(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Annals of the Carnegie Museum"

ANNALS 



CARNEGIE MUSEUM 



Volume XII 
1919 



W. J. HOLLAND, Editor 



Published by the Authority of the • 
Board of Trustees of the Carnegie Institute 
January and December, 1919 



PRESS OF 

THE NEW ERA PRINTING COMPANY 

LANCASTER, PA. 




TABLE OF CONTENTS. 



PAGES 

Title-page and Table of Contents i-iii 

List of Plates iv 

List of Figures in Text v 

List of Species and Varieties Xew to Science vi-viii 

Errata et Corrigenda : ix 

Editorial Notes r-3 ; 33-39 

Obituary Notes: Charles Rochester Eastman, Henry John 
Heinz, Andrew Arnold Lambing, Herbert Huntington Smith. 

By W. J. Holland 347-363 

L A Catalog of the Collection of Watches Belonging to Mr. H. J. 
Heinz of Pittsburgh and Deposited by him in the Carnegie 
Museum. By Douglas Stewart, W. J. Holland, and A. S. 

Coggeshall ^ 4- 32 

n. Report upon the Material Discovered in the Upper Eocene of 
the Uinta Basin by Earl Douglas in the Years 1908-1909, and 

by O. A. Peterson in 1912. By O. A. Peterson 40-168 

in. Contributions to the Study of the Fresh-water Fishes of the 

Island of Formosa. By Masamitsu Oshima 169-328 

IV. On Elephenor, A New Genus of Fishes from Japan. By David 

Starr Jordan 329-342 

V. A Description of Cypripedium passerinum. By Otto E. 

Jennings 343^344 

Index 345-369 



3 Vv3 7/ 



LIST OF PLATES. 



I. (Frontispiece) Watch which belonged to Lord Nelson (Heinz 
Collection). 
II-IV. Portable Sun-dials, 1618-1720 (Heinz Collection). 
\'-XXXn. Watches contained in the Heinz Collection. 
XXXIV. Canids from the Uinta. 
XXXV. Canids from the Uinta, 
XXXVI. Ungulates from the Uinta. 
XXXVII. Artiodactyls from the Uinta. 
XXXVIII, Artiodactyls from the Uinta. 
XXXIX. Achcenodon. 

XL. Protoreodon from the Uinta. 
XL I. Restoration of Skeleton of Protoreodon medius. 
XLII. Perissodactyls from the Uinta. 
XL II I. Helaletes and Epihipp us. 
XLIV. Perissodactyls from the Uinta. 
XLV. Prothyracodon uinlense. 
XLVI. Prothyracodon from the Uinta. 
XLVII. Artiodactyls from the Uinta. 

XLVI 1 1. Pseudobagrus taiwanensis; P. adiposalis; Liobagrus nantoensis. 
XLIX. Formosania gilberti; Labeo jordani. 

L. Scaphesthes tamusuiensis; Puntius snyderi; Spinibarbus hollandi. 
LI. Recumbent spine of S. hollandi; Gnathopogon iijimce; Phoxiscus 
kikiichii. 
LII. Cidtcr aokii; Macropodus filamentosus. 
LIII. Rhinogobius tahoamis; R. formosanus; Glossogobins parvus. 
LIV. Caristius japonicus; Elepheiior macropus. 
LV. Platyberyx opalescens; PteracUs velifera. 
LVI. Bentenia cesticola. 
LVII. Pterycombus brarna; PteracUs trichipterus. 
LVI II. PteracUs ocellatus; Centra phoUs goodei. 
LIX. Cypripediuin passerinum. 
LX. Portrait of Charles Rochester Eastman. 
LXI. Portrait of Andrew Arnold Lambing. 
LXII. Portrait of Herbert Huntington Smith. 
LXIII. Portrait of Henry John Heinz. 

iv 



LIST OF FIGURES IN TEXT. 



PAGES 

D'Orsay presented to the Carnegie Museum by Messrs. Healey & Co. of 
New York City. The first vehicle lighted by electric storage- 
batteries 37 

Fig. I. Oxycenodon dysclerus 44 

2- " " 44 

3. Paramys compressidens 60 

4. Ischyrotomus gidleyi 63 

5. Sciuravus altidens 64 

6. Prosciurus matthewi 64 

7. Pareumys milleri 66 

8. Hylomeryx annectens 68 

9- " " 69 

10. " " 70 

1 1 . Achcenodon insolens 80 

12. Parahyus vagus 82 

13. Protoreodon medius. Vertebral column as found in field 85 

14. Protylopus peter soni. Left cubonavicular, etc 90 

15. Protylopus annectens 91 

16. Diagram expressing relationship of Hypertragulidae lOi 

17. Isectolophiis annectens 119 

18. Hyrachyus grande 129 

19. Hyrachyus grande; Hyrachyus princeps 129 



GENERA, SPECIES, AND VARIETIES NEW TO 

SCIENCE DESCRIBED IN THIS VOLUME, 

OR REDESCRIBED. 



Plants. 

SPERMATOPHYTA. 

Family Orchidaceae. 

Cypripedium passerinum Richardson (Redescribed) p. 343, PI. LIX. 

Fishes. 

Family Silurid.e. 

Pseudobagrus taiwayiensis Oshima p. 180, PI. XLVIII, fig. i. Formosa. 

Pseudobagriis adiposalis Oshima p. 181, PI. XLVIII, fig. 2. Formosa. 

Liobagrus nantoensis Oshima p. 183, PI. XLVIII, fig. 3. Formosa. 

Family Homalopterid.e. 
Formosania (gen. nov.) gilberti Oshima . p. 194, PI. XLIX, figs. 1-2. Formosa. 

Family Cyprinid.e. 

Labeo jordani Oshima p. 204, PI. XLIX, fig. 3. Formosa. 

Acrossocheilus (gen. nov.) Oshima, p. 206. 

(Type Gymnostomus formosanus Regan.) 
Scapheslhes (gen. nov.) tamnstliensis Oshima. ... p. 209, PI. L, fig. i. Formosa. 

Puntius snyderi Oshima p. 216, PI. L, fig. 2. Formosa. 

Spinibarbus (gen. nov.) hollandi Oshima 

* p. 218, PI. L, fig. 3; LI, fig. I. Formosa. 

Gnathopog07i iijimce Oshima p. 219, PI, LI, fig. 2. Formosa. 

Phoxiscus (gen. nov.) kikuchii Oshima p. 216, PI, LI, fig. 3. Formosa. 

Aristichthys (gen. nov.) Oshima. . . . p. 246, (Type Leiiciscus nobilis (Gray) ). 

Cutter aotiii Oshima p. 250, PI. LI I, fig. i. Formosa. 

Cultricutus (gen. nov.) Oshima, p. 252. 

(Type Cutter teuciscutus Kner, non Basilewsky = Hemicutter kneri 
Kreycnberg.) 

Family Labvrinthici 

Macropodus Jitumentosus Oshima p. 278, PI. LII, fig. 2. Formosa. 

vi 



Genera, Species and Varieties New to Science. vii 

Family Gobiid.'E. 

Rhinogohius taiwanus Oshima p. 298, PI. LI 1 1, fig. i. Formosa. 

Rhinogohhis formosanus Oshima p. 300, PI. LIII, fig. 2. Formosa. 

Clossogobitis parvus Oshima p. 305, PI. LIII, fig. 3. 

Family Elephenorid.e fam. nov. Jordan. 

(Type Elephenor macropus (Bellotti) Jordan.) 

Elephenor gen. nov. Type Caristius macropus (Bellotti) Jordan., .pp. 329-334. 

Family Pter.\clid^. 

Centropholis goodei Jordan (nom. nov. for Pteraclis carolinus Good & Bean, 
non Cuv. & Valenciennes) pp. 332, 333. 

Mammals (Fossil). 
CARNIVORA. 
Family Oxy^nid^. 
Limnocyon douglassi Peterson, .p. 45, PI. XXXIV, figs. 11-14.. Uinta Eocene. 

Family Miacid^. 

Mimicyon (gen. nov.) longipes sp. nov. Peterson. 

p. 48, PI. XXXIV, figs. 6-10. Uinta Eocene. 
Prodaphcenus (?) robustus sp. nov. Peterson. 

p. 50, PI. XXXIV, figs. 3-5. Uinta Eocene. 
Pleurocyon (gen. nov.) magniis sp. nov. Peterson 

p. 52, PI. XXXV. Uinta Eocene. 
Pleurocyon mediiis sp. nov. Peterson p. 59. Uinta Eocene. 

RODENTIA. 

Family Ischyromyid^. 

Paramys compressidens sp. nov. Peterson p. 60, fig. 3. Uinta Eocene. 

Par amy s mediiis sp. nov. Peterson. 

p. 61, PI. XXXIV, figs. 15-22. Uinta Eocene. 

Ischyrotomus gidleyi sp. nov. Petenson p. 63, fig. 4. Uinta Eocene. 

Sciuravus altidens sp. nov. Peterson p. 64, fig. 5. Uinta Eocene. 

Prosciurus (?) robustus sp. nov. Peterson p. 65. fig. 6. Uinta Eocene. 

ARTIODACTVLA. 
Subfamily HoMACODONTiN^. 
Hylomeryx (gen. nov.) annectens sp. nov. Peterson. 

pp. 67-71, PI. XXXVI, figs. 5-6, text-figs. 8-10. Uinta Eocene. 
Sphenomeryx (gen. nov.) quadricuspis sp. nov. 

p. 71, PI. XXXVII, figs. 15-16. Uinta Eocene. 



viii Genera, Species and Varieties New to Science. 

Mesomeryx (gen. nov.) granger i sp. nov. Peterson. 

p. 73, PI. XXXVII, fig. 17. Uinta Eocene. 

Family AnoplotheriiD/E. 
Diplobimops (gen. nov.) matthewi, sp. nov., Peterson. 

p. 76, PI. XXXVII. Uinta Eocene. 

Family Agriochcerid^. 

Protor odon medius sp. nov. Peterson. 

p. 82, PI. XL, figs. 1-16; pi. XLI. Uinta Eocene- 
Family Camelid.^. 

Protylopus annectens sp. nov. Peterson. 

p. 91, PI. XXXVII, fig. 14, text-fig. 15. Uinta Eocene. 

Family Hypertragulid.^. 

Leptotragulus medius sp. nov. Peterson. 

p. 94, PI. XXXVII, figs. 1-4. Uinta Eocene. 

PERISSODACTYLA. 

IsECTOLOPHiD.E fam. nov. Peterson, p. 115. 

(To include Homogalax Hay {non = Systemodon Cope) Parisectolophus 
Peterson nom. 7iov. for Iseciolophus latidens Scott & Osborn, etc.) 

Homogalaxin^ subfam. nov. Peterson, p. 116. 
(Type gen. Homogalax Hay, non = Systemodon Cope.) 
IsECTOLOPHiN.E subfam. nov Peterson, p. 116. 
Isectolophus scotti sp. nov. Peterson 

p. 120, PI. XXXIV, fig. 23. Uinta Eocene. 
Pariseclolophus nom. nov. for Isectolophus latidens Scott & Osborn, Peterson. 

p. 121. Henry's Fork, Wyoming. 
Schizo'ophodon (gen. nov.) cuspidens. sp. nov. Peterson. 

p. 122, PI. XLIV, fig. 4. Uinta Eocene. 

Family Hyracodontid.e. 

Ilyrachyus grande sp. nov. Peterson, p. 129, text-figs. 18-19. Uinta Eocene. 
Prothyracodon uintense sp. nov. Peterson. 

p. 134, PI. XXXVl, fig. i; pi. XLV; Pi. XLX'I, figs. 10-16. Uinta Eocene. 

CHALICOTHEROIDEA. 
Eomoropus annectens sp. nov. Peterson. 

p. 139, PI. XXXVI, fig. 2. Uinta Eocene. 




\ 



ERRATA ET CORRIGENDA. 



P. 64, fifth line from top, for "near base " read near top. 

P. 65, third line from top, for " P-" read P-. 

P. 69, fifteenth line from top, for " protero " read postero. 

P. 76, fifteenth line from top, for " Plate XXXVII " read Plate XXXVIII. 

P. 86, third line from bottom, for " Culbertsoni " read culbertsoni. 

P. 92, sixteenth line from top, for " a paratype of " read belonging to. 

P. 95, fourth line from top, for " A. proavus " read L. proavus. 

P. 113, third line from bottom, for " to those " read do those. 

P. 124, First footnote belongs to p. 123. 

P. 128, second line from top, for " Fig. 3 " read Fig. 6. 

P. 138, thirteenth line from top, for '' cutitalis " read cubiialis. 

P. 141, eighth line from bottom, for " 103 " read 104. 

P. 142, third line from bottom, for " type " read paratype. 

P. 144, second and fourth lines from bottom, for " paratype " read type. 

P. 148, sixth line from top, for " upper " read lower. 

P. 154, fifth to sixteenth lines from bottom, for " 30 " read 3016. 

For " Capoeta," where found, read Capoeta. 

For " Distwchodon," where found, read " DistcEchodon.'' 



Publications of the Carnegie Museum Serial No. igo 



ANNALS 



OF THE 



CARNEGIE MUSEUM 



Vol. XII. Nos. 2-4 



W. J. HOLLAND, Editor 



Published by the Authority of the 

Board of Trustees of the Carnegie Institute 

October. 1919 



PRESS OF 

THE NEW ERA PRINTING COMPANY 

LANCASTER, PA. 



/ 




ANNALS "^ __ . '^ 



CARNEGIE MUSEUM 



VOLUME Xn. NO. I. 



Editorial Notes 



Last June a proposal made by the Alabama Museum of Natural 
History to undertake a thorough exploration of the Tertiary deposits 
of Alabama and adjoining states in conjunction with the Carnegie 
Museum, was received and accepted. It is not necessary here to out- 
line the details of the arrangement entered into by the two institutions. 
The field work was, however, to be in charge of Mr. Herbert H. Smith. 
Mr. Smith is a veteran collector, who enjoys an international repu- 
tation as probably one of the ablest and most painstaking field- 
naturalists of our day. The work has been begun and most grati- 
fying results are already reported. A number of localities, which 
hitherto have not been carefully investigated, prove to be very rich in 
specimens and species, some of which no doubt are new to science. 
They represent horizons from which up to the present time but very 
scanty collections have been made. Certain of the beds in Alabama 
have been quite thoroughly worked in times past and the faunules 
contained in them are known; but other horizons have been more or 
less neglected. While not overlooking such well-known deposits as 
the Claiborne, the joint expedition of the Carnegie Museum and the 
Alabama Museum of Natural History will devote much time to the 
investigation of those deposits which have as yet been only partially 
studied. It is hoped that by thus cooperating the two institutions 
may add considerably to the knowledge gained of the faunas of this 
most interesting geological region. 

1 



2 Annals of the Carnegie Museum. 

Letters received in August from Mr. Samuel M. Klages report that 
in spite of various difficulties he has succeeded in obtaining large series 
of birds and other natural history specimens, principally insects, in 
the Mana Valley, French Guiana, which he reports to be highly inter- 
esting and to contain a very large and varied avifauna. He found the 
country immediately in proximity to Cayenne rather poor in species, 
partly because of the wholesale destruction of birds by "pot-hunters," 
but on reaching the distant interior, where the ravages of the hungry 
are less visible, he discovered a great wealth of interesting forms, and 
hopes to be able to make a representative collection. This will possess 
much scientific interest, if for no other reason, because of its topotypic 
value, it being well known to ornithologists that many species of 
South American birds originally described in the writings of French 
and Dutch naturalists were obtained in this region. In these days in 
which writers are much addicted to the erection of subspecies and 
the description of so-called local varieties, we have deemed it desirable 
that at least one of the great American museums should endeavor to 
acquire a collection of the birds of Guiana, as complete as possible by 
comparison Avith which the value of so-called subspecies may be in a 
measure tested. 



Mr. H. J. Heinz has added to the collection deposited in the Heinz 
Room a number of interesting objects coming from the imperial palace 
at Pekin. According to the information which he has received con- 
cerning them, these things are pieces of elaborately carved or decorated 
furniture which affirmed were removed at the time of the establishment 
of the Republic in China, in order that certain rooms in the palace 
might be furnished in the latest European styles. Whether this was 
the exact motive for their removal from the imperial palace or not, 
they are certainly highly interesting as illustrating Chinese art. 

The building and installation of a large and costly series of cases 
for the display of the specimens in the Heinz Room is nearing com- 
pletion. The work has heavily taxed the time and thought of the 
Director of the Museum, who for the most part prepared the drawings 
and has attended to a multiplicity of structural details involved in the 
adaptation of the cases to the room and to the uses to which they are 
to be in:t. 



Editorial. 3 

Mr. Herbert DuPuy has within recent months added to the col- 
lection of old American silverware which he has kindly loaned to the 
Museum, thirteen pieces which possess great interest. He has also 
deposited a strip of Gobelin tapestry made toward the end of the 
seventeenth century. 

The initial paper in this number of the Annals consists of a catalog 
of the collection of watches deposited by Mr. H. J. Heinz in the Car- 
negie Museum, containing one hundred specimens illustrating the 
evolution of this form of time-piece and showing a number of beautiful 
specimens of the jeweler's art displayed in the adornment of watch- 
cases. Probably the most interesting specimen from an historic 
standpoint, is the watch which belonged to Lord Nelson, the hero of 
Trafalgar, which is shown as the frontispiece of the catalog, a small 
edition of which is being separately issued. The catalog has been 
prepared jointly by Mr. Douglas Stewart, who wrote the introduction 
and prepared the list of the specimens, and by Dr. W. J. Holland who 
edited the paper and attended to the preparation of the plates which 
illustrate it, the photographs having been made by Mr. A. S. Cogge- 
shall. 



.-J. 



I. A CATALOG OF THE COLLECTION OF WATCHES BE- ? 

LONGING TO MR. H. J. HEINZ OF PITTSBURGH | 

AND DEPOSITED BY HIM IN THE | 

CARNEGIE MUSEUM. | 

• 

By Douglas Stewart, \V. J. Holland, and A. S. Coggeshall. • 

INTRODUCTION. | 

• 

The necessity of some means of marking the passage of time must • 

early have impressed the minds of even primitive peoples. It is • 

impossible to clearly trace the evolution of chrononietric instruments, • 

but the passage of the sun across the heavens at comparatively regular | 

intervals doubtless gave the first impulse to their invention. The | 

earliest form was doubtless the sun-dial, not as we know it now, but i 

simply a stake driven into the ground, which measured time by the | 

shadow which it cast. It was called by the Greeks a gnomon (yvufxuv), I 

"the one who knows." The word dial, from the Latin dies, is more I 

familiar, gnomon now being used to designate that part of the sun-dial i 

which casts the shadow. As in the case of many other inventions, • 

that of the sun-dial is attributed to the Phoenicians. The first direct i 

> 

reference to it in literature is in Isaiah XXX-VIII, 8, which, in the I 

Revised Version, says: "Behold I will cause the shadow on the steps, i 

which is gone down on the dial of Ahaz with the sun, to return back- i 

ward ten steps; So the sun returned ten steps on the dial whereon it I 

was gone down." As chronologists assign the reign of Ahaz, King of i 

Judah, to the years 742-727 B. C, some idea of the antiquity of the i 

dial is given by this passage. The earliest dial of which we have an i 

accurate description is the hemicycle, or hemisphere, of the Chaldaean • 

astronomer, Berosus. This learned man, a priest of Bel, translated the • 

standard Babylonian work on astrology and astronomy into Greek. • 

The translation, which was completed in the reign of Antiochus II, i 

about the year 250 B. C, gives a full description of the dial. i 

Another very ancient method of determining the flight of time was i 

by means of the clepsydra or water-clock. The name is derived from • 

the Greek KkeiTTeLV, to steal, and u5cop, water, and refers to the grad- • 

4 I 



ANNALS CARNEGIE MUSEUM. Vol. XII. 



Plate 




Watch which belonged to Lord Nelson (No. 76.). 



The H. J. Heinz Collection of Watches. 5 

ual "stealing away of water." Time was measured by the amount 
of water discharged from a vessel through a small aperture, the quant- 
ity discharged in a given unit of time being first determined. In the 
Athenian courts it was customary to limit the length of arguments 
by this device; yEschines (389-314 B. C.) tells us that the "first water" 
was allowed to the accuser, the "second" to the accused, and the 
"third" to the judges. Many modifications of this instrument were 
employed, and the familiar hour-glass, also known to the Asiatics long 
before the time of Christ, was practically the same instrument, in 
which sand was substituted for water. 

All of these earlier methods for measuring time were woefully inac- 
curate, so that the invention of the clock was a great step in advance. 
The name of the inventor and the date of the invention of this val- 
uable instrument are both lost in the mists of mediaeval times. In 
the year 1120 A. D. in the rules of the Monastery of Citeaux, France, 
the sacristan is charged with the duty of "adjusting the clock, so that 
it may strike and awaken the monks for matins." In the latter part 
of the same rules it is ordered "to prolong their reading until the clock 
sounds."^ The bell was an important part of these early time- 
keepers. The word "clock" itself is most probably derived from the 
Celtic word for a hell, and in the Celtic, Scandinavian, and German 
tongues still preserves its original meaning (German, glocke; Danish, 
klokke; Gaelic, clog; Welsh, clock). 

Peter Henlein or Hele of Nuremberg, a noted clock-maker, seems 
to hold the undisputed honor of inventing the watch. He was born 
in 1480 and died in 1542. His ingenuity in substituting a spring to 
take the place of the ponderous weights of the clock made the watch, 
or portable clock, a possibility. This first spring was simply a straight 
band about a pillar. In the year 1658 Hooke applied a spiral spring 
to the balance-wheel of the watch in the same year that Huyghens 
first applied the pendulum to the clock. It does not seem necessary 
to attempt a history of the mechanical improvements which followed, 
as many technical works, fully illustrated, have been published. It 
was some time, however, before the mechanical improvements kept 
pace with the perfection which was lavished upon the ornamentation 
of the case. 

The bell was retained as an essential feature of these early time- 

1 See Dom. Augustin Calmet, " Commentaire litteral sur la regie de Saint Benolt," 
Vol. I, pp. 279-280. 



6 Annals of the Carnegie Museum. 

pieces. The word watch, from the Anglo-Saxon wacan, to waken, 
gives an indication of this. 

Though originating in Germany the watch-making craft soon ob- 
tained a foothold in France, and by 1600 A. D. the manufacture of 
time-pieces was a flourishing industry in that country, having its most 
important center in the town of Blols. The skill of the metal-workers 
of the seventeenth century found full scope in the ornamentation of 
watch-cases and many beautiful examples were produced, but they 
are now exceedingly rare in collections. From France some watch- 
makers emigrated to England, and many skilled workmen were pres- 
ently engaged there in this calling. 

The earliest watches were fitted with only one hand, which indicated 
the hour. The subdivisions of the hour were roughly calculated by 
the position of this hand between the points marking the hours. The 
figures were raised and many of the earlier watches were provided 
with a knob above each numeral so that the time could be ascertained 
in the dark by feeling. Watches with two hands did not appear until 
somewhat later, and it was not until the middle of the seventeenth 
century that dials painted in enamel became the fashion for the most 
costly watches. Shortly afterwards dials of gold and silver with 
figures in relief came into vogue. 

The development of the characteristic circular shape of the watch 
of today was gradual. The earliest types were square metal boxes, 
with the figures placed on a circle of a different metal fastened to the 
face. No. 19 of this collection is a good example of this type. This 
form of watch had no protecting lid, but the resultant injury to the 
dial soon necessitated the invention of a cover or outer case. The 
first lids were of perforated metal, thus permitting the figures under- 
neath to be seen. Watches at this time were expensive and could 
only be owned by the very wealthy. The extravagance of dress in 
the Elizabethan period had its effect upon watches. Gold and other 
precious metals, crystal, tortoise-shell, and enamel were lavishly used 
in the decoration of watch-cases and the most skillful artists were 
employed to make them. As the watch-glass had not yet come into 
general use, many of the watches had double or pair-cases. The outer 
case had no connection with the watch pi-oper, but was a box in which 
the watch was carried. Many beautiful examples of these outer gold 
cases ornamented with a chased or repousse design are exhibited in 
this collection. In this connection it is well to distinguish between 



The H. J Heinz Collection of Watches. 7 

these two methods of ornament. In repousse work, the metal is 
punched from the back, producing a design in bold relief, while in 
chasing the design is cut or engraved from the front and is the more 
delicate method. As these Elizabethan watches were not carried in 
the pocket, but were suspended about the neck by a ribbon, fashioned 
into bracelets, or set in brooches, every opportunity was afforded for 
their display. Both transparent and opaque enamel were extensively 
employed and many artists, noted as workers in enamel, as well as 
lapidaries, were engaged in designing and fabricating watch-cases. 

On account of the high price of watches, the portable, but inaccurate, 
sun-dial persisted in use, and Nos. i to 6 of this collection are rare 
examples of that instrument. In Shakespeare's "As You Like It" 
Jaques's speech, "And then he drew a dial from his poke," contains a 
reference to such a pocket-dial. 

The Puritan hatred for display reveals its influence even in watch- 
cases. During the period of the Revolution and Protectorate ex- 
travagance and beauty of ornament gave way to plainness in decora- 
tion and watches were carried concealed in the pocket. The fob 
(Provincial German fit ppe — a little pocket) now made its appearance, 
and Cromwell's watch, preserved in the British Museum, is one of 
the earliest to display this adjunct. 

The mechanical skill of the Swiss shortly after this period began to 
reveal itself in the manufacture of watches, and in modern times 
Switzerland became the recognized center of the industry, many 
thousands of persons being employed in it. With the introduction 
of machinery for making the parts of watches the trade gradually 
became of great importance in the United States. The American 
working man of today is able to buy for a few dollars a much more 
accurate time-piece than could have been possessed by the wealthiest 
courtier of the Elizabethan Era. It must be confessed, however, 
that in these instruments artistic design has been very largely sacri- 
ficed to utility. 

It is rather exceptional to find the name of the maker on early 
watches, but many of the cities had a distinctive trade-mark, as an 
"N" for Nuremberg, a pineapple for Augsburg, and a bear for Berne. 
By such marks, in many cases, it is at least possible to determine the 
place of manufacture. 

In England the "Hall-mark" is a sure guide to the date. These 
marks are impressed upon watch-cases, jewelry, and plate, made from 

•|^a-»..*-*-»~»»«-»~*..*-»-*~»..*-«-«-*..c~««»~»..«»*..*..*»*..«»«..«..»..«~»~«..«..*..*..«>.*..*~«..*..*..*..«..*..*..*..*..«..«> 



8 Annals of the Carnegie Museum. 

either gold or silver, after the quality of the metal has been ascertained 
by analysis at the Assay Halls of the Goldsmith's Company of London. 
The privilege of assaying and marking precious metals was conferred 
upon this honorable body by statute in 1300 A. D. and a charter was 
granted in A. D. 1327. The powers of the company have been con- 
firmed by subsequent acts of Parliament. The standard mark of the 
London Hall is a lion passant for sterling silver and this was also the 
mark on twenty-two-carat gold until the year 1845. The present 
mark for gold is a crown and the figures, which indicate the purity of 
the metal, such as " 18" for eighteen-carats. Prior to 1821 a crowned 
leopard's head was also used as hall-mark. The exact year of 
manufacture is indicated by a letter of the alphabet which is changed 
each year on the last day of May. The standard of purity in gold in 
the earlier watch-cases was much higher than at present, as it was not 
until 1798 that a lower standard than twenty-two carats was allowed. 
In that year the use of eighteen-carat gold was permitted. In 1854 
three more standards were introduced, fifteen, twelve, and nine carats 
of gold being admitted to use, 

Mr. H. J. Heinz, who has generously deposited this collection in the 
Carnegie Museum, has used extreme care in the selection of the speci- 
mens of which it is composed. A number of years ago he became 
interested in the subject and began the formation of a collection. His 
collection is much more extensive than that displayed, but he insists 
that only the best specimens shall be exhibited. Quality is empha- 
sized rather than quantity, though the collection is not by any means 
small. So far as possible the catalog is arranged in chronological order 
and the various steps in the- development of the watch from the por- 
table dial to watches of modern times may be traced. The collection 
has great educative value, not merely from the artistic standpoint, 
but from the mechanical standpoint as well, since it is possible by its 
help to study the evolution of watches as machines. For many reasons 
the collection is worthy to be regarded as one of the most important 
in the United States. 



^. 



1 The H. J. Heinz Collection of Watches. 9 

1 CATALOG 

• 

? I. Ivory universal portable sun-dial and compass. Made by Lien- 

I hart Miller at Augsburg, Germany, in the year i6i8. (92 X 

? 54 X 14 mm.)=' 

i (Plate II, fig. I.) 



2. Ivory universal portable sun-dial and compass. Made by Hanns 
Troeschel at Augsburg, Germany, in the year 162 1, (100 X 
72 X 16 mm.) 

(Plate II, fig. 2.) 



3. Ivory universal portable sun-dial, perpetual calendar, and compass. 
Made by Ch. Blond at Dieppe, France, about the year 1660. 
(69 X 58 X 13 mm.) 

(Plate III, fig. I.) 



4. Bronze portable sun-dial. Made by Johan Schretteger at Augs- 
burg, Germany, about the year 1660. (Diameter 57 mm.; 
thickness 13 mm.) 

(Plate III, fig. 2.) 



5. Bronze octagonal portable sun-dial. Made by Piochat in Paris, 
A. D. 1710. (77 X 64 X 9 mm.) 

(PlatelV, fig. I.) 

2 As in many cases the figures given on the plates accompanying this Catalog are 
necessarily reduced, we have given the principal dimensions, diameter over all 
across the face, and greatest thickness including the crystals or covers as they now 
are. It is proper to observe that crystals when broken may be replaced by others 
of a different curvature, and that this measurement, for purposes of comparison, 
may therefore at times prove to be misleading. 



-»-^ 



10 Annals of the Carnegie Museum. 

6. Bronze octagonal portable sun-dial* and compass. Made by 
Butterfield, an Englishman, who settled in Paris in the year 
1720. This dial is almost a duplicate of the preceding, even 
to the design of the gnomon, which is a bird. (77 X 65 X ii 
mm.) 

(Plate IV, fig. 2.) 



7. Silver "Nuremberg Egg".^ An elaborately engraved silver dial 
has a brass ring superimposed, on which the figures for the 
hours are engraved. The case is pierced and engraved, the 
engraving depicting a boar hunt. Made at Augsburg, Germany, 
during the early part of the seventeenth century. (Length 63, 
breadth 41, thickness 28 mm.) 

(Plate V, fig. 2.) 



8. Gilt brass "Nuremberg Egg", signed "H." Made in the seven- 
teenth century. (Length 48, breadth 30, thickness 23 ram.) 

(Plate V, fig. 3.) 



9. Watch in gilt metal case with a perforated and engraved lid, which 
makes the dial visible without raising the lid. The dial is of 
engraved metal. The case is surrounded by a band of pierced 
and engraved silver. The watch has an alarm-bell. Made by 
Angelo Rota in Rome between the years 1590 and 1600. The 
catgut movement of this watch is of slightly later date. (Diam- 
eter 60, thickness 31 mm.) 

(Plate V, fig. I.) 

3 The "Nuremberg Egg" is a name applied to watches of this flattened, oval 
form, which were first made at Nuremberg about 1600 A. D. 



^•»« 



The H. J. Heinz Collection of Watches. H 

lo. Watch in metal case with blue and white enameled dial, catgut 
movement, and alarm-bell. Made by P. Gregoire at Blois, 
France, about the year 1620. (Diameter 52, thickness 30 mm.) 
(Plate V, fig. 4.) 



1 1. Silver watch in a pierced and engraved case. The dial is of silver 

and bears the name of Grantham, though the works were made 

by Fromanteel and Clark of London, about the year 1680. 

The case is of a later date. (Diameter 50, thickness 34 mm.) 

(Plate VI, fig. I.) 



12. Silver watch with a pierced and engraved case. Made by Daniel 
Quare, of London, about the year iGSc* (Diameter 46, thick- 
ness 31 mm.) (Plate VI, fig. 2.) 



13. A silver watch with a double case. The outer and inner cases are 
pierced and engraved. The dial is of carved silver. Made by 
Paul Luttin in London, A. D. 1690. (Diameter 55, thickness 
38 mm.) (PlateVI, fig. 3.) 



14. Watch in a case enameled both inside and out in a Louis XIII 
design. Made by Thomas Tompion in London in the year 
1690.* (Diameter 46, thickness 29 mm.) 
(Plate VI, fig. 4-) 

* Quare was a famous English watchmaker and the inventor of the repeating 
watch. 

5 Tompion (1638-1713) was known as the "Father of English watchmaking," 
and did more to bring this craft to the fore in England than any man of his time. 
He was the first maker to number watches successively and thus make positive 
identification possible. This watch is number 234. Tompion and his pupil. 
George Graham, are buried in Westminster Abbey. 



12 Annals of the Carnegie Museum. 

15. Silver watch in an outer case of tortoise-shell pique. Made by 
Henry Jones, of London. Jones died in the year 1693.' 
(Diameter 53, thickness 30 mm.) 

(Plate VII, fig. I.) 



16. Silver watch with a pierced and engraved case, white enameled | 

dial. Contains an alarm-bell. Made by Duquesne in the | 

seventeenth century. (Diameter 60, thickness 31 mm.) | 

(Plate VII, fig. 2.) j 



17. Silver watch with a pierced and engraved case. An engraved 
metal dial with an ornamental brass border. Curious cover 
for winding holes, in the form of a mask. Made by Millegg in 
Vienna in the seventeenth century. (Diameter 53, thickness 
37 mm.) 

(Plate VII, fig. 3-) 



18. Silver watch contained in an outer protecting case. Highly deco- 
rated movement. A semicircular piece is removed from the 
upper half of the dial and through it is seen one-half of a disc 
which rotates underneath once in twenty-four hours. On one- 
half of the disc is a golden sun, one of the rays of which points to 
the hour from 6 A. M. to 6 P. M., and on the other half a silver 
moon performs the same service from 6 P. M. to 6 A. M. The 
figures are all on the upper half of the dial. The minutes are 
indicated by a hand in the usual way. Made by Paul Bramer, 
of Amsterdam, about the year 1700. (Diameter 58, thickness 
31 mm.) 

(Plate VIII, fig. 3.) 



' Henry Jones was a noted English clock- and watch-maker and the first English- 
man to construct a Torricellian tube, as the barometer was first called. 

The word pique designates a form of decoration made by driving silver pins through 
the outer surface of the case in a conventional design. 



The H. J. Heinz Collection of Watches. 13 

19. Square iron watch, ornamented with brass and cut iron. This is 
one of the earliest types of watches and was made by Johann 
Sigmund Schloer, at Regensburg, (?) Bavaria,'' probably in the 
seventeenth century. (Diameter 38, thickness 19 mm.) 

(Plate \TII, fig. 2.) 



20. Chatelaine of chased gold on grey agate. Period of Louis XV 
(1710-1774). (120 X 60 X 4 mm.) 

(Plate VIII, fig. I.) 



21. Gold and enamel watch. Miniature of a lady on the back, sur- 
rounded by a border of paste. Made by Esquivillon Freres 
and DeChoudens of Paris. Period of Louis XV (1710-1774)- 
(Diameter 36, thickness 20 mm.) 

(Plate IX, fig. 2.) 



22. Silver watch in a pierced and engraved case, containing an alarm- 
bell. White enameled dial. Made by Tasinon at Tournay, 
. France. Period of Louis XV (1710-1774). (Diameter 62, 
thickness 37 mm.) 

(Plate IX, fig. I.) 



23. Silver repeating watch with a pierced and engraved case; enameled 
dial. Made by Jean Fardoit in Paris in the early eighteenth 
century. (Diameter 52, thickness 31 mm.) 

(Plate IX, fig. 4.) 

' The inscription seems to be "Regenb," which we take to be an abbreviation for 
Regensburg. There is a small village named Regen in Bavaria, but it is more likely 
that Schloer was located in Ratisbon (Regensburg). 



I 
T 



.•-•-•-•-•^•^•■••-•-••••••••••-•-•-■-•••••••••••••-••••••••■•••••••••••••■••-•-••••-••••■••-••■•••••••••••••••••••■•••.••••••••••-••^ 

I 

• 
t 

14 Annals of the Carnegie Museum. T 

s 

I 

24. Brass watch with a white metal dial. The movement was made • 

by Gaudron in Paris (1710-1730), though the case is of a later • 

• 
date. Made for the Turkish market. (Diaineter 52, thickness • 

. 32 mm.) • 

(Plate IX, fig. 3.) I 



25. Silver double case watch. The outer case is in a repousse de- 
sign; signed by D. Cochin, a celebrated French silver worker 
(1660-1680). The movement was made at a later date (1718) 
by George Graham, of London.* (Diameter 55, thickness 31 
mm.) 

(Plate X, fig. 2.) 



26. Watch with a crystal case held by a silver gilt rim; chased gold 
dial, upon which are superimposed white plaques with black 
enameled figures. Green enamel decoration on dial. Made 
for the Turkish market by Julien LeRoy of Paris, about the 
year 1700.^ (Diameter 49, thickness 29 mm.) 

(Plate VII, fig. 4.) 



27. Gold watch with double case; white enameled dial ornamented 
with a gold and green enamel plaque. The outer case is of 
dark blue opaque enamel with a floral design in silver gilt 
applique. Made by Julien LeRoy in Paris for the Turkish 
market. Early eighteenth century. (Diameter 57, thickness 
28 mm.) 

(Plate X, fig. I.) 

* George Graham, known as "Honest George Graham," was one of the most 
eminent watchmakers of his time (1673-1751). A pupil of Thomas Tompion (see 
No. 14), he was the only other watchmaker to be honored by burial in Westminster 
Abbey. 

^ Julien Le Roy was a scientific watchmaker born in 1686 and died in 1759. He 
devised a form of repeating mechanism, which was soon substituted for tlie bell in 
use before. 



^•-•«« 



The H. J. Heinz Collection of Watches. 15 

28. Gold watch set with a beautiful enamel of a lady watering flowers 
surrounded by a border of blue enamel and vari-colored gold. 
Made by Esquivillon Freres and De Choudens in Paris. Period 
of Louis XY (1710-1774). (Diameter 53, thickness 22 mm.) 

(Plate X, fig. 3.) 



29. Double case silver watch. The back of the outer case covered 
with tortoise-shell upon which is painted "Peter and the Cock." 
Made by Charles Meadows in London In the year 1719. (Diam- 
eter 65, thickness 31 mm.) 

(Plate X, fig. 4.) 



30. Silver watch with an engraved gold border and silver dial. Made 
in Paris by L'Epine, about the year 1720. L'Epine was the 
father of Jean Antoine L'Epine, Court watchmaker to Louis 
XV of France. (Diameter 54, thickness li mm.) 

(Plate XIII, fig. I.) 



31. Silver watch with an outer case covered by leather with pique 
decoration. Silver dial. Made by Hilgerus Vogel at Cologne 
in the year 1720. (Diameter 57, thickness 33 mm.) 

(Plate XIII, fig. 4.) 



32. Silver watch with a silver dial. Outer case of tortoise-shell pique. 
Made by Francois Mercier, Paris, A. D. 1725. (Diameter 53, 
thickness 30 mm.) 



(Plate XII, fig. 5.) /V}0\^^^ 



\- 



16 Annals of the Carnegie Museum. 

33. Silver watch, four inches in diameter, with an outer protecting 
case covered with black tooled leather, silver mounted, the 
inner case beautifully pierced, engraved, and ornamented with 
silver gilt. The watch contains a bell upon which the hours 
and quarters are repeated, when the cord, suspended from the 
bottom of the case, is pulled. It also contains an alarm and 
indicates the days of the month. Made by Frantz Roth at 
Augsburg, Germany, about the year 1730. (Diameter 89, 
thickness 51 mm.) 

(Plate XI.) 



34. Double case silver watch. Both the Quter and inner cases elabor- 
ately pierced and engraved. Repeats the hours and quarters. 
Made by Andreas Colling at Augsburg, Germany, about the 
year 1730. (Diameter 53, thickness 29 mm.) 

(Plate XIII, fig. 2.) 



35. Gold repeating watch in a pierced and engraved case. Made by 
Charles Cabrier in London, in the year 1730. (Diameter 45, 
thickness 25 mm.) 

(plate XII, fig. I.) 



36. Enameled watch set with zircons; the back of beautiful enamel in 
two shades of blue with gold decorations; a fine pair of jewelled 
hands. The outer case is ornamented with bands of tortoise- 
shell with a glass protecting both front and back. Made by 
Phillip Terrot at Genoa, about the year 1730. (Diameter 58, 
thickness 25 mm.) 

(Plate XII, fig. 4.) 



t ? 

• • 

I The H. T. Heinz Collection of Watches. 17 ? 

? • -^ T 

• • 

• . . i 

• 37. Gold watch pierced and engraved. The outer case is ornamented • 
? . .... • 

• with a repousse design, depicting a warrior leaving for battle. • 

• _ _ • 

• It repeats the hours and quarters. Made by William Hawes, • 

• • 

• London, A. D. 1741. (Diameter 49, thickness 30 mm.) 4 

? (Plate XII, fig. 2.) ! 



Gold quarter repeating watch in a pierced and engraved case. 
Made by Benjamin Gray and Justin VuUiamy, London, A. D. 
1746.^° (Diameter 40, thickness 23 mm.) 
(Plate XII, fig. 3.) 



39. Gold double-case watch, the outer case pierced and decorated with 
a repousse design. Black enameled figures are super-imposed 
upon an engine-turned dial. Made by James Blackborow, 
London, A. D. 1734-1746. (Diameter 50, thickness 28 mm.) 
(Plate XIII, fig. 3.) 



40. Large silver watch designed for use in a sedan-chair. Case beauti- 
fully pierced and engraved. Black enameled figures on a silver 
dial; repeats and contains an alarm. Made by Antonius 
Heckel, Vienna, A. D. 1750. (Diameter 96, thickness 55 mm.) 
(Plate XIV.) 



41. Gold pair-case watch, decorated with a pierced, embossed, and 
engraved design. Repeating movement. Made by Beuster 
of London in the year 1750. (Diameter 48, thickness 29 mm.) 
(Plate XV, fig. 4.) 

1" Benjamin Gray was watchmaker to King George II. Justin Vulliamy emi- 
grated from Switzerland about 1730. He worked for Gray, married his daughter, 
and eventually became a partner in the business. Watches bearing the signature 
of these two men are invariably of exceptional quality. 



18 Annals of the Carnegie Museum. 

42. Pendant watch with copper case and enameled dial. Raised 
inscription "M. Dieu." Made in Rome by J. Clovi in the 
year 1741, though the case is dated 1751. (Height 77, diam- 
eter 53 mm.) 

(Plate XXIV, fig. I.) 



43. Silver watch with an outer case of opaque enamel. On the back 
of the case a man is depicted playing upon a pipe to a Watteau- 
like sheperdess, with Cupid in the background. Made by 
P. Charleson in London, A. D. 1758. (Diameter 54, thickness 
28 mm.) 

(Plate XXIII, fig. 3.) 



44. Gold watch, repousse case with an outer protecting case of glass 
and shagreen.^^ Made by Thomas White, of London, A. D. 
1759' (Diameter 52, thickness 25 mm.) 

(Plate XV, fig. 3-) 



45. Gold double-case watch; the outer case ornamented with a re- 
pousse design: "Maidens bringing offerings to Apollo." Made 
by Samuel Weldon, ol London, in the year 1759. (Diameter 
50, thickness 27 mm.) 



(Plate XVI, fig. I.) 



'1 Shagreen was a popular covering for watch-cases during the eighteenth century. 
True shagreen is a remarkably tough kind of leather made chiefly in Astrachan from 
the skin of the ass or horse. The peculiar roughness is produced by treading into 
the dampened skin, hard, round seeds which are shaken out when the skin is dry. 
The skin is then stained green in a solution of copper filings and salammoniac, 
dried a second time, and rubbed down. The depressions left by the seeds produce 
the characteristic spotted appearance. 



'•»-l> 



The H. J. Heinz Collection of Watches. 19 

46. Watch of vari-colored gold. The back ornamented by a design 
in quatre couleur. Made by Gudin in Paris. Period of Louis 
XV, about the year 1760. (Diameter 40, thickness 17 mm.) 

(Plate XVI, fig. 2.) 



47. Pair-case watch, the outer case ornamented with a marine view 
in Battersea enamel.^- The inner case is of gold with a cipher 
C. C. engraved on the back, surmounted by a unicorn head. 
Signed G. C. Made in London, about the year 1760. (Diam- 
eter 58, thickness 25 mm.) 

(Plate XVL fig. 3-) 



Gilt metal watch with an engraved silver dial. The movement 
is by Daniel Torin of London, about the year 1760. The 
elaborate case is of French workmanship and was made for 
the Turkish market. At the back of the case a gold plaque 
depicting a boy playing with a bird is placed upon a painted 
background, while above the head a bird, controlled by the 
movement, flies back and forth. (Diameter 58, thickness 32 
mm.) 

(Plate XV, fig. I.) 



49. Silver verge ^^ watch contained in an outer case of tortoise-shell 
pique. Made in London by John Wilter (1760-1784) for the- 
Dutch market. (Diameter 54, thickness 32 mm.) 

(Plate XV, fig. 2.) 

12 Battersea enamel dates from the year 1750, when Sir Theodore Janssen estab- 
lished a manufactory at York House, Battersea. 

13 The verge is the pin upon which are placed the plates which regulate the speed 
of the escape-wheel. This method was first used in the clock and later modified 
for use in the watch. 



20 Annals of the Carnegie Museum. 

50. Watch of vari-colored gold. The mechanism is visible through a 
crystal plate at the back of the case; beautifully made hands 
set with jewels. Made by vS. Triboulet, Geneva, A. D. 1760. 
(Diameter 45, thickness 23 mm.) 

(Plate XVII, fig. I.) 



51. Gold pair-case watch; the outer case repousse, classical design, 
with an ornamental border of green, blue, and white enamel. 
The dial is elaborately carved and discloses a calendar. Dutch 
manufacture, though signed "Smiht (sic) London." Made 
about 1760. (Diameter 50, thickness 25 mm.) 

(Plate XVH, fig. 2.) 



52. Pinchbeck^* watch; pierced and engraved case. Made by John 
Cock, of London, about the year 1760. (Diameter 47, thick- 
ness 33 mm.) 

(Plate XVII, fig. 3.) 



53. Silver watch with a pierced and engraved dial. The back of the 
case is pierced and ornamented with five repousse scenes from 
the legend of William Tell. Designed for use in a sedan-chair. 
Made by Philiph Votter, of Vienna, in the year 1763. (Diam- 
eter 95, thickness 55 mm.) 

(Plate XVIII.) 

" Pinchbeck was a favorite material for the manufacture of watch-cases. It is 
composed of an alloy, three parts zinc to four of copper. It took its name from the 
inventor, Charles Pinchbeck (1670-1732). The secret of this composition was 
jealously guarded by the inventor and his descendants for manj' years. 



.^ 



The H. J. Heinz Collection of Watches. 21 

54. Large silver- gilt double-case watch, both cases pierced and en- 
graved. In the outer case a large enamel "Mother watching 
a sleeping child" is set. The watch is provided with hour-, 
minute-, second-, and split-second-hands. Two tunes may be 
played upon a concealed musical attachment. Made by 
Timothy Williamson of Fleet Street, London, in the year 1769. 
Looted from the Chinese Imperial Palace in Pekin during the 
"Boxer Rebellion." (Diameter 167, thickness 60 mm.) 

(Plate XIX.) 



55. Gold pair-case repeating watch. The inner case is pierced and 
engraved; the outer case with a pierced and repousse design. 
Made by Hughes of London in the year 1770. (Diameter 48, 
thickness 25 mm.) 

(Plate XVII, fig. 4.) 



Gold watch of the Louis XVI period. The back is of gros bleu 
enamel with a superimposed floral design in platinum set with 
zircons, surrounded by a filigree gold border set with milk opals. 
French manufacture of about the year 1770. (Diameter 59, 
thickness 24 mm.) 

(Plate XXII, fig. I.) 



57. Pair-case gold watch. The inner case is without ornament. 
The outer case is ornamented with a repousse design. Made 
by Harry Potter in London, A. D. 1773. (Diameter 52, thick- 
ness 25 mm.) 

(Plate XX, fig. 2.) 



22 Annals of the Carnegie Museum. 

58. Gold watch with an outer protecting case of painted tortoise- 
shell. The back medallion depicts a maiden and Cupid. 
Made by N. Scott of Maidstone, England, about the year 1775. 
(Diameter 51, thickness 24 mm.) 

(Plate XX, fig. 3.) 



59. Gold watch with a beautifully enameled case depicting a lady and 
gentleman watching children at play in a park. Made by 
Paul Bannicker at Paris during the Louis XVI period. Upon 
the face of the watch a wreath of flowers terminating in clasped 
hands is painted, and the inscription " Maries Le 3 Juillet, 
1 775'" (Diameter 57, thickness 21 mm.) 
(Plate XX, fig. 4.) 



60. Pair-case gold watch. The outer case is ornamented with a 
repousse design. Made by James Noakes, of London, A. D. 
1776-1794. (Diameter 47, thickness 26 mm.) 
(Plate XXn, fig. 2.) 



61. A large watch in a pierced and engraved pinchbeck case; a border 
of imitation pearls and rubies at both sides of the case; musical 
attachment. Made by George Margetts in London, A. D. 
1779. (Diameter 90, thickness 42 mm.) 

(Plate XXL) 



62. Watch with a case of opaque enamel, the back decorated with a 
picture of a woman with a pot of flowers. Set with diamonds. 
Each side of the case is surrounded with a border of diamonds. 
Made in Switzerland by J. Fazy et fils, A. D. 1780-1785. (Di- 
ameter 49, thickness 19 mm.) 

(Plate XVI, fig. 4.) 



^. 



The H. J. Heinz Collection of Watches. 23 

63. Gold watch with the dial surrounded by a border of milk-opals 
set in dark blue enamel. The back is also surrounded by opals 
and is of beautiful translucent mauve enamel on a chased sur- 
face. A wreath of small diamonds, set in platinum, is cemented 
on the center of the back. Made by Joseph Martineau & Son, 
St. Martin's Court, London, in the year 1784. (Diameter 49, 
thickness 22 mm.) 

(Plate XXII, fig. 4.) 



64. Gold watch with the back of the case ornamented with an enamel 
picture of a woman and child with pet bird; surrounded with a 
border of blue and white enamel. Made at Paris by Pierre 
Gregson, Watchmaker to Louis XYI of France, A. D. 1780- 
1790. (Diameter 51, thickness 17 mm.) 

(Plate XXII, 'fig. 3.) 



65. Gold watch, surrounded front and back with a border of paste. 
In the back a single carbuncle is set, surrounding which is a 
white enamel circle with the inscription in black: " L'amitie 
fait le charme de la vie." Border is of rose diamonds and green 
enamel. Period of Louis X\T, A. D. 1774-1792. (Diameter 
35, thickness 18 mm.) 

(Plate XV, fig. 5.) 



66. Watch in a vari-colored gold case set with turquoise and garnets; 
gold dial with black enameled figures. Made by Pierre LeRoy, 
Paris, about the year 1780. (Diameter 38, thickness 14 mm.) 

(Plate XX, fig. I.) 



24 Annals of the Carnegie Museum. 

67. Silver watch upon the dial of which are two mechanical figures, 
which seem to strike the hours and quarters, though this is 
done by a concealed bell. Made at Geneva, Switzerland, by 
George Achard et fils, about the year 1780. (Diameter 54, 
thickness 25 mm.) 

(Plate XXVII, fig. I.) 



Gilt metal watch, the back enameled on copper and studded with 
jargons.^^ Made by Leonard Bordier at Geneva, Switzerland, 
about the year 1785. (Diameter 56, thickness 23 mm.) 

(Plate XXIII, fig. 4.) 



69. A pinchbeck watch with double case; the inner case pierced and 
engraved. The outer case is also pierced, engraved, and set 
with an enamel portrait of a lady. A pinchbeck chain with a 
topaz seal attached accompanies the watch. Made by Frazer 
of London in the year 1785. (Diameter 63, thickness 34 mm.) 

(Plate XXXI, fig. I.) 



70. Silver verge watch with an outer case decorated with a painting 
of a man bearing a gun, and accompanied by two dogs; the 
outer case protected by a covering of transparent horn. Made 
by Philip Phillips, of London, in the year 1776. (Diameter 60, 
thickness 26 mm.) 

(Plate XXIII, fig. I.) 

1' Jargon or jargoon, is the name given in Ceylon to the colorless or yellowish 
zircons, which are also sometimes called "Matura diamonds," because found in the 
province of Matura. Zircon is a mineral slightly harder than quartz and is a 
silicate of zirconium, though the name is also sometimes given to silicates of tin. 



The H. J. Heinz Collection of Watches. 25 

71. Watch contained in a bronze case ornamented with a conventional 
design of bronze pins piqu6. Blue and white enamel dial. 
Made by Tayot, of Paris, in the eighteenth century. (Diam- 
eter 59, thickness 42 mm.) 

(Plate XXIII, fig. 2.) 



72. Gold pair-case watch. The movement is by George Graham of 
London (1695-1751), though the case is of much later date, 
being Hall-marked, "London 1789." It is probable that the 
works have been taken from an earlier piece. The inner case 
is pierced and engraved, while the outer case contains a well- 
executed enamel of a lady with parrot. (Diameter 51, thick- 
ness 29 mm.) 

(Plate XXIV, fig. 3.) 



73. Gold watch with back of translucent blue enamel surrounded by a 
circle of pearls both front and back. Attached to the watch 
is an elaborate fob of gold, enamel, and pearls. Made in London 
by Jeffrys and Jones, A. D. 1 769-1 794. (Diameter 46, thick- 
ness 15 mm.) 

(Plate XXV.) 



74. Gold watch, the dial surrounded by a border of blue and white 
opaque enamel. The back is of royal blue translucent enamel, 
upon which is cemented a monogram in fine diamonds. Made 
by Rundle and Bridge of London, watchmakers to King George 
III of England, about the year 1790. (Diameter 49, thickness 
19 mm.) 

(Plate XXIV, fig. 2.) 



26 Annals of the Carnegie Museum. 

75. Large gold watch, three and one-half inches in diameter. Open 
face showing split-second-hand. Made by Bouvier Freres 
at Geneva in the year 1790. (Diameter 80, thickness 27 mm.) 

(Plate XXVI, fig. 3.) 



76. Gold repeating watch contained in an outer case of brown and 
white opaque enamel attached to a chatelaine of the same ma- 
terial. The movement was made by Peter Mackdonald in 
London in the years 1 790-1 794. This watch was formerly the 
property of Admiral, Lord Nelson, the Hero of Trafalgar. On 
the back of the watch an N is engraved, surmounted by a 
coronet, and also the letter B. The N is for Nelson and the B 
for Bronte. The Neapolitan title of Duke of Bronte was 
granted to Nelson in 1799. This watch was made some years 
before being presented to Nelson, the date of manufacture 
antedating the conferment of the title "Duke of Bronte." 
(Diameter 50, thickness 22 mm.) 

(Plate L Frontispiece.) 



