus kazika and Rheopresbe fujiyamce are respectively the young and adult of the same species. In the largest paratype of kazika, as in our specimen, the pectoral rays are already well branched; the type of fujiyamce, shows clear evidence of having been prickly; it has eight dorsal spines, and the two dorsals are in close contact. The fact that the type of R. Jujiyamee is a very large female, distended with eggs, accounts for the abdomen being longer than in smaller immature specimens. 278 MEMOIRS OF THE CARNEGIE MUSEUM. 424. [764] Cottus pollux Gunther. = Sculpin, Miller’s Thumb. Lake Biwa (Jordan); Nagano (Nakano); Himeji (Abe); Kamishibi near Fukui. The fin-rays vary considerably, both individually and geographically, but we are unable to refer our material to more than one species. Dorsal, VIII to X-I6 to 18; anal, 11 to 13; pectoral, 12 to 16; ventral, I, 3 or 4. 425. [774] Myoxocephalus raninus Jordan and Starks. Gisu-kazika = Gisu-Sculpin ; Gomo-kazika = Trifling Sculpin. Otaru market (Takayasu); Coast of Rikuchu (Awaya). Our smaller specimen is closely like the type; the larger (29.5 cm. long to caudal fin) likewise agrees, except in having a smaller eye (8.6 in head). 426. [783] Ainocottus ensiger Jordan and Starks. = Spear-sculpin. Ainocottus ensiger Jordan and Starks, Proc. U. S. N. M., XXVII, 1904, p. 283, fig. 23 (Hakodate). Ainocottus Jasciatus Pavlenko, Kazani, Trd. Obsch., XLII, 1910, p. 30. Kushiro and Nemuro (Tanaka). Pavlenko’s English diagnosis of his A. jasciatus contains nothing to indicate that A. jasciatus differs from A. ensiger. Dorsal rays, IX or X-12 to 15; anal, 10 to 13; pectorals, 18; ventrals, I, 3; pores, 38 or 39; second preopercular spine, on one side of one specimen, flattened and bifid; head, 2.2 to 2.3; orbit 4.9 to 5.6; interorbital, 6.4 to 7.4; snout 3.6 (given as 2.4 in description of type, obviously an error). The specimen from Kushiro is a nuptial male and differs widely from the females, which have hitherto been the only sex described. The lower sides are marked by brilliantly white blotches, as large as, or smaller than, the pupil; the ventral fins are black, crossed by four white bands, one basal and one terminal, the general color being that of the species of Megalocottus. The nasal spine is larger and doubled; the tubercles of the head are coarser and the spines are Aiore elevated and broken, slightly approaching those of the marine ^‘Oncocottus.” In addition to clusters of tubercular prickles above and below the lateral line, which are also present in the female, there is developed just above the lateral line an irregular row of cup-shaped scales, which are armed around the posterior border by a few strong spines. Pectoral rays 3 to 10 (counting from top) are armed along inner edge, except toward base, by very long strong spines; the width of a ray, including the spines, is about one-fifth of the large orbit. JOEDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 279 427. [786] Cottunculus brephocephalus Jordan and Starks. Bozu-kazika = Priest Sculpin (with allusion to the shaven head) . Cottunculus brephocephalus Jordan and Starks, Proc. U. S. N. M., XXVI, 1903, p. 689, fig.; XXVII, 1904, p. 289, fig. 25 (Suruga Bay).^ — Tanaka, Figs. Desc. Fishes Japan, XI, 1913, p. 187, pi. 51, figs. 195 and 196. Two adults from Misaki (Aoki). Dorsal rays, VI, 17; anal, 13. Head, 2.4; depth, 3.1, the back more elevated than in the two specimens heretofore recorded. Color brown, becoming pale about nape and dorsal fin, but elsewhere only very indistinctly marked with lighter. 428. [788] Gymnocanthus intermedius (Temminck and Schlegel). Aokazika = Green Sculpin. Takashima market (Takayasu). Dorsal rays, X, 13 or 14; anal, 15. 429. [789] Gymnocanthus herzensteini Jordan and Starks. Tsumaguro-kazika = Finger Sculpin. Kushiro and Nemuro (Tanaka). Dorsal, XI or X-15 or 16; anal, 17; pores 40 or 45; preopercular spine with four antler-like hooks above, those most anterior small on one side of one specimen; the tip of the spine proper bifid on one side only; the two most posterior hooks are imperfectly divided in one specimen, the posterior one directed upward along the side of the anterior one on one side, but directly backward on the other side, so as to be overlapped by the tip of the spine proper. Head, 2.8 or 3.0 in length; orbit, 4.2 or 4.3 in head; interorbital, 10.7 or 11.4; snout, 3.7 or 3.9; upper jaw, 2.3. 430. [792] Cottiusculus gonez Schmidt. Toyama (Yoshizawa); Fukui (Nonaka). The fish described and figured by Pavlenko (1910) as Blennicottus globiceps var. bryosus” from the Bay of Peter the Great on the Asiatic mainland is utterly unlike Blennicottus globiceps. It may be Cottiusculus gonez, or a related and perhaps undescribed species. Two of our eight specimens from the Sea of Japan have the nasal spines double and thus agree with the types of C. schmidti from Matsushima Bay, which consistently show this character, but do not differ in any other appreciable way. The statement that the gill-membranes of C. schmidti do not form a wide fold is not true, the error arising from the fact that all of the types had the head greatly 280 MEMOIRS OF THE CARNEGIE MUSEUM. distorted in preservation. We- await the collection of more material from the east coast of Japan before passing on the validity of C. schmidti. Dorsal soft rays, 10 to 12; anal rays, 9 to 12. 431 . [795 and 806] Alcichthys alcicornis (Herzenstein) . Nizi-kazika = Sham Sculpin. Bero zanclus Snyder, Proc. U. S. N. M., XL, 1911, p. 540; XLII, 1912, pi. 56, fig. 2. Otaru market (Takayasu); Noo; Nemuro (Tanaka). Bero zanclus is based on a half-grown specimen of this species, in which the preopercular spine had not yet developed the flat process on its posterior edge. With age this added spine breaks up into two, or occasionally as many as six separate points. In twelve specimens the fin-rays vary as follows: dorsal, IX or X-14 or 15; anal, 13 or 14 (the last ray doubled in both fins); pectorals, 15 to 17. Pores in lateral line, 35 to 38. Head, 2.6 to 2.8; depth, 4.9 to 5.4, about equal to greatest width of body; orbit, 4.3 to 4.6; bony interorbital width, 14.5 to 17; snout, 3.75 to 4.0; upper jaw, 2.0 to 2.1; least depth of caudal peduncle, 5.1 to 5.7. The color-pattern may consist solely of bars cut by half circles of the ground- color, or of bars plus numerous rounded whitish spots chiefly developed below the lateral line. Genus Furcina Jordan and Starks. Furcina Jordan and Starks, Proc. U. S. N. M., XXVII, 1904, p. 303. Ventral rays I, 2; body smooth, except for a few prickles along the lateral line anteriorly, and between these and the pectoral fin; penis trilobate, as in Pseudo- blennius; teeth in bands in jaws, vomer, and palatines; preopercular spines four, the first strong, turned upward, forked, the second well developed and sharp, near the base of first, the third rudimentary or obsolete, the fourth very small, turned downward; a fringed flap on upper orbital rim, and a simple or slightly divided flap at nape. 432. [796] Furcina ishikawse Jordan and Starks. 433. [797] Furcina osimae Jordan and Starks. This species differs widely from the type-species in having the first preopercular spine widely forked, instead of very narrowly divided at tip (or, rarely, even simple); the second preopercular spine shorter and weaker, not turned upward, smaller, instead of larger, than either fork of the first spine; the fin-rays fewer, 16 or 17 in the second dorsal, instead of 18 to 20, and 13 or 14, rather than 16 or 17, in the anal fin (in each case the last ray was counted as doubled). Neither species of Furcina is represented in the present collection. JOKDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 281 434. [800] Pseudoblennius percoides Gunther. Ana/ia2:e = Hole Goby. Tatoku Island (Mikimoto); Mikawa Bay (M. Ishikawa); Misaki (Aoki), Common southward. 435. [801] Pseudoblennius japonicus Steindachner. Toba (Jordan and Yamamoto); Mikawa Bay (Ishikawa); Misaki (Aoki). 436. [807] Vellitor centropomus (Richardson). Sui = Foip. Misaki (Aoki). Common at times about Misaki and southward, remarkable for its slender, pointed head, unlike that of other Sculpins. Family PARABEMBRADIDiE. Genus Parabembras Bleeker. f ; The genus Parabembras, which appears to be the sole member of a sharply distinguished family, has never been adequately characterized. It differs trenchantly from Bembras and other Bembrids in retaining the three strong anal spines, which are characteristic of the Scorpcenidm in general, also distinct nasal spines, and a more extensive squamation of the head. It is specialized in the development of two spines at the front of the second dorsal fin, which is completely separated from the first. The Parabembradidce are essentially intermediate between the Scorpcenidce and the Bembradidce. 437. [823] Parabembras curtus (Temminck and Schlegel). Uba-go^i = Nurse-kochi. A single specimen was found in the Osaka market (Jordan). Dorsal, IX-II, 7; anal. III, 5; pectoral, 22; ventrals, I, 5; caudal, 16 (15 branched); pores in lateral line to base of caudal 37; scale-rows, 4-37-8. Head, 2.4; depth, 5.15 in length to caudal. Orbit, 3.7; interorbital, 16; snout, 3.9; upper jaw, 2.95; least depth of caudal peduncle, 4.35 in head from tip of snout to end of long opercular membrane. The dorsal contour rises in a very gently concave curve from the tip of the sharply pointed snout to the origin of the dorsal, the curve being barely broken by the premaxillary processes; the contour is then concave between the origins of the two dorsal fins, and from the origin of the second dorsal to the upper edge of the deep compressed caudal peduncle; the ventral profile is a gentle curve from the tip of the strongly projecting mandible to the caudal fin. The width and depth of body are about equal at base of 282 MEMOIRS OF THE CARNEGIE MUSEUM. pectorals, but behind this point the body becomes more and more compressed, while before this point the head becomes flatter toward the greatly depressed snout, which is only half as high as wide at front of orbit. Orbit directed upward more than outward. As viewed from above the margin of the snout forms an arch wider than a semicircle, evenly rounded, except at the truncate tip, and broken posteriorly by two very long strong spines, of which the posterior is the lower. The suborbital ridge forms a wing-like edge, armed with three or four huge, sharp-edged spines in line with the even larger spine at the preopercular angle; above the spine the preopercular margin is weakly concave, below the spine it is strongly convex and entirely smooth; the interopercle, but not the preopercle, ends in a spine; the two faint diverging opercular ridges end in weak spines, the upper the longer and the stronger; the nasal spines are closely approximate and not strong; the edge of the flattish interorbital is armed with six moderately elevated spines, not counting one in the front and two on the posterior orbital margin; of the latter two the lower is near the anterior end of a ridge, which ends in a single or double spine at the upper end of the preopercular margin ; the parietal spine is weak, terminating a very .faint ridge; a ridge borders the lower edge of the acute scapular process and ends in a strong spine; behind this a few of the most anterior scales of the lateral line are weakly armed. Opercular region flat- tish; viewed from in front the head is hexagonal in outline, with the dorsal and lateral margins about twice as wide as the two lateral faces of each side. Maxillary extending to below front of pupil; its upper edge ensheathed by the sharp lower edge of the narrow suborbital; its posterior margin strongly emarginate, with the lower angle produced ventrad and mesad as a rounded lobe. The minute teeth form moderate bands on the jaws; posteriorly the entire premaxillary, anteriorly the entire mandibular band, are exposed; the vomerine teeth form two narrow lobes widely divergent from a common base at front of vomer; the palatine teeth, also small, are on a greatly elongate, elevated ridge, ending anteriorly opposite the posterior ends of the vomerine band. The gill-opening is free to below front of orbit; seven branchiostegals ; shoulder-girdle forming a sharp bony ridge; slit behind last gill-arch not as wide as pupil; gill-rakers 6-1-11, counting a few rudi- ments above and below. Dorsal spines strong, diverging at wide angles anteriorly; first dorsal evenly rounded in both directions from the fifth spine, which is con- tained 2.6 times in the head; both first and last spines small. Dorsals separated by an interspace about half as long as orbit. First spine of second dorsal strong and nearly as long as the first and longest soft ray (which is branched) , longer and much heavier than the second spine, contained 3.25 times in the head. Anal spines JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 283 strong and divergent, when depressed; second anal spine more than twice as long as the first, longer than the third (and also stronger), about two-thirds as long as the second, or first branched ray of the five soft rays, and contained 4.6 times in the head. Pectoral reaching to above anus, its length being contained 1.65 times in head; its lower margin nearly parallel with, and about half as long as, upper, when the fin is not stretched; lower rays simple, but not stronger, thickened, nor detached. Ventral fin lying flat and horizontal along the lower pectoral margin; the moderately strong spine is contained 1.8 times in total length of the fin, which measures 2.2 times in head; the fin does not reach the anus, which is in advance of anal fin a distance contained 2.2 times in orbit; base of ventral a little in advance of that of the pectoral, as in Bembras. The caudal fin very slightly and sym- metrically rounded. Scales large, evenly arranged, and rounded, their margins very weakly ctenoid. On the head the scales extend forward dorsally to the nostrils, being uniserially arranged along the interorbital laterally to the orbit, and ventrally to below the end of the maxillary. Lateral line nearly straight from its origin at upper end of gill-opening to the middle of caudal peduncle, thence between the seventh ray from the bottom and the ninth from the top, to the end of the caudal fin. Body pale amber-color, with dark scale-margins dorsally, but otherwise un- marked; fins whitish. Pyloric caeca present. Eggs minute. No air-bladder. Family BEMBRADID^E. In our opinion the Bembradidce, which are intimately connected with the Scorpcenidce through the Parabembridce, represent a family distinct from, and possibly not even directly related to, the still more aberrant PlatycephalidoB. We cannot therefore agree with Regan, who has united the two groups. 438. [821] Bembras japonicus Cuvier and Valenciennes. Aka-gochi = Bed Kochi. Osaka market (Jordan); Kochi (Wakiya); Misaki (Aoki). We here give a re-description of this rare species: Dorsal, XI, I, 11 ; anal, 14; pectoral, 17; ventral, I, 5; caudal, 14 (12 branched) ; pores 55 in lateral line, slightly fewer than the number of oblique scale-rows. Head, 2.9; depth, about 8 in length to caudal. Orbit, 4.0; interorbital, 15.0; snout, 2.9; upper jaw, 2.6; least depth of caudal peduncle, 6.5 in head. Tip of snout sharp, the premaxillary and its process entering the dorsal contour, wLich from behind the process ascends in a gentle curve (broken only by the superorbital serrations and the parietal spine) Regan, Ann. Mag. Nat. Hist. (8) XI, 1913, pp. 171, 177. 284 MEMOIRS OF THE CARNEGIE MUSEUM. to origin of dorsal, from which it descends gradually to the moderately attenuate caudal peduncle; ventral contour nearly straight. Head slightly depressed, its length at occiput 1.2 in the greatest width, its depth opposite front of orbit 1.3 in width at the same vertical; orbit directed equally outward and upward. No nasal spines; snout subspatulate when viewed from above, the sharp lateral margin of preorbital with three spines increasing in strength backward, the last concealing the upper edge of the maxillary; the sharp suborbital stay also entering the lateral profile, armed by four long spines directed backward, the first two below the margin of the pupil, the third below posterior orbital rim, and the fourth at front of preopercle; strongest preopercular spine in line with these suborbital spines, its length measured from front of preopercle two-fifths that of eye; two somewhat smaller and flatter spines directly below the strongest one; from the lower of these two the preopercular margin, armed by one or two obtuse angles, curves abruptly forward; one interopercular, one preopercular, and two opercular spines, forming a series increasing in strength upward; two strong opercular spines terminating two keeled ridges arising together near front of opercle, a strong spine on front of orbit, and seven to ten oblique serrations along the parallel sides of the deeply concave, three-grooved interorbital; a ridge from eye to upper edge of preopercle terminating in a spine followed by a similar spine; three ridges converging to spiny tips at origin of lateral line; the five anterior scales of the lateral line bearing spined keels ; top of snout ridged ; opercular region and cheeks below the suborbital ridge tumid; maxillary extending a little beyond the front margin of orbit; its wide posterior edge emarginate; jaws about equal anteriorly, but the mandibles narrower than the snout, so that the band of premaxillary teeth is almost entirely exposed in ventral aspect; outermost premaxillary teeth slightly enlarged; mandi- bular teeth somewhat smaller and in a narrower band than those on the pre- maxillary; vomerine teeth small, in two widely divergent bands connected anteriorly by a narrow half-ring of teeth; palatine teeth similar, on a very narrow, sharply elevated, and greatly elongate ridge; no mandibular barbel; gill-opening extending forward to below end of gape; seven branchiostegals ; edge of shoulder-girdle with an oblique bony ridge; last gill-slit about as wide as pupil; gill-rakers slender at angle, 1-1-9 in number, not counting bare rudiments, which do not grade into the developed raker. No pyloric coeca. Dorsal spines heteracanthous, but not widely divergent, high, the first 3.5 the third and highest 2.2 in head, the others graduated to the last, which is well separated by membrane from the first soft ray; spine at front of second dorsal, 4.9 in head; caudal barely emarginate, the upper edge longer than the lower. Pectoral with a produced lobe and a rounded lower margin. JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 285 composed of seventeen rays, of which the lower five are unbranched and somewhat thickened distally; ventral fins lying flat and horizontal against the body next to the lower edge of the peduncle, and curved, not reaching the anus; base of ventrals a little in advance of that of pectorals; length of pectoral, 1.7 of ventral, 2.0 in head. Anus close to anal fin. Scales sharply ctenoid, on the head extending forward only to the orbit, covering all of the opercles, excepting the preopercular rim and the cheeks below the suborbital stay. Body pale yellow, with a trace of a darker lateral streak; the upper sides anteriorly and the lateral line are marked with blackish spots; dorsal fins marked with round black spots; anal clear in one specimen, but in another with a black blotch at the base of each interradial membrane; pectoral streaked with brown medially; caudal with upper edge dark and with a broad subterminal bar rounded off ventrally and there intensified to form a large and conspicuous black spot. Young, 10.5 cm. long, have the color-markings more intense, the black spots generally distributed over the upper two-thirds of the body and head, becoming- faint on the snout; a dark cross-shade under the middle of each dorsal fin, and at the base of the caudal; spinous dorsal with a large dark blotch on its anterior half, while the posterior half is spotted like the second dorsal; anal and ventrals immaculate; the caudal, in addition to the markings described for the adult, with two darker spots near base and a black spot on lower margin; pectoral banded with brown spots above, and heavily marked with black dashes ventrally, the lowest two rays clear. Lateral line spiny at front, running not far above middle of sides of body, but rather high on front of caudal peduncle; then near base of caudal bent downward, and running out to end of caudal along upper edge of the fifth main caudal ray as counted from the lower edge of the fin. Family PLATYCEPHALIDiE. Key to Japanese Genera of Platycephalid.e. a. Head moderately depressed, with strong ridges and high sharp spines; only one enlarged si)ine on preopercular margin; vomerine teeth in two parallel longitudinal bands“^; palatine teeth in a band. b. {Onigociince subfam. nov.) Preopercle without an antrorse spine. c. Side of head unicarina'te; infraorbital ridge armed by close-set serrations; a rather strong antrorse spine on preorbital margin; orbit with a developed cirrhus; scales large, the pores in the lateral line fewer than 40. d. Scales of lateral line with weak basal keels anteriorly, but everywhere without spines; antorbital margin without small serrations, armed by a single spine; posterior half only of superciliary ridge serrate; ventral fins of moderate length, not extended to origin of anal. Onigocia}^^ {macrolepis) . The two bands are joined together anteriorly (as an abnormal variation ?) in one specimen of Inegocia japonica). Jordan and Thompson, Proc. U. S. N. M., XLVI, 1913, p. 70. 286 MEMOIRS OP THE CARNEGIE MUSEUM. dd. Scales of lateral line strongly spinecl anteriorly, weakly spined posteriorly; antorbital margin with serrations in addition to a strong spine; entire length of superciliary ridge serrate; ventral fins elongate, reaching beyond anal Wakiyus'-^^ (spinosus)^ cc. {Inegociinw, subfam. nov.) Side of head bicarinate; infraorbital ridge armed by well-spaced serrations; no strengthened antrorse spine on preorbital; orbit without a cirrhus; scales of moderate to small size. Scales of lateral line largely smooth, becoming weakly spined toward the head. e. All minor ridges of head denticulate or granulate; lowermost of the three preopercular spines strong, and persistent at all ages. Teeth villiform, non-depressible, in narrow bands; infraorbital ridge with four to six scarcely differentiated spines between preorbital and preopercular spines; a well defined ridge between orbit and occiput; inner edge of pre- maxillary produced inward and backward as a thick inflexible lobe; opercular margin with a membranous flap below preopercular spines; no orbital cirrhus. Insidiator^^^ (meerdervoorti) . ee. Ridges of head devoid of fine denticulations or granulations; lowermost of the three pre- opercular spines small, becoming obsolete with age. /. Teeth villiform, little specialized, becoming granular with age (especially in Inegocia); infra ridge with but two spines behind one on preorbital, the second turned upward; a well-defined ridge between orbit and occiput; inner edge of premaxillary dilated inward and backward as a thick inflexible lobe. g. Opercular margin with a membranous flap below preopercular spines; spine near center of preorbital obsolete (very rarely developed on one side); main preopercular spine short, about one-third as long as orbit, or shorter; twelve soft rays in dorsal and anal. ' Inegocia}^^ (japonica). gg. Opercular margin without a membranous flap; a sharp spine constantly present on preorbital; main preopercular spine of moderate length, about tw'o-fifths as long as orbit; eleven rays in dorsal and anal Coa'«s“® (crocodilus) . //. Teeth highly specialized, resembling those of a Synodus; those of upper jaw canine-like and depressible in a wide lobe anteriorly, minute and granular, except on innermost row on sides of jaws; vomerine teeth few, enlarged, very sharp and depressible; palatine teeth sharp, enlarged, and depressible along innermost row; infraorbital ridge with numerous differentiated spines, one on preorbital, two below front of eye (the posterior one en- larged), three below posterior part of orbit, the last very strong, and turned upward, and followed by three small spines; no continuous ridge from orbit to occiput; inner edge of premaxillary expanded inward, but not backward, as a thin and flexible lobe. Ratabidus^^^ {megacephalus) . bb. {Rogadiinoe subfam. nov.) Preopercle with a very strong antrorse spine on lower margin; orbit without cirrhus; ridges of head armed by close-set serrations or granulations; sides of head uni- carinate; opercular margin without membranous flap Rogadius (asper). aa. { Platycephalince) . Head greatly depressed, with feeble ridges and spines; two enlarged spines on preopercular margin; vomerine teeth in a transverse bilobed band; palatine teeth uniserial; opercular margin with a membranous flap Platycephalus (indicus), Jordan and Hubbs, new genus (Type: Platycephalus spinosus Temminck and Schlegel). Jordan and Snyder, Proc. U. S. N. M., XXIII, 1900, p. 368. Jordan and Thompson, loc. cit., p. 70. Cocius Jordan and Hubbs, new genus; Orthotype, Platycephalus crocodilus Tilesius. Ratabulus Jordan and Hubbs, new genus; Orthotype, Thysanophrys megacephalus Tanaka. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 287 439. [816] Onigocia macrolepis (Bleeker). Anesag()clii = ¥AdiQY Kochi; Om'f/oc/w = Devil Kochi. Misaki (Aoki) . 440. [815] Wakiyus spinosus (Temminck and Schlegel). This is the only Japanese species of the family not represented in the present collection. We have examined material previously recorded. 441. [818] Insidiator meerdervoorti (Bleeker). ilfe^/oc/ii = Big-eye Kochi. Wakanoura (Yamamoto); Toba market (Jordan and Yamamoto); Kobe and Yokohama markets (Jordan) ; Mikawa Bay (Ishikawa) ; Toyama (Yoshizawa) ; Miyazii, Noo. Dorsal soft rays, 10 to 12, usually 11; anal, 11. The closely related Insidiator detrusus represents this species on the Chinese coast. We have re-examined the type of that species. 442. [817] Inegocia japonica (Tilesius). Tokage-kochi = LizsiYd Kochi. Wakanoura (Yamamoto); Tokyo and Kobe markets (Jordan); Kagoshima Bay (Wakiya); Misaki (Aoki); Miyazu. 443. [819] Cocius crocodilus (Tilesius). /nc^ocM = Rice Kochi. Tokyo and Kobe markets (Jordan); Toba (Jordan and Yamamoto); Mikawa Bay (Ishikawa); Fukuoka (Hamada); Miyazu; Noo. 444. Ratabulus megacephalus (Tanaka). Haname-gochi ^Flower-eye Kochi. Kagoshima Bay (Wakiya). Mr. M. Kasawa has given us the following translation of Tanaka’s original description, published in Japanese: ‘Head, 2.56 in length of body without caudal fin; depth 8.5. Eye, 5.2 in head; interorbital space, 13; snout, 3; maxillary, 3.33; depth of caudal peduncle, 8.66; depth of head, 4.66; width of head, 2. Dorsal, IX-11; anal, 12; pectoral, 20, seven lower rays unbranched; caudal, 13 branched rays. Scales 71 from gill- opening to base of caudal counting downward and backward; 98 counting down- ward and forward; scales in transverse series 15-31 counting downward and backward, 10-23 counting downward and forward; preopercle with two spines; no tentacle on eye. Caudal fin somewhat rounded; pectoral reaching to middle of length of ventral; ventral to base of anal. Interorbital width 2.5 in diameter of eye; eye 1.75 in snout; snout 1.5 in postorbital. 288 MEMOIRS OF THE CARNEGIE MUSEUM. Color dark gray, the head, body, dorsal, pectoral, ventral, and caudal fins with small black spots; the spots a little larger than elsewhere on the first dorsal and caudal; front margin of first dorsal with a dark blotch; margin of caudal with dark streaks (not spots) ; anal fin without streaks, but with a dark margin pos- teriorly, and a pale white margin anteriorly. The type was found in the Tokyo market, where the species is not uncommon. Length 313 mm. from tip of snout to base of caudal. It is characterized by its long head, the proportionate width and depth of the head and the long snout.’ 445. [820] Rogadius asper (Cuvier and Valenciennes). M atsuha-gochi = VmQ Kochi. Kagoshima (Wakiya). 446. [814] Platycephalus indicus (Linnaeus). Kochi] Makochi^Ti'WQ Kochi. Tokyo market, Mikawa. Fukuoka, Osaka, Kobe, Yokoyama, Toba. Every- where common in southern Japan. The largest species, constantly in the markets. We have also a specimen from Soo-chow, China, sent by Dr. Cora D. Reeves. Family HOPLICHTHYIDTl. 447. [824] Hoplichthys langsdorfi (Cuvier and Valenciennes). Natsuhari-gochi = '^ummQY Needle-kochi. Kagoshima Bay (Wakiya); Misaki (Aoki). Jordan and Thompson have given a useful analysis of the Japanese species of Hoplichthys. Family TRIGLIDtE. 448. [827] Chelidonichthys kumu (Lesson and Garnot). = Gurnard. Tokyo and Osaka markets (Jordan); Toba (Jordan and Yamamoto); Tatoku Island (Mikimoto); Mikawa Bay (M. Ishikawa); Choshi (C. Ishikawa); Toyama (Yoshizawa); Misaki (Aoki); Miyazu. Dorsal IX-15 or 16; anal, 14 or 15 (the last ray doubled). Interorbital width, 6.4 to 7.4 in length of head. 449 [829] Lepidotrigla alata (Houttuyn). = Gurnard. Tokyo, Shizuoka, and Osaka markets (Jordan). Dorsal, IX, 15 or 16; (last branched); scales of lateral line 60. Inner surface of pectoral fin violet blackish, except on the uppermost and three lowermost rays, which are whitish; a large lemon-yellow blotch medially. A ripe female is 163 mm. long to caudal. JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 289 450. [830] Lepidotrigla giintheri Hilgendorf. Kanado. Misaki (Aoki); Tokyo market (Jordan); Toyama (S. Yoshizawa); Fukui. Dorsal, VIII-15; anal, 14 (rarely) or 15 (the last ray as double); the last dorsal spine very small in one specimen. Inner surface of pectoral fin red at base, yellowish on the first upper and lowermost three rays, dusky or olive elsewhere, becoming blotched dorsally and inky black, with fewer or more bright blue streaks, ventrally. Young, about 6 cm. long to caudal, have the head rougher, with stronger spines, but show the specific characters as well as do the adults. They show an irregular double dark band under each dorsal fin, and a basal and broader sub- terminal band on the caudal. 451. [832] Lepidotrigla japonica (Bleeker). Toge-kanagoshira^ Thorn Gurnard. Yokohama and Tokyo markets (Jordan) ; Toba (Jordan and Yamamoto) ; Wakanoura (Yamamoto); Kagoshima Bay (Wakiya); Mikawa Bay (M. Ishikawa). One hundred and twenty specimens in all. Very common southward. Scales about 55 to end of last vertebra. The rostral lobes are strongly spinous in many individuals, particularly large ones. Color red in formalin, with a greenish tinge dorsally. 452. [833] Lepidotrigla strauchi Steindachner. Kanagashira = Gurnard . Tokyo and Osaka markets (Jordan); Toba (Jordan and Yamamoto); Toyama (S. Yoshizawa); Misaki (Aoki); Miyazu. With age the dorsal spot, black in the young, becomes faint. Inner surface of pectoral dusky, except on the first upper and lowermost three rays, which are yellow; the dusky color often becoming black on a blotch located medioventrally where the rays show bluish white spots. 453. [834] Lepidotrigla abyssalis Jordan and Starks. Misaki (Aoki). Dorsal rays, VIII, 14 or 15; anal, 14 or 15. The uppermost detached pectoral ray may not reach the ventral tip, though not falling so far short as in L. strauchi. The anterior profile of the snout, as seen from above, may be gently concave. Inner side of pectoral smutty black, unspotted, except on the uppermost ray and lowermost three rays.* 290 MEMOIRS OP THE CARNEGIE MUSEUM. 454. [831] Lepidotrigla kishinouyi Snyder. Misaki (Aoki). A single specimen agreeing well with a paratype. Dorsal, VIII-16; anal, 16, counting the last ray as double (the paratype has VIII-16 dorsal and 15 anal rays). Inner side of pectoral fin dusky, except for the pale uppermost and lowermost three rays, becoming black in a bright blue spotted blotch ventrally. We do not feel certain of the relationships or even the validity of this species. Family DACTYLOPTERTDA5. 455. [840] Daicocus peterseni (Nystrom). H oshi-semi-hobo = Star Cicada-gurnard. Misaki (Aoki). Our eleven young specimens seem referable to this species; 52 to 76 mm. long to caudal fin. The head is contained 2.9 to 3.1 times in the standard length; the snout, eye, and postorbital are each about equal. The body is spotted as in the adult, but is further marked with broad diffuse bars below each dorsal fin. Family AGONIDdE. Key TO Genera Confused with Occa. a. Dorsal spines 7 to 10; pectoral fin narrow, rounded (with 14 or 15 rays); ventral fins of male enlarged, the rays thickened, and provided with free keel-like membranes on their outer edges; snout narrow (approaching that of Brachyopsis), a little wider at base than long, as measured to tip of mandible. b. Dorsal spines 7 to 9; anal fin short, having only 10 or 11 rays;^^^ ventral fins of male greatly elongate, reaching past front of anal; plates more strongly spined, those of the dorsal as well as lateral series with sharp spines; suborbital stay with a spine; vertebrae 37 Occa {verrucosa). bb. Dorsal spines 9 or 10; anal fin longer, of 13 to 15 rays; ventral fins of male not very much larger than those of female, not nearly reaching anal; plates much smoother, especially toward belly and caudal base, those of the dorsal series scarcely spined; suborbital stay without a spine; ver- tebrae 39 Ocella (dodecaedron) . aa. Dorsal spines 12 to 15; pectoral fin broad, subtruncate above (with 18 or 19 rays); ventral fins of male not enlarged, without keel-like membranes on outer edges; snout much broader than long, as measured to tip of mandible; anal rays 15 to 17. c. Dorsal spines usually 12 (sometimes 13) ; suborbital stay with a sharp spine; suborbital bicarinate; frontal and parietal prominences connected b}" a continuous ridge; supraorbital ridge forming a shelf somewhat overhanging the surrounding groove Iburina (iburia), cc. Dorsal spines thirteen to fifteen; suborbital stay spineless; suborbital unicarinate; frontal and parietal prominences entirely separated; supraorbital ridge low, not overhanging the sur- rounding groove Iburiella (kasawce)^ The last double ray counted as one as in all descri])tions in this pajicr. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 291 OccELLA Jordan and Hubbs, gen. nov. Type: Agonus dodecaedron Tilesius. This genus, and the others which follow, are defined in the key given above. 456. [846] Occella dodecaedron (Tilesius). Shachi-uwo = Csipstsin-iish. Iburina Jordan and Hubbs, gen. nov. Type: Occa iburia Jordan and Starks. 457. [847] Iburina iburia (Jordan and Starks). Yori-shachi-uwo = Twisted Capstan-fish. Neither this species, nor the preceding, of which we have material at hand, is represented in the present collection. Iburiella Jordan and Hubbs, gen. nov. Type: Iburiella kasawce, sp. nov. This genus combines certain of the characters of Occella and Iburina, with which it is compared above in the key. In the length of the spinous dorsal fin it approaches Tilesina, which genus we now think may be referred to the same subfamily ( Brachyopsince) . 458. [847A] Iburiella kasaw® Jordan and Hubbs, sp. nov. (Plate XI, fig. 1). Type a fine specimen 133 mm. in length, seined by Snyder and Sindo at Tomakomai, near Mororan, Japan, in 1906. It was taken with a series of Iburina iburia, already recorded by Snyder. The type is in the Carnegie Museum (Cat. of Fishes No. 7906). A single paratype, 85 mm. long, was collected by Tanaka at Kushiro, Hokkaido, in 1922. Dorsal, XV (XHI)^^^-9 (8); anal, 15 (16)'^^; pectorals, 18; ventrals, I, 2; caudal 13 (counting small rays). Pores in lateral line, 43, the first twenty armed with spines. Head (from tip of snout to end of opercular membrane), 4.3 (4.0) in standard length. Greatest depth (between spines of dorsal and ventral series) 1.7 (2.4) in head; depth of caudal peduncle, 7.6 (8.0); eye, 6.0 (6.6); orbit, 5.1 (4.9) equal to snout; interorbital width, 4.1 (4.2); snout and orbit, 2.6 (2.45); width across opercles, 1.2 (1.3); width of snout at base, 2.35; upper jaw, 3.05 (3.0). Body everywhere, except near caudal base, forming in cross-section a somewhat depressed octagon with each face concave. The dorsal and the two lateral series Proc. U. S. N. M., XLII, 1912, p. 436. The counts and measurements in parentheses are those of the paratype. 292 MEMOIRS OF THE CARNEGIE MUSEUM. of plates bear strong spines, which become weak on the caudal peduncle and beneath the pectoral fin; the ventral series comprises plates bearing low rounded keels, each ending in a small point. The spines of the body are marked with strong ridges, which diverge outward and backward from the spines on most of the plates, but which radiate in all directions from the prominences on the tubercles of the caudal peduncle and breast. The plates number 44 (43) in the dorsal series, the last 6 (or 7) being more or less completely joined to their fellows; 38 to 40 in the ipiper lateral series ; 38 or 39 (36 to 39) in the lower lateral series ; 36 (34) paired and 3 (4) median in the ventral series. Breast armed with large rounded flattish tubercles, one of which, median in position, is separated from each ventral fin by two or three others It is preceded by four, arranged in the form of a triangle or square, and surrounded by smaller less definite tubercles; a chain of large bucklers around the gill-opening, giving off a branch to lower end of the pectoral base; an irregular row of small tubercles crowded in between the bucklers of the two ventral series before the anal fin, but no plates surround the anus. All fleshy regions between the bony plates (on the trunk only the tail being completely armoured) covered with numerous thick-set fleshy flaps, which are in part modified into small tubercles ventrally and into small flat spines in the area covered by the pectoral fin. Head completely smooth, on all faces being largely covered with skin thickly studded with small to minute pads, most of which bear small, more or less spiny, points. The radially striate bony prominences on frontal, parietal, and cheek regions not connected with one another, the whole head being relatively free from long trenchant ridges; suborbital stay without trace of spine; super- ciliary ridge greatly depressed, and not dilated as a shelf over the surrounded groove; preopercle with three blunt spines, of which the uppermost is rather strong, the lowermost almost rudimentary; suborbital bearing but one keel; pre- orbital sculptured, with two keels directed downward and two forward, none ending in sharp spines; the strongly convergent nasal spines very weak; the striae on almost all exposed bones and spines of head bearing fine granulations. Lower jaw hooked upward in front of upper, so that the anterior mandibular teeth are exposed, and directed backward as much as upward. Snout relatively extremely broad and short, its length from eye being narrower than interorbital space; its length along midline from opposite front of orbits only three-tenths its basal width. Gill-membranes narrowly joined, but free from isthmus; gill-rakers fleshy tubercles, 1 -|- 7 on first arch. Upper jaw when in place largely ensheathed, extending almost to below front of pupil, bearing a thin flat tentacle near its end. Teeth all weak, forming moderate bands on the jaws, a large transverse patch on the vomer, JORDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 293 and evident on the palatines. Occipital region flattish, interorbital but weakly concave. Pectoral fin rather more sharply truncated than is indicated in the figure. Body brown or slaty, colorless below; back posteriorly crossed by alternate bars of dark and light; fins marked with blackish (brown on spinous dorsal) and clear areas. In the paratype the upper outer margin of the second dorsal is blackish ; the spot on the anal fin extends to the margin of the fin; the caudal is black about the lighter (though dusky) area, which is longitudinally divided into two parts. This species is named for Mr. Masunosuke Kazawa of Sapporo, a graduate student of Stanford University, engaged in the study of the fishes of the Hokkaido. 