77. Pinchbeck watch contained in an outer case of shagreen. Made 
by John Morier in London. Morier was admitted to the 
Clockmakers' Company in the year 1799.^^ (Diameter 46, 
thickness 22 mm.) 

(Plate XXVL fig. I.) 



78. Open-face pinchbeck watch; surround,ed on each side by a border 
of paste. The back is ornamented with an enamel. French 
or Swiss manufacture of the late eighteenth century. (Diam- 
eter 51, thickness 22 mm.) 

(Plate XXVn, fig. 3-) 

1* The shagreen cover on this watch is the best example of that material in the 
collection. 



The H. J. Heinz Collection of Watches. 27 

79. Watch of vari-colored gold, set with garnets and demantoids.^^ 
French or Swiss manufacture of the early nineteenth century. 
(Diameter 35, thickness 16 mm.) 

(Plate XXVII, fig. 2.) 



80, Watch in case of vari-colored gold set with garnets and turquoise. 
Made by Luigi Duchene et fils in Switzerland in the late eigh- 
teenth or early nineteenth century. (Diameter 40, thickness 

15 mm.) 

(PlateXXVII, fig. 4.) 



81. Gold open-face watch with a silver dial. The back is elaborately 
enameled in black, blue, gray, and yellow. Swiss manufacture 
of the early nineteenth century. (Diameter 41, thickness 12 
mm.) 

(Plate XXIX. fig. 2.) 



82. Gold watch surrounded on the front by a border of pearls. The 
back also has a border of pearls around the outer edge upon a 
base of blue enamel; a picture framed in pearls is cleverly 
executed in gold, seed-pearls, and enamel. Made by Patry 
and Chaudoir at Geneva, Switzerland, about the year 1800. 
(Diameter 52, thickness 22 mm.) 

(Plate XXVI, fig. 2.) 

1' Demantoids are green garnets, the word "garnet" being in jewelers' parlance 
restricted to the red variety. 



28 ' Annals of the Carnegie Museum. 

83. Silver watch with various dials showing the hours and minutes, 
the day of the month, a compass and a curious painting of 
Adam and Eve, about which a serpent revolves to mark the 
seconds. French manufacture of the early nineteenth cen- 
tury.^* (Diameter 60, thickness 19 mm.) 

(Plate XXIX, fig. 3.) 



Chatelaine of gold and enamel with a circular enamel picture of 
two children and a bird upon its nest. French manufacture 
of the period of the First Empire. (62 X 40 X 3 mm.) 

(PlateXXVIII, fig. 3.) 



85. Gold watch with a silver dial upon which a miniature gold clock 
with a pendulum is shown. Made by Perrin Freres at Paris. 
Period of First Empire. (Diameter 56, thickness 17 mm.) 

(Plate XXVIII, fig. I.) 



I 86. Gold mechanical repeating watch. The dial is cut out in the 4 

I center disclosing two seated figures of gold, which seem to I 

I strike the quarters, while a movable figure at the top ad- | 

I vances to strike the hours. In reality the striking is accom- I 

I plished by a concealed mechanism. French or Swiss manu- i 

I facture of the early nineteenth century. (Diameter 47, thick- i 

i ness 20 mm.) i . 

{ (PlateXXX, fig. I.) ? 

i — i 

i ISA similar watch, number 291, is in the collection of Mrs. George A. Hearn at » 

• the Metropolitan Museum, New York City. t 

• • 






The H. J. Heinz Collection of Watches. 29 

87, Gold mechanical musical watch. The face is painted with a 
woodland scene with a village in the distance. Beautifully 
made figures of vari-colored gold perform when the mechanism 
is set in motion. A maiden and youth dance to the music, a 
woman grinds an organ, while a seated boy industriously beats 
upon a kettledrum. French or Swiss manufacture of the early 
nineteenth century. (Diameter 57, thickness 20 mm.) 

(Plate XXX, fig. 3.) 



Gold open-face watch and white dial with black figures. The 
back has a scene painted to represent a kitchen. Figures of 
vari-colored gold and silver are set in motion by the mechanism 
concealed in the watch. A woman turns a spinning-wheel, a 
little girl stirs some food in a bowl, while the spit, upon which 
a fowl is roasting, is turned by a dog in a treadmill. French 
or Swiss manufacture. Early nineteenth century. (Diameter 
56, thickness 27 mm.) 

(Plate XXIX, fig. I.) 



89. Gold open-face watch with a border of small zircons. The back 
is set with an enamel of a lady listening to a youth playing 
upon a pipe. Made by Racine at Paris about the year 1800. 
(Diameter 51, thickness 19 mm.) 

(Plate XXVIII, fig. 2.) 






30 



Annals of the Carnegie Museum. 



90. Silver watch with plain open face, made by Abraham Louis Bro- 
guet (1746-1823) at Paris. In the back is set a repousse 
design by D. Cochin, depicting Esther at the Court of Ahas- 
uerus.'* (Diameter 54, thickness 20 mm.) 

(Plate XXX, fig. 2.) 



91. Gilt metal watch. Surrounded front and back by a border of 
paste. In the back an enamel of a young girl with flowers is 
set. Made by Abraham Louis Breguet, Paris, A. D. 1 746-1 823. 
(Diameter 49, thickness 23 mm.) 

(Plate XXXIIL fig. 3.) 



92. Silver open-face watch surrounded by two gold bands. Made by 
Abraham Louis Breguet, Paris, A. D. 1746-1823.2° (Diameter 
62, thickness 14 mm.) 

(Plate XXXL fig. 2.) 



93. Gold watch with silver dial; enameled on the back. Made by 
Abraham Vacheron Girod in the early part of the nineteenth 
century. (Diameter 43, thickness 16 mm.) 

(Plate XXIX, fig. 4.) 

" Abraham Louis Breguet was the most famous watchmaker of his time, a man 
of rare attainments and inventive genius. The silver back of this watch has been 
taken from a much eariier piece, as Cochin was a famous silversmith who worked 
between the years 1660 and 1680. 

2« Many of the watches made by this workman were contained in exceedingly 
plain cases as he was always willing to conform to the wishes of his customers in 
regard to cases. 



^«. 



The H. J. Heinz Collection of Watches. 31 

94. Silver watch in a case without ornament; white enamel dial with 
twenty-four figures. The hours from seven A. M. to six P. M. 
are indicated by red Arabic numerals, while those from seven 
P. M. to six A. M. are indicated by black Roman numerals. 
Made by Louis Antoine Breguet, a son of the famous Breguet, 
in Paris in the year 1830. (Diameter 60, thickness 23 mm.) 

(Plate XXXII, fig. 3.) 



95. Gold engine-turned watch. Surrounded by a border of white 
enamel. Made by Louis Antoine Breguet, in Paris about the 
year 1830. (Diameter 55, thickness 13 mm.) 

(Plate XXXII, fig. I.) 



96. Oval watch with a case enameled inside and out. The outer 
picture represents Venus disarming Cupid. Viennese work of 
the nineteenth century. (Long axis 48, shorter axis 40, thick- 
ness 24 mm.) 

(Plate XXX, fig. 4.) 



97. Watch set in a case carved from a single carbuncle, in the shape of 
a cockle shell. The dial is elaborately decorated in gold and 
enamel set with diamonds, covered by rock crystal. Made by 
Froment-Meurice. French, nineteenth century. (Height 46, 
width 53, thickness 24 mm.) 

(Plate XXXIII, fig. 4-) 



32 Annals of the Carnegie Museum. 

98. Blue translucent enamel box containing a miniature watch, 
ornamented with pearls. From a lid of vari-colored gold a 
small feathered bird emerges and moves its head and wings, 
while a bird-song is produced by a mechanism concealed within 
the box. The box also contains a musical attachment. French 
or Swiss manufacture of the nineteenth century. (95 X 60 X 
26 mm.) 

(Plate XXXI, fig. 3.) 



99. Watch with a silver dial and gold case ornamented with Niello 
enamel in black and white. Made by F. Pernetti in Switzer- 
land, in the nineteenth century. (Diameter 46, thickness 14 

mm.) 

(Plate XXXII, fig. 2.) 



100. Watch completely studded with pearls, with the exception of 
the dial, which is of vari-colored gold. The figures are set with 
seed-pearls, and the key is studded with pearls with a small 
garnet in the center. Made by Bautte and Moynier at Geneva, 
Switzerland, between the years 1820 and 1825. (Diameter 
45, thickness 12 mm.) 

(Plate XXXIII, figs, i, 2.) 






ANNALS CARNEGIE MUSEUM, Vol. XII 



Platf" 





)f;^.-^t 







I. Portable sun-dial, A. D. 1618. (No. i.) 2. Suii-dial and compass, 1621. (No. 2.) 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate III 



jf fi^irL't ^T I ,.'. ,f ' i 





I Sun-dial, calendar, and compass, A. D. 1660. (No. 3.) 
2. Bronze portable sun-dial, A. D. 1660. (No. 4.) 



ANNALS CARNEGIE MUSEUM, Voi. XII. 



Plato IV 





I. Bronze sun-dial, A. D.1710. (No. 5.) 2. Bronze sun-dial, A. D .1720. (N0.6.) 



ANNALS CARNEGIE MUSEUM, Vol. XII 



Plate V. 




I. Italian watch, A. D. 1590-1600. (No. 9.) 2. "Nuremberg Egg." (No. 7.) 

3. "Nuremberg Egg." (No. 8.) 4. Watch made at Blois, France, cjVco 1620. (No. 10.) 



ANNALS CARNEGIE MUSEUM, Vol XII. 



Plate VI. 




I. English watch, circa 1680. (No. 11.) 2. Watch by Quare, London, 1680. (No. 12.) 
3. Watch by Luttin, London, 1690. (No. 13.) 4. Watch by Tompion, 1690. (No. 14.) 



ANNALS CARNEGIE MUSEUM. Vol, XII, 



Plate VII. 




1. Watch in case of tortoise-shell pi(iue, 1693. (No. 15.) 

2. Watch by Duquesne, seventeenth century. (No. t6.) 

3. Watch by Millegg, Vienna, seventeenth century. (No. 17.) 

4. Watch by LeRoy, Paris, circa 1700. (No. 26.) 



ANNALS CARNEGIE MUSEUM, Vol. XII 



Plato VIII. 




1. Chatelaine, gold on agate, period of Louis XV. (No. 20.) 

2. Early type of watch by J. S. Schloer, Regensburg. (No. 19.) 

3. Watch by Briimer, Amsterdam, circa 1700. (No. 18.) 



ANNALS CARNEGIE MUSEUM, Vol. XII 



Plate IX. 




1. Watch by Tasinon, Tournay, France, period of Louis XV. i .NO. _>j. i 

2. Gold and enamel watch, by Esquivillon Freres, eighteenth century. (No. 21.) 

3. Brass watch for Turkish market, Paris, eighteenth century. (No. 24.) 

4. Silver repeating watch, by Fardoit, Paris, early eighteenth century. (No. 23.) 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate X 




1. Gold watch by Le Roy, Paris, early eighteenth century. (No. 

2. Silver watch, case by D. Cochin, Paris, 1660-1680. (No. 25.) 

3. Enamelled gold watch, Paris, period of Louis XV. (No. 28.) 

4. Silver watch by Meadows, London, A. D. 1719- (No. 29.) 



ANNALS CARNEGIE MUSEUM, Vol. XII 



Plate XI 




Watch by Frantz Roth, Augsburg, circa A. D. 1730. (No. 33.) 



ANNALS CARNEGIE MUSEUM, Vol. XI 



Plate XII 




1. Gold repeating watch by Cabrier, London, 1730. (No. 35.) 

2. Gold watch by William Hawes, London, 1741. (No. 37.) 

3. Gold watch by Gray and Vulliamy, London, 1746. (No. 38.) 

4. Watch by P. Terrot, Genoa, 1730. Enamelled, set with zircons. (No. 36.) 

5. Watch by F. Mercier, Paris, 1725. Outer case tortoise-shell pique. (No. 32. 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XIII 




1. Watch by L'Epine, Paris, 1720, gold border, silver dial. (No. 30.) 

2. Watch by Andreas Colling, Augsburg, circa 1730. (No. 34) 

3. Watch by James Blackborow, London, 1 734-1746. (No. 39.) 

4. Silver watch with outer leather cover, by Vogel, Cologne, 1720. (No. 31.) 



ANNALS CARNEGIE MUSEUM, Vol, XII. 



Plate XIV. 














Large watch for use in sedan-chair, by Heckel, Vienna, A. D. 1750. (No. 40.) 



ANNALS .CARNEGIE MUSEUM, Vol. XII. 



Platn XV. 




1. Gilt watch and key, London, circa 1760. (No. 48.) 

2. Silver verge watch by John Wilter, London, A. 1). 1 760-1 784. (No. 49.) 

3. Gold watch by Thomas White, London, A. D. 1759. (No. 44.) 

4. Gold repeating watch by Beuster, London, A. D. 1750. (No. 41.) 

5. Gold watch with carbuncle in back, period of Louis XVL (No. 65.) 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XVI. 




1. Gold watch by Samuel Weldon, London, A. D. 1759. (No. 45.) 

2. Watch of vari-colored gold by Gudin, Paris, circa 1760. (No. 46.) 

3. Watch with outer case of Battersea Enamel, London, circa 1760. (No. 47.) 

4. Enamelled watch set with diamonds, Swiss, Fazy et fils, 1780-85. (No. 62.) 



ANNALS CARNEGIE MUSEUM, Vol. X( 



Plat.. XVII. 




1. Watch by Triboulet, Geneva, A. D. 1760. (No. 50.) 

2. Gold pair-case watch, Dutch, circa 1760. (No. 51.) 

3. Pinchbeck watch by John Cock, London, circa 1760. (No. 52.) 
4- Gold repeating watch by Hughes, London, A. D. 1770. (No. 55.) 



ANNALS CARNEGIE MUSEUM, Vol. XI 



Plate XVIII. 




Silver watch tor use in sedan-cliair by Pliilii)li X'ottcr. X'ienna. 17O3. (No- .S,5-' 



o 




ANNALS CARNEGIE MUSEUM, Vol. XII 



Plate XX. 




1. Vari-colored gold watch, set with turquiose, P. LeRoy, Paris, 1780. (No. 66.) 

2. Pair-case gold watch by Harry Potter, London, A. D. 1773. (No. 57.) 

3. Gold watch with outer case of tortoise-shell, N. Scott, Maidstone. (No. 58.) 

4. Enamelled gold watch by P. Bannicker, Paris, 1775. (No. 59.) 



ANNALS CARNEGIE MUSEUM, Vol. XII 



Plate XXI 




Large watch in pierced pinchback case, G. Margetts, London. A. D. 1779. (No. 61. 



ANNALS CARNEGIE MUSEUM, Vol. XI 



Plate XXI 







1. Gold watch of Louis XVI period, French, circa 1770. (No. 56.) 

2. Gold pair-case watch, by Jas. Noakes, London, 1776-1 794- (No. Oo.) 

3. Gold watch by Jos. Martineau & Son, London, 1784- (No. 63.) 

4. Gold watch by Pierre Gregson, Paris, 1 780-1 790. (No. 64.) 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XXIII. 




1. Silver verge watch by Philip Phillips, London, 1776. (No. 70.) 

2. Watch in bronze case pique by Tayot, Paris, eighteenth century. (No. 71.) 

3. Silver watch with outer enamel case, Charleson, London, 1758. (No. 43.) 

4. Gilt metal watch by L. Bordicr, Geneva, circa 1785. (No. 68.) 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XXIV. 




1. Pendant watch by J. Clovi, Rome, A. D. 1741- (No. 42.) 

2. Gold watch by Rundle & Bridge, London, A. D. 1790. (No. 74-) 

3. Gold pair-case watch, outer case enamelled. (No. 72.) 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XXV. 




Watch and fob by Jeffrys & Jones, London, 1769-94. (No. 73.) 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XXVI. 




1. Watch in outer case of shagreen, J. Morier, London. (No. 77.) 

2. Gold watch bordered with pearls by Patry & Chaudoir, Geneva, 1800. (No. 82.) 
3- Large gold watch by Bouvier F~reres, Geneva, A. D. 1790. (No. 75.) 



ANNALS CARNEGIE MUSEUM, Vol. XII, 



Pkitc- XXVII. 




1. Watch by George Achard et fils, Geneva, A. D. 1780. (No. 67.) 

2. Gold watch, Swiss or French, early nineteenth century. (No. 79.) 

3. Pinchbeck watch, French or Swiss, late eighteenth century. (No. 78.) 

4. Watch by Luigi Duchene et fils, Switzerland, late eighteenth centurJ^ (No. 80.) 



ANNALS CARNEGIE MUSEUM, Vol. XII 



Plate XXVIII 




1. Gold watch by Pcrrin Freres, period of First Empire. (No. 85.) 

2. Watch by Racine, Paris, circa A. D. 1800. (No. 89.) 

3. Chatelaine, gold and enamel, French, period of First Empire. (No. 84.) 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XXIX 




1. Watch, l"'rc'iuli or Swiss, caiiv' iiiiulct'iuli rciilur\-. (No. S8.; 

2. Swiss watfli, carh- iiinctci-iilli cciUmN-. (No. 8i.) 

3. French watch, early nineteenth centnr>'. (No. 83.) 

4. Watch by A. V. Girod, early nineteenth century. (No. 93.) 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XXX. 




1. Watch, Frencli or SvvivSs, early nineteenth century. (No. 86.) 

2. Watch by A. L. Breguet, 1746-1823. (No. 90.) 

3. Gold musical watch, French or Swiss, early nineteenth century. (No. 87.) 

4. Viennese watch, early nineteenth century. (No. 96.) 



ANNALS CARNEGIE MUSEUM. Vol. XII 



Plato XXXI. 




i*» 




1. Watch by Frazer, London, 1785. (No. 69.) 

2. Watch by A. L. Breguet, Paris, 1746-1823. (No. 92.) 

3. Musical box and watch, Swiss or French, early nineteenth century. (No. 98.) 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XXXII 




1. Watch by L. A. Breguet, Paris, circa 1830. (No. 95.) 

2. Watch by Pernetti, Swiss, nineteenth century. (No. 99.) 

3. Watch by L. A. Breguet, Paris, 1830. (No. 94.) 



ANNALS CARNEGIE MUSEUM, Vol XI 



Plate XXXIII. 




I, 2. Watch studded with pearls by Bautte & Moynier, Geneva, 1820. (No. 100.) 

3- Watch by A. L. Breguet, Paris, 1746-1823. (No. 91.) 

4- Watch set in carbuncle, by Froment-Meurice, France, nineteenth century. (No. 97.) 






ANNALS 

OF THE 

CARNEGIE MUSEUM 

VOLUME XIL NOS. 2-4. 




Editorial Notes. 



The exigencies of the period, which began when the United 
States were forced into the world war by Germany, left their impress 
upon the entire life of the nation, and affected every individual and 
institution. The Carnegie Museum did not escape this influence. 
In view of all the circumstances and the painful economies which 
seemed to be necessitated, the Director of the Museum, who is also 
the Editor of its publications, felt constrained to suspend the issue of 
the Annals and Memoirs of the Museum, until such time as the 
institution might be relieved in a measure from the financial embar- 
rassment into which it had been brought in the spring of 1918 by the 
reduction in its revenues, which took place at that time. During the 
year 1918 no parts of either of the above-named publications were 
placed in the hands of the printer. It is a matter of congratulation 
that in May, 1919, we were apprised of the fact that the funds avail- 
able for the work of the Museum for the current fiscal year had been 
restored to the same amount as that which had been granted in a num- 
ber of the preceding years, and accordingly steps have been taken to 
issue as speedily as possible a number of valuable and important 
papers on various subjects which should, and under other circum- 
stances would, have been published in 191 8. The Editor sincerely 
hopes that the temporary delay in the issue of these contributions to 
science may not have worked entire disappointment to the authors 
of the same, and that he may receive their pardon in view of the fact 
that the circumstances were not of his making or choosing. 
3 — DEC. 2, 1919 33 



34 Annals of the Carnegie Museum. 

From July, 1918, to the middle of April, 1919, Mr. Douglas Stewart, 
who has long been associated with the Director of the Carnegie 
Museum as an administrative assistant, and also has discharged the 
functions of the Curatorship of Mineralogy, and taken charge of the 
collections in several of the other sections, was absent from Pitts- 
burgh. Mr. Stewart had volunteered his services to the American 
Red Cross in Washington, and there was occupied in the work of the 
Bureau of Prisoners Relief for most of his time. Letters testifying 
in unqualified terms to the value of his services in this connection, 
and to the gratitude of the Officers of the American Red Cross for 
the kindness of the Trustees of the Carnegie Institute in having al- 
lowed him to undertake the work have been received, and are cherished 
in the archives of the Museum. 



Mr. O. J. MuRiE, who at the time the United States entered into 
the war, was assistant curator of the collections of mammals in our 
possession, was called to the colors, and entered the Aviation Service. 
It was not his privilege to be called into duty across seas, and he has 
recently been honorably discharged. He writes that it is his 
present intention not to devote himself to curatorial duties, as he 
prefers life in the open, and he has made arrangements to undertake 
the breeding of fur-bearing animals in the Northwest, in association 
with his brothers. 



Mr. John Link, who entered an Ofificers Training Camp in June, 
1918, to better qualify himself for military service, was in the fall of 
the year discharged because of a minor physical defect, and imme- 
diately returned to his post at the Museum. 



Since the last number of the Annals of the Museum was issued, 
the hand of death has removed from us a number of those whose 
services to the institution have made their connection with it memor- 
able. On Christmas Eve, December 24, 1918, IMonsignor A. A. 
Lambing passed away, full of years and honors. He was one of the 
original Trustees of the Institute, having been appointed by Mr. 
Carnegie, had served for a considerable time as a member of the 
Committee upon the Museum, and almost continuously as Honorary 
Curator of our Historical Collections. The sudden death from pneu- 
monia on May 14, 1919, of Mr. Henry J. Heinz, has robbed us of another 



Editorial. 35 

devoted friend, who did much to add to the collections of the Museum, 
and long served as Honorary Curator of Textiles, Timepieces, and 
Ivory Carvings. His loss is irreparable. Professor Charles Rochester 
Eastman and Herbert Huntington Smith are two others whose names 
are indelibly linked with the development of the Carnegie Museum, 
and they both were summoned from life under tragic circumstances. 
Brief biographies of these friends of the institution and former asso- 
ciates of the Director will be given elsewhere in this issue of the 
Annals. 



In spite of the financial embarrassment under which we labored 
during the fiscal year, which closed on March 31, 191 9, there were 
some notable additions made to the collections in the Museum during 
that period. 

One hundred and twenty-six mammals were added, seventy-nine 
representing the fauna of North America, the remainder being from 
South America. We have over four thousand five hundred mammals 
in the Museum. 

The collection of birds was increased by the addition of six thousand 
and eighty-five specimens. Thirty-two specimens represent species 
from North America, the remainder are specimens from portions of 
South America, which have hitherto been little explored, or only 
imperfectly represented in the museums of the world. Our birds 
number about seventy thousand specimens. 

Our collection of recent reptiles was notably augmented as the result 
of the journey undertaken by Dr. L. E. Grififin through New Mexico, 
Arizona, and southern California. There are not far from eight 
thousand specimens now in the collection, principally from North 
America, although South America is also represented, and we have 
some material from other continents. 

The acquisition of a representative collection of the fresh-water fishes 
of the island of Formosa through the kindness of Mr. Matsumitsu 
Oshima of the Institute of Science which is maintained by the govern- 
ment of that island, is noteworthy. Our collection of fishes aggre- 
gates over fifteen thousand specimens. 

Numerous mollusca, principally from the Mississippi Valley were 
acquired, and a great deal of important work was done during the 
year by Drs. Ortmann and Sterki in determining and classifying 
material already obtained in former years. The collection of moUusks 



36 Annals of the Carnegie Museum. 

now numbers about six thousand seven hundred species, represented 
by about thirty thousand sets of specimens. 

About sixteen thousand insects were obtained mainly from South 
America and tropical Africa. The Director completed for the American 
Museum of Natural History an extensive paper upon the lepidoptera 
of the Congo brought back from that country by the Lang-Chapin 
Expedition. A number of new species and new varieties are described. 
There are over a million of insects in our collections, representing 
approximately one hundred thousand species. 

The work in the Herbarium was largely restricted during the past 
year to the classification and arrangement of material acquired in 
former years, but some important accessions were made by purchase and 
exchange. A very important paper upon the Oligocene fossil plants 
collected some years ago by Mr. Earl Douglas near Missoula, Mon- 
tana, has been prepared by Professor O. E. Jennings, and submitted 
for publication. Dr. Jennings informs me that there are about two 
hundred and fifty thousand specimens in the herbarium, representing 
approximately fifty-five thousand species. 

Nearly one thousand mineralogical specimens were received during 
the year, principally as gifts, including that of the Andriessen Col- 
lection, presented to the Museum by Mr. Richard Hartje, Jr., and the 
collection made by the late Mr. E. L. Dunbar of Pittsburgh, pre- 
sented by his daughter. Miss Fannie K. Dunbar. Our collection of 
minerals is one of the largest in the state. 

In the field of paleontology important work was done at the National 
Dinosaur Monument in Utah, where Mr. Earl Douglass has continued 
the work of excavation commenced some years ago. The force in the 
laboratory has succeeded in freeing from the matrix a great deal of 
interesting and valuable material, which we hope to soon describe 
and publish to the world. Some of it undoubtedly represents forms 
hitherto unknown to science. Our paleontological collections are 
very extensive, exceeded in size and importance by only one, or pos- 
sibly two others in North America. 

One of our undertakings was the thorough exploration of the Indian 
mound at Guyasuta, the ancestral home of the Darlington family 
near the city, the Museum having been invited by Mrs. S. A. Amnion 
and her sister, Miss Darlington, to do this work. It is a well-confirmed 
tradition that the late General O'Hara caused the interment in this 
mound of the remains of the celebrated Indian chief, Guyasuta, who 



Editorial. 



37 



accompanied George Washington on his memorable journey to the 
French commandant at Fort LeBoeuf. During the later years of the 
life of the old Seneca chief, General O'Hara built for him a log cabin 
near the spot and provided him with food and clothing. Evidence 
of an intrusive burial in the mound was discovered, and the remains 
have been carefully preserved, and when the proper time comes, it is 
hoped to obtain permission to reinter them in Highland Park at a 
point overlooking his old home, marking the spot by a suitable me- 
morial not only to him but to the tribe to which he belonged, which 
has almost vanished. 

We have received many important gifts and loans of ethnological 
material, among which may be mentioned the entire collection of the 
late Robert C. Hall, which, having become the property of Mr. 
Edward B. Lee, has by him been kindly placed in our custody. 

Our collection of coins and medals was enriched by the gift to the 
Museum by Mrs. S. A. Amnion of the beautiful assemblage of medals 
formed by her father, the late William M. Darlington, Esq. A large 
number of the medals struck by the authority of the American Con- 
gress at various dates in the last century, thus became our property. 




D'Orsay presented to the Carnegie Museum by Messrs. Healey & Co. of N. Y. City. 
The first vehicle lighted by electric storage batteries. 



38 Annals of the Carnegie Museum. 

Notable additions to the ceramics in the Museum have been made, 
among which may be mentioned the gift of one hundred and forty- 
three specimens of early colonial and eighteenth century English 
pottery and glass-ware presented by Mrs. Jessie Porterfield Heasley. 

A gift which is much appreciated is that made by Messrs. Healey & 
Company, the famous carriage-builders of New York City, who 
presented to us a D'Orsay, made under the personal supervision of 
General Healey, which is a masterpiece of the art of the blacksmith 
and carriage-builder, and which has historic interest because it is 
known to be the first vehicle manufactured in America which was 
illuminated by electric light from storage-batteries, and was one of the 
first, if not the very first, to be provided with rubber tires. 

Numerous interesting additions were made to our carvings in wood 
and ivory, and art-work in the metals. 

Important accessions of letters and documents were added to our 
rapidly growing accumulation of historical papers. 

The foregoing paragraphs serve to throw some light on the activities 
of the Museum during the past year. Fuller and more detailed 
information is contained in the Annual Report of the Director, pre- 
sented to the Trustees in April, 1919. 



The fraternity between men of science in foreign lands and our- 
selves has been illustrated in a pleasing manner on several occasions 
during the past years. 

The following letter was received from Monsieur Edmond Perrier, 
the Director of the National Museum of Natural History in Paris, 
and Ex- President of the Academic des Sciences, shortly after the United 
States had begun hostilities against Germany. 

"Museum National d'Histoire Naturelle 

"Paris, le 22 Avril, 1917. 

"Le Museum d'Histoire Naturelle de Paris au Musee Car- 
negie. 

" Dr. \V. J. Holland, Directeur, 

" AIo?i cher Collegiie. 

" Au moment ou on va dresser avec accompagnement de salves d'ar- 
tillerie ])our cette fin le drapeau americain et le drapeau tricolor 
accoles I'un a I'autrc au sommet de la tour Eiffel, le Museum National 



Editorial. 39 

d'Histoire Xaturcllc, que vous avez comble, tient k vous adresser 
avec emotion un salut confraternel, et je suis heureux d'etre son inter- 
prete. Ensemble nos deux pays combattent pour assurer la paix 
au monde, pour detruire ce qui reste des institutions du moyen-^ge 
organisees pour la conquete et le butin, et leur substituer des institu- 
■ tions pacifiques basees sur la bienveillante solidarite qui doit unir tous 
les hommes dignes de ce nom. Cette commune conception de la vie 
est le gage que les Etats Unis et la France ne se separeront plus. 
"Soyez, confrere, de vos collegues de I'lnstitut Carnegie, institut 
de paix, notre interprete tres aime. 

"Edmond Perrier, 
" Directenr die Museum d'Histoire Naturelle.'^ 

Another communication, which was inspired by the signing of the 
armistice in November, and which came to hand in January, is the 
following: 

"MusEO Nacional de Ciencias Naturales Madrid 

(HiPODROMO). 

"Nov. 15, 1918. 
To the Director of the Carnegie Museum, Pittsburgh. 

Dear Sir: Please let us congratulate very warmly your Museum, 
as representative of American Science, for the end of the great war, 
a so glorious end for your country and for the cause of universal 
freedom and peace. 

Yours very friendly, 

Ign. Bolivar, 

Director. 
Joaquin Gonz. Hidalgo 

Luis Lozano Eduardo H. Pacheco 

Lucas Fr. Navarro Ricardo Ga. Mercet 

Angel Cabrera Candido Bolivar 

Antonio de Zulueta Romualdo Gonzalez Fragoso." 



11. REPORT UPON THE MATERIAL DISCOVERED IN 

THE UPPER EOCENE OF THE UINTA BASIN BY 

EARL DOUGLAS IN THE YEARS 1908-1909, 

AND BY O. A. PETERSON IN 1912. 

By O. a. Peterson. 

Introduction. 

The present report may be considered as the continuation of the 
work done by Messrs. Earl Douglass, O. A. Peterson, and C. W. 
Gilmore who have already written upon the Titanotheroidea and the 
turtles, and incidentally have briefly discussed the geology of the 
Uinta Basin. ^ A more complete and detailed account of the geology 
of the Uinta Tertiary is deferred pending a further and more complete 
study in the near future. The collection of fossil insects and plants 
will also be taken up in later publications. 

The illustrations are from drawings made by Mr. Sydney C. Pren- 
tice. 

I. FISHES AND REPTILES OF THE UINTA EOCENE. 

(The Turtles belonging to the collections made by Mr. Earl Douglas 
and the author of this paper have received full treatment by Mr. 
C. W. Gilmore in his paper, entitled "The Fossil Turtles of the 
Uinta Formation," Memoirs Carn. Mus., VII, 1915, pp. 101-161.) 

II. PISCES. 

Above the typical Green River Shales of the Uinta Basin remains of 
fishes, such as the scales of gars and isolated bones, are often found in 
different localities and horizons. These remains are generally dis- 
covered in heavy bedded sandstones of stream origin and are very 

^ Douglass, Earl, "Preliminary Description of Some New Titanotheres from the 
Uinta Deposits," Ann. Car. Mus., Vol. VI, 1909, pp. 304-311; "Geology of the 
Uinta Formation," Bull. Geol. Surv. of America, Vol. XXV, 1914, pp. 417-420. 
Peterson, O. A., "A New Titanothere from the Uinta Eocene," Ann. Car. Mus., 
Vol. IX, 1914, pp. 29-52; "A Small Titanothere from the Lower Uinta Beds," 
I.e., pp. 53-57; "Some Undescribed Remains of the Uinta Titanothere Dolicho- 
rhinus," I.e., pp. 129-138. Gilmore, C. W., "The Fossil Turtles of the Uinta 
Formation," Mem. Car. Mus., Vol. VII, No. 2, 1915, pp. 101-161. 

40 



Peterson: Material Discovered in Uinta Basin. 41 

seldom so complete as to permit of exact identification. A mandible, 
No. 2382, found in Horizon B appears to belong to the Halecomorphi 
of the family Amiidce, and agrees best with Pappichthys (P. plicatus) 
Cope, " Tertiary Vertebrata ", p. 59, Plate IV, Fig. i. A second speci- 
men. No. 3031, with a great number of scales in position, found in 
horizon C, appears to belong to Amia, while a third. No. 2368, from 
horizon C (near base) are fragments of both rami, which are longer, 
slenderer, more rod-like, but with a single tooth-row, as in No. 2382, 
referred to Pappichthys. This specimen may or may not pertain to 
the same group. 

III. REPTILIA. 

CROCODILIA. 

The fauna of the Uinta sediments includes a number of species 
of Crocodilia which continued from earlier epochs. These remains 
are generally found in sandstone of more or less coarse texture and 
of stream origin, as is the case with the remains of the fishes. In 
the collections made by the Carnegie Museum in 1908 and 1912 are 
a number of specimens which represent horizons A, B, and C of the 
Uinta. C, M., No. 2971, a skull which was found in the upper part 
of horizon A near White River, Utah, compares best with the de- 
scription and illustrations of Crocodiltis clavis by Professor Cope, 
" Tertiary Vertebrata", pp. 157-159, Plate XXII. A lower jaw, CM., 
No. 2972, found in horizon B, is also referred to C. clavis, while No. 
2988 is a pair of lower jaws found in horizon C, and represents an 
animal larger than either Nos. 2971 or 2972. This specimen may 
very possibly represent a new species, but reliable comparisons and 
diagnoses of characters cannot be made because of the mutilated 
condition of the specimen. 

IV. MAMMALIA. 

Order CARNIVORA (Fera;). 
Family MESON YCHID^E Cope. 
Genus Harpagolestes Wortman. 

I. Harpagolestes uintensis (Scott and Osborn). 

Proc. Amer. Philos. Soc, Vol. XXIV, 1887, p. 225. 

To this species is referred a lower jaw, together with other parts of 
the skeleton of one individual, C. M. No. 2961. The remains were 
found in horizon B of the Uinta formation close to the place where 



42 Annals of the Carnegie Museum. 

the hypotype of H. uintensis was found at the eastern end of the 
Uinta Basin near Wagonhound Bend Canyon on White River, Utah. 
The specimen represents an animal as large as, or slightly larger, than 
the type of H. iiintensis, and is here recorded for the convenience of 
the student, because it differs from that specimen, and also from 
H. imvianis Matthew, by the presence of a minute anterior cusp on 
Pg-. P2- and also M3- are of larger size than is the case in the type 
of H. uintensis. Since the type of the latter species consists of loose 
teeth, which may possibly not all belong to the same specimen, no 
great stress should at this time be laid on the differences noted. 
The presence of the anterior cusp on P3- may also be an individual 
character of the present specimen. The erection of a new species is 
therefore not thought prudent. 

Family OXY.'ENID.E Cope. 
Genus Oxy.enodon Matthew. 
Oxyaenodon dysodus Matthew. 

Hycenodon sp. Osborn, Bull. Am. Mus. Nat. Hist., VII, 1895, p. 78, Fig. 3. 
Oxyanodon dysodus Matthew, Bull. Am. Mus. Nat. Hist., XII, 1900, p. 49 

(type of genus Oxycenodon). 
Oxycenodon dysodus H.\Y, U. S. Geol. Surv., Bull. 179, 1902, p. 759. 

(Specimen No. 1893 A.M.N.H.) 

Oxyaenodon dysclerus Hay. 
Oxyanodon dysodus Wortman (mow Matthew), Bull. Am. Mus. Nat. Hist., XII, 

1900, p. 145. 
Limnocyon dysodus Wortman, Am. Journ. Sci. (4), XIII, 1902, p. 206. 
Oxycenodon dysclerus Hay, U. S. Geol. Surv., Bull. No. 179, 1902, p. 769; Matthew, 

Mem. Amer. Mus. Nat. Hist., IX, Pt. VI, 1909, p. 412 and 433. 

(Specimen No. 2515 A.M.N.H.) 

The above synonymy carefully worked out shows the status of the 
two species of Oxycenodon as known at the present time. 

On comparing Oxycenodon dysodus Matthew with material repre- 
senting Limnocyon the two appear to agree in the general structure 
of the lower jaw. They have a very thick jaw with a heavy symphysis, 
which extends backw-ard even with the posterior face of Pg-. The 
dentition of the type of Oxycenodon dysodus IMatthew is very imperfect 
and furnishes, unfortunately, small opportunity for comparison. 
It represents an animal larger than the type of Oxycenodon dysclerus, 
and further differs from the latter by having the chin gently turned 



Peterson: Material Discovered in Uinta Basin. 43 

upward, while in 0. dysclerus the chin is turned u])ward more abruptly. 
Although 0. dysodus may pertain to a different genus (? Limnocyon) 
it seems preferable to retain it in Oxya;nodon, pending the discovery of 
more material of both Oxyccnodon and Limnocyon in that horizon of the 
Uinta formation in which the type was found. This decision is mainly 
based upon the fact that a specimen representing a new species of 
Limnocyon, vide injra, was found during the summer of 1912 at the 
same locality where Oxycsnodon dysodus was obtained,- and which 
differs from the latter. In this new form the mandible is proportion- 
ally shorter, especially in the region of the premolars. The first 
upper premolar is obliquely placed in the alveolar border, and the 
premolars are crowded, while in 0. dysodus P- is directly fore-and-aft 
and the premolars occupy proportionally a greater space. These 
differences between the two specimens compared may be of generic 
importance. 

M- in Oxyccnodon dysclerus has the paracone and metacone placed 
closer together than in Limnocyon. In the latter genus P- and P- 
have smaller heels and M- has one instead of two median tubercles. 
Recently acquired material, representing the limbs of Oxyccnodon 
dysclerus, described below, further show that the limbs are longer in 
proportion to the skull than is the case in Limnocyon. 

2. Oxyaenodon dysclerus Hay (Plate XXXIV, Figs. 1-2). 

U. S. Geological Survey, Bull. No. 179, 1902, p. 759. (For synonymy see ante.) 

This species is represented by a well-preserved skull with the lower 
jaws attached, portions of the vertebral column, and the fore-limb 
of one individual. No. 3051. The specimen was found by the writer 
on White River, Uinta County, Utah, in the same locality and in the 
same horizon in which the type was obtained. It was found in a 
fine-grained sandstone concretion in which it is still imbedded in half 
relief, the skull and portions of the fore-limb having been more com- 
pletely worked out for further detailed study and illustration. 

The differences between the present specimen and the type (No. 
2515 of the American Museum) are very slight, judging from the 
descriptions, the illustration, and actual comparison. 

The axis has a high neural spine and the cervical centra have stout 
transverse processes, which project rather strongly backward. 

2 Mr. Peterson collected in 1893-95 the material upon which Osborn, Matthew, 
Hay, and Wortman wrote. (Editor.) 



44 



Annals of the Carnegie Museum. 



The left fore-limb is quite completely preserved, and, as observed 
above, is longer in proportion to the length of the skull than in Linino- 
cyon. The scapula is long and rather narrow, more nearly recalling such 




Fig. I. Oxycenodon dysderus. Carnegie Museum No. 3051. X 2/3. 

a recent form as Viverra zibetha, the coracoid process and the meta- 
cromion being, however, proportionally more developed than in the 
latter. The spine, which is very high, rises close to the glenoid cavity 
and separates the pre- and post-scapular fossae in nearly equal pro- 
portions. 

The humerus has received considerable crushing, especially in the 
upper portion of the bone, but its length is not impaired. The bone 




Fig. 2. Oxycenodon dysderus. Carnegie Museum No. 3051. X 2/3. 

is slender, the deltoid is quite prominent and extends well down on 
the shaft, but does not terminate as abruptly below as in the Miacidas. 
The entepicondyle and entepicondylar foramen are quite large, 
having approximately the same proportions as in Limnocyon veriis. 
The trochlea is also similar to the latter, except the inter-trochlear 



Peterson: Material Discovered in Uinta Basin. 45 

ridge, which seems to be less developed in the present species. This, 
however, may be partly due to crushing. The supinator ridge appears 
to be somewhat less developed than in L. verus. 

The proportionate length of the ulna and radius is the same as in 
Limnocyon verus. The fore-arm answers quite well to the description 
given of L. verus by Matthew, except that the olecranon process of 
the present specimen is shorter. 

The fore-foot is preserved with the exception of the ungual pha- 
langes, the trapezium, a portion of the trapezoid, and Mc. i. The 
carpus is low and broad, the metacarpals quite long in comparison 
with the length of the radius and ulna, and the phalanges are long, 
those of the proximal row with the shafts quite convex dorsally. 
The scaphoid and centrale, or what I take to be a portion of the cen- 
trale, appear to be united, and the lunar is also apparently fused, or 
nearly fused, with the scaphoid. The cuneiform is not complete, 
but appears to be high, with the articulation for the ulna excavated 
anteroposteriorly and having a large concave facet for the pisiform. 
The latter is of large size, with the termination of the tuberosity con- 
siderably expanded, especially supero-inferiorly. The unciform is 
large and the facet for the cuneiform is very oblique. The metacar- 
pals are moderately spread, their shafts have an oval cross-section, 
and are more expanded laterally near the distal end. The median 
phalanges are depressed. The unguals are not represented in the 
specimen. 

Genus Limnocyon Marsh. 
3. Limnocyon douglassi sp. nov.' (Plate XXXIV, Figs. 11-14). 

Type: Fragments of skull, lower jaws, and other portions of skeleton, 
C. M. No. 3373- 

Horizon: Uinta Eocene, horizon B. 

Locality: Eastern end of Uinta Basin, Uinta County, Utah. 

Specific Characters: Premolars relatively small; Pj suddenly en- 
larged; P- with anterior and posterior accessory cusps proportionally 
small. Type representing art animal slightly larger than Limnocyon 
potens of the Washakie Eocene. 

A portion of the right maxillary, with P-, P-, and My are very 
nearly complete; the roots of M- and the anterior premolars are pre- 
served in the specimen. There is also a fragment representing the 

5 In honor of Mr. Earl Douglass who found the type in 1908. 



46 Annals of the Carnegie Museum. 

left maxillary, the base of the zygomatic arch, both upper canines, 
one incisor tooth, and other fragments of teeth. The upper incisor is 
robust, laterally compressed, with a short and thick fang and the 
crown heavily enameled and striated. The canine is likewise flattened 
and the crown deeply striated. The maxillary is broken off in the 
region of P-, but it is evident that the tooth had two roots and occu- 
pied a crowded and oblique position in the alveolar border, differing 
thus from L. venis. P- is represented by two stout fangs, while P- 
has the crown complete. The main cusp of the latter is conical and 
rather blunt. At the antero-internal angle there is a minute basal 
tubercle, while posteriorly there is a basal heel of considerable size as 
in L. verus. P- has the triangular outline due to the prominent 
deuterocone characteristic of the genus. The anterior external tub- 
ercle of this tooth is unusually well developed, while that of the oppo- 
site tooth, though better developed than in L. verus, is of considerably 
smaller size. The median and postero-external cusps are much worn 
down, indicating the senility of the animal. On comparing P- with 
that of Limnocyon potens Matthew, it is at once observed that the 
anterior and posterior cusps of the crown are proportionally smaller 
than in that species. 

All the inferior premolars are crowded in the specimen under con- 
sideration and they carry deep vertical striae. Pj- is represented only 
by its roots in the left mandible, this tooth in the right mandible 
having dropped out and the alveole having been partly closed before 
the death of the animal. The second, third, and fourth premolars 
have only a slight indication of an anterior basal tubercle as in L. 
potens,^ while posteriorly there is a considerable cusp. Cingula are 
feebly or not at all represented. The trigonid of Mj is much worn 
and is otherwise incomplete posteriorly; the heel is rather large and 
slightly basin-shaped. The trigonid of My is entirely broken off, 
but the heel is quite complete, and shows a sharp inner face and a more 
obtuse elevation along the external margin of the heel. 

The vertebral column is represented only by a few centra, which 
appear to be very small in proportion to the remains of the head. 
The tail was robust, as is shown by two or three centra from the proxi- 
mal region. 

The limb-bones, even more than the remains of the vertebral 
column, appear disproportionately small, and it was with some hesita- 

■• In L. polens the anterior premolars are larger, while P^ is proportionally smaller. 



Peterson: Material Discovered in Uinta Basin. 47 

tion that I originally referred them to the same individual. However, 
when compared with Livniocyon verus described by Dr. Matthew 
{Mem. Amer. Mus. Nat. Hist., Vol. IX, Part VI, 1909, pp. 433-447) 
the differences are not so great, the limb-bones of the present specimen 
being perhaps somewhat smaller in proportion. 

The pelvis is represented by fragments of the ilia and the ischium. 
The fragments indicate that the pelvis was quite elongated and narrow. 
The ilium is suddenly expanded at the contact with the sacrum, espe- 
cially along the superior border, and the great sacro-sciatic notch is 
short. Immediately anterior to the acetabulum are heavy rugosities 
for muscular attachments. The ischium has a considerable inward 
twist when the pelvis is placed in position; the shaft is rather com- 
pressed laterally and the ischial spine is located well back of the aceta- 
bulum, similar to, though relatively of smaller size than in Tritem- 
nodon agilis Matthew. 

The femur is quite complete and presents characters much in 
accord with Limnocyon verus. The distal end appears to be somewhat 
broader than in the latter species, which may, however, be partly due 
to crushing. The upper end of the shaft has an inward turn similar 
to that in Limnocyon verus; the lesser trochanter is possibly located 
somewhat more posteriorly than in the latter species, but the third 
trochanter is approximately in the same position and of the same 
proportionate size. Distally the bone is, as already stated, broadly 
expanded, which is in part, at least, due to crushing. However, the 
rotular trochlea appears to be proportionally broader than in Lim- 
nocyon verus. Above the external condyle on the postero-fibular angle 
is a curious deep groove in the neighborhood of the attachment for 
the gastrocnemius. The intercondyloid notch is quite broad. 

The tibia and fibula are poorly represented. It is, however, plain 
that the shaft of the fibula is quite stout and oval in cross-section. 

Both calcanea are represented with the greater portion of the 
tuber calcis broken off. The bone is rather small, not much expanded 
laterally, due to the small development of the lesser process. The 
groove for the interosseous ligament is quite broad, as in many recent 
FercB, but the bone as a whole is perhaps most suggestive of the 
Miacince. The peroneal tubercle is large, with the tendinal groove 
for the peroneus longus fairly well developed. The facet for the 
cuboid is oblique and on the plantar face is a truncated tuberosity for 
muscular attachment. 



48 Annals of the Carnegie Museum. 

The astragalus is represented by the head, which is broad as in L. 

verus. 

Measurements. 

Length of upper molar-premolar series 54 mm. 

Length of mandible, condyle to point of symphysis 131 " 

Depth of mandible at Pj 25 " 

Depth of mandible at M 2 20 " 

Length of lower molar-preraolar series 55 " 

Length of lower premolars 32 " 

Length of lower molars i and 2 23 " 

Length of femur 122 " 

Family MIACID^ Cope. 
Genus Mimocyon gen. nov. 

4. Mimocyon longipes sp. nov. (Plate XXXIV, Figs. 6-10.) 

Type: Fragments of the left lower jaw with Pj and My in place, 
fragment of Pj? and greater portion of the lower canine of same side; 
the distal end of the humerus, proximal end of the ulna, a section of 
the shaft of the radius, the distal articulation of the tibia, the tarsus, 
and proximal ends of three metatarsals, C. M. No. 3022. 

Horizon: Uinta Eocene, Horizon C, near base. 

Locality: Six miles east of Myton, Uinta County, Utah. 

Generic Characters: Antero-posterior diameter of P^ and ify ^Qual. 
Anterior and posterior accessory cusps small and heel large. Trigonid 
low and small, heel large and basin-shaped. Tarsus high. 

Description of the Type. 

The ramus is not deep and is rather thin transversely. Py had in 
all probability two roots. The succeeding alveoli of P^- and P-g- indi- 
cate that they were of considerable antero-posterior diameter and 
were distinctly two-rooted. Py is not reduced in size, its antero- 
posterior diameter equals that of My. The principal cusp of the 
crown is not compressed laterally, is rather high, and has distinct 
ridges in front and behind. There is a small anterior basal cusp, which 
more appropriately might be regarded as a heavy cingulum, a promi- 
nent basal heel, and a small accessory tubercle posteriorly. The 
cingulum is well developed and completely surrounds the tooth, in 
this respect answering the description of Miacis vulpinus (Scott). ^ 
My has an unusually small and low trigonid, which is very little greater 

^ Proc. Atner. Assoc. Adv. Sci., 36th meeting, New York, 1887, p. 255. 



Peterson: Material Discovered in Uinta Basin. 49 

in the antero-postcrior diameter than the heel. The anterior and 
internal tubercles of the trigonid are of moderate development, while 
the external is, as usual, the largest of the three. The heel is unusually 
large and decidedly basin-shaped. Externally this tooth has a well- 
developed cingulum, but internally it is smooth. M-j was of consid- 
erable size, judging from the alveolus. M-g- was evidently present, 
but whether or not it had two roots cannot be determined from the type 
specimen. 

The distal end of the humerus is comparatively broad, and in this 
respect it suggests such forms as Miacis parvivorus and Oodectes 
proximus, described by Matthew.^ The entepicondyle and the entepi- 
condylar foramen are also of large size and the trochlear portion for 
the radius convex, as in M. parvivorus. The trochlea for the greater 
sigmoid cavity of the ulna is wide and the supinator ridge does not 
extend high up on the shaft. 

A fragment of the proximal end of the ulna is preserved with the 
type. Tie olecranon process appears to have a moderate expansion 
antero-posteriorly and the anterior border is compressed laterally 
immediately above the greater sigmoid cavity. The latter is quite 
wide and oblique, with a prominent coronoid process and a concave 
lesser sigmoid cavity, which indicates power to rotate the head of the 
radius. The shaft appears to have a considerably backward curve, 
and is compressed laterally, expanded antero-posteriorly, and deeply 
channeled on the ulnar face. 

The distal trochlea of the tibia is not broad, and there is a shallow 
groove separating the internal malleolar facet from the external 
portion of the trochlea. The astragalus is high and narrow, with a 
long neck unlike that of Miacis parvivorus and more suggestive of the 
astragalus of Cercoleptes. 

The trochlear groove of the astragalus is shallow, the internal ridge 
being low, but more developed than in the form from the Bridger and 
in older genera. The head is laterally broad and there is present a 
minute astragalar foramen. The tuber of the calcaneum is compressed 
laterally and terminates above in a prominent tuberosity, which is 
very slightly grooved. The lesser process is not greatly expanded 
and the facet for the cuboid is very oblique. The dorsal face of the 
calcaneum is broken ofT. The navicular is proportionall}' narrow, 

^ "Carnivora and Insectivora of the Bridger Basis," Mem Amer. Mus. Nat. Hist., 
Vol. IX, 1909, pp. 368, 378. 
4 — DEC. 2, 1919. 



50 Annals of the Carnegie Museum. 

especially when the transverse diameter of the head of the astragalus 
is taken into consideration. The fact, however, is, that the head of the 
astragalus extends laterally, as in the Oxyaenids, and articulates with 
the cuboid perhaps even to a greater extent than in the recent Kinka- 
jou. The cuboid is very nearly as high proportionally as in Cer- 
coleptes and has, as indicated above, a facet for the astragalus on the 
proximal tibial angle. The facet for the distal end of the calcaneum 
is oblique to conform to the corresponding facet on the calcaneum 
described above. There is present a heavy plantar tuberosity which 
again suggests similarity to Cercoleptes. The cuneiforms are high 
and narrow, the entocuneiform being of considerable size and has a 
large facet for Alt. I, indicating that digit to be approximately pro- 
portionate in size to that in Cercoleptes. The metatarsals are repre- 
sented only by the heads of II, IV, and V and a portion of the shafts of 

III. 

Measurements. 

Vertical diameter of ramus at My approximately 20 mm. 

Length of molar-premolar dentition approximately 60 

Antero-posterior diameter of P4 11 

Antero-posterior diameter of My n 

Transverse diameter of distal trochlea of humerus 19 

Greater transverse diameter of dorsal end of humerus approximately .... 30 

Greatest length of calcaneum approximately 42 

Greatest height of astragalus 25 

Transverse diameter of trochlea of astragalus 10 

Greatest height of cuboid 19 

Transverse diameter of tarsus distally 24 

The genus and species described above is most closely allied to 

Miacis. That it belongs to the branch of the Cynoid MiacincB 

appears to be quite certain from the remains at hand, but whether or 

not its successors are to be found in any of the known genera from 

later geological formations "will not be known until the discovery of 

more perfect material. The study of the limited material at hand 

suggests that the genus represents an aberrant form, especially when 

the oxyaenid and cercoleptid features of the pes are compared with the 

dentition. 

Genus Prodaph^nus Matthew. 

5. Prodaphaenus (?) robustus sp. nov. (Plate XXXIV, Figs. 3-5). 

Type: Fragments of lower jaws with P:^ My and M2- in place; 

fragments of vertebrae, greater portion of right humerus, and head of 

femur, C. M. No. 3023. 



Peterson: Material Discovered in Uinta Basin. 51 

Locality: Six miles east of^Myton, Uinta County, Utah. 

Horizon: Uinta Eocene, Horizon C, near base. 

Specific Characters: Antero-posterior diameter of P^ and ilfy subequal. 
P-^ with anterior basal cusp; absence of posterior accessory cusp, hut a 
cutting heel of considerable size present. Mj with broad and high 
trigonid; small heel, not basin-shaped. M^ of proportionally small 
size, trigonid low, and heel small. M-^ one-rooted. Animals larger 
than Prodaphxnus scotti. 