459. [848] Brachyopsis rostrata (Tilesius). Shichiro-uwo. A series from Kushiro, Hokkaido (Tanaka). 460. [850] Pallasina eryngia Jordan and Richardson. = Goat-fish. Toyama, (S. Yoshizawa). Our specimen has the barbel as remarkably long as in the types. 461. [852] Draciscus sachi ’Jordan and Snyder. *8/wzc/w = Capstan. Kushiro (Tanaka); Noo. Dorsal rays VIII or IX-14 (the last one double); anal, 16 or 17. 462. [863] Sarritor leptorhynchus (Gilbert). Fukui. Family LIPARID^. 463. [895] Crystallias matsushimse Jordan and Snyder. Abachan. A specimen 34.5 cm. long, -collected by Tanaka at Kushiro, Hokkaido, agrees in most respects with Gilbert and Burke’s redescription of Crystallias matsushimce,^^* but differs in the much smaller size of the head, and of the various parts of the head in reference to the total length. This difference is doubtless to be accounted for by the veritably huge size of our specimen. Measurements in hundredths of length without caudal (34.5 cm.). Length of head, .22; diameter of eye, .04; length of snout, .085; interocular width, .095; preoral length of snout, .05; length of gill-slit, .045; distance from tip of mandible to front of disk, .07; to anus, .175; diameter of disk, .085; depth of body, .325; longest pectoral ray of upper lobe, .18; of lower lobe, .095; length of caudal, .14; length of attachment between caudal and anal, .09. A smaller specimen from the Sea of Japan, taken by Nonaka at Fukui, agrees better with the description of the type. Proc. U. S. N. M., XLII, 1912, p. 376. 294 MEMOIRS OF THE CARNEGIE MUSEUM. Family ECHENEIDtE. Genus Remorina Jordan and Evermann. Body unusually robust, the ventral contour strongly curved; lower jaw without a produced flap; pectoral fin broadly rounded, scaly, thick and firm, especially on upper edge, but not bony; dorsal and anal fins small, the anal the longer; caudal rounded; disk unusually large, with twelve or thirteen laminae. 464. [902] Remorina albescens (Temminck and Schlegel). Shir o-koban = White Koban, or Remora. Echeneis albescens (Temminck and Schlegel), Fauna Japonica, Pisces, 1850, p. 272, pi. CXX, fig. 3. — Gunther, Cat. Fishes Brit. Mus., II, 1860, p. 377. — Jordan and Evermann, Bull. U. S. N. M., XLVII, pt. 3, 1898, p. 2272. Echeneis clypeata Gunther, Ann. Mag. Nat. Hist., (3) Vol. V, 1860, p. 401; Cat. Fishes Brit. Mus., II, 1860, p. 376. — Tanaka, Fig. Desc. Fishes Japan, XVIII, 1914, pi. 90; XIX, 1915, p. 334. One specimen of this very distinct Echeneid was obtained by Aoki at Misaki. It is 118 mm. long to caudal fin. The body is pale, with very irregular dark markings, rather thickly set. Genus Remoropsis Gill. Body moderately slender, but not greatly attenuate as in Echeneis naucrates; lower jaw without a specialized flap; pectoral truncate, its rays flexible; dorsal and anal fins long, the dorsal especially so, having about thirty rays; caudal truncate; disk of moderate size, with about 16 (14 to 17) laminae. 465. [903] Remoropsis brachyptera (Lowe). Kuro-koban = Black Kohan. One specimen; Misaki (Aoki), Laminae 16. Genus Remora Gill. 466. [904] Remora remora Linnaeus. /lo5an = Suck-fish. Two young specimens; Misaki (Aoki). Laminae 17 or 18. Family BOTHIDA^. Genus Scoops Jordan and Starks. This genus seems sufficiently distinct from both Psettina and Engyprosopon, differing from the former in having the teeth biserial and from the latter in having the interorbital space greatly widened. JORDAN AND HUBBS! JAPANESE FISHES COLLECTED 1922. 295 467. [909] Scaeops grandisquama (Temminck and Schlegel). Daruma-garei = Daruma^^^-flounder. Kobe market (Jordan). 468. [910] Psettina iijimae (Jordan and Starks). Kagoshima Bay (Wakiya). 469. [911] Laeops lanceolata Franz. Lceops lanceolata Franz, Abh. Bayer. Akad. Wiss., Vol. I, 1910, Suppl. 4, p. 62, pi. 8, fig. 60 (Fukuura; Zushi, on Sagami Bay). — Hubbs, Proc. U. S. N. M., XLVIII, 1915, p. 460. Toba (Jordan and Yamamoto); Fukui (Nonaka). The first two dorsal rays are separated from the rest of the fin, as in Lceops kitaharcB'-^^ and LcBoptichthys Jragilis}^’’ Dorsal rays, 106 to 108; anal, 86 to 88; scales in lateral line, about 98 to 113; head, 6.25 to 6.6; depth, 2.65 to 2.8; eye, 2.9 to 3.2. The specimens agree better with the characters of L. lanceolata than with those of kitaharcE (See Hubbs, 1. c., p. 460), but in some respects are intermediate. It is quite probable that the two species will prove inseparable, in which case Smith and Pope’s name will be used. 470. Arnoglossus tenuis Gunther. Arnoglossus tenuis Gunther, Challenger Reports, Shore-Fishes, 1880, p. 55 (Hong Kong). We have seventeen specimens of an Arnoglossus from the market at Kobe (Jordan), which are very different from the two species of the genus now known from Japan. Arnoglossus violaceus Franz^^* has 100 scales in the lateral line, and Arnoglossus japonicus Hubbs^^® has 64 scales, the teeth enlarged anteriorly, and the second dorsal ray considerably elongated. The material from Kobe we identify with Arnoglossus tenuis originally described by Gunther from Hong Kong. Dorsal rays, 90 to 93; anal rays, 67 to 72; head, 4.0 to 4.1; depth, 2.5 to 2.7; upper orbit, 3.1 to 3.3; upper jaw, 2.6 to 2.7; scales in the lateral line, 48 to 54. Daruma is a small squat demigod. Lambdopsetta kitaharce Smith and Pope, Proc. U. S. N. M., 1906, p. 496, fig. 12. — Jordan, Tanaka, and Snyder, Jour. Coll. Sci., Tokyo, XXXIII, 1910, p. 317, fig. 265. — Lceops kitaharce Hubbs, Proc. U. S. N. M., XLVIII, 1915, p. 460. Hubbs, 1. c., p. 460, pi. 26, fig. 4. Abh. Bayer. Akad. Wiss., Vol. I, 1910, Suppl. 4, p. 61, pi. 7, fig. 56. Proc. U. S. N. M., XLVIII, 1915, p. 454, pi. 25, fig. 2. 296 MEMOIRS OF THE CARNEGIE MUSEUM. Body elongate-elliptical, the gentle dorsal contour being broken only by the pre- maxillary processes. Lower eye in advance of upper; interorbital narrow and trenchant, its sigmoid ridge forming also the front margin of the lower orbit; nostrils in a horizontal line, the anterior one bearing a short, broadly rounded flap; jaws about equal anteriorly, the mandibular knob projecting a little; gape moderate, curved and oblique; upper jaw extending backward beyond front of lower orbit, but not to below pupil. Gill-rakers represented by one or two bare rudiments on upper limb, and by seven to nine developed ones on lower limb, the longest about one-fourth as long as orbit. Scales very deciduous, none remaining on our specimens. No color-markings, other than the dark margins of the scale-pockets. Our specimens vary in length from 37 to 62 mm. A still smaller specimen, 27 mm. long to caudal, was found in the stomach of a Platycephalid taken at Kobe. Family PARALICHTHYIDAH. 471. [916] Pseudorhombus oligodon (Bleeker). Kobe market (Jordan). Dorsal, 74 to 78; anal, 64 to 66; pores, 85 to 88; gill-rakers, 2-|-8 = 10; h*ead, 3.3; depth, 2.2. , 472. [917] Pseudorhombus arsius (Buchanan Hamilton). Shizuoka (Jordan); Mikawa Bay (M. Ishikawa); Fukui (Nonaka). Dorsal rays, 76 or 77; anal, 58 to 60; pores, 78 to 82; gill-rakers, 11 or 12 below angle; depth, 2.1 to 2.2. Body marked rather indistinctly with non-ocellated blotches, one on lateral line just behind end of arch, and sometimes another near middle of straight portion of lateral line; one near caudal base; bare traces of others above and below the lateral line. In addition to the blotches, fine black specks may be present over the body. Our largest specimen is 268 mm. long to the caudal fin. 473. [918] Pseudorhombus cinnamomeus (Temminck and Schlegel). Ganzo-birame ^Fedse Halibut. Pseudorhombus misakius Jordan and Starks, Proc. U. S. N. M., XXXI, 1906. p. 175, figs. 4, 5. Tokyo (Jordan); Choshi (C. Ishikawa); Noo. Dorsal, 81 or 82; anal, 63 or 64; pores, 80 or 83; head, 3.5 to 3.6; depth, 1.9. The dark spot on the lateral line near the end of the arch is not invariably ocellated. ^'■’“See Snyder, Proc. U. S. N. M., XLII, 1912, p. 439. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 297 474. [919] Pseudorhombus ocellifer Regan. Shizuoka, Osaka, and Kobe markets (Jordan); Kagoshima Bay (Wakiya); Misaki (Aoki); Fukui (Nonaka); Toyama (Yoshizawa); Miyazu; Noo. Dorsal rays, 71 to 77; anal, 53 to 57; pores, 64 to 73; gill-rakers, 4 to 7 15 to 18; head, 3.1 to 3.3; depth, 2.1 to 2.2. 475. [921] Tarphops oligolepis (Bleeker). Toba (Jordan and Yamamoto); Kobe market (Jordan); Tatoku Island (Mikimoto) . Dorsal, 64; anal, 49; scales in lateral line, 40. 476. [924] Paralichthys olivaceus (Temminck and Schlegel). = Halibut (Wide-eye); Magarei = Ti'\iQ Flounder. Sapporo market (Majima); Takashima market (Takayasu); Tokyo and Kobe markets (Jordan); Toba (Jordan and Yamamoto); Misaki (Aoki); Mikawa Bay (M. Ishikawa); Choshi (C. Ishikawa); Fukui (Nonaka). Very common southward, the most abundant and valuable of the flat-fishes. In this species the fin-rays are well branched posteriorly. Dorsal rays 70 to 78; anal, 54 to 60. To the synonymy of this species should apparently be added Platessa per- cocephala Basilewsky, Pseudorhombus swinhonis Gunther, and Paralichthys olivaceus var. coreanicus Schmidt. A specimen from Hong Kong has 83 dorsal and 61 anal rays; the type of P. swinhonis, from Chifu, has 69 dorsal and 51 anal rays. The range of variation in fin-rays in China and Japan consequently appears to be quite similar. It is quite probable, however, that the fin-rays will be found to show an increase in average number toward the north. Family PLEURONECTIDiE. 477. [926] Verasper variegatus (Temminck and Schlegel). Medaka-garei = Mhmoyf Flounder; = Star-flounder. Toba (Jordan and Yamamoto). 478. [928] Xystrias grigorjewi (Herzenstein). Caterpillar flounder. Osaka market (Jordan); Toyama (Yoshizawa); Miyazu. Dorsal rays, 91 ; anal, 77. The markings on the body vary from barely evident blotches to strongly ocellated spots. 298 MEMOIRS OF THE CARNEGIE MUSEUM. 479. [930 and 932] Hippoglossoides dubius (Schmidt). d.6Rra-^am’ = Fat Flounder. Sapporo market (Majima); Kushiro (Tanaka); Osaka market (Jordan), said to have been shipped to Osaka from near Aomori; Noo. Dorsal rays, 79 to 86; anal, 61 to 67; gill-rakers, x+11 to 16. Cynopsetta dubia Schmidt seems identical with Hippoglossoides katakurcB Snyder and to be referable to Hippoglossoides. 480. [937] Cleisthenes pinetorum Jordan and Starks. Noo; Fukiii (Nonaka). Mr. Hubbs has lately shown that Proto psetta herzensteini is congeneric with Cleisthenes pinetorum of the east coast of Japan, representing that species in the Sea of Japan, and differing chiefly in the number of gill-rakers. The material at hand confirms these findings, the gill-rakers before the angle of the first arch numbering in seven specimens: 17, 17, 18, 18, 18, 19, 19. Protopsetta is ap- parently a synonym of Cleisthenes. 481. [934] Atheresthes evermanni Jordan and Starks. Kushiro (Tanaka). Dorsal rays, 107; anal, 86 or 87; head, 3.7; depth, 2.8. 482. [938] Alaeops plinthus Jordan and Starks. Misaki (Aoki). 483. [939] Dexistes rikuzenius Jordan and Starks. Migi. Fukui (Nonaka). Dorsal rays, 72; anal, 61; no fin-rays branched; pores in lateral line, about 70. Araias ariommus Jordan and Starks is the same species as already indicated. 484. [940] Pleuronichthys cornutus (Temminck and Schlegel). Meita-garei = Cloak-flounder. Tokyo and Kobe markets (Jordan); Toba (Jordan and Yamamoto); Kago- shima (Wakiya); Mikawa Bay (M. Ishikawa); Alisaki (Aoki); Fukui (Nonaka); Noo. Everywhere abundant, reaching a small size. Dorsal rays, 73 to 81; anal rays, 52 to 60; head, 3.8 to 4.2; depth, 1.8 to 2.3; eye, 3.0 to 3.35. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 299 485. [941 and 942] Lepidopsetta mochigarei Snyder. Mochi-garei = Rice-flounder. Nemuro and Kiishiro (Tanaka). All the records of Lepidopsetta hilineata from the Sea of Japan refer apparently to the present species. Dorsal, 71 to 77; anal, 58 to 61; pores, 100 to 104; gill-rakers, 24-4 to 6; head, 3.3 to 4.0; depth, 2.1. Genus Limandella Jordan and Starks. This group seems fully entitled to generic separation from Limanda] jaws very unsymmetrical, as in most pleuronectine flounders. 486. [943] Limandella angustirostris (Kitahara). Kitsunegarei = Fox-iloundei\ Sapporo market (Majima); Takashima (Takayasu); Fukui (Nonaka); Noo. 487. [944 and 945] Limandella yokohamse (Gunther). Kuragoshira; Makogarei = True Little Flounder, or Dab. Takashima (Takayasu); Nemuro (Tanaka); Tokyo and Kobe markets (Jordan); Mikawa (M. Ishikawa); Fukui (Nonaka). Limanda schrencki seems to be the same species. Dorsal rays, 61 to 73; anal rays, 48 to 54; pores, 70 to 82; gill-rakers, 3-]- 6 or 7 = 9 or 10. The variation in fin-rays in our series covers the supposedly distinctive numbers accredited to the nominal species, L. schrencki. 488. [946] Limanda punctatissima (Steindachner). i^an-(/am = War-flounder. Hippoglossoides (Hippoglossina) punctatissimus Steindachner, Ichth. Beit., 1879, p. 49 (Situngsb. Akad. Wiss. Wien., LXXX, 1879). (Hakodate). Limanda iridorum Jordan and Starks, Proc. U. S. N. M., XXXI, 1906, p. 206, fig. 14, Mororan. — Jordan, Tanaka and Snyder, Jour. Coll. Sci., Tokyo, XXXIII, 1913, p. 327, fig. 276.— Hubbs, Proc. U. S. N. M., XLVIII, 1915, p. 484. Takashima (Takayasu); Nemuro (Tanaka). The description of this species by Steindachner, based on a specimen from Hakodate, has been entirely overlooked. Body covered with fine beaded lengthwise streaks, and with small dark spots. Head, 3.2 in length; depth, 2.1; dorsal, 61; anal, 48; scales, 73. 300 MEMOIRS OF THE CARNEGIE MUSEUM. 489. [954] Kareius bicoloratus (Basilewsky). I shigarei = Rock-Rounder. Sapporo market (Majima) ; Takashima market (Takayasu) ; Tokyo and Osaka markets (Jordan); Mikawa (M. Ishikawa); Fukui (Nonaka). Common northward. Young, 24 mm. long to caudal fin, are fully transformed, and show all adult features other than the bony plates. At 39 mm. they show traces of the developing bony plates, but are otherwise naked, while somewhat larger specimens (55 to 150 mm. long) have the body covered with small, thin, imbedded scales, visible to the naked eye. Still larger specimens agree with the current descriptions of the species in lacking evident scales. 490. [955] Clidoderma asperrimum (Temminck and Schlegel). Same-gar ei = Sturgeon-flounder. Kushiro and Nemuro (Tanaka). Dorsal rays, 86 to 91; anal, 66 to 69; head, 3.2 to 3.5; depth, 1.7. Common northward. 491. [956] Microstomus stelleri Schmidt. Ba6a-master~gobj. Shizuoka and Kobe (Jordan); Toba (Jordan and Yamamoto); Mikawa Bay (M. Ishikawa); Fukui (Nonaka). Dorsal rays, VIII-16 or 17; anal, 13 or 14. Mem. Car. Mils., VI, 1914, p. 289. JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922 309 525. [1034] Ainosus geneionemus (Hilgendorf). Toba market (Jordan and Yamamoto). Two adults^ 80 to 82 mm. long to the caudal base, agree well with Jordan and Snyder^s description of half-grown specimens, except that the spot on the posterior part of the dorsal is less prominent. In this species the tongue is not strictly truncate, as described, but is slightly emarginate. 526. [1035] Triffinopogon japonicus Rendahl. Shoki-haze = Summer Goby . According to Rendahl the Japanese species of Trioeno'pogon is distinct from the Chinese Trimnupogon barbatus Gunther. We have but one example of this species, taken by Hamada near Fukuoka. 527. [1036] Sicyopterus japonicus (Tanaka). = Priest-goby (Shaven Head). Five specimens from the River Tonda in Kishu, collected by H. Kuroiwa, Sept. 17, 1922. We have other specimens from near Uwajima in lyo, Shikoku, in which region it is common. Body above anal fin with horizontal rows of spots. 528. [1037] Tridentiger obscurus (Temminck and Schlegel). Chichibu. Ctenogobius atriceps Jordan and Thompson, Ann. Car. Mus., VI, 1914, p. 287 (Nagoya). (Not of Regan.) Lake Togo, Tottori (Inomata); fresh-water at Kumamoto (Wakiya); Kachi River (Jordan); Toba (Jordan and Yamamoto); Mikawa Bay (M. Ishikawa); Himeji (Abe); Tonda River, Kishu (Kuroiwa); Aomori (Beppu); Lake Biwa (Jordan); Morioka, Rikuchu (Awaya); Noo, Lake Mikata, Fukui. We have re-examined the material recorded by Jordan and Thompson from Nagoya as Ctenogobius atriceps, and have found their specimens to be referable rather to Tridentiger obscurus. This little fish is generally very abundant in estuaries. 529. [1039] Tridentiger bifasciatus Steindachner. Shima-haze = Striped Goby. Hachi River at Nagoya (Jordan); Mikawa Bay (M. Ishikawa). 530. [1049] Luciogobius elongatus Regan. Noo, 310 MEMOIRS OF THE CARNEGIE MUSEUM. 531. [1051] Leucopsarion petersi Hilgendorf. A series of this diminutive Goby was collected by Y. Tsuchiga at Isobemura, near Yamada in Ise. Genus T^enioides Lacepede. Two species of this genus occur in Japan. On comparison we find that they agree in having the body naked, the dorsal and anal fins united for their full height to the caudal, the soft rays in each fin more than forty in number, a pair of small mental barbels, and the length of the head contained one and one-half times in the distance between the insertion of the ventral fin and the anus. They differ widely however, in other respects, as the following key will indicate: Key to the Japanese Species of the Genus T^nioides. a. Dorsal fin beginning scarcely behind the moderate pectoral, which is more than two-thirds length of ventrals, which extend half-way to front of anal; head about six times in length to base of caudal (shown too short in Schlegel’s plate), two-thirds length of caudal fin; depth of body 10 in length; dorsal rays VI, 44; anal, 41; mouth very oblique, its angle about forty-five degrees with the vertical. lacepedei. aa. Dorsal fin beginning well behind the very short pectoral fin, which is about two-fifths length of ventral, the latter extending two-thirds distance to origin of anal; head 7.5 in length, longer than the caudal fin; depth seventeen times in length; dorsal rays VI, 46; anal, 44; cleft of mouth almost vertical; cheeks not tumid snyderi 532. [1054A] Taenioides lacepedei (Temminck and Schlegel). We have a specimen of a Tmnioides, which is obviously the species described by Temminck and Schlegel, but certainly not the one called T. lacepedei by Jordan and Snyder. This was collected by Hamada near Fukaoka. Dorsal rays VI, 44; anal, 41 (counting the last ray as a double one). 533. [1054] Tffinioides snyderi Jordan and Hubbs, sp. nov. Warazube = Straw-Goby Twnioides lacepedei Jordan and Snyder, Proc. U. S. N. M., XXIV, 1901, p. 128, fig. 33 (not Amblyopus lacepedei Temminck and Schlegel, which is the species listed above). We are unable to identify this species with any which have been described from East Indian waters. Its characters are intimated in the key inserted above and in Jordan and Snyder’s description. Tcenioides abhotti, described by Jordan and Starks (Proc. U. S. N. M., XXXI, 1907) from Port Arthur, is plainly different, having a much longer pectoral, reaching much beyond the ventrals, a longer head. JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 311 Family PARAPERCID.E. 534. [1055] Parapercis pulchella (Temminck and Schlegel). Tora-gisu = Tiger-gisu. Misaki (Aoki). 535. [1056] Parapercis ommatura Jordan and Snyder. Toba (Jordan and Yamamoto); Kobe (Jordan). 536. [1058] Neopercis sexfasciata (Temminck and Schlegel). Yokohama, Tokyo, and Kobe markets (Jordan); Misaki (Aoki); Toyama (Yoshizawa); Fukui; Shizuoka, where it is used for making fish-curd. Color in formalin: general tone pink on dorsal half of body, yellow on lower sides, grading to whitish below; the colors due to the spots on the center of the scales, the margins of the scales being broadly gray, darker above, pale below. Head with narrow lines of brilliant yellow, breaking up into spots on cheeks and opercles, the most prominent line running from the suborbital edge backward and upward to eye, then backward along the lower orbital margin. Upper lip yellowish posteriorly, becoming dusky toward symphysis. Soft dorsal with three longi- tudinal streaks, the proximal and marginal ones yellow, the lower median orange and the upper median red; the dorsals blackish near base, where the branches of the body-bars extend a short distance upon the fin. Caudal barred upon upper edge with black and white, mottled reddish and dusky medially, becoming blackish toward the extreme white margin ventrally. Anal with the front of each inter- radial membrane yellow. Pectoral mostly pale pinkish, but yellowish toward the broad black mark at base of fin and along the lower edge; ventral pale dusky. 537. [1059] Neopercis multifasciata (Doderlein). Oki-tora-gisu = Off-shore Tiger-gisu. Kagoshima Bay (Wakiya); Misaki (Aoki). Family TRICHODONTIDTl. 538. [1066] Arctoscopus japonicus (Steindachner). Hata-hata^ Grasshopper. Kushiro (Tanaka) ; Toyama (Yoshikawa) ; Noo. The number of dorsal spines, which has been used as one of the chief generic characters of Arctoscopus, varies from nine to fourteen, and hence may be as numerous as in Trichodon. The spines, however, are much more flexible and weaker than in that genus, and the two dorsal fins are much more widely separated. The pectoral rays are also much less thickened, and the dermal folds on the lips much smaller. , 312 MEMOIRS OF THE CARNEGIE MUSEUM. Family URANOSCOPIDiE. Zalescopus Jordan and Hubbs, gen. nov. Type: Zalescopus tosce Jordan and Hubbs. This genus is closely related to Uranoscopus, with which it agrees in nearly all characters, but from which it differs in having the nuchal region densely covered with small imbedded scales, and the suborbitals wider. In Uranoscopus the nape is naked. The known species are uniform in coloration, or nearly so, whereas those of Uranoscopus mostly show a highly variegated color-pattern. Key to the Species of Zalescopus. a. Respiratory valve prolonged into a very slender non-fimbriate, black filament, nearly two-fifths length of head; top of head slightly concave; scapular spines stronger and longer, less tuberculate; preopercular spines four, stronger; dorsal rays, IV-14; scale-rows all directed downward and back- ward tosce. aa. Respiratory valve merely pointed, not prolonged into a retractile filament; top of head (excluding minor concavities) about flat; spines of scapular plate shorter and more strongly tuberculate; pre- opercular spines five or six; dorsal rays, IV-13; several scale-rows below interval between the dorsal fins usually directed downward and forward satsumoe. 539. [1066B] Zalescopus tosae Jordan and Hubbs, sp. nov. (Plate XI; fig. 3.) The type alone is known; it is 138 mm. long to the caudal base, and was collected at Kochi in Tosa on the island of Shikoku by Ybjiro Wakiya. C. M. Cat. Fishes, No. 7945. This species is very distinct from the other Japanese species, with which it is compared in the foregoing key. It corresponds in most respects with descriptions of Uranoscopus crassiceps (Alcock), a species inhabiting considerable depths in the Indian Ocean. It appears, however, to have a smaller head and a shorter retractile filament, and perhaps further differs in the scaly nape (a character not described by Alcock). Dorsal rays, IV-14; anal, 13; caudal, 12 (10 branched) ; pectorals, 18; ventrals, l, 5. Head from middle of margin of upper lip to tip of opercular fringe 2.8 in total length to caudal base; greatest depth of body, 4.15; least depth, 3.45 in head; width of head, 1.15; depth below occiput, 1.45; length of orbit, 5.4; of snout, 6.3; least interorbital width, 5.65; length of interorbital fossa, from front of lip, 3.85; extreme width of gape, 2.35; length of upper jaw, 2.45. Form of head and body and course of contours as in the next species, except that the top of the head is slightly concave. Armature of the head and the surface sculpturing are as de- scribed for Z. satsumce, except that the tubercles are somewhat coarser, the pre- orbital definitely ridged, the preopercular spines only four in number, spines of JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 313 scapular plate freer of tubercles and longer, the largest about as long as orbit, and directed almost straight backward, the humeral spine less deeply grooved, and directed almost as much backward. Upper lip obscurely, lower lip strongly but simply, fimbriate; respiratory valve smooth-edged, produced into a very slender, flat, unfringed retractile filament of blackish color, nearly two-fifths as long as head. Teeth, opercular, branchial, and gular membranes as described for Z. satsumce; the process on the inner edge of the shoulder-girdle much more obtuse. Squamation as in Z. satsumce, except that none of the scale-rows are reversed in direction. Fins as described for Z. satsumce, except that the fourth dorsal ray is half as long as head, the pectoral and ventral rays less thickened. Color uniformly purplish brown over entire body, even on belly behind ventral fins, but much darker above than below; first dorsal jet-black, except for a narrow white base; second dorsal dusky, with a darker streak on each membrane, and traces of spots along the sides of the rays; caudal light dusky with narrow borders, blackish above, and whitish behind and below; pectoral darkish on both sides, except along the exserted tips of the lower rays; anal and ventral each with considerable dark pigment, that on the ventral located in a long indefinite blotch near the inner margin of the fin. 540. [1067] Zalescopus satsumae Jordan and Hubbs, sp. nov. IS irame-okose = Star-Okose. {Okose = Poison-fish) . Uranoscopus oligolepis Tanaka, Fig. Desc. Fishes Jap., XVIII, 1914, pi. 88, fig. 286; XIX, 1915, p. 327 (description and figure, but not name, nor synonymy). Type 166 mm. long to caudal fin, collected by Wakiya in Kagoshima Bay, Province of Satsuma, C. M. Cat. Fishes, No. 7949. A paratype, 182 mm. long, with the same data, is preserved in the Stanford Collection. A specimen 120 mm. long, taken by Aoki at Misaki, probably belongs to the same species. It agrees with the specimen from Kagoshima in all important respects, except that the color is darker and more variegated dorsally, especially toward the nape, where distinct blackish reticulations on a light ground-color may be seen. Zalescopus satsumce seemingly is closely related only to Z. tosce, to which it bears a striking similarity. The two species, however, apparently differ widely in structural features, as indicated in the foregoing key. Dorsal rays, IV-13; anal, 13; caudal 12 (10 branched); pectorals, 18; ventrals, I, 5. Head from middle of margin of upper lip to tip of opercular fringe, 2.65 (2.6) in total length to caudal base; greatest depth of body, 4.1 (3.75); least depth of caudal peduncle, 3.9 (3.5) in head; width of head, 1.12 (1.15); depth below 314 MEMOIRS OF THE CARNEGIE MUSEUM. occiput, 1.6 (1.5); length of orbit, 5.5 (5.6); length of snout, 5.5 (5.6); least inter- orbital width, 5.3 (5.55); length of interorbital fossa from front of premaxillaries, 3.65 (3.8); extreme width of gape, 2.15 (2.2); length of upper jaw, 2.45. Ventral contour nearly straight, rising from pubic spines to chin at an angle of about 120°; dorsal contour very slightly curved from tip of snout to dorsal fin, and along base of second dorsal; head and trunk depressed; tail compressed, becoming thinner toward caudal. Head angulated ventrolaterally, rounded dorsolaterally, flat on top. Bones of head widely exposed, covered with moderately coarse sculpturing, composed of tubercles, which are at once united to form a reticulation, and aligned to form radial ridges; the sutures marked by rather sharp channels; no definite occipital lobes developed; preopercular limb joined with opercle by broad bony ridges; preopercle with five or six spines, and subopercle with one spine at lower margin; preorbital projecting sharply over upper jaw, with coarsely emarginate borders, but without definite ridges (except in smaller specimen from Misaki) ; suborbitals widely expanded, almost covering upper part of cheeks, separated from preopercle by a naked strip, only one-fifth to one-sixth as wide as the orbit; inter- orbital fossa widely dilated, rather than constricted anteriorly, not extended backward to line joining posterior margins of orbits; scapular plate covered with tubercles to, or almost to, extreme tips of spines, of which one to three of small size point inward and backward, the largest one (shorter than orbit) outward and backward; humeral spine long and sharp, smooth, except for longitudinal ridges and grooves, directed backward and a little upward, its length one-half greater than that of orbit. Upper lip obscurely, lower lip rather strongly, but simply, fimbriate; lower (main) respiratory valve smooth-edged, obtusely pointed, not prolonged into a retractile filament. Teeth in jaws pointed, larger in the lower than in the upper jaw, in two series on front of mandible, in one series laterally (where they are few in number) very large, spaced, sharp, triangular, with com- pressed bases; premaxillary teeth in two series widely separated by a band near symphysis and more narrowly by one series on median (anterior) half of sides, converging to meet at end of band; teeth in a patch on each side of vomer and on each side of palatine. Opercular membrane finely fringed; gill-membranes largely free from isthmus, forming a fold in front of the three pubic spines, and covered by a fold of the gular membrane; inner edge of shoulder-girdle with a rounded fleshy lobe. Scales arranged on oblique ridges' (about fifty to fifty-six to caudal base) much as in Uranoscopus, but several rows below interval between dorsal bases directed reversely, that is, downward and forward; ridges continuous and strong on caudal peduncle; entire area between lateral lines dorsally covered JORDAN AND HUBBSI JAPANESE FISHES COLLECTED 1922. 315 by small imbedded scales. Body naked on belly and backward to a line running from base of humeral spine toward the anal origin, and in a strip, narrowing pos- teriorly, along anal base. First dorsal small, extremely flexible, a rounded triangle in outline, narrowly separated from the second or main dorsal, in which the rays are abruptly lengthened to the fourth (or third), which is about two-fifths as long as the head, and about twice as high as the anal; caudal sharply truncate, when half expanded; posterior margin of pectoral truncate above, rounding off gradually below into the lower margin, the fin-margin moderately incised between the thickened tips of the lower rays; first three ventral rays much thickened. Color purplish brown, somewhat puncticulate, but practically uniform in effect dorsally, becoming whitish below (the specimen from Misaki shows dark reticulations above); first dorsal jet-black, except for a clear base, which is very narrow medially; second dorsal light dusky, with olive spots along the edges of the rays; caudal similarly, but less distinctly, marked; pectoral on both faces dusky, with the free tips of the thickened rays whitish, and with traces of dark reticulations just inside this light border; anal and ventrals whitish. Genus Uranoscopus Linnaeus. Key to the Japanese Species of Uranoscopus. a. Respiratory valve smooth, merely pointed, not prolonged into a retractile filament in the adult^^^; interorbital fossa smaller, not extended backward to line joining posterior rims of orbits, and normally more or less constricted anteriorly; suborbitals smaller, the largest little or no longer than orbit; preopercular spines three (a fourth rarely interpolated); head less than one-third the standard length. Occipital lobes well developed; sculpturing of exposed bones of head so fine as to feel velvety to the touch; dorsal rays, IV or V-13 or 14 (usually 14)^^^; scale-rows 40 to 64, (usually more than 50) everywhere aligned along definite folds; upper half of body abruptly darker than lower parts, and marked with more or less rosette-like spots “of rosy or yellowish;” the pattern similar, but finer, on top of head japonicus. aa. Respiratory valve prolonged into a slightly fimbriate retractile filament at all ages; interorbital fossa larger, extending backward to a line joining the posterior orbital margins, and widely dilated anteriorly; suborbitals wider, the largest nearly twice as long as orbit; preopercular spines four (rarely six, in U. bicinclus)-, sculpturing of head coarser, feeling granular to the touch; head more than one-third the standard length. h. Retractile filament broad and flat, shorter than orbit, rarely reduced in adult to a sharp angle; occipital lobes well developed; head only moderately rough; dorsal rays, IV or V-12 or 13; anal, 13; scales larger, the rows 36 to 47, everywhere aligned along definite folds; upper half of sides darker, marked with irregular, but entire {i. e., not rosette-like), “rosy” spots, the pattern ob- solescent on head oligolepis. The young, 7 to 8 cm. long to caudal fin, have a retractile filament of moderate length, pale in color, membranous and foliaceous, which becomes reduced to an angle of the valve at a length of about 9 cm. The last soft ray is divided to base in both dorsal and anal fins, as in fishes generally. Throughout this paper this double ray, having but one interneural, is counted as one. 316 MEMOIRS OF THE CARNEGIE MUSEUM. hb. Retractile filament thick, about twice as long as orbit; occipital lobes obsolescent; head extremely rough; traces of light spots on young only; the body crossed by two blackish bars, dorsally more or less disrupted into streaks; top of head transversely streaked; a large blotch on cheeks. bicinctus. 541. [1068] Uranoscopus japonicus Houttuyn. Teminondae-okose — Observatory-Sciilpin. Tokyo and Osaka markets (Jordan); Wakanoura (Yamamoto); Kochi; Kago- shima Bay (Wakiya); Toyama (Yoshizawa); Misaki (Aoki); Miyazu; Noo. Gen- erally common in the markets. The scale-rows are highly irregular, being often fused or divided, varying from 40 to 64. None of the rows, however, show any tendency to be reversed in direction, as they are in Zalescopus satsumce. 542. [1067] Uranoscopus oligolepis Bleeker. Mikawa Bay (M. Ishikawa); Fukui (Nonaka); Tokyo market (Jordan and Snyder, 1900, not previously recorded). Tanaka has confused this species with Zalescopus satsunm. 543. [1069] Uranoscopus bicinctus Temminck and Schlegel. Megane-uwo — Spectacle-fish. This species was not obtained in our collection made in 1922, but we have examined specimens taken by Jordan and Snyder in 1900 at Tokyo, Misaki, and Wakanoura. 544. [1071] Gnathagnus elongatus (Temminck and Schlegel). Ao-mishirna = Blue Lookout-fish. Miyazu, Noo. Family CALLIONYMIDTl. 545. [1077] Calliurichthys doryssus Jordan and Fowler. Attmerf = Dragonet. Toba market (Jordan and Yamamoto). 546. [1078] Calliurichthys japonicus (Houttuyn). Fome-j/oc/u’ = Bride-kochi. Tokyo market (Jordan) ; Wakanoura (Yamamoto) ; Kagoshima Bay (Wakiya) ; Misaki (Aoki). 547. [1081] Callionymus lunatus Temminck and Schlegel. Numeri-kochi. Tokyo market (Jordan); Mikawa Bay (M. Ishikawa); Misaki (Aoki); Noo; Obama. JORDAN AND HUBBS: JAPANESE FISHES COLLECTED 1922. 317 In females the spinous dorsal may be entirely black; in males the first dorsal spine may be long enough to reach, when depressed, beyond the end of the second dorsal base; the other spines, of which the fourth is longest, are never produced. These small fishes, like the species of Platycephalidw, go in the markets under the general name of Kochi {gochi in composition) . 548. [1082] Callionymus richardsoni Bleeker. Nezumi-gochi = Rsit-kochi. Tokyo, Yokohama, and Kobe markets (Jordan) ; Toba (Jordan and Yama- moto); Mikawa Bay (M. Ishikawa); Fukui (Nonaka). Jordan and Fowler^^^ have transferred the name Callionymus valenciervnie Temminck and Schlegel (altering the spelling to valenciennesi) to the iiresent species, and renamed the true valenciennei as a new species, Callionymus jlagris. The chief distinctive feature of “(7. hagris” namely the produced dorsal spines, is specifically mentioned and figured for valenciennei by Schlegel. The valid name for the present species appears to be C. richardsoni. The name Callionymus punctatus Langsdorff, applied to the species by Richardson, has no standing, having been based upon the following remark by Valenciennes: “Dans son etat de dessication il parait uniformement brun, avec du noir a la dorsale; mais il y a lieu de croire que ce poisson, frais, avait quelques taches ou points, car M. Langsdorff, qui le rapportait au genre des platycephales, lui avait donne Tepithete de punctatus.” In other words, Langsdorff had merely misidentified the fish as Platycephalus punctatus = Cocius crocodilus. In the young of both sexes the spinous dorsal is marked with an ocellus. In the females the ocellus persists throughout life, while in the male it usually dis- appears, being gradually replaced by a black border on the fin. Both ocellus and border exist together in half-grown, and even occasionally in large adult, males. 549. [1083] Callionymus beniteguri Jordan and Snyder. Tobi-numeri = Hawk-Dragonet ; Beni-tiguri — Red Dragonet. Toba (Jordan and Yamamoto); Kobe (Jordan); Tatoku Island (Mikimoto); Mikawa Bay (M. Ishikawa). 550. [1086] Callionymus valenciennei Temminck and Schlegel. Callionymus valenciennei Temminck and Schlegel, Fauna Japonica, Pisces, 1845, p. 153, pi. 78, fig. 3. Proc. U. S. N. M., XXV, 1903, p. 950, fig. 6. 318 MEMOIRS OF THE CARNEGIE MUSEUM. Callionymus flagris J ORDAi \> MU) V»V ‘ U J >V )) )P m ))))))) 1 \V>.O.M>>M))Uu)U)V V. ’ V/; M ) ) M ) M J M M VuV,V, Vj > V.’. V lr^y,,vf,v,v,v,v/.v,>/’: I V, ’, ’ ’, ’, ’, ', ’. ’,'>’> > ’ ’■ '' ?U) m%V V* WiVA' Psychichthrys and Oncorlnjnchiis. 334 MEMOIRS OF THE CARNEGIE MUSEUM. EXPLANATION OF PLATE VI. Fig. 1. Oncorhynchus ishikawce Jordan and McGregor, sp. nov. Type. Lake Biwa at Otsu, Japan. (C. M. Cat. Fishes, No. 7786.) Fig. 2. Oncorhyrichus viacroslovius {GuniXxGv) . Lake Biwa at Otsn, Japan. (C. M. Cat. Fishes, No. 7791.) PhG. 3. Oncorhynchus rnacrostomus (Giinther). Type. Lake Hakone, Japan. (C. M. Cat. Fishes, No. 7790.) MEMOIRS CARNEGIE MUSEUM, Vol. X, No. 2 Plate VI ,?.>JJ-> mm Oncorhynchiis. 336 MEMOIRS OF THE CARNEGIE MUSEUM. EXPLANATION OF PLATE VIE Fig. 1. Oncorhynchus rhodurus Jordan and McGregor, sp. nov. Type. Lake Hakone, Sagami, Japan. (C. M. Cat. Fishes, No. 7794.) Fig. 2. Salvelmus pluvms (Hilgendorf). Shinshu near Nagano, Japan. (C. M. Cat. Fishes, No. 7796.) Fig. 3. Salvelmus irnbrius Jordan and McGregor, sp. nov. Type. Hamada, Iwami, Japan. (C. M. Cat. Fishes, No. 7797.) MEMOIRS CARNEGIE MUSEUM, VoL X., No 2 Plate VII .WS’.V,',’ Wmmmm^ tiwm /vAV"*’ V/'r Oncorliynchus and Salvelinus. 338 MEMOIRS OF THE CARNEGIE MUSEUM. Fig. 1. Fig. 2. Fig. 3. Fig. 4. Fig. 5. Fig. 6. Fig. 7. Fig. 8. Fig. 9. Fig. 10. Fig. II. Fig. 12. Fig. 13. Fig. 14. Fig. 15. Fig. 16. Fig. 17. Fig. 18. EXPLANATION OF PLATE VIII. (Scales of Japanese Salmonidie.) Oncorhynchus rhodurus Jordan and McGregor, Lake Hakone. Oncorhynchus rhodums Jordan and McGregor, Lake Hakone. Oncorhynchus gorbuscha (Walhanm), Hokkaido. Oncorhynchus adonis Jordan and McGregor, Lake Hakone. Oncorhynclms hawaniurcc Jordan and McGregor, Ugo. Oncorhynchus ishikau'w Jordan and McGregor, Lake Hakone. Ilucho perryi (Brevoort), Naoetsn, Echigo. Oncorhynchus macrostomus (Giinther), Shihukavva. Oncorhynchus niacrostonms (Giinther), Shibnkawa. Plecoglossi(s allivelis Temininck and Schlegel, Korea. Oncorhynchus keta (Walbaiim), Sapporo. Plecoglossusf allivelis Temininck and Schlegel, Kumamoto. Salvelinus leucoma’nis (Pallas), Petropavlovsk. Salvelinus phivius (Hilgendorf), Shinshu. Salvelinus imbrius Jordan and McGregor, Iwame. Salvelinus malma (Walbaum), Pt. Hope, Alaska. Salvelhiiis spectabilis Girard, Montana. Oncorlnjnchus forrnosanus (Jordan and Oshima), Saramoa, Formosa. MEMOIRS CARNEGIE MUSEUM, VoL X, No. 2. Plate VIII. Scales of Japanese Sahnonldoe. ■ i .'.'j fj . • . /v. ’-V /l■^- 'V1»J '• . i/'y ■■' ',• i .r . Z' ' aW^ ' ■ 340 MEMOIRS OF THE CARNEGIE MUSEUM. EXPLANATION OF PLATE IX. Fig. 1. XeGtma osa/ea; Jordan and Kasawa, sp. nov. Type. Osaka, Japan. (C. M. Cat. Fishes, No. 7808.) Fig. 2. Gnathopogon majimcc Jordan and Hublis, sp. nov. Type. Ping-yang River, Korea. (C. M. Cat. Fishes, No. 7816.) Fig. 3. Belligohio eristigma Jordan and Hubbs, sp. nov. Type. Okayama, Japan. (C. M. Cat. Fishes, No. 7820.) Fig. 4. Ocycrms japonicus (Dbderlein), gen. nov. Tokyo market, Japan. (C. M. Cat. Fishes, No. 7859.) MEMOIRS CARNEGIE MUSEUM, Vol. X, No. 2 Plate IX Netuina, Gtiathopogon, Belligobio, Ocycrius. 342 MEMOIRS OF THE CARNEGIE MUSEUM. EXPLANATION OF PLATE X. Fig. 1. Liopempheris sasakii Jordan and Hubbs, sp. nov. Type. Toba, Shiina, Japan. (C. M. Cat. Fishe.s, No. 7860.) Fig. 2. M alakichthys wakiyce Jordan and Hubb.s, sp. nov. Type. Kagoshima, Japan. (C. M. Cat. Fishes, No. 7863.) Fig. 3. Brnchirus bellus Jordan and Hubbs, sp. nov. Type. Misaki, Sagami, Japan. (C. M. Cat. Fishes, No. 7894.) MEMOIRS CARNEGIE MUSEUM, Vol. X, No. 2. Plate X. Liopempheris, Malakichthys, Brachirus. « s T-: , ’ , . :v».' >1.^ , . ■ f •:> •\ > \ ' i ^“• *1, •; ’2;'v v^o5^c5<^' Iburiella, Encceura, Zalescopiis. 346 MEMOIRS OF THE CARNEGIE MUSEUM. EXPLANATION OF PLATE NIL Fig. 1. Zestichthys tanakce Jordan and Hubbs, sp. nov. Type. Kushiro, Japan. (C. M. Cat. Fishes. No. 7951.) Fig. 2. Allolepis hollandi Jordan and Hubbs, sp. nov. Type. Fukui, Japan. (C. M. Cat. Fishes, No. 7952.) Fig. 3. Monomitopus kurnce Jordan and Hubbs, sp. nov. Type. Misaki, Japan. (C. M. Cat. Fishes. No. 7954.) MEMOIRS CARNEGIE MUSEUM, VoL X, No. 2. Plate XII Zestichthys, Allolepis, M onomitopus. Publications of the Carnegie Museum, Serial Nos. 124'125 MEMOIES OF THE OAENEGIE MUSEUM VOL. X Nos. 3-4. W. J. HOLLAND, Editor A NEARLY COMPLETE ARTICULATED SKELETON OF CAMA- RASAURUS, A SAURISCHIAN DINOSAUR FROM THE DINOSAUR NATIONAL MONUMENT, UTAH. OSTEOLOGY OF ORNITHOPODOUS DINOSAURS FROM THE DINOSAUR NATIONAL MONUMENT, UTAH. Part. I. On a Skeleton of Camptosaurus medius Marsh. Part II. On a Skeleton of Dryosaurus altus Marsh: Part III. On a Skeleton of Laosaurus gracilis Marsh. By CHARLES W. GILMORE TITLE-PAGES, ETC., TOGETHER WITH INDEX OF VOL. X. PITTSBURGH Published by the Authority op the Board op Trustees op the CARNEGIE INSTITUTE July 10, 1925 MEMOIRS OF THE CARNEGIE MUSEUM Vol. X. No. 3. A NEARLY COMPLETE ARTICULATED SKELETON OF RASAURUS, A SAURISCHIAN DINOSAUR FROM THE DINOSAUR NATIONAL MONUMENT, UTAH. By Charles W. Gilmore. (Plates XIII-XVII.) INTRODUCTORY. In 1899 at the time of establishing a Department of Vertebrate Paleontology, the Carnegie Museum began the systematic exploration of the Morrison formation for remains of its dinosaurian fauna, and with but few short interruptions these explorations were continuously carried on up to the close of the year 1922. A very important period in this field of exploration was marked by the dis- covery in 1909 by Mr. Earl Douglass of an extensive deposit of Morrison fossils in northeastern Utah, since set aside as a part of the National Park system and designated as the Dinosaur National Monument. In the thirteen consecutive years during which this quarry was operated by the Carnegie Museum a great mass of materials, about three hundred tons in all, was collected and shipped to the museum. In these collections were many partially articulated skelotons of both large and small dinosaurs, but especially important was the recovery of a considerable series of well preserved skulls, the rarest and most sought for portions of the dinosaurian skeleton. The great diversity of forms represented, together with their unusual perfec- tion and excellence of preservation makes this one of the most remarkable fossil deposits that has ever been discovered in the Morrison formation. The quarry 347 348 MEMOIRS OF THE CARNEGIE MUSEUM. was esjiecially rich in skeletons of the large Sauropod types, but it remained for the final year of work in excavation to disclose the most perfect skeleton of a sauropod dinosaur, which has ever been discovered. So few bones are missing from this individual that in all essentials it may be regarded as being complete. With the acciuisition of this specimen it can be said, and without fear of contradiction, that the Carnegie Museum now has the largest assemblage of sauropod dinosaur skeletons of any institution in the world. In the present paper it is proposed to give a preliminary description of the superb skeleton mentioned above in order that it may be immediately available to students of the Dinosauria. The description is preliminary in that the final prepa- ration of the specimen was not completed at the time of my study, and undoulrtedly other details of structure will be disclosed in the concluding work. I wish here to exiiress to Dr. Douglas Stewart, Director of the Carnegie Museum, my great appreciation for the privilege of describing this unrivaled siiecimen and also for having made the necessary arrangements whereby it was [lossible for me to undertake this agreeable task. In the laboratory the bones of the skeleton have been skilfully worked out in relief under the direction of Mr. Arthur Coggeshall, by Messrs. Louis Coggeshall, A. Agostini, and Roy Kay. The excellence of the drawings and restoration illustrating this paper are all due to the artistic work of Mr. Sidney Prentice, draughtsman in the Carnegie Museum. Occurrence and Preparation of the Skeleton. The specimen (No. 11,338, C. M. Cat. Vert. Foss.) was received at the Museum in four large blocks of sandstone. Upon bringing these blocks together into their original relationship as found in the quarry, it was seen that the skeleton was pi'actically intact from the tip of the nose to the end of the tail, with the ribs of one side, pelvis, limbs, and feet in practically their natural positions. Some of the bones of the opposite side, however, have been shifted out of position and a few are missing. The animal lies on its right side with the neck bent strongly upward in a sigmoid curve as shown in Plate XIII. The skull retained its normal position, i. e. with its longitudinal axis at an angle with the longer axis of the neck (See fig. 4) . The tail likewise curves strongly upward in the anterior caudal region, sweeping forward above the line of dorsals, the distal portion cutting across the upwardly extended neck at about its middle, but at a higher level in the matrix. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 349 The vertebral column is preserved with the vertebrie in sequence from the skull back to the tip of the tail. It would appear that the caudal series is comi:)lete, except for the loss of three centra, the ninth, tenth, and eleventh, which are repre- sented by their articulated neural processes. In all the vertebral column consists of eighty-two vertebrae, divided as follows: cervicals, twelve; dorsals, twelve; sacrals, five; caudals, fifty-three. On the right side there are twelve thoracic ribs in regular sequence and apparently articulated with their respective vertebrae. These remain quite regularly spaced and give a clear conception of the extent of the body cavitjc The ribs of the left side are nearly all missing, only the first two remaining in position (See pi. XV). Originally other ribs of this side were present, but owing to their fragile condition they could not be preserved. On the left side all of the cervical ribs posterior to those for the atlas are present, but those of the opposite side, if preserved, remain buried in the matrix. The articulated fore limbs with the greater portions of both feet occupy their relative positions, the left lying above and slightly in front of the other. The preservation of the right scapula and coracoid in place beneath the ribs furnishes the first evidence in the Sauropoda as to the position and angulation of the shoulder blade in relation to the ribs and vertebral column. The proper articulated position of the scapula in the sauropod dinosaurs has long been a debated question among paleontologists and one upon which there is a considerable diversity of opinion. The evidence furnished by this articulated skeleton will be of the greatest help in arriving at a satisfactory understanding of its proper position and angulation. The right half of the pelvic arch is complete and properly articulated with the sacrum (See Plate XIV) but the pubis is all that remains of the left side and it was shifted somewhat out of position. The right hind limb and foot are complete. The head of the femur lies in its proper position in the acetabulum with the tibia and fibula bent backward. The left femur was found out of position below the neck, but the tibia and foot were retained in their natural place above the opposite limb, as shown in Plate XV. The articulated right hind limb furnishes indisputable evidence in favor of those who have supported the view that these animals walked in an upright quadrupedal attitude, and it should quiet for all time those who advocate a crawling, lizard-like posture for the sauropod dinosaurs. The important bones missing from this specimen are the left ilium, left ischium, part of one sternal plate, left coracoid, all but two thoracic ribs of the left side, a few anterior chevrons and some of the smaller bones of the fore and hind feet. 350 MEMOIRS OF THE CARNEGIE MUSEUM. A sheet of black carbonized matter found beneath and between the ribs of the right side may represent the carbonized skin and body tissues. Careful micro- scopical examination, however, failed to disclose a scale-pattern, such as has been found with sauropod remains in England. The position of the skeleton is that of an animal which died a natural death, for such disarrangement as exists can be attributed to the natural shifting of the bones rather than to tearing apart by any of the contemporary carnivora. The skeleton as shown in Plate XV has been worked out in deep relief. The few displaced bones were re-articulated, and the tail has been somewhat straightened, but otherwise the bones remain nearly as they were found. This treatment of this specimen is highly commendable since it preserves for all time the original evidence as to the proper articulation of the fore and hind limbs, as well as making clear several lesser points in the anatomy of these reptiles. The position of most of the bones of this skeleton is shown in Plate XIV, reproduced here from a drawing made before they were disturbed. The skeleton measured along the vertebral column has a greatest length of about seventeen feet, with a height at the hips of slightly less than five feet. That it is an immature individual is abundantly shown by its small size, the non- coalescence of the sutures and lack of rugose muscular areas on the limb and pelvic bones. The specimen is unhesitatingly identified as belonging to the genus Cainara- sauriis Cope, as recently characterized by Osborn and Mook, and which now, according to those authorities, includes the genus Morosaurus of Marsh. This conclusion was reached after a careful comparison of the skeletal parts with those of Camarasaurus and Morosaurus, which have been described and figured, and although dealing with a smaller individual than any of these, such close similarities were found in the outlines, proportions, and general massiveness of the individual bones as to leave no doubt as to its generic affiliations. In view of the very complete nature of the specimen under discussion it appears desirable to append here a synopsis of the principal osteological features of this genus. The characterization to follow has to a great extent been derived from the monographic work of Camarasaurus supremus by Osborn and Mook but includes such emendations and corrections as a study of this nearly perfect skeleton makes possible. ^ ^ ^ Genus Camarasaurus Cope. Cope, E. D., Pal. Bull. No. 25, pp. 5-10, 1877; Osborn and Mook, Memoirs Amer. Mus. Nat. Hist., N. S., Vol. Ill, Pt. Ill, 1922, pp. 247-377, Pis. LX to LXXXV, text figs. 1 to 118. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 351 OSTEOLOGICAL FEATURES. Skull abbreviated, deep, with open fenestration; jaws massive, deep; lachrymal elongate, with lachrymal foramen; jugal abbreviated antero-posteriorly; quadrato- jugal extensive and strongly in contact with squamosal; teeth large, spoon-shaped. Dental formula: premaxillary teeth 4; maxillary teeth 12-13; dentary teeth 13. Ring of bony sclerotic plates. Hyoid arch. Vertebral column composed of eighty-two vertebrae, divided as follows: twelve cervicals; twelve dorsals; five sacrals; fifty-three caudals. Proatlas present. Anterior cervicals with low simple spines with broad flattened tops. Cervicals beginning with the sixth have divided spines; cervicals wide, depressed; dorsal vertebrae stout; spines posterior to sixth simple, low, broad; anterior spines strongly divided; all dorsal centra depressed, of medium subequal length, and all opisthoccnlus ; lamination and fenestration more or less of distinct type; sacrum with short spines fused into a plate in adults; tendency toward retardation in the inclusion of sacrodorsal; anterior caudals with short spines having expanded summits; development of caudal ribs slight, disappearing posteriorly between ninth and twelfth vertebrae; anterior centra short, distal caudals never elongate. Ribs of thorax long, slender; cervical ribs, except posterior two, long, at- tenuated, the longest extending the length of three succeeding vertebrae. Pelvis: pubis short, massive; pubic foramen either open or closed; ischium characteristically of light construction with long slender shaft without distal expansion. Pectoral arch: scapula with broadly expanded ends; coracoid subcircular; sternal plates suboval; thickened border anterior. Carpus: usually with one ossified carpale, the radiale, always above meta- carpals I and II; flattened ossicle may represent a second carpale. Manus with five digits; metacarpals relatively slender; digit one bears the only clawed ungual; phalangial formula, 1 = 2, II = 1, III = 1, IV = 1, V=l. Femur relatively broad; fourth trochanter usually on proximal half, occasionally on middle. Tibia stout with heavy recurved outer process; tibia and fibula slightly less than two-thirds length of femur. Tarsus, with one ossified tarsal, the astragulus; small, rounded ossification may represent calcaneum. Pes with five functional digits; first three bear unguals; first metatarsal short, very stout; Met. II, III, and IV comparatively slender subequal in length; Met. V 352 MEMOIRS OF THE CARNEGIE MUSEUM. shortened, with flattened expanded upper extremity; phalangial formula 1 = 2; 11 = 3; III = 4; IV=1; V = 0? Cope proposed two species of Camarasaurus, C. supremus and C. leptodirus. The latter is regarded by Osborn and Mook as conspecific with C. supremus. Under the genus Morosaurus, Marsh proposed five species: Morosaurus impar, M. grandis, M. rohustus, M. lentus, and M. agilis. He gave but few distinctive char- acteristics and in his Dinosaurs of North America made no attempt to distinguish them, except to illustrate a few bones of the different species. Morosaurus impar is the genotype, but as Williston^ has pointed out, it is clearly a synonym of M. grandis. Riggs- correctly interprets the synonomy of the two species as follows: “while the specimen upon which the former species was based must remain the generic type, that of the latter being much more com- plete and better known will naturally be referred to in comparisons.” M. rohustus was established upon a single ilium of large proportions, and it may eventually fall within the species C. supremus. M. lentus is based on considerable portions of the skeleton of an immature individual, and according to Riggs^ who has examined the ty])e materials, “is not to be distinguished by the sutural articulation between the centrum and the neural arch of the vertebrae as Marsh’s figures would suggest, but by the massiveness of all parts of the skeleton and the depression of the vertebral pedicles so that the neural arch rests directly upon the centrum.” M. agilis, as first pointed out by Gilmore"* and further confirmed by direct comiiarison of the tyiie materials with the present specimen, does not belong in the genus Camarasaurus and may be dismissed from further consideration in that connection. From the above Irrief review of the seven species originally proposed only four may be considered valid at this time. These are: C. supremus Cope, C. impar (Marsh), C. rohustus (Marsh), and C. lentus (Marsh). A revision of these species is far beyond the scope of the present paper, and is a ])iece of work provided for in the monographic study of the Sauropodous dinosaurs which Prof. Henry F. Osborn and his associates now have under way. Until such time as the Morosaurus type materials are restudied and the species adequately characterized, the reference of new materials _to any of them is at- tended with much uncertainty. The assignment of the present specimen to any of these species would be more or less doubtful, although its small size, stoutness ’ Kansas Univ. Quart, vol. VII, p. 173. - Field Columbian Museum, Pub. 63, vol. I, 1901, p. 275. Lor. cit. p. 275. '* Proc. U. S. National Museum, vol. 32, 1907, pi. 163. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 353 of the skeletal parts, together with depressed pedicles of the dorsal vertebrae resting directly upon the centra, and pubis with an open pubic foramen, constitute an assemblage of characters in common with the type of Camarasaurus lentus that apparently indicate the affinities of the present specimen to lie in that species. OSTEOLOGICAL DESCRIPTION. The Skull. The skull of specimen No. 11,338, C. M. is the most perfect cranium of Camarasaurus {Morosaurus) that has yet been discovered. It has been slightly compressed, but the distortion is so little as to be almost negligible, and it gives a very clear conception of its original shape and proportions. Since it pertains to an individual not yet mature nearly all of the sutures are clearly indicated. The skull has not been completely extracted from the sandstone matrix but with the exception of the palate the structure from all other aspects is shown. Hitherto our knowledge of the Camarasaurian skull has been limited to the posterior part of the cranium of ‘‘Morosaurus” agilis (Marsh) described by Marsh*^ and by Gilmore*’; the more or less complete skull of Camarasaurus grandis (Marsh) briefly described by Osborn^; the posterior portion of a skull of Morosaurus grandis figured by Marsh* and the fragmentary skull of Camarasaurus supremus Cope described by Osborn and Mook.® This new skull, therefore, greatly extends our knowledge, especially in bringing about a better and more accurate conception of its detailed structure. The skull described by Osborn, although somewhat restored, closely approxi- mates the general form and proportions of the cranium now in hand, but it is now possible to correct certain errors of detail detected in the restored skull, although on the whole it is remarkably accurate, considering the limited knowledge of the cranium of Camarasaurus at the time of its reconstruction. Since the skull of Camarasaurus was very briefly described by Osborn, a detailed description of the Carnegie Museum specimen appears highly desirable. The skull of Camarasaurus can be distinguished at once from that of Diplo- docus by its abbreviated face with highly arched forehead or anteorbital region, and by the presence of large spoon-shaped cropping teeth. From the posterior aspect, however, the differences are not so clearly defined, as will be pointed out farther on. ® Amer. Jour. Sci., vol. 37, 1889, p. 334. Proc. U. S. National Museum, vol. 32, 1907, pp. 151-157. ’ Science, vol. 22, No. 560, 1905, pp. 374, 375; Nature, vol. 73, 1906, p. 283, fig. 2. * 16th Ann. Rept. U. S. Geol. Survey, pt. I, 1896, pi. XXX, fig. 2. ** Memoir Amer. Mus. Nat. Hist. N. S., vol. Ill, pfc. Ill, 1921, pp. 284-290. 354 MEMOIRS OF THE CARNEGIE MUSEUM. Viewed from the side the general structure of the skull is open and of slender construction as in the carnivorous dinosaurs. The posterior portion is deep dorso- ventrally and moderately wide transversely. The facial portion, as mentioned previously is shortened, slightly tapering toward the front when viewed from above with an olitusely rounded muzzle, as contrasted with the almost squarely truncated beak of Diplodocus. Fig. 1. Skull and lower jaws of Camarasaurus lentus (Marsh) No. 11,338, C. M. Viewed from the left side. Three-sevenths natural size, alsp, alisphenoid; an, angular; or, articular; hocpr, basi- oceipital process; d, dentary; /r, frontal; j, jugal; la, lachr3unal; rnx, maxillary; no, nasal; o, orbit; oc, occipital condyle; pa, parietal; pf, prefrontal; pmx, premaxillary; po, postorbital; poc, paroccipital process; pof, postfrontal; pr.at, proatlas; pt, pterygoid; q, quadrate; qj, qifadratojugal; sa, surangular; sq, squamosal; stf, supratemporal fossa; II, optic foramen. The plane of the occiput forms an obtuse angle with the fronto-parietal part of the skull. The occipital condyle, as Holland^” has so clearly pointed out in Diplodocus, is directed downward at nearly a right angle to the longer axis of the skull. Viewed from the back the skull is sub-rectangular in outline with the greatest diameter jiei pendicular. The sutural contacts of the occipital segment are rather obscure if not in most cases comiiletely obliterated. With the aid of the Memoirs of Carnegie Museum, vol. II, No. 6, lOOf), pp. 228-230. CHARLES W. GILMORE; SKELETON OF CAMARASAURUS, 355 type cranium of Morosaurus agilis Marsh in which most of the sutures are clearly defined it has been possible by direct comparison to work out the line of sutural contact between the parietal and the underlying supraoccipital and exoccipital bones with a considerable degree of assurance of their correctness as shown in figure 3. The other elements of the occiput could not be delimited. The basioccipital is composed of the subtrilobate condyle and the long, heavy, descending basioccipital processes. The exoccipitals contribute extensively to the formation of the occipital condyle and apparently exclude the basioccipital from participation in the boundary of the foramen magnum as is also clearly shown in M. agilis, Atlantosaurus montanus,^'^ and also in Antrodemus fragilis and Ceratosaurus nasicornisA Although the evidence is hardly extensive enough as yet upon which to base a positive assertion, it would seem that in the Saurischia the basioccipital is nearly always excluded from the foramen magnum whereas in the Ornithischia it always participates in the formation of the lower median boundary. This feature of the Ornithischia is well shown in Camptosaurus,^^ Stegosaurus, and in the Hadrosauridae. The supraoccipital cannot be differentiated, and the description to follow of the exoccipitals undoubtedly includes this bone. The exoccipitals are not as broad and strongly developed as in Diplodocus, and their outer extremities or paroccipital processes are more strongly deflected ventrally than in that genus. The combined elements from the posterior aspect are subtriangular in outline. The upper part of the triangular apex probably represents the coalesced supraoccipital bone. The exoccipitals probably meet broadly on the median line and entirely exclude the supraoccipital from the foramen magnum, as in M. agilis and as Holland has found the relations of these bones to be in the skull of Diplodocus carnegieiA'' Laterally these median elements articulate with the parietal and squamosals, the articulation with the latter being entirely with the paroccipital process, which extends outward, backward, and downward. On the median posterior surface of the supraoccipital area there is a pronounced vertical ridge developed as in Diplodocus and corresponding in position also to the more robust projection found in the skull of Antrodemus fragilis. The parietal is very short antero-posteriorly and enters very little into the composition of the cranial roof (See fig. 2). At the center it has an antero-posterior diameter “ 16th Ann. Rept. U. S. Geol. Surv. Pt. I, 1896, pi. XV, fig.s. 1 and 2. Bull. No. 110, U. S. Nat. Mus., 1920, p. 79. Proc. U. S. Nat. Mus.,vol. 36, 1909, fig. 4. Bull. No. 89, U. S. Nat. Mus., 1914, fig. 4. Op. cit. fig. 4. 356 MEMOIRS OF THE CARNEGIE MUSEUM. of only 20 mm. Anteriorly the coalesced parietals join the frontals by a nearly straight transverse suture that extends from the median anterior boundary of the supratemiioral fossa to join the postfrontal at a similar point on the opposite side. Viewed from above (See fig. 2) the parietal presents a narrow flattened median Fig. 2. Skull of Camarasaurus lentus (Marsh) No. 11,338 C. M. Superior view. One-half natural size, exoc, exoccipital;/?’, frontal; mx, maxillary; iin, nasal; pa, parietal; pj, prefrontal; pynx, premaxillaiy ; pr.at, proatlas; sq, squamosal. surface and two vertical iirocesses that extend in front and back of the supra- temporal fossa respectively. The iiosterior process is directed outward and slightly backward in relation to the central axis of the bone much as in the theropods. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 357 They rise slightly above the level of the median superior surface of the parietal forming a broad shallow concavity when viewed from the rear (See fig. 3). The parietals form all of the inner and most of the posterior boundaries of the supra- temporal fossa. The posterior lateral parietal process overlaps the squamosal at its outer extremity. In posterior aspect (See fig. 3) the lateral processes of the parietal are broad and wing-like at their outer extremities but rapidly narrow toward the median line. The anterior lateral process meets the postfrontal by a vertical process within the fossa and entirely excludes the frontal from participa- tion in the supratemporal boundaries. The frontals are united on the median line by a distinct interdigitating suture. They are broader than long and superiorly present shallow concave surfaces on either side of the median line which is slightly elevated. Laterally they extend outward to form the upper boundary of the orbit. This portion is interposed between the pre- and postfrontals. Over the orbital cavities the bone is thick and heavy, measuring eight millimeters in thickness. The exterior margins antero- posteriorly are shallowly concave, with a vertical rounded border which is somewhat roughened. Between the orbits the frontals have a greatest transverse diameter of 116 mm. The orbital contribution measures 31 mm. antero-posteriorly. The frontals, as mentioned above, do not participate in the boundaries of the supratemporal fossa. Posterior to the orbital border they have a wide sutural contact with the postfrontal. The fronto-nasal suture is rather obscure but its tentatively determined course is shown in fig. 2, but the evidence is very incon- clusive. As thus determined the frontals would have a greatest length on the median line of 65 millimeters. • The nasals are relatively short, broad posteriorly, with attenuated anterior extensions, which go forward above the anterior nares to lap the premaxillaries. Laterally, at the posterior end the nasals send a short pointed triangular process abruptly downward to meet the long slender ascending process of the maxillary, and thus completes the posterior boundary of the narial orifice (See fig. 1). Posteriorly they unite medially with the frontals, and laterally with the pre- f rentals and lachrymals. The upper surfaces of these bones are broadly rounded from side to side. The pointed anterior ends appear to be separated by the slender processes of the premaxillaries. The nasals have a greatest length in this specimen of about 112 mm,, a greatest transverse diameter of about 96 mm. Across the center of the nares the width diminishes to 19 mm. The postfrontal and postorbital are firmly coalesced and there no longer 358 MEMOIRS OF THE CARNEGIE MUSEUM. remains any indication of their sutural junction. This bone undoubtedly repre- sents, as in the Theropoda, a complex of two elements. On the median internal side, best seen on the left side of the present specimen (fig. 1 alsp.) at the back of the orbit, the outer end of the alisphenoid is seen to articulate. In those dino- saurian skulls in which the postorbital and postfrontal bones are found as distinct elements, this end of the alisphenoid is always received in a cupped depression on the inner side of the postorbital. On this evidence it would appear, therefore, that the line of sutural articulation of these tvu bones must be above this contact with the alisphenoid. This complex has the usual triradiate shape, a short heavy process which extends inward and articulates with the parietal and frontal and forms the outer two-thirds of the anterior wall of the supratemporal fossa; a smaller and more slender posteriorly directed process that meets the squamosal to form the short upper temjioral bar; the thii’d, much the longest one of the three, extends downward and forward and joins the posterior process of the jugal by a long oblique s(![uamous suture, thus forming the postorbital bar, which separates the orbital from the infratemporal opening. This bar is trihedral in cross-section, and it has the strong diagonal direction found in Diplodocus rather than the more vertical position found in tlie Ornithischia and the carnivorous Dinosauria. The prefrontal is a small triangular element that forms the upper anterior boundary of the orliit. Its widened jiosterior end and internal border articulates with the frontal and nasal. Proceeding forward it rapidly narrows to a sharp extremity that turns strongly downward to lap along the posterior upper side of the lachrymal (See fig. 1). The lachrymal in Camarasaurus is especially elongated and slender, and quite unlike any other dinosaurian lachrymal with which I am acquainted. In the articulated skull it stands nearly vertical in relation to its longer axis. The distal end is comparatively thin transversely but widened antero-posteriorly. It articulates with the jugal and maxillary above their junction, the contact being more extensive with the latter than Muth the former bone (See Plate XVI). Pro- ceeding superiorly in lateral view the lachrymal contracts antero-posteriorly into a narrow bar (See fig. 1), but from the middle upward, from a posterior view it widens transversely, the upper extremity being intercalated between the pre- frontal, nasal, and the ascending maxillary process. On the posterior side of the upper third the bone is perforated by a vertically elongated lachrymal foramen. This bone forms the bar separating the large anteorbital fenestra from the orbital cavity. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 359 The jugal is an exceedingly abbreviated; triradiate bone that participates but slightly in the lower boundary of the orbit. The anterior end has a wide dorso- ventral contact with the maxillary being overlapped by that bone. A slender tapering process extends backward and upward at an angle of 45° and joins the postorbital by a long diagonal squamous suture. This superior process is unusually short. . The third or inferior process is directed slightly downward and strongly backward to join the quadratojugal beneath the infratemporal fossa. The extent Fig. 3. Skull and lower jaws of Camarasaurus Icntus (Marsh) No. 11,338, C. M. Posterior view. One-half natural size, an, angular; ar, articular; bocpr, basioccipital processes; exoc, e.xoccipital ; oc, occipital condyle; pa, parietal; poc, paroccipital process; pr.at, proatlas; q, quadrate; qj, quadratojugal; sa, surangular; soc, supraoccipital; sq, squamosal; stf, supratemporal fossa. or manner of articulation of these two bones cannot be determined in this specimen, but the bar is long and very slender as shown in figure 1. The squamosal fits very snugly over the head of the quadrate and it nearly excludes the quadrate from contact with the paroccipital process, as in Antro- demus. The relationship of these bones is slightly obscure. The squamosal is a small, irregularly shaped element that is strongly interposed between the other 360 MEMOIRS OF THE CARNEGIE MUSEUM. elements of the upper posterior angle of the skull. Its inner extension meets the parietal and posteriorly it appears to meet the paroccipital process. It seems to contribute to the boundary of the supratemporal fossa, though the exact outlines of the bone are not entirely clear in this specimen. The quadrate bones are perfectly preserved. Neither one has been fully un- covered, but from a study of both a very clear conception of the entire structure has been obtained. Although it has an appearance of stoutness, the quadrate is relatively light in construction due to deep excavations on both front and back surfaces. The distal end which articulates with the lower jaw is massive and stout; it is concave on its anterior surface and slightly convex on the lower posterior face; the end is obliquely truncated, the longer side being internal. Distally it presents two articular faces, the inner being at right angles to the long diameter of the shaft, the outer being oblique in both anterior and external directions. The quadrate has a greatest length of 120 mm. as contrasted with the quadrate of Camarasaurus supremus described by Osborn and Mook‘® with a greatest extent of 270 mm. Viewed from the side the upper part of the quadrate curves strongly backward. The extent of the quadratojugal articulation cannot be certainly determined, although it is tentatively regarded as having the extent shown in figure 1. On the inner anterior side a vertically wide but thin plate extends forward at the back of the infratemporal fossa to articulate with the pterygoid as shown in figure 1. Its inferior border is convex. The internal side has not been un- covered, but on the basis of the (juadrate described by Osborn and Mook it probably turns up on the lower internal side forming a pocket-like excavation for the lodg- ment of a process from the pterygoid. A somewhat similar condition prevails in the skull of Antrodejrius. In the front between the outer anterior border and the thin, inner, anteriorly directed process for the pterygoid, the quadrate is cpiite deeply excavated. On the opposite or posterior face there is also a deep longi- tudinal pocket, so that the bone between these two excavations is very thin. The quadratojugal cannot be entirely delimited in this specimen though it seems to be extensive dorso-ventrally, resembling the Theropods in this respect, especially such forms as Antrodemus and Tyrannosaurus. From a lateral view it appears to articulate with the squamosal near the top of the quadrate, as in the carnivorous dinosaurs and thus to entirely exclude the top of the quadrate. From analogy it would also seem that such was the case. It unites by a squamous suture with the external side of the quadrate, a thin posterior portion a little Memoir.s Aiuer. Mus. Nat. Hist., n. s. vol. Ill, pt. Ill, 1921, p. 288. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 361 above the distal end remaining free and forming the outer wall of tlie deep longi- tudinal pocket or recess in the liack of the quadrate, when viewed from the back (See fig. 3). The distal end extends below the level of the outer articular end of the quadrate forming an obtuse notch between the two bones. A forwai‘dly directed process from the lower end meets the jugal, the two forming the long slender bar, which bounds the lower border of the infratemporal fossa. The premaxillary is a heavy rectangular bone, with a long, thin superior process which extends upward and backward from the anterior superior border and with its fellow of the opjiosite side divides the external narial orifice into right Fig. 4. Skull and neck of Camarasaurus lentus (Marsh) (No. 11,338, C. M. Cat. Vert. Foss.) Viewed from left side. One-fifth natural size. Shows relations of skull and anterior cervicals as originally found in the matrix. and left nares. This process unites above the center of the nares with the slender anterior process from the nasals by a long lapping sutural contact. Though thin transversely, this process widens rapidly antero-posteriorly toward its point of origin. On the anterior margin at the base of this process, there is a decided offset between it and the heavier dentigerous portion which gives the muzzle a decidedly dished profile, as contrasted with the regularly convex nose of the carni- vorous dinosaurs. On the inner side approaching the anterior end the contact with its fellow of the opposite side is evidently a flattened surface, that is oblique to the longer axis of the bone. Posteriorly it unites with the anterior end of the 362 MEMOIRS OF THE CARNEGIE MUSEUM. maxillary by a nearly straight vertical suture. In Camarasaurus the premaxillary carries four teeth, the longest and largest of the entire upper dental series. The articulated premaxillae show a decided depression following the median suture below the base of the superior process (See fig. 2). This depression fades out more ventrally and the muzzle is evenly rounded from side to side above the alveolar border. The heavy block-like portion of the premaxillary has a greatest diameter, measured around the curve antero-posteriorly, of 75 mm.; its greatest depth from dental to nasal border is 80 mm. Attention should be called to the striking similarity of this bone to the premaxillary in Antrodemus. The maxillary is heavy and especially broad vertically in its anterior portion, becoming narrower posteriorly. The superior bar of the maxillary is stout and especially produced upward and slightly backward from the main mass of the bone. It gradually widens antero-posteriorly at the upper extremity preceding its junction with the attenuated lachrymal. Toward the distal part of this process a thin transverse flange is developed which is directed inward and forward and joins its fellow of the opposite side on the median line. From their junction they extend anteriorly as a thin median septum to meet a posteriorly directed pair from the premaxillae. Between this septum and the lower external face is a shelf- like offset that forms the floor of the narial orifice. The inferior mass of the maxillary presents a broad smooth lateral surface that is pitted here and there with foramina. The inferior edge of the maxillary is deflected slightly outward, forming a border for the protection of the immature teeth. The posterior end unites wholly with the jugal by a nearly vertical concave suture. Measured along the outer side of the alveolar border the maxillary has a greatest length of 142 mm. The left maxillary has eight fully erupted spoon-shaped teeth; the right has nine. The alisphenoid is visible on the left side of the skull which has been divested of its matrix to the median line, and thus exposes this side of the brain case. Its full outline cannot be traced, but as in all other dinosaurs its outwardly directed branch, meets the inner side of the postorbital probably in a cupped depression on the inner side of that bone. Above it is certainly in contact with the frontal and possibly also with the parietal. Anteriorly it overlaps what appears to be the orbitosphenoid. It is perforated by a large foramen thought to be the optic (See fig. 1, II). I am unable to differentiate the other elements of the brain case in the present partially prepared condition of the skull. The posterior end of the left pterygoid is exposed showing it to send upward a high, broad, nearly vertical plate forward of its junction with the quadrate. This portion of the pterygoid is plainly visible within the orbit (See fig. 1) and its CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 363 anterior end is to be seen projecting forward of the lachrymal within the ante- orbital fenestra. A short, stout, transversely directed process joining the pterygoid and maxillary on the inside immediately below the lachrymal probably represents the ectopterygoid bone. The sutural divisions, however, cannot certainly be determined. The presphenoid is partly exposed. The Lower Jaw. The dentary in Camarasaurus is especially massive and affords a most striking- contrast to the slenderer dentary of Diplodocus. It is deepest vertically at the symphysis, which measures 83 mm. from the alveolar to the ventral border. This vertical diameter gradually grows less posteriorly. The front of the chin is broadly rounded from side to side, the two dentaries uniting by a strong median symphysis. The chin is slightly receding. The external surface, especially the anterior half, is dotted with pits, probably vascular foramina as in the premaxillary and maxillary above. Antero-posteriorly the dentary has a greatest length taken at the center of about 164 mm. As in the maxillary the superior edge sets out from the teeth. The posterior end overlaps the angular below and the surangular abpve. Viewed laterally the dentary is deeply excavated at the posterior end at its junction with the surangular (See fig. 1). Below the dentary-surangular suture, it sends back a wide, triangular process, which extends backward un,derneath the sur- angular and overlaps the angular. Posterior to the last tooth the superior l^order meets the coronoid. In this specimen the dentary carries thirteen teeth, evidently the complete series. These occupy a longitudinal space of 128 mm. The angular forms the lower portion of the posterior half of the ramus. It underlies the more robust surangular and the articular. Anteriorly it is overlapped by the dentary. The surangular with the small coronoid forms the whole of the upper external face of the ramus posterior to the dentary. Narrow at its posterior extremity it gradually widens anteriorly to the coronoid where it again becomes narrower as it approaches the dentary. A small external mandibular foramen perforates the bone as in Camptosaurus. The articular is a small block-like bone, longer than wide, with a shallowly fossate upper surface that constitutes the principal articulation for the quadrate. Viewed posteriorly (See ar. fig. 3) it is subtriangular in outline, wide above, and obtusely pointed below. Laterally it is lapped by the surangular and angular. The articular projects posteriorly slightly beyond the posterior termination of these elements (See fig. 1) and thus forms the profile of this end of the ramus. Its anterior and internal relationships are largely hidden by the enveloping matrix. 