This specimen appears to differ both from Miacis and Uintacyon'' 
and may represent a new genus, nevertheless it is thought best to 
provisionally place it in Prodaphcenus having regard to the incom- 
pleteness of the type. The unfortunate absence of the upper teeth 
prevents comparison with Prodaphcenus scotti. It is purposely kept 
sepacate from the latter on account of its greater size and also to save 
a possible confusion later on, since it may not even pertain to the 
same genus. The description and measurements of '^Miacis'' 
vulpinus Scott and Osborn^ do not appear to agree with the present 
specii-ticn. The structure of the lower jaw and the teeth do not com- 
pare well with Procynodictis vulpiceps Wortman and Matthew.^ 

The general contour of the mandible is perhaps more like Uintacyon 
than Miacis. It is certainly heavier, deeper, and I should judge^" 
relatively shorter than Prodaphcenus (Miacis) uintensis (Osborn) 
the symphysis being especially heavy. The canine, as indicated by 
the root, is heavy and laterally compressed. Py is single-rooted, Po- 
is two-rooted, but, as in Procynodictis, apparently of considerably 
smaller size than P-jj. Py suggests that of Miacis hargeri; it is a large 
tooth with a prominent anterior basal tubercle, a slight and rather 
smooth cingulum externally, entirely smooth internally, and a large 
cutting heel. My has a high and broad trigonid with the anterior 
tubercle well developed, as in Procynodictis vulpiceps and later forms. 
The heel is rather slightly developed both antero-posteriorly and 
transversely. It is of a trenchant type with a moderately heavy 
cingulum, extending along the inner face to the base of the postero- 
internal tubercle of the trigonid. The antero-posterior diameter of 

^ Dr. W. D. Matthew, Mem. Amer. Mus. Nat. Hist., Vol. IX, 1909, pp. 326-377. 

« Proc. Amer. Philos. Soc, Vol. XXIV, 1887, p. 255. 

^ Bull. Amer. Mus. Nat. Hist., Vol. XIV, 1899, p. 121. 

1" No contact has been established between the anterior and posterior portions 
of the mandible, but there would appear to be little doubt that the two pieces 
pertain to the same individual. 



52 Annals of the Carnegie Museum. 

the tooth is but very little greater than that of P4. There is a slight 
cingulum on the external face of the trigonid, but internally the tooth 
is smooth, except the cingulum of the heel just described. Mg- is 
much reduced in size and in this respect suggests the condition in 
Procynodictis and Cynodictis rather than in Daphcenus and Daphce- 
nodon. The trigonid of this tooth is, however, lower than in the illus- 
tration given of Procynodictis and suggests more nearly that found in 
later forms (Cynodictis); the heel is slightly basin-shaped. M3- was a 
small one-rooted tooth which in the type is represented only by the 
alveolus. 

Except the mandible just described, the humerus, minus the proximal 
end, is the only part of the type worthy of description. The shaft is 
quite heavy and rather long, and the deltoid ridge is robust and ex- 
tends downwards more than half the length of the shaft, terminating 
abruptly, which plainly furnishes the cynoid characters given by 
Matthew & Wortman. The ulnar border of the distal end of the 
shaft is not complete, but enough is preserved to show that the supi- 
nator ridge was ciuite prominent and extends well up, in this respect 
being unlike the supinator ridge of Mimocyon longipes, which is much 
less extended upwardly and more like w^hat is found in most species of 
Miacis and of Lycarion {Vulpaviis) hargeri. The distal trochlea, 
though broad, is not as broad in proportion as in Mimocyon, and the 
middle portion of the articulation for the radius is much less convex 
and more nearly approaches what is seen in the genera of later for- 
mations {Daphcenus and Daphamodon). The entepicondyle and the 
radial border of the entepicondylar foramen is broken off. There 
was apparently no supra-trochlear perforation. 

Measurements. 

Length of mandible, canine to M3 approximately 65 mm. 

Depth of mandible, canine to P2 approximately 22 

Transverse diameter of mandible at symphysis opposite Pj 10 

Length of molar-premolar series approximately 58 

Length of premolar series approximately 36 

Length of molar series approximately 23 

Genus Pleurocyon gen. nov. 
6. Pleurocyon magnus sp. nov. (Plate XXXV). 

Type: Symphysis of lower jaws with left ramus nearly complete 
and a portion of the dentition in place, C. INI. No. 2928. 



Peterson: Material Discovered in Uinta Basin. 53 

Paratype: Fragment of right maxillary and jugal, right ramus with 
canine in place, fragments of upper and lower teeth, and a consider- 
able portion of the skeleton, C. M. No. 3006. 

Horizon: Uinta Eocene, Horizon C, near base. 

Locality: Six miles southeast of IVIyton, Uinta County, Utah. 

Generic Characters: Mandibular rami proportionally short and deep. 
Lower molars decreasing in size from first to third, tuber culo-sectorial, 
with long trenchant heels, low trigonids, and the metaconids of moderate 
size. 

Description of the Type. 

In comparing some fragments of the upper teeth of the paratype, 
No. 3006, with the description and illustrations of the superior dentition 
of Vnlpavus and Lycarion by Wortman" and Matthew^^ it is at once 
seen that the canine of the present form is more compressed laterally. 
There is also a total absence of the internal and posterior ledge-like 
elevation, which rises from the cingulum (hypocone) on the internal 
face of the molars of Miacis and Lycarion, and in this respect it is more 
like the condition found in Oodectes, except that the deuterocone is 
proportionally more developed antero-posteriorly and has a tendency 
to be sub-divided into two tubercles in the form now being described. 
In other words, it appears that the hypo- and proto-cones have been 
united as indicated in the illustration, PI. XXXV, Fig. 11, and that 
there is one intermediate tubercle of small size. Whether or not there 
were two intermediate tubercles cannot be determined from the 
material at hand. The postero-external tubercle is of well propor- 
tioned size. The tooth is broken on what I take to be its antero- 
external angle. 

The lower dentition of the type. No. 2928, is much better preserved 
and furnishes more satisfactory means of comparison. There were 
most likely three lower incisors, though this cannot be fully deter- 
mined from the type. Judging from the roots, the second incisor was 
of rather small size, while I3- was quite large. The canine, premolars 
Y, 2". the anterior portion of P4 and M3- unfortunately were not re- 
covered. The root of the canine presents a long oval, placed nearly 
in a direct antero-posterior position on the axis of the jaw. The 

11 Amer. Journ. Set., (4), Vol. XI, 1901, p. 341, Vulpavus; pp. 342-445, Lycarion 
hargeri (Wortman). 

12 Mem. Amer. Mus. Nat. Hist., Vol. IX, 1909, p. 343, Lycarion; pp. 344, 346, 380, 
Vulpavus palustris Marsh. 



54 Annals of the Carnegie Museum. 

first lower premolar is indicated by a single root separated from the 
canine by a short diastema, but is continuous with the series back 
of it. Pa^ has two roots closely crowded together. P3- is rather cur- 
iously oxyaenid in structure, the main cusp consisting of a low conical 
tubercle with a prominent conical heel and the anterior and internal 
faces surrounded by a smoothly rounded cingulum; externally the 
tooth is practically smooth. As already stated, the anterior face of Pj 
is not present in the type; in the paratype, however, this tooth is 
present, though isolated (see PI. XXXV, Figs. 12-13). The oxyaenid 
feature of this tooth is again repeated, the main cusp being compara- 
tively of rather small size, while the heel is large and of a trenchant 
character. The anterior basal tubercle, though small, is well defined, 
but there is no posterior accessory cusp, in this respect suggesting 
such forms as Prodaphccnus (?) rohiistus, p. 50, or " Miacis'' vulpinus 
Scott. 1^ The heel is, however, proportionally larger than in the two 
species mentioned. There is a poorly developed cingulum on the 
internal angle of the heel, otherwise the tooth is smooth. Mj has a 
low trigonid with the external tubercle the largest, the internal of 
moderate size, and the anterior small and low, this again agreeing 
in a general way with Scott's description of " Miacis'' vulpinus. 
The heel of My in the present species, though trenchant as in " Miacis" 
vulpinus, is of large size, with an extended inner ledge, while that 
of Scott's species is low and small. In comparing My of the present 
genus with the miacids generally, it is apparent that Vulpavus pro- 
fectus with its low trigonids bears a closer similarity to it than any of 
the other genera. The basin-shaped heel of Vulpavus is, however, 
totally unlike the large trenchant heel of the genus under consider- 
ation, which in this respect is perhaps most suggestive of Oodectes her- 
pestoides Wortman. In the present genus M^- is very little smaller 
than Mj and the two are similar in every respect, except a somewhat 
smaller sized anterior tubercle of the trigonid on M^- There are no 
cingula on either of the two teeth just described. Mg is considerably 
smaller than the preceding molars and is implanted by two fangs. 

The only portion of the skull preserved, besides the fragments of the 
teeth described, is a mutilated fragment of the right maxillary, with 
a portion of the zygomatic arch of the jugal attached. This fragment 
may or may not pertain to the paratype here described. The alveolar 

^^"Amphicyon" (?) vulpinum, sp. n. Scott and Osborn, Proc. Amer. Phil. Soc, 
XXIV. 1887, p. 255. 



Peterson: Material Discovered in Uinta Basin. 55 

portion of the maxillary apparently had considerable vertical diameter 
and the zygomatic arch is rather slender. 

The general contour of the horizontal ramus of the type as well as 
of the paratype is not unlike that of Oodectes. The diameter of the 
jaw is small transversely and great vertically, with the lower border 
suddenly rising opposite Mj and gradually tapering towards the 
symphysis (see PI. XXXV, Figs. lo, 15). The latter is curiously sug- 
gestive of Sinopa, having a shallow, quite long, and loose symphysis. 
In fact the entire jaw is like that of the latter genus, but the canine is 
heavier, the premolars set closer together, the cusps of all the teeth 
lower, the heels of molars not basin-shaped, and M^j- too much reduced 
(added to this the different structure of the limbs, vide infra) to permit 
it to be regarded as belonging to Sinopa. 

There are two mental foramina of moderately large size, the ante- 
rior under P2- and the posterior under the posterior portion of P^-. 
In the type the base of the ascending ramus is very close to the dental 
series, so that the posterior fang of M3- is placed at a considerable 
angle to the position of the roots in the rest of the cheek-teeth. 
This is not the case in the paratype, where there is a wider space 
between M-g- and the ascending ramus. Whether or not this is a 
specific character cannot be determined from the material at hand. 
The ascending ramus of the rami of both type and paratype are broken 
off. 

M EASUREMENTS. 

Type Paratype. 

No. 2928. No. 3006 

Transverse diameter of upper molar 12 mm. 

Antero-posterior diameter of upper molar 7 

" " " " horizontal ramus from incisor to 

and including M3 65 mm. 75 

Vertical diameter of ramus at Py 17 " 20 

" M^ 23 " 26 

Length of molar-premolar series 51 " 59 

" " premolar series 28 " 33* 

" molar series 25 " 26* 

Antero-posterior diameter of My 9-5 

" Mt 9 

Vertebral Column: The vertebral column of the paratype. No. 3006, 
is represented by a few fragments of the centra, which present no 
features worthy of mention, except that the caudal region was long 
and robust, as usual in the Eocene Ferce. 

* Indicates the measurements to be approximately correct. 



56 Annals of the Carnegie Museum. 

Limbs: The limbs consist of the greater portion of the shaft of the 
left humerus with the distal end mutilated, the ulna with the distal 
end wanting, fragments of the shaft of the radius, and the fifth meta- 
carpal of the left manus. There are also fragments of the pelvis, the 
shafts of both femora, right and left tibiae, fragments of the shaft of a 
fibula, a calcaneum, cuboid, ento- and meso-cuneiforms, all the meta- 
tarsals except the fifth, a few phalanges of the proximal and median 
rows, and the proximal portion of one ungual phalanx. 

The limbs of Pleurocyon, as represented in the paratype, No. 3006, 
are proportionally large, in this respect suggesting Palcearctonyx 
meadi Matthew. The fragmentary humerus in detailed structure is 
perhaps more like that of Viilpavus Marsh or Miacis Cope. As in the 
latter genera, the deltoid and supinator crests are prominent, the 
entepicondylar foramen is of large size and the entepicondyle was no 
doubt also of large size. The proximal end of the bone is broken off 
and the distal end is badly mutilated. It is, however, to be seen that 
the articular trochlea for the upper portion of the greater sigmoid 
notch of the ulna is rather deep and narrow. Neither the olecranon or 
the supratrochlear fossae are deep or high. Whether or not there is a 
supratrochlear foramen cannot be determined from the specimen. 

The shaft of the ulna is compressed laterally, and is rather straight 
The lower half of the ulnar, face is broadly and quite deeply channeled. 
Directly in front it presents a prominent and sharp crest for the 
attachment of the interosseous membrane, and well down on the 
radial face there is a second prominent crest, which helps to furnish 
support for the pronator muscle. The upper portion of the shaft of 
the ulna is again channeled on its outer face, while radially the shaft 
is more or less smooth. The coronoid process of the sigmoid cavity 
is broken off in the specimen, but the broken surface indicates that it 
was not large. The lesser sigmoid cavity is very shallow, while 
immediately below, and radial to the coronoid process, there is an 
unusually deep and large cavity for the attachment of the lateral 
ligament. The upper portion of the greater sigmoid cavity is com- 
pressed in order to meet the requirements of the deep and narrow 
articulation of the humerus described above. The olecranon process 
is short and truncated. The crescentic groove, over which passes 
the tendon of the triceps, is narrow, not very deep, and rather obliquely 
placed, due to the prominent anconeus process and the small develop- 
ment of the inner anterior margin of the groove. 



Peterson: Material Discovered in Uinta Basin. 57 

The fragments of the radius indicate that the head is oval, the 
articulation with the ulna flat, the vertical groove on the anterior face 
of the head deep and well defined. Some distance below the bicipital 
tubercle the shaft is round-oval in cross-section. The bicipital tub- 
ercle is large and apparently well separated from the head. 

Metacarpal V, the only bone which has been recognized as belonging 
to the manus, is short and quite heavy. Both the head and the distal 
end are much expanded, and the shaft has a decided forward curve, 
especially when viewed from the ulnar side. The manus was probably 
short and broad. 

Measurements. 

Humerus, total length of the fragment 149 mm. 

" antero-posterior diameter of shaft at lower extremity of deltoid 

ridge 27 

" transverse diameter of shaft at lower extremity of deltoid ridge. 16 

Ulna total length of the fragment 169 

" " olecranon process, anterior measurement 23 

" antero posterior diameter of olecranon process 28 

" greatest transverse " " " " 13 

Mc. V greatest length 42 

Hind Limb: A mutilated fragment of the ilium (all of that portion 
of the pelvis at hand) has the form of a heavy trihedral bar with the 
gluteal surface deeply excavated. A fragment of the ischium indicates 
that the shaft of this is deep and laterally compressed. The ischial 
spine is prominent, and terminates immediately back of the aceta- 
bulum, as is generally the case in the Ferce. 

Both femora are represented, but they are flattened by crushing 
and furnish few reliable characters. The shaft is rather heavy and 
there is a third trochanter, as in the Aliacidce generally. The lesser 
trochanter appears to be located on the postero-internal angle of the 
shaft. The digital fossa appears to extend well down on the shaft. 

Both tibiae are represented, the left with the distal end partly pre- 
served. The bones are crushed flat and have lost many of the original 
characters. There is a well-marked rugosity, which extends far 
down on the heavy shaft, indicating the position and prominence of 
the cnemial crest. The distal end is much flattened by crushing. 
The astragalar trochlea appears rather flat; the oblique ridge, which 
separates the external and internal condyles of the astragalus, is light; 
the internal malleolus is shown to be of fairly large size, and the grooves 
for the different tendons are well developed. 



58 Annals of the Carnegie Museum. 

A few fragments, which I judge to belong to the fibula, show that 
bone to have had a heavy shaft. 

The pes is fairly well represented. In its proportions it approaches 
Vulpavus profectus Matthew, though the metatarsals may possibly 
be relatively somewhat shorter than in that genus. The tuber of 
the calcaneum is not long, but is quite heavy, and has an ill-defined 
groove for the plantar tendon. The lesser process of the distal end 
is quite widely expanded; on the fibular face of the greater process is 
located the peroneal tubercle, which is large, and has a groove for 
the peroneus longus fully as large proportionally as in Cercoleptes 
caudivolviiliis. There is no facet for the fibula; the facet for the cuboid 
is triangular in general outline and not as oblique as in Vulpavus 
profectus, described by Matthew, but is perhaps more nearly like that 
in Cercoleptes. The astragalus and ectocuneiform were unfortunately 
not recovered. The entocuneiform is of large size and has a very 
oblique facet for the navicular. This great obliquity continues in a 
similar manner downward over the superior fibular face of the 
mesocuneiform, so that the facet for the entocuneiform and the facet 
for the navicular on the mesocuneiform form an unusually acute 
angle. The cuboid is low as compared with Mimocyon, or with such 
a recent form as Cercoleptes. There is not any evidence of an articu- 
lation for the astragalus on the proximal tibial angle as seen in the 
latter genera. The proximal portion of the tibial face has, however, 
a large rough area, which no doubt joined the navicular, but probably 
did not come in contact with the side of the head of the astragalus. 
The facet below this area is plane, subtriangular in outline, of consider- 
able size, and articulates with the ectocuneiform. The plantar tuber- 
osity, though quite well developed, is not nearly as large as in Cer- 
coleptes or Mimocyon, and the groove of the peroneus longus is 
consequently smaller and shallower than in the latter genera. The 
articulation for the metatarsals is quite oblique and subtriangular in 
outline. The metatarsals have apparently a close resemblance to 
those in Vulpavus profectus, the first being stouter and much shorter 
than the other four. Mt. V is not present in the material studied. 
As in the Miacidce, the proximal row of phalanges are long, somewhat 
depressed, and the shafts slightly curved dorsad. The median row of 
phalanges are short, with only a very slight indication of asymmetry 
and at their lower extremities bent slightly dorsad. The ungual 
phalanges are represented only by the proximal portion of one phalanx, 



Peterson: Material Discovered in Uinta Basin. 59 

which is high and latcralh' compressed, but whether or not it was cleft 
or whether it was retractile cannot be determined. 

M EASUREMENTS. 

Tibia, approximate length ' i6o mm. 

Pes, length, calcaneum to ungual phalanx 158 " 

transverse diameter from cuboid to entocuneiform, approxi- 
mately 33 

greatest length of calcaneum 50 

greatest height of cuboid 18 

length of Mt. 1 33 

length of Mt. Ill 48 

length of Mt. IV 56 

length of proximal phalanx digit (? IV) 30 

length of median " " (? IV) 18 

7. Pleurocyon medius sp. no v. 

In the America-n Museum collection of Uinta material are two frag- 
mentary specimens, No. 1969, a fragment of a lower jaw with My in 
place, and No. 1992, a lower jaw without teeth, and a number of 
fragments of limb-bones. These pertain to a considerably smaller 
species, which may be called Pleurocyon viedius, the second species 
known from the Uinta formation. 

The genus as described above should undoubtedly be placed in the 
family Miacidcc, as defined by Dr. Matthew.^^ A careful study of the 
type makes it possible to further place the genus in Matthew's series 
"B," the "Cercoleptoidei" of the subfamily MiacincB {I.e., p. 346). 
In certain respects the genus is perhaps most nearly like Vnlpavus, 
having, as that genus, low trigonids, the lower molars uniform in their 
general characters, ^^ and decreasing in size from the first to the third. 
However, instead of having the broad basin-like heels of the molars 
as in Vulpaviis, the present genus has molars with long trenchant 
heels, more like what is observed in Oodectes. From the latter the 
present genus differs in having the posterior basal cusps of the lower 
premolars larger and better defined, the trigonids of the molars lower, 
and the whole animal of much larger size. The paratype presents 
many characters common to the Miacidce which have already been 
mentioned. 

'^"Carnivora and Insectivora of the Bridger Basin," Mem. Amer. Mus. Nat. 
Hist., Vol. IX, 1909, pp. 344-345. 

15 M3 is unfortunately lost, but judging from the space it occupied, it was of 
somewhat large size. 



60 Annals of the Carnegie Museum. 

Incert^ Sedis. 

A fragmentary skeleton, C. M. No. 2386, from the upper B or lower 
C of the Uinta sediments is provisionally referred to the family 
MiacidcE. 

From the proportions of the fragments of the skull and from what 
is known of the limb-bones, I should judge the animal to have had a 
head in its proportions somewhat like Vulpavus, or possibly as small 
as that of Palcearctonyx. The limb-bones are, however, not as robust 
as in the latter, and more nearly suggest Vulpavus. The femur is 
slightly longer than that of V. ovatns described by Matthew. The 
illustration of the hind foot of V. projectus Matthew {I.e., p. 389, 
Fig. 31) is quite suggestive of the remains of the hind foot of the 
present specimen. The material very likely represents a species new 
to science, but I refrain from proposing a name at the present time, 
even though the specimen may be worthy of being named. 

Order RODENTIA {Glues). 
Family ISCHYROMYID^ Alston. 
Genus Paramys Leidy. 
8. Paramys compressidens sp. nov. 

Type: Lower jaw with cheek-teeth. C. M. No. 2920. 
Horizon: Uinta Eocene (Horizon C). 
Locality: Six miles east of Myton, Utah. 

Characters of Type Specimen.^^ Smaller than P. rohnstns or P. 
{Ischyrotomus) petersoni Matthew. Teeth narrower and jaw shallower 
and slenderer than in P. robustus. A greater prominence and better 

definition of the connecting crest between the 
two principal outer cusps of the molars than 
in P. {Ischyrotomus) petersoni. 

' This species resembles P. robustus and 
Fig. 3. Paramys compres- ,ti \ • • 1 

sidens. Carnegie Museum ^- i.i^chyrolomus) petersoni in the general 
No. 2920. X i/i. smoothness of the cusps, the shallow me- 

dian valley, the absence of the external in- 
termediate cusp, which is completely fused with the connecting loph, 
and the relatively small size of the anterior outer cusp of P4' (Gidley). 

" Mr. James W. Gidley of the U. S. National Museum, Washington, D. C, has 
kindly compared the Uinta Eocene rodent remains, published in this paper, with 
type material from different institutions now (1916) in the National Miiseum under- 
going a study preparatory to the forthcoming work on the Recent and Fossil 
Rodents by Messrs. Gerritt S. Miller, Jr., and James W. Gidley. 



Peterson: Material Discovered in Uinta Basin. 61 

Measurements. 

Length of cheek-dentition i8 mm. 

Vertical diameter of ramus at Mv 13 mm. 

9. Paramys medius sp. nov. (Plate XXXIV, Figs. 15-22). 

Type: Fragment of maxillary with three teeth, and fragment of 
lower jaw, C. M. No. 3048. 

Horizon: Uinta Eocene Horizon C. 

Locality: Six miles east of My ton, Utah. 

' Characters of Type: Of same size as Paramys delicatus, but with 
broader summits of the unzvorn crowns of the molars and with the inner 
and outer faces more vertical. The hypocone is also somewhat more 
set off from the protocone, and the styles of the outer cusps and the hypocone 
are more prominent than in the Bridger species. The lorhikled appear- 
ance of the enamel suggests that this species is referable to Taramys ' 
(Gidley). 

The total length of the three upper teeth is 13.4 mm. 

A few fragments of vertebrae and limb bones together with portions 
of hind feet representing three individuals C. M. Nos. 3374, 3374a, 
and 3376, found by Mr. Earl Douglass at the base of horizon C near 
Kennedy's Hole, were referred by myself provisionally to Paramys 
uintensis. Mr. Gidley who has recently very kindly compared this 
material with other specimens in the U. S. National Museum, advises 
me, that, if these remains are to be referred to any described species, 
he would place them with P. medius described above rather than with 
P. uintensis "since the former is about the size of P. delicatus, and this 
foot agrees also in size with that of P. delicatus, being only slightly 
more robust." The specimens are apparently somewhat large to be 
regarded as P. sciuroides, and too small and apparently of different 
proportions from P. robustus Marsh or P. (Ischyrotomus) peter soni 
Matthew.^'' The hind foot as a whole appears to be relatively broader 
and shorter, while the distal end of the metatarsals is more suddenly 
expanded and the articulations are more rounded than in the foot of 
P. robustus figured by Dr. Matthew.^^ 

The calcaneum and astragalus appear to be very similar to these 
bones in P. robustus. Among recent forms they most nearly suggest 

^'' Bull. Amer. Mus. Nat. Hist., Vol. XXVII, 1910, p. 49. 

'' In a letter of May 22, 1916, Mr. Gidley says in this connection: "While this 
foot resembles in a general way those of P. robustus and P. delicatus, the differences 
are sufficiently great to make its reference to this genus very doubtful." 



62 Annals of the Carnegie Museum. 

those of Ardomys {MarmotaY^ but with the inner keel of the astragalar 
trochlea heavier. The peroneal tubercle of the calcaneum is located 
well up, as in Marmota and Sciurns. The cuboid is possibly somewhat 
broader and lower than in P. robnstiis, while the entocuneiform appears 
to be higher, and upon comparison totally unlike that given on p. 49 
of Dr. Matthew's paper. In the latter species this bone apparently 
does not extend as high as the proximal face of the navicular. Added 
to this, regard must be paid to Matthew's statement on the same page, 
that the presence of the pre-hallux is not demonstrated in P. robustus.^ 
In the present specimen, the entocuneiform extends high above the 
navicular and terminates in a broadly rounded tubercle considerably 
greater in its diameters when compared with those of the marmot or 
Sciurns. From this round and rather smooth head I would judge that 
there was present a plantar sesamoid or pre-hallux in the case of this 
individual. The proportional size and shape of the metatarsals seem 
to agree quite well with those of the marmot, while the distal articula- 
tion is distinctly more hemispherical, in this respect disagreeing with 
those of P. rohustus illustrated by Matthew. The phalanges are 
also broader and possibly more depressed than in P. rohustus. 

Measurements. 

Length from top of astragalus to distal end of Mt. Ill 60 mm. 

Length from top of astragalus to proximal end of Mt. Ill 30 

Transverse diameter of tarsus, approximately 23 

Greatest length of astragalus 20 

Transverse diameter of trochlea 12 

Length of Mt. I 20 

Length of Mt. II 29 

Length of Mt. Ill 31 

" See Palmer "North American Fauna," U. S. Dept. Agri. Bull. No. 23, p. 400. 

2" In a letter from Mr. Gidley dated May 22, 1916, he states: "In the foot of P. 
robustus figured by Matthew, the top of the entocuneiform is broken off. But in 
the other specimen figured, which he referred to P. delicatus, this bone is complete, 
and shows the same backwardly directed ascending process on its proximal end as 
in your specimen. No. 3376. It is a little more strongly developed, however, in the 
latter. This development of the entocuneiform is usual in rodents having penta- 
dactyl feet with unreduced digits (Compare squirrels and Aplodontia). . . . 
Although the distal facet of the astragalus is broken off in your specimen, there 
seems to be evidence of its having had a pre-hallux. Matthew is certainly mistaken 
in his observation regarding this element in their specimen of P. robustus. The 
astragalar facet and the navicular both show evidence of a pre-hallux having been 
present." 



Peterson: Material Discovered in Uinta Basin. 63 

Subgenus Ischyrotomus Matthew. 21 

10. Ischyrotomus gidleyi sp. no v. 

Type: Fragment of lower jaw with My and Mt^ in place, fragment of 
caudal. C. M. No. 3461. 

Horizon: Uinta Eocene, Base of Horizon C. 

Locality: Four miles northeast of Well No. 2, eastern end of Uinta 
Basin. 

Characters of Type Specimens: Teeth relatively large: a well-formed 
basin on the inner half of the molar crown with a large exit internally. 

While the type has a suggestion of Sciuravus Marsh^^ in which genus 

this type was first placed, Mr. Gidley says: "I do not think this species 

can well be referred to Sciuravus, since the proportions and general 

features of the molars are more nearly those of the 

Paramys-groxxp. It differs from the typical Paramys, 

however, in the comparatively higher tooth-crowns 

and especially in the greater development of the Fi^- 4- Ischy- 

continuous external loph, which is deeply infolded ''«''"««^ sidleyi. 

on the outer side, ... a characteristic feature of 

seum No. 3461. 

Ischyrotomus.'" The dentition in proportion to the -^ 2/1 * 

fragment is considerably larger than that of 

Sciuravus, illustrated by Matthew {I.e., p. 59). My and Mir of the 

present species each have a well-formed basin on the inner half of the 

crowns, from which extends a large internal exit, as stated above, and 

also indicated by Fig. 4. This specimen may possibly represent a new 

genus. 

Measurements. 

Vertical diameter of ramus at Mo 6.5 mm. 

Antero-posterior diameter of Mx and Mo 7 

Antero-posterior diameter of Mj 3.5 

Transverse diameter of My 3. 

Antero-posterior diameter of Mo 3.5 

Transverse diameter of Mo 3. 

21 In the Bulletin of the American Museum of Natural History, Vol. XXVIII, 
p. 50, Dr. Matthew erected the subgenus Ischyrotomus. Mr. Gidley is of the opinion 
that Matthew's subgenus Ischyrotomus should be given full generic rank and that 
to it should be referred Paramys robustus and Paramys compressidens. 

* Outer side of teeth face toward the top of the page. 

^'^ Amer. Jour. Sci., Vol. II, 1871, p. 46; Ibid., Vol. XXIIIi 1907, pp. 124; 130. 
Bull. Amer. Mus. Nat. Hist., Vol. XXVIII, 1910, p. 59. 



64 Annals of the Carnegie Museum. 

Genus Sciuravus Marsh. 
II. Sciuravus altidens sp. nov. 

Type: Fragment of left maxillary with two cheek-teeth in position 
and the root of the third. C. M. No. 2348. 

Horizon: Uinta Eocene, Horizon B, near base. 

Locality: Between Bonanza and Kennedy's Hole, Uinta Basin, 
Utah. 

Characters of Type Specimen: Position of tritocone placed well inter- 
nally and in line with the deuterocone; the presence of a small meta- 
conule on M-; a small mesostyle on P-; type specimen representing a 
somewhat larger animal than Sciuravus nitidus. 

The most noteworthy differences which I am able to find between 
the present specimen and that of Sciuravus nitidus, illustrated by 
Dr. Matthew, is that the postero-internal tubercle (tritocone) is placed 
more internally and more nearly in an antero-posterior line with the 
deuterocone. There is also in the present specimen a metaconule 
indicated on M-. This is plainly shown in Fig. 5, and is not indicated 
in the illustration, or mentioned in the text, of Matthew's paper. 
The specimen is of about the same, or slightly larger, size than S. 
nitidus. 

Of this specimen Mr. Gidley has the following to say: "This 
specimen (No. 2348) is somewhat larger than S. nitidus, and further 
differs from that species in having the inner row, or rather the inner 




Fig. 5. Sciuravus altidens. Carnegie FiG. 6. Prosciurus Matthew. Carnegie 
Museum No. 2348. X 3/1. Museum No. 2925. X i/i. 

base of the crown in the upper cheek-teeth, considerably heightened, 
giving a decidedly bowed outline to the inner walls of these teeth. 
There is also a small mesostyle on P4 (absent in P4 of S. nitidus), and 
this cusp is relatively more prominent in the molars than in those of 
S. nitidus." 

* Outer side of teeth face toward the top of the page. 



Peterson: Material Discovered in Uinta Basin. 65 

Measurements. 

Total antero-posterior diameter of P^ and M- 5 mm. 

Antero-posterior diameter of P^ 2 " 

Transverse diameter of P- 2.3 " 

Transverse diameter of M- 3.4 " 

Antero-posterior diameter of M' 2.3 " 

Genus Prosciurus (?) Matthew.-^ 

12. Prosciurus (?) robustus sp. nov. 

Fragments of lower jaws of two individuals: C. M. Nos. 2925 and 
2926 from Horizon C of the Uinta, six miles east of Myton, Utah, are 
here provisionally referred to Matthew's genus of the lower Oligocene, 
see Fig. 6. 

In a note from Mr. Gidley, dated January 20, 1916, he says: "There 
are some peculiarities about the two forms I have provisionally re- 
ferred to Prosciurus which indicate that better material may justify 
making a new genus for these species. "^^ 

Unfortunately there are no lower teeth with the type of Proscmnis, 
and consequently no basis of comparison except size. The present 
specimens appear to be at least twice the size of Prosciurus vetustus 
Matthew. To place the specimens with the latter species would seem 
to be out of the question. On the basis of its large size and the geologi- 
cal horizon in which it is found, the name Prosciurus (?) robustus sp. 
nov. may be proposed: No. 2925 is the type and No. 2926 is the 
paratype. This is mainly in order to have all the rodents from the 
Uinta more completely recorded. 

M EASUREMENTS. 

Type Paratype 
No. 2925. No. 2926. 

Antero-posterior diameter of M 2 6 mm. 

Transverse diameter of M^ 4 mm. 

Antero-posterior diameter of M3 7 mm. 

Transverse diameter fo M3 5 mm. 

23 Bull. Amer. Mus. Nat. Hist., Vol. XIX, 1903, p. 213. 

2^ In a later communication Mr. Gidley says that he thinks that it is possible 
that the species they represent may after all have been derived from the Leptotomiis 
group of Paramys. This cannot be determined, however, until the upper teeth 
are known. 



5 — DEC. 2, 1919. 



66 Annals of the Carnegie Museum. 

Family MURID^ Alston. 
Genus Pareumys gen. nov. 
13. Pareumys milleri"^ gen. et sp. nov. 

Type: Lower jaw fragment with My and M 3^ in position. C. M. No. 
2938. 

Horizon: Uinta Eocene, Horizon C. 
Locality: Six miles east of My ton, Utah. 

Principal Characters revealed by the Type: M-^ and M-^ slightly longer 
than broad, distinctly quadritubercular with a heavy posterior marginal 
crest which extends well inward. Longitudinal crests connecting anterior 
and posterior tubercles. Mj slightly longer than M^- Animal abotit 
the same size as Eiimys elegans of the Oligocene. 

In comparing the above genus with Eumys elegans 
Leidy from the Oligocene it is to be observed that M2- 
and M 3^ are more nearly subequal in size than in the 

Fig. 7. Pareii- latter. In the present genus the molars suggest 
mys milleri.* Car- , , „ , ^ . , , , ^ 

those of huniys, but are simpler, and are character- 
negie Museum 

No 2Q10 X 3/1 '^^^'^ ^^ ^^^ absence of a fossette directly anterior to 
the protoconid; by the absence of a cross-crest from 
the protoconid to the internal median valley of the molars, and by the 
fact that the transverse diameter of the posterior portion of M-^ is 
greater than in Eumys. 

Measurements. 

Antero-posterior diameter of M 2 i-6 mm. 

Transverse " " Mj i-6 

Antero-posterior " " M3 2. 

Transverse " - " M^^ 1.2 

ARTIODACTYLA.'-<^ 
Subfamily HOMACODONTIN/E. 

Small hunoselenodont Artiodactyls of the middle and upper Eocene 
with tetradactyl manus and pes; vestigial pollcx (Bunomeryx montanus). 
Dentition |-' y- |-' f- {Bunomeryx and Hylomeryx, vide infra). Upper 
molars with small protoconule, or the latter and protocone united into a 

26 In recognition of the work on the Glires (Rodentia by Mr. Gerrit S. Miller, Jr.). 
Bull. Amer. Mus. Nat. Hist., Vol. X, 1898, p. 97. 

* Outerside of teeth face toward the top of the page. 

26 This section of this paper, dealing with the artiodactyla, was read before the 
meeting of the Paleontological Society at Pittsburgh, 19 17-18. 



Peterson: Material Discovered in Uinta Basin. 67 

cross-crest. Absence of hypocone on M-. Presence or absence of 
"hypocone" on M- and M-. A feebly developed paraconid on Mj 
{in Homacodon only). A strong tendency towards the quadricuspid 
selenodon! structure of the molar dentition. 

Genus Bunomeryx Wortman. 
14. Bunomeryx elegans Wortman (Plate XXXVII, Fig. 18). 

Bull. Amer. Mus. Nat. Hist. Vol. X, 1898, p. 97. 

Three specimens, C. M. Nos. 2951, 2949, and 3063, are referred to 
this species. The specimens are represented by lower jaws; they are 
smaller than the type specimen of B. montanus (see PI. XXXVI, 
Figs. 3-4) and in this respect more nearly agree with B. elegans 
Wortman. There is, however, a shorter diastema between P2- and P-g 
than in the latter species, and the accessory cusps of P3- are better 
developed. The tooth regarded as the canine by Wortman is of a 
trenchant character, with the apex a little recurved and elevated 
slightly above the crown of the succeeding tooth. The tooth has the 
appearance of a sub-caniniform premolar, in which event Bunomeryx 
would have four instead of three lower premolars. The tooth in 
front of this so-called canine is equally large and its crown also of the 
same detailed structure. It would therefore appear that Py in Buno- 
meryx was mistaken for a canine, while there are only two incisors. 
The latter are suddenly reduced in size and have typical fan-shaped 
crowns. 

Genus Hylomeryx gen. nov. 

Type: Anterior portion of skull and lower jaws, C. M. No. 2335. 

Paratype: Fragments of upper teeth, portion of cast of brain, and 
both mandibular rami, C. M. No. 2944. 

Horizon: Near the base of Horizon C, Uinta Eocene. 

Locality: Eastern border of Uinta Basin, near Vernal on the Uinta 
Railroad Stage-road, Uinta Basin, Utah. 

Generic Characters: I ^' Cy- Pf* ilff. Homacodon-like dentition, but 
with short diastema between P- and P-. Premolars heavier and of 
relatively greater antero-posterior diameter than in Homacodon. Cin- 
guliim of the premolars rather slightly, or not at all, indicated. Absence 
of antero- and poster o-external cuspules on P- as in Bunomeryx. Pro- 
tocone and protocunule united into a cross-crest on M- and M-. Pj with 
distinct internal tubercle; Py with deuteroconid; antero-internal tuj^ercle 



68 Annals of the Carnegie Museum. 

and heel as in Biinomeryx. Protocone and protoconule united into a 
cross-crest on M- and M-. P-g- ivith distinct internal tubercle; P4- with 
deuteroconid, antero-internal tubercle, and heel as in Bunomeryx, and a 
greater advance towards the selenodont pattern of the inferior molars than 
in Homacodon. 

The principal differences between Bunomeryx and Hylomeryx are as 
follows: the antero-external angle of M- and M- is more developed, 
giving these teeth a more perfectly quadrate outline in the present 
genus. Furthermore there is present in Hylomeryx a distinct hypocone 
on M- as in Homacodon, while in Bunomeryx there is no hypocone. 
P- and P~ of Hylomeryx are apparently also proportionally larger. 

15. Hylomeryx annectens sp. nov. (Plate XXXVI, Figs. 5-6). 
General Description of the Material Constituting the Type. 

The skull is considerably depressed by crushing, but it is possible 
to make out the outline of the nasals, which are long, rather slender, 
and extend backward to opposite the anterior border of the orbits. 
The frontal meets the nasal in a decided zigzag cross-line on the face, 
analogous to what is seen in such forms as Limnenetes anceps among 
the Oreodonts or Stenofnyliis among the Cameloids. There is no 




Fig. 8. Hylomeryx anjteclens. Carnegie Museum No. 2335. X i/i. 

lachrymal pit or vacuity, and the infraorbital foramen is small and 
located above P- as in Homacodon vagans illustrated by Dr. Sinclair.-^ 
The anterior border of the orbit is opposite the median portion of M-. 
When the skull in its present crushed condition is considered, it is 
possible to imagine that it, when normal, was similar to that of Homa- 
codon in the specimen illustrated by Sinclair. 

^'' Bull. Amer. Mtis. Nat. Hist., Vol. XXXIII, 1914, p. 285. Fig. 19. 



J 




Peterson: Material Discovered in Uinta Basin. 69 

A partial cast of the brain-case indicates a fairly large-sized brain. 
The lower jaw is well proportioned, there being considerable depth to 
the ramus throughout. The symphysis is quite strong and extends 
back opposite P-. 

From the study of the upper dentition alone one would perhaps 
hesitate in referring the Uinta remains to a separate genus. Thus 
the roots of the superior canines, which are still in position, reveal just 
such a powerful and recurved apex as is indicated in Sinclair's illus- 
tration. P-, which however is separated by a short diastema from 
P-, is larger than the preceding tooth, and has a more gentle slope of 
the posterior border; otherwise there is little or no difference between 
the two. P- agrees with the description and illustration by Sinclair 
just referred to, except that the deuterocone is somewhat better 
developed. P- appears to have the cingulum very little, if at all, 
developed; the antero- and protero-external cus- 
pules are absent; nevertheless on the whole the 
tooth closely resembles that of the specimen of 
Homacodon described by Sinclair. The most pro- 
nounced modification of the molars is the union 
into a cross-crest of the antero-internal and me- 
dian tubercles, which is especially well accom- 
plished on M- and M- of the type, but which 
cannot be said to have taken on an advanced 
selenodont pattern, as is seen in Bimomeryx Wort- 
man.28 On the contrary these crests are very Fig. 9. Hylomeryx 

, , ... . , , ... • 1- • r annectens. Carnegie 

penssodactyl-like with the slightest indication 01 ,^ 

, , Museum No. 2944. 

the intermediate tubercle on an unworn or x i i I2. 
slightly worn tooth, (see Fig. 9, M- and M-). 

This condition is quite clearly foreshadowed in Homacodon where the 
intermediate closely crowds the antero-internal tubercle. 

When the study and comparison of the inferior dentition is taken 
up, it is clear that a considerable modification toward the selenodont 
pattern of the tooth-structure has taken place in these Uinta remains, 
which at once separates them from the Bridger form. In the paratype, 
No. 2944, both rami are preserved (the left with the symphysis com- 
plete), containing two incisors represented by moderately large 
roots subequal in size.-^ The canine is a large trenchant tooth with 

^^ Bull. Amer. Mus. Nat. Hist., Vol. X, 1898, p. 100, Fig. 2. 

25 Unfortunately there are as yet no incisors known in Homacodon, but I judge 
that genus to have two, as in the present form. 



70 



Annals of the Carnegie Museum. 



the apex slightly recurved. P- is single-rooted as in Homacodon 
and has a trenchant crown. P- has a considerably greater fore- and- 
aft diameter, but is otherwise quite similar to the preceding tooth. 
P- differs from that in Homacodon by the presence of an internal 




Fig. 10. Hylotneryx annectetjs. Carnegie Museum No. 2944. X i 1/2. 

tubercle situated well back, which may be regarded as the deutero- 
conid. P- is entirely different from that tooth in Homacodon and 
more nearly suggests that in Bunomeryx. Thus the deuteroconid is 
prominent,^" there is a decided antero-internal tubercle and a heel 
differing both from Professor Marsh's illustration {Amer. Jour. Sci., 
Vol. XLVTII, 1894, p. 262, Fig. 6), and Sinclair's description {I.e., 
p. 284). Marsh has stated that this tooth in Homacodon vagans has a 
simple crown, while Sinclair, without criticizing Marsh, says that there 
is no deuteroconid, but a prominent anterior tubercle and almost as 
prominent a posterior basal tubercle rising from a cingulum-like heel. 
The lower molars are of a type further advanced toward the seleno- 
dont structure than those of Homacodon. 

M EASUREMENTS. 

Type Paratype 
No. 2335. No. 2926. 
Antero-posterior diameter of skull from anterior border of 

orbit to canine 32 mm. 

Length of cheek dentition 39 " 38 mm. 

Length of superior premolar series 22 

Length of superior molar series 16 " 

Antero-posterior diameter of P- 3 " 

^ In the type, which pertains to a younger individual than the paratype, the 
deuterocone and antero-internal tubercle are less prominent; less distinctly separ- 
ated from the protoconid, and situated higher up on the crown. 



Peterson: Material Discovered in Uinta Basin. 



71 



Antero-posterior diameter of P'^ 5 

P^ 5 

Transverse diameter of P^ 5 

P^ 5 

Antero-posterior diameter of " 5 



Transverse . 



Antero-posterior 



M' 



M^ 



M' 



5-5 
6 

6.5 
6 

5-5 
6 



Transverse 

Vertical diameter of mandibular ramus at P3 9 

Vertical diameter of mandibular ramus at M3 12 

Antero-posterior diameter of inferior premolar series 

Antero-posterior diameter of inferior molar series 19 



5. 5 mm. 
6.5 " 
7- " 
6 

5-5 " 



20 
18 



For other measurements see text-figures 8 and 9, and PI. xxxvi, figs. 
5,6. 

Genus Sphenomeryx gen. nov. 

16. Sphenomeryx quadricupsis gen. et sp. nov. (Plate XXXVII, 
Figs. 15-16). 

Type: Fragments of upper and lower jaws with teeth, C. M. No. 
2346. 

Paratypes: Surface fragments of two or three individuals found 
together, representing brain-casts and other mutilated parts of limb- 
and foot-bones, together with a fragment of a lower jaw with P- and 
the molars represented, C. M. Nos. 2914, 2915, 2926. 

Horizon: Uinta Eocene. Base of Horizon C. 

Locality: Type found about two miles south of Kennedy's Hole to 
the west of old Vernal- Dragon Stage- road in eastern end of Uinta 
Basin. Paratypes found six miles east of Myton, Utah. 

Generic Characters, as shown by the Type and Paratypes: M- without 
postero-internal tubercle, but instead a heavy ledge or cingulum, which 
extends around the internal face of the tooth. A transverse ridge-like 
protocone on -wear of the tooth as in Mesomeryx. Paracone conical, 
metacone subconical. External faces of para- and metacones sharply 
convex from side to side, as in Mesomeryx. Small parastyle, meso- 
and metastyles absent, and instead a heavy cingulum on the external face, 
as in Hylomeryx and Mesomeryx. Premolars heavy; deuterocone of P- 
very small and placed well back. Lower teeth like those in Bunomeryx 
and Hylomeryx. Animals of same size as Hylomeryx. 






72 



Annals of the Carnegie Museum. 



The present genus, as well as Hylomeryx, has the para- and meta- 
cones of M- more nearly conical than is the case in either Bunomeryx 
or Mesomeryx. In the latter there are, however, no mesostyles, while 
in Bunomeryx montanus they are fairly well represented. The large 
premolars of the present genus suggest Hylomeryx, while the minute 
deuterocone on P- and the absence of the postero-internal tubercle 
of M- separates the genus from both Bunomeryx and Hylomeryx. 
The present new genus is apparently the first of the Homacodontince 
marked by the absence of this hypocone on M-. The heavy internal 
ledge of the postero-internal angle of the crown fills out this portion 
of the tooth so that it has an outline practically as quadrate as that of 
M^ in Hylomeryx. I regard the present genus as being very close 
to Hylomeryx because of the similar meta- and paracones, the absence 
of the mesostyle, and the robust premolars. The protocone and 
protoconule have undoubtedly united into a solid cross-ridge on M-. 
This supposition is expressed by the specific name, while the generic 
name serves to express the inter-relationship of this form with prac- 
tically all the known genera of this subfamily. 

The brain-casts of No. 2915 and 2926 are of quite large size, with 
well marked convolutions, a distinct or rather deeply constricted area 
between the cerebrum and cerebellum. The medulla oblongata is 
also quite large. In No. 2915 there is still adhering to the cast a 
portion of the parietals, which display a sagittal crest of considerable 
prominence and length. The lower jaw of the paratype. No. 2915, 
shows the same generally robust structure, which is seen in the type. 

M EASUREMENTS. 

Type No. 2346^ 

Antero-posterior diameter of P- to M- inclusive 16.5 mm. 

" p3 6 

Transverse " " " 4 

" P^ 6 

Antero-posterior " " " 5 

"M' 5.5 " 

Transverse " " " 6 

Antero-posterior " " P3, M-^ approximately 32 

" P4 6 " 

Transverse " " " 3-5 

"Ml 4 

Antero-posterior " " " 5-5 

" M^ 5-5 " 

Transverse " " " 4 

" M^ 4 " 

Antero-posterior " " " approximately 7 



Peterson: Material Discovered in Uinta Basin. 73 

Genus Mesomeryx gen. nov. 

Type: Fragment of maxillary with cheek-teeth in place, C. M. No. 
31S9. 

Horizon: Uinta Eocene. Lower C. 

Locality: Near eastern border of the Upper Eocene sediments, two 
miles east of Dragon-Vernal Stage-road, Uinta Basin, Utah. 

Generic Characters: Molars with sharp external convexity of para- 
and metacones ; the latter tubercles distinctly connected by a fore-and-aft 
ridge near the external face of the tooth, prominent parastyle, weak meta- 
style, absence of mesostyle. Protocone formed into an oblique cross- 
crest on wear. Subselenodont structure of postero-internal tubercle. 
P- with weak, and P- with strong, deuterocone. 

17. Mesomeryx grangerP^ sp. nov. (Plate XXXVII, Fig. 17). 

General Description of the Type Specimen: The type represents an 
animal smaller than Bunonieryx elegans Wortman. The dentition 
has, however, advanced a step further than in the latter genus. This 
is especially seen in the molar teeth. The protocone is united with 
the protoconule, so that the two form an oblique forward and outward 
extended ridge on the surface of a worn tooth as in Hylomeryx. On 
close examination this ridge is seen to have a slight constriction, but 
whether or not there was a separation between protocone and pro- 
toconule, as in the Homacodonts generally, cannot be stated (most 
likely there was only a solid cross-ridge). There is no postero-internal 
tubercle on either M- or M-. The present genus appears in this 
respect to resemble the one just described, but in the present form the 
para- and metacones are decidedh' less conical and the whole structure 
of the crown more selenodont. If this postero-internal tubercle (hypo- 
cone?) had ever existed in this phylum, as for instance in Homacodon, 
Hylomeryx, Bunomeryx, and Sphenomeryx, it had already been com- 
pletely crowded out, absorbed, or otherwise replaced by the meta- 
conule in the Bridger or earlier genera. From what we have just seen 
in the true Homacodon phylum it would appear to add much strength 
to Dr. Wortman's views of the process of the development of the 
quadricuspid selenodont from the quinquicuspid bunodont molar.^^ 
There is a well-developed cingulum posteriorly on both M- and M-, 

'1 In recognition of Mr. Walter Granger of the American Museum of Natural 
History, New York. 

^^ Bull. Amer. Mus. Nat. Hist., Vol. X, 1898, p. loi. 



74 Annals of the Carnegie Museum. 

which unites with the rather well-developed parastyle, while in the 
posterior molars the cingulum is rather poorly developed, as is also 
the metastyle. There is no mesostyle, but a well-developed ridge 
near the outer face of the crown, which effects a connection between 
para- and metacones, unlike what is seen in Bunomeryx or Hylomeryx. 
Internally at the exit of the median valley there is (especially on M-) 
a heavy, smooth cingulum, which is most suggestive of similar cingula 
on the molars of Helohyus and of such later Tertiary selenodont forms 
as Dromomeryx, Palceomeryx, and Dicrocerus. P- has the proto- 
and deuterocone of equal size and a heavy cingulum anteriorly and 
posteriorly, which terminate externally in small basal tubercles. 
P- has a large protocone with a quite trenchant ridge extending back- 
ward and sloping gradually. The tooth is surrounded by a cingulum 
especially well-developed internally, so that one may say there is a 
rudimentary deuterocone. This deuterocone or tubercle is located 
further forward on the tooth than in Bunomeryx or Hylomeryx. 
P- is represented by two roots. 

Measurements. 

to M^ i6 mm. 

4-5 

3-7 

5 

3-5 

- 4-5 

- 5 

" 6 

• 5 

Phylogeny. 

There is but little doubt that Mesomeryx belongs to the subfamily 
HomacodontincB. From the characters of the molars one would not 
long hesitate in placing the genus near Sphenomeryx or Bunomeryx. 
The consolidation of the antero-internal and antero-median tubercles 
into a cross-ridge as in Hylomeryx, and the absence of the hypocone 
on M- as in Sphenomeryx, may in this genus be looked upon as repre- 
senting a line, which paralleled the evolutionary stages in the Homa- 
codon-Sphenomeryx phylum. 

It_is quite evident that the foregoing genera are most nearly related 
to the Bridger genus Ho^nacodon. That these upper Eocene Artio- 
dactyls of America hold a position relatively well differentiated from 



Antero-posterior 


diameter P- 


" 


pa 


Transverse 


P- 


" 


p4 


Antero-posterior 


pi 


" 


M 


Transverse 


M 


" 


M 


Antero-posterior 


M 



Peterson: Material Discovered in Uinta Basin. 75 

the Hypcrtragulids, the Camelids, the Oreodonts, and other Artio- 
dactyls must also be admitted. In my judgment these Uinta genera 
as well as Ilomacodon constitute a sub-family distinct from the Dicho- 
bunids of Europe.^^ They represent an American branch, whose com- 
mon ancestors, no doubt, also gave rise to those of Europe. Upon 
the whole the details of structure in the dentition differ considerably 
in the genera representing the two regions. This is especially notice- 
able when such European genera as Mouillacitherium and Metrio- 
therium^'^ are compared with the Uinta forms. The protocone sends a 
spur backwards, which has a tendency to close the cross-valley in the 
European genera, while the valley is clear in the American genera, 
analogous to what is seen in Oxacron and Ccenotheriuni of Europe. 
The hypocone of the European genera appears to have a greater 
functional value, that is: it is of proportionally larger development 
and apparently expresses a greater degree of permanency, which is 
especially emphasized in the Oxacron-Ccenotherium phylum. These 
phyla are, however, not recognized by Stehlin, and others, as having 
any especial relation to the true dichobunids. 

If we regard the Homacodon-Hylomeryx-Btaiomeryx-Sphenomeryx 
and the Mesomeryx phyla as at all closely related, we have clearly a 
tendency toward the quadricuspid condition in the American forms. 
In Dichohune according to Stehlin's reconstruction {I.e., p. 604; 607) 
we have a skull proportionally longer, lower, and narrower, with the 
premaxillaries heavy, in order to support the large incisors. The upper 
canine has specialized in size and shape, so that it differs much from 
that in Homacodon and Hylomeryx. 

So far as I am aware, the HomacodontincB are not represented in the 
Oligocene or later epochs of North America. Dr. Sinclair's opinion 
{I.e., pp. 294-295), that several divergent lines of the bunodont Artio- 
dactyls are already established in the lower and middle Eocene of 
North America is altogether quite likely. It is also probable that 
some of the lower Eocene genera, already partially known, may prove 
to be in the line of the HomacodontincB. 

33 From our present knowledge of Bunophorus (Bull. Amer. Mus. Nat. Hist., 
Vol. XXXIII, 1914, p. 273) it is perhaps premature to here include this Wasatch 
genus, which may, however, be a forerunner of Homacodon ,as Dr. Sinclair suggests. 

34 Stehlin, G. H., Abhand. Schweiz. Paleont. Gesellsohaft, Vol. XXXIII, 1906. 
pp. 628, 661. 



76 Annals of the Carnegie Museum. 

Family ANOPLOTHERIID^. 

A great surprise encountered in the study of the collection from the 
Uinta Eocene made in 191 2 is the discovery of a genus, which repre- 
sents the family Anoplotheriidce hitherto only known from the old 
world. Only surface fragments were found, which are apparently 
mixed with remains pertaining to other typical Uinta genera, but 
enough has been brought together from the lot to certainly establish 
the features, especially of the limb and foot structure, of an Ameri- 
can anoplothere. There was found a second individual in the same 
locality and horizon, which also consists of foot-bones and other frag- 
ments of the skeleton. ^^ These fragmentary remains (especially those 
of the limbs and feet) appear to agree most closely with the genus 
Diplohiine of the European Tertiary. 

Genus Diplobunops gen. nov. 
18. Diplobunops matthewi^^ gp ^ov. (Plate XXXVII). 

Type: Numerous fragments of the skeleton, C. M. No. 2974. 
Paratype: Foot-bones and other fragments of the skeleton, C. M. 

No. 3394- 

Horizon: Uinta Eocene, Horizon C. 