364 MEMOIRS OF THE CARNEGIE MUSEUM. On the inner anterior side of the articular a thin bone, protruding posteriorly from the matrix, may represent the prearticular. The coronoid forms a low convex coronoid process, which in the articulated skull passes upward behind the maxillary and when the jaws are closed is not visible from a lateral view. The other elements of the lower jaw are hidden by the matrix, which has not been removed from between the jaws. External Openings in the Skull. Viewed from the side the openings in the skull of Camarasaurus beginning posteriorly are the lateral temporal fenestra, the orbital opening, the anteorbital fenestra, and the anterior nares. Viewed superiorly the paired supratemporal fossse are the only openings. I can find no trace of a pineal foramen, although Osborn^^ observed it in skulls of this genus belonging to the American Museum of Natural History. The lateral tcjnporal fenestra is an obliquely elongated opening that is stirrup- shaped in outline, narrow above and wide below. It is bounded above by the posttemporal bar formed by the curved anterior process of the quadrato-jugal; anteriorly by the postorbital bar, formed about equally by the processes of the postorbital and jugal bones; ventrally by the quadratojugal and jugal. Its greatest oblique diameter is 116 mm.; its greatest antero-posterior diameter is 78 mm. The orbital cavity like the temporal fenestra is elongated obliquely, but widest at the top and narrowed to an acute angle at the bottom. The restored skull figured by Osborn^* is in error so far as the correct outline of the orbit is concerned, being too large and with the lower portion especially wide and out of proper pro- portion. The doubtful correctness of this region was fully recognized by Osborn. Dorsally the upper boundary is formed by the postfrontal, frontal, and prefrontal; anteriorly by the prefrontal, but principally by the lachrymal; ventrally by the jugal, and posteriorly by the descending and ascending branches of the postorbital and j ugal respectively. The orbit has a greatest supero-inferior diameter of 1 12 mm. ; a greatest transverse diameter of 80 mm. The anteorbital fenestra is strikingly larger than in Diplodocus resembling more nearly the conditions found in the carnivorous dinosaurs. It is acutely pear- shaped in outline. Its upper, anterior, and inferior boundaries are formed by the maxillary and its suiierior process; the posterior by the lachrymal. The greatest diameter supero-inferiorly is 60 mm.; greatest transverse diameter is about 35 mm. Nature, vol. 73, 1906, p. 284. Loc. cit. fig. 2. CHARLES W. GILMORE; SKELETON OF CAMARASAURUS. 365 The external nares are very large and face more directly forward than upward. There is no roofing over of the nares by the processes of the premaxillaries as found in many reptilian skulls. This opening is bounded above by the slender processes of the nasal and premaxillary bones; anteriorly by the premaxillary; below by the premaxillary and maxillary, and posteriorly by the maxillary process and nasals. It has a greatest oblique diameter of about 125 mm. ; an antero-posterior diameter of about 65 mm. The supralemporal fossa is suboval in outline with its greatest diameter at right angles to the longer axis of the skull. The external wall is much lower than its other boundaries so that the fossa looks strongly outward as well as upward. The inner and posterior wall is formed almost exclusively by the parietal, the squamosal entering only slightly into its outer posterior and external boundaries. The inner third of the anterior wall is formed by the parietal, the outer half by the postfrontal postorbital complex. Beneath these bones the alisphenoid may also contribute to the inferior boundary. The greatest oblique transverse diameter of this fossa is 55 mm.; the greatest antero-posterior diameter is 28 mm. Teeth. The dental formula of Camarasaurus, as clearly shown by this specimen, is as follows : Upper jaw: premaxillary 4; maxillary 8-9; total 12-13. Lower jaw: dentary 13. Osborn and Mook^® found only eight teeth in the maxillary of a specimen from the “Bone Cabin” quarry in Wyoming. It therefore appears that the number of teeth in the upper jaw may be found to vary slightly among different individuals, as is known to be the case in the carnivorous dinosaurs. In the present specimen there are eight maxillary teeth on the left side and nine on the right. Only the external view of the dental series is to be observed in this specimen at the present time. Nearly all of the teeth are fully erupted and functional in this individual. The teeth are homodont, small posteriorly, but becoming steadily larger toward the front in both upper and lower series. On the whole the teeth of the upper jaw seem to be slightly more robust than those of the dentary. Other- wise from an external view I fail to detect any peculiarities for distinguishing upper from lower. The unworn teeth have obtuse points that have the appear- ance of raking slightly backward, brought about by a concavely cut out condition on the posterior-superior border or posterior-inferior border of the crown, whether it be upper or lower tooth. This condition prevails on the lateral maxillary teeth. Op. cit. p. 290. 19 366 MEMOIRS OF THE CARNEGIE MUSEUM. as in all of those of the dentary. The premaxillary teeth have the two borders of the crown evenly convex coming to a blunt median point. Externally the crowns of the teeth are broadly convex. These surfaces are composed of a finely pitted type of enamel. On the external surface a vertical groove sets off a narrow posterior strip from the main mass of the tooth. It is presumed that the roots are cylindrical and the interior face of the crowns are spoon-shaped, as described by Osborn and Mook in Camarasaurus suprernus and by Marsh in 'Apatosaurus. The lower teeth bite within the upper in the articulated jaws. The most anterior premaxillary tooth has a greatest length measured from the alveolar border of 31 mm.; the most posterior maxillary tooth of the right side projects only 10 mm. All of the lateral teeth rake decidedly forward, in both upper and lower series and all curve slightly inward. MEASUREMENTS OF SKULL AND LOWER .JAWS. Greatest length of skull over all 330 mm. Greatest width across tops of ((uadrates - 147 mm. Greatest width above center of orbits 115 mm. Greatest width above center of nares 15 mm. Distance from i)osterior border of orbit to posterior extremity of scpiamosal, about 68 mm. Distance from anterior border of orbit to anterior extremity of premaxillEe, about 190 mm. Distance from extremity of premaxilhe to distal extremitj'' of quadrate 286 mm. Distance from distal end of quadrate to top of skull, about 175 mm. Height of skull through center of orbit 165 mm. Height of skull with lower jaw, taken at posterior end 205 mm. Height of skull with lower jaw, taken at center of orbit 215 mm. Height of skull with lower jaw, taken at front of superior process of premaxillary, about .... 200 mm. Lower Jaws. Greatest length of ramus over all 280 mm. Greatest depth of ramus below quadrate 25 mm. Greatest depth of ramus through coronoid process 67 mm. Greatest depth of ramus at anterior end 80 mm. Greatest breadth of rami, articulated with quadrates, about 125 mm. Sclerotic Ring. In the matrix filling the right orbit, the nearly complete but slightly dis- arranged sclerotic ring was preserved as shown in plate XVI. the second occurrence in the Sauropodous dinosaurs, and the first time found in the genus Camarasaurus. This ring is present in a skull of Diplodocus No. 11,255 in the Carnegie Museum, and since they are now known in the Ceratopsidse, Hadrosaurida?, and the Sauro- jioda there is reason for believing that sooner or later they will be found in all dinosaurian reptiles. In the present siiecimen the preservation is such that neither the number of scelrotic plates nor the method of their arrangement can be determined. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 367 Hyoid Arch. That there is a well developed hyoid apparatus in the Sauropod dinosauria is shown by the preservation in the matrix beneath the lower jaws of three rod-like bones, which undoubtedly represent elements of the hyoid arch. The longest of these measiu'es 165 millimeters in length. Tw’o of them are jiaired and probably represent the thyrohyal bones. Vertebral Column. The vertebral column as determined from this continuous articulated series ■has the following formula: cervicals 12; dorsals 12; sacrals 5; caudals 53. The cervical formula does not include the proatlas, which is apparently present and attached to the skull as shown in figure 3. In giving the formula as above the cervical, dorsal, and sacral series may now be considered as absolutely determined, since they are based on a continuous articulated series beginning with the atlas and ending with the forty-seventh caudal. At this point there is a slight disarrangement but since these disturbed caudals regularly diminish in length posteriorly it appears to indicate that none are missing. A total number of 53 is actually present, which agrees with the esti- mate of Osborn and Mook of the complete series in Camarasaurus supremus. There is reason for believing that the total number of caudal vertebrae will not be constant, but may vary with the individual, even within a species. The total number of presacrals disagrees with the tentative determination made by Osborn and Mook for C. supremus, and their interpretation of the division of these between neck and thoracic regions does not coincide with the evidence furnished by the present specimen. Their estimate of the formula is: cervicals 13; dorsals 10-11; sacrals 5-4; caudals 53. In consideration of the fact that they were dealing with disarticulated series, the accuracy of the results obtained is re- markable. The determination of the number of dorsals is made on the evidence of there being 12 undoubted thoracic ribs preserved in regular sequence on the lower or right side of the skeleton, as shown in Plate XIV. The vertebral column in Camarasaurus as compared with other known Sauro- poda shows a reduced number of cervicals and an increased number of dorsals, except for Haplocanthosaurus which, if correctly determined by Hatcher, has a greater number of dorsals than any known member of this group. These facts are graphically set forth in the table below". Cervicals Dorsals Sacrals Caudals Camarasaurus lentus (Marsh) 12 12 5 53 Apatosaurus louisce Holland 15 10 5 82 Diplodocus carnegiei Hatcher 15 10 5 73 Haplocanthosaurus priscus Hatcher (Unknown) 14 5 — 368 MEMOIRS OP THE CARNEGIE MUSEUM. The vertebral series in Camarasaurus is characterized by the compactness and stoutness of the individual vertebr®; the dorsals, especially the posterior members, by their low and wide spines, and the opisthocoelous character of the centra of the presacral series; the caudals by their short spines and weakly de- veloped transverse processes, and the absence of a long attenuated distal extension of the tail. The vertebrae agree with other Sauropods in having pleurocentral cavities in the presacrals, and divided spines in the postero cervical and antero dorsal regions. As pointed out by Osborn and Mook^® the vertebrae of Camarasaurus are much “more compact than those of Apatosaurus and much stouter than those of Amphi- coelias, Barosaurus, and Diplodocus.” Cervical Vertebra. The atlas is composed of four elements, intercentrum, odontoid, and two neurapophyses. They articulate in the usual manner forming a cup for the recep- tion of the occipital condyle. That a proatlas was present is apparently indicated by two fragmentary bones attached by matrix to the posterior part of the skull immediately above the foramen magnum and crushed down over that opening, so as to obscure its outlines, as shown in figure 3. They are too poorly preserved to permit of description. The axis is distinctive from the other cervicals, and differs from those of both C. supremus Cope and C. impar Marsh in having a more depressed spinous process. This process, which anteriorly is low, rises at a low angle, as contrasted with the steep inclination in the two above mentioned species. The centrum is of moderate length, with an elongated pleurocentral cavity. The odontoid is free from the anterior end, which is squarely truncated. No indication is found of a separate axial intercentrum such as exists in the type of Morosaurus agilis Marsh. The left diapophysis is broken, but enough remains to show it to have been very small. The cervical ribs for the axis and atlas, if such bones were present, are both missing. Beginning with the third cervical, the remaining vertebrae of the neck are only partly exposed. The dorsal surfaces and the centra of the left side are all uncovered, as shown in Plate XIV. The long, attenuated, cervical ribs remain articulated with their respective vertebra, and somewhat screen the underlying centra. This specimen shows the almost complete cervical rib-series for the first time in the Memoirs Amer. Mas. Nat. Hist., No. 8, vol. 3, pt. 3, 1921, p. 290. Gilmore, C. W., Proc. U. S. Nat. Museum, vol. 32, 1907, fig. 7. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 369 genus Camarasaurus. In so far as comparison is possible, the cervicals resemble one another except in degree of development . From the third posteriorly they gradually increase in size in all dimensions. The spines on cervicals 3, 4, 5, and 6 are single, low, stout processes with flattened summits. The spines grow pro- gressively larger posteriorly with the first notching on cervical 7; this notching becomes deeper and deeper until on cervical eleven the spine is divided into two metapophyses. In relation to the arch the spines are posterior not median as found in other genera of the Sauropoda. The summits of the spines on cervicals 3, 4, 5, and 6 are wider than long, whereas posteriorly their dimensions are reversed. Postspinal fossae are present behind and between the posterior zygapophyses, but deepest in those vertebrae with divided spines. , The prezygapophyses are large, flat, and wide apart with articular faces inclined toward one another. The zygapophysial laminae, so far as they can be observed, seem to be arranged as shown in C. supremus, which have been described in great detail by Osborn and Mook. The diapophyses, relatively short anteriorly, grow progressively longer and heavier posteriorly. They project nearly straight out from the arch and in all instances exceed the spread of the zygapophyses. Their truncated outer ends are triangular in section. None have become coalesced with the tuberculum of the ribs. Dorsal Vertebra. The dorsal series in Camarasaurus, as positively shown by this articulated vertebral column, consists of twelve vertebrae, exclusive of the sacro-dorsal. Osborn and Mook were correct in their conclusions as to the total number of presacral vertebrae, but erred by the inclusion of two dorsals in the cervical series. That the division between cervical and dorsal regions takes place as given above is indicated, not so much in the structure of the vertelirae themselves, as by the presence of twelve typical thoracic ribs on the right side, and with the first two ribs of the left side properly articulated by both capitulum and tuberculum with the proper facets on the first and second dorsal vertebrae respectively. The cervical rib preceding it was found nearly in position and bears no resemblance to the first thoracic rib. The parapophysis on the first dorsal is situated on the side of the centrum in front of the pleurocoel. In the second dorsal the parapophysis is slightly higher ; on the third dorsal on the lower part of the arch ; on the fourth dorsal it occupies a slightly higher position; and on the fifth dorsal has apparently reached its maximum height, which is retained throughout the remainder of the series. 370 MEMOIRS OF THE CARNEGIE MUSEUM. The dorsal vertebrse have suffered considerably from crushing, especially the transverse jirocesses and lateral laminae of the left side. The spines posterior to the eighth have also suffered the loss of parts. At the time of preparing this manuscript the processes of the dorsals had not been completely prepared, so that a complete account of these bones cannot be given at this time. In so far as the centra can be observed in their articulated condition they appear to be in accord with other members of the genus in being opisthoccelous. Pleurocentral cavities are present in all and situated on the upper part of the sides of the centra. The anterior dorsals unlike those of Diplodocus and Apato- saurus are not lengthened, but are about subequal in length to those that follow. The measurements given below are made along the sides of the articulated centra and do not include the length of the ball. Xo. in series 1 2 3 4 o 6 7 8 9 10 11 12 mm. mm. mm. mm. mm. mm. mm. mm. mm. mm. mm. mm. Length 70 82 80 80 82 80 80 80 80 82 76 70 Height over all .. 227 230 240 255 260 The spine of the first dorsal is deeply cleft as in the posterior cervicals, but posteriorly they become more and more shallow, that of the sixth dorsal showing only a shallow notch. The seventh spine has an evenly low convex outline when viewed from the front. These spines have the characteristic low broad massive structure found in the other members of the genus. The lamination in so far as it can be compared at this time, is in close accord with the structure found in Camarasaurus supremus, which has been minutely described by Osborn and Mook. Sacrum. Tlie sacrum in the present specimen is composed of five vertebrae, which have only been fully prepared on the ventral side. Coossification does not appear to have taken place between any of these centra, but all are joined to the ilia by sacral ribs. Viewed from below the four anterior centra are about subequal in length as well as in transverse dimensions; the fifth or sacro-caudal is somewhat shorter. This specimen differs from the sacrum of Camarasaurus supremus de- scribed by Osborn and Mook in which the caudo-sacral is subequal in size with those preceding it. The proximal ends of the sacral ribs are coalesced with the centra, although their sutural articulation remains clearly defined. They are broad near the centra, constricted at the center, and greatly expanded at their distal ends where they meet one another to form the sacricostal yoke. Their outer ends have not yet become coalesced in this specimen. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 371 The first rib articulates intervertebrally between the centra of the twelfth dorsal and first sacral, but more especially with the latter; the ribs of the succeeding three are articulated largely on the anterior half of the centra, but the rib of the fifth unites centrally in relation to that centrum. A very similar condition pre- vails in the sacrum of Morosaurus grandis figured by Marsh. As in Camarasaurus supremus Cope and M. grandis Marsh, the ventral surfaces of the centra are transversely broadly rounded. The sacricostal yokes are stout; on the left side, from which the ilium is missing, this yoke presents a rounded longitudinally concave surface, which, as shown on the opposite side, articulates with the ilium and enters slightly into the formation of the superior boundaries of the acetabulum. The first sacral rib contributes but slightly to the formation of the yoke, and its union with the second at the distal end is not so strong as between those more posterior. It is also thinner antero- posteriorly. The presence of five vertebrae, entering into the formation of the sacrum, would, according to Marsh’s former interpretation, exclude this specimen from the genus Morosaurus, which was regarded as having only four sacrals.^® Osborn and Mook, however, have shown that in all probability there are five in all of the Sauropoda, but whether ankylosed or not depends upon the age of the individual. It is quite apparent that in this specimen we have the usual three primary sacrals with a sacro-dorsal in front and a modified caudal or sacrocaudal behind. MEASUREMENTS. Length of five articulated sacral centra 420 min. Length of four posterior sacral centra 332 mm. Breadth across sacricostal yoke, anterior end 255 mm. Breadth across sacricostal yoke, posterior end 300 mm. Length of sacral spines 280 mm. Caudal Vertebra. The total number of caudal vertebrae in the tail of Camarasaurus may now be quite certainly given as 53. In the present specimen 53 are present, the anterior 47 being preserved in regular articulated sequence. At this point, however, there is slight disarrangement, and while it cannot be positively asserted that all of the remaining elements are present, the regularity of their diminishing lengths seems to indicate that none are missing. The estimate of a total of 53 in C. supremus as determined by Osborn and Mook from incomplete series is therefore correct. Dinosaurs of North America, 1896, pi. 31, fig. 8. hoc. cit. p. 241. 372 MEMOIRS OF THE CARNEGIE MUSEUM. Only the left side of the centra and transverse processes of the eight anterior caiidals are exposed, but from the eighth posteriorly all have been worked out in full relief. The centra of caudals nine, ten, and eleven are missing, but the neural arches and spines remain articulated in their proper sequence. Transverse processes, or caudal ribs, are present on the first eight, but have disappeared entirely on the twelfth. The missing centra therefore, do not allow of an exact determination on this point. In a larger specimen of Camarasaurus (No. 584, C. M. Cat. Vert. Foss.), they persist as far back as the thirteenth caudal. In Apatosaurus these processes are found for the last time on caudal fifteen, and in H aplocanthosaurus on the thirteenth. These processes rapidly decrease in size posteriorly, as in the other genera mentioned. The caudal centra in the anterior part of the tail are amphiccelous, relatively short and high, as in H aplocanthosaurus. These centra are somewhat constricted medially. In length they remain about subequal as far back as the twenty-second caudal, but the vertical height, diminishing rapidly at first, becomes more gradual thereafter and continues to the very tip of the tail. As shown in the table of measurements, the reduction in length from the twenty-second vertebra posteriorly averages about 2 mm. to the vertebra. The thirty-seventh is exactly one-half the length of the first and the fifty-third is about one-fifth as long. The pointed terminal caudal is nearly twice the length of the vertebra preceding it. On the anterior third of the tail the inferior surface of each caudal is relatively broad, but this condition gradually changes posteriorly where they become regularly rounded. The chevrons are attached intercentrally, but more especially with the beveled surface of the anterior vertebra of each pair. The chevron-facets are not distinctly defined and on the anterior vertebrae in the absence of the chevrons it is quite impossible to certainly determine which vertebra of the series carried the first. The neural arches are not disclosed in advance of the ninth vertebra at this time. The arch of the ninth is low and simple in structure, with a plate-like neural spine and apparently without terminal expansion. It is inclined slightly backward. The spinous processes decrease regularly in height and other dimensions posteriorly, becoming more rounded and gradually assuming a more nearly horizontal position in the posterior part of the series. The neural arches persist as far back as the forty-eighth caudal. The position of the arch on the anterior half of the centrum, at least from the twelfth vertebra posteriorly, appears to distinguish the caudals of Camarasaurus from those of H aplocanthosaurus, where they arise more nearly in the middle of the centrum. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 373 The anterior half of the tail is higher than wide, but the dimensions become more nearly equal in the distal portion, where the vertebrae are nearly round in cross-section. The presence of a pointed distal caudal positively precludes the development of a long, slender, whip-lash extension, such as has been found in Diplodocus and Apatosaurus. Measured along the side of the centra the complete series has a total length of about 23.55 mm. or about 7 ft., 9 inches. MEASUREMENTS OF CAUDAL VERTEBRA. No. in series Length of Centrum Height over all No. in series Length of Centrum Height over all No. in series Length of Centrum Height over all mm. mm. mm. mm. mm. mm. 1 64 24 55 95 47 22 — 2 63 25 55 90 48 19 — 3 62 26 53 85 49 18 11 4 61 27 52 80 50 15 9 5 62 28 50 72 51 14 9 6 62 29 49 63 52 13 8 7 65 30 47 60 53 20 7 8 59 31 45 58 9 - 32 43 57 10 - 33 41 50 11 34 39 — 12 55 152 35 37 — 13 57 145 36 35 45 14 57 140 37 32 43 15 58 135 38 30 38 16 58 130 39 29 36 17 58 130 40 29 35 18 60 120 41 27 — 19 59 115 42 26 — 20 60 110 43 25 — 21 58 — 44 25 — 22 58 105 45 24 — ’ 23 56 102 46 22 21 Chevrons. The total number of chevron-bones in the tail of Camarasaurus cannot be certainly determined from this specimen. All are missing back to the fourteenth caudal, but from this point to the thirtieth the articulated series is preserved in situ. That there were at least three more is shown by those present but slightly displaced on the thirteenth to the thirty-second caudal. Assuming that Osborn and Mook are correct in regarding the first caudal as carrying the first chevron, there would be thirteen missing from the anterior part of the tail, or, with the 374 MEMOIRS OF THE CARNEGIE MUSEUM. nineteen actually present, the complete series of chevrons would consist of thirty- two elements. The fourteenth chevron is Y-shaped, with a shaft which curves backward. It has a greatest length of 90 mm. Posteriorly they become progressively shorter, with a more and more decided bending back of the distal extremity. On the twenty-second vertebra the chevron presents a widened blade of putty-knife shape, but beginning with the twenty-fourth this blade begins to develop an an- terior projection, and in the last of the series the anterior and posterior projections are about subequal. The twenty-third chevron is 40 mm. long; the twenty-seventh is 28 mm. in length. Cervical Ribs. The cervical ribs of Camarasaurus as shown by this specimen for the first time, constitute one of the characteristic features of its skeletal anatomy. Refer- ence is made to their great length and the slenderness of their shafts. The rib of the seventh vertebra, which is the longest of the series attains a length of 375 milli- meters, and its attenuated end reaches a point near the forward end of the centrum of cervical 11. The fifth rib measures 318 mm.; the sixth rib measures 325 mm. in length and its distal termination is slightly in advance of the diapophyses of cervical 9. The anterior ribs are long, but posteriorly they shorten rapidly. This type of cervical rib is in striking contrast to the heavy and relatively short ribs in Apatosaurus, or the relatively short, but more slender ribs of Diplodocus, none of which overlap more than the succeeding cervical. In large adult individuals of Camarasaurus the cervical ribs are usually anky- losed by their tubercular processes with the diapophyses of the vertebrae, but in this specimen all remain distinct, another indication of its immaturity. There is no appreciable extension of the rib forward in front of the tubercular process as in Diplodocus, and to a less extent in Apatosaurus, this end being quite squarely truncate. The shafts of the ribs extend backward at right angles to the tubercular and capitular processes and taper to a long, slender rod whose distal extremities have a tendency to turn outward. Eleven of the cervical ribs of the left side were preserved articulated, beginning with the axis. These have the following lengths commencing with the fifth: fifth, 318 mm.; sixth, 325 mm.; seventh, 375 mm.; eighth, 320 mm. ; ninth, 170 mm.; tenth, 146 mm.; twelfth, 130 mm. The first thoracic rib increases to a total length of about 510 mm. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 375 Thoracic Ribs. On the lower or right side of the skeleton, undisturbed in the matrix, are twelve thoracic ribs. These are preserved in regular sequence and show for the first time in the genus Camarasaurus the total number of thoracic ribs as well as a correct idea of their proper arrangement. The proximal ends extend into the matrix beneath the line of dorsal vertebrie and are not to be observed at this time. The second rib, due to a fracture passing through the vertebral column, is properly articulated with the second dorsal, and it is presumed from the regularity of their spacing that all remain articulated with their respective vertebrae. All of these ribs, with the possible exception of the first, are considerably flattened and give no clue as to the proper curvature of the thoracic walls. It is the presence of these twelve undoubted thoracic ribs which certainly de- termines the number of dorsal vertebrae in this genus. This count does not include any dorso-sacral rib, and it seems very doubtful that this vertebra had a free rib. The first rib is straight and relatively short, its distal extremity ending behind the scapula, at a point opposite the most constricted part of the shaft. The second rib of the left side has a greatest length measured over the curve of 620 mm. Since the right scapula in this specimen is regarded as having been retained in its proper angle in relation to the vertebrae and thoracic ribs, in any articulated skeleton the distal end of this rib would not be visible below the lower border of the blade, in a lateral view. The upper or exposed part is wide transversely, but near the middle of the shaft is subround, becoming flattened and widened antero-posteriorly as the truncated distal end is approached. The character of this end strongly suggests that it was attached to the sternum by a cartilaginous rib. It certainly bears no resemblance to the sharply pointed transitional ribs found between the true cervical and dorsal ribs in the Camptosauridce, and which give rise to dif- ferences of opinion as to which series they should properly be assigned. The first and second ribs of the left side are also present and articulated with the diapophyses, but they have suffered loss of distal portions of the shafts. The right rib, however, is complete and is nearly eight inches (200 mm.) longer than the first. Proceeding posteriorly from the second the ribs become progressively longer, and heavier. In these respects they reach their maximum development in the fifth. The sixth to the eighth rib remain about subequal in length, but their 'shafts grow progressively narrower. These ribs, including the next two, are very slender. The ninth is slightly shorter than the tenth. The remaining ribs shorten rapidly by decided steps. The distal end of the eleventh terminates slightly below the level of the point of the articulated ilium. The twelfth has the 376 MEMOIRS OF THE CARNEGIE MUSEUM. proportions of the eleventh, but measures 100 mm. shorter. It curves forward in front of the iliac border, and suggests that in an articulated skeleton its distal portion would be in front of the ilium and not behind it, although that is the position of similar ribs in mounted skeletons of Apatosaurus and Diplodocus. Brown^"* has shown that the posterior ribs in Monoclonius also curve forward to positions in front of the ilia. The Pectoral Girdle. The pectoral girdle is represented by both scapulae, coracoid, and one complete sternal plate, and the anterior portion of the other. The scapula is of the typical Camarasaurus type, being relatively short, with broadly expanded proximal end and with the upper or distal end strongly expanded in an anterior direction. The coracoid is subquadrangular as in the other members of the genus. Especially interesting is the preservation of the right scapula and coracoid in somewhere near their natural relations in reference to the thorax and the vertebral column. Since all the other bones of the right side, have been retained in their proper articulated relationships, there seems no reason for not regarding the scapula as being similarly preserved, except that it is quite apparent (See Plate XIV) that the anterior part of the body has been thrust upward away from the attached fore limbs. Assuming these deductions to be correct, this specimen furnishes the first evidence obtained of the articulation of the shoulder-blade in the sauropodous dinosaurs. This is especially important as there has been a great diversity of opinion as to its position and angulation as expressed in articulated skeletons, and also in pictorial restorations of these animals. The latest and most radical departure from previously held views being the restoration by Osborn and Mook of Camara- saurus supremus^^ in which the scapula has been given a more nearly perpendicular position bringing about a great elevation of the shoulders. The evidence furnished by the present articulated specimen showing the scapula as found in position, is quite opposed to such a posture. It shows the scapula as having the blade in a much more horizontal position. The whole scapula is probably lower on the side of the thorax than shown in any restoration or mounted skeleton of this group of dinosaurs. The first rib crosses the upper border at about the center of the narrowest part of the shaft, and the upper truncated end reaches posterior to the fifth rib. This end is 370 millimeters distant from the lower anterior border of the ilium of the same side. Bull. Amer. Mus. Nat. Hist. vol. 37, 1917, p. 294, pi. XIII. Memoirs Amer. Mus. Nat. Hist., ii. s., vol. 9, pis. 83-84, also fig. 28. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 377 Though highly desirable to have verification of this evidence, past experience has shown that information furnished by articulated skeletons in situ, is far more to be relied upon than any number of expert opinions, based on analogies deduced from the study of living reptiles. This information concerning the articulation of the scapula is one of the outstanding contributions of the present specimen to a more exact knowledge of sauropod anatomy. Sternal Plates. The sternum is represented by one complete plate provisionally identified as the right and the anterior portion of the opposite. Both were found out of position in the rock above the caudal vertebrae. The complete plate shown in fig. 5, is suboval in outline, as in C. supremus. This plate has a thickened end which looks forward and outward and it is this border which probably articulated with the process of the coracoid, the distal is thinner and with many indentures for the attachment of the cartilaginous ribs. The internal and external borders are subequal in thickness, but the side regarded as median has a straighter border. While there may be some doubt as to which side this plate belongs, the thickened extremity certainly identifies it as being the anterior end, as shown by the sternal plates of Monodonius found in position by Brown. The dorsal surface of the plate is shallowly concave from side to side, while the ventral surface is slightly convex in the same diameter. The slight constriction of the lateral borders near the posterior ends seems to show that Marsh has incorrectly identified the posterior ends of the sternal plates of C. grandisas illustrated by him in fig. 30, p. 179, in “Dinosaurs of North America.’’ This plate has a greatest antero-posterior diameter of 283 mm.; a greatest diameter of 200 mm. Fore Limb and Foot. All of the bones of both fore limbs are preserved and to a large extent remain properly articulated. The head of the right limb remains in the glenoid fossa. Brown, B., Bull. Amer. Mus. Nat. Hist. vol. 37, 1917, p. 291, figs. 3 and 4. a Fig. 5. Right sternal plate of Camarasaurus lentus (No. 11,338 C. M. Cat. Vert. Foss.). Superior view, one-fourth natural size, a, anterior end; b, margin next to median line; c, coracoid border; p, posterior end. 378 MEMOIRS OF THE CARNEGIE MUSEUM. The left limb lies in the sandstone matrix above and slightly forward of the right. The humeri have their anterior surfaces directed strongly toward one another, but whether this fact will contribute any information as to the proper pose of the front limbs I am unprepared to say at this time. The humeri are stout, with expanded ends more especially the proximal, with prominently developed deltoid crest. The condyles are rather weakly developed, and the olecranon fossa is shallow. In length they are slightly more than half as long as in the type of Morosaurus grandis. The width of the proximal end, as compared with the length, is considerably greater than in either Camarasaurus supremus Cope, or C. grandis (Marsh). The ulna is moderately stout at the proximal end deeply concave anteriorly for the reception of the proximal portion of the radius. In the shaft the ulna is stouter than the radius, but tapers toward the distal end, so that the distal ends of these two bones are about subequal in size. The articulated radius crosses from the front at the proximal end to the side at the distal end of the ulna. The two slightly expanded ends of this bone are about equal in size. In relation to the humerus the lower limb bones of this specimen are relatively longer than in the type of Morosaurus grandis, but whether this is a constant difference or due to the immaturity of the individual yet remains to be determined. The lack of rugoseness on the ends of the limb bones and the moderate development of their processes clearly indicates the immaturity of the present specimen. Carpus and Forefoot. Both articulated fore feet are present, but neither one is completely pre- served, though from a study of both it has been possible to determine nearly the entire structure of the manus. The carpus in Camarasaurus, as is now quite certainly established, consists of a single large, compact, flattened, block-like radiale. Such a bone is present in both limbs, and in both instances occupy precisely similar positions in relation to the other elements of the limb and foot, i. e. above metacarpals I and II and below the distal end of the radius. Additional evidence in support of this view is furnished by an articulated forefoot in the U. S. National Museum assigned to Morosaurus agilis by Marsh, and figured by him.^^ It has a single element similarly articulated. That it has Camarasaurus affinities is clearly shown by comparison with the feet of the present specimen. In 1901 Riggs^* described and figured a manus identified 16th Ann. Kept. U. S. Geol. Survey, pt. I, 1896, pi. 37, fig. 1. Field Columbian Museum, Pub. 63, vol. I, 1901, Geol. Sur. p. 276, pi. 41, fig. 1. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 379 as pertaining to Morosaurus grandis also having a single carpal and occupying the same relative position as in the feet of the specimen here discussed. In restoring this foot, however, Riggs introduced a second large element. Osborn^® has illustrated feet ascribed to Morosaurus, but in each instance there are two large carpal elements of about equal proportions. The justification for this arrangement is not presented, so that in the light of the present evidence it would appear either that Osborn was in error, or else the feet were wrongly identified. In this connection it should be mentioned that a small flattened ossification was found near the left wrist of the present specimen, but, as it was evidently out of position, one cannot be altogether certain where it belongs in the carpus. On the evidence of the four articulated fore feet discussed above, it would now seem to be an established fact that the carpus in Camarasaurus consists of a single element, the radiale, which always articulates with metacarpals I and II, and possibly a smaller bone whose precise position is yet unknown. This feature of the carpus in Camarasaurus will be of assistance in distinguishing it from Apatosaurus with its single large medially located disk-like carpal and from Diplodocus with its two block-like carpal elements. The metacarpus consists of five functional metacarpals. All are present in each foot of this specimen, the right having the palmar view exposed; the left with part of the front aspect. The proximal ends vary in form, the median three being roughly triangular and all are closely applied to one another at this end. The distal ends are all expanded transversely. Metacarpal I is the shortest and stoutest of the series with the distal end slightly oblique and shallow and broadly grooved. Metacarpal II is longest. Metacarpal V is reduced. When articulated the metacarpals are broadly rounded viewed from the front, and it would seem they would carry the weight evenly. Only the first digit is provided with a claw-like ungual. The evidence fur- nished by many articulated fore feet in various museums, supported by the present specimen, further establishes the fact that in the Sauropoda there is but one ungual, and that a relatively small one, carried on the pollex. The digital formula as established on this specimen would be 2, 1, 1, I, 1. Digit I of the right pes has two articulated phalanges, the last one a slightly curved ungual; digit II and III of this foot carry a single phalanx each; those of IV and V are missing. The left foot has proximal phalanges on the I and V digits possibly a disarranged phalangial may pertain to digit IV. A small button-like ossification in the matrix lateral to Bull. Amer. Mus. Nat. Hist. vol. 14, 1901, fig. 3; Idem. vol. 20, 1904, fig. 1. 29 380 MEMOIRS OF THE CARNEGIE MUSEUM. the proximal phalanx of digit V may represent the terminal phalanx of that toe, but of this one cannot be certain. Riggs3o especially drawn attention to the oblique direction of the claw of the first digit brought about by the bevelling of the inferior facet of metacarpal I and the two facets of the proximal phalanx in such a manner as to give the articulated claw an oblique outward and backward direction. This condition is fully substantiated in the present specimen. Pelvic Girdle. The pelvic arch is represented by the articulated ilium, pubis, and ischium of the right side, and the pubis of the left side. The left ilium and ischium are missing. In the present position of the skeleton the right ilium is almost entirely hidden from view by the overlying sacrum, except a portion of its anterior extremity and the acetabular border. The parts exposed are in conformity with those of the described ilia of this genus. The pubis is short and in its general conformation is quite in accord with other pubes of Camarasaurus. It lacks the massiveness of this bone in the larger species, but this difference is no doubt due to the immaturity of the individual. The pubic foramen is open posteriorly as in C. lentus whereas in C. supremus Cope and C. grandis (Marsh) it is entirely enclosed. Since similar conditions are found among specimens of Camptosaurus, where an open foramen is always associated with juvenile characteristics, I am inclined to the opinion that this difference is to be explained in the same way. The right ischium has the characteristic long, slender shaft without distal expansion so distinctive of the genus Camarasaurus. The proximal end is expanded and closely approximates the form of the C. {Morosaurus) lentus ischium figured by Marsh. These bones of the pelvis are less massive, and lack the heavy and rugose development of their articular ends and surfaces found in the larger members of the genus. Hind Limb and Foot. The right hind limb and foot are complete and articulated, except for the loss of a few smaller elements of the pes. The femur of the left leg is shifted out of position and was found lying in the matrix below the posterior part of the neck. The fibula of this side was displaced, but the tibia and nearly complete foot was preserved in its natural relations to the remainder of the skeleton. Op. cit. p. 276. 16th Ann. Kept. U. S. Geol. Survey, pt. I, 1896, pi. 36, fig. 1. CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 381 The head of the right femur remains in the acetabulum and furnishes irre- futable evidence in support of those advocating an upright mammalian-like articu- lation of the hind limb in the sauropodous dinosaurs, as opposed to the lizard-like posture advocated by Tornier/^ Hay^^ and others. The head of the femur is inserted straight into the acetabulum at right angles to the longer axis of the shaft. ' That it could not be articulated otherwise is indicated by measurements, which show the antero-posterior diameter of the acetabulum (175 mm.) to be less than the transverse diameter of the proximal end of the femur (220 mm.). This fact absolutely disposes of the proposed articulation of the femur suggested by Hay. The bone of the distal half of the right femur has been somewhat altered by crushing, but the left is quite perfectly preserved. It is relatively short and broad, as compared with those of other sauropod genera. The fourth trochanter is moderately developed, and is entirely on the proximal half of the bone, whereas in C, supremus the apex of its crest is exactly in the middle of the shaft. The condyles have a moderate development, and there is a decided anterior inter- condylar groove. The tibia is massive, and slightly less than two-thirds the length of the femur. The proximal end is broad transversely. The external flange, which is developed in front of the fibula, is thickened and curves decidedly backward. The distal end is less expanded than the proximal end and has the usual notching for the astragulus. The right fibula rem.ains articulated, but it has not been sufficiently developed to show its characteristics. Both astragali are preserved and both are free from the tibi*. Viewed from above the inner end is obtusely pointed, the posterior border nearly straight, the anterior irregularly convex. The inferior surface is slightly convex antero- posterioiiy. The external end is grooved longitudinally for the fibula. A small rounded ossification attached by matrix close to the outer end of the left astragalus may represent either the calcaneum or a tarsale of the distal row. The hind foot of the left side (See PL XV) is the most perfectly preserved. The five metatarsals are articulated, and the phalanges of the first three digits are all present and but slightly disturbed. It shows that the first three toes of the pes bore terminal claws, with a phalangial formula of 2, 3, and 4. The other phalanges are missing, while the right foot shows a proximal phalanx on digit IV. The metatarsals closely resemble those of Morosaurus grandis, figured by Marsh Sitzungsber. Geo., Naturforsch. Freundc. zu. Berlin, 1909, pp. 293-209, pi. 2. Proc. Wash. Acad. Sci., XII, 1910, No. 1, pp. 1-25, text-figs. 1-7. 382 MEMOIRS OF THE CARNEGIE MUSEUM. {Dinosaurs of North America, pi. 29, fig. 2) except for their much smaller size. Metatarsal I is shortened and the stoutest of the series. COMPARATIVE MEASUREMENTS. Camarasaurus lentils C. grandis C. supremus No. 11338, C. M. Yale Mus. Nat. Hist. Right Left Pectoral Girdle and Fore Limbs: Scapula, length 1 1 1093 1665 Scapula, width widest portion of blade 200 ! 635 970 Scapula, width narrowest portion of blade 88 340 Coracoid, length 725 Humerus, length 450 426 864 1135 Humerus, transverse diameter proximal end — 210 381 500 Humerus, transverse diameter distal end — 160 260 360 Idna, length 300 310 628 — Ulna, transverse diameter proximal end — 108 Ulna, transverse diameter distal end 64 65 Radius, length 298 292 — Radius, transverse diameter proximal end — 73 Radius, transverse diameter distal end — 72 Metacarpal I, length 107 104 — Metacarpal II, length 140 — 245 Metacarpal III, length 135 134 — Metacarpal IV, length 120 122 Metacarpal V, length 1 19 107 — Sternal bone, length 283 — 675 Sternal bone, greatest width 200 — 330 Pelvic Girdle and Hind Limb: Ilium, length 1150 Pubis, length 385 410 980 Pubis, width proximal end 185 155 415 Pubis, width distal end 125 148 460 Ischium, length 380 — 488 1160 Ischium, width of proximal end antero-posteriorly 110 — Ischium, width of distal end, transversely 56 — . 185 Femur, length 555 580 1050 1270 Femur, transverse diameter proximal end — 208 340 450 Femur, transverse diameter distal end 154 182 290 440 Tibia, length 350 350 1040 Tibia, transverse diameter proximal end 135 145 415 Tibia, transverse diameter distal end 119 115 355 Astragalus, transverse width 109 110 380 Metatarsal I, length 70 70 — Metatarsal II, length — 90 — Metatarsal III, length — 88 225 Metatarsal IV, length — 80 Metatarsal V, length — 60 — CHARLES W. GILMORE: SKELETON OF CAMARASAURUS. 383 The linear proportions of the several segments of the skeleton are as follows: mm. In Skull and jaws 330 13.2 12 Cervical vertebr® 1020 40.4 12 Dorsal vertebrae 952 38.0 5 Sacral vertebrae 420 16.8 53 Caudal vertebrae 2355 94.2 Length of combined axial skeleton 5077 203.0 Estimated height of back bone at hips 1240 49.6 Reconstructed Skeleton of Camarasaurus. On Plate XVII is presented a reconstruction of the skeleton of Camarasaurus based on the magnificent specimen described in the preceding pages. It bears the distinction of being the first restoration ever made of a sauropod dinosaur based exclusively upon the remains of a single individual. The restoration represents the animal in a quadrupedal pose as viewed from the left side. It at once reveals the unusual proportions of Camarasaurus, in having a relatively short neck, long body and deep thoracic cavity. It is to be regretted however, that such a disparity of proportions are found between it and the artistic reconstructions and restorations of Camarasaurus supremus so recently presented by Osborn and Mook,^^ as to render the latter unfit for further use in depicting the characteristic proportions of the genus Camarasaurus. On the evidence of this complete articulated back bone, the relative propor- tions of the several segments of the vertebral column may now be considered as absolutely determined. One of the unusual features brought out in the Osborn-Mook restoration was the elevation of the shoulders, making it the highest point in the backbone. This was brought about by placing the scapula in a nearly vertical lizard-like position. The evidence afforded by the present specimen does not support such a conclusion and in the present restoration the scapula is given a more horizontal attitude overlying the sides of the ribs in conformity with the evidence of the right scapula, which was found in place, as shown in Plate XIII. Anything less than a vertical position of the shoulder-blade necessarily results in a lowering of the shoulders below the level of the sacrum, thus bringing the contour of the back into conformity with other sauropod restorations, such as those of Apatosaurus and Diplodocus. Past experience has shown that evidence fur- nished by articulated parts found in situ are far more to be relied upon for structural features than any number of “expert opinions,” and this specimen appears to furnish another example of that fact. Memoirs Amer. Mus. Nat. Hist., vol. 3, pt. Ill, 1921, pis. 83, 84, and 85 and text fig. 28. 384 MEMOIRS OF THE CARNEGIE MUSEUM. Another departure in this reconstruction from the usual restoration is the elevation of the anterior caudal region, which carries the tail well out over the ischia before it commences to droop toward the ground. This innovation was adopted upon the evidence furnished by the mounted skeleton of Apatosaurus louisce Holland in the Carnegie Museum, where a similar elevation is developed from the actual articulation of the bones. The correctness of this pose in Apato- saurus is further substantiated by a study of two Diplodocus skeletons, both of which indicate that a similar condition exists in that genus. Reference is first made to the mounted skeleton of Diplodocus carnegiei Hatcher in the Carnegie Museum, which is obviously incorrect in this region, as shown by the V-shaped space between the centra of the posterior sacral and first caudal; the non-articulation of their zygapophyses ; and the wide space between the tops of their respective spinous processes. A skeleton of Diplodocus now being prepared in the U. S. National Museum in which the sacrum and anterior caudals were found articulated shows that the faces of the centra mentioned above are parallel and that the tops of the spines of the vertebrae are uniformly spaced, and, when viewed from the side, the curve of the caudals corresponds precisely with the mounted Apatosaurus tail so skillfully articulated by Mr. A. S. Coggeshall. From the above evidence it was considered reasonable to believe that Camara- saurus probably had a similar upward arcuation of the tail. The thoracic cavity is unusually deep, and, although some allowance has been . made for their having been lengthened by the straightening out of their natural curves, it may be that they are still too long, as represented in the restoration. Another interesting feature brought out in this reconstruction is the great length and slenderness of the cervical ribs. In the matter of length they rival those of the recently described Uintasaurus douglassi Holland.'*^ In the present pose the neck curves upward from the body with the head slightly elevated in relation to the neck. The relatively low spines of the vertebrae, the stout limbs, and the slender ischia are all features characteristic of the genus, and, while the actual articulation of the bones of a skeleton oi ’ Camarasaurus will probably show the present reconstruction to be in error regarding some details, it is believed to give a fairly accurate representation of this interesting animal. Annals of the Carnegie Museum, XV, 1924, pp. 119 to 138. Pis. X to XIII. MEMOIRS CARNEGIE MUSEUM, VoL X. Plate XIII. Skeleton of Camarasaurus lentus (Marsh). (C. M. Cat. Vert. Foss., No. 11,338.) In position as fonnd. One-fifteenth natural size. Photographed by Arthur S. Coggeshall. MEMOIRS CARWfiGIE MUSEUM, VoL X. Plate XIV. Skeleton of Camarasaurus lenius (Marsh). (C. M. Cat. Vert. Foss., No. 11,338.) One-twelfth natural size. Drawing by Sidney Prentice, designed as a key to plate XIII. Left femur at base of neck, left thoracic ribs, left scapula, left tibia, foot, and left pubis not indicated in this figure. Sternal plate, originally at left of tail, shown beneath the neck. MEMOIRS CARNEGIE MUSEUM, VoL X. ^ Plate XV. Skeleton of Camarasduriis lentus (Marsh). (C. M. Cat. Vert. Foss., Xo. 11,338.) From a photograph by Arthur S. Coggeshall, showing the skeleton as now mounted and displayed. About one-eighteenth natural size. A few displaced bones have been re- articulated; the left ilium of another individual has been introduced; the tail has been straightened. MEMOIRS CARNEGIE MUSEUM, Vol. X Plate XVL right of orl)ital cavity. About two-thirds natural size. to be made entirely MEMOIRS CARNEGIE MUSEUM, Vol. X. Plate XVII. This restoration is based on the nearly complete skeleton shown in plates XIII and XIV. It is the first restoration of a sauropod dinosaur to be made entirely on the evidence of a single individual. Restoration by Mr. Sidney Prentice under the direction of the author. About one-tenth natural size. 7 SIOOTHAD SmOMaM ' , ' ■ ■■ ^ , — TT- rtn-N iriintn-- - m - i ri fi -|- t iri^ilffiiy^TiVrilh^Ti^illfirifg-fMaf v; :si MEMOIRS OF THE CARNEGIE MUSEUM VoL. X. No. 4. OSTEOLOGY OF ORNITHOPODOUS DINOSAURS FROM THE DINOSAUR NATIONAL MONUMENT, UTAH. By Charles W. Gilmore. Part I. On a Skeleton of Camptosaurus medius Marsh. Part 11. On a Skeleton of Dryosaurus altus Marsh. Part HI. On a Skeleton of Laosaurus gracilis Marsh. (Plate XVIII.) INTRODUCTION. Three partly articulated skeletons pertaining to the Ornithischian genera Camptosaurus, Dryosaurus, and Laosaurus, in the paleontological collection of the Carnegie Museum, are of unusual interest in that they furnish information concerning the skeletal anatomy of these genera, which has long been obscure, if not previously unknown. All three of these specimens were collected by Mr. Earl Douglass and his assistants from the famous Dinosaur National Monument (Quarry, discovered by Douglass in 1909, near Jensen in northeastern Utah. The present paper gives the results of a study of these specimens. Part 1. On a skeleton of Camptosaurus medius Marsh. Since the skeletal anatomy of the genus Camptosaurus has lieen set forth in considerable detail in a previous paper \ it seems only necessary at this time to supplement that description by such new features as are disclosed by a study of a partially articulated skeleton of Camptosaurus medius (No. 11,337, Carnegie Museum) . * Gilmore, C. W., Proc. U. S. Nat. Mus., vol. 36, 1909, pp. 197-332. 385 386 MEMOIRS OF THE CARNEGIE MUSEUM. The skeleton, consisting of the greater portion of the articulated back-bone and ribs, together with the more or less disarranged limb and pelvic bones, was received at the Museum in a single large block of sandstone. The vertebral column with attached ilia and ribs has been worked out in relief, while the slightly dis- placed limb and pelvic elements, with the exception of the left fore limb and foot, have been entirely extracted from the matrix. The skeleton is lying on its right side with the neck curved strongly backward above the anterior dorsal region (See Plate XVIII).. The vertebral column is completely articulated from the axis back to the fourteenth caudal vertebra. This is the first camptosaurian siiecimen discovered, in which the complete presacral series can be positively determined. The ribs are distended, a few slightly disarranged, but nearly all are articulated with their respective vertebrae. None of the disturbed bones were far removed from the vertebral column in the rock, and since this was the only small dinosaurian siiecimen found in this jiart of the quarry there is no reason for not regarding all of the associated bones as belonging to a single individual. The following bones of the skeleton are present: eight cervicals; seventeen dorsals; five sacrals; thirteen caudals; both ilia; both pubes; both ischia; both scapulae; left coracoid; both humeri, left radius, ulna, carpus, and manus; both femora; both tibiae; both fibulae; thirty-one thoracic ribs (several incomplete); three posterior cervical ribs; two chevrons, and ossified tendons. The important parts missing are the skull, atlas, hind feet, a few anterior cervical ribs, chevrons, and the distal half of the tail. The medium size of the specimen together with characters found in the pelvis, such as the vertical narrowness of the ilium, with shortened oblique border of the supero-posterior end, the general slenderness of the ischia with moderately expanded distal extremities, indicate that this specimen should be assigned to the species Camptosaurus 7nedius Marsh. Camptosaurus medius Marsh. Camptosaurus medius Marsh, Amer. Jour. Sci. (3), vol. 48, 1894, p. 85, pi. III. The status of this species is not altogether satisfactorily established, and it never can be until the type has been fully prepared and described. Although based on an adequate specimen. Marsh’s original description consisted of a few lines without definition, accompanied by figures of the skull. In revising the genus in 1909,^ C. medius was retained as a valid species and an attempt was made to characterize it. Owing to the unprepared state of the greater portion of the type materials, which were studied at that time, the results obtained were far " Op. cit. GILMORE; ORNITIIOPODOUS DINOSAURS FROM UTAH. 387 from satisfactory. Attention was called to the composite nature of the illustrated skull, and the statement made that it could not be relied upon for specific dif- ferentiation. At present the species rests upon its intermediate size and characters found in the pelvic bones. In my previous study of this species it was anticipated that other and more important specific characters might be disclosed by a study of the other portions of the type specimen, when they should become available. I am not so sure of that now. A study of the excellently preserved skeleton now before me, of the proper size and with a pelvis typically like C. medius, fails to reveal any differences in the other parts of the skeleton which can be regarded as of specific importance. If a similar condition is eventually found in the type specimen, it will certainly permit the suggestion that perhaps after all the observed differences represent sexual characters only, and that C. medius may be the female of one of the larger species of the genus. Description of Specimen No. 11,337, C. M. Cat. Vert. Foss. Vertebral column: — The articulated presacral series in the specimen now before me consists of 25 vertebrae, commencing with the axis and continuing back- ward in sequence to include the vertebra usually designated as the sacro-dorsal. This articulated series has a total length of about 1206 mm. of which the cervicals contribute 391 mm. The first eight vertebrae are regarded as belonging to the neck. The point of division between cervical and dorsal series is marked by the sudden transposition of the capitular facet from the anterior lateral surface of the centrum on cervical nine to a point well up on the side of the arch beneath the transversely extended and strongly developed diapophysis on the succeeding vertebra. Thus it is positively shown, for the first time that the dorsal series series consists of 17 vertebrae. It now becomes necssary to amend the vertebral formula as formerly determined by me in 1909,^ by the addition of one more dorsal making 17 in all. The vertebral formula will now stand as follows: Cer- vicals-9; dorsals-17; sacrals-4 or 5; caudals-44+. In the European representative of this group, Igaunodon hernissartensis , Dollo‘‘ recognizes 9 cervicals, 17 dorsals, and 1 lumbar, a total of 27 presacrals or one more than in Carnptosaurus. ^ Op. cit., p. 224. ^'Bull. Bruxelles. Mus. Roy, d’Hist. Nat. de Belgique, vul. 2, 1883, p. 245. 388 MEMOIRS OF THE CARNEGIE MUSEUM. It may be that the vertebra here designated as sacro-dorsal (Plate XVIII, dll) does not bear a rib and in that event lumbar would be the more apiiropriate designa- tion. In this specimen the diapophysis of the left side has been freed from the matrix, but no trace of a rib was found, although the vertebrae preceding it are all articulated with double-headed ribs. If present, the rib carried by this vertebra would lie single-headed, as shown by the disa})pearance of the capitular facet. The suspected jiresence of a single-headed rib presents nothing new in dinosaurian anatomy, as Thescelosaurus neglectus Gilmore^ from the Uiiper Cretaceous has no less than four single-headed ribs preceding the sacral series, and the closely related Dryusaurus (See p. 400) also has a posterior single-headed rib. In the i)resent specimen there are five vertebra?, which are strongly joined to the ilia, and, while these are all regarded as belonging to the sacrum, it is quite evident that the ])osterior one is a modified caudal. It has a shortened centrum that is not suturally united with the centrum preceding it. The other sacrals, including the sacro-dorsal, have their centra suturally united, though none have liecome coalesced. This feature, including the distinct neurocentral sutures of the presacral series, clearly indicates the immaturity of the individual. The five sacral vertelirse have a combined length of 243 millimeters. I am unable to de- termine whether the centra were joined by the peg-and-notch articulation, which is so characteristic of Camptosaurus dispar Marsh and jiresent to a less degree in C. bruivni Gilmore. The other features of the sacral vertebrae are not to be observed in this specimen, since they are either covered by the overlying ilium or remain liuried in the matrix. There are thirteen anterior caudal vertebrae articulated in seipience with the sacrum. The first shows a distinct chevron-facet, and, although the chevron-bone is missing, it is (piite evident that this vertebi-a bore the first of the series, whereas in C. nanus and C. browni the first chevron is carried on the second caudal® and this is apparently the condition found in C. dispar, as clearly indicated in Marsh’s restoratioid of that species. Flattened transverse processes are present on the first eleven vertebra? counting from the sacrum, whereas in C. browni, and C. dispar they continue backward to the twelfth and thirteenth caudals respectively. It woidd appear' that the distal extent of these processes vary with the individual and thei-cfore do not constitute a specific distinction. ® Proc. U. S. Nat. Mus., vol. 49, 191.5, ]). 609. ® Proc. U. S. Nat. Mus., vol. 36, 1909, p. 239. ^ 10th .4,1111. Reiit. U. S. Geol. Surv. for 1894-95, jit. 1, 1896, pi. 56 GILMORE: ORNITHOrODOUS DINOSAURS FROM UTAH. 389 A com]:)arison of tlie vertebrae of Camptosaurus medius, as represented in the Carnegie Museum specimen, vath those of the other described species, insofar as they can be compared, fails to disclose any siiecific differences, excejit their intermediate size. Ribs\ — The importance of the present specimen is further emjihasized by the presence of nearly all of the thoracic ribs, 31 in all, the greater number remaining articulated with their respective vertebra?. This is the only s])ecimen known at this time, which gives absolute information as to the ]iro])er sequence of the dorsal ri]3S, and it will hereafter be the standard for interpreting and coordinating the scattered and isolated ribs of others of its kind. All of the ribs vdth the exception of the last (if the seventeenth dorsal bears a rib) are double-headed. The first is hardly more than an attenuated cervical rib, without distal expansion, and probably not connected with the sternum. It has a greatest length of IGO mm., measured from the tuberculum. The second rib is progressively longer and more rol^ust, and also without distal expansion. The ends of this pair of ribs are incompletely preserved, l)ut from the taper of the shaft it is quite apparent that it was pointed. Its shaft is more curved than the first and the articular end has assumed the shape of the more typical ril^s of the median thoracic series (See Plate XVIII). In all probability it was not joined to the sternal plates. The third, hov^ever, is develoj^ed into a fully shaped thoracic rib having a truncated slightly expanded distal end, which unquestionably articu- lated with a cartilaginous rib. The center of the rib is flattened, but the lower half is more rounded than those succeeding it. The fourth, fifth, and sixth are about subecpial in length, the longest and heaviest of the series. It was previously thought from fragmentary evidence that the seventh dorsal rib was the heaviest, but in this specimen the fifth has that distinction. The shafts are flattened, broad, with a strong median longitudinal swelling on the external side toward the distal end. The ribs posteriorly become progressively shorter, with a gradual narrowing of the shaft, and the flattened branch carrying the capitulum of the anterior ribs becomes more rounded. The space between the capitulum and tuberculum gradually shortens and the tuberculum is gradually reduced to a small and very weak articular attachment. The posterior members of the series which lie within the forward processes of the ilia are directed strongly forward. Unfortunately the distal ends of all the posterioi- half of the series are incomplete and their precise length cannot be determined. That Camptosaurus had a broad, rounded back is clearly indicateil by the strong curve of the ribs beyond the transverse processes with which they articulate. 390 MEMOIRS OF THE CARNEGIE MUSEUM. Shoulder-Girdle, Fore Limb, and Foot: — Both scapulse and one coracoid are all that remains of the pectoral arch. Fortunately the right scapula (See Plate XVIII) has been retained in the position in which it was found and gives the first evidence of the proper position and relative angulation of the shoulder blade, thus defining the chest. The horizontal position of the scapula in the hadrosaurian dinosaurs has long been established upon the evidence of numerous specimens having these bones preserved in position. It is, therefore, of interest to find in the bipedal Camptosaurus a somewhat similar ]3osition of the scapula, as shown in Plate XVIII. It is shown by this specimen that these bones are incorrectly articulated in the three mounted skeletons of this genus now exhiliited in the American Museum of Natural History and in the United States National Museum.® The blade has a more horizontal position on the sides of the ribs, though not so extreme in this respect as in the Hadrosauridie. Viewed from the side, the upper or distal extremity falls below the level of the vertebral centra in the articulated skeleton, and the truncated distal end reaches a point slightly jiosterior to the fifth rib. In this ]iosition the anterior or proximal end falls below the first dorsal, instead of the posterior cervicals, as in the mounted skeletons mentioned above. The left scaiiula, as shown in figure 1, is perfectly pre- served and jiresents an outline of the upper extremity of the blade somewhat different from that found in C. hroumi and C. nanus. This scajiula has a greatest length of 290 mm. with a greatest breadth of the expanded blade of 117 mm. In the articulated skeleton as found it was pushed out of its pro])er position as clearly shovTi in Plate XVIII. The articulated left fore limb, carpus, and foot are beautifully preserved, as shown in figure 2. The strong flexure of the elbow -joint is indicated and also the fact that in the hanging pose of the fore legs, when walking erect on the hind limbs, the palms of the feet would be directed strongly toward one another. The carpus and foot are complete, except for the loss of the ungual phalanx of digit III. It agrees in every iiarticular with the descrilied® foot of Cam.ptosaurus dispar, oxce]it for its smaller size. ® IToc. U. S. Nat. Mils., vot 36, 1909, pi. 19; vol. 41, 1912, pis. 56, 57. ^ Proc. U. S. Nat. Mus., vol. 36, 1909, pp. 251-256. Fig. 1. Left scapula and coracoid of Camp- tosaurus medius Marsh (No. 11,337 C. M. Cat. Vert. Foss.), viewed from the side. One- fourth natural size. % GILMORE: ORNITHOPODOUS DINOSAURS FROM UTAH. 391 The importance of the present limb and foot lies in its verification of the structure of previously described fore feet, and the more accurate information furnished regarding the proper articulation of the bones at the elbow-joint, so clearly demonstrated in figure 2. The compact ossified carpus, with smooth, well defined articulating surfaces for the radius and ulna, short and stout metacarpals, are indicative of a foot, the func- tion of which was for support, rather than for use as a grasping or prehensile ajipend- age. It presents a striking contrast to the largely unossified carpus and elongated metacarpus of the bipedal, herbivorous hadrosaurian dinosaurs of the Upper Cretaceous. From a posterior view of the carpus only seven carpal elements can be detected. Three of these form the proximal row, the others represent carpals Fig. 2. Left fore limb and foot of Camptosaunis medius (No. 11,337 C. M. Cat. Vert. Foss.). One-fourth natural size, viewed from the internal side. The palmar side of the foot is shown, h. humerus; ra, radius; ul, ulna; r, radiale; in, intermedium; u, ulnare; C2, C3, C4, C5, carpals two to five respectively; me I metacarpal one. I, II, III, IV, V, digits one to five. two, three, four, and five of the distal row. Carpal one, as known from other specimens, is not visible from a posterior aspect of the foot. The humerus, radius, and ulna, except for their intermediate size, show no characteristics distinguishing them from the homologous elements . of the other described species. Pelvic Girdle and Hind Lhnhs: — Both ilia remain attached to the sacrum, but only the left is available for comparison, and it has suffered the loss of portions of the anterior and posterior extremities. With our present knowledge of the CamptosauridcB, this is one of the few characteristic bones of the skeleton. The proportions of the ilium are of the narrow depressed type found in Camptosaurus medius and C. hrowni, which reached its extreme development in C. depressus. 392 MEMOIRS OP THE CARNEGIE MUSEUM. The narrowness of the ilium above the acetabular border as compared to its length, the relatively narrow preacetabiilar notch, and the apparently short oblicpie siipero-posterior border are in close accord with the ilium of the type of Camptosaurus mcdius, and these features at once distinguish it from the ilia of C. dispar, and C. nanus. From C. broumi, however, except for its smaller size, the distinction is not so clearly established. The piupubis is relatively shorter than in the type of C. medius with a more squarely truncated end. The postpubic portion is long, slender, and terminates with slight expansion of the distal end. The pubic foramen exists as a notch, whereas in the type it is entirely inclosed, but this is probably a characteristic of age. The ischium agrees perfectly with the type of the species in being slender, with a small distal hammer-like expansion, as contrasted with the more robust development in C. dispar and C. hroivni. The bones of the hind limbs, except for their smaller size, show no distinctive features, and, since these elements of Camptosaurus have been fully described, it ajiiiears unnecessary to mention them further at this time. Summary The important facts determined from a study of this well preserved and partially articulated specimen may be summarized as follows: First, that the complete presacral series in Camptosaurus consists of twenty-six vertebrae; second, that the most jiosterior presacral may not bear ribs and should therefore be con- sidered a lumbar as in Iguanodon; third, that the third thoracic rib is the first one of the series to be attached to the sternum ;< fourth, that the first caudal carries the first chevron; fifth, that the scapula assumes a more horizontal position and occupies a lower and more posterior place on the side of the ribs than has been given it in articulated skeletons; sixth, that while the manus can be rotated into other jiositions, the normal hanging pose would be with the palms of the hands directed strongly toward one another. Six species of Caynptosaurus have been described from North America: C. dispar, C. medius, C. amplus, and C. nanus by Marsh; C. browni and C. depressus by Gilmore. The genotyiie, C. dispar, is satisfactorily established; C. nanus is a good species, distinguished at once by its small size; C. medius, the type of which yet remains to be fully described, appears to be a distinct species, distinguished at present by characters found in the pelvis, which with its smaller size and slenderer structure, permit the suggestion that it may eventually be found to represent the female of the larger C. dispar, and that its fully adult development may be repre- sented in C. broivni. While for the present it may be well to retain the last men- GILMORE: ORNITHOPODOUS DINOSAURS FROM UTAH. 393 tioned species, after several years reflection and with a wider knowledge of dino- saurian anatomy, I am inclined to the opinion that no good reason ever existed for its establishment. C. amplus at present is distinguished by its very large size. Camptosaurus depressus, although founded on a rather poor specimen, is probably justified on the ground of its geological occurrence in the Lakota formation. That the present specimen, here attributed to Camptosaurus medius, is not fully adult is abundantly indicated by the distinct sutures and especially by the noncoalescence of the sacral vertebrae. This immature condition may account for the intermediate proportions of the animal, as is so clearly demonstrated in the table of comparative measurements which follows: Comparative Measurements Camptosaurus nanus Type, 2210, U. S. N. M. Camptosaurus medius No. 1 1337, Carnegie Mus. Camptosaurus browni Type No.4282,U.S. Nat. Mus. Camptosaurus dispar Type No. 1877, Yale Mus. Camptosaurus dispar Paratype, No. 1877a, Yale Mus. mm. mm. mm. mm. mm. Pectoral Girdle and Fore Limbs: Scapula, length 187 290e 482 - - Scapula width, widest portion of blade 59 117 175 - - Coracoid, length 35 68 115 - - Humerus, length 143 227 360 337 - Ulna, length 102 IGO 262 260 - Radius, length 45 143 232 245 - Metacarpal I, length - 10 - - - Metacarpal II, length - 35 - - - Metacarpal III, length - 45' - - - Metacarpal IV, length - 41 - - - Metacarpal V, length - 22 - - - Pelvic Girdle and Hind Limbs: Ilium, length 244 400e 618e - - Ilium, vertical height above middle of acetabulum 45 70 115 - 132 Pubis, length - 224 - - 348 Pubis, width of preacetabular expansion - 55 - 72 Pubis, length of postacetabular process - 320 - - 528 Ischium, length - 380 545 - 600 Ischium, greatest width of distal end - 50 79 - 96 Femur, length 258 395 - 565 - Tibia, length 235 360 - 555 - Fibula, length 207 320 - - - Sacrum, combined length of 4 sacral centra 120 200 288 323 - Neck, length of combined nine cervicals 20 Ge 39 le - 565 - e = estimated. 