Locality: Six miles east of Myton, Utah. 

Principal Characters obtained from the Type and Paratype: radius 
and ulna short and stout; carpus and tarsus relatively low and broad; 
articulation of proximal phalanges convex, distally carnivore-like; 
unguals high, claw-like, compressed posteriorly and superiorly, and 
suddenly expanded along the plantar border in front of the subungual 
process; animals slightly smaller than Diplobune quercyi of Europe. 

The fragments of the skull and lower jaws which were found together 
with the [Paratype of Diplobunops, C. M. No. 3394, are so nearly 
similar to corresponding parts of other Uinta protoreodonts that it is 
only provisionally that they are referred to the paratype of this new 
genus. Indeed these skull and jaw fragments were first given a separ- 
ate catalog number and referred to Protagriochcerus annectens Scott.^^ 
A closer and more critical study of the specimen in connection with 
Diplobune quercyi reveals two hypotheses: of which the first and most 
probable is, that the remains of a new species of Protagriochcerus in 

'^ The fragments of the skull and lower jaws are provisionally referred here. 

5^ In recognition of Dr. W. D. Matthew. 

^' Trans. Wagner Free Institute of Science, Vol. VI, 1899, p. 100. 



Peterson: Material Discovered in Uinta Basin. 77 

some way got mixed with the limb- and foot-bones of Diplobunops; 
the second is that possibly the present new genus and the protoreodonts 
possessed a dentition more closely similar in structure than might have 
been anticipated. 

The upper dentition is almost completely demolished in the speci- 
men under consideration. However, a few important characters are 
ascertainable. The alveolar border is longer than in the type of 
Protagriochcenis anftectens, which is due to the relatively longer pre- 
molar region, a character which suggests Diplohune. The root of the 
upper canine indicates that it is like the same tooth in Protagriochcerus, 
but F- is more isolated and recalls the condition in some species of 
Agriochcerus. The inner portions of M- and M- are preserved and 
show that the postero-internal crescents have more acute angles on 
the inner face than in Diplohune and thus are more like these teeth in 
ProtagriochKriis, and also that the intermediate tubercle (protoconule) 
is smaller than in the European genus. 

Although the mandible of Diplobiine is longer and slenderer, the 
premolars longer and better developed behind than in the present 
specimen, there is a surprising similarity in the dentition of the two 
forms, so far as comparison can be made by means of the material 
at hand. While the internal tubercles of the molars are grooved on 
the inner face, I judge that the anterior internal tubercles were not 
twinned, at least not to the same extent as those in Diplobiine. In the 
latter genus the inner faces of the molars are less rugose than in the 
specimen under description, while in some species of Protoreodon 
this feature appears to be similar to that in Diplobunops. M- has a 
well developed fifth cusp as in Diplobiine and in the Oreodonts 
generally. 

The distal end of the scapula has a general resemblance to that bone 
in Diplobtme. The distal end of the humerus has the same low and 
broad anconeal fossa, the large entepicondyle, the broad trochlea 
with the great convexity of the intertrochlear ridge, and the sharp 
and well defined external ridge, which articulates with a corresponding 
facet of the radius. The latter has an expanded head as in Diplohune. 
The ulna is short and stout, with a remarkably short and heavy 
olecranon process. 

The material at hand clearly indicates that the carpus is lower and 
relatively broader than in Diplohune. The scaphoid is broader, but 
the antero-posterior and vertical diameters are less, and the distal 



78 Annals of the Carnegie Museum. 

articulating faces less distinct from one another than in the European 
genus. The lunar upon the whole is perhaps more suggestive of 
Anoplotherium inasmuch as the proximal articulation extends back- 
ward and downward in a similarly gentle slope to very nearly the 
palmar face without the sudden downward pitch, which is seen in the 
posterior half of this articulation in Diplobune. The distal articula- 
tions for the magnum and unciform, on the other hand, are more 
subequal than in Anoplotherium and in this respect the lunar in the 
present genus is perhaps more like Diplobune. The dorsal face of the 
cuneiform in the latter genus is of uniform height, while in the present 
genus it is highest radially and decreases in the ulnar direction, due 
mainly to the upward turn of the unciform facet in the ulnar region. 
(See PL XXXVII, Fig. 17.) The posterior portion of the lunar 
facet on the unciform of No. 3394 is extremely convex from side to 
side, and terminates supero-radially in a blunt cone, quite unlike 
what is seen in either Diplobune or Anoplotherium, but the general 
characters of the unciform are more nearly like those of that bone in 
Diplobime. 

The metapodials which have been associated with the type have 
their distal articulation for the proximal phalanx very convex and 
carnivore-like. Mc. II is represented by the upper end in both type 
and paratype. This bone is slenderer in proportion than in Diplobune, 
but as in the latter genus there is a facet for Mc. I. The phalanges 
of the proximal and median rows are broad and depressed, while a 
terminal phalanx, belonging to specimen No. 3394, indicates that these 
elements are high, laterally compressed, and claw-like, with a sudden 
broadening along the plantar borders in front of the sub-ungual 
process, and that there is a large nutrient foramen on either side near 
the plantar face of the bone. The bone closely resembles that in 
Diplobune. 

Enough is preserved of the astragalus to indicate that it was low 
and broad as in the European anoplotheres. The calcaneum is better 
preserved and only in the more minute details does it differ from that 
of Diplobune. In the latter genus the internal or tibial face of the 
tuber calcis is less convex, the peroneal tubercle less developed, the 
eminence on the dorsal border which articulates with the fibula is 
smaller, and the facet for the cuboid is less extensive both laterally 
and antero-posteriorly. The cuboid is also very suggestive of Diplo- 
bune, though broader, lower, and having the astragalar and calcaneal 



Peterson: Material Discovered in Uinta Basin. 79 

facets more unequal in size; tliat for the calcaneum being much the 
broader of the two. 

The genus is provisionally referred to the subfamily Anoplotherince 
pending the discovery of more complete material. From the unusually 
low tarsus together with the hemispherical or carnivore-like distal 
articulation of the metapodials it is altogether likely that the American 
genus represents a distinct subfamily (Diplobunopsincs) which may 
be more satisfactorily differentiated from European diplobunids 
upon further discovery and verification of dental and cranial characters. 

Measurements. 

Type No. 2974. 

Scapula, distal end antero-posteriorly approximately 45 mm. 

transverse 25 " 

Humerus, distal end greatest transverse diameter 62 " 

greatest antero-posterior diameter 24 " 

Radius, proximal end greatest transverse diameter 32 " 

greatest antero-posterior diameter 20 " 

Ulna olecranon process, length 19 " 

length of shaft, approximately 112 " 

Carpus greatest transverse diameter 42 " 

vertical diameter at ulnar face 9 " 

Scaphoid transverse diameter 13 " 

antero-posterior 19 " 

Lunar transverse diameter 14 " 

antero-posterior diameter 18 " 

Cuneiform, transverse diameter 18 " 

Calcaneum, total length, approximately 60 " 

Cuboid transverse diameter 26 ." 

vertical diameter at fibular face 9 " 

Family ACH^NODONTID^. 

Family Characters: Dentition: f • \' f • f ; hunodont. Orbits not en- 
closed posteriorly; limbs short; feet tetradactyl,^^ animals the size of a 
wild boar to nearly that of an Hippopotamus. 

In this family is included Achcenodon Cope and Parahyus Marsh. 

Genus Ach^nodon Cope. 
19. Achaenodon insolens Cope. 

Palaeontological Bulletin, No. 17, 1873, p. 2. 

Two mandibular rami, C. M. Nos. 2309 and 3183, are referred to 
this species. No. 2309 is of the same size as the type described by 

'' The characters of the limbs and feet are obtained from Professor Osborn's 
publication (Bull. Amer. Mus. Nat. Hist.. Vol. VH, 1895, p. 105). The skull and 
limbs have not as yet been found together in any representative of this family. 



80 Annals of the Carnegie Museum. 

Cope, while No. 3183 is somewhat smaller. These specimens are 
easily separated from the type of Achcenodon robustiis Osborn by 
their slenderer and shallower form and by the relatively longer molars. 
Furthermore the heel of P^ in A. insolens, as exhibited by the Car- 
negie Museum specimen No. 2309 (see Fig. 11) is more strongly de- 




FiG. II. Achcenodon insolens. Carnegie Miiseun, No. 2309. X 1/4. 

veloped than in A. robust us. The premolars in A. robustiis are appar- 
ently larger than in A. insolens, which, however, may perhaps be 
partly due to imperfection of the premolar teeth in No. 2309, which 
are extensively repaired with plaster. The coronoid process of A. 
insolens appears to be more everted and the summit sharper than in 
A. robustus. 

Both of these mandibular rami were found by Mr. Earl Douglass in 
Horizon B of the Uinta sediments. 

20. Achaenodon uintense (Osborn). 

Bull. Amor. Mus. Nat. Hist., Vol. VII, 1895, p. 102. 

This species is represented by two crania, C. M. Nos. 2160 and 3182, 
found by ]\Ir. Earl Douglass in the same horizon (B.) and locality 
(eastern portion of Uinta Basin) in which the type of the species was 
obtained. In a previous publication, Protelotherium was referred to 
Achccnodon.^'^ As the result of recent careful study of the Uinta and 
Washakie forms I am strengthened in the opinion that they should 
be kept under one generic name. When a liberal allowance for the 
complete premaxillary and a correction of the frontal region (See 
PI. XXXIX, Fig. 1) in A. robustus is made, there does not appear to 

'^ Memoirs Carnegie Museum, Vol. IV, 1909, p. 145. 



Peterson: Material Discovered in Uinta Basin. 



81 



be as grciit a difference in the length of the face between A. robustiis 
and A. nintense as appear in the original illustration and as stated by 
Professor Osborn.^ 

This character together with the position of the orbit was until 
very recently regarded by the present writer as possibly having 
generic importance. A thorough review of the type of Achcenodon 
robustiis reveals the fact that the specimen has received considerable 
lateral crushing in the region of the frontals and anterior portions 
of the parietals, which no doubt is at least a partial cause for the 
apparently greater elongation of the cranium than in A. uintense: 
When the orbit of the left side is carefully studied in relation to the 
top of the skull it becomes plain that the characters in the type of 
A. robustiis and the Carnegie Museum specimens of A. nintense 
are almost identical; that is, the orbit is situated nearer the dorsal face 
of the frontal than is the case in the illustration by Professor Osborn.^^ 
This is corrected in PI. XXXIX, Fig. i of this paper. The end of the 
muzzle in the Princeton specimen is, however, apparently heavier 
than in A. nintense, which is possibly also partly due to crushing. 
The chief points of difference between these species then are: the 
much greater development of the posterior accessory tubercle of M- 
the thicker premolars and the larger size of ^. nintense when compared 
v/ith A. robustiis. (See illustrations of M^ PI., XXXIX, Figs. 2 
and 4.) 

In skull No. 3182 the occiput is perfectly preserved and for the 
first time gives us an accurate conception of this region. In Professor 
Osborn's paper on A. nintense {I.e., p. 104) the occipital plate is repre- 
sented as fan-like, while the specimen in the Carnegie Museum shows 
that it has a more evenly rounded appearance from side to side (see 
PI. XLVn, Fig. 3). 

Genus Parahyus Marsh. 
21. Parahyus vagus Marsh. 

Amer. Jour. Sci., Vol. XII, 1876, p. 402; Amer. Jour Sci., Vol. XLVIII, 1894, p. 
261. 

From casts of the types kindly communicated by the authorities of 
the Peabody Museum of Natural History, it is possible to determine a 
few significant characters, which may be regarded as possessing generic 

^"i.c, p. 103. 

^1 Contributions from The E. M. Museum of Geology and Archaeology of Prince^, 
ton College. Bull. No. 3, 1883, pi. VI. 

6 DEC. 15, I919. 



r-'-' 



82 



Annals of the Carnegie Museum. 



importance. When compared with Achcenodon the teeth of Parahyns 
vagus are proportionally larger in relation to the size of the jaw, 




Fig. 12. Parahyus vagus. From a cast in the Carnegie Museum. No. 3448. 
Original type specimen in Peabody Museum of Natural History. 

which is much slenderer and has a different contour from that of 
Achcenodon, the under border being straighter fore-and-aft. Dentition 
I?3". Cy, P3, M3-. Premolars more compressed laterally than in 
Achcenodon, especially A. uintense. The dentition is proportionally 
longer, M-g- being one third narrower while of the same actual length 
as in A. rohustus. The cross- valleys between the posterior and 
anterior tubercles of the molars are wider and the heel of Mg- is more 
distinctly separated from the main body of the tooth than in Achce- 
nodon. 

In the type of Parahyus aberrans Marsh the external tubercles of 
the upper molar possess more conical symmetry than in Achcenodon, 
and, as in the lower teeth of the type of Parahyus vagus, the tubercles 
of the upper tooth of Parahyus aberrans are also separated by more 
clearly defined valleys than in Achcenodon. 

Family AGRIOCHCERID.E. 

Genus Protoreodon Scott and Osborn. 

22. Protoreodon medius sp. nov. (Plate XL, Figs. 1-16). 
Type: Greater portion of the skeleton, C. M. No. 2962. 
Horizon: Uinta Eocene, Horizon C. 

Locality: Six miles east of My ton, on the Duchesne River, Utah. 
Besides the type specimen there are numerous fragmentar\- remains, 



Peterson: Material Discovered in Uinta Basin. 83 

which are more or less doubtfully referred to the above species. Of 
these may be mentioned No. 2917, a pair of lower jaws mutilated in 
front, No. 2933, a left lower jaw, No. 3020, a portion of a skeleton 
including nearly the complete tail, and No. 3038, also a portion of a 
skeleton. C. M. Nos. 2987 and 3067 may possibly belong to a dif- 
ferent (new) species on account of the rather small skull and large 
teeth, but as the specimens pertain to young individuals, I i)refcr to 
place them with P. medms. 

Specific Characters: If- Cy- P^' P- isolated by a diastema, upper 
molars with reduced intermediate cusps. Pes relatively long. Animal 
considerably larger than Protoreodon parvus, P. pumilus, P. paradoxicus, 
or P. minor. 

The cranial region is long and the face short as in P. paradoxicus 
Scott, but the infra-orbital foramen appears to be located further 
back than in either P. paradoxicus or P. parvus. The premaxillary 
is of well-proportioned size, and, as already stated, there are three 
incisors present in the skull. The premolars have the same develop- 
ment as in P. parvus, but the upper molars have advanced a step, 
the anterior intermediate cusp being reduced. 

The present species is probably from a later horizon than those 
heretofore described. In fact it appears that even the fragmentary 
remains from this locality, referred to P. parvus and other known species 
of the Uinta, are possibly further advanced, especially with regard 
to the reduction of the anterior intermediate cusp of the molars, which 
is a mere remnant in nearly all of the specimens, in which the upper 
molars are preserved. 

The new species here proposed is by far the most abundantly 
represented in this new fossil locality of the Duchesne Valley, the 
smaller species being comparatively few in number. 

In general detail the skeleton answers very well to the descriptions 
furnished by Professor Scott and need not here be repeated. However, 
the illustrations herewith reproduced will serve as a correct guide to 
proportions, as the type of P. medius is represented by practically 
all the main parts of the skeleton, thus enabling us for the first time to 
effect a restoration of an Oreodent from the Uinta with approximate 
correctness. 

Unfortunately the carpus is not represented in the type. There 
are, however, two metacarpals which give a fair idea of the proportions 
of the fore and hind foot. A feature at once noticeable is the relatively 



84 Annals of the Carnegie Museum. 

long pes, but whether this is a specific character, or whether it holds 
good throughout the genus Protoreodon, cannot be determined until 
more complete material of other species is secured. 

In this connection it is well to draw the attention of students to 
the fact that Professor Scott's figure 22 of Plate 3 of his Uinta report 
{Trans. Wag. Free Inst, of Science, Vol. VI, 1899) very probably repre- 
sents a hind foot of this species. The hind foot and other fragments 
of limbs illustrated in the " Mammalia of the Uinta Basin," Plate VII, ^^ 
also possibly pertain to Protoreodon medius. 

M EASUREMENTS. 

Type No. 2962. 

Length of skull (occiput to and including premaxillaries) 173 mm. 

" (occiput to end of nasals) 168 

" " cheek-dentition (upper) 73 

" canine to M- 29 

" " molar series 37 

Greatest length of mandible 127 

Length of cheek dentition (lower) 74 

Depth of ramus at M^ 31 

Length of humerus head to distal end 137 

Greatest transverse diameter humerus, distal end 27 

Length of Mc. IV, approximately 43 

Greatest antero- posterior diameter of humerus, distal end 19 

" length of femur 153 

" tibia : . . . 143 

" " " astragalus 26 

" " " calcaneum 50 

Height of tarsus including astragalus 37 

Length of Mt. IV, dorsal measurement 65 

Restoration of Protoreodon medius (Plate XLI). 

From the study of the articulated skeleton of Protoreodon medius 
it becomes plain that the genus combines some of the characters of 
Merycoidodon and of Agriochccrus of the White River Oligocene. Thus 
the neck, the trunk, and the caudal region almost duplicate those of 
Agriocha'rus, while the structure of the limbs, and especially of the 
foot, more nearly suggests Merycoidodon. Professor \V. B. Scott long 
ago pointed out very clearly the relationship between Protoreodon, 
AgriochKrus, and Merycoidodon.'^^ Indeed as early as 1875 Professor 

^2 Trans. Amer. Philos. Soc, 'Vol. XVI, 1889, PL \'II. 

''5 "The Mammalia of the Uinta Formation," Amer. Philos. Soc, Vol. XVI, 
Part II, 1889, pp. 487-503. 



Peterson: Material Discovered in Uinta Basin. 



85 



O. C. Marsh saw the close rehilionship between these Uinta and Oligo- 
cene genera and even referred one specimen to Agriocharus {A. 
puniiliis)^'^ 

The present restoration is effected from the remains of the type 
specimen No. 2962, with the exception of the posterior portion of the 
tail, which is reproduced from the paratype No. 3020. The portions 
on Plate XLI indicated in outlines are obtained from other individuals 
and are thought to be approximately correct. 

In this, connection mention should be made of the fact that the 
vertebral column was found dislocated at the first dorsal (see fig. 13), 




Fig. 13. Proloreodon medius. Carnegie Museum, No. 2962. X 1/4. Vertebral 
column in the original position as found in the field. 



and that one or possibly two vertebrae in this region are wanting. 
The posterior dorsals and all the lumbars, the sacrum, and the pelvis 
were found in position. The skull and atlas were in nearly their 
proper position with reference to the location of the neck, while the 
limbs and feet were more or less dislocated. 

From this it is plain that we cannot be positive in regard to the 
number of the vertebrae in the dorsal region. Thirteen dorsals, the 
same number as in Merycoidodon, are given in the illustration, which 
may or may not be correct. On the other hand there is no doubt in 
regard to the lumbar and the sacral series which are six and three 

'^* Amer. Jour. Sci., Vol. IX, 1875, p. 250. 



86 Annals of the Carnegie Museum. 

respectively. In Agriochcenis we have the same number, while in at 
least two complete skeletons of Merycoidodon {M. culbcrtsoni of the 
Carnegie Museum and M. gracilis of the U. S. National Museum) 
there are seven lumbars and four sacrals.^^ The caudal region is 
represented by twenty-three vertebrae and is equal to or perhaps even 
exceeds in length that of Agriochcerus. The thorax is not large and the 
clavicle was, no doubt, relatively larger than in Merycoidodon. 

The limbs are, as already stated, most suggestive of Merycoidodon, 
though longer and slenderer. The poUex is slightly larger than in 
Merycoidodon. There is no evidence of a hallux in the pes of the species 
under description. The animal was perhaps a better runner than the 
Oligocene genus. 

M EASUREMENTS. 

Total length of skeleton from premaxillary to tip of tail measured along 

the curves, approximately 1130 mm. 

Height of skeleton at fore limbs, approximately 414 

Height of skeleton at hind limbs 442 " 

Genus Protagrioch(erus Scott. 
23. Protagriochoerus annectens Scott. (Plate XL, Figs. 19-27). 

Trans. Wagner Free Institute of Science, Vol. VI, 1899, p. 100, PL IV, Figs. 26-29. 
C. M. No. 3016, the specimen referred to this genus, was found in 
horizon C, near Myton, Utah, and consists of the anterior portion of 
both maxillaries together with numerous other fragments of the skele- 
ton. The roots of the canine and P- agree in size and position with 
those of the type of Protagriocluvriis. The scapula is represented by 
the glenoid cavity and the coracoid, which are similar in structure to 
Protoreodon and the Oligocene oreodonts generally. The distal end 
of the humerus which is preserved is also identical with that of Pro- 
toreodon, and in a general way agrees with that element in the Oligocene 
oreodonts, but is entirely unlike the lower end of the humerus in 
Agriocha'riis. When compared with the latter the entepicondyle is 
less developed, the trochlea itself narrower, with the inner and outer 
condyles more nearly subequal in size and divided by a narrower and 
more [jrominent intertrochlear ridge. On the whole tliis portion of 

*'^ A communication recently received from Mr. Paul C. Miller of the Walker 
Museum, Chicago University, states that in a remarkably complete skeleton of 
Merycoidodon Culberlsoni in that institution recently freed from the matrix there 
are: seven cervicals, thirteen dorsals, seven lumbars, three sacrals and nineteen 
caudals. 



Peterson: Material Discovered in Uinta Basin. 87 

the humerus is not unlike that in Prolorcodon or the oreodonts in 
general. The olecranon process of the ulna is short and stout, but 
has apparently a smaller antero-posterior diameter and a less decided 
tendinal groove than is usual in the oreodonts of the Oligocene. 

Fragments of the hind limb present no noteworthy differences from 
corresponding parts in Protoreodon or the later oreodonts {Merycoi- 
dodon). Proportionally the distal end of the femur has possibly 
a somewhat greater transverse diameter than in the Oligocene oreo- 
donts and the rotular trochlea is shallower and broader. The head 
of the tibia is broad, corresponding to the distal end of the femur. 
The patella appears to have a greater dorsal convexity, or greater 
antero-posterior diameter, than that bone as described by Professor- 
Scott. At the same time it is possibly flatter than is usually the 
case in most oreodonts. 

In the type of Protagriochcerus the astragalus is represented only 
by the upper half, while in the present specimen the bone is very nearly 
complete.*'^ Though the bone is typically that of the oreodonts it is 
at once recognizable by its narrowness and greater height. This 
feature is apparently characteristic of all the known Uinta oreodonts. 
In Protagriochcerus and also in Protoreodon the cuboid facet of the 
astragalus is not concave from side to side as in Merycoidodon, but is 
practically flat, even more so than in Agriochcerus. On the other 
hand the posterior extent of the navicular facet of the astragalus in the 
latter genus reaches proportionally higher up on the posterior face of 
the astragalus than in Protagriochcerus. The bone as a whole is of 
quite different proportions in the two genera. 

With the exception of the proportionally greater length of the 
metatarsals in the present genus they are so similar to those in Mery- 
coidodon culbertsoni that a general description would fit in either case. 
The phalanges are also proportionally longer than in Merycoidodon, 
but otherwise quite similar. The ungual phalanges are slightly 
narrower and higher than those in Merycoidodon. So far as one may 
judge from the fragmentary material of the limb-structure of this 
specimen the high and narrow unguals appear to be the only parts 
which show ancestral features leading to the characteristic high and 
compressed terminal phalanges of Agriochcerus. On the other hand 

*8 The astragalus of Protagriochcerus anneclens of the Carnegie Museum collection 
fits admirably well on the navicular of the type specimen in the American Museum 
of Natural Historj'. 



88 Annals of the Carnegie Museum. 

when the ungual phalanges of the present specimen are compared 
with those of Protoreodon of the Uinta there is presented little or no 
difference in detailed structure. When the equality in size of Prot- 
agriochcerus and Agriochcerus is considered, there is revealed a remark- 
able difference in the general proportions as well as the detailed struc- 
ture of the limbs and especially of the feet^'' in the two genera. The 
radical change from the general slenderness of the Uinta form to the 
shorter and broader feet of the Oligocene genus, whatever caused this 
comparatively rapid change, could not well be due to the increase in 
size of the animal from the earlier to the later representative, as is 
claimed in other cases from earlier to later phyla of the Tertiary. 
While provisionally accepting Protoreodon as approximately in the 
ancestral line leading to Merycoidodon and also acknowledging the 
general mixture of similar characteristics between the Uinta protoreo- 
donts, the Oligocene oreodonts, and Agriochcerus, so ably worked 
out and presented by Professor Scott and referred to on a preceding 
page of this paper, it is not clear that we should acccept Protagrio- 
chcerus or any protoreodont of the Uinta Eocene, so far as at present 
known, as being directly ancestral to Agriochcerus. The material of 
this proposed genus of the upper Eocene is still entirely too fragmen- 
tary. A critical survey of the taxonomy precludes our saying more 
than that most of the known features of Protagriochcerus are indeed 
very close to Protoreodon, if indeed, it is not congeneric with the latter. 

Family CAM ELI D.^. 
Genus Protylopus Wortman. 
24. Protylopus petersoni Wortman. 

Bull. Amer. Mus. Nat. Hist., Vol. X, 1S98, pp. 104-110. 

Among the specimens of this species the greater portion of a skeleton, 
C. M. No. 2948, was found in the eastern part of the Uinta Basin at 
the base of horizon C. The cervical region is imperfectly preserved, 
but enough remains to make out some of the more important features, 
which arc for the first time described below. A brief description of the 
fore and hind foot is thought to be of value, especially since the tarsus 
presents some curious features. The material represents an old indi- 
vidual, of larger size than the type described by Dr. Wortman, and 

47 Agriochcerus gaudryi Osborn, the best preserved hind foot known, is used for 
this comparison. 



Peterson: Material Discovered in Uinta Basin. 89 

also slightly larger than the Princeton specimen decsribed and figured 
by Professor Scott.'"* 

The deuterocone of P- is relatively larger in the present specimen 
than in the type. The molars are so much worn down that an accurate 
comparison cannot be made with the type of the genus. 

The atlas resembles that of Po'ebrotherium as described by Scotf" 
that is to say, it is short and broad, has a posterior projection of the 
transverse processes, which extends well back of the articulation for 
the axis, a high and narrow neural arch with a very faint neural spine, 
deep emargination above the anterior cotyli and slightly indicated 
notches on the external margin. From the specimen in its present 
condition I am unable to say whether or not there is a posterior opening 
through the base of the transverse process as in Poehrotherium. This 
foramen is certainly not located on the dorsal surface of the transverse 
process, as in the Tylopoda. 

The axis is not as long in its general proportions as in Poebrotlicriiim. 
The centrum is strongly keeled throughout its length and the neural 
spine has perhaps even a greater development and extends further 
back of the zygapophyses than in Poehrotherium and is on the whole 
quite unlike that in the tylopods. The transverse process is very 
small and located well back on the centrum. The arterial canal is 
placed laterally and well forward and was evidently bordered an- 
teriorly by a narrow and thin bone which is broken off in the present 
specimen. 

The third and fourth cervical vertebrae are badly mutilated by crush- 
ing and portions of them are entirely lost. The centrum of the third 
cervical is very nearly as long as that of the axis, while that of the fourth 
is slightly shorter. There is evidently throughout a prorninent keel 
on the ventral face of the centra of these vertebrae, which is especially 
well indicated on the third and fifth cervicals. The entire length of the 
pedicle of the third cervical is probably pierced by the vertebrarterial 
canal. The region of the anterior exit is injured, but the posterior 
exit is located on the margin of the intervertebral notch and not inside 
the neural canal as in the camels, nor is there, any apparent tendency 
in that direction. 

A portion of the transverse process of the fifth vertebra is present 
and points forward and downwards, as is the case in Poebrotherium. 

"Wagner Free Institute, Vol. VI, 1899, pp. 22-47, PI- 2, Figs. 5-9. 
^' Journ. Morphology, Vol. V, 1891, p. 22. 



90 



Annals of the Carnegie Museum. 



Between the origin of the transverse process and the ventral keel 
there is located on the centrum a deep pit, which is apparently more 
emphasized than in Poebrotherinm. As in the latter, the vertebrarterial 
canal extends throughout the entire length of the side of the neural 
canal, and is plainly visible from the side, entirely unlike what is ob- 
served in the tylopods, and more like the condition found in the 
Bovidcc. The sixth and seventh cervicals are so badly mutilated that 
an accurate description of them is not possible. 

No dorsals are present, and there are only four posterior lumbars 
represented. The two vertebrae anterior to 
the last (the fifth and sixth) have the cen- 
tra quite long, with prominent ventral keels, 
and also well developed neural spines. The 
last lumbar, as usual, has a shorter and 
more depressed centrum. So far as compari- 
son is possible, these parts agree well with 
the description and illustration by Wort- 
man {I.e., p. 107). 

This individual also presents for the first 
time the complete metacarpals III and IV, 
the proximal end only of Mc. II is preserved, 
while Mc. V is lost. As Professor Scott has 
shown (I.e., p. 36) the metacarpals are rela- 
tively short and otherwise differ from those 
in Poebrotherinm. An illustration is here- 
with gi\en of the metacarpals and the 
metatarsals (See Fig. 14). 

Both hind feet are represented and the 
most curious feature is the characteristic 
bovine coossification of the cuboid and the 
navicular in the tarsus of both feet, which 
is hard to believe to be anything but purely 
pathological, especially since the entocunei- 
form of the right tarsus also has a tendency 
to become coossified with the cuboid.^" In 
another specimen (No. 2977) of slightly 
larger size and also fully adult, or old, both hind feet are preserved, 




Fig. 14. Proiylopus pcler- 
soni. Carnegie Museum 
No. 2948. X i/i. I. Left 
cubonavicular, dorsal face. 
2. Left Mc. Ill and IV, dor- 
sal face. 3. Left Mt. III. 
and IV. dorsal face. 



^^ In certain Antelopes as Calablepas gnu, the entocuneiform appears to be partly 
coossified with the cubonavicular bone. 



Peterson: Material Discovered in Uinta Basin. 



91 



presenting the normal conditions. I therefore conclude that the above 
described features are perhaps accidental and should only serve as a 
warning to the student. 

25. Protylopus annectens sp. nov. (Plate XXXVII, Fig. 14). 

Type: Portions of the skull and lower jaws together with other 
parts of the skeleton, C. M. No. 2932. 

Horizon: Uinta Eocene, Lower C. 

Locality: Six miles east of Myton, Utah. 

Specific Characters: A bifid posterior iving of the anterior inner cres- 
cents of the superior molars. Small basal pillars between the inner 
crescents of the molars. Deuterocone of P~ proportionally large. Para- 
style of ilf- especially prominent. Metacarpals proportionally short and 
heavy. Animals larger than Protylopus petersoni. 




Fig. 15. Protylopus annectens. Carnegie Museum No. 2932. X i/i. 

In comparing the present specimen with the excellent description 
by Professor Scott^^ of one specimen in the Princeton collection from 
the Uinta the only dififerences appear to be the absence of the posterior 
bifidity of the anterior inner crescents of the molars and the small 

^' Wagner Free Institute of Science, Trans., Yo\. VI, 1899, p. 24. 



92 Annals of the Carnegie Museum. 

pillar between the two inner crescents of M- in the Princeton speci- 
men.^- In discussing the differences between the molars of various 
specimens of Protylopus in the Princeton collection Scott says {I.e., 
p. 24): "In one case M^ has very small anterior and median external 
buttresses, and the two outer crescents are of similar size and shape, 
with equal prominent median ribs. In M- the buttresses are very 
much larger, and the rib of the postero-external crescent much less 
prominent than the antero-external one. The buttresses of J\I- 
are still larger, enclosing small fossettes, and the posterior buttress 
appears. On M- and M- a small pillar occurs between the two inner 
crescents." This description answers the condition found in the 
present specimen, so far as the buttresses of the external crescents 
and the minute pillars between the internal crescents of M~ and M- go. 
The absence or presence of the minute pillar on the internal cingulum 
of M--may well be a variable character. Altogether the Princeton 
specimen No. 1 1225 might provisionally be regarded as a paratype 
of Protylopus annectens. The illustration of Protylopus petersoni by 
Scott {I.e., PI. 2, Fig. 6) does not indicate this bifidity of the inner 
anterior crescent. 

The fragments of the skeleton furnish no noteworthy characters and 
add little or nothing to what we already know through the studies of 
Wortman and Scott. 

The species is interesting from the fact that it furnishes characters, 
which closely connect Protylopus with Eotylopns Matthew. ^^ From 
the study of the present material the view expressed by Matthew 
that Poehrotherium of the Oligocene is probably derived, not from 
Protylopus of the Uinta, but from some more advanced contemporary 
genus, is apparently much strengthened. But whether this contem- 
porary genus will yet be found in the Uinta or a more northern locality 
is a cjuestion which only the future can decide. 

Measurements. 

Antcro-posterior diameter of superior molar series 27 mm. 

" pi 7 

Transverse " " " 6 

^2 An examination of this Princeton specimen, No. 11225 shows that the teeth 
are very much worn, so that the inner anterior crescent may or may not be bifid. 
There is, however, no sign of the minute pillar on the internal cingulum of M ' , in 
fact the cingulum is rather less developed in the Princeton specimen. 

^^ Bull. Amer. Miis. Nat. Hist., Vol. XXVIII, 1910, p. 41. 



Peterson: Material Discovered in Uinta Basin. •93 

Measurements — Conlinucd. 

Transverse diameter of M ' 9 mm. 

Antero-posterior " " " 9.5 " 

" M? : 9 

Transverse " " " 10 " 

" M^ 10 " 

Antero-posterior " " " 10 *' 

Depth of inferior ramus at M j 15 

" ' " M.^ 17 

Antero-posterior diameter of inferior molar series 31 

" Mj 8 

Transverse " " Mj; 5 

" M^ 6 

Antero-posterior " " " ': . . . 8.5 " 

" M3 14 " 

Transverse " " " 6 

" diameter of distal end of humerus 17.5 " 

Antero-posterior diameter of distal end of humerus 12 " 

Length of Mc. II 50 

Family HYPERTRAGULID^. 
Genus Leptotragulus Scott and Osborn. 
26. Leptotragulus proavus Scott and Osborn. (Plate XXXVII, 
Figs. 5-13)- 

Proc. American Philosophical Society, Vol. XXIV, 1887, p. 258. 

A number of fragments of the upper and lower jaws in the Carnegie 
Museum, collected from the Uinta, reveal a few anatomical points not 
heretofore established. It is very evident that the genus possessed 
three lower incisors of sub-equal size, separated from the canine by a 
quite short diastema. In C. M. No. 3009 the canine tooth is just 
appearing through the alveolar border, and it presents a crown of 
quite large size with an oblong outline on cross-section, a rather sharp 
border, posterior and anterior, and a greater convexity on the internal 
face than on the external. The tooth has a procumbent position, as 
already stated by Professor Scott. I certainly am inclined to regard 
the tooth as a canine. This tooth is followed by a diastema back of 
which is a two-rooted premolar (See PI. XXXVTI, Figs. lo-ii). 
Unfortunately the ramus is broken back of this premolar, and we are 
unable to now definitely say whether or not it is isolated, but from 
Professor Scott's description and my own observation of the type, 
we are assured that there are three premolars in a continuous row back 



94 Annals of the Carnegie Museum. 

of the diastema. A portion of the last deciduous molar is in position 
and My and Mo- are fully erupted, but they have received very little 
wear. M-^ is still buried in the jaw. In two specimens, C. M. Nos. 
3195 and 3454, referred to this species, P4 is in position. In No. 
3195 Pj is just appearing through the alveolar border, while that of 
No. 3454 is well worn (See PI. XXXVII, Figs. 12-13). When P^ 
in these two specimens are compared with those of similar ages in 
Leptomeryx, the similarity is surprisingly close. Thus the antero- 
internal tubercle on P^f of a young or unworn tooth of Leptotragiilus 
is as well marked as, though rounder than, in Leptomeryx, and the 
two ridges, which extend posteriorly from the protoconid, are quite as 
well developed as in the Oligocene specimen No. 226 used for com- 
parison. It is further to be noticed that of the two ridges just de- 
scribed on P:f, the one which is external connects better with the heel 
of the tooth than the one which is internal. This condition is similar 
to that of the posterior premolars of Leptomeryx evansi noted by 
Dr. Matthew.5* 

The lower molars are quite similar to those in Leptomeryx, the cross- 
valleys being relatively somewhat wider, the heel of M^^^ more com- 
pressed transversely, and the internal tubercle considerably smaller 
(See PI. XXXVII, Fig. 12). 

M- and M- of No. 3009 arc not well preserved, but in a second young 
specimen. No. 2919, somewhat advanced in age, M- and "M- are 
perfect (See PI. XXXVII, Fig. 5). These teeth are more completely 
selenodont than in any known Uinta artiodactyl. The st\les on the 
external face are not strongly developed, though quite plain, while on 
the internal face there are prominent cingula. There are no anterior 
intermediate tubercles; in fact the upper molars quite closely suggest 
such Oligocene forms as Leptomeryx, Ilypertragulus, and JLteromeryXy 
as will be seen in the following paragraphs. 

27. Leptotragulus medius sp. nov. (Plate XXX\TI, Figs. 1-4). 

Type: Fragmentary skull and portion of mandible with M^j present; 
fragments of femur; distal end of calcaneum and astragalus, C. M. 
No. 2986. 

Paratype: Upper teeth and fragment of mandible with M;;. C. M. 
No. 3453- 

Llorizon: Uinta Eocene; Horizon C. 

^'^ Bull. Amcr. Mus. Nat. Hist., Vol. XIX, 1903, p. 223. 



Peterson: Material Discovered in Uinta Basin. 95 

Locality: Six miles east of Myton, Utah. 

Specific Characters: External faces of superior molars ivith relatively 
large styles; cingula on superior molars absent. Species of approxi- 
mately the same size as A. proavus. 

The premaxillaries and the front of the skull of the present type are 
unfortunately lost. The upper dentition of this genus is therefore 
not yet completely known. It is, however, \Ai\\n that there is a 
diastema in front of P-. The latter is represented only b>- the two 
roots. P- has three roots; in general features it is much as in Lep- 
tomeryx of the Oligocene, except that the rib on the external face of 
the protocone is somewhat less developed and the deuterocone is 
smaller. P- is proportionately somewhat small, but with the excep- 
tion of the smaller median rib on the external face and a slightly 
smaller deuterocone, this tooth, as the preceding, is most nearly like 
the same tooth in Leptomeryx. The main reason for separating this 
species from Leptotragulus proavus is found in the larger external 
pillars and the absence of the cingulum internally. The external 
faces of both anterior and posterior external crescents have prominent 
ribs as in Leptomeryx, and the internal faces are very nearly as vertical 
as in the latter. The inner crescents have an equally perfect formation 
as in the Oligocene genus, including the anterior and posterior cingula 
and the slight indication of the minute median tubercle. The chief 
difference between the molars of Leptotragulus and Leptomeryx are 
the greater brachyodonty and proportionally smaller size of M^ in 
the genus from the Uinta. ^^ 

M3- of this species does not differ in any particulars from that in 
L. proavus. From Leptomeryx of the middle Oligocene it differs by 
being more brachyodont and by its narrower and simpler heel. 

The skull is so badly mutilated that it does not permit of an accurate 
description. It is, however, possible to determine that the occipital 
plate is rather narrow, that there is a long and quite decided sagittal 
crest. The temporal crest is well defined and terminates in a promi- 
nent post-orbital process. The orbit is large, its anterior border; is 
directly above the posterior portion of M- as in Leptomeryx or Ilyper- 
tragulus, but is probably widely open posteriorly. The alveolar 
portion of the maxillary is rather high for an animal with brachyodont 
teeth, but in this respect it is also like the Oligocene genera Leptomeryx 

55 This comparison is based on the type specimen of Leptomeryx iransmontanus 
Douglass from the Upper Ohgocene of Montana. 



96 Annals of the Carnegie Museum. 

and Ilypcrtragnliis. The exit of the infra-orbital foramen cannot be 
definitely located. 

The femur is much mutilated, especially in the region of the shaft. 
The head cannot be regarded as small ; on the contrary, it is of pro- 
portionate size, well rounded, and set on a distinct neck. The pit 
for the ligamentum teres is shallo\\^ The great trochanter does not 
extend much above the head, it is not of great transverse diameter, 
though well proportioned fore-and-aft, and there is a deep trochanteric 
fossa. The second trochanter is quite prominent and the shaft in this 
region perhaps has a greater antero-posterior than transverse diameter. 
Distally the condyles are crowded close together, due to crushing. 
The rotular trochlea is very well defined, but its internal and external 
borders are nearly equally prominent. 

The distal end of the calcaneum is rather delicately constructed. 
The sustentacular facet is not prominent and the cuboid facet is very 
oblique fore-and-aft. The tuber is lost. The fibular facet is promi- 
nent and occupies a more lateral position than in Leptomeryx, and 
the ridge separating the latter facet from that of the astragalus is 
more prominent than in the Oligocene genus. The astragalus is high 
and narrow; its height being equal to that of this bone in Leptomeryx, 
but it is narrower. The proximal trochlea is quite broad and other- 
wise divided, as in Leptomeryx, but the proximal and distal portions 
are divided by a longer and more decided neck, and the cuboid articu- 
lation is decidedly narrower than in Leptomeryx. 

The limb and foot structure of Leptotragiilus is not well known. 
Professor Scott was not entirely certain that his paratype pertains to 
the same genus {I.e., p. 480). Nevertheless it appears quite certain 
that the genus possessed a tetradactyl manus and didactyl pes, such 
as Scott describes, when we recall the tetradactyl manus and didactyl 
pes of the hypertragulids of the Oligocene. 

When the dentition and certain cranial characters of Leptotragiilus are 
compared with these elements in the Oligocene genera, there appears to 
be greater assurance of the taxonomic position of this Uinta genus. 
In the first place there does not appear to be the slightest indication 
of an elongation or a lateral compression of the upper cheek-teeth in 
Leptotragiilus, or other Uinta genera, such as we should expect from 
what we see in the Oligocene Poehrotherium, or in camels of later 
horizons. In the second place we see that, while the upper teeth of 
Leptotraguiiis, as now known, arc very closely like those of Leptomeryx 



Peterson: Material Discovered in Uinta Basin. 97 

of the Oligocene, there are two and possibly three highly important 
features of Leptotragulus which differ from Leptomeryx and more 
closely suggest those of Hypertragidus, also of the Oligocene. These 
characters are: (a) the subequal size of the incisors; {b) the larger 
canine; (c) the reduction of the premolars to three (if that be the cor- 
rect number in Leptotragulus). While the Uinta genus is apparently 
related to both of these Oligocene genera it can hardly be seriously 
considered to be a direct ancestor in either case. 

Practically equal grounds exist for considering Leptomeryx as a 
direct descendant of Leptotragulus. In this Oligocene genus we have 
the brachyodont dentition characteristic of Leptotragulus; molars 
with a mesostyle, and an internal cingulum.^^ 

Taking into consideration the known characters of Leptotragulus, 
it is perhaps most likely that the genus, if found later than the Uinta, 
continued in the Oligocene as an independent line, which after all 
remained rather more closely related to the true hypertragulids and 
did not constitute a transition towards the tylopods. From what we 
now know of Leptotragulus it is quite doubtful whether the Oligocene 
species "Leptotragulus'' profectus Matthew^^ belongs to this genus. 
Very recently^* Dr. Matthew has orally expressed himself as doubting 
that "Leptotragulus" profectus from the Lower Oligocene of Montana 
is in the line of the early camels. 

M EASUREMENTS. 

Antero-posterior diameter of upper teeth P- to and including M-. ... 35 mm. 
Antero- posterior diameter of P- to M- 16 

" pi 6 

Transverse " " " 4-5 

"P^ 5 " 

Antero-posterior " " " 5 

"Ml 5-5 " 

Transverse " " " 7 

" M- 8 

Antero-posterior " " " 7 

" M^ 7 

Transverse " " " 9 

Diameter from alveolar border at M- to orbital border of molar 14 

Height of astragalus 15 

Breadth of astragalus, lower end 8 

^^ The cingulum of the upper molars of L. mediiis is heavier in proportion to 
that of the specimens in the Carnegie Museum referred to Leptotragulus proavus. 
" Bull. Amer. Mus. Nat. Hist., Vol. XIX, 1903, p. 224. 
58 During the meeting of the Paleontological Society at Pittsburgh, Pa., 1917-18. 

7 — DEC. IS, 1919. 



98 Annals of the Carnegie Museum. 

Genus Leptoreodon Wortman. 

Bull. Amer. Mus. Nat. Hist., Vol. X, 1898, p. 94. 

This genus is represented in the collection of the Carnegie Museum 
by a number of fragmentary specimens, which are not sufficiently 
complete to add anything of importance to what we know from the 
studies of Dr. Wortmann and Professor Scott. 

28. Leptoreodon marshi Wortman. 

Bull. Amer. Mus. Nat. Hist., Vol. X, 1898, p. 98. 

From a recent reexamination of the type of Leptoreodon marshi 
in the American Museum it appears that the crescents of the molars 
were quite high in an unworn state, and that the anterior intermediate 
tubercle was probably not present, as Dr. Wortman was inclined to 
believe. ^^ The type specimen should be further cleared from its 
matrix and good crown-views published of the upper and lower 
dentition. 

With regard to the genus Camelomeryx proposed by Prof. Scott, 
I feel that it is so nearly like Leptoreodon, that, until more material 
is obtained, I should hesitate to separate them generically. 

Genus Oromeryx Marsh. 
Amer. Jour. Sci., Vol. XLVHI, 1894, p. 269. 

29. Oromeryx sp. 

To Oromeryx is provisionally referred an immature specimen, 
C. M. No. 3027, found in horizon C, six miles east of Myton, Utah. 
The specimen consists of a number of mutilated fragments representing 
various parts of the skeleton. The most important of these is a 
maxillary, which contains three deciduous and two permanent teeth. 
The deciduous teeth do not possess any characters of diagnostic value, 
but the permanent molars have that curious broadness of the anterior 
inner crescents and the sudden transverse reduction of the posterior 
inner crescent seen in Professor Marsh's illustration and mentioned 
by Professor Scott.*" The internal cingulum is weak, but has the 
mammillary structure seen in the type of Oromeryx. The t\-pe of 
Oromeryx plicatns, which has recently been examined, consists of right 
and left upper teeth; those of the right side are all broken externally 
except P-. The specimen pertains to an animal fully adult; but shows 

5' Prof. Scott ascertained from unworn specimens that the anterior intermediate 
cusp of the upper molars is not present. 

60 Trans. Wagner Free Institute of Scietice, Vol. VI, 1899, p. 83. 



Peterson: Material Discovered in Uinta Basin. 99 

that the internal cinguluni had a mammillary structure. The external 
face of M- is not complete, as is indicated in Marsh's illustration. 
P- has a prominent posterior heel-like extension, which is not fully 
represented in the figure. The different styles of the external faces, 
and especially the mesostyle of the molars, have sharper angles than 
represented in the illustration. The anterior inner crescent has a 
tendency to have the outline triangular, as in Protylopus annectens. 
The illustration of Oromeryx is a composite from the right and left sides. 
The specimens probably pertain to the same individual, although the 
cinguluni and internal basal tubercle of the median valley is larger on 
the left than on the right side. 

In this connection it is proper to publish an observation on the 
status of the type specimen of " Paramsryx'' Marsh, ^^ recently (191 5) 
made by the writer, while in New Haven. 

So far as I was able to ascertain, the type of '' Param^ryx" Icevis 
Marsh consists of a mixed lot of surface fragments representing dif- 
ferent individuals. The upper tooth and astragalus figured as be- 
longing to the same species are clearly too disproportionate in size, as 
Professor Scott has already pointed out, '''^" and most certainly pertain 
to separate genera. With the astragalus and the upper tooth figured 
by Marsh are now two other teeth and various other fragments. 
In this same general lot are also fragments of the lower jaw having 
about the right size to go with the astragalus mentioned, and altogether 
too large to go with the upper tooth represented in Figure 20 by Marsh. 
Furthermore, the figured tooth represents an animal with teeth very 
little Ivorn, while the other upper molars in the same general lot of 
material are considerably worn. The tooth figured appears to have 
a less marked internal cingulum, especially on the anterior lobe, than 
is indicated in the illustration. The lower teeth associated with this 
figured tooth may or may not belong with the same specimen. These 
teeth have basal pillars in the external valleys similar to those in 
Leptotragiiliis. 

Parameryx sulcatus Marsh seems to be based on a young jaw with 
all the teeth lost, except the anterior portion of Mg^,. This tooth is 
just appearing through the alveolar border. The deep groove which 
Marsh distinguished as his chief specific character is apparently due to 
the crushing of the specimen. 

The tarsus referred to this genus by IVIarsh is high and narrow, the 

" Amer. Jour. Sci., XLVIII, 1894, p. 269, figs. 20-21. 

'''"Wagner Free Institute of Science, Transactions, Vol. VI, 1899, p. 48. 



100 Annals of the Carnegie Museum. 

distal end of the cuboid small and displaying no sign of a fifth digit, and 
the pes was therefore didactyl, as Marsh says. 

This type material seems to me to be an unfortunate mixture of 
Lcptotragiihis, probably of Protoreodon, and possibly also of Oromeryx. 
If the tooth designated by Marsh as Parameryx should after all prove 
to be a second molar of Oromeryx it would further complicate matters, 
as we would be forced to accept Parameryx, and relegate Oromeryx 
to the synonj-my, inasmuch as Parameryx was, strictly speaking, first 
mentioned by Professor Marsh in his address. "^^^ 

That the Hypertragulids of the Uinta Eocene represented by Lep- 
totragiiliis, Leptoreodon, and Oro?neryx are three closely related genera, 
which naturally fall into the subfamily Leptotragulina, appears to 
admit of little or no question. But whether or not any or all of these 
Uinta forms are directly ancestral to the forms from the American 
Oligocene we are not yet in a position to state. All who have studied 
them agree that they have many anatomical features in common with 
the Oligocene genera. However, in the cases which afford the best 
opportunities for comparison, there appear, grouped together in a single 
form from the Uinta, characters which appear to be held in common by 
any one of two or three dififerent genera from the Oligocene. These 
characters of the specimens from the Uinta are collectively such 
that the animals, which the remains represent, may be said to have 
had a common ancestor, but it is not possible to say that they point 
to any particular genus of the Oligocene'^- as a derivative. Future dis- 
coveries and studies may possibly prove that all the species "of the 
Uinta hypertragulids have individually characters which agaifl may 
be recognized in the different Oligocene genera, and we know, for in- 
stance, that Leptotragulus has certain features in common with Hetero- 
meryx, Leptomeryx, and Hypertragulus. This mixture, or combination 
of characters, in Leptotragulus which are revealed by the Oligocene 
genera, is then, as I regard it, the expression of a relationship springing 
from a common descent from the old Pecoran stock. Certain new 
or more recently acquired features characteristic of the Hypertragalidce 
of the Oligocene, express progression toward types which had their 
origin and gradual development, as these forms were dispersed from 
centers where the most rapid moulding and development of new char- 

^^^ Amer. Jour. Sci., Vol. XIV, 1877, p. 364. 

^2 The same may well be said of the Uinta protorcodonts, as pointed out by 
others. 



Peterson: Material Discovered in Uinta Basin. 101 

acters took place. ''^ Such an hypothesis makes it possible to admit 
that the sudden appearance in a given horizon of the remains of new 
forms more or less like those of the indigenous species, and sometimes 
the actual mingling of them in the same geological horizon, represent 



Whi(PRli/er 
Oligocene 


? 

~ - , Heteromenpc stem /fyper-laaulua stem Le/ftomerux stem. 


Uints Eocene 


Leplotraoulua jstenv \ / ^^.^-'^^ 


Eurltf Pecora- \ 



Fig. i6. Diagram expressing the relationship of the HypertraguHdae. 

invasions, which gradually filled the places formerly occupied by an 
older fauna. 

FERISSOBACTYLA. 

Subfamily Hyracotheriin^. 

Genus Epihippus Marsh. 

30. Epihippus gracilis IMarsh. (Plate XLIII, Figs. 4-8). 

Amer. Jour. Sci., Vol. II, 1871, pp. 37-38 (Anchitherium gracilis)/'* 
This species is represented by five individuals all in fragmentary 
condition. The best specimens, C. M. No. 2923 and No. 3398, each 
consist of a few fragments of the skeleton. The specimens, especially 
No. 3398, compare very closely with the description and illustrations 
of Epihippus uintensis by Granger. The latter author, however, 
was unable to distinguish between E. gracilis and E. uintensis.^^ 
The accompanying illustration is given in order to further elucidate 
the characters of the species (See Pis. XLII-XLIll). It is evident 
that P- has the antero-external angle more acutely developed than 
in the molars, which is due to some extent to the well developed 
parastyle. The massive mesostyle of the molars answers well to the 
description given by Granger. 

The glenoid cavity of the scapula is quite ovate in outline and the 
coracoid process is prominent, -while the spine rises comparatively 

"Dr. W. D. Matthew, The Popular Science Monthly, Nov., 1910, pp. 473-478. 
Acad, of Science, New York, Vol. XXIV, 1915, p. 100. 

'■• Later Professor Marsh established the name Epihippus which was published 
in a foot-note in Popular Science Monthly, April, 1878, and Dr. O. P. Hay has desig- 
nated the type, while Mr. Granger has recently pointed out the characters of the 
genus (Amer. Mus. Nat. Hist., Vol. XXIV, 1908, p. 232.) 

^^L.c, p. 258, PI. XVIII, Figs. 4-5. 



102 Annals of the Carnegie Museum. 

close to the border of the glenoid cavity. The proximal end of the 
radius is deeply grooved for the articulation of the humsrus and the 
shaft is more suddenly constricted ^'' below the head and more rod-like 
throughout than in Mesohippiis. Distally the articulations for the 
scaphoid, lunar, and cuneiform are also more distinctly separated 
by ridges than in the latter genus. The facst for tha cuneiform is 
located laterally in the specimen under description. 

No. 2923 represents an animal somewhat heavier thaan the one de- 
scribed above. The length of the tibia is completely represented, 
and shows a stouter bone than in Epihippus parvus. The hind foot 
was possibly shorter and stouter. The lateral digits are apparently 
also slightly more reduced than in Epihippus parvus. 

31. Epihippus parvus Granger*^" (Plate XLII, Figs. 12-16). 

A specimen in the Carnegie Museum, No. 3397, referred to this 
species, consists of the right upper cheek-dentition, the lower jaws 
with Py-^, My-o of the left side; P^-g- and My of the right side; frag- 
ments of the hind limb and the greater portion of the right hind foot. 

The specimen reveals an animal of slightly larger size than the type 
of the species. The characters of the teeth agree very closely with 
those of the type in the American Museum of Natural History, except 
P~, which lacks the antero-internal tubercle, which is present in the 
type. That this is a variable character in the species, it is cjuite 
reasonable to suppose. A distinct feature is the relatively large size of 
the tooth, when compared with P- of such a form as Orohippus 
progressus Granger from the upper Bridger (see p. 250, and PI. XVIII, 
Fig. I). 

The best preser\ed parts besides the upper and lower jaws are the 
different portions of the hind limb. The pelvis is relatively slender, 
the acetabulum well-formed, and relatively deeper than in Mesohippus. 
The proximal end of the femur is broken off, but the distal portion 
presents a slenderness which is proportional to the rest of the limb. 

The tibia and fibula are completely separated. The latter bone is 
not represented, while the tibia is very nearly complete. There is a 
prominent spine, separating the femoral articulations of the tibia, 
which appears somewhat abnormal, esj^ecially in its anterior region. 
The cnemial crest is sharp, not very much extended beyond the main 

^^ It is possible lliat this sudden constriction may be due, at least in part, to 
crushing. 

6' Bttll. Amcr. Mjis. Nal. Hist., Yo\. XXIV, 1908, p. 258, PL X\'III, Fig. 3. 



Peterson: Material Discovered in Uinta Basin. 103 

surface of the shaft, and reaches downward about one-third of the 
length of the bone. The distal trochlea is deeply excavated, obliciue, 
and the internal malleolus is well developed. 

The pes was found in an articulated position with the distal end of 
the tibia in place. The long and slender structure of the limb, when 
compared with Mcsohippus, is at once observed. The astragalus is 
high and laterally compressed, the navicular and cuneiform are quite 
high, the median metatarsal is especially slender, while the lateral 
metatarsals are relatively heavier than in the Oligocene genus. There 
is no evidence present of the first digit. The structure of each bone 
of the pes is wonderfully similar in detail to that of the same bones 
in Mesohippiis from the Oligocene. 

PSEUDOTAPIRS OF THE XoRTH AMERICAN EoCENE.*^^ 

While working upon the fossil tapirs of the Uinta Eocene in the 
collection of the Carnegie Museum it became evident that a more 
intensive study of the pseudotapirs of the earlier Eocene formations 
was necessary. Types and other material representing " Systemodon" 
= Homogalax^^ of the Wasatch and Wind River, Heptodon of the Wind 
River, Helaletes hoops, H. nanus, Dilophodon minusculus " Isectoloplins'' 
latidens and Desmatotherium guyotii of the Bridger, and Isectolophus 
annectens of the Uinta were kindly submitted for this study by the 
authorities of the Peabody Museum of Natural History, the American 
Museum of Natural History and the Natural Science Museum of 

*8 An abstract of this portion of the present paper was read before the Pittsburgh 
meeting of the Paleontological Society, 1917-18. 

*' "On The Names of Certain North American \"ertebrates," Science, Vol. IX, 
1899, p. 593. 

In 1908 {Bull. Amer. Mus. Nat. Hist., Vol. XV, p. 241) Mr. Granger was under 
the impression that the genus from the Bighorn, known as " Systemodon," species 
tapiriniis (1881, Amer. Naturalist) though a much mutilated type, should be accepted 
Since that time extensive and thorough field work by the American Museum parties 
in New Mexican localities reveals the fact that not a single specimen of the 
Bighorn " Systemodon" was found, and Mr. Granger is now inclined to believe, 
according to a communication dated by him January 20, 1916, that this form 
''does not occur there and that the type of tapirinus is a horse {Hyracotherium, or 
Eohippus as we call the American species)." In this connection I may further 
quote a portion of Mr. Granger's letter: "It seems too bad to have to give up the 
name Systemodon, it has become so firmly fixed in the literature, but there is no way 
out of it that I can see. If by any chance we could use it, it would not be for the 
form which we understand as Systemodon, but for a Hyracolhere, which I now feel 
sure that it is." 



104 Annals of the Carnegie Museum. 

Princeton University. The result of my work convinced me (i)_that 
there are two distinct lines of tapir-like animals, or pseudotapirs, in 
the Eocene of North America, which hold a position parallel to the 
Lopliiodontidce of the European Tertiary; (2) that the decision reached 
by Scott, Osborn, Wortman and Earl that the Helaletids are not true 
ancestors of Tapirus is confirmed; (3) that the view that Isectolophus 
may be regarded as in the ancestral line of the recent tapirs is not 
substantiated, but that this Uinta genus represents a distinct side- 
or secondary line of pseudo-tapirs in the American Eocene; and (4), 
that, with the possible exception of Dilophodon minusculus, we have 
not yet discovered in the American Eocene the true ancestor of the 
Oligocene and recent tapirs. 

Subfamily TAPIROIDEA Gill. 

Family HELALETID.E Osborn and Wortman. ^^ 

This family includes Heptodon of the Wasatch and Wind River 
Eocene; Helaletes boops, H. nanus, Dilophodon minusculus of the 
Bridger Eocene, and probably Colodon of the Oligocene. Dcsmato- 
thcrium of the Bridger Eocene should be excluded and placed in the 
line of the Rhinoceroses, vide p. 127, or more correctly in the family 
Hyracodontida. 

Subfamily HELALETIN.E. Wortman and Earl.'i 

Small perissodactyls with large anterior nares and air-sinuses analogous 
to those in the recent tapir; cross-crests of loiver molars perfectly developed; 
no connecting crests betiveen proto-, and hypoconids; hypoconulid small or 
absent; incisors subeqiial in size; limbs long and slender and of more 
nearly equine than tapiroid structure; astragalar trochlea very oblique. 

Genus Helaletes Marsh. 
32. Helaletes boops Marsh. (Plate XLII, Figs. 1-9; Plate XLIII, 

Figs. 1-3). 
Amer. Jour. Sci., Vol. IV, 1872, p. 218. 

Type: Considerable portion of the skull, lower jaws and other frag- 
ments of the skeleton. No. 11807, Peabody Museum Catalogue. 

Horizon: Bridger Eocene. 

Locality: Grizzly Butte, Wyoming. Yale Expedition of 1S71. 

^^ Bull. Amer. Mus. Nat. HisL, Vol. IV, 1892, p. 127. 

"' L.c, Vol. V, 1893, P- 173- 



Peterson: Material Discovered in Uinta Basin. 105 

Principal Generic Characters: J^f , C\, P?|-", Mf . Premaxillaries 
extended well foncard and anterior nares oblique, quite large, and con- 
tinued backii'ard into a distinct air-sinus analogous to that in the recent 
tapir. Nasals somewhat reduced in length. Skull and lower jaws 
gently constricted back of the incisors and canines. Superior canine 
isolated by diastemata. Infra-orbital foramen large and located well 
back. Sagittal crest prominent. P~ much reduced. P- with slight 
indication of two internal tubercles. P- with large single deuterocone. 
Upper molars ivith sharp cross-crests and deep valleys. Parastyle 
small and sessile, m^tacone trenchant and the external face not convex. 

Symphysis of mandibular rami solidly codssified. Hypoconulid of 
Mj small. Absence of connecting crest between proto- and hypoconids 
Limbs slender. Astragalus high, narroiv, trochlea oblique, equine-like.''^ 

Specific Characters: P- with little or no indication of a tctracone; 
styles at the exit of the median valley of the superior molars relatively 
large; cingula on anterior and posterior faces of upper molars prominent; 
animals about half the size of a sh'eep. 

General Description of the Type. Skull and Lower Jaivs: 
Since Professor Marsh's preliminary description of Helaletes boops 
{I.e., p. 218) was published, the entire specimen has been skilfully 
prepared and it is now possible to furnish a more complete description. 

From the generic characters it is at once evident that these remains, 
and especially the skull, possess remarkable similarities to the tapir. 
We have for instance the prominent sagittal crest, the broadness of the 
anterior portion of the frontals, and posterior portion of the nasals, 
the reduced antero-posterior diameter of the latter bones, the large 
premaxillaries, the oblique borders of the anterior nares, the large 
infra-orbital foramen and, above all else, the characteristic large 
excavation on the side of the face formed at the expense of the maxil- 
lary and the nasals (See PI. XLIII, Fig. i). This air-sinus, so far as 
can be determined from the crushed specimen, appears to directly 
communicate with the anterior nares, as in Palccotherium medium, 
and the tapir. As already stated the base of the nasal has a consider- 
able lateral contact along the anterior termination of the frontal and 

"2 See Leidy's description and illustration of " Hyrachyus" nanus in the "Extinct 
Vertebrate Fauna," pp. 68-69, PI- XXVI, Fig. 11. Also Scott; Contrib. E. M 
Museum, Bull. No. 3, 1883, p. 51. PI. VIII, Fig. 4. 

"^ The character of the astragalus was already pointed out by Prof. Marsh in his 
original description (p. 218). 



106 Annals of the Carnegie Museum. 

meets the maxillary at the posterior termination of the border of the 
air-sinus. A short distance in front of the fronto-nasal suture the 
nasal is suddenly depressed. Anterior to this depression the bone is 
broken off and lost, but is indicated approximately correctly by a 
dotted line (See PI. XUII, Fig. i). The anterior palatine foramen 
appears to be .of large size. The palatines are not well preserved, 
while the zygomatic arches and the base of the skull, except the de- 
tached occipital condyles, are entirely wanting. In the region of the 
orbit the skull is much mutilated. 

The superior incisors are subequal in size, slightly spaced, and their 
crowns are broad and somewhat fan-shaped. The diastemata in front 
and behind the superior canine are well indicated in the accompanying 
illustrations. The small size of the canine, which may be a sexual 
feature, is also plainly seen in the illustration cited above. P- is 
much reduced, its crown rather simple, and it is inserted by a single 
fang. The rest of the premolars have a single internal cone (deuter- 
ocone), except P^, which appears to have a slight indication of two 
internal tubercles on the worn surface. The ectoloph of these teeth 
is dixided into four vertical ridges consisting of the external tubercles 
and the anterior and posterior styles, while directly anteriorly and 
posteriorly, there are well-developed basal cingula. 

The cross-crests of the molars are quite perfect, rather sharp, and 
directed inward and backward. The metacone is very characteristic, 
presenting a flat external face, thus forming a cutting lobe, which 
extends from the protoloph backward, even with the posterior face 
of the tooth. The paracone is more convex externally. The para- 
style is rather small and quite sessile on the antero-external angle of 
the metaloph. There are well developed cingula directly anterior 
and posterior, while internally there are no cingula on proto- or hypo- 
cones. At the internal exit of the median valley there is a blunt style, 
and on the ectoloph, especially on M-3- and M^, there is a smaller st3'le 
located between the para- and metacones.'^'* 

The lower jaw also presents a number of tapiroid features. The 
inferior border is quite convex fore-and-aft. The vertical ramus in the 
region of the coronoid process is also directed well forward, and the 
temi)oral fossa is deep. The horizontal ramus is, however, somewhat 

"■' Prof. Marsh mentioned this externally located small tubercle and regards it 
as a specific character differentiating from Helaletes nanus. A very careful examina- 
tion of the type of the latter reveals an even smaller tubercle similarly located. 



Peterson: Material Discovered in Uinta Basin. 107 

dec]-), and not of the thick and round character met with in tlie recent 
tapir. The symphysis is solidly fused and there is a long and sharp 
edentulous border between the front and the cheek-teeth. The 
actual contact is lost between the anterior and posterior portions of the 
lower jaws. 

Unfortunately the symphysis of the jaws is broken back of the front 
teeth so that their characters or number cannot be stated. This is 
also true of the canine and the anterior premolars. F^, however, is 
completely preserved. Its crown is submolariform, the tetraconid 
being absent. Externally the base of the crown has a light cingulum, 
while internally the tooth is smooth. The molars are well represented; 
their cross-crests are perfectly formed and sharp. As in the upper 
molars, there are anterior and posterior cingula, while internally and 
externally the teeth are more or less smooth. The hypoconulid 
of Mtt may be said to be only a heavy cingulum. 

M EASUREMENTS. 

Total length of skull fragment 123 mm. 

Length of skull from end of premaxillary to anterior border of orbit, 

approximately 73 

Length of alveolar border U to and including M'* approximately 95 " 

Length of diastema between incisors and canine 8 

Length of diastema between canine and P- 16 

Length of premolar series 26 

Length of molar series 30 " 

Antero-posterior diameter of canine at base of crown 5 " 

Transverse diameter of canine at base of crown 3 " 

Antero-posterior diameter of P- 4.5 " 

Transverse " " " 3 " 

" P^ 7 

Antero-posterior " " " 7 " 

" P^ 7-5 " 

Transverse " " " approximately 8 " 

" P' 9-5 " 

Antero-posterior " " " 8 

" ' "Ml 9 

Transverse " " " 10 

" M? II 

Antero-posterior " " " 10 

" M-* 9 

Transverse " " " 10.5 

Depth of lower jaw at M ' , approximately 19 

" " " " M,, " 21 

Antero-posterior diameter of P.j 8 



108 Annals of the Carnegie Museum. 

Transverse diameter of P4 5.5 mm. 

" My 9 

Antero-posterior " " " 6.5 " 

" M2 ■ 10 

Transverse " " " 6.5 " 

" M3 6.5 " 

Antero-posterior " " " 12 " 

Vertebral Column: The vertebral column is represented by fragments 
of the cervicals, the centrum of one dorsal, fragments of four lumbars, 
and the anterior portion of the sacrum. 

From a fragment of the atlas it is evident that this bone was quite 
elongate and high, and that the base of the transverse process was 
pierced by a foramen. The centra of the anterior cervicals have a 
sharp ventral keel, which terminates posteriorly in an enlarged rounded 
tubercle, while two posterior cervicals (the fifth and sixth?), have 
heavy and quite high neural spines. 

The centra of the lumbar vertebrae are heavy, broad, and depressed, 
especially those in the posterior region; they have a sharper ventral 
keel and are much less opisthoccelous than in the recent tapir, but 
the posterior face of the transverse process of the last lumbar is pro- 
vided with a large articular face, which meets a corresponding face on 
the anterior extremity of the pleurapophysis of the first sacral vertebra 
as in the recent genus. 

The centra of the sacrum decrease rapidly in their vertical dimen- 
sion from before backward while transversely they appear to maintain 
a greater uniformity, which is again a tapiroid character. The pleur- 
apophysis of the first sacral only supported the ilium, while in the 
tapir the second also takes a considerable part in this function. 

Limbs: The fore and hind limbs are represented by numerous frag- 
ments. The glenoid cavity of the scapula is quite concave and 
presents a prominent descending process on its anterior margin. 
The coracoid process is very prominent and occupies a position slightly 
separated from the glenoid cavity, similar to that in the tapir, while 
the spine, judging from the fragment with which we are dealing, 
formed a more prominent median ridge on the neck and possibly rises 
more rapidly and nearer the glenoid cavity than in the recent genus. 

The inter-trochlear ridge of the distal articulation of the humerus 
is shifted well over to the ulnar side and is prominently developed, 
as in the tapir. The entepicondyle is also similarly developed, but 
the supinator ridge is proportionally somewhat less prominent. 



Peterson: Material Discovered in Uinta Basin. 109 

From the remains of the radius and ulna it is at once evident that 
this animal had the lower fore limb proportionally much longer than 
the tapir, while the articulating surface for the humerus is quite 
similar. 

The carpus is represented only by the anterior portion of the mag- 
num, which bears a striking resemblance to the corresponding portion 
in the recent tapir. The facet for the scaphoid rises gently in the 
ulnar direction to form an acute angle with the nearly vertical facet 
for the lunar and unciform on the ulnar face. The lunar facet does 
not appear to be interrupted by a sulcus before reaching the elevated 
posterior portion, as in the tapir, but is continuous from the dorsal 
face backwards. The superior and inferior facets for the trapezoid 
are not as well separated as in the tapir. In the tapir there is a deeply 
excavated area on the radial face of the magnum which distinctly 
separates these two facets. The facet for Mc. II is plainly represented 
on the lower radial angle just as in the recent genus. 

On comparing the magnum with the illustration of Isectolophus 
annectens by Osborn {I.e., p. 522, PI. X, Fig. 3) there appears to be a 
difference in the angle of the articulations for the scaphoid, the lunar, 
and the unciform. In Isectolophus the scaphoid articulation on the 
magnum rises more, and in the lunar and unciform less, steeply than 
in Helaletes. This may or may not be due to faulty drawing. 

The heads of metacarpals III, IV, and V are preserved and here 
again there is a similarity to the corresponding portions of the tapir. 
Mc. Ill was apparently enlarged in an equal ratio with that in the 
recent genus. The only points of difference, capable of comparison 
between the two, are the suddenly interrupted facet for the unciform 
on the head of Mc. Ill, and the two separate facets for the same bone 
(unciform) on Mc. IV in Helaletes, whereas in the tapir the facet on 
Mc. Ill, is more continuous and the one on Mc. IV is solidly united. 
The antero-posterior diameter of the shaft of Mc. Ill appears to be 
relatively greater in Helaletes than in Isectolophus. Mc. IV has 
facets on the radial angle of the head corresponding to those described 
on the ulnar angle of Mc. III. 

M EASUREMENTS. 

Greatest antero-posterior diameter of glenoid cavity of scapula 21 mm. 

Greatest transverse diameter of glenoid cavity of scapula 17 

Greatest transverse diameter of distal end of humerus 23 

Greatest antero-posterior diameter of distal end of humerus 20 " 



110 Annals of the Carnegie Museum. 

Total length of fragment of radius. . no mm. 

Transverse diameter of head of radius 18.5 

Antero-posterior diameter of head of radius 12 

Greatest height of magnum 10.5 

Transverse diameter of magnum 8 

" " at head of metacarpals III, IV and V in position. 18 

The blade of the ilium rises suddenly from the great sacro-sciatic 
notch and the gluteal surface presents a great antero-posterior extent , 
as in Paloplotherium'^ and in the recent tapir; the point of the ilium, 
however, extends proportionally further forward than in the recent 
genus. The acetabulum is deep and surrounded by a heavy border; 
the ligamentary pit is quite large and deep and the cotyloid notch is 
well developed and rather broad. The spine of the ischium is small, 
though quite plain, and is located opposite the posterior border of the 
acetabulum, as in the tapir. The obturator foramen is apparently 
large, judging from the broad expansion between the ischium and 
pubis near the acetabulum. 

The great trochanter of the femur rises high above the head, which 
again suggests Paloplotherium viiuiis and shows a marked difference 
from the condition in the tapir, more nearly suggesting the condition 
found in Equus. The digital fossa is deep and also extends above the 
top of the transverse line of the head, while in the tapir it is below this 
line. The bone is mutilated in the region of the lesser trochanter, but 
enough is preserved to indicate that this process is of some prominence. 
The shaft is also broken in the region of the third trochanter, but it is 
shown to have considerable vertical and lateral diameters (See PL 
XLII, Fig. 3). The tibial border of the rotular trochlea is very little 
more developed than the fibular border, a tapiroid rather than an 
equine character. 

The tendinal groove on the outer face of the head of the tibia, the 
groove for the lateral ligament from the femur, is larger than in the tapir 
and the cnemial crest is less prominent, otherwise the head of the tibia 
is generally similar in the two genera. The distal end also differs 
from that in Tapirtis by having the trochlea more oblique, an equine 
rather than a tapiroid character. 

From the distal end of the fibula, the only portion of that bone 
prescr\-ed, I judge that the shaft is considerably reduced. On the dis- 

'^ Sec Blainville (Osteographic) Palceolhcrium minus, PI. \'I. Palaolherium 
minus Cuvicr, is now accepted as Iselonging to the genus Paloplotherium Owen. 



Peterson: IMaterial Discovered in Uinta Basin. Ill 

tal posterior angle there is a distinctly rounded facet which fits neatly 
into a corresponding pit-like facet just above the articulation for (he 
astragalus. 

The high and narrow astragalus with its oblique trochlear groove 
was pointed out by Professor Marsh in his original description. This 
important equine feature agrees with Heptodon.a.nd is altogether dif- 
ferent from what is seen in Paloplotherium minus and in the recent 
tapir, in which the bone is broader, shorter, and the trochlear groove 
more directly fore-and-aft. The articulation for the navicular is 
more concave from side to side than in the recent genus. 

The calcaneum of Ilelaletes hoops agrees with that of Ileptodon and 
is longer and slenderer than in the tapir. The bone as a whole, how- 
ever, has a general similarity to that of the latter genus, except the 
round pit-like articulation for the fibula at the base of the tiib3r calcis 
directly dorsad, already referred to in the description of the fibula. In 
the recent tapir, the fibula touches the calcaneum, but this facet is 
only a rounded surface, which is more or less continuous with the facet 
for the astragalus, and is situated on the fibular angle of the latter 
articulation. 

A fragment of the navicular plainly indicates this bone to be pro- 
portionally high when compared with that of the tapir. The cuboid 
is also high and the antero-posterior and transverse diameters are 
proportionally less than in the recent genus. The articulation for Mt. 
IV also differs from that of Ta pirns in being single. 

The pes is tridactyl as in the tapir. This is plainly indicated by 
the heads of the metatarsals of the right side which are all represented. 
Mt. Ill is enlarged in an equal proportion to Mc. Ill of the manus and 
the lateral digits, in particular Mt. II is as much, if not more, reduced, 
than in the tapir. There is a large facet for the entocuneiform on the 
tibial angle of the head of the latter bone, but no evidence of facets for 
metatarsals I or V. The distal trochlea of the median digit is quite 
symmetrical and the metapodial keel is more sharply defined than in 
the recent tapir, but is, as in the latter, confined to the posterior face 
of the articulation. On the whole the pes is very similar to that of 
Heptodon from the Wind River, described by Osborn and Wortman. 

The median phalanges are broad and depressed, unlike the more 
elongated corresponding bones of Heptodon from the Wind River, 
while the lateral phalanges are more rounded. The phalanges as a 
whole are suggestive of the Equidce rather more than those of the true 
tapirs. 



112 



Annals of the Carnegie Museum. 



M EASUREMENTS. 

Length of ilium from border of acetabulum to point of ilium approxi- 
mately 100 

Greatest transverse diameter of femur at head"^ 42 

Antero-posterior diameter of head of femur 18 

Vertical 

Transverse 

Antero-posterior 



tibia. 



19 

31 
30 



" distal end of tibia 18 

Transverse " " " " " " 20 

Height of tarsus, end of tuber calcis to distal face of cuboid 68 

Height of astragalus 27 

Greatest transverse diameter of astragalus 22 

Transverse diameter of trochlea of astragalus, approximately 12 

Total length of calcaneum 154 

Transverse diameter of calcaneum at lesser process 20 

Greatest height of cuboid 17 

Antero-posterior diameter of cuboid 16 

Transverse " " " 11 

" " tarsus at heads of metatarsals 20 

Transverse diameter of head of Mt. HI 12 

Antero-posterior diameter of head of Mt. HI 13.5 

Transverse " " " " " H 6 

Antero-posterior " " " " " " 11 

Measured at facet for entocuneiform. 

Antero-posterior diameter of head of Mt. IV, approximately 10 

Transverse diameter of head of Mt. IV, approximately 7 



33. Helaletes nanus Marsh. 

Amer. Jour. Sci., Vol. IV, 1872, p. 218. 

Type: " Lophiodon" nanus Marsh."'' Fragments of both upper 
jaws with all the cheek-teeth represented on right side, No. 1 1080, 
Yale Museum Catalogue. 

Hypolype: " Ilyrachyus" nanus Leidy.'' P- lower jaw and lower 
dentitions. ? Philadelphia Academy of Natural Sciences. 

Horizon: Middle Eocene. 

Locality: Grizzly Butte, Wyoming. One of the hypotypes, a lower 
jaw from near the Lodge-Pole Trail Crossing, Dry Creek Valley, 
Wyoming. 

'^ The bone is slightly crushed in this region. 

" L.c, Vol. I, 1871, p. 37. 

'8 "Extinct Vertebrate Fauna." . . . Report U. S. Geol. Surv. of the Terri- 
tories, Vol. I, 1873, p. 67, PI. XXVI, Fig. 11; PI. VI, Fig. 42; PL XXVII. Figs. 21 
and 22. 



Peterson: Material Discovered in Uinta Basin. 113 

Specific Characters: P~ with tetartocone more distinctly separated 
from the deiiterocone than in H. hoops; styles at the exit of the median 
valley of the superior molars smaller; cingula on anterior and posterior 
faces of upper molars smaller, but larger on the ectoloph near the posterior 
angle. Animals of same size as II. boops. 

For the present it is thought most prudent to continue to keep the 
above type specimen under a separate species as established by Marsh. 
The writer is, however, under the impression that the features of the 
specimen which vary from those of //. boops may ultimately be re- 
garded as only representing individual variation. In that case 
Ilelaletes nanus becomes the type of the genus and //. boops Marsh 
and " Ilyrachyus" nanus Leidy hypotypes. 

Genus Dilophodon Scott. 
34. Dilophodon minusculus Scott. (Plate XLIV, Fig. 5). 

Contribution from the E. M. Museum, Princeton, New Jersey, Bull. No. 3, 1883, 
p. 46, PI. VIII, Fig. 4. 

Type: Right lower jaw, No. 10,019, Princeton Museum Catalogue. 

Horizon: Middle Eocene. 

Locality: Henry's Fork? Wyoming. 

Generic Characters ascertained from the type. /.;, Cj, P-g, ilf-^-. 
Diastema of the loiver jaiv proportionally shorter than in Ilelaletes boops 
and rami more sharply constricted back of the incisors and in front of the 
cheek-teeth; ilfg- ivithoiit hypoconulid; canine small; animal slightly 
smaller than Ilelaletes. 

There can be comparatively little or no doubt that Dilophodon 
minusculus pertains to a distinct genus, as originally determined by 
Professor Scott. The shorter diastema of the ramus suggests a dif- 
ferent structure of that region than in Ilelaletes. This character 
together with the absence of a hypoconulid on M^ of the type of 
Dilophodon is regarded, by the writer as of sufficient generic value, 
pending the discovery of more complete material. In accepting the 
genus we avoid attributing two important characters to the genus 
Helaletes which do not appear to belong to the latter. Beside this 
fact, it is even possible that Dilophodon represents a third line of tapirs, 
which may represent a closer relation to the Oligocene and recent 
tapirs than to those of the Eocene, which are better known. 

Although the dentition of Ilelaletes, Hyrachyus, and Colonoceras 
resemble one another, the differences between them are obvious on 

8^ — DEC. 16, I919. 



114 Annals of the Carnegie Museum. 

comparison of complete specimens." The reduced nasals, the large 
anterior nares and air-sinus of Heptodon and Helaletes are specializa- 
tions of an early origin, which obtained not only in these genera, but 
in all likelihood was also extended to the true ancestors of the recent 
tapirs, and may thus be regarded as an additional proof of the con- 
servativeness of these Perissodactyls. 

Not only did Helaletes have the abbreviated nasals, but the incisors 
were of subequal size, P- much reduced, and in Heptodon we have the 
long cursorial limbs, truly a combination of specializations too con- 
flicting and requiring too great a modification to fit them to be regarded 
as the true ancestors of the recent tapirs, even be they so far removed in 
time as the Eocene. On the other hand we have in the genus Colodon 
of the Oligocene a form which appears eminently fitted to represent 
the line of pseudo-tapirs in the later Tertiary, according to the deter- 
minations already reached by Osborn, Wortman, and Earl. 

Dilophodon should only be provisionally included in the sub-family 
Helalitince, since it is too imperfectly known, and may still prove to 
be more closely related to the true tapirs. 

When these pseudo-tapirs of the American Eocene are referred to 
as Lophiodonts it should be in a super-family sense (Lophiodontoidea), 
although with a greater restriction than that proposed by Dr. Gill.^" 
In persisting in regarding these American genera as belonging strictly 
to the European family Lophiodontidse^^ the taxonomy of the whole 
group is confused rather than cleared. That there is a relation be- 
tween the European and American forms in question is not to be 
denied, especially when their dental structure is considered. But is 
this similarity much greater than that in other separate families; 
e.g., the Palaeotheres, the Chalicotheres, and the Titanotheres? 
When comparison of the limb structures of Lophiodon (L. isselensis)^'^ 
and Heptodon or Helaletes is made we meet with an important degree 
of adaptive radiation. In Lophiodon, as is well known, the limbs are 

"' See Amer. Jour. Sci., Vol. IV, 1897, p. 161, Fig. 2; Bull. Amer. Mas. Nat. Hist., 
Vol. VII, 1895, p. 370; Me7n. Amer. Mtis. Nat. Hist., Vol. I, 1898, p. 128a, PI. 
Xllla. 

8" Smithsonian Mis. Coll. No. 230, 1872, p, 88. Certain genera of suids, anoplo- 
theres, and Cory phodon, -which were included by Dr. Gill, should, of course, be ex- 
cluded. 

81 Osborn, H. F., and Wortman, J. L., Bull. Amer. Mus. Nat. Hist., Vol. VII, 
189s, p. 358; Osborn 's "Age of Mammals," Classification Table, p. 557. 

82 Filhol, M. Henry, Mem. Soc. Geol. de France, Tom. V, 1888. 



Peterson: Material Discovered in Uinta Basin. 115 

short and thick, while in Ilcptodon and lielaleles they are long and 
slender. In Lophiodon the femur has a curved shaft with the third 
trochanter located well down, the distal end broad; low and broad 
condyles and rotular trochlea; tuber of calcaneuni short and thick; 
astragalus broad, low, with its trochlea directly fore-and-aft. In 
Heptodon and Helaletes the hind limb is, as-we have seen, long, slender, 
and in structure quite equine. Such a broad inclusion of characters 
would necessitate, to my mind, the inclusion, of other early groups of 
the Perissodactyls more or less interrelated {e.g., the Pala'otheres, 
the early Equidse, and even certain phyla of the Rhinocerotoi'lea) 
in this group an arrangement obviously out of reason, especially in 
view of our present accepted order in the phylogeny. 

If the fragmentary material of the small lophiodont described and 
figured by Cuvier,^^ again referred to and figured by Blainville,*^ and 
also referred to under the name " Colodon'' miyiimus by Gaudry,** 
belongs to the same species, it appears that it differs from both Hela- 
letes and Colodon (C. longipes Osborn and Wortman).^*^ These 
European remains appear to be on the whole more tapir-like, i.e., they 
have a lower and broader tarsus, and the astragalar trochlea less 
oblique than in the He ptodo7i- Colodon series of America. We may 
well question the propriety of including them in the same family. 

ISECTOLOPHID.E jam. nov. 

Small Perissodactyls of the Eocene with dentition f , \, f, f . P- with 
one or tivo well defined internal cones. P- with one internal cone. Para- 
and metacones subequal and conic. With or without diastema. Cross- 
crests of lower molars more or less perfectly developed. Connecting crests 
hetiveen proto- and hypoconids. Large hypoconuUd on M^. 

This proposed family includes Homogalax {" Systemodon") Hay as 
defined by Osborn, Wortman^^ and Granger ** and Parisectolophus 
{" Isectolophus") latidens, vide infra. A third genus Schizolophodon, 

*^ Cuvier, Oss. Foss., Tom. II, 1825, p. 194, Figs. 20-25. 

S"" Blainville, Osteographie, Tom. IV, Y, page 100, PL III (fourth specimen). 

85 Gaudry, Bull. Soc. Geol. de France, Tom. XXV, 1897, pp. 318-319. PI- IX, 
Fig. 4- 

8« Osborn, H. F., and Wortman, J. L., Bull. Amer. Mus. Nat. Hist., Vol. VI, 1894, 
p. 214. 

8" Bull. Amer. Mus. Nat. Hist., Vol. IV, 1892, p. 124. 

88 0^. cit.. Vol. XV, 1908, p. 241. (See note in the present paper, page 103, in 
regard to Mr. Granger's identification of Systemodon and Homogalax.) 



IIG Annals of the Carnegie Museum. 

see page 122, is provisionally included, while the so-called tertiary 
tapir Leptolophiodon (Isectolophns) annectens (Rutimyer) vide p. 126, 
is doubtfully included. 

This family may be subdivided into two subfamilies, the Ilonw- 
galaxincc, nom. nov., defined by Osborn & Wortmann under the name 
SystemodontincB {op. cit., p. 124) which cannot be used since the genus 
Systemodon has been transferred to the Equidce, and the Isectolophince, 
subfam. nov. 

Subfamily HOMOGALAXIN^ nom. nov. 
(Type Gen. Homogalax Hay, non Systemodon Cope). 

''Dentition: f, \, ^, |. Superior dental series continuous. First 
lower premolar contiguous to canine, followed by narrow diastema. 
Third 'and fourth superior premolars with two external cusps and a 
single internal lobe. Paracone and metacone subequal, conic. Proto- 
loph and metaloph complete. Large third lobe upon third lower molar."^° 

ISECTOLOPHIN.-E subfam. nov. 
Small Perissodactyls of the middle and upper Eocene with dentition 
-|> T' T' I- Pai'd- and metacones subequal and subconic. P- with two 
internal cones. P~ with one internal cone. With or without diastema 
hack of loiver canine. Cross-crests of lower molars more or less perfectly 
developed. Connecting crests hetiveen proto- and hypoconids. Large 
hypoconulid on M^. 

Genus Isectolophus Scott and Osborn. 
35. Isectolophus annectens Scott & Osborn (Plate XLI\', Fig. i). 

Proc. Amer. Philos. Soc, Vol. XXIV, 1887. p. 260. "The Mammalia of the 
Uinta Formation." Trans. Amer. Philos. Soc, Vol. X\'I, 1889, p. 520, PI. X, 
Figs. 1-8. 

Type: Upper and lower teeth No. 10400, Princeton Museum Cata- 
logue. 

Llorizon: Uinta Eocene. Near base of Horizon C? 

Locality: White River, Utah. 

Hypotypes:^^ A crushed skull with fragments of the lower jaws and 

* Osborn & Wortman (I. c.) defined the subfamily Syslemodonlince basing their 
definition upon material representing Homogalax, and, as Systemodon is now referred 
to the Equidce, I propose the change here made. 

^^ The definition is quoted from Osborn & Wortman. 

^' Fragments of lower jaws. No. 1828 in the American Museum Collection, which 
were found in horizon C, near Wagon Hound Canyon on White River, I^Tinta County, 
Utah, are here provisionally associated as hypotypcs. 



Peterson: Material Discovered in Uinta Basin. 117 

fragnienls of limb-bones, C. M. No. 3030. Right maxillary with 
cheek-teeth in place, C. M. No. 2337. 

Horizon: Uinta Eocene. Near base of Horizon C. 

Locality: On the Duchesne River, six miles east of Myton, Utah, 
and eastern end of Uinta Basin near Kennedy's Hole. 

Generic Characters: {I. ? f , Cy, -Pf, M^. Long diastema back of lower 
canine.^^ P- with two internal tubercles. P- ivith one internal tubercle. 
Meta- and paraconcs of equal size. Ectoloph extended well back of 
protoloph. Paraslyle large and well separated from paracone. Cross- 
crests of molars comparatively sharp, ilf-g- imth large hypoconulid; 
connecting-crests between proto- and hypoconids of molars. "Lunar 
with subequal magnum and cuneiform facets. Cuboid broad ivith an 
extensive astragalar facet. Manus and pes. Digits, 4 — ? J."^^ 

Provisional Specific Characters: Long symphysis back of lower canine; 
the latter tooth proportionally small. Symphysis long and laterally 
constricted.^'^ 

The mashed cranium, No. 3030 and the maxillary, No. 2337, in 
the Carnegie Museum set definitely at rest the question of the upper 
premolar teeth of Isectolophus. Unfortunately the skull of No. 3030 
when found was completely disintegrated in front of P-. No part 
of the front of either upper or lower jaws was obtained. 

The present material agrees very well with the description and espe- 
cially with the illustrations by Professor Osborn and serves well as 
a hypotype of this genus. 

In reexamining the type of Isectolophus annectens I find a portion of 
a premolar, which Osborn mistakenly referred to P-, but which in 
reality compares exactly with P- of the specimen in the Carnegie 
Museum. Hatcher has correctly identified this tooth-fragment.'* 
There is a single tooth with the type specimen, which is apparently 
M- of the right side. The upper teeth described by Osborn as M- 
and M- agree better with M- and M- of the Carnegie Museum speci- 

'2 As shown by the specimen of Isectolophus in the collection of the American 
Museum of Natural History, No. 1828. 

*' The generic characters inside of quotation marks are according to Osborn 
(Trans. Amer. Philos. Soc, Vol. XVI, 1889, p. 519), and are based upon what is 
now regarded as a separate species, Isectolophus scotti. 

'^ If it be hereafter found that complete specimens of Isectolophus annectens 
have this long diastema, then Isectolophus scotti, vide p. 120, must be raised to 
generic rank. 

^^ "Recent and Fossil Tapirs," Amer. Jour. Sci., \'ol. I, 1896, p. 177. 



118 Annals of the Carnegie Museum. 

mens, and reveal a mistake, by oversight in the text, since they are 
correctly indicated in his illustration {op. cit., PI. X, Fig. l). 

Detailed Description and Comparison of the Material. 

P- of the Carnegie Museum specimen has two external tubercles 
set close together and a small blunt deuterocone not unlike some of 
the early tertiary forms {Homogalax protapirinum (Wortman) of the 
Big Horn Wasatch) except that the latter species has the deuterocone 
located slightly further back and the tooth itself has proportionally a 
greater transverse diameter than in the present specimen. ^^ P- has 
two internal tubercles, which are slightly better developed than in 
Pai'isectolopliiis latidois. The proto- and tritocones are situated close 
together, the parastyle is ciuite prominent and is separated from the 
inner tubercles by a deep concavity of the ectoloph. This portion of 
P- is lost in PariscctoJophus latidens. P- agrees in all its detailed 
structure, with that tooth in Parisectolophiis latidens. Hatcher has 
already pointed out that there is only one internal tubercle of P^- in 
Parisectolophiis latidens^'^ and that Osborn's generic definition of 
Isectolophus is partly erroneous. ^^ The ectoloph of P- in Isectolophns 
consists of three subequal swellings, the trito- and protocones, and 
the parastyle is identical w'ith that in the Bridger species. The 
molars increase gradually from the first to the last, and they differ 
in no important degree from those of Parisectolophiis latidens. 

A right maxillary. No. 2337, with the cheek-teeth represented, found 
by Mr. Earl Douglass in Horizon C, at the Devil's Play Ground near 
Kennedy's Hole, Uinta Basin, is interesting from the fact that a de- 
tached small premolar crown, most probably P-, was found with 
the specimen and is here for the first time recorded. P- of this speci- 
men is immediately in front of P-, and is two-rooted and the crown, 
if correctly determined, is rather small, blunt, and conical in shape, 
with considerable wear from the opposite tooth. The teeth back of 
P- are identical with, though smaller than, those described in the fore- 
going pages. From this specimen it also appears quite clear that the 
infra-orbital foramen is located above the anterior portion of P-. 
The maxillary is broken off immediately in front of P-. 

The inferior dentition of specimen No. 3030 is represented by the 

'^ P- is lacking in the type specimen of Parisectolophiis. 

^'' Amer. Jour. ScL, Vol. I, 1896, p. 177. 

'J« Trans. ,\mcr. Philos. Soc, Vol. XVI, 1889, p. 519. 



Peterson: Material Discovered in Uinta Basin. 



119 



greater portion of P:j-, and fragments of the molars. As already 
stated, the specimen was found in a mutilated state, there being 
neither upper or lower incisors present. The specimen in the Ameri- 
can Museum, No. 1828, already referred to, supplements the present 
specimen admirably. The symphysis of this specimen does not fit 
upon the back part of the jaw, due to loss of contact. If the speci- 
men (See Fig. 17), pertains to the same individual, which appears quite 
likely, it is clear that there are two well-marked species of Isectolophus 
in the Uinta sediments. One of these has a long diastema back of the 
canine and the symphysis itself is constricted transversely as in 
Triplopus, while the other species has no diastema and no constriction 
of the symphysis. 

The incisors and canine of the specimen in the American Museum 
are represented by roots only. These roots are of subequal size and 
have about one-half the diameter of the canine. Both incisors and 
canines had apparently a procumbent position. Back of the canine 
there is, as already stated, a long diatema with sharp superior border. 
The premolars are not present, while the molars are well-represented 
in this specimen. In comparing the molars with those of the type of 
Isectolophus it is at once clear that they pertain to Isectolophus annec- 
tens from their size and detailed structure. 

P-j- of the specimen in the Carnegie Museum, No. 3030, is quite 




Fig. 17. Isectolophus annectens. No. 1828 American Museum of Natural History. 

X2/3. 



molariform, except for the lack of the tetartoconid. The cross-crests 
of the lower molars are perfect, as is the case in the specimen in the 
American Museum, and the third lobe (hypoconulid) of M-j is large, 
as in the type. The only noteworthy difference between No. 3030 



120 Annals of the Carnegie Museum. 

of the Carnegie Museum collection and the type of Isectolophus, as 
well as the American Museum specimen, is the presence of two small 
mammillary tubercles rising from the cingulum on the antero-external 
portion of Mg- in the specimen in the Carnegie Museum. These are 
absent both in the type and in the New York specimen, Xo. 1828, and 
may well be only an individual character (See Fig. 17). 

The transverse diameter of the distal trochlea of the humerus is 
unusually small. This fact seems to agree with Osborn's description 
of the head of the radius {op. cit., p. 521). The trochlea is rather deep, 
which is due to the very prominent capitellum. There is a decided 
tubercle on the dorso-ulnar face, just above the capitellum, but there 
is little or no articular surface on the ulnar side of the capitellum, such 
as is seen in the tapir, or developed more prominently in the horse. 
36. Isectolophus scotti^s sp. nov. (Plate XXXIV, Fig. 23). 

Type: Fragments of upper and lower jaws, vertebrae, limb- and foot- 
bones, described by Professor Osborn as the Paratype of Isectolophus 
annectens}^^ Natural Science Museum of Princeton University, 
No. 1 040 1. 

Paratypes: Lower jaw fragment, No. 10399, Natural Science Museum 
of Princeton University, described by Professor Osborn as Paratype of 
Isectolophus annectens (/. c, p. 520, PI. X, Fig. 4). Fragments of upper 
and lower jaws with Mf present. Fragments of vertebrse. No. 31 13, 
Carnegie Museum. 

Ilorizon: Uinta Eocene. Near base of Horizon C. 

Locality: White River, Uinta County, Utah. 

Specific Characters: No diastema back of loivzr canine: symphysis 
short and not contracted laterally; canine of proportionally large size; 
animals smaller than Isectolophus annectens. 

Pending the discovery of complete upper and lower jaws of Isec- 
tolophus annectens the above determination must be regarded as 
pro\isional. If the contour of the lower jaw together with the long 
symphysis, as exhibited in the fragmentary specimen in the American 
Museum, No. 1928 (See Fig. 17) proves to be the same as in /. 
annectens, then the new species here erected must take generic rank, 
as already stated in footnote 94. 

In the specimen, C. M. No. 31 13 (See PI. XXXIV, Fig. 23) the 
position of the dentition in the alveolar border indicates the same con- 
's In recognition of Prof. W. B. Scott, of Princeton University. 

'01 Trans. Amer. Philos. Soc, Vol. XVI, 1889, pp. 521-522, PI. X, Figs. 3-8. 



Peterson: Material Discovered in Uinta Basin. 121 

ditiou described and illustrated by Osborn {op. cit., p. 521); /". e., that 
the second and third premolars have two roots and Py one root. 
The root of the canine indicates that tooth to be relatively larij;er that 
in the American Museum specimen, No. 1828. 

Parisectolophus nom. nov. 
(Type Parisectolophus latidens (Scott & Osborn).) 
37. Parisectolophus latidens (Scott & Osborn). 

Isectolophus latidens Osborn, "The Mammalia of the Uinta Formation," 

Trans. Amer. Phil. Soc, Vol. XVI, 1889, p. 513 el seq. 
Helaletes lalidens Scott & Osborn, E. M. Museum Bull., Princeton Univ., 1878, 
p. 54. No. 3- 

(Plate XLIV, Figs. 2-3). 

Type specimen: Upper and lower jaws, No. 10251, Natural Science 
Museum, Princeton University. {Cf. Osborn, L.c, p. 518.) 

Horizon: Middle Eocene. 

Locality: Henry's Fork? Wyoming. 

Principal Characters: /-j, C-y, P-f, ikff, P-, with tivo internal tubercles; 
P- -liith one internal tubercle; light cingula on anterior, posterior, and 
external faces of premolars and molars. Mela- and paracones subcorneal; 
paracone larger than metacone; parastyle large and luell-separated from 
paracone; cross-crests of molars obtuse and valleys shalloiv; extremely 
short diastema between Py and P-j. Inferior canines comparatively 
robust. Hypoconidid of M-^ well-developed; connecting crests between 
proto- and hypoconids of molars. Lower jaws quite thick and slightly 
constricted back of canine and P 1. 

The most noteworthy differences between this genus and Helaletes, 
as exhibited by the type specimen, are the following: in Parisectolophus, 
the metacone is more subconical, the parastyle larger and more widely 
separated, the cross-crests of the upper and lower molars more obtuse, 
the hypoconulid of M^ longer, and the diastema between the incisors 
and cheek-teeth absent (See PI. XLIV, Figs. 2-3). 

Parisectolophus latidens differs from Isectolophus in comparatively 
few nevertheless well-marked and rather important characters, so 
far as comparison can be made. First, the molars of Isectolophus are 
considerably advanced, because of the greater posterior enlargement 
of the ectoloph, making the para- and metacones equal in size; sec- 
ondly the cross-crests of the molars are slightly higher and sharper; 
the cingulum of the upper molars are much heavier. 



122 Annals of the Carnegie Museum. 

Fig. 13 of Plate II in Leidy's " Extinct Vertebrate Fauna of the 
Bridger Tertiary Formation of Wyoming Territory" represents Hy- 
rachyus modestiis, ^hxch. seems somewhat similar to Parisectolophus, and 
possibly may belong to this genus. However, the metacone appears 
to be located on a more direct antero-posterior line with the paracone, 
and the junction of the metaloph with the ectoloph is somewhat 
further forward than in the type of Parisectolophus described and 
figured. 

Genus Schizolophodon gen. nov. 
38. Schizolophodon cuspidens sp. nov. (Plate XLIV, Fig. 4). 

Type: A pair of lower jaws with premolars and molars fairly well 
preserved. C. M. No. 3045. 

Paratype: Right lower jaw and symphysis with roots of canines 
and the anterior cheek-teeth of the right ramus, C. INI. No. 3010. 

Horizon: Uinta Eocene, Horizon C. 

Locality: Duchesne River, Six miles East of Myton, Utah. 

Principal Characters: Lower molars with incomplete cross-crests, 
especially on M-^ and the anterior cross-crest of M^- Hypoconulid of 
M-^ comparatively small. Animals about the same size as Isectolophus 
annectens. 

Detailed Description of the Material: The horizontal ramus is heavy 
and quite deep; the symphysis is heavy, but the rami are not con- 
stricted back of the canine, and there are two or more mental foramina, 
of which the most anterior is directly below Py. The vertical ramus is 
not present in either type or paratype. 

The incisors, of which there appear to be three on either side, are 
only represented by roots and are rather small. 

The canines are also broken off, and are proportionally small^"^ 
when compared with those in Parisectolophus latidens from the Bridger. 
There is no diastema between the canine and the cheek-teeth. Py has 
a single fang and a simple crown with a small basal shelf postero- 
intcrnally. P2- is quite suddenly enlarged, especially in the fore-and- 
aft direction. The anterior portion of the crown consists of a blunt 
trihedral elevation with the apex directed forward, while the posterior 
portion (the heel) is low, partly due to wear, the tetartoconid being 
absent. P..- is considerably farther advanced towards the molar pat- 

'"' In the paratype the canines are apparently of larger size. 



Peterson: Material Discovered in Uinta Basin. 123 

tern. The proto- and deiiteroconids are rather closeh' united and 
have received much wear. The anterior margin of the crown has 
developed into a crescentic ridge, which extends from the protoconid 
forwards and inwards. The metaconid has formed a somewhat sim- 
ihir, though less crescentic, ridge-connection with the deuteroconid. 
This portion of the crown has also received considerable wear. The 
tetartoconid is absent. P^ is only represented by the posterior half 
of the crown and its detailed structure is practically a repetition of the 
tooth in advance of it, except that the crest, which connects the meta- 
conid with the anterior cross-crest, forms the junction with the latter 
rather midway between the proto- and deuteroconid, instead of directly 
with the deuteroconid as in Pj. In the paratype of Isectolophns annec- 
tens this character is observed to be similar to that just described. 
There is no tetartoconid, but the rise of the fold of the enamel in this 
region of the tooth furnishes the margin of a distinct pit. The 
cingulum is not well-developed on the premolar series. My is repre- 
sented by the posterior portion of the crown in the type, while in the 
paratype only the antero-external angle is preserved. The tooth 
in both specimens has received much wear, so that its characters are 
rather unreliable. IXW is also much worn, but it is still possible to 
detect that the anterior cross-crest was interrupted by a longitudinal 
valley, while the posterior cross-crest is more complete. Directly in 
front the cingulum has formed a heavy ledge, which constitutes a 
conspicuous addition to the base of the crown. M^- may be said to 
have five distinct tubercles on the crown: in addition to this there is 
again repeated the anterior ledge (parastylid) shown on M^-. There 
are no cingula on the internal face of the molars, while externally they 
are moderately developed. 

M EASUREMENTS. 

Type Paratype 
No. 3045- No. 3010. 

Total length of the jaw fragment 128 mm. 

Depth of ramus at Pj 22 " 24 mm. 

Depth of ramus at M^j 31 " 31 

Length of cheek-teeth including canine 90 " 88* " 

Length of premolars 38 " 35* " 

Length of molars 45 " 45 

Antero-posterior diameter of Pf 7 " 7 

Transverse diameter of " 4 " 4 

Antero-posterior diameter of P/ 9 

Transverse " . " " 5 " 

Antero-posterior " " P3 10 " 



124 Annals of the Carnegie Museum. 

Transverse diameter of P-, ■ 6 mm. 



Antero-posterior 



Transverse 

Anter-posterior 

Transverse 



" P4, approximately 10 mm. 9 

" Mx " II 

" M2 13 

" " 9 

" M:i 18 



The close relationship existing between Isectolophus and Parisec- 
tolophiis was in reality recognized by Professor Osborn^"- and Mr. 
Hatcher;^"^ the former included Helaletes latidens under Isectolophus, 
and the latter expressed the opinion that this same specimen should be 
made the type of a new genus "if it can be shown to differ generically 
from Helaletes^ Dr. Wortman and Mr. Earl referred to this same 
Bridger specimen as Isectolophus latidens. 

With the more complete knowledge which we now possess of the 
characters of the type of Helaletes it is obvious ( i ) that the genus should 
be separated from Parisectolophus as well as from Isectolophus; (2) 
that many cranial features as well as certain characters of the teeth in 
Helaletes agree as well, or better with those of the recent tapirs than is 
the case with the characters with which we are acquainted in Homo- 
galax, Parisectolophus, or Isectolophus, (3) that the generic separation 
instituted by Wortman and Earl^"* between Ilcptodon and Ilomagalax 
should be accepted. 

According to my view of the facts in the case Heptodon may be 
regarded as in the ancestral line leading to Helaletes, while Ilomogalax 
is equally clearly in the line leading to Parisectolophus and most 
probably to Isectolophus. But we cannot seriously claim that either 
Helaletes, Parisectolophus, or Isectolophus are direct ancestors of the 
modern tapirs. 

Wortman, Earl, and Hatcher have questioned the view that the 
true ancestor of the tapirs is Isectolophus}'^^ Wortman and Earl say: 
(/. c, p. 171) "The crests of the inferior true molars are nearly cres- 
centoid in form, which character we should not expect to find in a 

* The jaw is crushed in this region and the measurement is not rehable. 

102 '"phe Mammaha of the Uinta Formation," Trans. Amer. Philos. Soc, Vol. 
XVI, 1889, p. 519. 

103 "Recent and Fossil Tapirs," Amer. Jour. Sci., Vol. I, 1896, p. 178. 

^oi Bull. Amer. Mus. Nat. Hist., Vol. V, 1893, pp. 178-180; I.e., Vol. \TII, 1896, 
pp. 85-86 (Wortman). 

105 Wortman, J. L., and Earl, Charles, Bull. Amer. Mn.s\ Nat. Hist., Vol. V, 1893, 
p. 171; Hatcher, J. B., Amer. Jour. Sci., Vol. I, 1896, p. 178. 



Peterson: Material Discovered in Uinta Basin. 125 

Uinta tapir. The last lower molar has also a very large third lobe. 
Some of the characters above adduced as to the teeth of Isectolophus 
annectens may show by later discoveries that this species is not in the 
direct line leading to the true tapir." On page i8o of the same pub- 
lication is found, by the same authors, the following: "In fact all the 
known species of both Systemodon [Ilomogalax] and Ileplodon are 
extremely slender forms, as compared with their supposed Miocene 
[= Oligocene] successors, and if we derive the true tapirs and pseudo- 
tapirs from any of the known species of either of these genera we must 
suppose a considerable modification of their foot-structure to reach 
the condition found in their Miocene [= Oligocene] relatives. The 
dentition of these early Wasatch and Wind River Tapiroids, however, 
is well adapted for further evolution into later Miocene [= Oligocene] 
types, but in their foot-structure we find it otherwise." 

Mr. Hatcher objects to regarding Isectolophus as the ancestor of the 
tapirs as follows: (/. c, p. 178) "Themetacone \n Protapiriis is placed 
farther in, and is less prominent and not so convex externally as in 
Isectolophus, while the same element in recent Tapirs is more prominent 
and has a more external position than in Isectolophus. Thus, accord- 
ing to our present phylogenetic arrangement we should have to allow 
for first a gradual shifting inward of the position of this cone followed 
by a period when it commenced to move outward to its normal position 
in modern Tapirs, a rather extreme case of oscillation, but not entirely 
inconsistent with what Scott has shown to have taken place in the 
equine series." 

If we accept the phyletic development of the superior premolar den- 
tition of the three Oligocene species of Protapirus,^°^ namely, P. 
simplex, P. validus, and P. obliquidens, as well-established, we have 
not yet, to my mind, established any satisfactory evidence that there 
has been found in the American Eocene an ancestor for these forms. 
In Helaletes the deuterocone of P- is already slightly divided at the 
apex, while in Isectolophus and Parisectolophiis there are two distinct 
internal tubercles. On the other hand the upper premolars of Pro- 
tapiriis simplex, a much later form of the Oligocene, has only one inter- 
nal cone, the paraconule of P- being even less developed than in 
Ilomogalax from the Big Horn Eocene. From the comparative char- 

1"^ Hatcher, J. B., Amer. Jour. Set., Vol. I, 1896, p. 179. (Hatcher questions the 
propriety of referring the American species of the Ohgocene tapiroids to the Eur- 
opean genus Prolapirus.) 



126 Annals of the Carnegie Museum. 

acters of the upper dentition, of which we have just spoken, taking no 
account of the limb and foot structure of the Eocene pseudotapirs, 
we find difificulty in accepting as natural such a fluctuation as we 
must admit to have occurred were we to place Isectolophns or Parisec- 
tolophus in the direct line of ancestry to Protapirus. These Eocene 
tapirs of North x^inerica, if they continued into the Oligocene, had as 
their successors, not Protapirus, but some form contemporaneous 
with the genus Colodou. The true Eocene ancestors of the Oligocene 
and recent tapirs cannot logically be said to be represented by the 
remains heretofore found in the Eocene of North America. 

Isectolophns, Parisectolophus, and possibly Schizolophodon^'^'' appear 
to form a second independent line of Eocene tapirs paralleling that 
represented by Heptodon, Helaletes, and Colodou. 