394 MEMOIRS OF THE CARNEGIE MUSEUM. Part II. On A Skeleton of Dryosaurus altus Marsh. A jiartly articulated skeleton (No. 3,392, C. M. Cat. Vert. Foss.) is identified as belonging to Dryosaurus altus Marsh. This specimen is not yet fully prepared, but the skeletal parts available for study contribute so much to a better understanding of this little known form, that it seems advisable to describe it in advance of its complete preparation. At present the skeleton is in three principal sections. The skull and lower jaw, articulated with the first five cervical vertebrie, constitutes the first section; the second consists of an articulated series of seven anterior dorsal vertebrae with the proximal portions of the ribs of both sides in position, together with the right scapula, coracoid, and humerus; the third is made up of six posterior dorsal vertebrae articulated with the sacrum, and portions of the pelvic arch. In addition the proximal third of the right femur, proximal end of the right tibia, one complete metatarsal and distal portions of two others have been completely freed from the rock. The articulated left hind foot, with some of the limli and other identified bones, still remain unprepared in a block of sandstone. Dryosaurus altus Marsh. Laosaurus altus Marsh, Amer. Jour. Sci. (3) vol. IG, 1878, pp. 415-416, pi. IX. Dryosaurus altus Marsh, Amer. Jour. Sci. (3) vol. 48, 1894, p. 86. The genus Dryosaurus was established by Marsh on the type of the species Laosaurus altus. At the time of publishing the original description, a tooth, the articulated pelvis, hind limb, and foot were figured. These are the only bones of the type material which have ever been illustrated, and in the many years which have intervened since its establishment, nothing further has been contributed to our knowledge of this genus and species. The brief original description, charac- terization of the genus, and illustrations were republished by Marsh in various places, but there were no additions and but few emendations. It is, therefore, of great interest to find a specimen in the collections of the Carnegie Museum which contributes to a further elucidation of its skeletal anatomy. The final characterization of the genus Dryosaurus by Marsh'” is as follows: Premaxillaries edentulous with horny beak. Teeth of moderate size. A sipiraorbital fossa. Cervicals long and biconcave. No lumbars. Six ossified vertebrae in sacrum, without peg and notch articulation. Sternum unossified. Limb bones hollow. Forelimbs very small. Five digits in manus. Prejiubis long and narrow; postpubis elongate and slender. Posterior limbs very long. Femur shorter than tibia. Meta- tarsals long and hollow. First digit in pes complete; fifth metatarsal reiiresented by short splint only. 16th Ann. Rept. U. S. Geol. Surv., pt. 1, 1896, p. 201. GILMORE: ORNITITOPODOUS DINOSAURS FROM UTAH. 395 The above definition, when compared with that given i)y the same author" for the closely allied genus Laosaurus, shows but two important distinctions, i. e. “cervicals long and biconcave” as contrasted with “cervicals short and flat,” and “prepubis long and narrow” as opposed to “prepubis very short and pointed.” A comparison of Dryosaurus and Laosaurus s]:)ecimens in the Carnegie Museum unfortunately does not permit of a verification of these characters. On the other hand they show such close resemblances in skull and other features as to raise the question of their generic distinctness. It remains, therefore, for the tyjie specimens in the Yale Museum to be fully prepared and described before judgment can be passed on the query here raised. Should it be shown that these two genera are not distinct, Dryosaurus v-ould become a synonym of Laosaurus, which has priority by several years. The large size of the present individual, larger than any described species of Laosaurus, the presence of a relatively long prepubis (though incomplete), and other lesser features are in accord with Marsh’s definition and illustrations of Dryosaurus altus, the genotype, and for these reasons it is so referred. Fig. 3. Skull of Dryosaurus altus Marsh. No. 3,392 C. M. Cat. Vert. Foss. Viewed from the left side. One-half natural size. Slightly restored after Laosaurus. anf, anteorliital fossa; an, angular; d, dentary;/, frontal; inj, infratemporal fossa;, fw, .jugal; la, lachrymal; mx, maxillary; nq, nasal; o, orbit; p, parietal; pd, predentary; pf, postfrontal-f postorbital comjilex; pmx, premaxillary; qj, quadratojugal; qu, quadrate; s, supraorbital; sof, supraorbital fossa; sq, squamosal; stf, supratemporal fossa; s«r, sur- angular. Unknown or obscure sutures are indicated by broken lines; restored parts are unshaded. Description of No. 3,392, C. M. Cat. Vert. Foss. The SJcuU: — In the s])ecimen now before me the left side of the skull and lower jaws are preserved. They are little distorted by crushing and give for the first time a very clear conception of the lateral profile of the skull. In figure 3, is " Loc. cit. p. 201. 390 MEMOIRS OF THE CARNEGIE MUSEUM. shown a side view of this skiill and lower jaws. Slight corrections have been made where distortion was apparent, and some restoration was necessary where parts were missing or obscure in the specimen. Viewed from the side (See fig. 3), the skull of Dryosaurus with the jaw in jiosition is relatively short, bluntly pointed in front, moderatel}^ high behind, and with a large subcircular orbit. The orbit is about one-fourth the total length of the skull and is placed well toward its middle. The infratemporal fossa is sub- triangular in outline with the apex directed ventrally. Especially noteworthy is the presence of a large enclosed supraorbital fossa, and a relatively large anteorbital foramen or fossa, which lies wholly within the maxillary bone. Marsh'^ recognized the presence of a supraorbital fossa in the Dryosaurus material studied liy him, but that is about the only mention made of cranial characters. In Marsh’s restoration‘s of the skeleton of Laosaurus, the skull is depicted, but how much of this figure is based on actual cranial materials and how much is conjectural has never been stated. If the restored skull of Laosaurus is well founded, it furnishes important characters for the more distinct separation of these two closely allied genera. When compared with the Dryosaurus skull now before me, many differences are to be observed, so that at this time I have grave doulits of its authenticity. The supraorbital bar, which forms the external boundary of the supraorbital fossa, is relatively heavy, and presents a striking contrast to the slenderer and incomplete bar found in the allied Camptosaurus.^* This bar in Dryosaurus is probably formed by two supraoiTital bones, an anterior and a posterior, which have coalesced where they meet above the orbit. In Iguanodon hernissartensis, Dollo has found two such elements, but up to this time only an anterior supra- orbital bone has been recognized in Camptosaurus. The lateral fenestra in the maxillary, so far as I can discover, has not before been observed in a member of the American Ornithischia. A small anteorbital vacuity is present in Iguanodon. A skull of Laosaurus gracilis, which I shall presently describe, has a similar fossa, but none was shown in Marsh’s restoration of the skull and skeleton of Laosaurus consors mentioned above. If at all comparable with the lateral fenestra of the theropod skull, it would represent the second ante- orbital fenestra, which in those forms also lies wholly within the maxillary bone. It certainly is not to be correlated with the first, as that opening is always bounded posteriorly by the lachrymal. Slight displacement of the left maxillary, which Amer. Jour. Sci. (3) vol. 48, 1894, p. 88. 16th Ann. llept. U. S. Gcol. Survey, Pt. I, 1896, PI. 57. Gilmore, Charle.s W., Proc. U. S. Nat. Mus., vol. 36, 1909, figs. 2 and 3. GILMOHE: ORNITHOPODOUS dinosaurs from UTAH. 397 apparently follows the lines of sutural contact, clearly discloses the outline of this as well as of a few other cranial elements, but for the most part they cannot be certainly delimited. Whenever in doubt, the cranial sutures in the restored skull (See fig. 3) have been indicated by broken lines. The quadrate curves strongly backward as in Camptusaurus. Superiorly it articulates with the squamosal, but the distal portion is so poorly preserved that its features can not be observed. . It has been restored after the cpiadrate of the Laosaurus gracilis (No. 11,340, C. M. Cat. Vert. Foss.) described in the present paper on pp. 403-409. The squamosal is strongly produced backward and downward behind the top of the quadrate, forming the whole of the outer postero-superior angle of the skull. Whether a portion of this hook-like projection may be a part of the outer end of the paroccipital process cannot be determined. Enough remains of the frontal region to show that it was broad and flat between the supraorbital fossai. The upper parietal surface closely resembles that of Pamptusaurus in being narrow between the supratemporal fossae. The sutural contacts of this aspect of the skull cannot be seen. The lateral border of the frontal is higher above the orbit than in Camptosaurus, and on this account the supraorbital fossa is more conspicuously exposed in a lateral view. It is presumed that the postfrontal and postorbital bones have coalesced to form a complex as in many other dinosaurian genera. The jugal in lateral view forms a relatively wide bar, especially below the infratemporal fossa. Its narrow contribution to the lower boundary of this fossa is in striking contrast to the much more extensive participation in Camptosaurus. In the antero-posterior shortening of the skull posterior to the orbit, Dryosaurus resembles the cranium of Iguanodori. Presumably the jugal meets the lachrymal anteriorly, but the limits of that bone cannot be differentiated in this specimen. The form of the maxillary, as mentioned previously, is c|uite certainly deter- mined, except for its upward extent. It appears to be in contact with the nasal above and to exclude the posterior process of the premaxillary from contact with the prefrontal. The premaxillary is largely missing, only the posteriorly directed process remains, which is interposed between the forward end of the maxillary and the nasal bones. The nose has been wholly restored, as shown in figure 3. The left ramus is nearly complete and presents an accurate conception of its size and contour. Unfortunately none of its sutures can be detected, so that this 398 MEMOIRS OF THE CARNEGIE MUSEUM. siiecimen contributes nothing concerning its more detailed structure. The preden- tary is present, l^ut Iradly crushed and broken. Teeth are present in both upper and lower jaws, but their preservation is such that neither their number nor details of structure can be determined. The iirincipal measurements of the skull and lower jaw are given in the following table: Me.\surements of Skull No. 3,393 C. M. Greatest length of skull, estimated 185 mm. Distance from posterior border of orbit to e.xtremity of scpiamosal 58 mm. Distance from distal end of quadrate to toj) of skull, estimated 90 mm. Height of skull with ramus, at posterior end 112 mm. Height of skull with ramus, taken at center of orbit 103 mm. Height of skull, taken at center of maxillary 55 mm. Length of quadrate, estimated 73 mm. Greatest length of ramus 172 mm. .Liitero-posterior diameter of orbit 48 mm. Vertical diameter of orbit, taken at center 30 mm. Antero-ijosterior diameter of supraorbital fossa ' 36 mm. Vertical diameter of infratemporal fossa 43 mm. Vertebrw. — The vertebral column is reiiresented by parts of 24 vertebrae, separated into three articulated series. The neck is reiiresented by five vertebrae articulated 4vith the skull and commencing with the atlas; the dorsal region by seven anterior dorsals and six posterior dorsals, the latter articulated 4vith six vertebrae which comprise the sacrum. Because of their poor preservation it cannot be determined whether the two dorsal series were continuous or not. The cervicals have not been fully prepared and but little more than an oblique latero-superior view is to be obtained. Insofar as they can be compared they are strikingly like those of Camptosaurus mediiis. The only differences observed are the more slender zygopophyses and the presence of an incipient spinous process on cervicals three and four. These spines are absent in Camptosaurus, but Marsh indicates them on the cervicals of Laosaurus consors in his restoration of that animal. The atlas remains almost completely hidden in the matrix. The axis, especially the neural spine, is clearly seen, and, while it resembles the spine of the axis of Camptosaurus, it differs in having the upper crest more depressed, with a long convex profile, as contrasted with the concave border in that genus. These five cervical vertebrae have an estimated length of 205 millimeters. The series of seven anterior dorsals is thought to begin with the fourth. If correct in this estimate and if Dryosaurus has the same vertebral formula as Camptosaurus, there would be but one missing between the two dorsal series. The poor preservation of the vertebi’ae, however, does not allow of an accurate GILMORE: ORNITHOPODOUS DINOSAURS FROM UTAH. 399 determination on this point. The badly crushed condition of the vertebra3, com- bined with a lack of definition, as to their outline and other details renders an attempted description of but little value. Their principal features are shown in figure 4. The proximal portions of most of the thoracic ribs of both sides re- main articulated with their respective vertebrae, as they are in the posterior series. These seven dorsals have an estimated length of about 360 mm. Fig. 4. Anterior dorsal vertebral with articulated ribs of Dryosaurus alius Marsh. Suiierior view. (No' 3,392, C. M. Cat. Vert. Foss.), about one-sixth natural size, h, right humerus; sc, right scapula and coracoid; t, ossified tendons. The third series (See fig. 5) may be continuous with the second group de- scribed above, but of this one cannot be certain. The preservation of these verte- brae is slightly better than that in the preceding series, but like them only the superior view is available for study (See fig. 5) . ^ ^ Fig. 5. Posterior dorsal vertebra?, sacrum, and jielvis with articulated ribs of ■Dryosaurus alius Marsh. Superior view. (No. 3,392, C. M. Cat. Vert. Foss.), about one-sixth natural size, dor, last dorsal vertebra; il, ilia; p, prepubis; r, ribs; Si, S2, S3, S4, S5, Sg, sacral vertebra; one to six respectively; sf, sacral foramina; t, ossified tendons. The principal features observed in the posterior presacral series is that all of the dorsals preceding the sacrum bear ribs; the spinous processes grow pro- gressively longer and become wider antero-posteriorly, as the sacrum is approached ; the diapophysial, and parapophysial facets gradually approach one another as in Camptosaurus; the transverse processes of the last two dorsals are directed 400 MEMOIRS OF THE CARNEGIE MUSEUM. decidedly forward; the posterior dorsal bears a single-headed rib, all others are double-headed; several of the posterior ribs seem to be coossified with the dia- pophysis, especially the last; the most posterior ribs are directed strongly forward toward the anterior ends of the ilia; the vertebrae are strongly bound together by a series of ossified tendons running along on either side of the spinous processes. There are six vertebrae articulated with the ilia by short, stout, sacral ribs. These vertebrae may all properly be considered sacral, as originally determined by Marsh, and according to him an equal number is found in Laosaurus. Although as many as seven vertebrae have been found united by suture in Camptosaurus,^^ never more than five are directly joined with the ilia by their diapophyses and sacral ribs. This feature therefore constitutes an important structural difference distinguishing the Laosauridoc from the Camptosauridce. It cannot be determined from the present specimen whether the sacral vertebrae are coossified or not. It is estimated that this series of 12 articulated dorsal and sacral vertebrae have a combined length of about 545 mm. Pectoral Arch, Fore Limb, and Foot: — The pectoral arch is represented by the right scapula and coracoid, preserved practically in position in relation to the anterior thoracic region, as shown in figure 4. The scapula and coracoid, insofar as they can be compared, bear a strikingly close resemblance to the homologous bones in Camptosaurus. Unfortunately their full outlines cannot be observed due to lack of complete preparation. The complete length of the two bones is about 245 millimeters. The scapula, as in Camptosaurus medius, previously described, occupies a more or less horizontal position on the upper anterior side of the thorax, and, if the vertebrae of this second series has been correctly identified as to position in the vertebral column, its upper or distal extremity reaches a point slightly posterior to the fifth rib, as in the skeleton of Camptosaurus mentioned above. The right humerus is retained in the matrix immediately below the scapula, as shown in figure 4. Only the distal half has been prepared, but it has the same antero-posteriorly compressed, straightened shaft found in Laosaurus; the obliquely truncated distal end with feebly developed condyles bring especially characteristic. The total length of this bone is 190 millimeters, the same dimensions given by Marsld*^ for the humerus of the type specimen of Dryosaurus altus. The distal end has a greatest transverse diameter of 42 millimeters. A small block of matrix contains a considerable number of the bones of a manus. These elements are somewhat disarranged and their contours are so Proc. U. S. Nat. Mus., vol. 36, 1909, p. 234, fig. 17. Amer. Jour. Sci. (3) vol. 16, 1878, p. 415. GILMORE: ORNITHOPODOUS DINOSAURS FROM UTAH. 401 poorly defined that their detailed structure cannot be determined. Tentatively the following elements are recognized: the distal end of the radius, below which are several carpal elements; and below the carpus four metacarpals which from left to right have the following lengths: Met. V, 18 mm.; Met. IV, 26 mm.; Met. II, 23 mm. ; a few phalangial bones are also present. Pelvic Arch, Hind Limb, and Foot: — The pelvic arch has the greater portion of the right ilium, the anterior superior half of the left ilium, and the left prepubis preserved. The ilia are in position and articulated with the sacrum, and thus give a very clear conception of the superior aspect, as shown in figure 5. The ilium is elongate antero-posteriorly and very narrow vertically. In the latter particular it is relatively narrower than the most depressed type of ilia found in the genus Camptosaurus. Among American dinosaurs it most closely resembles ■ the ilium of Thescelosaurus^’’ from the Upper Cretaceous. Its upper plate-like extension rises very little above the level of the attached diapophyses from the sacrals, whereas in Camptosaurus it is everywhere high above these processes. Viewed from above (See fig. 5) the articulated ilia approach one another closest opposite the second sacral. Anteriorly the long pointed preacetabular processes are strongly divergent; posteriorly their divergence is more gradual, but the total width considerably exceeds the anterior end. On the inner side, behind the aceta- bulum, a relatively thin but wide shelf of bone is given off at right angles to the main portion of the ilium and articulates along its inner margin with the ribs of the two posterior sacral vertebrae. Viewed from above the shelf is V-shaped and unusually wide, the vertical plate of the ilium at the posterior end rises only 6 mm. above the level of the superior surface of the horizontal plate. The posterior termination of the right ilium, external to the shelf, is bluntly pointed. 4 These ilia clearly show the restored ilium of Dnjosaurus by Marsh to be too deep and plate-like above, and posterior to the acetabulum. This same criticism is also probably applicable to his restoration of the ilium of Laosaurus consorsP The right ilium measures 340 mm. from end to end. A lateral view of the restored ilium is shown in figure 6. The articulated ilia have a greatest Fig. 6. Right ilium, vertebrae, and ossified tendons of Dnjosaurus alhis Marsh (No. 3,392, C. M. Cat. Vert. Foss.). Lateral view. One-sixth natural size. Proc. U. S. Nat. Mus., vol. 49, 1915, p. 607. Marsh, 0. C., Dinosaurs of North America, 16th Ann. Rept., U. S. Geol. Surv., pt. 1, 1896, pi. 55, figs. 3 and 4. 402 MEMOIRS OF THE CARNEGIE MUSEUM. expanse anteriorly of 165 mm.; at the middle 100 mm., posteriorly, estimated, 240 mm. The left prepubis (See figure 5) lacks the anterior end and practically all of the slender postpubic portion. It is especially unfortunate that the anterior extremity is missing, for it cannot be determined whether the end was bluntly rounded, or whether it was squarely truncated; but it is quite apparent that the termination could not have been sharply pointed, as figured by Marsh in Laosaurus. It is this character which most definitely separates Dryosaurus from Laosaurus at the present time. The upper border of this bone is thickened and rounded, the lower being thin and sharp. The pubic foramen is closed as in the type. The larger size, narrow blade, and absence of a pointed extremity definitely distin- guishes this prepubis from the short, tapering, sharply pointed prepubis of Laosaurus. The proximal fourth of the right femur has been entirely extracted from the matrix and is in a fine state of preservation. In general aspect it closely resembles the femur of Cainptosaurus, the only outstanding difference being the thickened and much heavier lesser trochanter, as contrasted with the compressed blade-like process in that genus. The greatest transverse diameter of the proximal end measures 80 millimeters. The pes is represented by one nearly complete lateral metatarsal and the distal halves of two others of the right foot. The articulated left foot with a portion of the limb still remains unprepared in a block of matrix. These show the foot to be especially elongated and slender as compared with Camptosaurus. jMetatarsal II has a greatest length of 163 mm.; the proximal end is compressed transversely, but wide antero-posteriorly. The proximal end is distinctly concave antero-posteriorly. The distal portion of the shaft shows no flattened appositional surface and was probably not closely applied to metatarsal III. The distal end is also transversely compressed with a prominent articular surface for the toe, which faces obliquely outward and downward. Metatarsal III has a distinctly grooved distal articular end. It is the heaviest and probably the longest bone of the foot. The left element shows it to have a greatest length of about 170 millimeters. Metatarsal IV is compressed transversely, with a decided flattening of the inner surface for close articulation with the median metatarsal. It would appear from these bones that the lateral elements in the articulated foot, as illustrated by Marsh, are too widely divergent at their distal extremities. 16th Ann. Kept. U. S. Geol. Surv., pt. 1, 1896, pi. 55, fig. 4. GILMORE ORNITIIOPODOUS DINOSAURS FROM UTAH. 403 Part III. On a Skeleton of Laosaurus gracilis Marsh. A third siiecimeii in the Carnegie Miiseuin, consisting of a badly crushed skull and jiartly articulated skeleton of a diminutive bipedal dinosaur, is pro- visionally identified as pertaining to the little known species, Laosaurus gracilis Marsh. Its tentative identification is due to my inability to comjiare it directly with the type specimen in the Yale Museum, which is necessary on account of the very meager original description, and from the fact that no part of the type of the species has ever been illustrated. Marsh proposed three species under this genus: Laosaurus celer, L. gracilis, and L. consors. About the only distinguishing feature given by him for their separation is that of size. Laosaurus consors is the largest and best established, because of his illustration of a tooth and the articulated pelvis, hind limb, and foot.^° In a later publication'’^ a complete skeletal restoration of this species was published. L. celer, the genotype, is of intermediate size, and L. gracilis is the smallest. Whether all three represent valid species, or not, must await a restudy of the type specimens. The genus Laosaurus is well established, and should it be shown, as previously suggested, that Dryosaurus is congeneric, Laosaurus would stand on the ground of priority. That the present specimen is referable to Laosaurus gracilis seems to be indicated by its very small size; liy a scapula and coracoid, which resemble those of L. consors figured by Marsh; and posterior dorsal vertebrae, which agree closely with the measurements given of a ‘‘lumbar” of tlie type specimen. The Ornithopod affinities of this specimen are abundantly shown by the skull, with a supraorbital fossa; a maxillary fossa or foramen; quadrate, jugal, and lower jaw of characteristic shape and proportions; the jiresence of curved femora with pendant fourth trochanter; and the presence of the unmistakable pectoral girdle of this group. The original description given below constitutes practically all that has been written descriptive of this species, and, if the specimen in the Carnegie Museum is correctl}^ referred, it constitutes a most welcome contribution to a better under- standing of this species. Laosaurus gracilis Marsh. Laosaurus gracilis Marsh, Amer. Jour. Sci., vol. 15, 1878, p. 244. The original description is as follows: “A second species much smaller than the above [Laosaurus celer] is ““ Amer. Jour. Sci. (3) vol. IS, 1894, pi. v, fig. 4. 16th Ann. Kept. U. S. Geol. Surv., pt. 1, 1896, pi. 57. 404 MEMOIRS OF THE CARNEGIE MUSEUM. represented by well preserved remains of various parts of the skeleton. Its size is indicated by the following measurements; Lengtli of lumbar vertebrae 16 mm. Transverse diameter of anterior end 18 mm. Transverse diameter of posterior end 17 mm. Length of median caudal vertebrae 16 mm. Transverse diameter anterior face 12 mm. Greatest diameter of proximal end of ulna 17 mm. This reptile is the smallest known dinosaur with the exception of the diminutive species of Nanosaurus {N. agilis and Ah victor)." Description of Specimen No. 11,340, C. M. Cat. Vert. Foss. The specimen consists of a badly crushed skull and lower jaws, articulated with the first three or four vertebrae of the neck, the latter still enveloped by the refractory matrix. Originally, I am informed, these cervical vertebral were in sequence with the remaining part of the vertebral column preserved in a separate block of sandstone, but the contact has now been lost. The second series begins in the median cervical region and is apparently continuous posterior to, and in- cluding, one or more sacral vertebrae. The anterior dorsal region is hidden by the nearly perfect pectoral arch with both humeri in position. Posterior to the arch the column has suffered distortion and partial loss of centra, so that it cannot be determined whether the presacral series is complete or not. In the rock at the posterior end of the vertebral column are a number of fragments of bone, which may represent the pelvic bones, though none could be positively identified as such. In addition the hind limbs are represented by the distal portions of both femora and the proximal ends of the articulated tibia and fibula of the left limb. The skeleton lay upon its back with ribs and fore limbs widely distended. It has been partially worked out in relief from this side and consequently, with the exception of those bones completely freed from the matrix, only the vertebral centra and a few ribs are to be observed in ventral view. The short time at my disposal did not permit of the complete preparation of this specimen, but, in view of the rarity of Laosaurus remains, it was thought advisable to prepare a preliminary description, in order to make immediately available to students of the Dinosauria such information as was furnished by the skull and pectoral arch, with brief mention of such other structural features as may be observed in its present unprepared condition. The Skull: — The skull of No. 11, 340, C. M. Cat. Vert. Foss., is so badly crushed and misshapen, that in the attempted restoration, shown in figure 7, there is much left to conjecture in so far as its natural profile is concerned. The general GILMORE: ORNITHOPODOUS DINOSAURS FROM UTAH. 405 structure and shape of the skull appear very close to the cranium of Dryosaurus previously described, as may be seen by comparing figures 3 and 7, the greater de- pression of the nasal region being the one outstanding difference. This bending down of the nasals in front of the orbits gives the muzzle a bluntly wedged-shaped appearance, which is undoubtedly exaggerated in the original by crushing. The nose of the skull has been pinched off and is missing. Fig. 7. Reconstructed skull of Lnosaunis gracilis Marsh (No. 11,340 C. M. Cat. Vert. Foss.) Viewed from the left side, about natural size, anj, anteorbital fossa; d, dentary; /, frontal; ju, jugal; mx, maxillary; na, nasal; o, orbit; pf, postfrontal + postorbital complex; qu, quadrate; s, supraorbital; stf, infratemporal fossa; sof, supraorbital fossa; sg, squamosal; unshaded portions indicate restoration of parts not present in the fossil. On the left side a slender supraorbital bar clearly forms the outer boundary of the elongated supraorbital fossa as in Dryosaurus. It is relatively slenderer than in that genus. The orbits are large, with a greater diameter antero-posteriorly than vertically. On the right side the complete quadrate is present and articulated below with the lower mandible. Its upper extremity is capped by a remnant of the squamosal. It is strongly curved from end to end as in Camptosaurus. As preserved, the quadrate is in close contact anteriorly with the postorbital bar and the wide triangular plate-like jugal, but it would seem that on this side the whole postorbital bar has been crushed backward against the quadrate, thus entirely obscuring the outlines of the infratemporal fossa from a lateral view. That this fossa is present is clearly indicated on the opposite side, where the complete postorbital bar remains in natural position and behind it the anterior outline of the fossa is clearly discernible. The postorbital + postfrontal probably forms a complex, as in Camptosaurus and Dryosaurus, at least no sutural separation of these elements can be detected. 40G MEMOIRS OF THE CARNEGIE MUSEUM. The ventral position of this complex forms a flattened, moderately wide bar, which extends dovTiward to join the jugal. The jugal, especially on the right side, is fairly complete. It is narrow belov' the middle of the orbit, but posteriorly Avidens rapidly, forming a subtriangular jilate, the apex of which is directed doMm- MTird and backward. The maxillary bone is perforated by a small elongated foramen, or anteorbital fossa, which lies v’holl}^ within the maxillary, as in the skull of Dryosaurus. The frontal region is liroad and flat. Between the inner liordei’s of the supraorbital fossa the frontals have a greatest transverse extent of 28 mm. The parietal region is so badly crushed and broken that its details are oliscured. The nasals are wide posteriorly, and tui’ii strongly downwartl from their union with the frontals. Their median suture is distinct. The superior surfaces of the conjoined nasals appear to be decidedly concave for a consideralile distance along their median junction. None of the detailed features of the lovur mandible are to be observed. In general outline, so far as iireserved, it resembles the lovur ja\v of Dryosaurus. Measurements. Greatest length of skull, estimated 82 mm. Greatest height of skull with ramus, posterior eud 47 mm. Greatest height of skull with ramus, center of orbit 34 mm. Autero-posterior diameter of orbit 22 mm. Vertical diameter of orbit 17 mm. Autero-posterior diameter of suiiraorbital fossa 20 mm. Length of quadrate 33 mm. Vertehrcc: — The vertebral centra in their general ventral aspect closely re- semble those of the articulated skeleton of Camptosaurus niedius, excejit for their very much smaller size. The cervical centra have the median lateral surfaces pinched in, forming lateral depressions; the capitular facets on the anterior lateral sides widen the forward end, and ventrall}^ there is an angular keel which widens at either end. Posteriorly the ti'ansversc constriction of the centra becomes less and less, so that in the iiosterior dorsal region they become broadly rounded. The whole vertebral column, v'lien viewed from lielow, grows gradually heavier and stronger postei’ioiiy, i-eaching its maximum development in the anterior sacral region. In the present condition of the specimen there is little more to be said of the vertebral column. If the last comidete centrum is correctly identified as being the sacro-doi’sal, it is estimated that the comiilete presacral series would have a total length of about 315 mm. or about 12 inches. GILMORE; ORNITIIOrODOUS DINOSAURS FROM UTAH. 407 The close agreement in the dimensions as given below of the most ])osterior centrum of the present specimen to the so-called “lumbar” of the type of L. gracilis apparentl}" indicates the correctness of its reference to that species. Type of No. 11,340 L. gracilis Carnegie Mus. mm. mm. Length of centra 16 14 Transverse diameter, anterior end 18 16 Transverse diameter, posterior end ' 17 Id.,') Hind Limbs: — The left femur, which lacks its proximal third, has a curved shaft, as in the other members of the family Laosauridw. On the inner hind margin of the shaft a compressed fourth trochanter of the pendant tyjie is developed. The inner condyle is robust with a decided projection backward. There is a shallow anterior intercondylar groove as in Camptosaurus. The greatest transverse diameter of the distal end is 27 mm. The articulated proximal ends of the left tibia and fibula resemble the corre- sponding parts in Camptosaurus. This end of the tibia has a greatest diameter of 21 mm.; the fibula measures 15 mm. The limb bones are hollow, but have thickened walls. It is estimated that the complete length of this animal would not have exceeded two and one-half feet, with an estimated height at the hips of about 12 inches. Fig. 8. Pectoral arch and luimeri of Laosaurus gracilis (No. 11,340, C. M. Cat. Vert. Foss.) ventral view. About one-half natural size, co, coracoid; gif, glenoid fossa; h, humerus; sc, scapula; st, sternal plates. Pectoral Arch and Fore Limb: — The pectoral arch in specimen No. 11,340, C. M. Cat. Vert. Foss, is in an especially fine state of preservation, showing clearly for the first time the relative relationships of these bones in the Laosauridm. The scapulae lie in nearly normal position in relation to the back-lione, with the articu- lated coracoids turned in to meet in close apposition along the median line. Tlie right scapula has a greatest length of 78 mm. Immediately posterior to them are the relatively large paired sternal plates, and on either side, with the proximal ends but slightly withdrawn from the glenoid fossa tlie complete humeri extend outward at right angles (See figure 8) . 408 MEMOIRS OF THE CARNEGIE MUSEUM. Humerus'. — The humerus in Laosaurus is slender and but little expanded at distal and proximal ends. These bones are relatively straighter than in Campto- saurus, with deltoid crest but feebly developed, condyles ill defined, and with head but slightly indicated. It appears to differ from Camptosaurus in having the greatest diameter of the proximal and distal ends in the same plane, whereas in Camptosaurus they are twisted so as to be at a slight angle to one another. The distal end is obliquely truncated, as in Dryosaurus, and to a less degree in Camptosaurus. Measurements. Right Left Greatest length of humerus Gl mm. 63 mm. Greatest width of proximal end 14 mm. 14 mm. Greatest width of distal end 11 mm. 12 mm. Sternal Plates: — One of the interesting features of the present specimen is the presence of ossified sternal plates (See figure 8). In the characterization of the genus Laosaurus Marsh definitely states “sternum unossified,” which statement this specimen now demonstrates to be an error. Furthermore, now that osseous sternal plates have been found in Laosaurus, Monoclonius, Triceratops, Thespesius, Stegosaurus, and Thescelosaurus, it may be expected that these elements will eventually be found in all predentate dinosaurs. In specimen No. 11,340, C. M. Cat. Vert. Foss., it is especially fortunate that these bones have been so preserved that they not only show their proper mutual relationships, but also their actual positions in relation to the rest of the pectoral girdle, as clearly indicated in figure 8. The sternum of Laosaurus, except for its very much smaller size, exhibits a striking similarity in form to that of Monoclonius. It consists of two long flat plates each having a thick external and a thin internal border. The anterior end is thickened, especially on the external side, for articulation with the coracoid; the wider distal end seems to be thin. Indentures for the attachment of the carti- laginous sternal ribs have not been observed. These plates have a greatest length of 24 mm. At the anterior end these articulated bones have a greatest transverse diameter of 22 mm.; breadth of the posterior ends 28 mm. The evidence furnished by this specimen as to the proper articulation of the sternal plates, shows quite conclusively that these bones are reversed in the mounted skeleton of Diplodocus in the Carnegie Museum. The articulated pectoral arch of this specimen has been illustrated by Holland, who at that time expressed Brown, B., Bull. Amcr. Mus., vol. 37, 1917, pp. 291-292, figs. 3 and 4. Holland, W. J., Memoirs of the Carnegie Miiseiun, Vol. II, no. 6, 1906, p. 257, fig. 25. GILMORE: ORNITHOPODOUS DINOSAURS FROM UTAH. 409 doubt as to the correctness of their articulation, and it was after much study that the decision was reached that the narrowed but thickened extremities represented the border best suited for the attachment of the cartilaginous ribs. In the absence of definite evidence at that time, this decision represented a most natural con- clusion. In the paper cited above Holland calls attention to the fact that Marsh in his ‘‘Dinosaurs of North America,” plate XXII, figure 1, represents the sternal plates of Brontosaurus one way, and in the same publication on page 179, figure 30, those of Morosaurus in an opposite direction. From the evidence now at hand it is very evident that the position of the plates in Brontosaurus is correct, and that those of Morosaurus have been reversed and therefore erroneously placed by Marsh. 410 MEMOIRS OF THE CARNEGIE MUSEUM. EXPLANATION OF PLATE XVIIL Ui’PEH Figuke: Skeleton of Camptosaurus viedius Marsh (No. 11,337, C. M. Cat. Vert. Foss.), viewed from the left side, showing tlie position of the bones of tlie skeleton as now prepared for exliibition. One block containing caudal vertebiue and another carrying left fore limb and foot did not lend themselves to photography and are not shown in this figure. One-eleventh natural size. Reproduced from a photograph by A. S. Coggeshall. Lower Figure: Outline of the skeleton of Camptosaurus medius Marsh. Designed as a key to the upper figure. One-eleventh natural size, ax, axis; cl, and cl3, caudals 1 and 13 respectively; dl, first dorsal; dl6, sixteenth dorsal; dl7, last presacral, or lumbar vertebra; fe, femur; ft, fibula; h, humerus; il, ilium; is, ischium; p, jiubis; sc, scapula; ti, tibia. MEMOIRS CARNEGIE MUSEUM, Vol. X. Plate XVIII. INDEX abbreviatus, Gonorhynchus, 153 abdominalis, Abudefdiif, 68 Abe, 'Mr. Ryohei, 94, 99 Ablennes hians, 18 Aboma lactipes, 307 tsushimse, 307 Abudefdiif abdominalis, 68 imparipinnis, 68 saxatilis, 249 sindonis, 68 sordidus, 68 aburaco, Hexagrammos, 276 abyssalis, Lepidotrigla, 289 Acahara, 177 Acahara hakonensis, 178, 182 jusanensis, 180 phalacrocorax, 180 Acanthias, 5 Acanthias blainvillii, 106 megalops, 106 mitsukurii, 105 squalus, 104 sucklii, 103 vulgaris, 105 Acanthidium, 107 Acanthidium pusillum, 107 Acantliocepola krusensterni, 248 Acantliochsetodon septentrionalis, 253 Acanthocybium, 213 Acanthocybium petus, 34 Sara, 34, 213 solandri, 34 acanthogenys, Zacco, 184 Acanthogobius flavimanus, 308 Acanthuridae, 65, 253 Acanthurus achilles, 65 argenteus? 