Since we now know the upper dentition of Isectolophns better than 
heretofore, it is obvious that Professor Osborn's inclination to refer 
Riitimeyer's Lophiodon annectens to Isectolophns,^"^ and the acceptance 
of this reference by Trouessart^"^ and others, cannot now consistently 
be continued. Riitimeyer's illustrations '^° indicate different outlines 
and configurations of the grinding surfaces, besides greater height 
of the cross-crests and a greater amount of cement in the different ele- 
ments of which the teeth are composed; in the latter respect apparently 
more rhinocerotic than is the case in Isectolophns. The name Lep- 
tolophiodon annectens (Riitimeyer) might be proposed for this European 
form, which must in any event be excluded from the American genus. 
Furthermore, it is questionable whether this European genus should 
even be included in the Isectolophincc. 

"^ The imperfect cross-crests of the molars of Schizolophodon suggest a conser- 
vative type, which recalls the condition in Homogalax or Eohippus, and may belong 
in a separate subfamily. 

^°^ Amer. Naturalist, Vol. XXVI, 1892, p. 763. 

1°^ Catalogus Mammalium, p. 765 (riithneyeri). 

^^° Abhand. Schw. Pal. Ges., 1891, p. 26, PI. I, Figs. ii-i3- Note: Specimens 
represented by Fig. 12 in Rutimej^er's illustrations (PI. I) are perhaps most sug- 
gestive of the American genus Isectolophns, but according to the illustrations, the 
cross-crests are more curved, the ectoloph thicker, and the teeth contain more 
cement. 



Peterson: Material Discovered in I^inta Basin. 127 

Superfamily RHINOCEROTOIDEA. 

Family HYRACODONTID.4^. 

Subfamily HYRACHYIN.4^. 

Genus Desmatotherium Scott. 

39. Desmatotherium guyotii Scott (Plate XLIV, Fig. 6). 

Contributions from The E. M. Museum, Princeton, N. J., Bull. No. 3, 1883, p. 
46, Pl.'.VIII, Figs. 1-3. 

Type: Upper jaws, No. 10166, Princeton Museum Catalogue. 

Horizon: Eocene Formation (Bridger Beds). 

Locality: Henry's Fork, Wyoming. 

Generic Characters: Professor Scott defines the genus as follows: 
"Lophiodonts closely allied to Ilyrachyiis, having the molar teeth 
constructed exactly as in that genus, but differing from it in the pattern 
of the third and fourth upper premolars, which have two internal cusps 
instead of one. Dental formula: ^ c. 5^^ Pm. -^fy M -^." 

In his specific determination Scott states that the postero-internal 
cusp of P- "is very small and situated somewhat exterior to the 
antero-internal cusp." His illustration (/. c, PI. VHI, Fig. 3) also 
appears to agree with this statement. Both are erroneous and clearly 
a mistake, since his detailed description of P- apparently agrees better 
with the actual type specimen. Upon the reexamination of the 
specimen (See PI. XLIV, Fig. 6) it is clear that the tetartocone is on a 
more nearly even antero-posterior line with the deuterocone, and 
that the cusp represented in Professor Scott's illustration is an en- 
largement, or an intermediate tubercle located on the posterior cross- 
crest. The tetartocone is very much ground down by the opposition 
of the inferior tooth. "^ The true condition is better shown on P^ of 
the left side, which is completely preserved, and from which the draw- 
ing represented by the present illustration is partly restored. 

Scott states (/. c, p. 48) that the true molars of Desmatotherium 
are, with few minor differences, very similar to those in Hyrachyiis, 
and the matter here needs no further discussion, except to say that 
these two genera are indeed very closely related. 

Comparison and Relationship. 
With regard to the taxonomic position of Desmatotherium I am 
strongly impressed with the idea that the genus is not directly ances- 

"■ It is possible that this tubercle was broken off and healed over before the 
death of the animal. 



\^ 



128 Annals of the Carnegie Museum. 

tral to the tapirs, but that it is closer to the Rhinoceroses. Again 
referring to /. c. Pla e XLIV, Fig. 3, it is quite clear that the premolars 
have unmistakably advanced towards such Ol'igocene genera as 
Leptaceratherium trigonodum, Ccvnopus mile, or C. platycephalum. 
This implies a very early origin of the "atypical" premolar structure, 
to which Professor Osborn calls attention in his work on the Rhi- 
noceroses referred to above. Nor is there any apparent reason for 
excluding Ilyrachyus agrarius or Colonoceras agrestis as- also possibly 
belonging to the Rhinoceroses, the latter genus to Diceratherium as 
Marsh originally suggested."^ The Rhinocerotidce had, in the Uinta, 
the Bridger, or even earlier Tertiary time, most likely made more than 
the initial start towards their varied specilizations seen in later time. 
I am inclined to the opinion, that, if we had found in America the true 
middle Eocene ancestry of the various types of Rhinoceroses, in the 
later epochs we would have met with even further advancements than 
are recorded in the Bridger or later Eocene."'* The Amynodonts of 
the Uinta are certainally surprisingly advanced along their line. 
When the dentition of Desmatotherium is compared with that of 
Isectolophus and Parisectolophus and the earlier genus Homogalax the 
tendency toward the vertical increase of the crowns, or greater de- 
velopment of the cement is clearly observed in Desmatotherium. 
Besides the early origin of the tetartocone of P-, the cross-crests are 
visibly of greater prominence and the valleys deeper than in the teeth 
of Parisectolophus or even the Uinta genus Isectolophus. Altogether 
the teeth, especially the molars, are more nearly like those in Hyra- 
chyus, Colonoceras, or Hclaletes. In Ilyrachyus there is, however, 
present a crista which is only faintly or not at all indicated in Des- 
matotherium. In Colonoceras the crista is slightly better developed 
than in Desmatotherium and the posterior cross-crests of the superior 
premolars are also located further foward in Colo?ioceras.^^* Scott's 
genus Desmatotherium should be placed with Ilyrachyus in a line 
nearer to the Rhinocerotidce. It certainly appears to be more closely 
related to that group that to the Tapiroidea, where it has been 
heretofore placed. 

"- Amcr. Jour. Sci. (3), Vol. XIV, 1877, p. 362. 

"^ Europe during this time was nearer, or perhaps more accessible, to the center 
of dispersion of the true Rhinoceroses, according to the works of Schlosser, Abel, 
and others. 

"'' If these features in Colonoceras are constant they may be regarded as of con- 
siderable importance. 



Peterson: Material Discovered in Uinta Basin. 



129 



Genus Hyrachyus Leidy. 

Proc. Acad. Nat. Sci. Philad., 1871, p. 229. 

Hyrachyus is represented by two individuals C. M. Nos. 2908 and 
3112. The former consists of a pair of lower jaws minus the angle 
and the incisor teeth, and was found in the upper portion of horizon 
A, near White River, Uinta Basin, Utah, while No. 31 12 is a fragment 




Fig. 18. Hyrachyus grande sp. nov. Type Carnegie Museum No. 2908. X 3/10. 

of a lower jaw with Pj and My in place and a fragment of the maxillary 
with only the roots of the premolars present. The latter was found 
in horizon B, southeast of Kennedy's Hole, Uinta Basin, Utah. 
While the lower jaws, No. 2908, represent an animal larger than any 
pertaining to that genus heretofore found, and may probably belong 
to a new species, or possibly even a new genus, there are unfortunately 
no characters present indicating any marked advance over those 
found in the Bridger genus. As stated above, the incisors are lost. 




Fife. 19. I. Hyrachiiys grande sp. nov. Type. Carnegie Museum No. 2908. X 

2/5. 2. Hyrachyus princeps Marsh. M3- of type specimen in Peabody 

Museum of Natural History. X 2/5. 

but from their alveoli it is possible to quite definitely determine that 
they were all of subequal size, the two intermediates possibly some- 
what larger than the laterals. The canine is possibly slightly reduced 
when compared with Hyrachyus agrarius, but this may well be a 
sexual character and not of specific importance. The premolars are 
all of the Hyrachyus pattern, with practically no indication of a further 
9 — DEC. 16, 1919. 



130 Annals of the Carnegie Museum. 

step towards the development of a posterior cross-crest, such as is 
found in Trigonias oshorni from the base of the Oligocene. Deciduous 
Py is present in the right ramus (see Fig. 19). My and My are much 
worn, the crowns furnishing no detailed characters, while M^ practi- 
cally agrees in all details with the same tooth of the known species 
of the genus, except in its larger size. The ramus appears to be 
proportionally slenderer than in Hyrachyus agrarius. Hyrachyus 
princeps Marsh, which also is a large form, is smaller than the 
specimen under description. If the specimen does not belong 
to the latter species ^^^ it may be called Hyrachyus grande sp. nov. 
simply to indicate its extraordinarily large size (See Figs. 18 and 19). 
The specimen No. 31 12 is slightly larger than Hyrachyus agrarius 
Leidy, and appears moreover to differ from the latter by a proportion- 
,ally greater development of the metaconid and a slight basal elevation 
indicating the tetartoconid. 

Measurements. 

No. 2908. No. 3112. 
Length of ramus from anterior point of symphysis to and in- 
cluding M3 215 mm. 

Depth of ramus at Py 40 

Mg- 54 " 

Antero-posterior diameter of cheek-teeth 161 

" P^ 21 " 15 mm. 

Transverse " " " 15 " n 

Antero-posterior " " My 26 " 22 

Transverse " " " 19 " I4 

Subfamily AMYNODONTIN.F:. 
Genus Amynodon Marsh (Plate XLVII, Figs. 5-7). 

Amer. Jour. Sci., Vol. XIV, 1877, pp. 251-252. 

This genus is represented by a number of individuals. The material 
has a considerable range in size and undoubtedly represents two or 
probably three species. Unfortunately the fragmentary condition 
of the greater number of the specimens does not admit of an accurate 
identification. The smaller individuals are therefore provisionally"® 

11^ There is unfortunately no other basis of comparison between these two speci- 
mens besides M 3^. In H. princeps the cingulum of the anterior face is heavier and 
the anterior extension of the metalophid is somewhat less developed than in the 
Uinta specimen. These may or may not be valid characters. 

116 All the specimens referred to A. advenum are too small when compared with 
the measurements of the type specimen by Marsh. One specimen in our Collec- 
tion, No. 3217, is especially small and may represent a new species. 



Peterson: Material Discovered in Uinta Basin. 131 

placed in Amynodou adveiiiim (Marsh)"" while the larger are referred to 
Amynodon intermedium Osborn."^ Of the smaller specimens those 
which have the upper and lower premolars preserved show three in 
either jaw, which according to Osborn's reidentification (/. c, 1889, 
page 507) would, except in size, agree with Amynodon advenum. 
The best preserved specimen representing the larger species, is C. M. 
No. 3200, the greater portion of a skull of an adult, perfectly symmetri- 
cal. Unfortunately, however, the dentition is represented only by 
the roots of the posterior premolars, the roots of the molars, and a 
portion of the crown of M^ (PI. XLVII, Fig. 5). This portion of M^- 
and the general large size of the teeth agree with A. intermedium and 
an outline drawing of the dentition of the type as figured by Osborn 
(/. ("., PI. X, Fig. 10) is here added in connection with the palatal view, 
in order to better aid the student in comparative work (See PI. XLVII, 
Fig. 6). A side view of the specimen in the Carnegie Museum is 
given in Fig. 7 on the same plate, in order to supplement the figures 
given by Osborn of the dentition and maxillary region, the only por- 
tions of the type specimen preserved. The description of the skull, 
which belongs to the articulated young skeleton exhibited in the 
American Museum of Natural History, referred to A. intermedium 
by Osborn"^ appears to agree quite well with the present cranium. 

Subfamily HYRACODONTIN^. 
Genus Prothyracodon Scott and Osborn. 
40. Prothyracodon obliquidens Scott & Osborn (Plate XLVI, Figs. 
1-9). 

Proc. Amer. Philos. Soc, Vol. XXIV, Sept. 2, 1887, p. 260. 

Generic Characters: If, c\, P^^, M^. Incisors subequal in size; 
upper canine followed by a short diastema, loiver canine close to Ij; 
a general advance of the cheek dentition toivards the Ilyracodonts and 
Rhinoceroses of the Oligocene; that is, a decided development of the crista, 
crochet, and anticrochet together with the Rhinocerotic structure of M- 
(Prothyracodon nintense, vide infra). Fore limb proportionally shorter 
than in Triplopns. Animals larger than Triplopus. . 

It is a matter of surprise to find such differences in the proportionate 

1'^ Awer. Jour. Sci., Vol. IX, 1875, p. 244. 

118 "The Mammalia of the Uinta Formation," Trans. Amer. Philos. Soc, Vol. 
XVr, 1889, p. 508. 

^^^ Bull. Amer. Mus. Nat. Hist., Vol. VII, 1895, p. 95. 



132 Annals of the Carnegie Museum. 

length of the humerus and radius of Professor Cope's genus of the 
Washakie Eocene Triplopus cubitaUs and the recently acciuired 
material of Prothyracodon from the Uinta deposits. It is thus seen 
that the genus from the Uinta has the fore limb much more like that 
of the Oligocene genus Ilyracodon than is the case with the genus 
from the Washakie. The type of the latter was, according to Pro- 
fessor Cope, "cut from a block of calcareous sandstone" and Dr. 
Matthew of the American Museum of Natural History assures the 
writer that there is no doubt as to the correct association of the parts 
of the type specimen upon which Cope founded the genus. ^'^° A re- 
stud}' of the type of Triplopus cnbitalis, compared with the material 
in the Carnegie Museum, results in definitely placing the form from the 
Uinta in a separate genus, Prothyracodon, as was originally done by 
Scott and Osborn. The generic rank of the form from the Uinta was 
called into question by Professor Osborn in a later publication.^-^ 

Prothyracodon obliquidens is represented in the Carnegie Museum 
by a number of individuals. These are all more or less fragmentary, 
but serve to throw further light on the limb-structure of this cursorial 
Rhinoceros from the Uinta Eocene. C. M. No. 2942 is represented 
by both fore limbs and is of approximately the same size as the type 
of Prothyracodon obliquidcns. A second specimen, C. AE No. 3199, 
not fully adult, and slightly smaller, is represented by both fore and 
hind limbs. The former was found by the writer in the lower portion 
of horizon C, six miles East of Myton, Utah, while the latter was 
found by Mr. Earl Douglass in horizon C further east in the Uinta 
Basin. The scapula of No. 2942 is represented only by a fragment of 
the proximal end (See Pi. XLVI, Fig. 5). It agrees with the descrip- 
tion and illustration by Professor Cope.^-'^ In a specimen, which 
probably pertains to another species, described on page 134, the scapula 
Is fairly well-preserved. This bone is quite elongated, with a long neck. 
The spine, however, rises more rapidly than that in Hyracodon of the 
Oligocene and differs further from that genus by having apparenth' the 
acromion process situated lower down or nearer the glenoid cavity,'-^ 
and by having a proportionally shorter and broader blade. 

'-» Tertiary Vertebrata, p. 684, PI. L\'Ia. 

'-'"The Mammalia of the Uinta F"ormation," Trans. Aincr. Philos. Soc, \'ol. 
X\T, 1889, p. 524. 

'22 Tertiary Vertebrata, p. 684, PI. LVIa. 

'23 The spine in this region is broken, but what still remains is proportionally 
higher and differs otherwise from that in Hyracodon and is more suggestive of 
Mesohi l^pHs. 



Peterson: Material Discovered in Uinta Basin. 133 

The humerus is laterally compressed proximally, which is in part 
due to crushing. The greater tuberosity rises considerably above 
the head. There is a large deltoid groove and the deltoid ridge is 
prominent, but does not have the large and rugose tuberosity seen in 
Ilyracodon, and the supinator ridge is also less developed. The distal 
trochlea is relatively slightly narrower than in the latter genus, but 
resembles it closely in height, its very prominent and narrow inter- 
condylar ridge, and narrow outer condyle (PI. XLVT, Fig. 4). 

The radius is only very slightly longer than the humerus, and in 
'this respect is quite similar to the same bone in Hyracodon, though 
slenderer. Its head is not greatly expanded and the shaft is broad 
and rather compressed antero-posteriorly, with a considerable bow 
in the same direction, while distally it is expanded both laterally and 
antero-posteriorly, with deeply excavated facets for the scaphoid 
and the lunar. The shaft of the ulna is more reduced than in Hyra- 
codon of the Oligocene, but, as in that genus, it is at no place coossified 
with the radius. 

The carpus agrees with the description given by Osborn (/. c, 
pp. 527, 547) except the trapezium, which according to Osborn is 
greatly reduced. This is probably a mistake, since the trapezium of 
one individual (No. 2336) of the Carnegie Museum is of considerable 
size, and the large facet on the radial palmar angle of the trapezoid 
of the specimen under description indicates a bone proportionally 
quite as large as in Hyracodon. 

The metacarpals are all complete, and plainly show that they are 
much shorter proportionally than in Mesohippus, the genus with 
which Professor Osborn compared the pes of Triplopus. Mc. I is 
entirely absent, Mc. II and IV are reduced in size, but not quite as 
much as in Hyracodon, \n\\\\q their length in comparison with Mc. Ill 
is fully as much or even more reduced than in that genus. Mc. V 
is reduced to about the same extent as in Hyracodon (See Plate XLVI, 
Fig. 2). 

The median phalanx of the proximal row is broad and depressed, 
while those of the lateral digits are higher. The ungual phalanges 
are also depressed, somewhat pointed and cleft, suggesting the features 
of the early horses to a remarkable degree. 

With No. 3199, already referred to above, there is unfortunately 
only preserved a fragment of the pelvis, the femur, and the metatarsals, 
while the entire length of the tibia, except the epiphysis of the prox- 
imal end, and the greater portion of the tarsus is present. 



134 Annals of the Carnegie Museum. 

The length of the tibia is very little, if any, greater than the radius. 
The bone is quite symmetrically formed with a prominent cnemial 
keel well confined to the proximal end, unlike what occurs in Hyracodon, 
in which it extends somewhat further down. The shaft has a slightly 
backward curve, while the distal trochlea is deeply excavated and is 
very oblique, in this respect closely suggesting the Hyracotheres. 

The astragalus of No. 3199 has received slight lateral crushing, 
which makes it appear narrow in comparison with the specimen illus- 
trated by Professor Osborn (/. c, PI. XI, Fig. 9). The pes as a whole is 
in reality slightly smaller than in the type, which is, no doubt, due to^ 
the incomplete maturity of the specimen at hand. The different 
bones of the tarsus agree quite well with the description of the material 
in Princeton University furnished us by Osborn (/. c, p. 549), and 
needs no further description, except to say that the entocuneiform, 
though quite large, and with the plantar process somewhat like that 
of Hyracodon described and figured by Professor Scott, ^-^ has not 
formed an articulating facet with the plantar process of Mt. Ill, and 
the process itself has not attained the prominence seen on the ento- 
cuneiform of the Oligocene genus (See Plate XLVI, Figs. 7-8). 
41. Prothyracodon uintense sp. nov. (Plate XXXVI, Fig. i; Plate 
XLV; Plate XLVI, Figs. 10-16). 

Type: Skull and lower jaws of young individual C. M. No. 30076!.^-^ 

Horizon: Uinta Eocene, horizon C. 

Locality: Six miles east of My ton, Utah. 

With the type specimen are provisionally associated three other 
specimens in the Carnegie Museum as probably pertaining to the 
same species. Of these No. 2990 consists of fragments of the lower 
jaws, limbs, and foot-bones; No. 3097, portion of vertebral column, 
scapulae, and humerus, and No. 3399, fragments of vertebrae. Of 
these specimens No. 3007 was found together with the type, but is of 
an adult individual, the others were found in the same general locality 
and in the same horizon. 

Specific Character: Protoloph of molars with a ivell-markcd aut:crochet. 
M- with smooth posterior face, no spur of the ectoloph-at all indicated. 
Teeth proportionally large, zvhen compared with P. obliqiiidcns. Animals 
larger than P. obliqnidcns. 

'-■•"Osteology von Hyracodon Lcidy," Festschrift fiir Carl rioii;cnl)aur, Leipzig, 
1887, p. 377. PI- I. Fig- 7- 

'25 The type specimen was found together with the remains of other individuals 
of different genera. 



Peterson: Material Discovered in Uinta Basin. 135 

In the type, No. 3007a, there are three lower incisors, which are 
procumbent in position, of subequal size, with laterally expanded or 
fan-shaped crowns (PI. XXXVI, Fig. i). The canine is placed close 
to l-r^ and is but very little larger than the incisors in the type, wliile in 
an adult specimen No. 2990^-^ it is slightly larger in proportion. On 
the whole the tooth in this species is relatively larger than in the 
Oligocene genus Ilyracodon and it is also more oval in cross-section. 
In the type there are four upper and lower milk-teeth, while in the 
adult specimen. No. 2990, Py is absent.^'" In the Uinta genus there 
is not developed the strong rib near the median portion of the ectoloph 
on the lower molars and premolars, nor is the heavy cingulum seen in 
Ilyracodon present. There is in the type specimen a decided swelling 
on the posterior face of the protoloph, represerlting the antecrochet, 
and the prefossette is well indicated, but the crista is little or not at 
all represented, while in Hyracodon the crista is much better developed 
and the antecrochet is prominent. M- is just appearing through the 
alveolar border. This young tooth has the posterior face of the ecto- 
loph perfectly smooth, there being no spur of the ectoloph represented 
as in Triplopus cuhitalis or Prothyracodon obliqiddens. This feature 
was thought to be due to the immaturity of this tooth in the Carnegie 
Museum specimen, but it is perhaps more probable that the present 
well calcified crown would not before its final eruption add the char- 
acteristic spur seen in Hyracodon of the Oligocene or in the contem- 
porary species of the Uinta deposits. Too much stress, however, 
should not be laid upon this highly interesting Rhinocerotic feature of 
M- in the specimen at hand, until the discovery of fully adult upper 
dentitions. If this character is found in M- of fully adult specimens, 
P. uintense should be placed in a distinct genus. 

The base of the skull appears to be proportionally broader than in 
Hyracodon, which may be due to crushing. Whether or not there is a 
tympanic bulla cannot be determined. Unfortunately the important 
region of the external ear is also too much mutilated to determine 
whether it is like that in Hyracodon or whether it is closed inferiorly 
as in Triplopus cuhitalis. There is a long and well defined sagittal 
crest and the supra-orbital ridges are also indicated in this young 
individual. 

'2° No. 2990 is provisionally referred to P. uinlense on account of its large size, 
which is taken as an indication that it pertains to the same species. 

12" The first premolar above and below are very rudimentary, and their absence 
or presence in this genus is perhaps not of great phyletic importance. 



136 Annals of the Carnegie Museum. 

The postglenoid process is unusually heavy. In a specimen, No. 
3201, referred to P. ohliquidens the post-tympanic and paroccipital 
processes are separated, as shown in Plate XLVI, Fig. i, and answers 
quite well to Cope's description. Furthermore the actual specimens 
have been compared, and are found to be quite alike in this respect. 
There are a number of similarities in the contour of the present speci- 
men and the young specimen on which Cope established his Triplopns 
c II bi talis. 

These similarities may be regarded as of comparatively little 
importance since both specimens pertain to young animals. 

In comparing Prothyracodon with Doctor Koch's Prohyracodon^-^ 
as described and illustrated in Doctor O. Abel's work^-^ it is at once 
seen that M- and M^ of Prohyracodon, though mutilated, show the 
proto- and metalophs to be at a more nearly right angle to the ectoloph, 
the proto- and metalophs are also of more nearly subequal size than in 
the American genus. Furthermore there is in Prothyracodon a 
better defined antecrochet, a proportionally longer M-, and M- has a 
more nearly triangular outline, than in the specimen preserved at 
Budapest. 

M EASUREMENTS. 

Total length of skull from condyle to and including d. P- 165 mm. 

Antero-posterior diameter of deciduous upper dentition 42 

" M' and M- 36 

. " " " 18 

Transverse " " M- 16 

" M^ 19 

Antero-posterior " " " 22 



deciduous lower dentition 42 

permanent My and M2 32 

Mt 15 



Transverse 



" M2 •. 10 

Antero-posterior " " " 16 

On comparing the atlas of No. 3007 with that of Ilyracodon, the 
similarities are remarkably close. Thus it is seen, that, as in the 
latter genus, the bone is high^^^ and rather short, with a backward 

128 Koch, A., Termeszetrazi Fuzetek, Budapest, XX, 1897, pp. 490-500, Pis. 
XII-XIII. 

^'^^ AbhancU. der K. K. Geolog. Reichsanslall, Bd. XX, 1910, pp. 24-25, PI. II, 
Fig. I. 

131 The atlas of Prothyracodon appears to be proportionally higher than in 
Hyracodon, which is to a certain extent due to crushing. 



Peterson: Material Discovered in Uinta Basin. 137 

projecting transverse process, i)erforated by a canal of considerable 
size, which again appears on the under surface of the transverse 
process and forms a deep atlantal groove (not foramen), at the an- 
terior base of the same process. The anterior cotyles are deep and 
broadly excavated above and below, as in Ilyracodon, but the accessory 
facets for the base of the occipital condyle are less developed than in 
the latter genus. The above description also agrees with that of 
Triplopus cuhitalis by Professor Cope. 

The axis of No. 3399, a specimen approximately of the same size as 
No. 3007, is considerably longer than the atlas. The bone again 
agrees with the description of Triplopus cubitalis by Cope,"^ and of 
Hyrachyus by Scott. ^^^ Thus the bone is relatively longer and 
slenderer than in the Rhinoceroses and suggests the axis of the horses 
through its prominent spine and odontoid process, ventral keel, and 
the oblique and concave posterior face of the centrum. 

The remaining cervicals, which are represented in specimen No. 
3007, appear to gradually shorten from the axis backward, and agree 
generally, so far as comparison can be made, with those in contem- 
poraneous Uinta species as well as Hyracodon. 

Measurements. 

No. 3007. No. 3399. 
Transverse diameter of anterior cotyle 39 mm. 

Vertical " " " " 22 " 

Greatest vertical diameter of atlas 33 " 

" length of axis, odontoid process not included 48 mm. 

Length of odontoid process 14 

Besides the specimens described above, there are in the Carnegie 
Museum a number of individuals from the same locality and horizon, 
which perhaps pertain to a third species intermediate in size between 
Prothyracodon obliquidens and P. uintense. It is thought best to defer 
adding more species until more complete specimens are obtained. 
Two species of Prothyracodon from the Uinta were originally proposed 
by Scott and Osborn, which Osborn united in 1889 (/. c, p. 525). 

I do not hesitate in expressing my agreement with earlier workers 
(Scott, Osborn, Wortman, Earl and others), in regard to the phylogeny 
of Prothyracodon. From the evidence at hand there is compara- 
tive certainty that some genus closely allied to Hyrachyus should 

I'l Tertiary Vertebrata, p. 683. 

132 "Die Osteologie von Hyracodon Leidy," Festschrift fur Carl Gegcnbatir, p. 363. 



138 Annals of the Carnegie Museum. 

be regarded as the Bridger representative of this line of cursorial 
Rhinoceroses.^^^ The Washakie genus Triplopus and the Uinta form 
Prothyracodon uintense^^^ obviously represent independent lines, 
which may or may not be represented in the Oligocene,"^ while P . 
obliquidens is in all the obtainable characters so very closely related 
to the Oligocene genus Hyracodon that one cannot deny the phylo- 
genetic relationship here displayed. It is very unfortunate that we 
have not in the recently acquired collection from the Uinta a skull 
sufficiently complete in the region of the tympanum and the external 
ear to verify Professor Cope's studies of Triplopus cubitalis. Con- 
sidering all the known characters of Prothyracodon which are so very 
suggestive of Hyracodon, I believe that the Uinta form did not have 
the meatus closed inferiorly as in Triplopus cubitalis. This would 
substantiate Cope's position in placing Triplopus in a separate syste- 
matic position. Together with this equine feature of the external 
ear in Triplopus we now know that the limbs were also proportionally 
longer than in the Uinta genus. The genus apparently does, in fact, 
represent a subfamily {Triplopodincc) of the Hyracodontidce which 
holds an equal rank to the Prothyracodon-Hyracodon phylum. 

Together with the highly Rhinocerotic feature of M- in Prothyra- 
codon uintense the first upper and lower premolars are altogether too 
much reduced in size^^'' to be seriously regarded as a forerunner of the 
rhinoceroses of the Oligocene. Furthermore the lower canine and 
incisor series are typically those of Hyracodon, plainly excluding this 
species from the true rhinoceroses of the later Tertiary. In my opinion 
it is altogether possible, that, if this line continued in later epochs, we 
may find a Hyracodon-like form in the Oligocene with M- reduced to 
the characteristic features of the Rhinocerotidcc so strongly suggested 
in the Uinta genus. 

1'' The actual type of Hyrachyus implicalus Cope I have not seen, but from the 
splendid illustrations (Tertiary Vertebrata, PI. LVIII, Figs. 6, 6a, 7) by Professor 
■Cope, it appears to be in this line. Its dentition seems to be advanced in the direc- 
tion of the Hyracodonts. 

134 \Yith the exception of M', Prothyracodon uintense bears a closer relation to 
Hyracodon nebrascencis than does Triplopus cubitalis. 

1^5 The different forms ol Hyracodon of the lower and upper Oligocene are as yet 
comparatively little known. 

'28 Both the upper and lower first premolars may well be absent in fully adult 
specimens of P. uintense as they are, in fact, seen to be in individuals in the Car- 
negie Museum. 



Peterson: Material Discovered in Uinta Basin. 139 

CHALICOTHEROIDEA}^'' 
Subfamily SCHIZOTHERIIN^? Holland and Peterson. 
Genus EomoRopus Osborn. 
42. Eomoropus annectens sp. nov. (Plate XXXVI, Fig. 2). 

Type: Portion of skull with the cheek-dentition in position, C. M. 
No. 3109. 

Horizon: Uinta Eocene, Horizon B. 

Locality: Eastern portion of Uinta Basin, near Wagon-hound Bend, 
on White River, Utah. 

Specific Characters: Skull comparatively narroi.v across the orbit and 
palate, long in the cranial region, less suddenly contracted laterally in 
the region of the orbit, and a lighter post-glenoid process •when compared 
icith E. amarorum. The type also indicates a considerable smaller 
animal than the latter. 

From the character available for comparison with the type specimen 
of Eomoropus amarorum (Cope) the present species is, as already said, 
of considerably smaller size, especially in the dentition. Eomoropus 
amarorum differs from E. annectens by the relatively greater measure- 
ments across the maxillary from the inner face of M- to the lower 
external face of the jugal, as seen in the illustrations given by Pro- 
fessors Cope^^^ and Osborn,"^ as well as on the actual comparison of 
the specimens by the writer. The entire posterior portion of the skull 
of E. annectens, except the postglenoid process, was weathered out of 
the sandstone in which the maxillaries w^ere found. The process 
appears to be in its natural position with relation to the zygomatic 
arch and maxillaries. From this fact it is presumed that there is a 
relatively greater distance between M- and the postglenoid process 
in the present species than in E. amarorum. 

There were probably only three premolars present. Through their 
relatively small size and the details of structure the premolars suggest 
those of Schizotheriiim priscum (Gaudry) in a remarkable manner. 
The molars on the other hand are proportionally shorter and broader. 
Furthermore the parastyle is more loosely connected with the para- 
cone, and the vertical ridge on the external face of the paracone is 

"'^Professor Osborn has recently (Bull. Amer. Nat. Hist., Vol. XXXII, 1913, 
pp. 261-274) placed Triplopiis aynaroriiin Cope in a distinct genus of the Super- 
family Chalicotheroidea. 

138 Tertiary Vertebrata, PI. LVIIIa, Fig. 20. 

"^ L.c, page 262, Fig. 2a. 



140 Annals of the Carnegie Museum. 

more prominent. The general construction of the molars in the two 
genera is, however, strikingly similar. On comparing the present 
specimen with other European genera, it may be said that Eomoropus 
annectens and Macrotherinni grande Lartet resemble each other, 
more especially in the proportions of the molars. It thus appears 
that if Eomoropus is in the ancestral line leading directly to Moropus 
the molar series underwent a considerable lengthening during the 
Oligocene period. On the other hand, if the postglenoid process, as 
found, is in its correct position in the type of E. annectens, which seems 
quite likely, then this region of the skull had already taken on the 
characters found in Aloropiis to a greater degree than seems to be the 
case in E. amarorum. More perfect remains must be found and con- 
sulted before anything final in regard to the true phylogeny of these 
Eocene Chalicotheres of America can be stated. 

Measurements. 

Length of skull P- to postglenoid process 129 mm. 

Length of skull from AT- to postglenoid process 58 " 

Transverse diameter from internal face of M- to inferior border of jugal 

opposite the orbit 31 " 

Length of the cheek-dentition 73 ' " 

Length of the premolars 29 

" " molars 47 

Antero-posterior diameter of P— 10 " 

Transverse diameter of P- 9.5 " 

Antero-posterior diameter of P- 11 " 

Transverse " " " 12.5 " 

Antero-posterior diameter of P* 10 " 



Transverse 

Antero-posterior 

Transverse 

Antero-posterior 

Transverse 

Antero-posterior 

Transverse 

Carnegie Museum, 
May 29, 1916. 



13 

Ml 14 

" 13-5 

M^ 17.5 

" 18 

M-3 17.5 

" 19.8 



SYNOPSLS OF CONTENTS OF PAPER. 

Introduction 40 

Pisces 40 

Reptilia 41 

Mammalia, Mesonychida; 41 

Miacidae 48 



Peterson: Material Discovered in Uinta Basin. 141 

Rodentia (Glires) Ischyioinyida' 60 

Muridae 66 

Artiodactyla, Homacodontinae 66 

Anoplotheriidae 76 

Achsenodontidse 79 

Agriochceridae 82 

Camelidae 88 

Hj'pertragulidae 93 

Perissodactyla, Hyracotheriinae loi 

Pseudotapirs of the North American Eocene 103 

Helaletidse 103 

Isectolophidas fam. nov 115 

Isectolophina? subfam. nov 116 

Rhinocerotoidea, Hyracodontidae 127 

Hyrachyinae 127 

Amynodontina; 130 

Hyracodontinae 131 

Chahcotheroidea, Schizotherinae 139 



142 



Annals of the Carnegie Museum. 



Fig. 


I 




Fig. 


2 




Fig. 
Fig. 


3 

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. 


Fig. 


19. 


Fig. 


20. 


Fig. 


21. 


Fig. 


22. 


Fig. 


2 


3- 



Plate XXXIV. 

OxycBfiodon dyclerus. Scapula and humerus. C,M-. No. 3051. 
Oxycenodon dyclerus. Radius, ulna, and nianus. C,M., No. 3051. 
Prodaphcenus? robustus, type. Humerus, anterior face. 
ProdaphcBnus? robustus, type. Crown view of lower teeth 
ProdaphcEniis? robustus, type. Side view of lower jaw. 
Mimocyon longipes, type. Crown view of lower teeth. 
Mimocyon longipes, type. Side view of lower jaw. 
Mimocyon longipes, type Dorsal view of pes. 
Mimocyon longipes, type. Anterior view of humerus. 
Mimocyon longipes, type. Side view of ulna. 
Limnocyon douglassi, type. Crown view of lower teeth, right. 
Limnocyon douglassi, type. Side view of lower jaw, right. 
Limnocyon douglassi, type. Side view of maxillary, right. 
Limnocyon douglassi, type. Crown view of upper teeth, right. 
Paramys medius, type. Crown view of upper teeth. 

Lower end of tibia, side. 

Trochlea, distal end of tibia. 

Patella. 

Trochlea of astragalus. 

Pes, dorsal view. 

Proximal phalanx, dorsal view. 

Median phalanx, dorsal view. 

Alveolar border of lower jaw. 



Paramys medius, type 

Paramys m.edius, type 

Paramys medius, type 

Paramys m.edius, type 

Paramys medius, type 

Paramys medius, type 

Paramys medius, type 

Iseclolophus scotti paratype 
C. M. No. 3113- 
Fig. 24. Schizolophodon cuspidens, type 

crown view of lower teeth. 
All figures one half natural size, except Fig. 15, which is natural size. 



Alveolar border of lower jaw and 



ANNALS CARNEGIE MUSEUM, Vol, XII. 



Plate XXXIV. 




Canids from the Uinta. 
(For explanation see opposite page.) 



144 



Annals of the Carnegie Museum. 



Plate XXXV. 



Fig. I 
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. 



Pleurocyon magnum, paratype. 
Pleurocyon magnum, paratype. 
Pleurocyon magnum, paratype. 
Pleurocyon magttum, paratj'pe. 
Pleurocyon magnum, paratype. 
Pleurocyon magnum, paratype. 
Pleurocyon magnum, paratype. 
Pleurocyon magnum, paratype. 
Pleurocyon magnum, paratype. 
Pleurocyon magnum, paratype. 
Pleurocyon magnum, paratype. 
Pleurocyon m,agnum, paratype. 
Pleurocyon magnum, paratype. 



Dorsal view of pes. 

Anterior view of femur. 

Anterior view of humerus. 

Side view of ulna. 

Anterior view, head of radius. 

Articulating surface, head of radius. 

Dorsal view of Mc. V. 

Side view of upper canine tooth. 

Anterior view of tibia. 

Right lower jaw from the side. 

Upper molar, crown view. 

P4, external view. 

P4, crown view. 

Alveolar border of jaw and crown 



Pleurocyon magnum, paratype. 

view of lower teeth, left ramus. 
Pleurocyon magnum, paratype. External view of left ramus. 
All figures are one-half natural size. 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XXXV. 




Canids from the Uinta. 
(For explanation see opposite page.) 



146 



Annals of the Carnegie Museum. 



Plate XXXVI. 



Fig. I. Prothyracodon uiritense, type. Alveolar border of lower jaw and crowns 

of dentition. 
Fig. 2. Eomoropus anneclens, type. Palate and crown view of upper teeth. 
Fig. 3. Biinonieryx montanus . Palate and crown view of upper teeth. 
Fig. 4. Bunomeryx montanus. Crowns of lower teeth. 

Fig. 5. Hylomeryx anneclens, type. Palate and crown view of upper teeth. 
Fig. 6. Hylomeryx anneclens, type. Crowns of lower teeth. 
Figs. I and 2 natural size. Figs. 3, 4, 5, and 6 are slightly more than one and a 
third of nature. 




O to 



^ X 

^ O 



148 Annals of the Carnegie Museum. 



Plate XXXVII. 

Fig. I. Leplolragulus mediiis, type. Left side of skull. 

Fig. 2. Leplolragulus medius, type. Crowns of upper teeth. 

Fig. 3. Leplolragulus medius, type. Astragalus, dorsal face. 

Fig. 4. Leplolragulus medius, type. Calcaneum, dorsal face. 

Fig. 5. Leplolragulus proavus. Crowns of upper teeth. No. 2919. 

Fig. 6. Leplolragulus proavus. Crowns of upper teeth, No. 2919. 

Fig. 7. Leplolragulus proavus. Lower jaw, side view. No. 2919. 

Fig. 8. Leplolragulus proavus. Upper jaw, side view, No. 2919. 

Fig. 9. Leplolragulus proavus. Upper teeth, outer face, No. 3009. 

Fig. 10. Leplolragulus proavus. Alveolar border of lower jaw and crowns of 

teeth. No. 3009. 

Fig. II. Leplolragulus proavus. Outer face of left jaw. No. 3009. 

Fig. 12. Leplolragulus proavus. Crowns of lower teeth. No. 3195. 

Fig. 13. Leplolragulus proavus. Outer face of lower jaw, No. 3195. 

Fig. 14. Protylopus anneclens, type. Palate. 

Fig. 15. Sphenomeryx quadricuspis, type. Crown, upper molar. 

Fig. 16. Sphenomeryx quadricuspis, type. Palate. 

Fig. 17. Mesomeryx grangeri, type. Crowns of upper teeth. 

Fig. 18. Bunomeryx elegans. Outer face, right lower jaw. C. M., No. 2951. 
All figures natural size except Fig. 15, which is twice natural size. 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XXXVI 








^flN-- 





10 





Artiodactyls from the Uinta. 
(For explanation see opposite page.) 



150 



Annals of the Carnegie Museum. 



Plate XXXVIII. 



Fig. I 
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. I 
Fig. 19. 
Fig. 20. 
Fig. 21. 
Fig. 22. 
Fig. 23. 
Fig. 24. 
Fig. 25. 
Fig. 26. 
Fig. 27. 
All figure: 



Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, tj'pe 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
Diplobunops mallhewi, type 
8. Diplobunops mallhewi, type 



Diplobunops mallhewi 
Diplobunops mallhewi 
Diplobunops mallhewi 
Diplobunops mallhewi 
Diplobunops mallhewi 
Diplobunops mallhewi 
Diplobunops mallhewi, 
Diplobunops 7nallhe%vi 



type 
type 
type 
type 
type 



Radius ulna and carpus, dorsal face. 

Patella, anterior face. 

Femur, distal end. 

Calcaneum, fibular face. 

Calcaneum, dorsal face. 

Humerus, anterior face. 

Cuboid, dorsal face. 

Metapodial, dorsal face. 

Proximal phalanx, dorsal face. 

Mc. II, dorsal face. 

Mc. Ill, dorsal face. 

Scaphoid proximal face. 

Lunar proximal face. 

Cuneiform proximal face. 

Scaphoid dorsal face. 

Lunar, dorsal face. 

Cuneiform, dorsal face. 

Scaphoid distal view. 

Lunar distal view. 

Cuneiform distal view. 

Tarsus, dorsal face. 

Scapula, distal end. 

Distal end of tibia, anterior face. 



paratype. Maxillary, outer face. 

paratype. Crowns of upper teeth. 

paratype. Lower jaw, external face. 
Diplobunops mallhewi, paratype. Lingual phalanx, dorsal face. 
; one-half natural size except Figs. 24, 25, and 26, which are 3/4 natural 
size. 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XXXVI 




Artiodactyls from the Uinta. 
(For explanation see opposite page.) 



152 Annals of the Carnegie Museum. 



Plate XXXIX. 

Fig. I. Achcrnodon robust us, type. Skull and jaws, right side. 
Fig. 2. Achccnodon robustus, type. Crowns of upper teeth. 
Fig. 3. Achanodon robustus, type. Crowns of lower teeth. 
Fig. 4. Achccnodon uinlense. Crown of M-. C. M. No. 3182. 
All figures 3/8 of nature. 



o 




I 



154 



Annals of the Carnegie Museum. 



Plate XL. 



Fig. I. 
Fig. 2. 
Fig. 3. 
Fig. 4. 
Fig. s. 
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. 
Fig. 19. 

Fig. 20. 

Fig. 21. 
Fig. 22. 

Fig. 23. 

Fig. 24. 

Fig. 25. 

Fig. 26. 

Fig. 27. 



Proloreodon medius, type. 
Protoreodon medius, type. 
Proloreodon medius, type. 
Protoreodon medius, type. 
Protoreodon medius, type. 
Protoreodon medius, type. 
Protoreodon tnedius, type. 
Protoreodon medius, type. 
Protoreodon medius, type. 
Protoreodon medius, type. 
Protoreodon medius, type. 
Protoreodon medius, type. 
Protoreodon medius, type. 
Protoreodon m.edius, type. 
Protoreodon medius, type. 
Protoreodon medius, type. 
Protoreodon minor. 
Protoreodon minor 



Skull and lower jaws from the side. 

Atlas, from the side. 

Atlas, from above. 

Atlas, from below. 

Femur, posterior face. 

Tibia, tibial face. 

Calcaneum, tibial face. 

Astragalus, dorsal face. 

Humerus, radial face. 

Radius and ulna, anterior face. 

Radius and ulna, radial face. 

Pelvis, from the left side. 

Caudals from the right side. 

Carpus, dorsal face. 

Pes, dorsal face. 

Metacarpals, dorsal face. 
Skull, jaws, and neck. C. M. No. 3032. 
Fore limbs, from the right side. CM. No. 3032. 



Protagriochoerus annectens. 

C. M. No. 30. 
Protagriochoerus annectens. 

C. M. No. 30. 
Protagriochceus annectens. 
Protagriochoerus annectens. 

C. M. No. 30. 
Protagriochoerus annectens. 

C. M. No. 30. 
Protagriochoerus annectens. 

C. M. No. 30. 
Protagriochoerus annectens. 

C. M. No. 30. 
Protagriochoerus annectens. 

C. M. No. 3016. 
Protagriochoerus annectens. 



Proximal phalanx, dorsal face. 

Median phalanx, dorsal face. 

Aletatarsal, dorsal face. C. M. No. 30. 
Humerus, anterior face of trochlea. 

Femur, anterior face of trochlea. 

Ungual phalanx, dorsal face. 

Ungual phalanx, from the side. 

Astragalus, from the front. 

Cuboid, from the front. C. M. No. 3016. 



All figures 1/3 natural size, except Figs. 19-27, which are one-half natural size. 



ANNALS CARNEGIE MUSEUM, Vol. XII 



Plate XL. 




Protoreodon from the Uinta. 
(For explanation see opposite page.) 



156 Annals of the Carnegie Museum, 



Plate XL I. 
Pfotoreodon medium, type, restoration of skeleton, one-sixth natural size. 



158 



Annals of the Carnegie Museum. 



Fig. I. Helaleles hoops, type. 

Fig. 2. Helaletes boops, type. 

Fig. 3. Helaleles boops, type. 

Fig. 4. Helaleles boops, type. 

Fig. 5. Helaletes boops, type. 

Fig. 6. Helaletes boops, type. 

Fig. 7. Helaletes boops, type. 

Fig. 8. Helaletes boops, type. 

Fig. 9. Helaletes boops, type. 

Fig. 10. Epihippus gracilis. 

Fig. II. Epihippus gracilis. 

Fig. 12. Epihippus parvus. 

Fig. 13. Epihippus parvus. 

Fig. 14. Epihippus parvus. 

Fig. 15. Epihippus parvus. 

Fig. 16. Epihippus parvus. 



Plate XLII. 

Tibia and fibula, fibular face. 

Pes, darsal face. 

Femur, anterior face. 

Radius and ulna, radial face. 

Calcaneum, anterior face. 

Metacarpals and magnum, anterior face. 

Astragalus, plantar face. 

Astragalus, dorsal face. 

Pelvis, external face. 
Pes, dorsal face. C. M. No. 2923. 
Tibia anterior face. C. M. No. 2923. 
Hind limb, tibial face. C. M. No. 3397. 
Pes, dorsal face. C. M. No. 3397. 
Tibia, anterior face. C. M. No. 3397. 
Crowns of upper dentition. C. M. No. 3397. 
Crowns of lower dentition. C. M. No. 3397. 



All figures one-half natural size. 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XLII. 




Perissodactyls from the Uinta. 
(For explanation see opposite page.) 



160 



Annals of the Carnegie Museum. 



Plate XL 1 1 1. 



Helaleles boops, type. Skull and lower jaws. 



Fig. I. 

Fig. 2. Helaletes boops, type. 

Fig. 3. Helaletes boops, type. 

Fig. 4. Epihippus gracilis. 

Fig. 5. Epihippus gracilis. 

Fig. 6. Epihippus gracilis. 

Fig. 7. Epihippus gracilis. 

Fig. 8. Epihippus gracilis. 



Palate and crowns of upper teeth. 
Crowns of lower teeth. 
Crowns of upper teeth. C. M. No. 3398. 
Crowns of lower teeth. C. M. No. 3398. 
Scapula, distal end. C. M. No. 3398. 
Radius, anterior face. C. M. No. 3398. 
Lunar, dorsal face. C. IVL No. 3398. 



All figures natural size. 



162 Annals of the Carnegie Museum. 



Plate XLIV. 

Fig. I. Iseclolophus annedens. Crowns of upper dentition. C. M. No. 3030. 
Fig. 2. Parisectolophus latidens, type. Alveolar border of lower jaw and crowns 

of inferior dentition. 
Fig. 3. Parisectolophus latidens, type. Crown view of upper dentition. 
Fig. 4. Schizolophodon cuspidens, type. Lower jaws. 
Fig. 5. Dilophodon minusculus, type. Alveolar border of lower jaw and crown 

view of dentition. 
Fig. 6. Desmalolherium gnyolii, type. Palate and crowns of upper dentition. 
All figures natural size except Figs. 2 and 3 which are X 8/9 of nature. 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XLIV. 




Perissodactyls from the Uinta. 
(For explanation see opposite page.) 



164 Annals of the Carnegie Museum. 



Plate XLV. 
Pyothyracodon idntense, type. Palatal view, natural size. 



166 



Annals of the Carnegie Museum. 



Plate XLVI. 



Fig. 


2. 


Fig. 


3- 


Fig. 


4- 


Fig. 


5- 


Fig. 


6. 


Fig. 


7- 


Fig. 


8. 


Fig. 


9- 



Fig. I. Prothyracodon obliquidens. 

C. M. No. 3201. 
Prothyracodon obliquidens. 
Prothyracodon obliquidens. 
Prothyracodon obliquidens. 
Prothyracodon obliquidens. 
Prothyracodon obliquidens. 

C. M. No. 3199. 
Prothyracodon obliquidens. 
Prothyracodon obliquidens. 
Prothyracodon obliquidens. 

C. M. No. 2942. 
Prothyracodon uintense, 

C. M. No. 2990. 
Prothyracodon uintense. 
Prothyracodon uintense. 
Prothyracodon uintense. 
Prothyracodon uintense. 
Prothyracodon uintense. 
Prothyracodon uintense 

C. M. No. 2990. 
All figures one-half natural size except Fig 



Skull and lower jaws, from the side. 



Fig. 10 

Fig. II. 
Fig. 12. 
Fig. 13. 
Fig. 14. 
Fig. 15. 
Fig. 16. 



Manus, dorsal face. C. M. No. 2942. 
Pisiform, dorsal face. C. M. No. 2942. 
Humerus, anterior face. C. M. No. 2942. 
Scapula, proximal end. C. M. No. 2942. 
Metatarsals III & IV, dorsal face. 

Pes, tibial face. C. Al. No. 3199. 

Pes, dorsal face. C. M. No. 3199. 

Radius and ulna, anterior face. 

paratype. Radius and ulna, ulnar face. 



Tibia, anterior face. 
Mc. IV,, dorsal face. 
Mt. Ill, dorsal face. 
Mt. IV, dorsal face. 



C. M. No. 2990. 
C. M. No. 2990. 
C. M. No. 2990. 
C. M. No. 2990. 



Median phalanx, dorsal face. C. M. No. 2990. 
Proximal phalanx, dorsal face. 

I, which is one-third of nature. 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate XLVi 




Prolhyracodon from the Uinta. 
(For explanation see opposite page.) 



168 



Annals of the Carnegie Museum. 



Plate XLVII. 



Fig. I. Achanodon uintense. Skull, palatal view. 
Fig. 2. Achanodon uintense. Skull from the side. 
Fig. 3. AchcBnodon uintense. Skull from the back. 
Fig. 4. AchcBnodon uintense. Lower jaws, outer fa.ce. 
Fig. 5. Amynodon intermedium. Skull, palatal view. C. M. No. 3200. 
Fig. 6. Amynodon intermedium, type. Crowns of upper dentition. 
Fig. 7. Amynodon intermedium,. Skull, from the side. Same as Fig. 5. 
Fig. 8. AchcBnodon robustus, type. Skull, top view. 

All figures one-sixth natural size. Fig. 6 is redrawn from Professor Osborn's 
paper. Trans. Amer. Philos. Soc, XVI, 1889, Part II, PI. X, Fig. 10. 



ANNALS CARNEGIE MUSEUM, Vol. XM. 



Plate XLVI 




Artiodactyls from the Uinta. 
(For explanation see opposite page.) 



III. CONTRIBUTIONS TO THE STUDY OF THE FRESH 
WATER FISHES OF THE ISLAND OF FORMOSA. 

By Masamitsu Oshima. 
Of the Institute of Science, Government of Formosa. 

(Plates XLVIII-LIII.) 

Introduction. 

The Island of Formosa, lies in the western Pacific Ocean, between 
the Southern and Eastern China Seas and is separated from the Chinese 
mainland by the Formosan Strait, which has a width of about ninety 
miles in its narrowest part. It is two hundred and five miles long and 
from sixty to eighty miles broad, having seven hundred and thirty-one 
miles of coast-line and an area of 13,429 square miles, being thus nearly 
of the same size as Kiushiu, the southernmost island of Japan proper. 

The island is traversed throughout its length by a fine mountain 
range, which reaches an altitude of from 8,000 to 12,000 feet, Mt. 
Niitaka (14,270 ft.) stands in the middle and Mt. Silvia (12,480 ft.) 
in the north. In addition, along the eastern shore there are coast 
ranges of considerable height, the bases of which form magnificent 
cliffs from fifteen to twenty-five hundred feet high. Thus the middle 
p(i.rt and the eastern side of the island are mountainous, while the west 
is covered by fertile plains. 

Because of the above mentioned topography, there are very few 
rivers on the eastern coast, while on the western coast there are many 
of considerable length, namely, Tamusui, Hozan, Koro, Daian, Daito, 
Dakusui, Seira, Shinkobi, Sobun, and Shimo-Tamusui, of which the 
last-named is the largest. In addition to these, there is a lake called 
Jitsugetsutan (Lake Candidius), which has a small outlet into the 
River Dakusui. 

Until a comparatively recent date the Island of Formosa was a 
"terra incognita" to the naturalist, and the fresh-water fishes, which 
are especially valuable as confirming the geographical relationship 
between isolated islands and continents, were quite unknown. 

When Albert Giinther published his " Catalogue of Fishes " in 1859- 
1870, he had only sixteen species of Formosan fresh-water fishes to 
12 — DEC. 16, 1919. 169 



170 



Annals of the CARNf:GiE Museum. 



enumerate. x\fter a lapse of more than twenty years Jordan and 
Evermann reported one hundred and eighty-six species of Formosan 
fishes, including twenty-seven which were found in the fresh waters. 
Since that publication the efforts of Regan, Jordan and Richardson, 
and Boulenger have raised the total number of Formosan fresh- water 
fishes from thirty-two to forty-three. 

In the present paper is given a record of an extensive collection of the 
fresh-water fishes of the Island of Formosa, chiefly made by Mr. 
Takeo Aoki, my assistant, during the years 1915-1917, making an 
addition of seven new genera, fifteen new, and eighteen unrecorded 
species. 

New Genera. 



1. Formosania. 

2. Spinibarbus. 

3. Scaphesthes. 



4. Acrossocheilus. 

5. Phoxiscus. 

6. Aristichthvs. 



7. C id trie III US. 



New Species. 



Pseudobagrus taiwanensis. 
Pseudobagrus adiposalis. 
Liohagnis nantoensis. 
Formosania gilberti. 
Labeo jordani (introduced). 
Piintiiis snyderi. 
Spinibarbus hollandi. 



8. Scaphesthes tamusuiensis. 

9. Gnathopogon iijimcB. 

10. Phoxiscus kikuchii. 

1 1. Ciilter aokii. 

12. Macro podiis filamentosus. 

13. Rhino gobius taiwanus. 

14. Rhinogobius formosanus. 



15. Glossogobius parvus. 



Species Not Hitherto Recorded. 



1. Parasalanx ariakensis. 

2. Cobitis tcenia. 

3. Capoeta semifasciolata. 

4. Pseudorasbora parva. 

5. Distcechodon tumirostris. 

6. Rhodeiis oceUatus. 

7. Zacco temminckii. 

8. Hypophthalmichthys molitrix 

(introduced). 

9. Aristichthys nobilis (introd.). 



10 Ciiltriculus kneri. 

1 1 . Oryzias latipes. 

12. Gambusia affinis (introd.). 

13. Mugil cephalus. 

14. Mugil carinatiis. 

15. PolyacaiitJius opcrciilatus. 

16. Sicyopterus japomcus. 

17. B litis biitis. 

18. Glossogobius granniieponnis. 



The Fresh Water Fishes of the Island of Formosa. 171 

It is very difficult to draw a hard and fast line between true fresh- 
water fishes and brackish-water fishes. Therefore in the following 
pages the descriptions of all the fishes which were found in the fresh 
waters are given. For the sake of completeness I have added the 
descriptions of five known species, namely: Purasalanx aciiticeps, 
Liobagrus fornwsaiuts, Ischikaiiia macrolepis, Anguilla sinensis, and 
Glossogohius ahacopiis, of which I have not seen specimens. 

The typical portions of the collections, including the type specimens, 
and the others which are described in the present paper, are pre- 
served in the Carnegie Museum in Pittsburgh. A second set, in- 
cluding cotypes of the new species, is deposited in the Museum of 
Leland Stanford Junior University. The remainder is reserved for 
the Institute of Science, Government of Formosa. 

It has been my good fortune during a visit to Leland Stanford Junior 
University to be able to examine types and other specimens, and to 
use the reference books and literature quite freely. For this privilege 
I wish to express my sincere gratitude to the authorities of the Uni- 
versity. In the preparation of the present paper I have received the 
kind and valuable assistance of Dr. David Starr Jordan, Dr. Charles 
Henry Gilbert, and Prof. John Otterbein Snyder, for whose courtesy 
I express my hearty thanks. 

Family SALMONID^E. 
Genus Plecoglossus Temminck & Schlegel. 

1846. Plecoglossus Temminck & Schlegel, Fauna Japonica, Poiss., p. 229. 
(Type Plecoglossus allivelis Temminck & Schlegel.) 

Body moderately elongate, covered with very small scales. Mouth 
wide, the premaxillaries with a few, small, conical, pointed teeth. 
Maxillaries and lower jaw with teeth of peculiar form, lamelliform, 
broad, truncate, serrate, movable, seated in folds of the skin; mandibles 
each ending in a small knob, not jointed at the symphysis. Mucous 
membrane of interior of mouth between terminal halves of the man- 
dible forming a peculiar organ, raised in folds with two pouches in 
front and one behind. Tongue very small, with minute teeth, its tip 
toothless; no teeth on vomer, palatines with teeth. Pyloric caeca 
very numerous. Eggs small. Small fishes inhabiting the clear 
streams of Japan and Formosa, migratory like the salmon, and among 
the very finest of food-fishes. One species is known (Jordan & Snyder). 

Distribution: Japan proper; Corea; Formosa. 



172 Annals of the Carnegie Museum. 

I. Plecoglossus altivelis (Temminck & Schlegel). 
1846. Plecoglossus altivelis Temminck & Schlegel, Fauna Japonica Poiss., 
p. 229, PI. CV, Fig. i; no locality. — Gunther, Cat. Fish., VI, 1866, p. 165; 
Japan; Formosa. — Ishikawa, Zool. Mag. Tokyo, VII, 1895, p. 129; Japan. — 
Jordan & Snyder, Proc. U. S. Nat. Mus., XXIII, 1901, p. 349; Lake Biwa. — 
Proc. U. S. Nat. Mus., XXIII, 1901, p. 744; Numata, Tsushima.— Proc. 
U. S. Nat. Mus., XXIV, 1902, p. 584; Ishikari River, Niigata; Aomori; 
Same; Matsushima; Sendai; Morioka; Tokyo; Tama River; Daij^a River; 
Gifu; Lake Biwa; Osaka; Wakanoura; Kobe; Hiroshima; Kurume; Nagasaki; 
Tamusui River, Formosa. — Jordan, Ann. Zool. Jap., IV, 1902, p. 75, — 
Jordan & Evermann, Proc. U. S. Nat. Mus., XXV, 1903, p. 323; Formosa. — 
Jordan & Richardson, Mem. Carneg. Mus., IV, no. 4, 1909, p. 167; For- 
mosa. — Smith & Pope, Proc. U. S. Nat. Mus., XXXI, 1905, p. 463; Gifu. — 
Snyder, Proc. U. S. Nat. Mus., XLII, 1912, p. 403; Tsuruga'.— Jordan; 
Snyder, and Tanaka, Journ. Coll. Sci., Tokyo, XXXIII, 1013, p. 44, 
Japan; Hokkaido; Corea; Formosa. — ^Jordan & Metz, Mem. Carneg. 
Mus., ^T, 1913, p. 10, Fusan, Corea. 

Ayu (Japan); Kyarihii (Formosa). 

Head 4.88 in length; depth 4.27; D. 10, A. 17; P. 14; V. 8; width of 
head 2 in its length; eye 5.33 in head; interorbital space 2.76; snout 
2.76; maxillary 2; scales in the lateral line about 150, in an oblique 
series between origin of dorsal and lateral line 20, between the latter 
and the middle of belly 20, between lateral line and the root of ventral 
13; pectoral 1.22 in head; ventral 1.22; gill-rakers 16 + 19, 

Body oblong, a little compressed, curvature of the dorsal profile 
stronger than the ventral; head rather small, triangular, its dorsal 
surface slightly convex; snout rather long, acutely pointed; inter- 
orbital space slightly convex; mouth large, oblique, its angle reaching 
a vertical through the posterior border of orbit; lips thick; premaxil- 
lary well-developed; lower jaw shorter than the upper; palatines with 
minute teeth; si.x premaxillary teeth, minute and sharply pointed.; 
maxillary teeth modified into serrated plates, twelve on both jaws; 
eyes moderate, superior and anterior; nostrils close together; gill- 
rakers short and slender. 

Origin of the dorsal nearer tip of snout than base of caudal, opposite 
the ventral, rather high, each ray rather stiff; adipose dorsal very 
small, inserted above posterior third of the base of anal; pectoral as 
long as the ventral, not reaching the latter; ventral inserted below 
origin of dorsal, not reaching the vent; anal fin elongate, rather low, 
outer margin concave; caudal peduncle slender, its depth 2.18 in the 
length of head. 



Thk Fresh Water Fishes ov the Island of Formosa. 173 

Body covered with small cycloid scales; head naked; lateral line 
complete, extending along the middle of the sides, a little decurved. 

Color light l)luish green above, paler below; belly yellowish; dorsals, 
pectorals, adipose dorsal, and caudal fin dusky; ventrals and anal 
whitish. 

Total length 273 mm. 

Described from a specimen from Tamusui River near Shinten, 
collected by T. Aoki in December, 1915 (Female). 

Habitat: Tamusui River; Choso River; Taiko, Giran. 

Remarks: The anal fin of the male is higher and shorter than that 
of the female. The present species is distributed only in the rivers 
of the northern part of the island, mainly in the Tamusui River. 

Measurements of Plecoglossus allivelis. 



Locality 




.a 

ft 


a 


<■ 


fc 


>' 


•si 




u 

V S 

a'S 


3 

s 


>> 


Scales. 


as 


Tamusui River 


.S-37 


4-2.S 


10 


17 


13 


8 


1.88 


2.94 


2.76 


5 


21-150-20 


283 


Tamusui River 


4.88 


4.27 


10 


17 


14 


8 


2 


2.76 


2.76 


5-33 


20-150-20 


273 


Choso River 


4-30 


4-52J II 


17 


14 


8 


2 


2.90 


2.75 


5 


21-155-20 


172 


Taiko 


4.44 


4.441 II 


17 


IS 


8 1 2.15I 2.89I 2.80 


5 


20-145-21 


140 



Family SALANGID.E. 
Genus Parasalanx Regan. 

1908 Parasalanx Regan, Ann. Mag. Nat. Hist. (8), II, p. 444. (Type Para- 
salanx gracillitmis Regan.) 

Body slender, translucent, elongate, cylindrical, flattened anteriorly, 
compressed posteriorly, naked. Head elongate, much depressed, 
with long, flat, pointed snout; premaxillaries forming an anterior 
triangular expansion; lower jaw not projecting, ending in a distinct, 
more or or less movable presymphysial bone, with double series of 
teeth; tongue toothless. Dorsal fin partly above the anal; adipose 
fin present, small and low; caudal fin forked. 

Distribution: China; Formosa; Japan. 



Key to the Formosan Species. 

a. Depth of body ii times in length aciiliceps. 

aa. Depth of body more than 15 times in length ariakensis. 



174 Annals of the Carnegie Museum. 

2. Parasalanx acuticeps (Regan). 

1908. Salanx acuticeps Regan, Ann. Mag. Nat. Hist. (8), II, p. 360; Lake Can- 

didius, Formosa. 
1908. Parasalanx actiticeps Regan, Ann. Mag. Nat. Hist. (8), II, p. 446; Lake 

Candidius, Formosa. — Jordan & Richardson, Mem. Carneg. Mus., Vol. 

IV, no. 4, 1909, p. 167; Lake Candidius (after Regan). 

Depth of body 11 in length, length of head 5.5 to 5.66. Head 
three times as long as broad; snout acutely pointed, shorter than post- 
orbital part of head; diameter of eye eight times in the length of head. 
Lower jaw not projecting, with a toothed predentary bone and with 
anterior canines which perforate the roof of the mouth; tongue tooth- 
less. Dorsals 13-14. Anal 26-27, originating below the second ray 
of dorsal. Pectoral with nine or ten rays; origin of pelvic nearer to 
anal than to base of pectoral. 

Two specimens, 115 mm. in total length. (Regan). 

Habitat: Lake Candidius (Regan). 

Remarks: Not seen. 

3. Parasalanx ariakensis (Kishinouye). 

1902. Salanx ariakensis (Kishinouye) Jordan & Sn^'der, Proc. U. S. Nat. Mus., 
XXIV, p. 592; Ariake Sea, Japan. — -Jordan, Snyder & Tan.\ka, Journ. 
Coll. Sci., XXXIII, 1913, p. 47; Ariake Sea. 

1908. Hemisalanx ariakensis Regan, Ann. Mag. Nat. Hist. (8), II, p. 445; 
Ariake Sea (after Jordan & Snyder). 

Head 5.85 in length; depth at insertion of the anal 15 in length; 
depth of caudal peduncle 4.5 in head; eye 6; intcrorbital space 3; snout 
2.166 D. 13; A. 28; P. 9; V. 7. 

Body elongate, cylindrical, strongly flattened anteriorly, posterior 
part compressed, highest in front of the insertion of anal; head flat, 
much broader than body, height 2.5 in its width; snout spatulate, pre- 
maxillaries forming an anterior triangular expansion, tip sharply 
pointed; jaws subequal; teeth on both jaws and vomers in a single row, 
those on the upper jaw stronger and set apart, recurved and canine- 
like; vomerine teeth minute; three pairs of sharp, recurved canine-like 
teeth near the tip of lower jaw, which are received into a pouch-like 
concavity of the upper jaw behind premaxillaries; tongue toothless; 
eyes lateral, ])rominent. 

Dorsal fin inserted on posterior four-fifteenths of body, partly 
iibovc the anal; adipose dorsal low, originating at the end of base of 
anal; pectoral subhorizontal; origin of ventral midway between tip of 



The Fresh Water Fishes of the Island of Formosa. 175 



snout and base of caudal, anterior ray longest; caudal fin deeply 
forked, the tip of each lobe shar])K- i>()in(ed. 

Head and body naked. 

Color white, except eyes; two longitudinal series of small black spots 
along the ventral median line. 

Total length 80 mm. 

Described from a specimen fron: Tamusui River near Taihoku, 
collected by Oshima in February, 1917. 

Habitat: Tamusui River. 

Remarks: All the characters of the present species agree quite well 
with those of the cotype in the Stanford University collections. 

In the year 1908, Regan noted in his paper (Ann. Mag. Nat. Hist. 

(8), II, p. 445) that Salanx ariakensis (Kishinouye) described by 

Jordan & Snyder might belong to the genus Ilemisalaivx . However, 

the lower jaw of the present species does not project and the premaxil- 

laries form an anterior triangular expansion. Such being the case, 

it seems reasonable to include Salanx ariakensis in the genus Parasal- 

anx. 

Measurements of Parasalanx ariakensis. 



Locality. 


-0 


0. 

V 

Q 


a 


<:' 


p^ 


> 


-0 

"^ 


3 


c 




S3 


jl 


Tamusui River 

Tamusui River 


5-85 

5-83 

5-92 

6 

6.16 

5-33 
5-42 


IS 

15-75 

15.60 

15 
17-75 

16.57 
17-33 


13 
13 
13 
13 
13 

13 
13 


28 
26 
28 
27 
28 

28 
29 


9 
9 

8 
9 
9 

9 
9 


7 
7 
7 
7 
7 

7 

7 


2.60 

2.75 
2.60 
2.60 
2.75 

3 
2.71 


2.166 
2.40 

2-33 
2.166 

2.20 

2.33 
2.375 


6 

6 

6 

6.5 

55 

6.33 

7 


3 
3-66 

3-5 
3-5 
3-33 

3-66 
3-5 


80 
70 


Tamusui River 


77 


Tamusui River 

Tamusui Rivpr 

Ariake Sea 
(Cotype; No. 8574); S. U. . 
(Cotype; No. 8574); S. U. . 


78 
74 

125 
112 



Family SILURID.^. 
Artificial Key to the Formosan Genera. 
I. Dorsal fin spineless; anal very long. 

a. Dorsal fin many-rayed, very long, uniformly composed of feeble rays; 

four pairs of barbels Clarias. 

aa. Dorsal fin very short, rudimentary; two pairs of barbels. .Parasilurus. 
J I. Dorsal fin with a pungent spine; anal moderate. 

a. Adipose dorsal fin not adnate, free behind; mental barbels as usual, 
median pair not notably distant; dorsal spine smooth or denticulated; 

pectoral spine denticulated behind Pseudohagrus. 

aa. Adipose dorsal fin adnate to the back and connected with the caudal; 
median mental barbels far apart; dorsal and pectoral spines smooth, 
sharp, and imbedded in the skin Liobagrus. 



176 Annals of the Carnegie Museum. 

Genus Parasilurus Bleeker. 

1856. Glanis Agassiz, Proc. Amer. Acad., p. 333 (Name pre-occupied by Giants 

Gronow, 1854). (Type Glanis aristotelis Agassiz.) 
1863. Parasilurus Bleeker, Nederl. Tydschr. Dierk., p. 114. (Type Silurus 

asotus Linnaeus.) 

Bod}^ elongate, the profile of the back almost horizontal. Head 
depressed and covered with soft skin; eyes anterior and subcutaneous; 
mouth broad, transverse; four barbels, two maxillary, which are very 
long, two mental, which are short; teeth cardiform or villiform, in broad 
bands in the jaw and on vomer; no teeth on palatines. Gill-opening 
wide, not confluent with isthmus, and narrowly jointed together. 
Dorsal small, without spine, and anterior; adipose fin absent; anal 
more or less united with the caudal, very long; pectorals with spine; 
ventral behind dorsal. Air-bladder not inclosed in bone. (Jordan & 
Fowler.) 

Distribution: India; East Indies; Formosa; China; Corea; Amur 
Province; Japan. 

4. Parasilurus asotus (Linnaeus). 
Namadzu (Japan); Ryamhii (Formosa). 

1758. Silurus asotus Linn^us, Sj^st. Nat. Ed. X, p. 501; Asia.^BLOCH & Schnei- 
der, Syst. Ichth., 1801, p. 375. — Basilewsky, Nouv. Mem. Soc. Nat. Mosc, 
X, 1855, p. 240, PI. 3, Fig. 4; China. — Gunther, Cat. Fish., V, 1864, p. 33; 
Japan; China. — Ann. Mus. St. Petersb., 1896, p. 11; Huihsien, China. — 
ISHiKAWA, Prel. Cat., 1897, p. 23; Japan. — GDnther, Ann. Mag. Nat. 
Hist. (7), I, 1898, p. 261; Newchang. — Popta, Z06I. Anz., XXXII, 1907, 
p. 250; Kiautschau, China. 

1846. Silurus xanthosteus Richardson, Ichthyol. China, p. 281; Canton; Chusan. 
— Voy. Sulph., Fishes, p, 133, PI. 56, Fig. 12-14. 

1846. Silurus japonicus Schlegel, Fauna Japonica, Poiss., p. 226, PI. CIV, Fig. 
i; Higo; Satsuma; Nagasaki. — Bleeker, Verh. Batav. Genootsch., XXV, 
1855. PP- 30 and 51. 

1846. Silurus sinensis Richardson, Ichthyol. China, p. 281; Chusan. 

1867. Silurus (Parasilurus) asotus Kner, Novara, Fisch, III, p. 303; Shanghai. 

1901. Parasilurus asotus Jordan & Snyder, Ann. Zool. Jap., Ill, p. 45; Yoko- 
hama. — Abbott, Proc. U. S. Nat. Mus., XXIII, 1901, p. 83; Pei-ho, China. — 
Jordan & Fowtler, Proc. U. S. Nat. Mus., XXVI, 1903, p. 903; Tokj-o; 
Niigata; Morioka; Tama River; Kawatana; Sendai; Ichinoseki; Chikugo 
River; Tsuchiura; Lake Biwa; Formosa. — Jordan & Richardson, Mem. 
Carneg. Mus., IV, no. 4, 1909, p. 163; Formosa. — Berg, Ichthyol. Amur., 
1909, p. 175; Amur Province. — Snyder, Proc. U. S. Nat. Mus., XLII, 
1912, p. 403; Tokyo; Takamatsu River. — Jordan & Metz, Mem. Carneg. 
Mus., VI, No. 2, 1913, p. 12; Corea. 

1903. Glanis asotus Jordan & Evermann, Proc. U. S. Nat. Mus., XX\', p. 320; 
Tamusui River, Formosa. 



The Fresh Water Fishes of the Island of Formosa. 177 

1913. Parasihinis asoliis Jordan, Sn^'der, & Tanaka, Journ. Coll. Sci. Tokyo, 
XXXIII, p. 58; Japan. 

Head 4.83 in length; depth 6.20; U. 5; A. 75; P. 1.12; V. 11 ; width 
of head 1.33 in its length; eye 8 in head, 3 in snout, 4 in interorbital 
space; interorbital space 2 in head; pectoral 1.66; ventral 2.14. 

Body elongate, posterior part compressed, anterior part more or 
less depressed; tail long and tapering; head flat and broad; snout 
depressed, broadly rounded anteriorly; lips thin and smooth, lower 
jaw somewhat protruding; eyes oval, supra-lateral, and anterior; 
nostrils far apart, distance between anterior and the posterior nostril 
equal to the distance between the latter and eye, anterior nostril in a 
short tube; mouth very broad and superior; teeth sharp, in broad 
villiform bands in the jaws, vomers, and palatines; barbels four, two 
maxillary, nearly as long as head, two mental, very short, about 4 in 
head; interorbital space very broad, its middle part somewhat de- 
pressed; gill-openings large; gill-rakers i + 10, rather short, and 
widely set; gill-membranes entirely separated. 

Dorsal fin shorter than ventral, distance to pectoral twice as long 
as the distance to ventral; anal fin very long, united with caudal be- 
hind, of uniform height, its origin much in advance of the middle of the 
length; pectoral armed with a short strong spine, both edges of which 



Measurements of Parasilurus asoius. 



Locality. 



Jitsugetsutan. . 
Tamusui River. 

Inzanpo 

Ritakukan .... 



g 


J3 

Q 


Q 




p^ 


> 






3 


a 






u 
C 
> 




4-83 


6.20 


5 


75 


I, 12 


II 


1.33 


8 


2.64 


2 


1.66 


2.14 


330 


5 


6-5 


5 


78 


I, II 


II 


1.4 


8.5 


3 


2-5 


1.83 


2-5 


242 


4-57 


6.4 


5 


75 


1, 12 


II 


1-33 


7-33 


3.6 


2 


2 


2.2 


180 


4-33 


6.23 


5 


73 


I, 12 


II 


1-73 


7 


3 


2.4 


1.72 


2.28 


150 



bear denticulations, inner ones stronger; ventral fins shorter than the 
pectoral, reaching behind the origin of anal; caudal fin very slightly 
emarginate, each lobe obtusely rounded; anal papilla present. 

Body smooth, naked; lateral line distinct, extending along the 
middle of sides, continuous. 

Color in formalin uniformly dark gray, lower surface of head and 
abdomen whitish. 

Total length 330 mm. 

Described from a specimen from Jitsugetsutan (Lake Candidius), 
collected by T. Aoki in September, 1916. 
13 — DEC. 17, 1919. 



178 Annals of the Carnegie Museum. 

Habitat: Distributed throughout the island. My specimens are 
from Jitsugetsutan (Lake Candidius); Tamusui River; Inzanpo; 
Ritakukan. 

Genus Pseudobagrus Bleeker. 

i860. Pseudobagrus Bleeker, Act. Soc. Indo-Nederl., VII, p. 87. (Type Bagrus 
aurantiacus Tcmminck & Schlegel). 

Body moderately elongate. Head broad and compressed, covered 

above by moderately thick, smooth skin; eyes moderate or rather 

small; snout broad, obtuse; mouth broad, tranverse, and with bands of 

villiform teeth in the jaws; a continuous transverse band of teeth on 

the roof of the mouth; nostrils remote, the anterior usually in a small 

tube; eight barbels, the maxillaries longest, and the mentals more or 

less evenly distributed. Dorsal fin short, with five to seven rays, and 

like the pectoral with a stout si)ine; caudal rounded or subtruncate; 

anal with twenty or more radii; ventrals broad, with six rays. (Jordan 

& Fowler.) 

Synopsis of the Formosan Species. 

A. Depth of body 6.4,5-7.40 in length; origin of the dorsal in advance of the tip of 

the pectoral. 
a. Maxillary barbels long, reaching the tip of operculum; adipose dorsal 
shorter than the anal; anal fin 15-iayed; pectoral I, 7; caudal fin slightly 

emarginate laiivatius. 

aa. Maxillary barbels short, not reaching the tip of operculum; adipose 
dorsal longer than the anal; anal fin 18-19-rayed; pectoral I, 8-9; caudal 
fin rounded adiposalis. 

B. Depth of body 4.5-5.33 in length; origin of the dorsal above the tip of the pec- 

toral. 
a. Maxillary l^arbels long, reaching beyond the tip of operculum; adijsose 
dorsal longer than the anal; anal fin 15-rayed; pectoral I, 7; caudal fin 
slightly emarginate brevianalis. 

5. Pseudobagrus brevianalis Regan. 
Sankakuko (Formosa). 
1908. Pseudobagrus brevianalis Regan, Ann. Mag. Nat. Hist. (8), \'ol. I, p. 149; 
Lake Candidius, Formosa. 

Head 4 in length; depth 5.33; D. I, 7; A. 15; P. I, 7; V. 6; width of 
head 1.33 in its length; eye S.5 in head; snout 2.5; inter-orbital space 
2.33; pectoral 1.75; ventral 2.28. 

Body elongate, higher in front, posterior part compressed, cross- 
section of the anterior part triangular; head rather Hat, its top gradu- 
ally inclining anteriorly; snout broad, obtuseh' rounded anteriorly, 



The Fresh Water Fishes of the Island of Formosa. 179 

projecting bcjoiul the U)\vcr jaw; mouth inferior, transverse, crescent- 
shaped, with fleshy tliick lips, the upper more or less papillose; lower 
jaw much thinner, and distinct at the angle of the mouth only; both 
jaws with broad bands of villiform teeth; a subcrescentic narrow band 
of villiform teeth on the palate; eight barbels, four on the snout, the 
other four on the mentum; maxillaries long, one and one-half times in 
the length of head, scarcely reaching the base of the pectoral, nasal 
barbels as long as median mentals; eyes small, covered with thin skin, 
lateral and superior; nostrils separated, the anterior in a short tube, 
situated in a shallow groove just behind the upper lip, posterior nos- 
trils behind the root of nasal barbels; gill-openings large, extending to 
upper part of the base of the pectorals; gill-membranes not confluent 
with the skin of isthmus, their posterior margins entirely free; gill- 
rakers 3 + 8, slender. 

Origin of the dorsal above the tip of pectoral, its spine rather slender, 
nearly two-thirds as long as the dorsal; adipose dorsal rather short, 
above the anal, and inserted in front of anus, slightly longer than the 
anal; pectoral with a spine, its inner edge strongly serrated; ventral 
short, tip reaching the anus; anal fin rather short; anal papilla, not well 
developed, very short; caudal fin emarginate, each lobe rounded at 
the tip; caudal jieduncle elongate, slightly higher posteriorly, its 
deepest part twice in length of head. 

Body naked, smooth; lateral line continuous, extending along the 
middle of the sides. 

Color in formalin uniformly brownish gray, without any markings; 
belly and throat whitish. 

Total length 130 mm. 

The present description is from a specimen from Jitsugetsutan 
(Lake Candidius), collected by T. Aoki in August, 1916. 

Habitat: Jitsugetsutan (Lake Candidius); Dainansho, Nanto. 

Remarks: Regan's type specimen is provided with a l6-i8-rayed 
anal fin, instead of being 15-rayed as in the present specimen. 



Measurements 


OF Pseudobagrus brevianalis. 








Locality. 


Head. 
Depth. 


Q 


< 


eu 


>' 





9 


in 


£3 


1 



c 
u 
> 




Jitsugetsutan 

Dainansho 


4 1 5-33 
3-751 4-5 


1,7 
1.7 


15 
15 


1.7 
1-7 


6 
6 


1.32 
I-3I 


2.5 2.33 
2.66 2.43 


1-75 
1-7 


2.28 
2 


130 
78 



180 Annals of the Carnegie Museum. 

6. Pseudobagrus taiwanensis sp. nov. (Plate XLVIII, Fig. i). 

Head 3.875 in lengtli; depth 7.14; D. I, 7; A. 15; P. I, 7; V. 6; width 
of head 1.4 in its length; snout 2.75 in head; interorbital space 3; eye 
8; pectoral 1.55; ventral 2.1 1. 

Body elongate, depth rather uniform, tail compressed; head broad, 
depressed; snout flattened, bluntly rounded anteriorly; upper jaw 
projecting beyond the lower; mouth inferior, transverse, crescent- 
shaped; lips moderately thick, fleshy, lower lip distinct at the angle 
of the mouth only; jaws with broad bands of villiform teeth; palate 
with a transverse, crescent-shaped band of villiform teeth; 8 barbels, 
four on the snout, the other four on the mentum, maxillaries the longest, 
reaching tip of operculum; nasal barbel as long as median mental; 
eyes small, laterally superior, covered with thin skin; nostrils separated, 
the anterior tubular, in contact with posterior margin of the upper lip; 
posterior nostrils just behind the root of nasal barbel; interorbital 
space broad, middle part depressed longitudinally; gill-openings 
large, extending backwards to the base of pectoral; gill-membranes 
entirely separated; gill-rakers 3 + 10, proximal ones on the lower arm 
minute. 

Origin of the dorsal in advance of the tip of pectoral, with a sharp 
spine; adipose dorsal very short and low, originating behind the base 
of anal, much shorter than the latter, and ending in front of the tip of 
the anal; pectoral fin rather elongate, armed with a sharp spine, its 
inner edge strongly denticulated; ventral fin scarcely reaching the 
base of anal, overlapping anus and the well-developed anal papilla; 
anal fin rather short; caudal fin slightly emarginate, the tip of each 
lobe rounded; caudal peduncle elongate, deeper posteriorly, its depth 
about twice in the length of head. 

Body smooth; lateral line continuous, nearly straight, extending 
along the middle of the sides. 

Color in formalin uniformly dark brown, belly and lower part of 
head whitish. 

Total length 152 mm. 

Described from a specimen from Tozen River near Taiclui, collected 
by T. Aoki in December, 1916. 

Habitat: Tozen River; Taito River; Shinchiku. 

Reniarhs: The nearest relative of the present'species is Pseudobagrus 
brevianalis Regan. These two species differ as follows: 



ANNALS CARNEGIE MUSEUM. Vol. XII. 



Plate XLVIIi. 




Fig. I. Pscudobagyiis laiuancnsis Oshima, 8p. nov. 
Fig. 2. P seudobagrus adiposalis Oshima, sp. nov. 
Fig. 3. Diobagrus nanloensis Oshima, sp. nov. 



\'-- 



The Fresh Water Fishes of the Island of Formosa. 181 

Depth of body 6.42 to 7.14; maxillary barbels long, reaching the tip of oper- 
culum; adipose dorsal shorter than the anal; origin of the dorsal in advance of 

the tip of pectoral taiwanensis. 

Depth of body 4.5 to 5.33; maxillary barbel short, reaching only to the base of 
the pectoral; adipose dorsal longer than the anal; origin of the dorsal above the 
tip of pectoral hrevianalis. 

Measurements of Pseudobagrus taiwanensis. 



Locality. 


T3 

n! 


0. 



Q 


< 


P^' 


> 


^1 


3 

c 

C/2 


S3 


i 




Tozen River 


3-87 
3.58 
3-91 


7.14 
6.83 
6.42 


1.7 
1.7 
T 6 


15 
15 
15 


1.7 
1.7 
1.7 


6 
6 
6 


1.40 
1-33 
1-33 


2.75 
2.60 

2.75 


3 
3 
3 


8. 
8. 

7.5 


152 
52 
50 


Shinchiku 


Daito River 







7. Pseudobagrus adiposalis sp. nov. (Plate XLVIII, Fig. 2). 
Sankakufu (Formosa). 

Head 4.28 in length; depth 7.16; D. I, 7; A. 19; P. I, 8; V. 6; width 
of head 1.5 in its length; interorbital space three times in head; snout 
2.8; width of mouth 2.5; pectoral 1.33; ventral 1.86; eye eight times in 
head; three times in interorbital space. 

Body elongate, depth rather uniform, tail compressed; head 
broad, triangular, depressed and smooth; snout flattened, obtusely 
rounded anteriorly, projecting beyond the lower jaw; mouth inferior, 
transverse, crescent-shaped; lips thick, more or less papillose; jaws 
with broad bands of villiform teeth; eyes small, lateral and superior, 
covered with thin skin; nostrils separated, the anterior tubular, in a 
shallow pit behind the upper lip, the posterior in contact with the root 
of the nasal barbel; eight barbels, four on the snout, the other four on 
the mentum, the maxillary barbels the longest, reaching beyond the 
posterior margin of orbit, the median mental barbels the shortest; 
interorbital space broad, somewhat elevated; gill-openings large, 
reaching upward beyond the base of pectoral; gill-membranes deeply 
notched, entirely separated from each other; gill-rakers 3 + 10; 
slender. 

Dorsal fin inserted on anterior third of the distance between tip of 
snout and base of caudal, armed with a sharp spine; adipose dorsal 
very long, inserted behind the base of ventral, reaching beyond the 
posterior end of base of anal, its height gradually increasing posteriorly; 
the pectoral with a strong spine which is strongly serrated behind; 



182 



Annals of the Carnegie Museum. 



ventral fin broad, entirely behind the dorsal, reaching the anal 
papilla; anal fin well-developed, shorter than the adipose dorsal, 
inserted behind the origin of the latter, elongate and higher in front; 
caudal fin slightly diverging posteriorly, its tip very slightly emargi- 
nate; caudal peduncle much compressed, deeper posteriorly; depth 
twice in head. 

Body smooth; lateral line continuous, straight, extending along 
the middle of the sides. 

Color in formalin dark reddish gray, darker above, paler below; 
lower parts whitish. 

Total length 172 mm. 

Described from a specimen from Tamusui River near Shinten, 
collected by T. Aoki in December, 191 5. 

Habitat: Tamusui River; Taito River; Sobun River. 

Remarks: The present species is closely related to Pseudobagrus 
atirantiacus^ from Japan. The specific differences may be expressed 
as follows: 

a '. Anal fin 20-22-raycd; pectoral I, 7; gill-rakers 3 -h 7; the dorsal inserted 
above tip of pectoral; adipo e dorsal shorter than the anal; maxillary barbel 

long, reaching the tip of operculum aurantiacus. 

aa" . Anal fin 18-19-rayed; pectoral I, 8 or I, 9; gill-rakers 3 + 10; origin of the 
dorsal in advance of the tip of pectoral; adipose dorsal much longer than the 
anal; maxillary barbel short, not reaching the base of pectoral. . . .adiposalis. 

Measurements of Pseudobagrus adiposalis. 



Locality. 



Tamusui River . 
Tamusui River. 
Tamusui River . 
Tamusui River. 
Tamusui River. 

Heirinbi 

Sobun River. . . 





ji 










^1 


0— ■ 








0. 



Q 


< 


Om 


> 


It 


1.4 





c 
in 


>> 


4-5 


7 


I- 7 


18 


1.9 


6 


3 


3 


6.5 


4.28 


7.16 


I. 7 


19 


1. 8 


6 


^■S 


3 


2.8 


8 


4.16 


7 


I. 7 


18 


1.9 


6 


1.4 


3 


3 


7 


4-25 


6.6 


1. 7 


19 


1. 9 


6 


1-75 


3 


3 


7 


4-5 


7 


1. 7 


19 


1. 9 


6 


1-43 


3 


3 


6 


4.26 


6.74 


1.7 


19 


1, 8 


6 


1-43 


3 


2.58 


6 


4.42 


7 


1. 7 


18 


1.9 


6 


1-33 


3 


3 


9 






165 
172 

147 
190 
160 
141 
310 



Genus Liobagrus Hilgendorf. 

1878. Liobagrus Hilgendorf, Sitzungs. Ge.sellsch. Frcund. Berlin, p. i. (Type 
Liobagrus reinii Hilgendorf.) 

Body elongate, with compressed tail and rounded caudal. Head 
broad and depressed; top of head smooth, the humeral process smooth; 

1 Bagrus aurantiacus Temminck & Schlegel, Fauna Japonica, Poiss., 1846, p. 227, 
PI. CIV, Fig. 2; Japan. 

Pseudobagrus aurantiacus Bkn^kcr, Act. Soc. Sci. Indo-Nocrl., \TII, 1S60, p. 85. 



The Fresh Water Fishes of the Island of Formosa. 183 

eyes small and covered with thin skin, anterior in position; snout 
broad, obtuse, and projecting; teeth in jaws only, in broad villiform 
bands; eight barbels, the median mental barbels widely separated. 
Dorsal fin placed anteriorly; dorsal and pectoral spines smooth, 
sharp, and imbedded in skin; adipose fin long and low, joined to the 
caudal as in Notrius; ventral fins small, not reaching the anal, which 
has fifteen rays (Jordan & Fowler). 

Distribution: Formosa; China; Corea; Japan. 

Synopsis of the Formosan Species. 
a. Dorsal fin with seven soft rays; anal with twelve; upper jaw slightly longer 

than the lower nanto'ensis. 

aa. Dorsal fin with five soft rays; anal with fifteen; jaws equal anteriorly. 

formosaniis. 

8. Liobagrus nantoensis sp. nov. (Plate XLVIII, Fig. 3). 

Head 4.33 in length; depth 6; D. I, 6; A. 12; P. I, 7; V. 6; width of 
head once in its length; snout three times in head; interorbital space 
2.5; pectoral 1.25; ventral 1.8. 

Body compressed, dorsal profile abruptly inclined anteriorly at the 
base of dorsal fin; head large, round, flattened, with a median shallow 
groove, both sides of which are slightly swollen; interorbital space 
depressed; snout very short, anterior margin broadly rounded, pro- 
jecting beyond lower jaw; mouth anterior, transverse, with thick 
fleshy lips; upper jaw with a band of villiform teeth, similar bands on 
the lower jaw, crescent-shaped, narrower, but longer than that of 
upper jaw; no teeth on palatine and vomer; eight barbels, four on snout, 
others on mentum, all barbels thick at the base, nasal barbel nearly 
as long as inner mental, maxillary barbels slightly shorter than outer 
mentals, scarcely reaching the base of pectoral; nostrils superior, 
separated, anterior nostril in a short tube, the posterior in contact 
with root of nasal barbel; eyes very small, superior, imbedded in skin; 
gill-openings rather large; gill-membranes entirely separated. 

Dorsal fin small, its spine nearly half as high as the fin, hidden 
beneath skin; adipose dorsal low and long, opposite to the anal; pec- 
toral armed with a sharp smooth spine, which is hidden beneath the 
skin, tip of the fin reaching beyond the middle of dorsal; ventral fin 
entirely behind the dorsal, reaching beyond anus; the anal much 
shorter than adipose dorsal; caudal fin rather long, its tip rounded; 
depth of caudal peduncle uniform. Body smooth; lateral line indis- 
tinct. Color in formalin brownish gray, fins somewhat paler. 



184 



Annals of the Carnegie Museum. 



Total length 88 mm. 

Described from a specimen from Dainansho, Nanto, collected by 
T. Aoki in December, 1916. 

Habitat: Dainansho, Nanto. 

Remarks: The present species differs from Liohagrus formosanus 
Regan from Lake Candidius in having a 7-rayed dorsal, a 12-rayed 
anal, and a slightly protruded upper jaw. 





Measurements 


OF Liobagriis 


nantoaisis 










Locality. 


-6 


Depth. 
D, 


< 


Ph' 


> 


■5I 




215 


3 


a 


u 
>> 

W 


fl 


Dainansho 




4-33 


6 


1,7 


12 


1.7 


6 


I 


2.5 


3 


II 


88 


Dainansho 




4-50 


5.61 


1,7 


12 


1,7 


6 


1.23 


2.66 


3 


10 


48 



9. Liobagrus formosanus Regan. 

1908. Liobagrus formosanus Regan, Ann. Mag. Nat. Hist. (8), II, p. 360; Lake 
Candidius, Formosa. — Jordan & Richardson, Mem. Carneg. Mus., IV, 
no. 4, 1909, p. 168; Lake Candidius (after Regan). 

Depth of body six times in the length, length of head 4.33. Head a 
little longer than broad; interocular width nearly three times in the 
length of head. Jaws equal anteriorly; premaxillary band of teeth 
apparently two and one-half times as long as broad; posterior mandibu- 
lar barbel extending to basal part of pectoral. Dorsal I, 5; spine 
one-fourth the length of head. Pectoral spine half the length of the 
fin, which is eight- tenths the length of the head. Anal 15. Caudal 
rounded. Grayish; fins dusky; anal and caudal with a narrow pale 
edge. 

i\ single specimen, 37 mm. in total length; Lake Candidius, For- 
mosa (Regan). 

Habitat: Jitsugetsutan (Lake Candidius). Not seen. 

Remarks: According to Regan the present species is closely related 
to Liobagrus andersoni Regan^ from Corea. 

Genus Clarl4s Gronovius. 

1763. Clarias Gronovius, Zoophyl., p. 100 (non binomial). 

1777. Clarias Scopoli, Intr. Hist. Nat., p. 455. (Type Clarias oronlis Gunthcr.) 
1803. Macro pier onotus L.acepede, Hist. Nat. Poiss., V, p. 84. (Type Macrop- 
leronoliis charmuth Lacepede.) 

Adipose fin none; dorsal long, extending from the nape to the 

^Liobagrus andersoni Regan, Proc. Z06I. Soc. London, 1908, p. 61; Kimhoa, 
Corea. 



The Fresh Water Fishes of the Island of Formosa. 185 

caudal; anal long. Jaws with a band of villiform teeth; a band of 
villiform or granular teeth across the vomer; cleft of mouth trans- 
verse, anterior, of moderate width; barbels eight, one pair of nasal, 
one of maxillar}', and two pairs of mandibular barbels. Eyes small, 
with a free orbital margin. The upper and lateral parts of the head 
are osseous, or covered with only a very thin skin. A dendritic, 
accessory branchial organ is attached to the convex side of the second 
and fourth branchial arches, and received in a cavity behind the gill- 
cavity proper. Ventrals six-rayed; only the pectoral has a pungent 
spine. 

Distribution: Africa; Syria; Bengal; India; Sumatra; Java; Borneo; 
Ceylon; Luzon; Mindanao; Formosa; China; Cochin-China. 

lo. Clarias fuscus (Lacepede). 
Tosa or Tause (Formosa). 

1803. Macropleronotus fuscus Lacepede, Hist. Nat. Poiss., V, p. 88, pi. 2, fig. 2. 
1846. Clarias pulcaris Richardson, Voy. Sulph., Fish., p. 135, pi. 62, fig. 56; 

Canton, China. 
1864. Clarias fuscus Gunther, Cat. Fish., V, p. 18; China. — Jordan & Evermann, 

Proc. U. S. Nat. Mus., XXV, 1903, p. 321; Taihoku, Formosa. — Jordan & 

Richardson, Mem. Carneg. Mus., IV, no. 4, 1909, p. 168; Taihoku; Takao. 

— Vaillant, Bull. Mus. Hist. Nat., VI, 1904, p. 297; Tongking. 
1897. Clarias fuscus Rutter, Proc. Acad. Nat. Sc. Philad., Jan., p. 57; Swatow, 

China. 
1908. Clarias sauleri Regan, Ann. Mag. Nat. Hist. (8), I, p. 151; Kagi; Formosa. 

Head 4. 11 in length; depth 5.66; D. 58; A. 42; P. I, 9; V. 6; width 
of head 1.2 in its length; eye twelve times in head, five times in snout, 
six and one-half times in interorbital space, which is contained 1.75 
in head; pectoral 1.33; ventral 2.66. 

Body compressed, higher in front; head flattened, inclined anteriorly, 
with two oval depressions along the median longitudinal line, one on 
the interorbital space, the other on occiput; tail long and tapering; 
snout flattened, truncated in front; mouth slightly inferior, transverse; 
lips granulated, upper lip much thicker than the lower; jaws with a 
band of villiform teeth, upper jaw protruding; vomer with a crescent- 
shaped band of villiform teeth, which is narrower than that of the inter- 
maxillary; eyes very small, supra-lateral; nostrils separated, anterior 
nostril in contact with upper lip, in a short tube, posterior nostril 
superior, just behind the base of rostral barbel; eight barbels, one 
pair rostral, one maxillary, two mandibular, of which the maxillary 



186 



Annals of the Carnegie Museum. 



pair are the longest, about 1.2 in the length of head; gill-openings low; 
gill-membranes entirely separated; gill-rakers 4 + 14, slender and 
pointed. 

Dorsal fin very long, originating at anterior third of body without 
caudal, extending posteriorly to the base of the caudal; anal fin long, 
inserted slightly posterior to a point midway between tip of snout and 
base of caudal; pectoral fin armed with a strong, short spine, which 
has a fine serration below the skin; the ventral very small, its tip 
reaching beyond base of anal fin; caudal fin separated from the dorsal 
and anal; rather long, the tip rounded. 

Body naked, with numerous undulating vertical striations on the 
sides; lateral line descends a little at the commencement and runs 
straight at mid-height. 

Color in formalin uniformly dark brown, lower parts whitish; 
sides with about eight vertical rows of minute white spots, descending 
from the back to the lateral line, and two or more longitudinal rows of 
the same below the lateral line. 

Total length 192 mm. 

The present description from a specimen from Jitsugetsutan, 
collected by T. Aoki, in August, 1916. 

Habitat: Very common in the fresh waters of Formosa. My 
specimens came from Jitsugetsutan (Lake Candidius) ; Taihoku; 
Tamusui River near Shinten; Maruyama near Taihoku; Giran. 

Measurements of Clarias fuscus. 



Locality. 



■6 


J3 












P, 





< 


CL, 


> 


X 


Q 










4. 1 1 


5-66 


58 


42 


1.9 


6 


4.28 


5-45 


55 


36 


1.9 


6 


4.44 


5-83 


54 


35 


1.9 


6 


4-35 


5-32 


53 


41 


1.9 


b 


4.09 


6 


52 


39 


1.9 


6 


4.27 


5-50 


55 


40 


1.9 


6 









Jitsugetsutan. 
Jitsugetsutan. 
Maruyama. . 
Maruyama . . 
Maruyama. . . 
Giran 



1.20J 2 
i.i6| 1. 75 
1. 14 1.83 
1. 12 1.86 
1.20 2 
1.60I 1.8 



3 |I2 
2.33 10 
2.28 11.66 
2.36 10.33 
2.55 10 
2.50 10.33 



192 
100 
162 
132 
no 
218 



Remarks: In the year 1908, Mr. C. Tate Regan described a new 
catfish from Kagi, Formosa, giving it the name Clarias saiiteri, and 
made the following statement: "C. saiiteri is close to the Chinese 
C. juscus Lacepede, which differs notably in having villiform teeth 
on the palate." It is observed, however, that large specimens of 
C. Juscus are pro\ide<l with more or less enlarged vomerine teeth, 



The Fresh Water Fishes of the Island of Formosa. 187 

instead of being villiform, as in small specimens. Thus the size of the 
vomerine teeth is an individual difference, not specific. Therefore 
it is unreasonable to separate C. saiiteri from C. fiiscus. 

Family COBITID.^. 

Artificial Key to the Formosan Genera. 

a. No erectile spines below the eye; ten or twelve barbels; four about the 

mandible; lateral line medium Misgurnus. 

ua. An erectile spine below the eye; six barbels, only on the upper jaw; lateral line 
incomplete Cobitis. 

Genus Misgurnus Lacepede. 
1803. Misgurnus Lacepede, Hist. Nat. Poiss., p. 16. (Type Cobitis fossilis 
Linnaeus.) 

Body elongate, compressed. Head triangular, elongate, compressed; 
snout projecting; mouth inferior, with fleshy lips; barbels ten or twelve, 
of which four are mandibular; eye small. Gill-openings lateral; 
lateral line complete. No spine below the eye. Body with small 
scales, except on the head, which is naked. Origin of the dorsal about 
in the middle of the length of the fish, over the ventrals; anal entirely 
behind dorsal; pectorals more or less equal to the head; caudal nearly 
equal to head, and rounded. Air-bladder in a bony capsule (Jordan & 
Fowler). 

Distribution: Europe; Bengal; India; Siam; Indo-China; China; 
Formosa; Amur Province; Japan. 

Synopsis of the Formosan Species. 
a. Scales relatively small, 140-150 in lateral series; body slender, the depth 7-8 in 
length; barbels short, the longest, about 2.5 in head; color dark gray, above 
spotted and marbled with dark, base of the caudal above with a black spot. 

anguillicaudalus. 

aa. Scales relatively large, 1 06-11 5 in lateral series; body plump, the depth 5.86- 

6.50 in length; barbels long, the longest about 1.8 in head; color relatively 

plain, obtusely speckled with minute dark spots, not marbled, base of the 

caudal above with no black spot decemcirrosus. 

II. Misgurnus anguillicaudatus (Cantor). 
Dojo (Japan); Horyu (Formosa). 

1842. Cobitis anguillicaudatus Cantor, Ann. Mag. Nat. Hist. IX, p. 485. — - 
Richardson, Voy. Sulph., Fish., 1846, p. 143, PI. 55, Figs. 9 and 10; China. 

1868. Misgurnus anguillicaudalus Gltnther, Cat. Fish. VH, p. 345; China; 
Japan; Formosa. — Ann. Mag. Nat. Hist., Sept., 1873, p. 250; Shanghai. — 



188 Annals of the Carnegie INIuseum. 

Peters, Monatb. Konigl. Akad. Berlin, 1880, p. 926. — Sauvage, Nouv. 

Arch. Mus., 1881, p. 190. — Gunther, Ann. Mus. St. Petersburg, 1896, 

p. 19; Kansu; China. — • Rutter, Proc. Acad. Nat. Sc. Philad., 1897, p. 60; 

Swatow. — Fowler, Proc. Acad. Nat. Sc. Philad., 1899, p. 179; Tan-lan-ho, 

China. — Jordan & Snyder, Proc. U. S. Nat. Mus., XXIII, 1901, p. 340; 

Japan. — Regan, Ann. Mag. Nat. Hist. (7), XIII, 1904, p. 192; Yunnan-Fu, 

China. — Vaillant, Bull. Mus. Hist. Nat., 1904, VI, p. 298; Tongking. — 

Jordan & Snyder, Proc. U. S. Nat. Mus., XXX, 1906, p. 834; Japan. — 

Snyder, Proc. U. S. Nat. Mus., XLII, 1912, p. 404; Hakodate; Tokyo; 

Takamatsu River. — Jordan, Snyder, & Tanaka, Journ. Coll. Sci. Tokyo, 

XXXIII, 1913, p. 60; Japan. 
1846. Cohilis maculata Temminck & Schlegel, Fauna Japonica, Poiss., p. 221, 

PI. cm, Fig. 2; near Nagasaki. 
1846. Cobilis nibripinnis Temminck & Schlegel, Fauna Japonica, Poiss., p. 220, 

PI. cm, Fig. i; near Nagasaki. 
1846. Cobilis micropus Cuv. & Val., Hist. Nat. Poiss., XVTII, p. 29, China. 
1846. Cobilis psammismiis Richardson, Ichthyol. China, p. 300; Canton, China. 
i860. Cobitichthys enalios Bleeker, Act. Soc. Indo-Neerl., VIII; Japan. — IV, 

p. 88, PI. II, Fig. 4; Japan. 
i860. Cobiiichlhys dechachraus Bleeker, Act. Soc. Sci. Indo-Neerl., VIII; Japan. 

—IV, p. 89, PI. II, Fig. 2; Tokyo. 
1868. Misgurnus dechachraus Gunther, Cat. Fish., \TI, p. 346; Tokyo. 
1878. Misgurnus crossochilus Sauvage, Bull. Sc. Philom., Jan., p. 4; Koaton, 

Cochinchina. 
1907. Misgurnus fossilis anguilUcaudalus Berg, Proc. U. S. Nat. IVIus., XXXII, 

p. 435; Amur Province. 

Head 6 in length; depth 7; D. 9; A. 7; P. I, 9; V. 6; width of head 
2 in its length; eye 3 in snout, 1.66 in interorbital space; snout 2.5 in 
head; pectoral 1.6; ventrals 2; scales about 140. 

Body elongate, compressed; head small, triangular, compressed; 
snout rather long, obtuse, somewhat produced; eyes small, anterior 
and superior; mouth inferior, with thick fleshy lips; ten barbels, four 
belonging to the mandible; the longest barbel 2.5 in head; nostrils 
close together, in front of eye, the anterior in a short tube; interorbital 
space slightly convex; gill-openings lateral; gill-membranes joined 
below in front of the base of the pectoral. 

Origin of dorsal about midway between tip of snout and tip of caudal, 
a little in front of the ventral; anal fin entirely behind the dorsal, 
nearer the origin of the ventral than the base of caudal; pectoral fin 
short and low; the caudal oblong, broad, rounded, nearly as long as 
head; caudal peduncle long and deep, its depth about two-thirds of 
the length of head. 

Head naked, trunk covered with small cycloid scales; lateral line 
extends along the middle of the sides. 



The Fresh Water Fishes of the Island of F"ormosa. 189 



Color in formalin dark gray, upper half of the body spotted and 
marbled with darker; base of the caudal above with a black sjjot; 
dorsal and caudal fins with several rows of dark small spots; sides of 
body with a number of obscure longitudinal black stripes; belly to- 
gether with pectorals, ventrals, and anal whitish. 

Total length 107 mm. 

Described from a specimen from Giran, collected on December 2, 
1916. 

Habitat: Very abundant in all the rivers and rice-fields of the island. 

My specimens came from the Tamusui River near Taihoku; Maru- 

yama; Giran; Jitsugetsutan (Lake Candidius) ; Rato and Raupi; 

Giran. 

Measurements of Misguniiis anguillicaudalus. 



Locality. 



-0 


0. 


Q 


< 


fe 


> 


•s'S 


— 

4J *j 


X 


a 













CJi 


6 


7 


9 


7 


1.9 


6 


2 


4-25 


6 


6.87 


7 


6 


1.9 


6 


1.6 


4-33 


5-6 


7.29 


8 


7 


I. 9 


6 


1.69 


4.80 


5.42! 7.6 


9 


7 


1.9 


6 


1-75 


4 


5-561 8 


8 


7 


I. 9 


6 


1-75 


4 


6 


7-4 


8 


6 


1. 9 


6 


1.86 


4 



c J:; 



Giran 

Giran 

Maruyama . 
Maruyama . 
Maruyama. 
Maruyama. 



2.40 
2.66 
2.30 
2.28J 8 
2.33 8.66 
2.i6l 7.33 



140 
140 
150 
148 



107 

120 

150 

90 

90 



149 



12. Misgurnus decemcirrosus (Basilewsky). 

1855. Cobitis decemcirrosus B.\silewsky, Mem. Soc. Nat. Moscow, p. 239, PL 7; 

near Peking. 
1888. Misgurnus mizolepis Gunther, Ann. Mag. Nat. Hist., June, p. 434; Kiu- 

Kiang, China. 
1901. Misgurnus anguillicaudalus Abbott, Vvoc. U. S. Nat. Mus., XXIII, p. 489; 

Tientsin, China. — Jordan & Evermann, Proc. U. S. Nat. Mus., XXV, 1903, 

p. 321; Tiaholu, Formosa. 
1906. Misgurnus decemcirrosus, Jordan & Snyder, Proc. LI. S. Nat. Mus., XXX, 

p. 834; Tientsin, China. — ^Jordan & Rich.\rdson, Mem. Carneg. Mus., IV, 

No. 4, 1909, p. 169; Formosa. 

Head 7 in length; depth 6.2; D. 7; h. 6; P. I, 9; V. 6; width of head 
2.33 in its length; eye 7 in head, 3 in snout; 2 in interorbital space; 
snout 2.5 in head; pectoral 1.33; ventral 2; scales about 112 in lateral 
series, 23 in transverse series from origin of dorsal to the ventral. 

Body elongate and compressed; head small, triangular, compressed; 
snout long, pointed and produced; eyes small, superior, and nearer 
tip of snout than gill-opening; mouth inferior, with thick fleshy lips; 
ten barbels, four belonging to the mandible, the longest barbel 1.8 in 



190 Annals of the Carnegie Museum. 

head; nostrils in front of eye, separated a little, the anterior in a short 
tube; interorbital space convex, 1.66 in snout; cheeks slightly swollen; 
gill-openings lateral; gill-membranes joined below in front of the 
base of the pectoral. 

Origin of the dorsal about midway in the length of body including 
caudal, slightly in front of the origin of ventral; anal fin entirely 
behind the dorsal, nearer to origin of ventral than base of caudal; 
caudal squarish, with obtusely rounded tip, nearly as long as head; 
pectoral fin short and low; the ventral short; caudal peduncle deep 
and long, its depth about three-fourths the length of head. 

Head naked; trunk covered with relatively large cycloid scales; 
lateral line extends along the middle of the sides. 

Color in formalin uniformly dark gray, obscurely speckled with 
small dark spots, not marbled; dorsal, anal, and caudal fins dusky, with 
numerous small dark spots; the pectoral and the ^■entral whitish, 
some dusky on the longer rays; belly yellowish white. 

Total length 145 mm. 