65 atramentatus, 65 blochii, 65 dussumieri, 65 elongatus, 65 glintheri, 65 . , 411 gut.tatus, 66 liypselopterus, 66 leucopareius, 65 matoides, 65, 253 olivaceus, 65 sandvicensis, 66 strigosus, 66 triostegus, 66 umbra, 65 xanthopterus, 65 Acheilognathus cyaiiostigma, 162 intermedia, 162 lanceolata, 162 limbata, 162 rhombea, 161 tabira, 162, 163 achilles, Acanthurus, 65 Achiridas, 301 acipenserinus, Matseocephalus, 22 acrages, Pseudotriakis, 102 Acropoma japonicum, 231 Acropomidse, 231 aculeatus, Balistapus, 85 Gasterosteus, 200 aculeatus, 201 Kentrocapros, 256 acuminatus, Clnetodon, 61 acus, Dalatias, 108 acuta, Ophisoma, 195 acutipinnis, Squalus, 107 acutirostris, Eurymyctera, 17 Acutomentum iracundus, 268 matsulDarse, 260 scythropus, 269 acutus, Fodiator, 19 adamsii, Minous, 275 adenomus, Diaphus, 12, 156 adonis, Oncorhynchus, 123, 127, 136, 332, 338 sequorea, Congermursena, 196 sequoreus, Leptoce]ihalus, 13 aerosa, Scaridea, 75 I 412 MEMOIRS OF THE CARNEGIE MUSEUM. Aetobatidse, 116 Aetobatus narinari, 5 tobijei, 116 affine, Alyctophiim, 11 affinis, Caranx? 38 ChilomycteiTis, 88 Eiithynnus, 220 Gambusia, 92 S^^naphobrancliiis, 191 agilis, CamarasaiiiTis, 352 Agonidse, 290 agoo, Cypselui'us, 204 Agostini, A., 348 agrammus, Agrammns, 276 Agrammus agrammus, 276 Agriosphyrsena, 30 ahula, Scarus, 75 Ainocottus ensiger, 278 fasciatus, 278 Ainosiis geneionemiis, 309 ajax, Carangoides, 41 akajei, Dasyatis, 114 Akitani, Mr. K., 95 Alieops plinthus, 298 alalaua, Priacanthiis, 47 alaliinga, Germo, 33, 217 alata, Lepidotrigla, 288 alba, Fliita, 190 “Albatross,” S. S., 1 alliescens, Echeneis, 294 Remorina, 77, 294 albiflora, Nibea, 243 albisella, Dascyllus, 67 alliofasciatus, Seliasticus, 271 alboiilumbeiis, Sphoeroides, 258 alboiiimctatus, Gobius, 78 albotseniatum, Sicydium, 79 albotamiatus, Lepidaplois, 68 albovittatiis, Stethojiilis, 69 Albiila, 6, 7 virgata, 6 Allnilidse, 6 Alcichtliys alcicornis, 280 alcicornis, Alcichtliys, 280 alcocki, Neoscopelus, 10 Promyllantor, 13 Aklrovandia, 9 kauaiensis, 9 proboscidea, 9 vcrticalis, 9 Alectis ciliaris, 41, 224 indicus, 41 alleteratus, Eiithynnus, 31, 32, 220 Alloconger, 192 Alloconger flavirostris, 195 Allolepis hollandi, 323, 346 aloha, Rooseveltia, 48 Alopias vulpes, 4 vuliiinus, 101 Alopiidse, 101 Alticus, 83 altirostris, Scorpsenopsis, 55 altivelis, Plecoglossus, 147, 338 altus, Dryosaurus, 394-402 Alutera liturosa, 86, 255 monoceros, 86, 255 scripta, 86 Ameiuridie, 91 Ameiurus nebulosus, 91 American Museum of Natural History, 1, 96 Amia, 42 Amioides, 44 Ammodytidse, 324 Amphicoclias, 368 Amphioxides iielagicus, 3 Amphiprionichthys unipinna, 54 ampins, Camptosaurus, 392, 393 Anampses cuvieri, 71 evermanni, 71 godeffroyi, 71 Anago, 191 Anago anago, 193 anastomella, Tylosurus, 206 Anchoviella, 8 aneitense, Thalassoma, 72 Anguilla japonica, 190, 194 anguillaris, Plotosus, 158 anguillicaudatus, Misgurnus, 160 Anguillidse, 190 anguineus, Chlamydoselachus, 99 angulosus, Canthidermis, 85 angustirostris, Limandella, 299 anomala, Psenopsis, 226 Triacanthodes, 253 anonyma, ?Solea, 302 Antennariidse, 90, 330 INDEX. 413 Antennaiius bigibbus, 90 commersoni, 90 drombus, 90 duescus, 90 laysanius, 90 leprosus, 90 nexilis, 90 nox, 330 sandvicensis, 90 sangiiifluuS; 330 scriptissimus, 330 tridens, 330 antennatus, Bathypterois, 10 anteorbitalis, Diaphus, 156 Lamprossa, 156 Anticitharus debilis, 25 Antigonia eos, 58 steindachneri, 58 Antimora microlepis, 22 antr^us, Hymenocephalus, 21 Antrodemus, 355, 360 Antrodemus fragilis, 355 Aoki, Mr. Kiimakichi, 95 Apatosaurus, 366, 368, 370, 372, 374 Aphareidae, 51 Aphareus flavivultus, 51 furcatus, 51 Aylocheilus, 198, 199 Aplodactylidse, 247 Apocryptes chinensis, 304 Apodes, 13 Apogon, 42, 43 Apogon kiensis, 231 lineatus, 230 maculiferus, 42 niger, 230 semilineatus, 230 Apogoniclitliys, 44 Apogonichthys carinatus, 230 waikiki, 42 Apogonidae, 42, 230 appendix, Entosphenus, 98 Aprion, 50 Aprioii microdon, 49 sieboldi, 49 virescens, 49 Apristurus platyrhynchus, 99 spongiceps, 3 Apsilus, 50 Apsilus microdon, 48 Aquatic Resources of the Hawaiian Islands, 1 aquilolo, Macropharyngodon, 70 Araias ariommus, 298 aratrum, Coelorhynchus, 21 Archamia, 43 Arctoscopus japonicus, 311 arcuatus, Chsetodontoplus, 62 Areliscus joyneri, 302 interruptus, 302 purpureomaculatus, 303 arenicola, Scseops, 25 arge, Upeneoides, 52 argentatus, Latilus, 248 argenteostriatus, Coris, 71 argenteus?, Acanthurus, 65 Pampus, 226 Argentina semifasciata, 152 Argentinidse, 152 argentiventris, Labracoglossa, 239 argentivittatus, Germo, 33 Argo steindachneri, 225 argus, ,Cephalopholis, 46 Ophicephalus, 207 argyrea, Synagrops, 43 Argyripnus, 8 ephippiatus, 8 argyromus, Myripristis, 27 Argyropelecus, 9 heathi, 9 argyrophanes, Saurida, 155 ariakensis, Parasalanx, 153 Ariid®, 157 Ariomma evermanni, 35 lurida, 35 ariommus, Araias, 298 Arisoma, 195 Arius maculatus, 157 armata, Hoplobrotula, 325 armiger, Rhechias, 14 Arnoglossus japonicus, 295 tenuis, 295 violacens, 295 Arothron ophryas, 87 arsius, Pseudorhombus, 296 Aseraggodes kobensis, 301 asotus, Parasilurus, 159 asper, Rogadius, 286, 288 asperella, Sebastapistes, 54 asperrimum, Clidoderma, 300 414 MEMOIRS OF THE CARNEGIE MUSEUM. Aspidontus trossulus, 318 Asterropteryx semipunctatiis, 77 Asterospondyli, 3 astrinius, Calliurichthys, 80 Astroconger, 192 Astroconger myriaster, 195 Astronesthidse, 8, 153 Astronesthes iijimai, 153 lucifer, 8 martensi, 153 Ateleopidse, 24 Ateleopus plicatellus, 24 ater, Enclielyurus, 83 aterrimus, Hymenocephalus, 21 Atheresthes evermanni, 298 Atherina bleekeri, 207 tsiirugse, 207 Atlierinidas, 29, 207 atherinoides, Hynnodus, 44 atherodon, Optonurus, 20 Atkinson, William Sackston, 97 Atlantosaiirus montanus, 355 atramentatiis, Acanthurus, 65 atraria, Cyclothone, 8 atriceps, Ctenogobius, 309 atrilatus, Pseudaspius, 177, 182 atripes, Enneapterygiiis, 82 atrisignis, Gypselurus, 20 atromarginatus, Dalatias, 108 Atule, 38 Atule polita, 38 Anlichthys japonicus, 199 Aulopidse, 153 Aiilopus japonicus, 153 Aiilorhynchidffi, 199 Aulostomi, 27 Aulostomidse, 27 Aiilostomus chinensis, 27 valentini, 28 aurantiacus, Pseudobagrus, 159 auratus, Carassius, 91, 190 aureolus, Canthidermis, 85 aureovittata, Seriola, 222 aureus, Cirrhitichthys, 247 auriflamma, Mulloides, 51 aurora, Pikea, 46 Auxis bisus, 220 hira, 221 rnaru, 220 rochei, 220 tapeinosoma, 31, 220, 221 thazard, 31 Awai, Mr. K., 94 Awaous, 78 awoara, Epinephelus, 236 axillaris, Hinalea, 69 lulis, 69 aygula, Coris, 70 azonus, Pleurogrammus, 276 azureus, Sectator, 51 azurio, Choerodon, 249 Bagridse, 159 baliiensis, Gypselurus, 20 baldwini, Hemipteronotus, 74 balia, Scaridea, 75 Balistapus aculeatus, 85 rectangulus, 85 Balistidse, 84, 254 ballieui, Goris, 71 Sebastapistes, 54 Thalassoma, 72 balteata, Hinalea, 69 Banjos banjos, 238 Banjosidae, 238 barbarus, Ganinoa, 100 Barbatula, 160 Barbatula oreas, 161 barberi, Brachirus, 56 barbouri, Bathyalopex, 118 barbus, Hemibarbus, 172 Barbus homogenes, 165 homozonus, 169 Barilius platypus, 184 Barosaurus, 368 bataviensis, Scar us, 76 Bathyalopex, 116 Bathyalopex barbouri, 118 bathybium, Euthyopteroma, 240 Bathygobius fuscus, 78 Bathypteroidse, 10 Bathypterois antennatus, 10 Batoidei, 5 beani, Serrivomer, 14 bedfordi, Gtenogobius, 305 Belligobio eristigma, 172, 173, 340 bellus, Brachirus, 274, 342 Belonidae, 18, 206 Bembradidse, 56, 283 Bembradium roseum, 56 Index:. 41 Bembras japonicus, 283 Bembrops filifera, 79 beniteguri, Callionymus, 317 bennetti, Scams, 76 bensasi, Upeneoides, 245 Beppu, Mr. Yoshimo, 94 bergi, Pseudaspius, 180 berndti, Gymnothorax, 16 Myripristis, 27 Polymixia, 26 bernissartensis, Iguanodon, 396 Bero zanclus, 280 Berycidse, 209 Berycoidei, 26 berycoides, Doderleinia, 236 Beryx splendens, 209 Bibliography of Hawaiian Fish Fauna, 2 bicinctus, Uranoscopus, 316 bicolor, Chsetodontoplus, 61 bicoloratus, Kareius, 300 bidens, Opsariichthys, 189 bifasciatus, Tridentiger, 309 Upeneus, 52 bifer, Julis, 73 bigibbus, Antennarius, 90 bilineata, Lepidopsetta, 299 bimacula, Cirrhitoidea, 53 bimaculatus, Cheilinus, 73 binoculata, Raja, 111 bipinnatulus, Elagatis, 37 Bishop Museum, Bernice Pauahi, 2 bispinosus, Holacanthus, 62 biwse, Cobitis, 160 Gnathopogon, 171 Gobio, 169 Leucogobio, 172 bixanthopterus, Caranx, 37, 39, 224 blackistoni, Salmo, 145 blainvillii, Acanthias, 106 bleekeri, Atherina, 207 Ebosia, 274 Enchelynassa, 15 Halichoeres, 251 Blennies, 82 Blennidse, 82 Blennius sordidus, 82 yatabei, 318 blochii, Acanthurus, 65 Boaxodon cyanescens, 47 Bola, 242 Boleophthalmus pectinirostris, 304 boops, Scombrops, 231 borborus. Scams, 75 Bothidse, 294 Bothrocara zesta, 321 Boulengerina, 41 bowersi, Gadomus, 20 Leptocephalus, 13 Bowersia ulaula, 49 Brachirus barberi, 56 bellus, 274, 342 chloreus, 56 jordani, 273 brachycephalus, Exonautes, 204 brachygramma, Foa, 42 Brachyopsis rostrata, 293 brachyptera, Remoropsis, 294 brachypterus, Parexocoetus, 19 Brachysomophis henshawi, 15 brachysomus, Synaphobranchus, 13 Bramidse, 35, 225 branchiostega, ?Coryphaena, 248 Branchiostegidse, 248 Brancliiostegus japonicus, 248 Branchiostoma, 3 Branchiostomidse, 3 brandti, Richardsonius, 177 braueri, Myctophum, 11, 12 Bregmaceros japonicus, 326 Bregmacerotidse, 326 Brephamia, 43 brephocephalus, Cottunculus, 279 Brevigobio, 161, 189 Brevigobio kawabatse, 161, 189 brevirostris, Naso, 66 Squalus, 106 Triacanthus, 254 brevis, Exallias, 83 brevispinosus, Pungitius, 203 brighami, Rooseveltia, 48 Brontosaurus, 409 Brook-lampreys, 98 Brotula marginalis, 83 multibarbata, 324 multicirrata, 83 Brotulidse, 83, 324 browni, Camptosaurus, 388, 392, 393 brunneus, Ectenias, 225 Glossogobius, 307 Scams, 75 410 MEMOIRS OF THE CARNEGIE MUSEUM. Bryan, William Alanson, 2 Bryttosus kawamcliari, 235 burragei, Macrourus, 21 bursa, Sufflamen, 85 Butterfly-fishes, 58 Butyrinus, 6 cacopsis, Scorpienopsis, 55 Chulorhynchus gilberti, 327 caerulaureus, Caesio, 239 ca'i'iileonotatus, Cfallionymus, 80 cieruleopimiatus, Caranx, 40 ca:'rulescens, Clnathopogon, 167 Leuciscus, 167 Squalius, 167 Caesio caerulaureus, 239 chrysozonus, 239 lunaris, 239 Caesionidae, 239 California State Fish and Game Com- mission, 96 californiensis, Cliilomycterus, 88 Callechelys luteus, 15 Callicanthus garretti, 07 lituratus, 67 metoposophron, 67 Callionymidae, 80, 316 Callionymus beniteguri, 317 caeruleonotatus, 80 corallinus, 80 flagris, 317, 318 lunatus, 316 punctatus, 317 ricliardsoni, 317 rubrovinctus, 80 valenciennei, 317 valenciennesi, 317 Calliurichtliys astrinius, 80 decoratus, 80 doryssus, 316 japonicus, 310 zanectes, 81 Callyodon, 75 calvarius, Silurus, 159 Camarasaurus agilis, 352, 353, 355 grandis, 352, 353, 371, 377, 378, 380, 382 imjiar, 352 lentils, 352, 354-384 leptodirus, 352 robustus, 352 supremus, 352, 353, 360, 367, 369 370, 371, 376, 378, 380, 382 Camarasaurus: carpus of, 378; caudal vertebrae, 371; cervical ribs, 374; cervical vertebrae, 368; chevrons, 373; dorsal vertebrae, 369; ex- ternal openings of skull, 364; fore foot, 378; hind limb and foot, 380; hyoid arch, 367; lower jaw, 363; pelvic girdle, 380; osteological features of, 351; ribs of, 374-375; sacrum, 370; sclerotic ring of eye, 366; skull, 358-363 ; sternal plates, 377; teeth, 365; vertebral column of, 367. Camptosauridae, 376 Camptosaurus, 363 Camptosaurus, ampins, 392, 393 browni, 388, 392, 393 depressus, 392 dispar, 388, 392, 393 medius, 385-393, 410 nanus, 388, 392, 393 candidius, Ctenogobius, 305 Rhinogobius, 305 canescens, Zanclus, 64 canina, Cyclothone, 8 Enchelynassa, 15 Snyderidia, 84 Caninoa barbarus, 100 canis, Mustelus, 100 Cantherines albopunctatus, 86 carolae, 86 howensis, 255 modestus, 254 nigromaculosus, 86 sandwichiensis, 86 tessellatus, 254 Canthidermis angulosus, 85 * aureolus, 85 rotundatus, 254 Canthigaster bitseniatus, 87 cinctus, 87 eiiilamprus, 87 jactator, 87 janthinus, 87 oahuensis, 87 psegma, 87 rivulatus, 87 INDEX. 417 Canthigasteridse, 87 capistratus, Sufflamen, 85 Capoeta elongata, 165 gracilis, 168 Caprodon schlegeli, 236 Caproidse, 58 Capropygia spilonota, 89 Caracanthidse, 54 Caracanthus maciilatus, 54 Carangidse, 1, 37, 222 Carangoides ajax, 41 eqimla, 224 evermanni, 41 gymnostethoides, 40 jordani, 40 Carangus hippoides, 39 rhabdotus, 39 Caranx affinis, 38 bixanthopterus, 37, 39, boops, 38 cseruleopinnatus, 40 dasson, 40 elacate, 40 forsteri, 39, 224 heberi, 39 hippoides, 39 ignobilis, 39 latus, 39 luna, 40 marginatus, 39 melampygiis, 39, 224 punctatiis, 40 rhabdotus, 39 stellatus, 40 thompsoni, 40 Carapidse, 323 Carapus, 84 Carapus sagamianus, 323 (?)sagamiiis, 323 Carassius auratus, 91, 190 Carcharhinidse, 3 Carcharias, 3 carcharias, Carcharodon, 5 Carcharinus insularum, 4 nesiotes, 4 melanopterus, 4 phorcys, 4 Carcharodon carcharias, 5, Cardinal-fishes, 42 cardinalis, Evynnis, 241 Kelloggella, 78 carinatus, Apogonichthys, 230 carlsmithi, Rhyacanthias, 46 carnegiei, Diplodociis, 355 Carnegie Museum, 1 carpio, Cyprinus, 91, 190 Carpus and fore foot of Camarasaurus, 378 castanea, Chloea, 307 castaneus, Gobius, 307 Catalog of the Fishes of Japan, 96 Catalufas, 47 Catalupha japonica, 227 umbra, 227 cataphractus, Gasterosteus, 200 Catfish, 91 catocala, Scorpsenopsis, 55 224 Catulus, preoccupied, 3 Caudal vertebrae of Camarasaurus, 371 Caulolepis longidens, 26 cayuga, Notropis, 169 celer, Laosaurus, 403 Centrarchidse, 92 Centrobranchus, 12 choerocephalus, 12 gracilicaudus, 12 Centrolophidse, 226 Centrolophus japonicus, 226 Centrophorus, 108 centropomus, Vellitor, 281 Centropyge diacantha, 62 potteri, 62 tutuilse, 62, 63 Centroscy Ilium nigrum, 5 ruscosum, 5 Centroscy mnus owstoni, 107 Cephalacanthidse, 57 Cephalopholis argus, 46 Cephalopterus, 6 cephalus, Mugil, 29, 207 Cepola schlegeli, 249 Cepolidse, 248 cerasina, Pseudojulis, 69 Ceratiidse, 90 Ceratosaurus nasicornis, 355 Cervical ribs of Camarasaurus, 374 Cervical vertebi-ae of Camarasaiu-us, 102 368 Cetorhinidse, 102 418 MEMOIRS OF THE CARNEGIE MUSEUM. Cetorhiiiiis maximiis, 102 Cha?nogobius macrognathos, 307 Chienomiigil chaptalii, 29 Choerojulis, 250 Chsetodon acuminatus, 61 corallicola, 60 epliiiipium, 59 fremblii, 59 lineolatus, 58 lunula, 59, 252 mautelliger, 59 miliaris, 59 oi'uatissimus, 59 puiictatofasciatus, 59 quadrimaculatus, 59 setifer, 58, 252 trifasciatus, 59 uniinaculatus, 59 ChaetodoutidiE, 58, 252 Cluetodontoplus arcuatiis, 62 bicolor, 61 Chsetodoiitops, 59 C4ia'to]iterus dubius, 49 Cha'tuiichthys hexanema, 308 clialcogramma, Theragra, 326 Chalinura ctenomelas, 20 challeugeri, Polyacaiithonotus, 199 Chamberlain, Henry, 93 Chamiisodon fimbi-iatus, 80 Chanida', 7 Chanos chanos, 7 cyprinella, 7 chanos, Chanos, 7 cha])talii, Chsenomugil, 29 Chascanopsetta iirorigera, 25 Chasmichthys gulosus, 308 ChaunacidiB, 9(), 330 Chaunax fimbriatiis, 330 umbrinus, 90 Cheilinoides jordani, 73 Cheiliniis bimaculatus, 73 hexagonatus, 73 siniiosus, 73 trilobatus, 73 zonurus, 73 Cheilio inermis, 71 cheilo, Uraspis, 40 Cheilodactylidre, 53 Cheilodastylus, 53 C'heilodipteiTis, 44 Chelidonichthys kumu, 288 Chelidoperca hirundinacea, 236 Chelon, 29 Chevrons of Camarasaurus, 373 chilensis, Sarda, 32 Chilomycterus affinis, 88 californiensis, 88 Chimsera, 116 Chimsera mitsukurii, 119 phantasma, 116, 119 purpurescens, 6, 117 Chima^ridse, 6, 116 Chimseroidei, 6 chinensis, Aiiocryptes, 304 Aulostomus, 27 Chlamydes laticeps, 78 Chlamydoselachidse, 99 Chlamydoselachus anguineus, 99 Chloea castanea, 307 nakamiirae, 307 senboc, 307 chloreus, Brachirus, 56 Chlorophthalmidse, 10 Chlorophthalmus proridens, 10 chlorospilus, Platophrys, 24 choerocephalus, Centrobranchus, 12 Choerodon azurio, 249 Chondrichthyes, 24 Chonophorus genivittatus, 78 stamineus, 79 Chromides, 67 Chromis elaphrus, 67 notatus, 249 verater, 67 chryseres, Chrionema, 79 Myripristis, 26 chryserydros, Upeneus, 52 chrysonemus, Upeneus, 52 Chrysophrys swinhonis, 240 chrysorhynchus, Diaphus, 12 chrysozonus, Csesio, 239 chua-tsi, Siniperca, 235 ciliaris, Alectis, 41, 224 cinctus, Canthigaster, 87 Plectorhynchus, 258 cinerascens, Kyphosus, 242 cinereus. Conger, 193 Murienesox, 198 cinnamomeus. Pseudorhombus, 296 cirrhifer. Monacanthus, 254 INDEX. 419 Cirrhitichthys aureus, 247 Cirrhitidae, 53, 247 Cirrhitiformes, 53 Cirrhitoidea bimacula, 53 Cirrhitus marmoratus, 53 cirrhosa, Scorpsenopsis, 272 CiiTostomi, 3 Citharoedus, 59 citrinellus, Tsenianotus, 55 citrinus, Hoplichthys, 57 clarescens, Vitraria, 78 Clarias fuscus, 92 Clariidse, 92 Cleisthenes herzensteini, 298 piiietorum, 298 Clidoderma asperrimum, 300 ClinidiB, 82 Clupanodon punctatus, 120 Clupea pallasii, 121 Clupeidse, 7, 121 clypeata, Echeneis, 294 coarctatus, Platophrj^s, 24 Cobitidse, 160 Cobitis biwse, 160 Cocius crocodilus, 286, 287 coelestis, Pomacentrus, 249 Coelocephalus, 22 Coelorhynchus aratrum, 21 doryssus, 21 g’ladius, 21 japonicus, 327 Coggeshall, Arthur S., 348 Coggeshall, Louis S., 348 Coilia ectenes, 122 coitor, Scisena, 242 colias, Pneumatophorus, 211 Scomber, 210 Collection of Fishes from Fiji with notes on Certain Hawaiian Fishes, 2 Collybus drachme, 35 Cololabis saira, 206 coloratus, Sebastapistes, 54 commersoni, Antennarius, 90 Cybium, 214 concatenatus, Lactophrys, 256 Conchoderma, 19 concolor, Lentipes, 79 Conger, 13, 191 Conger cinereus, 193 conger, 194 habenata, 195 japonicus, 194 vulgaris, 194 conger. Conger, 194 Congermuraena aecpiorea, 196 nasica, 196 Congrellus, 13 Congridse, 13 Congrina, 191 Congrina retrotincta, 197 Congriscus, 191 Congriscus megastomus, 193 Congrogadidse, 83 Congrogadus marginatus, 83 coniorta, Sebastapistes, 54 consors, Laosaurus, 396, 403 copei, Hemipteronotus, 74 Coradion modestum, 252 corallicola, Cha?todon, 60 Tifia, 60 Sebastapistes, 54 corallinus, Callionymus, 80 Samariscus, 25 Coreoperca herzi, 235 Coris argenteostriatus, 71 aygula, 70 ballieui, 71 eydouxi, 71 flavovittatus, 70 gaimardi, 70 greenovi, 70 lepomis, 70 pulcherrima, 70 rosea, 71 venusta, 71 Cornet-fishes, 28 cornuta, Lactoria, 256 Pleuronichthys, 298 ' coronatus. Hippocampus, 199 cornutus, Zanclus, 64 Coryphaena equisetis, 35 hippurus, 35, 225 ?Coryph8ena branchiostega, 248 Coryphsenidse, 35, 225 Coryphaenoides, 21 Coryphaenoides nasutus, 327 Coryphaenoididae, 20, 327 Cottidae, 277 420 MEMOIRS OF THE CARNEGIE MUSEUM. Cottiusculus gonez, 279 Cottuiiculiis brephocephalus, 279 Cottus kazika, 277 pollux, 278 craspediirus, Epinephelus, 235 crassilabris, Upeneus, 52 C'reaser and Hublis, 98 crocodiliis, Cocius, 286, 287 cnientatus, Priacaiithus, 47 ci'umenalis, Pelecanichthys, 25 crumenophthalma, Trachurops, 38 Cryptocentms filifer, 307 Crystallias matsiishimse, 293 Ctenochsetus striatus, 66 Ctenogobius atriceps, 309 bedfordi, 305 candidiiis, 305 liadroptems, 306 similis, 305 virgatulus, 306 ctenomelas, Chaliimra, 20 Cubiceps, 35 ciipido, Thalassoma, 251 curta, Hymenophysa, 160 ciirtus, Parabembras, 281 cuvieii, Anampses, 71 cyanescens, Boaxodon, 47 cyanostigma, Acheilognathus, 162 Cybiidse, 212 Cybium, 213 Pybium commersoni, 214 Cyclothone, 8 atraria, 8 canina, 8 rhodadenia, 8 cyclurus, Leptoscarus, 75 Cymolutes lecliisei, 74 Cynias manazo, 100 Cynoglossidse, 25, 302 Cynoglossus iniisita, 302 robustus, 302 cyplio, Salarias, 83 Cyprinidse, 91, 161 Cyprinodontidae, 92, 198 Cyprinus carpio, 91, 190 leuciscus, 177 Cypselurus agoo, 204 atrisignis, 20 liahiensis, 20 hirundo, 204 simus, 20 spilonotopterus, 20 Cyttomimus stelgis, 23 dabryi, Micropercops, 304 Dactyloptena orientalis, 57 Dactylopteridse, 290 Daicocus peterseni, 290 daimio, Pterogobius, 308 Dalatias acus, 108 atromarginatus, 108 dalwigkii, Physiculus, 327 Damsel-fishes, 67 Dascyllus albisella, 67 trimaculatus, 67 dasson, Caranx, 40 Dasson japonicus, 318 loxozonus, 318 trossulus, 318 Dasyatidse, 5, 114 Dasyatis akajei, 114 hawaiiensis, 5 lata, 5 sciera, 5 sp.?, 115 iishiei, 114 Dasycottus japonicus, 277 setiger, 277 Dasyscopelus, 10 pristilepis, 10 spinosus, 11 dea, Iniistius, 251 Deania, 107 Deania eglantina, 107 histricosa, 108 nocturnus, 108 rostrata, 108 debilis, Anticitharus, 25 Decaptems limdini, 38 mamadsi, 38, 223 muroadsi, 223 pinnatulus, 37 decoratus, Calliiirichthys, 80 delicatissimum, Longirostrum, 224 Dendrochims jordani, 273 dentex, Osmems, 149 Denticidse, 240 denticulatus, Metopomycter, 14 depauperatus, Hemirhamphus, 18 depressus, Camptosaums, 392 INDEX. 421 Description of Deep-sea Fishes from the Coast of Hawaii Killed by a Lava-flow from Manna Loa, 2 Devil-rays, 6 Dexistes kitaharoe, 300 rikiizeniiis, 298 diacantha, Centropyge, 62 diadema, Holocentriis, 27 diaphana, Lactoria, 256 Sternoptix, 9 Diaphus adenomus, 12, 156 anteorbitalis, 156 chrysorhynchiis, 12 gland ulifer, 157 latiis, 156 iirolampus, 12 Dibranchiis erythrinus, 91 stelliilatus, 91 diego, Pneiimatophoriis, 31, 211 Dinogimelliis grigorjewi, 320 Dinosaur National Monument, 347 Dinosaurs, Ornithopodous, 385-410 Sauropodous, 347-384 Diodon holacanthus, 88 hystrix, 88 nudifrons, 88 Diodontidae, 88 diphreutes, Heniochus, 253 Diplodocus, 354, 355, 358, 364, 368, 370, 374 carnegiei, 355 Diplophos, 9 pacificus, 9 Dipterygonotus leucogrammicus, 47 Discocephali, 77 dispar, Camptosaurus, 388, 392, 393 Ditrema temmincki, 249 dodecaedron, Occella, 291 Doderleinia, 50 berycoides, 236 doderleini, Epinephelus, 235 Dolphins, 35 dorobse, Leuciscus, 182 Dorosomidse, 120 Dorsal vertebrae of Camarasaurus, 369 doryssus, Calliurichthys, 316 Coelorhynchus, 21 drachme, Collybus, 35 Draciscus sachi, 293 Draconnetta hawaiiensis, 80 Draconettidae, 80 drombus, Antennarius, 90 Dryosaurus altus, 394-402 dubius, Chaetopterus, 49 Hippoglossoides, 298 Scarus, 75 ductor, Naucrates, 37 duescus, Antennarius, 90 Dules, 41 malo, 41 mato, 41 duperrey, Thalassoma, 72 dussumieri, Acanthurus, 65 Dussumieriidae, 7, 120 Duymaeria flagellifera, 250 dybowskii, Hypoptychus, 324 Ebosia bleekeri, 274 starksi, 273 Echeneidae, 77, 294 Echeneis albescens, 294 clypeata, 294 Echidna leihala, 17 nebulosa, 17 obscura, 17 psalion, 17 tritor, 17 vincta, 17 zebra, 17 zonata, 17 zonophaea, 17 echigonia, Lefua, 160 echigonius, Minous, 275 ectenes. Coilia, 122 Macrourus, 21 Ectenias brunneus, 225 ectenurus, Leptocephalus, 196 Rhyncoconger, 196 edentulus, Salarias, 83 Eels, 13 eglantina, Deania, 107 egregius, Vesposus, 24 eidolon, Psychichthys, 332 elacate, Caranx, 40 elaphrus, Chromis, 67 Elasmobranchii, 3 elegans, Dasson, 318 Kyphosus, 51 Sebastodes (Sebastocles), 270 422 MEMOIRS OF THE CARNEGIE MUSEUM. Elegatis bipinnatulus, 37 pinnatiilus, 37 elongatiis, Giiathagiius, 316 Liiciogoliiiis, 309 Eleotridie, 77, 303 Elect ris oxycephala, 304 sandwicensis, 77 elongata, Ilislia, 121 elongatiis, Acanthiirus, 05 Gnathopogon, 165 Elopidap 6, 119 Elo})s hawaiiensis, 6, 119 maclinata, 119 Einliiotocidai, 249 eniblemarius, Helicoleniis, 272 Ernmelas, glaucus, 200 Enimeliehthyidse, 47, 245 Encseura evides, 303, 344 Plnchelynassa bleekeri, 15 canina, 15 Enchelyiirus ater, 83 Enedrias iiebulosiis, 320 l^lngTaulidse, 8, 121 Engraulis japonicus, 121, 248 eiigyceros, Peristedion, 56, 57 Enneanectes, 82 Enneaiiterygius atripes, 82 ensifer, Holocentrus, 27 ensiger, Ainocottiis, 278 entargyreus, Novaculichthys, 73 Entospheiius aiipendix, 98 jaiionicus, 98 mitsikurii, 98 eos, Antigonia, 58 Stetliopristes, 23 eiihipiiiatus, ArgjniiRius, 8 eiihippium, Cha?todon, 59 Ei)igoniis, 45 Eiiinepheliday 235 Epinephelus, 45 E]iiiiephelus awoara, 236 craspediiriis, 235 doderleini, 235 eiiistictiis, 235 moara, 236 morrhua, 235 (liiernus, 46 septemfasciatns, 236 tsiriinenara, 230 epii)hanes, blviota, 77 Eptatretida3, 97 Eptatretus biirgeri, 97 okinoseaniis, 97 eqiiisetis, Corypluena, 35 equula, Carangoides, 224 ercodes, Gymnothorax, 16 Riidarius, 255 erebennus, Leptocephaliis, 194 Erilejiidae, 276 Erilejiis zonifer, 276 eristigma, Belligobio, 172, 173, 340 Erosa erosa, 275 eryngia, Pallasina, 293 erythra'iis, Holocentrus, 27 Ichthyocampus, 28 Erytln-ichthys, 47 erythriniis, Diliranchus, 91 Mulloides, 51 Pristiapogon, 42 Erythrocles scldegeli, 47, 245 scintillans, 47 erythrodon. Scams, 70 eso, Saurida, 155 esocinus, Pseudogobio, 174 Etelides, 50 Etelinus, 50 Etelinus marshi, 49 Etelis, 45, 50 Etelis eviims, 49, 50 Etmoptenis, 107 Etmopterus villosus, 5 Etrumeus micropiis, 7, 120 othonops, 7 Eugaleus, 3 eiigenius, Qiiisquilius, 77 Euleptorhamphus longirostris, 18 Eumegistus illiistris, 35 eurostus, Gymnothorax, 16 Eurymyctera acutirostris, 17 Euthynnus alleteratus, 31, 220 affinis, 220 lineatus, 220 pelamis, 31, 220 yaito, 220 Euthyopteroma bathybiiim, 240 virgatum, 240 evanidus, Pseiidocheilinus, 73 Eventognathi, 91 Evermann, Barton Warren, 1 INDEX. 423 evermanni, Anampses, 71 Atherestes, 298 Ariomma, 35 Carangoides, 41 Lepidamia, 42 Myctoplium, 11 evides, Encseura, 303, 344 Eviota epiplianes, 77 evolans, Zacco, 184, 185, 188 Evolaiitia microptera, 19 evurus, Etelis, 50 Evynnis cardinalis, 241 Exallias brevis, 83 excelsa, Loa, (Roa) 61, 252 ExoccEtidse, 19, 203 Exocoetiis rubescens, 19 splendens, 203 unicolor, 19 volitans, 19, 203 Exonautes bracliyceplialiis, 204 gilberti, 19 External openings of skull of Camara- saurus, 364 eydouxi, Coris, 70 farcimen, Gobiopterus, 77 fasciata, Oplegnatlius, 245 fasciatus, Ainocottus, 278 Trachidermus, 277 fermandezianus, Squalus, 107 fibulatum, Myctophum, 11 Fierasfer homei, 84 microdon, 84 Fierasferidae, 84 filifer, Cryptocentrus, 307 filifera, Bembrops, 79 fimbriatus, Champsodon, 80 Cliaunax, 330 fislieri. Hippocampus, 28 Xiphypops, 64 Fishes of Samoa with a Check-list of the Fishes of Oceania, 2 Fishing-frogs, 90 Fistularia petimba, 28, 200 serrata, 28, 200 Fistulariidae, 28, 200 flagellifera, Duym^ria, 250 flagris, Callionymus, 317 Flammeo, 27 flammeus, Sebastodes (Sebastosomus), 268 flavescens, Zebrasoma, 66 flavicaudus, Sphagebranchus, 14 flavimanus, Acanthogobius, 308 flavirostris, Alloconger, 195 Leptocephalus, 195 flavivultus, Aphareus, 51 flavomarginatus, Gymnothorax, 16 flavovittatus, Coris, 70 florealis, Spheroides, 86 fluctuans, Oncesthes, 319 Petroscirtes, 319 Fluta alba, 190 Flutidie, 190 Flying-fishes, 19 Flying Gurnards, 57 Foa, 44 Foa brachygramma, 42 Fodiator acutus, 19 rostratus, 19 fontinalis, Salvelinus, 144, 145 Forcipiger longirostris, 58 formosanus, Oncorhynchus, 133, 338 formosus, Scarus, 76 forsteri, Caranx, 39, 224 fowleri, Lumpenus, 320 Microdonophis, 14 Fowleria, 44 fragilis, Antrodemus, 355 Scepterias, 45 fraterculus, Upeneus, 52 fremblii, Chaetodon, 59 Frigate-mackerels, 31 fujiyamse, Rheopresbe, 278 fulvidraco, Pelteobagrus, 159 Fundulichthys, 199 Fundulichthys virescens, 177 Fundulus, 199 Fundulus virescens, 199 - Furcaria leucura, 67 ovalis, 67 furcatus, Aphareus, 51 Furcimanus nakamura?,, 320 Furcina ishikawse, 280 osimse, 280 fusca. Raja, 110 fuscescens, Rosicola, 260 Sebastodes, 261 Teuthis, 253 424 MEMOIRS OF THE CARNEGIE MUSEUM. fuscipilinis, Odontaiithias, 46 fuscolineatus, Sufflameii, 85 fuscus, Bathygobius, 78 Clarias, 92 Kyphosus, 51 Synod us, 154 Thalassoma, 72 Urolophus, 115 Gadidse, 326 Gadoimis bowersi, 20 melanopteiTis, 20 Gadus macrocephalus, 326 gaimardi, Coris, 70 galactacma, Sebastapistes, 54 Galeocerdo tigrinus, 4 galeodon, Lactoria, 89 Galeorhinidae, 3, 100 (Jaleorhiiius, 3 CJaleorhiniis jaiionicais, 3, 101 CJaleus, 3 Gambiisia affiiiis, 92 garretti, Callicaiithiis, 67 Tseniaiiotus, 55 Gasterosteidse, 200 Gasterosteus aculeatus, 200 aculeatus, 201 microcephalus, 202 williamsoni, 201 cataphractus, 200 insciilptiis, 200 loricatus, 200 obolarius, 200 santa-aniRU, 200 trachurus, 200 Gee, Mr. Gist, 95 Gempylidse, 34, 221 Gempylus serpens, 35 thyrsitoides, 35 geneioneinus, Ainosus, 309 genivittatus, Chonoiihorus, 78 geoffroyi, Macropharyngodon, 70 (Jerrno, 216 Germo alalunga, 33, 217 argentivittatus, 33 germo, 217 macropterus, 32, 33 pacificus, 217 sibi, 33, 218 germo, Germo, 32 gibber, Macrourus, 21 gibbifrons, Rupiscartes, 83 gibbosa, Scorpsena, 272 Scorpsenopsis, 55, 272 gibbosum, Ostracion, 256 gibbosus, Tetrosomus, 256 giganteus, Tylosiirus, 18 Gilbert, Charles Henry, 1, 2 gilberti, Coelorhynchus, 327 Exonautes, 19 Peristedion, 57 Scams, 76 ?Telescopias, 231 Gilmore, Charles W. : A Nearly Com- plete Articulated Skeleton of Camarasaurus, a Saurischian Dinosaur from the National Dinosaur Alonument, Utah, 347-384 Osteology of Ornithopodous Dino- saurs from the Dinosaur Na- tional Monument, Utah, 385- 410 Girella punctata, 241 Girellidse, 241 gissu, Pterothrissus, 119 giurinus, Gobius, 305 Rhinogobius, 305 gladius, Coelorhynchus, 21 Istiophorus, 30 Xiphias, 30 glandulifer, Diaphus, 157 Pantophos, 157 glauca, Isuropsis, 4 Prionace, 4, 101 glaucus, Emmelas, 260 I sums, 102 glossa, Plagiopsetta, 301 Glossamia, 43 Glossogobius brunneus, 307 Glyptocephahis ostroumowi, 301 sasre, 301 Gnathagnus elongatus, 316 Gnathanodon speciosus, 41 Gnatholepis knighti, 78 Gnathopogon biwse, 165, 171 cserulescens, 164, 167 chankanensis, 164 coreanus, 164 elongatus, 163, 165 INDEX. 425 Gnatliopog’on gracilis, 164, 168 lierzensteini, 164 iijimse, 165 ishikawse, 169 japonicus, 164, 171 jordani, 167 longifilis, 164, 169 majimse, 164, 167, 340 mayedse, 171 strigatus, 164 siiwse, 163, 166 tseniatus, 164 tsuchigse, 164, 170 Gobies, 77 Gobiidse, 77, 304 Gobio biwae, 169 mayedse, 171 Gobioidei, 77 Gobiopteriis farcimen, 77 Gobius albopimctaus, 78 castaneus, 307 giurinus, 305 sandvicensis, 78 yokohamse, 305 godeffroyi, Anampses, 71 goldsboroughi, Gymnothorax, 16 Gomphosus sandwichensis, 71 tricolor, 71 varius, 71 gonez, Cottiusculus, 279 Goniistius vittatus, 53 zebra, 247 zonatus, 247 Gonorhynchidse, 153 Gonorhynchus abbreviatus, 153 Gonostomidse, 8 gorbuscha, Oncorhynchus, 123, 129, 338 gracilicauda, Gymnothorax, 16 gracilicaudus, Centrobranchus, 12 gracilis, Capoeta, 168 Gnathopogon, 168 Laosaurus, 385, 403, 410 Saurida, 10 Grammicolepidse, 24 Grammatonotus laysanus, 47 grandis, Camarasaurus, 352 grandisquama, Scaeops, 295 grandisquamis, Saurida, 155 grandoculis, Monotaxis, 51 granulosus, Squalus, 108 Great White sharks, 5 greenovi, Coris, 70 grex, Pneumatophorus, 31, 211 grigorjewi, Dinogunellus, 320 Xystrias, 297 Grinnell, Fordyce, Jr., 1 grinnelli, Physiculus, 22 griseus, Malakichthys, 233 Mustelus, 101 guentheri, Lepidotrigla, 289 Leucogobio, 165 guntheri, Sebastodes, 261 Teuthis, 65 gulosus, Chasmichthys, 308 guttatus, Acanthurus, 66 Gymnocanthus lierzensteini, 279 intermedins, 279 Gymnodontes, 86 Gymnosarda, 213 Gymnosarda nuda, 215 unicolor, 31 gymnostethoides, Carangoides, 40 Gymnothorax berndti, 16 ercodes, 16 eurostus, 16 flavomarginatus, 16 goldsboroughi, 16 gracilicauda, 16 hilonis, 16 laysanus, 16 leucacme, 16 leucostictus, 16 meleagris, 16 mucifer, 16 nuttingi, 16 petelli, 16 pictus, 17 reticularis, 198 steindachneri, 16 thalassopterus, 16 undulatus, 16 vinolentus, 15 waialuse, 16 xanthostomus, 17 hadropterus, Ctenogobius, 306 hsematocheila, Liza, 208 hsematopterus, Lethrinus, 240 / 426 MEMOIRS OF THE CARNEGIE MUSEUM. liakoiiensis, Acahara, 178, 182 Leuciscus, 182 liakiiensis, Leuciscus, 178 Halichoeres bleekeri, 251 iriclesceiis, 70 lao, 70 oiTiatissimus, 70 poecilepterus, 250 p^UThogrammus, 250 tenuispinus, 250 llalieutaea retifera, 91 stellata, 330 Halocypselus, 19, 203 Halosauridie, 9 Halosauropsis, 9 “Hammer-head,” 4 Haplacanthasaurus, 367, 372 Hapalogeiiys mucroiiatus, 239 nigripiimis, 239 Haplocheilus, 198 Haplomi, 92 Hareiigula zunasi, 121 hasselti, Selar? 38 hawaiieiisis, Dasyatis, 5 Draconnetta, 80 Elops, 6, 119 Macrorhamiihosus, 28 Alalacocephalus, 22 Moriiigua, 15 Poecilopsetta, 24 Scseops, 25 Hazeus, 78, 304 Ilazeus otakii, 304 Head-fishes, 88 heat hi, Argyropelecus, 9 hebetatus, Macrourus, 21 Helicolenus emblemarius, 272 rufescens, 54 heliotropinus, Pseudoscarus, 76 helleri, Sphyrsena, 29 helvolus, Uraspis, 40 Hemibarbus barbus, 172 Hemicoris keleipionis, 71 remedius, 71 llemigTammocypris rasborella, 161, 189 Hemipteroiiotus baldwini, 74 copei, 74 jenkinsi, 74 umbrilatus, 74 Hemirhamphidse, 18, 205 Hemirhamphus depauperatus, 18 Hemisalanx prognathos, 153 Hemitremia steindachneri, 181 Hemiulis, 250 Heniochus diphreutes, 253 macrolepidotus, 61 henshawi, Brachysomophis, 15 Hepsetia insularum, 29 herzensteini, Cleisthenes, 298 Gymnocanthus, 279 Protopsetta, 298 herzi, Coreoperca, 235 Pungtimgia, 176 heterodon, Notropis, 169 Heterodontidse, 99 Heterodontus japonicus, 99 zebra, 99 heterognathos, Myrophis, 196 Heteromycteris japonica, 301 Heterosomata, 24 hexagonatus, Cheilinus, 73 Hexagrammidse, 276 Hexagrammos aburaco, 276 otakii, 276 hexanema, Chseturichthys, 308 hians, Ablennes, 18 Peristedion, 57 hilonis, Gymnothorax, 16 Hippocampus, 28 Hime japonica, 153 Hinalea axillaris, 69 balteata, 69 Hind limb and foot of Camarasaur 380 Hippocampidse, 28 Hippocampus coronatus, 199 fisheri, 28 hilonis, 28 japonicus, 199 Hippoglossina punctatissimus, 299 Hippoglossoides dubius, 298 punctatissimus, 299 hippoides, Carangus, 39 hippurus, Coryphsena, 35, 225 hira, Auxis, 221 hireguro, Microstomus, 301 hirundinacea, Chelidoperca, 236 hirundo, Cypselurus, 204 hispidus, Tetraodon, 87 Histiophorus, 30 INDEX. 427 Histiopteridae, 47, 245 Histiopterus typus, 47, 245 histricosa, Deania, 108 Holacaiithus bispinosus, 63 diacanthus, 62 diodoii, 88 fisheri, 64 tricolor, 61 hollandi, Allolepis, 323, 346 Myctophum, 11, 12 Holocentridae, 26, 209 Holocentriis diadema, 27 ensifer, 27 erythrseus, 27 macrourus, 21 microstomus, 27 punctatissimiis, 27 sammara, 27 scythrops, 27 spinifer, 27 spinosissimus, 209 xantherythrus, 27 Holocephali, 6 Holotrachys lima, 26 homei, Fierasfer, 84 homogenes, Barbus, 165 homozonus, Barbus, 169 Hoplichthyidse, 57, 288 Hoplichthys citrinus, 57 langsdorfi, 288 platophrys, 57 Hoplobrotula armata, 325 Hoplostethus mediterraneus, 209 hoshinonis, Synodus, 154 howensis, Cantherines, 255 Hubbs, Carl Leavitt, 93 Hubbs and Creaser, 98 Hucho hucho, 146 perryi, 145, 338 hucho, Hucho, 146 Hymenocephalus antraeus, 21 aterrimus, 21 striatulus, 21 tenuis, 92 Hymenophysa curta, 160 Hynnodus atherinoides, 44 Hyoid arch of Camarasaurus, 367 Hypentelium, 174 Hyperoglyphe, 226 Hypodytes rubripinnis, 275 hypomelas, Stemonidium, 14 Hypomesus japonicus, 151, 152 olidus, 151 pretiosus, 151 Hypoptychus dybowskii, 324 steindachneri, 324 Hyporhamphus kurumeus, 205 pacificus, 18 sajori, 205 Hypostomides, 28 hypsclopterus, Acanthurus, 66 hystrix, Diodon, 88 iburia, Iburina, 291 Iburiella kasawae, 290, 291, 344 Iburina iburia, 290, 291 Ichthyocampus erythra3us, 28 Ichthyophis, 17 ignobilis, Caranx, 39 Iguanodon, 392 Iguanodon bernissartensis, 396 iijimai, Astronesthes, 153 Psettina, 295 Ilisha elongata, 121 illustris, Eumegistus, 35, 36 imbrius, Salvelinus, 140, 142, 144, 145, 336, 338 immaculatum, Ostracion, 256 impar, Camarasaurus, 352 imparipinnis, Abudefduf, 68 Imperial University of Kyoto, 94 incipiens, Naso, 66 incisum, Pteropsaron, 80 indicus, Platycephalus, 286, 288 Upeneus, 247 Inegocia japonica, 286, 287 Inermia vittata, 47 inhaghitsch, Salmo, 151 inermis, Cheilio, 71 Platophrys, 24 Scolopsis, 239 Sebastes, 263 Sebastodes, 261, 263 Sebastosomus, 261 Iniistius dea, 251 leucozonus, 74 mundicorpus, 74 niger, 74 pavoninus, 74 verater, 74 428 MEMOIRS OF THE CARNEGIE MUSEUM. Inimicus japonicus, 275 Inomata, Mr. S., 94 inoriiata, Raja, 112 Insidiator, meerdervoorti, 286, 287 insiilarum, Carchariiiiis, 4 Hepsetia, 29 intermedia, Acheilognathus, 162 Seriolina, 223 intermedins, Gymnocanthus, 279 interruptiis, Arelisciis, 302 innsita, Cynoglossiis, 302 Iraciindus signifer, 55 iracimdus, Aciitomentum, 268 Seliastodes, 268 Irex, 37 iridescens, Halicbanes, 70 iridornm, Limanda, 299 irradians, Leptoscarus, 74 iscliyiTis, Pseiidupeneiis, 247 Upeneus, 246 Ishikaiiia steenackeri, 189 Ishikawa, Mr. Masashi, 94 Dr. Chiyomatsu, 94 ishikaM'Se, Furcina, 280 Gnatho])ogon, 169 Oncorhynclms, 123, 124, 132, 136, 334, 338 Isospondyli, 6 Istio])horidse, 30, 222 IstioiihoiTis gladins, 30 orientalis, 222 Lsiiropsis giaiica, 4 Isiirus glaucus, 102 oxyrhynchiis, 102 izensis, Scorpaena, 272 jactator, Canthigaster, 87 japonica, Anguilla, 190 IIeterom,ycteris, 301 llime, 153 Inegocia, 286, 287 Leiiidotrigla, 289 Mobiila, 6 Nibea, 107, 244 Polymixia, 209 Pteroiilatea, 115 Rhinojilagusia, 302 Sillago, 248 Spliyrsena, 206 Squatina, 108 Synagrops, 231 japonicum, Acropoma, 231 japonicus, Aulichthys, 199 Aulopus, 153 Arctoscopus, 311 Bembras, 283 Branchiostegus, 248 Bregmaceros, 326 Galliurichtliys, 316 Centrolopliiis, 226 Coelorhynchus, 327 Conger, 194 Dasson, 318 Dasycottus, 277 Engraulis, 121, 248 Entosplienus, 98 Galeorhiims, 3, 101 Gnatliopogon, 171 Heterodontus, 99 Hippocampus, 199 Hypomesus, 151, 152 Inimicus, 275 Lateolalirax, 234 Leptocephalus, 194 Leptoscarus, 252 Leuciscus, 171 Monocentris, 209 Mulloides, 246 Ocycrius, 226, 340 Osticlithys, 209 Pemiiheris, 227, 229 Physiculus, 23, 326 Pneumatophorus, 31, 211, 212 Priacanthus, 232 Pristiophorus, 108 Pseudoblennius, 281 Pseudolabrus, 250 Scomber, 210 Sicyopterus, 309 Spratelloides, 121 Squalius, 171 Squalus, 105 Sy nodus, 154 Trachurus, 223 Triienopogon, 309 Trichiurus, 222 Uranoscopus, 316 Zebrias, 302 Zeus, 252 INDEX. 429 jenkinsi, Hemipteronotus, 74 Pomacentrus, 68 Scams, 76 Synaphobranchus, 191 Jenkinsiella macgregori, 15 Johnius, 242 Jordan, David Starr, 1, 2, 93, 122 Jordan, Eric Knight, 1 Jordan and Dickerson, Mary Cynthia, 2 Jordan and Metz, Charles William, 2 Jordan and Snyder, John Otterbein, 2 jordani, Brachirus, 273 Carangoides, 40 Cheilinoides, 73 Dendrochirus, 273 Gnathopogon, 167 Leucogobio, 167 Malthopsis, 91 Nyctimaster, 157 Pseiidoscarus, 76 Jordanicus sagamianus, 323 umbratilis, 84 jouyi, Leuciscus, 182 Moroco, 182 joyneri, Areliscus, 302 Sebastes, 264 Sebastodes, 264, 265 Sebastosomus, 264 Jugulares, 79 Julis axillaris, 69 bifer, 73 jusanensis, Leuciscus, 179, 180 kagoshimana, Scorpsena, 272 Scorpsenopsis, 272 kaianus, Sy nodus, 10 kaibarse, Pygosteus, 202 kailuse, Mursena, 15 kallosoma, Novaculichthys, 73 Kareius bicoloratus, 300 Kasawa, Mr. Masanosuke, 96 kasawse, Iburiella, 291, 344 Katsuwonidse, 219 Katsuwonus pelamis, 220 vagans, 219 kauaiensis, Aldrovandia, 9 kaupi, Physiculus, 327 kawabatse, Brevigobio, 161, 189 kawamebari, Bryttosus, 235 Kawamura, Dr. T., 95, 99 kawamurae, Oncorhynclius, 123, 128, 332, 338 Kay, LeRoy, 348 kazika, Cottus, 277 Rheopresbe, 278 keleipionis, Hemicoris, 71 Kelloggella cardinalis, 78 oligolepis, 78 kelloggi, Pseudanthias, 46 Scorpsenodes, 54 kenojei. Raja, 110, 113 Kentrocapros, 89 Kentrocapros aculeatus, 256 keta, Oncorhynclius, 129, 338 kiensis, Apogon, 231 Kishinoella, 216 Kishinoella rara, 219 kishinouyi, Lepidotrigla, 290 kisutch, Oncorhynchus, 123, 124, 131 kitaharai, Dexistes, 300 L^ops, 295 Lambdopsetta, 295 Microstomus, 300 Tanakius, 30() kiusiuanus, Leptoceiihalus, 194 knighti, Gnatholepis, 78 kobensis, Aseraggodes, 301 krusensterni, Acanthocepola, 248 Kuhlia malo, 41 mato, 41 rupestris, 232 tseniura, 42 Kuhliidse, 41, 232 Kujiensis, Raja, 109 Kuma, Dr. Toshiyasu, 93 kumse, Monomitopus, 324, 346 kumu, Chelidonichthys, 288 kurodei, Rhinogobius, 306 kurumeus, Hyporhamphus, 205 Rhodeus, 163, 242 Kyphosidse, 51, 242 Kyphosus cinerascens, 242 elegans, 51 fuscus, 51 sandvicensis, 51 Labracoglossa argent iventris, 239 Labracoglossidse, 239 Labridse, 249 Labyrinthici, 92 430 MEMOIRS OF THE CARNEGIE MUSEUM. lacepedei, Tsenioides, 310 lacrymatus, Tetraodoii, 87 lactipes, Aboma, 307 Lactophrys concatenatus, 256 tritropis, 256 Lactoria corniita, 256 diaphana, 256 galeodon, 89 schlemraeri, 89 Lady-fishes, 6 Laemonema rhodocliir, 22 Laeojis lanceolata, 295 Isevis, Miisteliis, 100 Lagocephaliis lunaris, 257 oceanicus, 86 siiadiceus, 257 Lambdojisetta kitaharse 295 Lamna nasus, 102 Lamnidse, 4, 102 Lampanyctus, 12, 13 omostigma, 12 Lamjietra planeri, 98 lampra, Muraena, 15 Lampreys, Sea-run, 98 Lampridae, 29 Lamiiris regius, 29 Lamprossa aiiteorliitalis, 156 lanceolata, Acheilognathiis, 162 Laeops, 295 lanceolatiis, Masturus, 89 Osmerus, 149 Spirinchus, 149 langsdorfi, Hoplichthys, 288 Lantern-fishes, 2 lao, Halichoeres, 70 Laosaurus celer, 403 consors, 396, 403 gracilis, 385, 403-410 lata, Dasyatis, 5 Lateolabrax jaiionicus, 234 laticeiis, Chlamydes, 78 latifrons, Ovoides, 87 Latiliis argentatus, 248 latipes, Oryzias, 199 latus, Caranx, 39 Diaphus, 156 Spams, 240 lauia. Scams, 76 Lava-flow from Mauna Loa, Deep-Sea Fishes from the Coast of Hawaii, Killed by, 2 laysanius, Antennarius, 90 Grammatonotus, 47 Gymnothorax, 16 Leather-jackets, 86 leclusei, Cymolutes, 74 Lefiia echigonia, 160 nikkonis, 160 leihala. Echidna, 17 Leiognathus michalis, 225 rivulata, 225 Leiiiranus semicinctus, 14 Lentipes concolor, 79 seminudus, 79 lentus, Camarasaiirus, 352 Lepidamia, 43 Lepidamia evermanni, 42 Lepidaplois albotseniatus, 68 modestus, 69 strophodes, 68 Lepidochsetodon, 59 Lepidopsetta bilineata, 299 mochigarei, 299 Lejiidotrigla alata, 288 abyssalis, 289 guentheri, 289 japonica, 289 kishinouyi, 290 strauchi, 289 lepomis, Coris, 70 leprosus, Antennarius, 90 Leptocardii, 3 Leptocephalidse, 13 Leptocephalus sequoreus, 13 bowersi, 13 ectenurus, 196 erebennus, 194 flavirostris, 195 japonicus, 194 kiusiuanus, 194 marginatus, 13, 195 myriaster, 194 nystromi, 195 retrotinctus, 196, 197 riukiuanus, 193, 195 leptorhynchus, Sarritor, 293 INDEX. 431 Leptoscariis cyclurus, 75 irradians, 74 japoniciis, 252 sandvicensis, 75 snyderi, 75 leptodims, Camarasaiirus, 352 Leptostomias macronema, 8 lepturus, Uroconger, 196 Lestidiiim nudum, 10 Lethenteron, 98 Lethrinus hsematopterus, 240 nematacanthus, 241 leucacme, Gymnothorax, 16 Leuciscus caerulescens, 167 dorobse, 182 hakuensis, 178 japonicus, 171 jouyi, 182 jusanensis, 179, 180 macropus, 185 minor, 185 phalacrocorax, 179, 180 platypus, 184 taczanowskii, 179 leuciscus, Cyprinus, 177 Leucogobio biwae, 172 guentheri, 168 jordani, 167 mayedse, 169, 171 leucogrammicus, Dipterygonotus, 47 leucomsenis, Salvelinus, 140, 142, 338 leucopareius, Acanthurus, 65 Leucopsarion petersi, 310 leucostictus, Gymnothorax, 16 leucozonus, Iniistius, 74 leucura, Furcaria, 67 leucurus, Uropterygius, 17 lima, Holotrachys, 26 Limanda iridorum, 299 punctatissima, 299 schrencki, 299 Limandella angustirostris, 299 yokohamse, 299 limbata, Acheilognathus, 162 limbatus, Trachinocephalus, 9 lineatus, Apogon, 230 Euthynnus, 220 Tetraodon, 87 lineolata, Sarda, 32 lineolatus, Chsetodon, 58 lineopunctatus, Xantliiclithys, 85 Linophora, 58 Liobagrus reini, 159 Liopempheris sasakii, 228, 342 vanicolensis, 229 Liparidse, 293 List of the Fishes of Hawaii, with notes and Descriptions of New Siiecies, 1-92 of the Fishes of Japan, Collected by D. S. Jordan, 1922, 93-346. littoralis, Scorpaenodes, 271 Sebastella, 271 lituratus, Callicanthus, 67 liturosa, Alutera, 86, 255 Liza haematocheila, 208 menada, 208 Lizard-fishes, 9 Loa, 61, 252 Loa excelsa, 61 lonchotus, Oxyurichthys, 78 longiceps, Monomitopus, 324 longicirrhus, Macrourus, 21 longidens, Caulolepis, 26 longifilis, Gnathopogon, 169 longirostris, Euleptorhamphus, 18 Forcipiger, 58 Longirostrum delicatissimum, 224 Lophiidse, 90, 330 Lophiomus miacanthus, 90 setigerus, 330 Lophobranchii, 28 loricatus, Gasterosteus, 200 Lotella phycis, 326 Lower jaw of Camarasaiirus, 363 loxozonus, Dasson, 318 lucifer, Astronesthes, 8 Luciogobius elongatus, 309 luetkeni, Myctophum, 11 Lumpenus fowleri, 320 luna, Caranx, 40 lunare, Thalassoma, 72 lunaris, Csesio, 239 lunatus, Callionymus, 316 lundini, Atule, 38 Decapterus, 38 lunula, Cluetodon, 59, 252 lurida, Ariomma, 35 lutescens, Thalassoma, 72 luteus, Callechelys, 15 432 MEMOIRS OF THE CARNEGIE MUSEUM. Liitianidie, 48, 237 Lutianus msselli, 237 vaigensis, 237 vitta, 237 Lycodapodidse, 84 Ij3^codontis, 16 Lycogramma zesta, 321 Maccullocliina, 44 Macdonaldia, 199 macgregori, Jenkinsiella, 15 maclinata, Elops, 119 Mackerel, 31 Mackerel-sharks, 4 macracantlms, Priacanthiis, 232 macrocephala, Merinthe, 55 macrocephaliis, Gadus, 326 Siiarus, 240 macrochir, Sebastolobiis, 259 macrognathos, Chaenogobius, 307 macrolepidotus, Heniochus, 61 Neoscopeliis, 10 macrolepis, Onigocia, 287 Saurida, 155 macronema, Leptostomias, 8 Macropharyngodon aquilolo, 70 geoffroyi, 70 Macropodus operciilaris, 207 macrops, Sy nodus, 154 macropterus, Germo, 32 Neothunniis, 219 macropiis, Barilius, 185 Leuciscus, 185 ]\'Iacrorhamphosida3, 28, 200 Macrorhamphosiis hawaiiensis, 28 sagifue, 200 macrorhynchus, Ophisurus, 198 macrostomus, Oncorhynchus, 123, 125, 134, 135, 334, 338 Macroiiridae, 20 Macrourus burragei, 21 ectenes, 21 gi liber, 21 hebetatiis, 21 holocentrus, 21 longicirrhus, 21 oliliquatus, 21 jiropinquus, 21 maculatus, Arius, 157 Caracanthus, 54 Sarcocheilichthys, 176 Silurus, 157 maculiferus, Apogon, 42 magnificus, Myrichthys, 15 Majima, Mr. T., 95 majimse, Gnathopogon, 167, 340 major, Pagrosomus, 241 makua, Ranzania, 89 Malacaiithidse, 53 Malacanthus parvipinnis, 53 Malacocephalus hawaiiensis, 22 Malakichthys griseus, 233 wakiyae, 233, 342 malma, Salveliniis, 140, 141, 144, 338 malo, Kuhlia, 41 Malthopsis jordani, 91 mitrigera, 91 Manabe, Prof. Y., 95 manazo, Cynias, 100 mancus, Platophrys, 24 Man-eaters, or Great White Sharks, 5 mantelliger, Chsetodon, 59 Mapo, 78 margaritatum, Myctophum, 11 marginalis, Brotiila, 83 marginatus, Caranx, 39 Congrogadus, 83 Leptocephalus, 13, 195 marina, Vulpeciila, 102 marmoratus, Cirrhitus, 53 Rupiscartes, 83 Sebasticus, 271 Uropterygius, 17 marshi, Etelinus, 49 martensi, Astronesthes, 153 maru, Auxis, 220 maruadsi, Decapterus, 38, 223 Masturus lanceolatus, 89 Mataeocephaliis acipenserinus, 22 matoides, Acanthurus, 65, 253 matsiibarse, Sebastes, 268 matsushimse, Crystallias, 293 mauritiana, Trachurus, 223 mauritianus, Selar, 38 Maiirolicidae, 8, 154 Maurolicus pennanti, 154 maximus, Cetorhinus, 102 INDEX. 433 mayedse, Gnathopogon, 171 Gobio, 171 Leucogobio, 169, 171 Maynea brimnea, 320 McGregor, Ernest Alexander, 93, 122 mebachi, Parathunnus, 218 Thunnus, 218 mediterraneiis, Hoplostethus, 209 medius, Camptosaurus, 385-393, 410 meeki, Priacanthus, 47, 232 meerdervoorti, Insidiator, 286, 287 Raja, 111, 113 megacephalus, Ratabulus, 286, 287 megastomiis, Congriscus, 193 Melamphaes unicornis, 26 Melamphaidse, 26 melampygus, Caranx, 39, 40, 224 Melanobranchus micronemus, 20 melanopterus, Carcharinus, 4 Gadomiis, 20 melanosticta, Sardinia, 121 Melanostoma, 43 meleagris, Gymnothorax, 16 Melichthys radula, 85 menada, Liza, 208 menesemus, Pristiapogon, 42 mento, Xanthichthys, 85 Merinthe macrocephala, 55 Metopomycter denticiilatus, 14 metoposophron, Callicanthiis, 67 miacanthus, Lophiomus, 90 Micracanthiis strigatus, 61 microcephaliis, Gasterosteus aculeatus, 202 microdon, Aprion, 49 Apsilus, 48 Fierasfer, 84 Salangichthys, 153 Ulaula, 49 Microdonophis fowleri, 14 microlepis, Antimora, 22 Monomitopus, 324 micronemus, Melanobranchus, 20 Micropercops dabryi, 304 Microphis pleurotsenia, 28 Micropterus salmoides, 92 micropus, Etrumeus, 7, 120 Microstomus hireguro, 301 kitaharse, 300 stelleri, 300 microstomus, Holocentrus, 27 Mikamo, Mr. K., 94 Mikimoto, Mr. K., 95 miliaris, Chsetodon, 59 miniatus, Scarus, 75 minor, Barilius, 185 Leuciscus, 185 Minous adamsii, 275 echigonius, 275 monodactylus, 275 Mionorus, 44 Miopsaras myops, 90 misakius, Pseudorliombus, 296 Misgurnus anguillicaudatus, 160 mitrigera, Malthopsis, 91 mitsukurii, Acanthias, 105 Chimsera, 119 Entosphenus, 98 Nibea, 243 Pliasmichthys, 119 Pteropodus, 270 Sciaena, 242 Sebastodes, 270 Squalus, 5, 104, 105, 106 Tetrapturus, 30, 222 Zacco, 187 Mitsukurina owstoni, 102 Mitsukurinidae, 102 moara, Epinephelus, 236 Mobula, 6 japonica, 6 Mobulidse, 6 mochigarei, Lepidopsetta, 299 modestum, Coradion, 252 modestus, Cantherines, 254 Lepidaplois, 69 Pseudaspius, 181 mola, Mola, 88 Mola mola, 88 Molida?, 88 molucca, Pempheris, 228 Monacanthidse, 86, 254 Monacanthus cirrhifer, 254 Monocentridse, 209 Monocentris japonicus, 209 monoceros, Alutera, 255 Monoclonius, 377, 408 monodactyla, Scorpsena, 275 monodactylus, Minous, 275 434 MEMOIRS OF THE CARNEGIE MUSEUM. Monomitopus kumse, 324, 346 longiceps, 324 microlepis, 324 Monotaxis grandociilis, 51 monopterygiiis, PleurogTammiis, 276 montanus, Atlantosaurus, 355 Moon-fishes, 29 morii, Sarcocheilichthys, 175 Moringiia hawaiiensis, 15 Moringiiidse, 15 Moriwaki, Mr. I, 95 Moroco, 177 Moroco boigi, 181 jonjd, 181, 182 steindachneri, 181, 182 yamamotis, 181, 182 Morosaurns, 379, 380, 409 morrhiia, Epinephelus, 235 Morrison formation, 347 mucifer, Gymnothorax, 16 mncronatns, Haiialogenys, 239 Mngil cephalus, 29, 207 Mugilidie, 29, 207 Mullets, 29 Mullidae, 51, 245 Mulloides auriflamma, 51 erythrinus, 51 ja])onicus, 246 pflugeri, 52 preorbitalis, 52 samoensis, 52 vanicolensis, 52 Mulliis iileurotaenia, 246 multiliarliata, Brotula, 324 multicelliis, Telestes, 177 miilticirrata, Brotula, 83 multifasciata, Neopercis, 311 multifasciatus, Upeneus, 52 multimaculatus, Pseudomonacantlius, 86 multiradiatus, Myri]iristis, 26 mundicor])Us, Iniistius, 74 Mupus, 226 .Murrena balearica, 196 kailua', 15 lampra, 15 Murapnesocidse, 14, 198 Muraenesox cinei’eus, 198 Muraeniday 15 Murakami, Mr. T., 95 murdjan, Myripristis, 26 muroadsi, Decapterus, 223 Mustelus canis, 100 griseus, 101 laevis, 100 stellatus, 100 mustelus, Squalus, 100 Myctophidae, 10, 156 Myctophum affine, 11 braueri, 11, 12 evermanni, 11 fibulatum, 11 hollandi, 11, 12 luetkeni, 11 margaritatum, 11 nitidulum, 11 reinhardti, 11, 12 Myliobatidae, 5 my ops, Miopsaras, 90 Trachinocephalus, 9, 155 Myoxocephalus raninus, 278 myriaster, Anguilla, 194 Astroconger, 195 Leptocephalus, 194 Myrichthys magnificus, 15 stypurus, 15 Myripristis argyromus, 27 berndti, 27 chry seres, 26 multiradiatus, 26 murdjan, 26 sealei, 26 symmetricus, 26 Myrophis heterognathos, 196 Myxinidae, 97 My XUS pacificus, 29 nagoyae, Rhinogobius, 305 nakamurae, Chloea, 307 Furcimanus, 320 Nakano, S., 99 Nannobrachium, 13 nigrum, 13 reinhardti, 13 nanum, Plectrogenium, 55 nanus, Camptosaurus, 388, 392, 393 narinari, Aetobatus, 5 Stoasodon, 116 Narke japonica, 109 nasica, Coiigermuraena, 196 INDEX. 435 nasiconiis, Ceratosaiiriis, 355 Naso brevirostris, 66 incipiens, 66 unicornis, 67, 253 nasus, Lamna, 102 nasutus, Coryphsenoides, 327 Naucrates ductor, 37 indicus, 37 Nealotus tripes, 221 Neamia, 44 neanis, Thalassoma, 72 nebulosa. Echidna, 17 nebulosus, Ameiurus, 91 Enedrias, 320 Zenopsis, 252 Needle-fishes, 18 neglecta, Scorpsena, 272 neglectus, Thescelosaurus, 388 nematacanthus, Lethrinus, 241 Nematognathi, 91 Nematoprora polygonifera, 14 Nemichthyidse, 14 Neobythites, 325 Neoditrema ransonnetii, 249 Neopercis multifasciata, 311 roseoviridis, 79 sexfasciata, 311 Neoscopelus alcocki, 10 macrolepidotus, 10 Neothunnus, 216 Neothunnus macropterus, 219 nerka, Oncorhynchus, 123, 126 nesiotes, Carcharinus, 4 Nettastomidse, 14 Netuma osakse, 96, 157, 340 thalassina, 157 nexilis, Antennarius, 90 nibe, Nibea, 243 Nibea albiflora, 243 japonica, 107, 244 mitsukurii, 243 nibe, 243 schlegeli, 107, 243 Nichols, John Treadwell, 1 niger, Apogon, 230 Iniistius, 74 Siifflamen, 254 nigripinnis, Hapalogenys, 239 nigromaculatus, Pseudomonacanthiis, 255 nigromaculosus, Cantherines, 254 nigrum, Centroscy Ilium, 5 Nannobrachium, 13 nikkonis, Lefua, 160 nimbaria, Vinciguerria, 9 Niphon spinosus, 234 niphonia, Sawara, 214 Niphonidae, 234 niphonius, Pseudopriacanthus, 233 nitidulum, Myctophum, 11 nocturnus, Deania, 108 Nomeidse, 35 Notacanthidae, 199 Notacanthus rissoanus, 199 notatus, Chromis, 249 Notes on Fishes of Hawaii, with De- scriptions of New Species, 2 Notropis cayuga, 169 heterodon, 169 Novaculichthys entargyreus, 73 kallosoma, 73 taeniourus, 73 tattoo, 73 woodi, 73 Nozawa, Prof. S., 95 nuchalis, Leiognathus, 225 nuda, Gymnosarda, 215 nudiceps, Pelteobagrus, 159 nudifrons, Diodon, 88 nudum, Lestidium, 10 nuttingi, Gymnothorax, 16 Polyipnus, 9 Nyctamia, 43 nycteris, Sufflamen, 85 Nyctimaster jordani, 157 nystromi, Rhynchocymba, 195 oahuensis, Canthigaster, 87 obliquatus, Macrourus, 21 obscura, Echidna, 17 obscurus, Odontobutis, 304 Tridentiger, 309 obtusata, Sphyraena, 206 Occa, 290 Occa verrucosa, 290 Occella dodecaedron, 291 Occocephalidae, 330 Oceania, Fishes of, 2 oceanicus, Lagocephalus, 86 Rhinoscopelus, 11 436 MEMOIRS OF THE CARNEGIE MUSEUM. ocellatiis, Rhodens, 163 ocellifer, Pseudorhombus, 297 octotsenia, Pseudocheilinus, 73 Ocycrius japonicus, 226, 340 Odontanthias fuscipimiis, 46 Odontoliutis oliscurus, 304 ()j2;cocephalid8e, 91 olidus, Hypomesus, 151 oligodon, Osmerus, 151 Pseudorhombus, 296 oligolepis, Kelloggella, 78 Tarjihops, 297 Uranosco])us, 313, 316 Oligoridse, 233 olivaceus, Acaiithurus, 65 Paralichthys, 297 ommatura, Parapercis, 311 omostigma, Lamiianyctus, 12 Oncesthes fluetuans, 319 Oiicoi-hynchus adonis, 123, 127, 136, i32, 338 foi-mosanus, 123, 338 gorbuscha, 123, 129, 338 ishikawie, 123, 124, 132, 136, 334, 338 kawamurse, 123, 128, 332, 338 keta, 129, 338 macrostomus, 123, 125, 134, 135, 334, 338 nerka, 123, 126 rhodurus, 123, 125, 134, 136, 137, 336, 338 tschawytscha, 124, 131 Oiiigocia macrolepis, 285, 287 opercularis, Macropodus, 207 Ophicephalidse, 92, 207 Olihicephalus argus, 207 sti-iatus, 92 Ophichthyidse, 14, 198 ()])hisurus macrorhynchus, 198 ophryas, Arothron, 87 Oplegnathidse, 245 Ol)legnathus fasciata, 245 liunctata, 245 Opsariichthys bidens, 189 ])achycephalus, 184, 187 platypus, 184 uncirostris, 188 Optomirus athcrodon, 20 Orcynus schlegeli, 216 oreas, Barbatula, 161 Oreias, 160 Oriental Steamship Company, (Toyo Kisen Kaisha), 93 orientalis, Dactyloptena, 57 Istiophorus, 222 Sarda, 32, 215 Scomberoides, 222 Thunnus, 32, 33 ornatissimus, Chastodon, 59 Halichoeres, 70 Ornitho])odous Dinosaurs, 385-410 Orthagoriscus oxyuropterus, 89 Orthrias, 160 Oryzias latipes, 199 osakiE, Netuma, 157, 340 Osbeckia scripta, 86 osimiB, Furcina, 280 Osmeridae, 149 Osmerus dentex, 149 lanceolatus, 149 oligodon, 151 Osiihronemidse, 207 Ostichthys japonicus, 209 pillwaxi, 26 Ostorhynchus, 42 Ostraciidae, 89 Ostracion immaculatum, 256 gibbosum, 256 lentiginosum, 89 oahuense, 89 sebae, 89 stellifer, 256 Ostracodermi, 89 ostroumowi, Glyptocejihalus, 301 Osurus schauinslandi, 79 Ota, Mr. T., 94 Otaki, Prof. K., 95 Otakia, 163 Otakia rasborina, 167 otakii, Hazeus, 304 Hexagrammos, 276 Othonias undovittatus, 244 othonops, Etrumeus, 7 ovalis, Furcaria, 67 Ovoides latifrons, 87 owstoni, Centroscymnus, 107 Mitsukurina, 102 Sebastodes (Primospina), 260 oxycephala, Eleotris, 304 INDEX. 437 oxyceplialus, Verreo, 69 oxyrhynchus, Isiirus, 102 Therapon, 237 Oxyurichthys lonchotus, 78 oxyuroptems, Orthagoriscus, 89 pachycephalus, Sebastodes (Ptero- podus), 270 Zacco, 184 pacificus, Diplophos, 9 Germo, 217 Hyporhamphus, 18 Myxiis, 29 Ruvettus, 34, 221 pachycephalus, Opsariichthys, 187 Zacco, 187, 188 Pagrosomus major, 241 pallasii, Clupea, 121 Pallasina eryngia, 293 Palometa, 226 paluca, Scarus, 76 Pampidse, 226 Pampus argenteus, 226 Panchax, 198 Pan-Pacific Educational Conference, 1 pantherinus, Platophr3^s, 24 Pantophos glandulifer, 157 papilio, Pegasus, 28 Parabembradidse, 281 Parabembras curtus, 281 Parachaeturichthys polynema, 308 ' Paracirrhites arcatus, 53 cinctus, 53 forsteri, 53 Parajulis, 250 Paralepididse, 10 Paralichthyidse, 296 Paralichthys olivaceus, 297 Paramyxine atami, 97 Parapercidse, 79, 311 Parapercis ommatura, 311 pterostigma, 79 pulchella, 311 Parapristipoma trilineatum, 238 Parasalanx ariakensis, 153 Parasilurus asotus, 159 Parathunnus, 216 Parathunnus mebachi, 218 sibi, 218 pardalis, Sphoeroides, 259 Pareioplitas, 54 Parexocoetus brachypterus, 19 Paronotus, 226 Parrot-fishes, 75 parva, Pseudogobio, 177 Pseudorasbora, 177 parvipinnis, Malacanthus, 53 Scorpsenodes, 54 parvisquamis, Sillago, 248 jiavoninus, Iniistius, 74 Pearl-fishes, 84 pectinirostris, Boleophthalmus, 304 Pediculati, 90 Pegasidie, 28 Pegasus papilio, 28 pelagicus, Amphioxides, 3 pelamis, Euthynnus, 31, 220 Katsuwonus, 220 Pelecanichthj^s crumenalis, 25 Peloropsis xenops, 55 Pelteobagrus fulvidraco, 159 nudiceps, 159 Pelvic girdle of Camarasaurus, 380 Pempheridae, 227 Pempheris japonicus, 227, 229 molucca, 228 pennanti, Maurolicus, 154 Pentanchida3, 100 Pentanchus profundicolus, 100 Percesoces, 29 Perciformes, 30 percoides, Pseudoblennius, 281 Percomorphi, 29 Periophthalmidie, 304 Peristediidae, 56 Peristedion engyceros, 56 gilbert! , 57 hians, 57 Perkinsia, 7 perryi, Hucho, 145, 338 . Salmo, 134 perspicillatus, Scarus, 75 peterseni, Daicocus, 290 petersi, Leucopsarion, 310 Petroscirtes fluctuans, 319 peruanus, Pneumatophorus, 211 petelli, Gymnothorax, 16 petimba, Fistularia, 28, 200 Petromyzonidae, 98 petus, Acanthocybium, 34 438 MEMOIRS OF THE CARNEGIE MUSEUM. pflaiimi, Rhinogobiiis, 306 pfliigeri, Mulloides, 52 phalacrocorax, Leucisciis, 179, 180 pliaiitasma, Chimsera, 116, 119 Phasmichthys mitsukiirii, 119 Pholidse, 320 phorcys, Carchariniis, 4 Phoxiniis septeiitrionalis, 178 steindachneri, 177, 181 phycis, Lotella, 326 Physiciilus dalwigkii, 327 fulvus, 327 grinnelli, 22 japonicus, 23, 326, 327 kaipii, 327 nematopiis, 327 rastrelliger, 327 pictiis, Gymnotliorax, 17 Plectorhynchiis, 238 Pikea aurora, 46 pilwaxi, Ostichthys, 26 piiietorum, Cleisthenes, 298 piiiguis, Sphyrsena, 206 pinnatiiliis, Decapterus, 37 Pisoodoiiophis zophistiiis, 198 Plagiopsetta giossa, 301 plaiieri, Lampetra, 98 Platophrys chlorospilus, 24 coarctatiis, 24 inermis, 24 mancus, 24 ])aiitlierinus, 24 platophrys, Iloiilichthys, 57 Platybelone ]ilatyura, 18 Platycephalidse, 285 Platycephalus iiidiciis, 286, 288 platypus, Barilius, 184, Zacco, 184, 188 Platyrhiiia sinensis, 109 Platyrhinidie, 109 platyrhynchus, A]iristurus, 99 Scylliorhinus, 99 platyura, Platybelone, 18 plebeius, Polynemus, 247 Plecoglossidai, 147 Plecoglossus alt i veils, 147, 338 Plectognathi, 84 Plectorhynchiis cinctus, 238 liictus, 238 Plectrogenium nanum, 55 Pleurogrammus azonus, 276 monopterygius, 276 Pleuronectidai, 24, 297 Pleuronichthys cornutus, 298 pleurostigma, Upeneus, 52 pleurotsenia, Microphis, 28 Upeneus, 246 plicatellus, Ateleopus, 24 plinthus, Alseops, 298 Plotosidse, 158 Plotosus anguillaris, 158 pluvius, Salvelinus, 140, 141, 144, 336, 338, Pneumatophorus, 210 Pneumatophorus australasicus, 211 colias, 211 diego, 31, 211 grex, 31, 211 japonicus, 31, 211, 212 peruanus, 211 tapeinocephalus, 210, 212 poecilepterus, Halich ceres, 250 Poecilopsetta hawaiiensis, 24 polita, Atule, 38 polhix, Cottus, 278 Polyacanthonotus challengeri, 199 Polydactylus, 30 polygonifera, Nematoprora, 14 Polyipnus, 9 nuttingi, 9 Polymixia berndti, 26 japonica, 209 Polymixiidse, 26, 209 polynema, Parachseturichthys, 308 Polynemidse, 30, 247 Polynemus plebeius, 247 sexfilis, 30 Pomacentridse, 67, 249 Pomacentrus coelestis, 249 jenkinsi, 68 Pomadasidse, 238 Pontinus spilistius, 55 Porcupine-fishes, 88 porphyreus, Upeneus, 52 potteri, Centropyge, 62 Prentice, Sidney, 348 preorbitalis, Mulloides, 52 pretiosus, Hypomesus, 151 Ruvettus, 221 Priacanthidie, 47, 232 INDEX. 439 Priacanthus alalaua, 47 cruentatus, 47 japonicus, 232 macracanthus, 232 meeki, 47, 232 pricei, Stephanolepis, 86 PriDiospina owstoni, 260 sasakii, 260 Prionace glauca, 4, 101 Prionodon, 4 Pristiapogon erythriniis, 42 menesemus, 42 snyderi, 42 pristilepis, Dasyscopeliis, 10 Pristiophoridse, 108 Pristiophorus japonicus, 108 Pristipomoides violescens, 48 Pristiurus, 3 proboscidea, Aldrovandia, 9 Proceedings of the U. S. National Mu- seum, 93 profundicolus, Pentanchus, 100 prometheus, Prometichthys, 34, 221 Prometichthys prometheus, 34, 221 solandri, 34 Promyllantor, 13 alcocki, 13 propinquus, Macrourus, 21 proridens, Chlorophthalmus, 10 prorigera, Chascanopsetta, 25 Protopsetta herzensteini, 298 psalion. Echidna, 17 psegma, Canthigaster, 87 Pseudanthias kelloggi, 46 Pseudaspius atrilatus, 177, 182 bergi, 180 modestus, 183 Pseudobagrus aurantiacus, 159 Pseudoblennius japonicus, 281 percoides, 281 Pseudocheilinus evanidus, 73 octotsenia, 73 Pseudogobio esocinus, 174 parva, 177 Pseudojulis cera.sina, 69 Pseudolabrus japonicus, 250 Pseudomonacanthus multimaculatus, 86 nigromaculatus, 255 Pseudoperialampus typus, 161 Pseudopriacanthus niiihonius, 233 Pseudorasbora, 175 Pseudorhombus arsius, 296 cinnamomeus, 296 misakius, 296 ocellifer, 297 oligodon, 296 swinhonis, 297 Pseudorasbora parva, 199 Pseudoscarus heliotropinus, 76 jordani, 76 sumbawensis, 76 troscheli, 76 vitriolinus, 76 Pseudotriakis acrages, 102 Pseudupeneus ischyrus, 247 Psenopsis anomala, 226 Psettina iijiime, 295 Psychichthys eidolon, 117, 332 Pterogobius daimio, 308 Pterois lunulata, 273 sphex, 56 Pteroplatea japonica, 115 Pteropodus mitsukurii, 270 Pteropsaron incisum, 80 pterostigma, Parapercis, 79 Pterothrissus gissu, 119 Puffers, 86 pulchella, Parapercis, 311 pulcherrima, Coris, 70 punctata, Girella, 241 Oplegnathus, 245 punctatissima, Limanda, 299 punctatissimus, Hippoglossina, 299 Hippoglossoides, 299 Holocentrus, 27 pimctatofasciatus, Chaetodon, 59 punctatus, Caranx, 40 Clupanodon, 120 Pungitius brevispinosus, 203 pungitius, 203 sinensis, 202 tymensis, 203 Pungtungia herzi, 176 hilgendorfi, 176 purpureomaculatus, Areliscus, 303 purpurescens, Chimaera, 6, 117 Seriola, 37, 223 purpureum, Thalassoma, 72 purpureus, Stolephorus, 8 \ 440 MEMOIRS OF THE CARNEGIE MUSEUM. Pygosteus, 200 Pygosteus kaibarie, 202 pungitius, 203 steindachneii, 202 undecimalis, 203 pyrrhogranimus, Haliclioeres, 250 qiiadiimaculatus, Chsetodon, 59 quernus, Epineplielus, 46 quinqueradiata, Seriola, 37, 222 Quisqiiilius eugenius, 77 Rachycentridse, 225 Rachycentroii canadum, 225 radiila, Melichthys, 85 Tseniopsetta, 24 Raia fusca, 110 Raja biiioculata, 111 fusca, 110 inornata, 112 isotrachys, 109 kenojei, 110, 112, 113 kujiensis, 109 meerdervoorti. 111, 113 sniirnovi, no. 111 teiigii, no tobie, no, 112, 113 Rajidae, 109 raniiius, Myoxocephalus, 278 ransonnetii, Neoditrema, 249 Ranzania makua, 89 truncata, 89 rara, Kishinoella, 219 rasborella, Hemigrammocypris, 161, 189 rasborina, Otakia, 167 Rastrelliger, 210 Ratabulus megacephalus, 286, 287 Record of Fishes obtained by D. S. Jordan, 93 rectangulus, Balistapus, 85 regius, Lampris, 29 reinhardti, Myctophum, 11, 12 Nyctimaster, 13 reini, Lioliagrus, 159 reniedius, Hemicoris, 71 I’emiger, Setarches, 55 Remora remora, 77, 294 remora, Remora, 77, 294 Remorina ahiescens, 77, 294 Remoropsis brachyptera, 294 reticularis, Gymnothorax, 198 reticulatus, Semicossyphus, 250 retifera, Halieutsea, 91 retrotincta, Congrina, 197 retrotinctus, Leptocephalus, 196 rhabdotus, Carangus, 39 Caranx, 39 Rliechias armiger, 14 Rhegnopteri, 30 Rheopresbe fujiyamoe, 278 kazika, 278 Rhinobatidse, 108 Rhinobatos schlegeli, 109 Rhinogobius candidius, 305 giurinus, 305 kurodei, 306 nagoyoe, 305 pflaumi, 306 similis, 305 taiwanus, 305 Rhinoplagusia japonica, 302 Rhinoscopelus oceanicus, 11 tenuiculus, 11 rhodadenia, Cyclothone, 8 Rhodeus kurumeus, 163 ocellatus, 163 rhodochir, Lsemonema, 22 rhodurus, Oncorhynchus, 123, 125, 134, 136, 137, 336, 338 rliombea, Acheilognathus, 161 rhombeum, Zebrasoma, 66 Rhyacanthias carlsmithi, xvi, 46, 236 Rhynchoconger, 192 Rhynchoconger ectenurus, 196 Rhynchocymba, 192 Rhynchocymba nystromi, 195 Ribs of Camarasaurus, cervical, 374 thoracic, 375 richardsoni, Callionymus, 317 Richardsonius brandti, 177 semotilus, 177 rikuzenius, Dexistes, 298 riukiuanus, Leptocephalus, 193 rivulata, Leiognathus, 225 rivulatus, Canthigaster, 87 Roa, 252 robustus, Camarasaurus, 352 Cynoglossus, 302 rochei, Auxis, 220 INDEX. 441 Rock-cod, 54 Rogadius asper, 286, 288 Rooseveltia aloha, 48 brighami, 48 rosea, Coris, 71 roseoviridis, Neopercis, 79 roseiim, Bembradium, 56 Rosicola fuscescens, 260 rostrata, Brachyopsis, 293 Deania, 108 rostratus, Fodiator, 19 rotundatus, Canthidermis, 254 riibescens, Exocoetus, 19 rubripes, Sphoeroides, 258 rubripinnis, Hypodytes, 275 rubrovinctus, Callionymus, 80 Rudarius ercodes, 255 Rudder-fishes, 51 rufescens, Helicolenus, 54 rupestris, Kiihlia, 232 Rupiscartes gibbifrons, 83 marmoratus, 83 variolosus, 83 ruscosiim, Centroscy Ilium, 5 russelli, Lutianus, 237 ruthise, Zanclus, 64 rutilus, Salarias, 83 Ruvettus pacificus, 34, 221 pretuosus, 221 tydemani, 221 sachi, Draciscus, 293 Sacrum of Camarasaurus, 370 sagamianus, Carapus, 323 Jordanicus, 323 sagamius, (?) Carapus, 323 sagifue, Macrorhamphosus, 200 Sail-fishes, 30 saira, Cololabis, 206 sajori, Hyporhamphus, 205 Salangichthys microdon, 153 Salangidse, 153 Salarias cypho, 83 edentulus, 83 rutilus, 83 zebra, 83 Salmo inghaghitsch, 151 perryi, 134 shasta, 139 salmoides, Micropterus, 92 Salrnonidse, l)y David Starr Jordan and Ernest Alexander McGregor, 122 Salvelinus fontinalis, 145 imbrius, 142, 145, 336, 338 leucomoenis, 338 malma, 141, 338 pluvius, 141, 336, 338 spectabilis, 140, 338 Samaridse, 301 Samariscus corallinus, 25 sammara, Holocentrus, 27 Samoa, The Fishes of, 2 samoensis, Mulloides, 52 sancti-petri, Scomberoides, 37 sandvicensis, Acanthurus, 66 Antennarius, 90 Eleotris, 77 Gobius, 78 Gomphosus, 71 Kyphosus, 51 Leptoscarus, 75 sanguineus, Verriculus, 69 Sara, Acanthocybium, 34, 213 Sarcocheilichthys maculatus, 176 morii, 175 variegatus, 175 Sarda, 213 Sarda chilensis, 32 lineolata, 32 orientalis, 32, 215 sarda, 32 Sardinia melanosticta, 121 Sarritor leptorhynchus, 293 Sasaki, Prof. M., 95 sasakii, Liopempheris, 228, 342 Primospina, 260 Sebastodes, 260 sasa3, Glyptoceiihalus, 301 satsumae, Lysoderma, 275 Sayonara, 236 Zaliscopus, 313 Saurida argyrophanes, 155 eso, 155 gracilis, 10 grandisquamis, 155 tiimbil, 155 Saurus, 10 Sawara, 213 Sawara niphonia, 214 saxatilis, Abudefduf, 249 442 MEMOIRS OF THE CARNEGIE MUSEUM. Sayonara satsumse, 236 Scseops arenicola, 25 grandisqiiama, 295 liawaiiensis, 25 xenandrus, 25 scaliRTim, Xesiirus, 253 Scaridae, 252 Scaridea aerosa, 75 balia, 75 zonarcha, 75 Scams ahula, 75 bataviensis, 76 bennet4i, 76 borboms, 75 bmnneus, 75 diibius, 75 erythrodon, 76 formosiis, 76 gilberti, 76 jenkinsi, 76 laiiia, 76 miniatus, 75 paluca, 76 perspicillatus, 75 Scepterias fragilis, 45 schauinslandi, Osurus, 79 schismatorhynchiis, Tylosums, 206 schlegeli, Caprodon, 236 Cepola, 249 Erythrocles, 47, 245 Nibea, 107, 243 Orcynus, 216 Rhiiiobatos, 109 Sebastodes, 267 Sebastosomus, 267 Syngnathus, 199 schlemmeri, Lactoria, 89 schrencki, Limanda, 299 Scisena coitor, 242 mitsiikurii, 242 umbra, 242 Scisenidse, 242 sciera, Dasyatis, 5 scintillans, Erythrocles, 47 Sclerodermi, 84 Sclerotic ring in eye of Camarasaurus, 366 Scolopsis inermis, 239 Scomber, 31, 210 Scomberoides orieiitalis, 222 sancti-petri, 37 tolooparah, 37 Scomberomorus, 213 Scomberomorus sinensis, 214 ScombresocidsD, 206 Scombridse, 31, 209 Scombropidse, 231 Scombrops bobps, 231 Scorpaenidse, 54, 259 Scorpsena gibbosa, 272 kagoshimana, 272 monodactyla, 275 neglecta, 272 Scorpsenodes kelloggi, 54 littoralis, 271 parvipinnis, 54 Scorpsenopsis altirostris, 55 cacopsis, 55 catocala, 55 cirrhosa, 272 gibbosa, 55, 272 kagoshimana, 273 Scorpion-fishes, 54 scripta, Alutera, 86 Osbeckia, 86 Scuticaria tigrina, 17 Scylliorhinid®, 3, 99 Scylliorhinus platyrhynchus, 99 scythrops, Holocentrus, 27 scythropus, Sebastodes (Acuto- mentum), 269 Sea-bass, 45 Sea-bats, 91 Sea-breams, 35 Sea-devils, 90 Sea-horses, 28 Seale, Alvin, 2 sealei, Myripristis, 26 Sea-moths, 28 Sea-run Lamprey, 98 Sea-toads, 90 sebae, Ostracion, 89 Sebastapistes asperella, 54 ballieui, 54 coloratus, 54 coniorta, 54 corallicola, 54 galactacma, 54 Sebastella littoralis, 271 INDEX. 443 Sebastes inermis, 263 joyneri, 264 matsubarae, 268 ventricosus, 263 Sebastichthys, 259 Sebasticus albofasciatus, 271 marmoratus, 271 tsuraara, 271 Sebastodes elegans, 260, 270 flammeus, 260, 268 fuscescens, 260, 261 guntheri, 260, 261 inermis, 260, 261 iracundus, 268 joyneri, 264, 265 mitsikurii, 270 nigrocinctus, 258 owstoni, 260 pachycephalus, 270 paucispinis, 258 sasakii, 260 schlegeli, 267 scythropus, 269 steindachneri, 260, 267 taczanowskii, 260, 267 thompsoni, 260, 265 tokionis, 260, 263 trivittatus, 270 ventricosus, 261, 263 vulpes, 270 Sebastolobus macrochir, 259 Sebastopsis, 271 Sebastosomus flammeus, 260 inermis, 260, 261 itinus, 260 joyneri, 264 schlegeli, 267 steindachneri, 260 taczanowskii, 260 thompsoni, 260 Sectator azureus, 51 Selar hasselti? 38 mauritianus, 38 Selenichthyes, 29 semicinctus, Leiuranus, 14 Semicossyphus reticulatus, 250 semifasciata, Argentina, 152 semilineatus, Apogon, 230 seminudus, Lentipes, 79 semipunctatus, Asterropteryx, 77 sernotilus, Ricliardsonius, 177 senboe, Chloea, 307 sentipellis, Trachonurus, 22 septemfasciatus, Epineplielus, 236 septentrionalis, Acanthochsetodon, 253 Phoximus, 178 Seriola aureovittata, 222 purpurascens, 37, 223 quinqueradiata, 37, 222 sparna, 37 Seriolina intermedia, 223 serpens, Gempylus, 35 Serranidse, 45, 236 serrata, Fistularia, 28, 200 Serri vomer beani, 14 servus, Therapon, 237 Setarches remiger, 55 setifer, Chsetodon, 58, 252 setiger, Dasycottus, 277 setigerus, Lophiomus, 330 sexfasciata, Neopercis, 311 sexfilis, Polynemus, 30 Sharks, Mackerel, 4 Thresher, 4 shasta, Salmo, 139 sibi, Germo, 33, 218 Parathunnus, 218 Thunnus, 218 Siboma, 177 Sicydium albotseniatum, 79 stimpsoni, 79 Sicyopterus japonicus, 309 sieboldi, Ulaula, 48, 49, 237 Zacco, 186 signifer, Iracundus, 55 Sillaginidse, 248 Sillago japonica, 248 parvisquamis, 248 Siluridse, 159 Silurus calvarius, 159 Silversides, 29 similis, Ctenogobius, 305 Rhinogobius, 305 simus, Cypselurus, 20 sindonis, Abudefduf, 68 sinensis, Pungitius, 202 Scomberomorus, 214 Siniperca chua-tsi, 235 sinuosus, Cheilinus, 73 sivicola, Watasea, 325 444 MEMOIRS OF THE CARNEGIE MUSEUM. Skull of CamarasaiiiTis, 353-3G3 Sleepers, 77 smiriiovi, Raja, 111 Snake-eels, 14 Snake-mackerels, 34 Snappers, 48 Snipe-eels, 14 snodgrassi, Sphju’sena, 30 Snyder, Prof. John OtJerbein, 93 snyderi, Leptoscarus, 75 Pristiapogon, 42 Tsenioides, 310 Snyderidia canina, 84 solandri, Acantliocybium, 34 ?Solea anonyma, 302 Soleida?, 30 i Soles, 25 Sorcerers, 14 sordidiis, Abudefdiif, 08 Blennius, 82 Sparkhu, 51, 240 Sparisomatidse, 74 sparna, Seriola, 37 Spams latus, 240 macroceplialus, 240 Spear-fishes, 30 speciosiis, Gnathonodon, 41 spectabilis, Salvelinus, 140, 144, 338 Sphagebranchns flavicaiidus, 14 Spheroides florealis, 86 sphex, Pterois, 56 Sphocroides alboplumbcus, 258 guttulatus, 258 oblongiis, 258 ocellatiis, 258 ])ardalis, 259 rnbripes, 258 stictonotus, 258 vermicularis, 259 Sphyrsena helleri, 29 japonica, 206 obtusata, 206 liinguis, 206 snodgrassi, 30 Sphyrienidse, 206 Sphyrna zygsena, 4, 101 Sphyrnidie, 4, 29, 101 spilistiiis, Pontinus, 55 spilonota, Capropj^gia, 89 spilonotopterus, Cypselurus, 20 spilosomus, Stephanolepis, 86 Spinax, 5 siickleyi, 103 spinifer, Holocentms, 27 spinosus, Dasyscopelus, 11 Niphon, 234 Wakijuis, 286, 287 Spirinchus lanceolatiis, 149 splendens, Beryx, 209 Exocoetus, 203 spongicejis, Aiiristurus, 3 Spratelloides, 8 japonicus, 121 Scpialidse, 5, 103 Squalius cserulescens, 167 japonicus, 171 Squalus acanthias, 104 acutipinnis, 107 blainvilhi, 107 brevirostris, 106 fernandezianus, 107 fernandinus, 107 granulosus, 108 japonicus, 105 mitsukurii, 5, 104, 105, 106 mustelus, 100 suckleyi, 103 sucklii, 103 wakiyje, 104 scpiamiceps, Chrionema, 79 Squainipennes, 58 Scpiirrel-fishes, 26 Squatina japonica, 108 Squatinidse, 108 stamineus, Chonophorus, 79 Stanford University, 96 starksi, Ebosia, 273 steenackeri, Ishikauia, 189 Stegosaurus, 408 steindachneri, Antigonia, 58 Argo, 225 Gymnothorax, 16 Hemitremia, 161 Moroco, 181, 182 Phoxinus, 177, 181 Pygosteus, 202 Sebastodes (Sebastosomus), 267 Taractes, 225 stelgis, Cyttominius, 23 stellata, Halieutsea, 330 INDEX. 445 stelkitus, Caranx, 40 Mustelus, 1()() Sternal plates of Camarasaiirus, 377 Steriioptychidie, 9 Sternoptix, 9 claiphana, 9 Stethojulis albovittatiis, G9 terina, 250 Stetliopi'istes eos, 23 Stewai’t, Douglas, 348 Stickseidse, 320 stimiisoni, Sicydium, 79 stelleri, Microstomus, 300 stellifer, Ostracion, 256 Triorus, 256 stellulatiis, Dibranchus, 91 Stemonidium hypomelas, 14 Stephanolepis price!, 86 spilosomiis, 86 Sting-rays, 5 Stoasodon narinari, 116 Stolephorus, 8 piirpureus, 8 Stomatidie, 8 strauchi, Lepidotrigla, 289 striatiilus, Hymenocephalns, 21 striatus, Ctenochietus, 66 Ophicephalus, 92 strictus, Symphiii'iis, 25 strigatus, Micracanthus, 61 strigosus, Acantliuriis, 66 Stromateidse, 226 strophodes, Lepidaplois, 68 stypurus, Myrichthys, 15 suckleyi, Spinax, 103 Squalus, 103 sucklii, Acanthias, 103 Squalus, 103 Sufflamen bursa, 85 capistratus, 85 fuscolineatus, 85 niger, 254 nycteris, 85 vidua, 85 sumbawensis, Pseudoscarus, 76 supremus, Camarasaurus, 352 Surgeon-fishes, 65 Surmullets, 51 suwse, Gnathopogon, 166 swinhonis. Pseudorhombus, 297 symmetricus, Myil])i'istis, 26 Symphurus strictus, 25 undatus, 25 Symiihysanodon, xvi Symphysanodon tyims, 236 Synagrops argyrea, 43 japonica, 231 Synaphobranchidse, 13, 191 Synaphobranchus affinis, 191 brachysomus, 13 jenkinsi, 191 taketie, 191 Synentognathi, 18 Syngnathidse, 28, 199 Syngnathus schlegeli, 199 Synodontidse, 9, 154 Sy nodus, 10 fuscus, 154 hoshinonis, 154 japonicus, 154 kaianus, 10 macrops, 154 varius, 10 tabira, Acheilognathus, 162 taczanowskii, Leuciscus, 179 Sebastodes, 267 Taenianotus citrinellus, 55 garretti, 55 Ta^nioides lacepedei, 310 snyderi, 310 Tseniopsetta radula, 24 tseniura, Kuhlia, 42 tseniourus, Novaculichthys, 73 Talus tumifrons, 241 Takayasu, Mr. S., 95 taketse, Synaphobranchus, 191 Tanaka, Prof. S., 94, 96 Tanaka, Baron Y., 99 tanakse, Zestichthys, 321, 346 Tanakius, 300 kitaharse, 300 tapeinocephalus, Pneumatophorus, 210, 212 tapeinosoma, Auxis, 31, 220, 221 Taractes steindachneri, 225 Tarphops oligolepis, 297 tattoo, Novaculichthys, 73 Teeth of Camarasaurus, 365 ?Telescopias gilbert!, 231 446 MEMOIRS OF THE CARNEGIE MUSEUM. Telestes multicellus, 177 temmincki, Ditrema, 249 Zacco, 186, 188 Ten-pounders, 6 tenuiculus, Rhinoscopelus, 11 tenuis, Arnoglossus, 295 Hymenocephalus, 92 tenuispinis, Halichoeres, 250 terina, Stethojulis, 250 tessellatus, Cantherines, 254 Tetraodon hispidus, 87 lacrymatus, 87 lineatus, 87 Tetraodontidse, 86, 257 Tetrapterus mitsukurii, 30, 222 Tetrosomus gibbosus, 256 Teuthidae, 253 Teuthis fuscescens, 253 guntheri, 65 thalassina, Netuma, 157 Tlialassoma aneitense, 72 ballieui, 72 cupido, 251 duperrey, 72 fuscum, 72 lunare, 72 lutescens, 72 neanis, 72 purpureum, 72 umbrostigma, 72 thalassopterus, Gymnothorax, 16 thazard, Auxis, 31, 220 Tlieragra chalcogramma, 326 ddierapon ox^u'hynchus, 237 servus, 238 Theraponidse, 237 Thescelosaurus, 388, 408 neglectus, 388 Thespesius, 408 Thompson, William Francis, 93 thompsoni, Caranx, 40 Sebastodes, 265 Thoracic ribs of Camarasaurus, 375 Thread-fishes, 30 Three New Hawaiian Fishes, 2 Thresher Sharks, 4 Thunnida!, 215 Thunnus, 216 Thunnus mebachi, 218 orientalis, 32, 33, 216 sibi, 218 thynnus, 32, 33 thynnus, Thunnus, 32, 33 thyrsitoides, Gempylus, 35 Tifia corallicola, 60 Tigoma, 177 tigrina, Scuticaria, 17 tigrinus, Galeocerdo, 4 tobse. Raja, 113 tokionis, Sebastodes, 263 tolooparah, Scomberoides, 37 Top-minnows, 92 Torpedinidse, 109 tosse, Zalescopus, 312, 344 Toyo Risen Kaisha (Oriental Steamship Company), 93 Trachidermus fasciatus, 277 Trachinocephalus, 9 limbatus, 9 my ops, 9, 155 Trachonurus sentipellis, 22 Trachurops, 38 crumenophthalma, 38 Trachurus, 38 Trachurus japonicus, 223 mauritiana, 223 trachurus, Gasterosteus, 200 tragula, Upeneoides, 245 Triacanthida;, 254 Triacanthodes anomalus, 253 Triacanthodidse, 253 Triacanthus brevirostris, 254 Trisenopogon barbatus, 209 japonicus, 309 Triceratops, 408 Trichiuridse, 222 Trichiurus japonicus, 222 Trichodon, 311 Trichodontidae, 311 tricolor, Gomphosus, 71 i Holacanthus, 61 tridens, Antennarius, 330 Tridentiger bifasciatus, 309 obscurus, 309 trifasciatus, Chaetodon, 59 Trigger-fishes, 84 Triglidae, 288 trilineatum, Parapristipoma, 238 trilobatus, Cheilinus, 73 trimaculatus, Dascyllus, 67 INDEX. 447 Trioms stellifer, 256 triostegus, Acanthurus, 66 tripes, Nealotus, 221 tritor, Echidna, 17 tritropis, Lactophrys, 256 trivittatus, Sebastodes (Pteropodus), 270 troscheli, Pseudoscarus, 76 trossulus, Aspidontus, 318 Dasson, 318 Trumpet-fishes, 27 truncata, Ranzania, 89 Trunk-fishes, 89 Trutta, 139 tschawytscha, Oncorhynchus, 124, 131 tsirimenara, Epinephelus, 236 Tsuchiga, Prof. Y., 95 tsuchigffi, Gnathopogon, 170 tsurugse, Atherina, 207 tsushimse, Aboma, 307 tumbil, Saurida, 155 tumifrons, Taius, 241 tutuilse, Centropyge, 62, 63 tydemani, Ruvettus, 221 Tylosurus anastomella, 206 giganteus, 18 schismatorhynchus, 206 typus, Histiopterus, 245 Pseudoperilampus, 161 Tyrannosaurus, 360 • Uintasaurus douglassi, 384 Ulaula microdon, 49 sieboldi, 49, 237 umbra, Acanthurus, 65 Catalupha, 227 Scisena, 242 umbratilis, Jordanicus, 84 umbrilatus, Hemipteronotus, 74 umbrinus, Chaunax, 90 umbrostigma, Thalassoma, 72 uncirostris, Opsariichthys, 188 undatus, Symphurus, 25 undovittatus, Othonias, 244 undulatus, Gymnothorax, 16 unicolor, Exocoetus, 19 Gymnosarda, 31 unicornis, Melamphaes, 26 Naso, 67, 253 unimaculatus, Chaetodon, 59 unipinna, Amphiprionichthys, 54 United States Fish Commission, 1 University of Michigan Museum, 96 Upeneoides arge, 52 bensasi, 245 tragula, 245 vittatus, 245 Upeneus bifasciatus, 52 chryserydros, 52 chrysonemus, 52 crassilabris, 52 fraterculus, 52 indicus, 247 ischyrus, 246 multifasciatus, 52 pleurostigma, 52 pleurotsenia, 246 porphyreus, 52 vittatus, 52 Uranoscopidae, 312 Uranoscopus bicinctus, 315, 316 japonicus, 315, 316 oligolepis, 313, 315, 316 Uraspis cheilio, 40 helvolus, 40 Uroconger, 192 Uroconger lepturus, 196 urolampus, Diaphus, 12 Urolophus fuscus, 115 Uropterygius, 17 leucurus, 17 marmoratus, 17 ushiei, Dasyatis, 114 vagans, Katsuwonus, 219 vaigensis, Lutianus, 237 valenciennei, Callionymus, 317 valentini, Aulostomus, 28 vanicolensis, Liopempheris, 229 Mulloides, 52 variegatus, Sarcocheilichthys, 175 Verasper, 297 variolosus, Rupiscartes, 83 varius, Gomphosus, 71 Sy nodus, 10 veliferum, Zebrasoma, 66 Vellitor centropomus, 281 ventricosus, Sebastes, 263 ventricosus, Sebastodes, 261, 263 venusta, Coris, 71 448 MEMOIRS OF THE CARNEGIE MUSEUM. Verasper variegatus, 297 verater, C'hromis, G7 Iiiiistius, 74 Verilinse, 50 vermicularis, Siihooroides, 259 verrens, Veternio, 13 Verreo oxycephalus, G9 Verriculus sanguineus, G9 Vertebral column of (’'amarasaurus, 3G7 verticalis, Aldrovandia, 9 Vesposus egi’egius, 24 Veternio, 13 verrens, 13 vidua, Sufflamen, 84 villosus, EtmopteiTis, 5 Vinciguerria, 9 nimbaria, 9 vincta. Echidna, 17 vinolentus, Gymnothorax, 15 violescens, Pristiiioinoides, 48 vircscens, Aprion, 49 Fundulichthys, 177 Fundulus, 199 virgata, Albula, G virgatulus, Ctenobogius, 30G virgatum, Euthyopteroma, 240 Vitraria clarescens, 78 vitriolinus, Pseudoscarus, 7G vitta, Lutianus, 237 vittata, Inermia, 47 vittatus, Goniistius, 53 Upeneus, 52 Upeneoides, 245 volitans, Exocoetus, 19, 203 vulgaris, Acantbias, 105 Vulpecula marina, 102 vulpes, Alopias, 4 Sebastodes, 270 vulpinus, Alopias, 101 waialuse, Gymnothorax, IG waikiki, Apogonichthys, 43 Wakiya, Dr. Yojiro, 94, 95 wakiyae, Malakichthys, 233, 342 Squalus, 104 Wakiyus spinosus, 28G, 287 Watasea sivicola, 325 woodi, Novaculichthys, 73 xantherythrus, ITolocentrus, 27 Xanthichthys lineopunctatus, 85 mento, 85 xanthopterus, Acanthurus, G5 xanthostomus, Gymnothorax, 17 xenandrus, Scseops, 25 Xenoberyces, 2G xenops, Peloropsis, 55 Xesurus scalprum, 253 Xiphias giadius, 30 Xiphiidse, 30 Xiphypops fisheri, G4 Xyrichthys niveilatus, 74 Xystrias grigorjewi, 297 yaito, Euthynnus, 220 Yamamoto, Prof. S., 95, 183 yamamotis, Aloroco, 182 yatabei, Blennius, 318 Yeto, Yoshio, 93 yokohama>, Gobius, 305 limandella, 299 Yoshizawa, Mr. S., 94 Zacco acanthogenys, 184 evolans, 185 mitsukurii, 187 pachycephalus, 184, 187, 188 platyinis, 184, 188 sieboldi, 18G temmincki, 184, 18G, 188 Zalarges, 9 Zalescopus satsuime, 313 tosse, 312, 344 Zanclus canescens, G4 cornutus, G4 ruthise, G4 zanclus, Bero, 280 Zanotocanthus, 199 zebra. Echidna, 17 Goniistius, 247 Heterodontus, 99 Salarias, 83 Zebrasoma flavescens, GG rhombeum, GG veliferum, GG Zebrias japonicus, 302 zebrinus, 302 zebrinus, Zebrias, 302 Zeidse, 23, 252 Zenopsis nebulosus, 252 Zeoidea, 23 zesta, Bothrocara, 321 Lycogramma, 321 Zestichthys, tanaka3, 321, 346 Zeus japonicus, 252 Zezera, 176 Zoarcidse, 320 zonarcha, Scaridea, 75 INDEX. zonata, Echidna, 17 zonatus, Goihistius, 247 zonifer, Erilepis, 276 zonurus, Cheiliniis, 73 zophistius, Piso6dono]iliis, 198 zunasi, Harengula, 121 zygsena, Spliyrna, 4 449 »l‘ r ifi 'V? 'rVl/V, 'f. 1C f. V 1 \ , ■.'*M N' / ‘W. K, * «• V S': ir I:,*' \''- s i' : •'» '.*-v' 'A / V.t. ■4 • . V V ‘i * 4^- >i*ji ir'.’l ,1 ^ 1 fjK 1 •'ii. I * < '.>1 )' t i V r)^ ) ^/)) >») > y h)^ 1 . >J) 3. > )>] )> )' . i:)Vl > ) ) ’ '! )