The present description is taken from a specimen from the fish- 
market of Taichu, collected by T. Aoki in December, 191 5. 

Habitat: Taihoku; Taichu. 

Remarks: In the year 1888 Giinther recorded a Chinese species of 
Misgurnus under the name of Misgiirniis mizolepis and stated that it 
has larger scales than any other species of the genus known to himself. 
According to his description, the scales of Misgurnus mizolepis are 
arranged in thirteen longitudinal rows between the dorsal fin and the 
lateral line, and ten between the lateral line and the ventral fin. 
Scales of Misgurnus decemcirrosus are also much larger than any other 
species of the genus, numbering 112 in the lateral series and 23 in an 
oblique series from the origin of the dorsal to the ventral. Moreover- 
as all other characters of that species agree quite well with those of 
M. mizolepis, there is no doubt that these two are the same species. 

Though Jordan & Snyder believe that M. decemcirrosus from 
Northern China differs from M. anguillicaudatus in ha\ing large 
scales (about 112 in lateral series), rather deep body, long barbels, 
and relatively plain color. Berg denies that the former is different from 
the latter. After examining a vast number of specimens of M. 
anguillicaudatus he comes to the conclusion that M. decemcirrosus is 
a nominal species, because there is no difference between Japanese 
and Nortli Chinese sjjecimens of M. anguillicaudatus in respect of the 



The Fresh Water Fishes of the Island of Formosa. 191 



number of scales and the depth of the body. In addition he asserts: 
"If we regard the specimens from near Peking as a distinct species we 
shall be obliged to regard many varieties of the same species found in 
other parts of China and in Japan also as distinct species, which is 
inadmissible in consequence of the known variability of anguilli- 
caudatus." 

It is certain that there is no specific difference between Chinese and 
Japanese M. anguillicaudatus ; however, it is also true that in China and 
Formosa there exists a large-scaled species of Misgurfius (Basilewsky's 
Cohitis decemcirrosus or Giinther's Misgurnus mizolepis) which is 
distinctly separate from common M. angiiillicaudaUis. According 
to Berg's statement all the specimens of Misgurnus which were 
examined by him belonged to M. anguillicaudatus, not M. decemcir- 
rosus, differing in having more than 145 scales in the lateral series. 
Not only is the large-scaled Misgurnus decemcirrosus not a nominal 
species, but it is proved that it is distributed in Northern and Southern 
China as well as in Formosa, distinctly differing from the common 
loach, Misgurnus anguillicatidatus. Therefore Berg's opinion with 
reference to M. decemcirrosus is incorrect. 

Measurements of Misgurnus decemcirrosus. 



Locality. 


s 

X 


.5" 
0. 

Q 


Q 


-1; 


cC 


> 








c 

CO 




CO 


J3 . 

tie 


Taichu 


6.S 


6.2 


7 


6 


1-9 


6 


2.,^.l 


.S-6 


2.50 


7 


112 145 


Taichu 


6 


5.86 


7 


6 


1.9 


6 


2 


3.6 


2.25 


7-33 


112 123 


Taihoku 


6 


6.5 


7 


6 


1.9 


6 


1.88 


3.75 2.66 


6 


115 108 


Taihoku 


6 


6.2 


7 


6 


1,10 


6 


2 


4 


2.50 


7 


106 


100 


Taihoku 


6 


6.2 


7 


6 


1,10 


6 


1.86 


4 


2.60 


6 


106 


Q8 


Taihoku 


6 


6 


7 


6 


1. 10 


6 


2 


4 


2.50 


7 


110 


lO.S 


Taihoku 


6 


6.5 


7 


6 


1,10 


6 


1.88 


3-66 


2.60 


6 


112 


105 



Genus CoBiTis Linnaeus. 
1738. Cobitis Artedi, Genera (non binomial). 

1758. Cobitis LiNN.«:us, Syst. Nat., Ed X, p. 303. (Type Cobitis la-nia Linnaeus.) 
1835. Acanthopsis Agassiz, Mem. See. Sci. Nat. Neuchatel, I, p. 36. (Type 
Cobitis tcenia Linnaeus.) 

Body elongate, more or less compressed, and the trunk not arched. 
Head elongate, compressed; eyes small; snout produced, blunt and 
rounded; mouth small, inferior, and with six barbels about the upper 
jaw; below the eye an erectile bifid spine. Dorsal fin about over 
the ventrals; anal behind dorsal; caudal rounded or truncate; pectorals 



192 Annals of the Carnegie Museum. 

less than the head; ventrals below dorsal. Air-bladder inclosed in a 
bony capsule. Lateral line incomplete. Small fresh-water fishes of 
Europe and Asia (Jordan & Fowler). 

Distribution: Europe; Assam; Bengal; China; Formosa; Corea; 
Amur province; Japan. 

13. Cobitis taenia Linnaeus. 

1758. Cobitis tania Linn.eus, Syst. Nat. Ed. X., p. 303; Europe. — Gunthep, 

Cat. Fish., VII, 1868, p. 362; Europe; Japan. — -Jord-^n & Fowler, Proc. 

U. S. Nat. Mus., XXVI, 1903, p. 771; Japan. 
1846. Cobitis ttenia japonica Schlegel, Fauna Japonica, Poiss., p. 222, PI. CIII, 

Fig. 2; near Nagasaki, 
1875. Cobitis sinensis Sauvage & De Thiersant, Ann. .Sci. Nat., Ser. 6, I, p. 8; 

Setchuan, China. — Fowler, Proc. Acad. Nat. Sci. Philad., 1899, p. 182; 

Tan-lan-ho, China. — Jordan & Metz, Mem. Carneg. Mus., \'I, 1913, p. 12; 

Suigen; Gensan; Fusan, Corea. 
1901. Cobitis biww Jordan &: Snyber, Proc. U. S. Nat. Mus., XXIII, p. 748; 

Lake Biwa (substitute for Cobitis japonica pre-occupied). — Jordan, 

Sn\T)er, & Tanaka, Journ. Coll. Sci. Tokyo, XXXIII, 1913, p. 62; Japan. 

Head 5 in length; depth 6.166; D. 8; A. 7; P. 10; V. 7; width of 
head a little over two in its length; snout two times in head; inter- 
orbital space 5.33; eye 5.5; pectoral 1.5; ventral 1.66. 

Body elongate and compressed; head elongate, strongly compressed, 
with conve.x upper profile; snout long, somewhat produced, anterior 
border bluntly rounded; eyes small, superior and lateral, located 
midway between tip of snout and gill-opening; mouth small, inferior, 
with fieshy lips, the lower bilobed; eight barbels, two of which are 
mandibular; nostrils nearer the eye than the tip of snout, close 
together, the anterior in a short tube; interorbital space narrow; gill- 
openings large, lateral; gill-membranes united below the base of the 
pectoral in front. 

Origin of dorsal nearer the base of caudal than the tip of snout, 
somewhat in advance of ventral, length of dorsal when depressed a 
little less than the length of head; pectoral a little longer than one- 
third the distance between its base and origin of ventral; ventral 
twice in the space between its origin and that of the ventral; anal fin 
entirely behind the dorsal, reaching two-thirds of the space between 
its origin and the base of caudal; caudal peduncle compressed, its 
depth slightly less than twice in head. 

Head naked, trunk covered with ver>' small cycloid scales; lateral 
line very short, extending a little beyond the middle of the pectoral. 



The Fresh Water Fishes of the Island of Formosa. 193 

Color in formalin pale grey above, lower parts and belly yellowish 
white; sides with two rows of dark blotches, the u])per one forming a 
continuous longitudinal band anteriorly, blotches of the lower row 
larger; between the two rows a narrow paler marbled brown streak, 
occupying the anterior half the interspace; eight blotches of dark 
brown between the nape and the origin of the dorsal, seven more 
between the latter and the base of the caudal; base of the caudal above 
with a jet-black spot; dorsal and caudal fins with several rows of dark 
small spots; other fins whitish; head marbled and spotted with brown 
above; back with mottlings; a black streak from eye to snout. 

Total Length 122 mm. 

Described from a specimen from Shinchiku, collected by T. Aoki 
on September 17, 1916. 

Habitat: Shinchiku; Jitsugetsutan (Lake Candidius) ; Rigyokutsu. 
Nanto; Maruyama, Giran. 

Remarks:T\\e markings of the present species are more or less 
variable. The other example from Shinchiku has the brown blotches 
on the sides distinctly separated into spots, while the example from 
Rigyokutsu has the upper blotches on the sides merged into continuous 
longitudinal bands. 

Measurements of Cobitis tcenia. 



Locality. 



<; 


PlJ 


>' 


■5'i 



V iS 


3 

a 
in 


7 


10 


7 


2.75 


6.75 


2 


7 


10 


7 


2.80 6 


2 


6 


10 


7 


2.60 6 


2 


6 


10 


7 


2.60 


6 


2 






Shinchiku. . . . . 

Shinchiku 

Jitsugsetsutan. 
Rigyokutsu . . . 



6.16 
6.18 
7-50 
6.88 



5-33 
5-66 

5 

5 



122 
82 
82 
67 



Family HOMALOPTERID.'E. 
Artificial Key to the Formosan Genera. 

a. Body elongate, rather high, anterior part depressed; mouth inferior; upper Up 
fleshy, with a distinct inner fo'd; barbels numerous, one pair of maxillary, two 
transverse series of minute ones on upper jaw, and three pairs on the lower 
jaw; dorsal fin inserted in advance of the ventral; pectoral with one simple 

outer ray, horizontal Formosania. 

aa. Body much depressed, nearly twice as broad as high; mouth inferior, with 
fringed upper lip; six barbels, two pairs on the extremity of the snout, the other 
at the angle of mouth, minute and subequal; dorsal fin inserted behind the 
origin of the ventral; pectoral with eleven simple outer rays, subhorizontal. 

Heminiyzon. 
14 — DEC 17, 1919. 



194 Annals of the Carnegie Museum. 

Genus Formosania gen. nov. 
Type Formosania gilherti Oshima. 

Body elongate, rather high, anterior part depressed; snout spatulate; 
mouth inferior; upper lip fleshy, with a distinct inner fold; lower jaw 
with a sharp horny edge; barbels numerous, inferior, upper jaw with 
two transverse rows of minute barbels and one pair of thick maxillary 
barbels, lower jaw with three pairs of small barbels; scales minute; 
lateral line continuous; dorsal fin high, inserted in front of the origin 
of the ventral; pectoral fin large, not reaching the ventral; the anal 
large, when depressed reaching the root of caudal; pectoral and ventral 
fins horizontal, one outer ray simple. 

Remarks: Well distinguished from other genera of the Ilomalop- 
terida: by the presence of numerous barbels, especially by the rostral 
barbels which are arranged in two series. 

Distribution: Formosa and China. 

14. Formosania gilberti sp. nov. (Plate XLIX, Figs. 1-2). 

Head 4.66 in length; depth 6.2; D. 2.8; A. 2, 5; P. 15; V. 9; width of 
head 1.22 in its length; eye 6 in head; interorbital space 2.63; snout 
1.75; ventral 1.31; pectoral longer than head; scales about 120 in the 
lateral line. 

Body elongate, lower surface flat, tail compressed, cross-section of 
body triangular, anterior part feebly depressed; head moderate, top 
nearly flat; snout spatulate, much longer than postocular part, shorter 
than width, anterior margin broadly rounded; skin of the snout ex- 
tending to the lower surface, but not overlapping upper lip; mouth 
inferior,- transverse; upper lip thick and fleshy, with a distinct inner 
fold; lower jaw much shorter than the upper, anterior margin rounded, 
horny; mentum with a broad fleshy tubercle, the tip of which is bilobed; 
barbels numerous, inferior, rostral barbels minute, arranged in two 
transverse rows, each with about five barbels, maxillary pair thick and 
longest, lower jaw with six barbels, posterior mental pair the longest, 
the others short and tubercle-like; eyes small, supra-lateral, posterior, 
nearer angle of gill-cover than tip of snout; nostrils close together, in 
front of eye; boundary between head and trunk distinct, occiput 
pointed posteriorly. 

Origin of dorsal nearer tip of snout than base of caudal, in advance 
of that of the ventral, rather high, the anterior ray longest; pectoral 
horizontal, flattened, large, outer margin broadK' rounded, not 



The Fresh Water Fishes of the Island of Formosa. 195 

reaching the ventral, one outer ray undivided; ventral inserted 
beneath the middle of base of dorsal, ovoid, middle ray the longest; 
anal fin large, when depressed scarcely reaching the root of caudal; 
caudal fin elongate, weakly emarginate, each lobe sharply pointed. 

Bod}' covered with minute scales, head and lower surface naked; 
lateral line nearly straight, extending along the middle of the sides. 

Color in formalin dark yellow, mottled with irregular dark brown 
blotches, lower surface white; head uniformly dark, with few yellowish 
markings; fin-rays of the dorsal with elongate black spots; caudal fin 
with a number of black cross-bars; other fins dusky, with numerous 
dark spots. 

Total length 117 mm. 

The present description is from a specimen from Tamusui River 
near Shinten, collected by T. Aoki in December, 1916. 

Habitat: Tamusui River (four specimens). 

Remarks: The present species is very closely related to Boulenger's 
Homaloptera stenosoma (misprinted Homalosoma) from Ningpo, 
China (Proc. Zool. Soc. London, March, 1901, p. 270), which seems to 
belong to the genus Formosania. The latter differs from F. gilberti 
in having seven-branched dorsal rays, a smaller number of scales in 
the lateral line, and slightly shorter head and pectorals. 





Measurements of 


Formosania gilberti. 










Locality. 


-6 




Q 


< 


fc 


>■ 


131 <u 




0— : 

U crt 

c3 


3 


a 


V 




at) B 


Tamusui River. . . . 
Tamusui River. . . . 


4.66 6.2 2.8 
4.72 5.36 2.8 


2.5 IS 
2.5 1 16 


9 
9 


1.22 
1. 21 


2.625 
2.666 


1-75 
1.60 


6 
6 


120 
130 


117 
90 



Genus Hemimyzon Regan. 

191 1. Hemimyzon Regan. Ann. Mag. Nat. Hist. (8), p. 31. (Type Homaloptera 
formosaniis Boulenger.) 

Body much depressed, nearly twice as broad as high; head disc- 
shaped, flattened; snout broad, with sharp anterior margin; mouth 
inferior, transverse, with fringed upper lip; edge of the lower jaw 
sharp, horny; six barbels, inferior, two pairs on the extremity of the 
snout, the others at the angle of mouth, minute and subequal; origin 
of ventral in advance of that of dorsal; the pectoral subhorizontal, 
eleven outer rays simple; ventral fin horizontal, four outer rays simple; 



196 Annals of the Carnegie Museum. 

the anal minute, not reaching the caudal; scales minute; lateral line 
nearly straight, continuous. 
Distrihiition: Formosa. 

15. Hemimyzon formosanus (Boulenger). 

1894. Homaloptera formosanus Boulenger, Ann. Mag. Nat. Hist., Ser. 6, XIV, 

p. 463; Central Formosa. 
.1911. Hemimyzon formosanus Regan, Ann. Mag. Nat. Hist. (8), VHI, p. 32; 
Formosa (after Boulenger). 

Head 4.25 in length; depth 6 in length and 1.5 in width; D. 2, 7; 
A. I, 5; P. 22; V. 15; width of head i in its length; eye 5.66 in head; 
snout 1. 71; interorbital space 2.3; pectoral 1.33 times as long as head; 
ventral very slightly shorter than the pectoral; scales about 70 in the 
lateral line. 

Body strongly depressed, low, tail compressed, middle part of body 
broadest, belly and throat flat; head flattened, disc-shaped; gill- 
opening rather large; snout broad, with rounded sharp anterior margin; 
interorbital space rather flat; mouth inferior, transverse, crescent- 
shaped; upper lip fringed with a fleshy inner fold; lower jaw shorter 
than the upper; lower lip with a smooth horny inner fold, its anterior 
margin sharp; six barbels, short, subequal, two pairs on the lower 
extremity of snout and the other at the angles of mouth; eyes small, 
superior and posterior, considerably nearer the angle of operculum 
than tip of snout; nostrils close together, superior, in front of eye, 
anterior nostril in a short tube. 

Origin of the dorsal very slightly nearer tip of snout than base of 
caudal, above anterior third of the base of ventral; pectoral fin sub- 
horizontal, with eleven simple rays, outer margin broadly rounded, 
extending beyond the origin of ventral; ventral fins horizontal, flat, 
distinctly separated, with four simple rays, not reaching the anus; the 
anal very small, entirely behind the donsal; caudal fin forked, the 
tip of each lobe pointed, lower lobe slightly longer than the upper. 

Scales minute; belly and throat naked; lateral line nearly straight, 
slightly upcurved above the pectoral, extending along the middle of 
the sides. 

Color in alcohol dark gray above, lower surface yellowish white; 
caudal fin with four dark brown cross-bars; other fins pro\'ided with 
a number of dark elongate spots. 

Total length 64 mm. 



The Fresh Water Fishes of the Island of Formosa. 197 



Described from a specimen from Taiko River, collected by T. Aoki 
in December, 1916. 

Habitat: Taiko River (five specimens). 





Measurements 


OF Hemimyzon formosanus. 








Locality. 




a 

a 

Q 


Q 


< 


CM 


>■ 


■si 

n 


u 
— 

_c3 


3 

C 


4) 

w 


J3 


Taiko River 

Taiko River 


4-25 
4 


6 

7 


2.7 
2.7 


1-5 
1-5 


22 
21 


15 
15 


1. 125 


2.5 
2 


1. 71 

1.6 


7 
6.5 


64 

43 



Family CYPRINID/E. 

Artificial Key to the Formosan Genera. 
I. Anal fin very short, with five or six branched rays. 

A. Lateral hne running along the median line of the tail; dorsal fin opposite to 

ventrals. 
a'. Dorsal fin with more than nine branched rays. 

b'. Pharyngeal teeth in three series, the outer molar-like; barbels 

two on each side; a strong dorsal spine Cyprinus. 

b". Pharyngeal teeth in a single series, barbels none; a strong dorsal 

spine Carassius. 

b'". Pharyngeal teeth in three series, not molar-like; without osseous 
dorsal rays. 
Snout more or less swollen; each lip with an inner transverse 
fold, which is covered with a deciduous horny substance form- 
ing a sharp edge; dorsal 13-20 Labeo. 

a" . Dorsal fin with no more than nine branched rays. 
b' . Pharyngeal teeth in three series. 

c' . Lower jaw transverse, with a narrow lip which is not con- 
tinuous, with a sharp inner transverse edge; barbels four. 

Acrossocheilus. 

c" . Mouth transverse, inferior; lower jaw not covered by lip, 

with a horny layer inside, rather sharp; barbels two, minute. 

Scaphesthes. 
c'" . Mouth arched, without inner fold or other peculiarities; 
lips fleshy. 
d' . A recumbent spine in front of the dorsal; barbels four. 

Spiyiibarbus. 
d" . No recumbent spine; barbels four, or two, or none. 

e' . Head with mucous cavities; scales large; lips thick; 

barbels two Hemibarbiis. 

e". Head without mucous cavities; scales moderate; 
lips thin. 

/'. Barbels four Barbodes. 

f". Barbels two Capoeta. 

/'". Barbels none Puntius. 



198 Annals of the Carnegie Museum. 

h". Pharyngeal teeth in two series. 

c'. Lips thin; mouth terminal, oblique; dorsal nearly over the 

ventral Gnathopogon. 

c" . Lips broad and papillose; mouth small, protracted down- 
wards; dorsal inserted well in advance of the ventral. 

Pseudogohio. 
b'". Pharyngeal teeth in a single series. 

c'. Mouth very small, transverse, directed upwards; mandible 
with a hard, trenchant eege, no hdLvh&ls. .Pseiidorashora. 
B. Lateral line running along the lower half of the tail; dorsal fin inserted 
behind the origin of ventrals. 
a'. Pharyngeal teeth in two series, 4, 4-4, 4; lateral line complete. 

Pararasbora. 
II. Anal fin short or of moderate length, with from seven to eleven branched rays, 
not extending forwards to below the dorsal fin. 

A. Dorsal fin with osseous ray. 

a'. Dorsal fin rather short, with two smooth spines and seven branched 
rays; barbels none; pharyngeal teeth in a double series. Pf^/crc/jo^/ow. 

B. Dorsal fin without osseous ray. 

a'. Lateral line incomplete; dorsal inserted behind the origin of the 
ventral; pharyngeal teeth in double series, slender and hooked, 5, 3- 

4, 4 Phoxiscus. 

a". Lateral line complete; dorsal inserted in advance of the origin of the 
ventral; pharyngeal teeth in double series, their outer surfaces 

deeply folded, 5, 2-2, 5 Ctenopharyngodon. 

III. Anal fin of moderate length, extending forwards to below the dorsal; lateral 
line, if complete, running into, or nearly into the middle of the tail; pharyn- 
geal teeth in a single series; no dorsal spine. 

A. Lateral line complete; anal fin with eight to ten rays Acheilognalhus. 

B. Lateral line incomplete; anal fin with about twelve rays Rhodeus. 

W. Anal fin of moderate length; lateral line running along the lower half of the 

tail; pharyngeal teeth in three series. 

A. Barbels, none. 

a'. Body moderately elongate and compressed; pharj^ngeal teeth i or 2, 
4, 4 or 5-5, or 4, 4, 2 or I ; dorsal fin inserted a little behind or oppo- 
site to the origin of ventrals Zacco. 

a". Body deep and strongly compressed; pharyngeal teeth 2, 4, 4-4, 4, 2; 
dorsal fin inserted above the interspace between ventrals and anal. 

Melzia . 

B. Barbels, two. 

a'. Body elongate, laterally compressed; pharyngeal teeth i, 4, 5-5, 4, i; 

dorsal fin inserted opposite to ventrals Candidia. 

\. Anal fin elongate; lateral line running along the middle of the tail; no dorsal 
spine; pharyngeal teeth in a single series, 4-4. 
a'. Entire abdominal edge carinated; gill-rakers continuous, forming a cres- 

centic horny membrane H y po phlhalmichlhys. 

a". Postvcntral edge carinated; gill-rakers separated, slender and long. 

Arislichlhys. 



The Fresh Water Fishes of the Island of Formosa. 199 

\'I. Anal fin elongate; abdomen, or a part of abdomen, compressed and carinated. 

A. Lateral line with no conspicuous curve, slightly decurved; entire, or a 

part of abdomen carinated. 
a'. Abdomen behind the ventral compressed to an edge; before ventrals 
rounded; anal elongate, of seventeen rays; first dorsal ray more or 

less enlarged and spine-like Ischikauia. 

a". Postventral edge carinated; rounded before ventrals; profile of the 
nape remarkably convex; lower jaw not protruding; anal fin elon- 
gate with more than twenty rays; dorsal fin armed with strong 

spines; scales rather large Chanodichthys. 

a'". Entire or postventral edge carinated; profile of the nape slightly 
convex; lower jaw protruding; anal fin elongate; with more than 
twenty rays; dorsal fin armed with strong spines; scales small. 

Culler. 

B. Lateral line abruptly bent downwards above the pectoral. 

a'. Abdominal edge entirely carinated; anal fin rather short, with fifteen 
to seventeen rays; dorsal fin with two smooth spines; scales large. 

Cullriculus. 
Genus Carassius Nilsson. 
1832. Carassius Nilsson, Prodromus Ichthyol. Scand. (Type Cyprinus carassius 
Linnaeus.) 
Body oblong, compressed and elevated. Mouth terminal, without 
barbels. Teeth 4-4, molar, but compressed. Scales large. Lateral 
line continuous. Dorsal fins very long, with third ray developed into 
a stout spine, which is serrated behind; anal short with a similar spine. 
Ventrals well forward. Large species of the fresh waters of Europe 
and Asia; often domesticated. (Jordan & Fowler.) 

Distribution: Central and Northern Europe; Siberia; China; Cochin- 
China; Formosa; Corea; Amur Province; Japan. 

16. Carassius auratus (Linnaeus). 
Funa (Japan); Chiira or Chirahii (Formosa). 

1758. Cyprinus auratus Linn^us, Syst. Nat. Ed. X, p. 323. — Gunther, Ann. 
Ac. St. Petersb., 1896, p. 12; Huihsien; Chang-tu-fu. — •Rich.\rdson, Ich- 
thyol. China, p. 293; Tse-Kiang. 

1863. Carassius auratus Bleeker, Atl. Ichthyol. Cypr., p. 74. — Gunther, Cat. 
Fish., VII, 1868, p. 32; China; Japan. — Bleeker, Cypr. China, 1871, p. 7; 
Pekin. — GDnther. Ann. Mag. Nat. Hist., Sept., 1873, p. 246; Shanghai. — 
Sauvage, Bull. Soc. Z06I. France, IX, 1884, p. i; Tonkin. — Rutter. 
Proc. Acad. Nat. Sc. Philad., 1897, p. 58; Swatow. — Fowler, Proc. Acad. 
Sc. Philad., 1899, p. 179; China. — Abbott, Proc. U. S. Nat. Mus., XXIII. 
1901, p. 484- Pei-ho, China. — -Jordan & Fowler, Proc. U. S. Nat. Mus., 
XXVI, 1903, p. 860; Japan. — Jordan & Everjviann, Proc. U. S. Nat. Mus., 
1903, p 321; Formosa. — Vaillant, Bull. Mus. Hist. Nat., VI, 1904, p. 298; 



200 



Annals of the Carnegie Museum. 



Ton-kin. — Jordan & Seale, Proc. U. S. Nat. Mus., XXIV, 1905, p. 519; 
Hongkong. — Snyder, Proc. U. S. Nat. Mus., XLII, 1912, p. 404; Niigata; 
Same; Takamatsu River; Yamaguchi; Dogo Island. — Jordan & Metz, 
Mem. Carneg. Mus., VI, No. 2, 1913, p. 14; Corea. — Jordan, SNi'DER & 
Tanaka, Journ. Coll. Sci. Tokyo, XXXIII, 1913, p. 76; Japan. 

1842. Cyprinus gibeloides Cantor, Ann. Mag. Nat. Hist., IX, p. 485. — Richard- 
son, Ichthyol. China, 1846, p. 292. 

1846. C ar as si us Ian gs dor Jit Scui^EGEL, Fauna Japonica, Poiss., p. 192, pi. 98, Fig. i; 
Japan. — Kner, Novara, Fisch, III, 1867, p. 346; Shanghai. — Peters, 
Monatsb. Ak. Berlin, 1880, p. 924. — Sauvage, Bull. Soc. Philom., 1881, 
p. 7; Swatow. — -Bull. Soc. Zool. France, 1884, p. i; Tonkin. 

1855. Carassius pekinensis Basii.'ewsky, Nouv. Mem. Soc. Nat. Mosc, X, p. 229, 
Tab. 3, Fig. 3; China. 

1855. Carassius discolor Basilewsky, I. c, p. 229; China. 

1855. Carassius coerideus Basilewsky, I. c, p. 229; China. 

1855. Cyprinus macrophthalmus Basilewsky, /. c, p. 230, Tab. V, Fig. 5. 

1846. Cyprinus carassioides Richardson, Ichthyol. China, p. 291. 

1846. Cyprinus burgeri Richardson, /. c, p. 292. 

1846. Cyprinus abbreviatus Richardson, /. c, p. 292; Canton. 

Head 3 in length; depth 2.25; D. II, 16; A. Ill, 5; P. 15; V. 9; scales 
6-28-8; snout 3 in head; eye 5; interorbital space 2.4; pectoral 1.66; 
ventral 1.66; teeth 4-4. 

Body stout, rather high, compressed, dorsal and ventral profiles 
about equally arched; head small, its top slightly depressed; snout 
blunt; mouth small, terminal and oblique; upper jaw protractile, 
more or less projecting; eyes large, superior and anterior; nostrils very 
large, in front of eyes, the anterior in a short tube. 

Origin of the dorsal nearer the tip of snout than base of caudal, one 
scale in advance of the origin of ventral, with strong spines, soft rays 
shortening posteriorly, the first dorsal ray longest, 1.83 in length of 



Measurements of Carassius auratus. 



Locality. 



•5 
p- 

Q 



< 


PlI 


>■ 


|2 




c 


HI, 5 


15 


9 


2.40 


3 


III, 5 


16 


9 


2.33 


3 


HI, 5 


16 


9 


2.50 3 


III, 5 


16 


9 


2.33 3-25 


HI, 5 


16 


8 


2.33 


3 



5 ,6-28-8 
4.3316-29-6 
4.6016-28-8 

4-33J5-27-7 
4.75I5-28-8 



^1 



Ritakukan 

Giran 

Giran 

Giran 

Shori 



3 

3-5 

3-25 

3-45 
3 



2.25! II, 16 
2.40I1II, 16 
2.42 III, 17 
2.32 III, 16 
2.23 III, 17 



210 
150 
152 
130 
107 



head; the anal squarish, armed with very strong spines, its rays longer 
in front, inserted nearer the base of caudal than that of the pectoral, 
its origin just l)elow the fourteenth soft dorsal ray; pectoral fin round, 



The Fresh Water Fishes of the Island of Formosa. 201 

its tip scarcely reaching ventral; ventral fin not reaching the anus; 
caudal peduncle deep and long, its depth twice in head. 

Bod}' covered with large cycloid scales; lateral line nearly straight, 
extending along the middle of sides from upper part of gill-opening to 
the base of caudal. 

Color in formalin uniformly dark gray, somewhat paler below; lower 
parts whitish; all the rays uniformly gray. 

Total length 210 mm. 

Described from a specimen from Ritakukansho, Giran. 

Habitat: Abundant in the fresh waters of Formosa, very common. 

Genus Cyprinus (Artedi) Linnaeus. 
1858. Cyprinus (Artedi) Linn/EUS, Syst. Nat. Ed. X, p. 320, (Type Cyprinus 
carpio Linnaeus.) 

Body robust, compressed. Mouth moderate, anterior, with four 
long barbels. Snout blunt, rounded. Teeth molar, broad and trun- 
cate, I, I, 3-3, I, I. Scales large. Lateral line continuous. Dorsal 
fin very long, with a stout spine, serrated behind; anal fin short, also 
with a spine. Large fishes of the fresh waters of Asia. (Jordan & 
Fowler.) 

Distribution: Temperate parts of Europe and Asia; introduced into 

North America. 

17. Cyprinus carpio Linnaeus. 

Koi (Japan); Taihii (Formosa). 

1758. Cyprinus carpio Linn^us, System. Nat., Ed. X, p. 320. — Gunther, Cat. 
Fish., VII, 1868, p. 25; Europe and Asia. — Peters, Monatsb. Ak. Berl. 
1880, p. 924and 1029; Hongkong. — Sauvage, Bull. Sec. Philom., 1881, p. 7; 
Swatow. — Bull. Soc. Zool. France, IX, 1884, p. i; Tonkin. — Gunther, 
Ann. Mag. Nat. Hist. (7), 1889, p. 224; Yang-tsze-kiang. — Ann. Ac. St. 
Petersb., 1896, p. 12; Cheng-tu-fu. — ^Rutter, Proc. Acad. Nat. Sc. Philad., 
1897, p. 57; Swatow. — Gunther, Ann. Mag. Nat. Hist. (7), I, 1898, p. 261; 
Newchang. — ^Abbott, Proc. U. S. Nat. Mus. XXIII, 1901, p. 484; Pei-ho. — 
Jordan & Evermann, Proc. U. S. Nat. Mus., XXVI, 1903, p. 321; For- 
mosa. — Regan, Ann. Mag. Nat. Hist. (7), XIII, 1904, p. 191; Yunnan. — 
Jordan & Richardson, Mem. Carneg. Mus., IV, No. 4, p. 169; Formosa. — 
Jordan & Seale, Proc. Davenport Acad. Sc, X, 1905, p. 3; Hongkong. — 
Proc. U. S. Nat. Mus., XXXIII, 1906, p. 537; Buitenzorg; Java. — Jordan & 
Metz, Mem. Carneg. Mus., VI, No. 2, 1913, p. 14; Corea. — Jordan, Snyder, 
and Tanaka, Journ. Coll. Sci., Tokyo, XXXIII, 1913, p. 76; Japan. 

1798. Cyprinus rubro-fuscus Lacepede, Hist. Nat. Poiss., V, p. 530, PI. 16, 
Fig. I. — Cuv. & Val., Hist. Nat. Poiss., 1846, XVI, p. 74. — Richardson, 
Ichthy. China, 1846, p. 288. 



202 Annals of the Carnegie Museum. 

1798. Cyprinus nigroauralus Lacepede, Hist. Nat. Poiss., V, p. 547, PL 16, Fig. 2. 

— Cuv. & Val., Hist. Nat. Poiss., 1846, XVI, p. 73. — Richardson, Ichthy. 

China, 1846, p. 290. 
1798. Cyprinus viridi-violaceus Lacepede, Hist. Nat. Poiss., V, p. 547, PI. 16, 

Fig- 3- — Cuv. & Val., Hist. Nat. Poiss., XVI, 1846, p. 75. — Richardson, 

Ichthy. China, 1846, p. 288. 
1846. Cyprinus Jlavipinnis Cuv. & Val., Hist. Nat. Poiss., XVI, p. 71. 
1846. Cyprinus viltalus Cuv. & Val., Hist. Nat. Poiss., XVI, p. 72. 
1846. Cyprinus alrovirens Richardson, Ichthy. China, p. 287. 
1846. Cyprinus flammans Richardson, Ichthy. China, p. 288. 
1846. Cyprinus acuminatus Richardson, Ichthy. China, p. 289. 
1846. Cyprinus sculponeatus Richardson, Ichthy. China, p. 290. 
1846. Cyprinus hccmatopterus Schlegel, Fauna Japonica, Poiss., p. 189, PI. 96. 
1846. Cyprinus melanolus Schlegel, Fauna Japonica, Poiss., p. 190, PI. 97, 

Fig. I. 
1846. Cyprinus coniroslris Schlegel, Fauna Japonica, Poiss., p. 191, PI. 97, Fig. 2. 
1855. Cyprinus chinensis Basilewsky, Nouv. Mem. Soc. Nat. Mosc, X, p. 227, 

Tab. 2, Fig. 3. 
1855. Cyprinus obesus Basilewsky, Nouv. Mem. Soc. Nat. Mosc, X, p. 228, 

Tab. I, Fig. 2. 
1863. Car pio flavi pinna Bleeker, Atl. Ichthyol. Cyprin., p. 74, Tab. 7, Fig. 3. 
1871. Car pio vulgaris Bleeker. Mem. Cyprin. Cliina, p. 6. 

Head 3 in length; depth 2.86; D. Ill, 19; A. Ill, 5; P. 16; V. 9. 
Scales 5-35-5; width of head 1.66 in its length; snout 2.5 in head; 
interorbital space 2.4; eye 7; pectoral 1.33; ventral 1.4; teeth i, i, 3-3, 
I, I. 

Body stout, more or less compressed; head moderate, triangular; 
snout obtusely rounded; mouth oblique, with fleshy thick lips, maxil- 
lar}- reaching a vertical through anterior margin of anterior nostril ; 
maxillary protractile; upper jaw more or less projecting; four barbels, 
two maxillary and two rostral, the former nearl}' twice as long as the 
latter, scarcely reaching the anterior border of orbit; eyes moderate, 
superior and slightly anterior; nostrils large, close together, in front 
of eye, the anterior in a short tube. 

Origin of dorsal midway between tip of snout and base of caudal, 
two scales before the origin of ventral, triangular and elongate, 
higher anteriorly, its longest ray about twice in the length of head; 
pectoral fin large, round, its tip exceeding the origin of ventral; 
ventrals inserted below the first soft dorsal ray, not reaching the anus; 
origin of anal nearer base of caudal than base of ventral, inserted 
below ihc fourteenth ray of dorsal; caudal fin deeply emarginate, tip 
of lobes sharply ])ointed; caudal peduncle long and deep, its depth 2.1 1 
in head. 



The Fresh Water Fishes of the Lsland of Formosa. 203 



Body covered wilh large scales with line concentric rings and 
radiated striations; lateral line straight, extending along the middle 
of sides from upper part of gill-opening to the base of caudal. 

Color in formalin uniformly dark grey above, paler below; sides of 
body below the lateral line yellowish white; lower parts of body 
together with the pectoral, ventral, and anal white; dorsal and caudal 
fins dusky; most of scales provided with a black spot on the base. 

Total length 280 mm. 

Described from a specimen from Taihoku, collected by Oshima in 
October, 1916. 

Habitat: The present species is common throughout the island; and 
is the most important food-fish among the Chinese people. 

Measurements of Cyprinus carpio. 



Locality. 



Taihoku 

Tensonpi, Giran. 

Inzanpo 

Maruyama 



a 


a 


Q 


< 


cC 


>' 


M 






3 

c 
en 

2.50 


u 




J3 . 
ME 


?, 


2.86 


III, 19 III, 5 


16 


9 


1.66 


2.40 


7 


5-35-5 


280 


3.61 


2.86 


III, iTlni, 5 


17 


9 


1.50 


2.66 


2.76 


5 


5-34-5 


193 


3-33 


2.94 


III, 17 III. 5 


16 


9 


1-57 


2.66 


2.47 


5 


6-33-5 


132 


b.3S 


3 


III, 18 


111,5 


16 


9 


1-45 


2.45 


2.53 


6 


5-35-5 


193 



Genus Labeo Cuvier. 
1 81 7. Labeo Cuvier, Regne Animal, p. 192. (Type Cyprinus niloliciis (Forskal) 

Geoffrey). 
1842. Rohita Cuv. & Val., Hist. Nat. Poiss., XVI, p. 242. (Type Cyprinus 

nandina Hamilton.) 

Body oblong, more or less compressed. Snout broadly rounded 
or obtusely pointed, prominent, mostly covered with tubercles or 
pores and sometimes having a lateral lobe or projection, its anterior 
pendulous border forming an entire, or superficially incised, rostral 
fold. Mouth moderate or large, protractile, inferior. Lips thick, 
continuous, the upper and the lower one fringed, lobed, or not lobed. 
The lower lip is distinctly separated by a deep groove from isthmus, 
or this postlabial groove is divided by a broad or narrow connection 
between isthmus and lip, and therefore restricted to behind the 
lateral part of the lower lip or even to the corner of the mouth. Lower 
lip with an inner transverse fold. Jaws with a fleshy covering carrying 
a deciduous horny sheath. A pair of rostral and maxillary barbels, 
one of them may be absent. Eye with a free circular rim. Dorsal 
elongate, without osseous rays, commencing before ventrals and 



204 Annals of the Carnegie Museum. 

ending before or above anal, with more than eight branched rays. 
Anal short, with five branched rays. Scales moderate or small. 
Lateral line running into the middle of the tail, sensory tubes un- 
divided. Gill-membranes broadly united with isthmus. Gill-rakers 
usually short. Phar^'ngeal teeth in three series, hooked, 5, 4, 2-2, 4, 5 
(Weber & Beaufort). 

Distribution: Sumatra; Java; Borneo; Africa; India; Ceylon; 
Burma; Cochin-China; China; Formosa. 

18. Labeo jordani sp. nov. (Plate XLIX, Fig. 3). 
Kenhii (Formosa). 

1903. Rohila decora Jordan & Evermann, Proc. U. S. Nat. Mus., XXV, p. 321; 
Formosa (not of Peters). 

Head 5 in length; depth 3; D. 3, 12; A. 3, 5; P. 15; V. 9; forty scales 
in the lateral line, eight scales in an oblique series between origin of 
dorsal and lateral line, nine scales between the latter and the middle of 
belly; width of head 1.83 in its length; snout 2.33 in head; inter- 
orbital space 2; eye 5; pectoral 1.076; ventral slightly longer than 
head; teeth 5, 4, 2-2, 4, 5. 

Body compressed, rather high, abdomen rounded; head short, 
lateral parts of the occiput slightly depressed; snout obtusely rounded 
anteriorly, overhanging the upper lip, with no tubercle nor lateral 
prolongation; mouth inferior, transverse, crescent-shaped, with thick 
lips; upper lip entirely fringed, with a distinct inner fold below; lower 
lip not fringed, with an inner fold, its edge sharp and covered with a 
horny substance; barbels two, rostral, nearly one-third as long as 
snout; maxillary barbels none; eye rather small, slightly anterior and 
superior; nostrils close together, in front of the eye; pharyngeal teeth 
high, their grinding surface flat, brown-colored; gill-rakers numerous, 
minute, setiform, and closely set. 

Origin of dorsal nearer tip of snout than base of caudal, its upper 
margin concave, its base covered with a series of pointed scales, an- 
terior ray longest; anal fin entirely behind the dorsal, nearer the base 
of caudal than origin of ventral; ventral long, with well-developed 
scaly flap, inserted below fourth divided dorsal ray; pectoral fin shorter 
than ventral, reaching three-fifths of the distance to ventral; caudal 
fin bilobed, tip of each lobe sharply pointed; caudal peduncle rather 
short, its depth 1.4 in head. 

Body covered with large scales with fine* concentric rings and 



The Fresh Water Fishes of the Island of Formosa. 205 

radiated striations; lateral line nearly straight, extending along the 
middle of sides from base of the caudal to the upper part of gill- 
opening. 

Color in alcohol uniformly dark grey above, sides and lower parts 
silvery; dorsal and caudal fins grayish; other fins dusky white; scales 
on the upper part of body with a black lunar-shaped spot near the 
base. 

Total length 340 mm. 

Described from a specimen from the hatchery at Shori. 

Habitat: A species introduced from China, propagated throughout 
the island, and bred artificially in ponds. 

Remarks: The nearest relative of the present species is Labeo decorus 
Peters from Hongkong. It difi"ers, however, in having an upper lip 
which is fringed at the sides only, and a distinctly fringed lower lip. 

Jordan and Evermann recorded a Formosan species of the genus 
Labeo under the name Rohita decora {— Labeo decorus Peters). The 
descriptions given by these authors are very inadequate. But the 
briefly described characters, except the number of rays of the pectoral, 
agree quite well with those of the present species. Moreover, as 
there is no record with reference to the occurrence of any species of 
Labeo in Formosa except the introduced species, it seems reasonable 
to unite Jordan & Evermann's Rohita decora, which distinctly differs 
from Peters' Labeo decorus,'^ with the present species. 

In the year 1910, Mr. Seno, Expert of the Fisheries Bureau of the 
Japanese Government, described the present species, giving it the 
name Labeo kontius (Jordan). (Cf. Report on the Fisheries of the Island 
of Formosa). The Indian species of that name distinctly differs from 
the Formosan species in having a higher body, longer head, tuber- 
culated snout with a fleshy lateral prolongation, fringed lower lip, 
and no barbels. 

Such being the case, I propose for the present species a new name 
Labeo jordani. 

"Kenhii" is one of the important fresh-water food-fishes in For- 
mosa, though it is not a native of the island. Early in the summer, 
newly hatched larvae of this fish are collected in the rivers near Swatow, 
South China, and are imported to Formosa. At first they are bred in 
small ponds, living mainly upon zoo-plankton, the growth of which is 
encouraged by human excrement. As soon as they are large enough to 
^ Labeo decorus Peters, Monatsb. Ak. Berlin, 1880, p. 1031; Hongkong. 



206 



Annals of the Carnegie Museum. 



be safe under natural conditions (30-60 mm. long), they are sold to the 
Chinese farmers. Usually, at the end of one year they grow to the 
length of 150-160 mm. According to the Chinese breeders' informa- 
tion this fish never spawns in Formosa. 





Measurements 


OF Labeo jordani. 






! "S 

Locality. 3 

1 ■* 


V 

Q 


Q 


< 


(U 


> 


■si 

n 


o — 
c3 


3 t> 
>. 

a cd 






Shori ; 5 

Shori ! 5 


3 
3-II 


3. 12 
3. 12 


3. 5 15 
3. 5 14 


9 1 1.83 2 
9 1-83 2 


2.33 5 8-40-9 
2.50 5 8-40-9 


340 

209 



AcROSSOCHEiLUS gen. nov. 
Type Gymnostomus formosanus Regan. 

Body elongate, compressed; head smooth, with many mucous 
cavities around the eye; snout obtuse, its tip not i^rojecting beyond 
the upper lip, without lateral lobe. Mouth inferior, transverse; upper 
lip fleshy, not fringed; lower lip thick, not continuous, distinct at the 
side of the mouth only; upper jaw projecting beyond the longer; 
anterior edge of the lower jaw sharp and naked. Four barbels, two 
maxillary and two rostral. Dorsal fin without osseous ray, with not 
more than nine rays, opposite the ventral; anal fin rather short, with 
five divided rays. Scales moderate, about forty in the lateral line; 
lateral line running along the middle of the tail. Pharyngeal teeth 
5, 3, 2-2, 3, 5. 

Distribution: Formosa; China. 

Remarks: The present genus is very closely related to Crossocheilus 
Van Hasselt. Four barbels and smooth upper lip are the character- 
istics of Acrossocheilus, which distinguish it from the latter. 

19. Acrossocheilus formosanus (Regan). 
Chopicn or Choppan (Formosa). 

1908. Gymnostomus formosanus Regan, Ann. Mag. Nat. Hist. (8), I, p. 140; 

Lake Candidius, Formosa. 
1908. Gymnostomus labiatiis Regan, Ann. Mag. Nat. Hist. (8), H, p. 358; Lake 

Candidius, Formosa. 

Head 4 in length; depth 4; D. 3, 8; A. 3. 5; P. 14; \'. t); width of 
head 1.8 in its length; interorbital space 3 in head; snout 2.5; eye 5; 
pectoral 1.16; ventral 1.25; forty scales in a longitudinal series, five 
and one-half in a transverse series from origin of dorsal to lateral line^ 



The Fresh Water Fishes of the Island of Formosa. 207 

fi\e between the latter and middle of belly, three and one-half scales 
between lateral line and origin of ventral; pharyngeal teeth 5, 3, 2-2, 
3, 5; gill-rakers 5 + 9. 

Body elongate, depth equal to the length of head, dorsal profile 
more convex than the ventral; head smooth, with man}- mucous cav- 
ities around the eye; snout not projecting beyond the upper lip, with 
strongly curved profile; sides of snout pitted, in the centre of each pit 
a small round tubercle; upper jaw projecting beyond the lower; 
mouth inferior, transverse, its angle not reaching the anterior border 
of orbit; upper lip fleshy, rather thin; lower lip thick, divided into two 
lobes by a median longitudinal notch, anterior border naked, with a 
sharp inner transverse horny edge; four barbels, the rostral two-thirds 
as long as the maxillary barbels, the latter reaching the posterior border 
of orbit; eye superior and slightly anterior; nostrils close together, 
in front of eye, anterior nostril in a short tube. 

Origin of dorsal slightly in advance of a point midway between tip 
of snout and base of caudal, opposite the ventral; pectoral fin not 
reaching the ventral; ventral inserted one scale behind the origin 
of dorsal; anal entirely behind the dorsal, its rays rather long, when 
depressed, tip of the anterior ray extending beyond the others; caudal 
fin strongly forked, tip of each lobe sharply pointed. 

Body covered with uniform scales; lateral line continuous, nearly 
straight, extending along the middle of the tail. 

Color in formalin olive-brown above the lateral line, yellowish 
gray beneath; sides with seven black vertical bars; lower parts of body 
pale reddish yellow; membrane of the dorsal with black streaks; 
caudal and pectoral fins gray; other fins dusky white. 

Total length 115 mm. 

The present description is from a specimen from Shinchiku, col- 
lected by T. Aoki in December, 191 6. 

Habitat: Tamusui River at Shinten and Heirinbi; Shinchiku; Tozen 
River; Horisha; Jitsugetsutan (Lake Candidius). 

Remarks: The type of the genus Gymnostomiis Heckel is Cyprinus 
ariza Buchanan-Hamilton. Although it is provided with moderately 
large scales (thirty-seven in the lateral line), it differs distinctly from 
the species which belong to Acrossocheiliis in having two small barbels 
instead of four. 

In February, 1908, Mr. C. Tate Regan described the present species 
naming it Gymnostomiis formosaniis (from Lake Candidius). But as 



208 



Annals of the Carnegie Museum. 



it is provided with four rather long barbels, it is incorrect to include it 
in the genus Gymnostomiis. 

In October of the same year Regan described Gymnostomus labiatus 
from the same locality. According to his description it differs from 
the former in having lower lips which are separated anteriorly by a 
deep narrow notch instead of being separated widely. In my speci- 
mens which came from the type locality and other places, the width of 
the inter-space between the lower lips is not definite; even in the 
specimens which came from the same locality there are deviations 
with regard to that character. Therefore it seems inadvisable to 
separate the two forms. 

Me.\surements of Acrossocheilus formosaniis. 



Locality 



> 


n 


9 


1-73 


9 


1-75 


9 


2 


9 


2 


9 


1.50 


9 


I.9I 


9 


1.80 


9 


1. 71 


9 


1. 71 


9 


1.86 



binE 

-J 



Heirinbi 

Heirinbi 

Tamusui River 
Tamusui River 
Jitsugetsutan . . 

Horisha 

Shinchiku 

Shinchiku .... 
Tozen River . . , 
Tozen River . . , 



4-25 
4.18 

4 

4.12 

4.40 

4 

4 

4-5 

4.16 

4.14 



3-64 
3-47 
3-50 

3,- 44 

3-20 

3-74 

4 

3-86 

3.66 

3-73 



3. 5 
3. 5 
3. 5 
3. 5 
3. 5 

2, 5 

3. 5 
3. 5 
3. 5 
3. 5 



2.77! 2.40 
2.80 2.25 



3 

2.66 

2.60 

3 

3 

2-75 

3 

2.66 



2.28 
2.50 
2.50 

2.75 
2.50 
2.60 
2.77 
2.4s 



4.33 6 -42-S 
4.50i6 -42-5 
4.5oi6 -41-5 
5 [55-42-5 
5 6 -41-6 
4.66|6 -41-5 

5 j 5 2-40-5 
4. SO 1 5 -42-6 1 1 25 
4 j52-4i-5ji20 
4.83I6 -42-6I135 



120 
140 
16S 
202 
165 
105 
115 



ScAPHESTHES gen. nov. 

Body elongate, slightly compressed. Snout rounded, tip of the 
skin extends downwards and partially overlaps the upper lip. Mouth 
transverse, inferior, lower jaw with no lip, its anterior margin sharp 
and covered with a horny brown layer. Barbels two, ninute, at the 
corner of the mouth. Dorsal fin without osseous rays, with not more 
than nine branched rays, opposite to the ventrals. Anal fin rather 
short. Scales large, less than fifty in a longitudinal series. Pharyn- 
geal teeth in three rows, 5, 3, 2-2, 3, 5. Lateral line running along 
the middle of the tail. 

Distributiofi: Formosa; Hainan. 

Re7iinrks: The type of the present genus is closely related to the 
species of Scaphiodon. It differs, however, in having larger scales and 
no osseous dorsal ray. 

In the year 1899, Boulenger described a species of Cyprinoid fish 



ANNALS CARNEGIE MUSEUM, Vol, XI 



Plate L 



■'''''/,, -m^ 












Scaphesthes lamusuiensis Oshinia, sp. nov. 
Pimlius snyderi Oshima, sp. nov. 
Spiiiihayhns hoUandi Oshima. si), nov. 



The Fresh Water Fishes of the Island of Formosa. 209 

from the interior of Hainan giving it the name Gymnostoniiis leptnnis/ 
Although there is no statement with regard to the barbels it seems to 
belong to the present genus, because of the peculiar shape of the mouth, 
large scales, and the absence of an osseous dorsal spine. The barbels of 
Scaphesthes are very small, and hidden beneath the labial fold. There- 
fore, they might have been overlooked by that author. 

20. Scaphesthes tamusuiensis sp. nov. (Plate L, Fig. i). 
Kooye (Formosa). 

Head 5 in length; depth 4.5; D. 3, 8; A. 3, 5; P. 17; V. 9; width of 
head 1.87 in its length; eye 4.33 in head; interorbital space 2.5; snout 
2.6; pectoral 1.2; ventral 1.33; scales forty-six in the lateral line, seven 
in an oblique series between origin of dorsal and lateral line, six from 
the latter to the middle of belly; pharyngeal teeth 5, 3, 2-2, 3, 5; 
gill-rakers 5 + 27. 

Body elongate, slightly compressed, abdomen rounded; head 
rather small, its top more or less convex; snout obtusely rounded; 
tip of the skin extends downwards and partially o\'erlaps the upper 
lip; mouth transverse, inferior, crescent-shaped; upper lip smooth, 
without labial fold; lower jaw with no lip, mandibular edge nearly 
straight, sharp, covered with a horny brown layer; barbels four, two 
rostral and two maxillary, very minute, the latter hidden in the deep 
lateral fissure behind the angle of mouth; eyes moderate, superior 
and anterior; nostrils close together in front of eyes, the anterior nos- 
tril covered with a flap; pharyngeal teeth in three rows, those of the 
outer row canine-like, slender, slightly curved, those of the inner rows 
are smaller; gill-rakers very minute; gill-openings large. 

Dorsal fin inserted nearer tip of snout than base of caudal, armed 
with smooth soft spines, anterior ray the longest; anal fin entirely 
behind the dorsal, slightly in advance of a point midway between 
origin of ventral and base of caudal; the ventral inserted behind the 
origin of dorsal; the pectoral reaches three-fifths the distance to ven- 
trals; caudal fin deeply forked, the lobes pointed; caudal peduncle 
elongate, its depth 2.2 in head. 

Body covered with uniform cycloid scales; the ventral with scaly 
flaps, no scaly pectoral flap; lateral line continuous, extending along 
the middle of the sides, slightly decurved anteriorly. 

■• Gymnostomns lepttirus Boulenger, Proc. Zool. Soc. London, 1899, p. 961, PI. 
LXIX, Fig. i; Hainan. 

15 DEC. 17, 1919. 



210 



Annals of the Carnegie Museum. 



Color in formalin grayish brown above, paler below the lateral 
line, lower surface silvery; dorsal and caudal gray, outer margin of 
the former with a series of black streaks; pectorals fuscous; other 
fins whitish. 

Total length 230 mm. 

Described from a specimen from Tamusui River near Shinten, 
collected by T. Aoki in December, 1916. 

Habitat: Tamusui River (Shinten, Heirinbi); Choso River; Giran 
River (Inzanpo, Kiburan). 

Remarks: Scaphesthes leptiinis from Hainan is closely related to the 
present species. It differs, however, in having a larger eye, the dorsal 
inserted midway between the end of snout and base of caudal, the 
ventral below middle of dorsal, six divided anal rays, and forty-nine 
scales in the lateral line. There is no statement with reference to the 
barbel in Boulenger's original description. 

Measurements of Scapheslhes tamusuiensis. 



Locality. 



Ph' 


> 





0— • 


1 

c 


V 


17 


9 


1.83 


2.50 


2.40 


4.8 


16 


10 


1.60 


2.50 


2.60 


4-. "5 


16 


10 


1-75 


2.33 


2.83 


4.66 


17 


10 


2.21 


3 


3-26 


4 


17 


10 


1.58 


2.50 


2.85 


4 


16 


10 


1.78 


2.66 


2.66 


3-66 


15 


10 


1.66 


2.50 


3 


3-66 


16 


10 


1.78 


2.40 


3 


3-66 






Tamusui River 
Tamusui River 
Tamusui River 
Choso River. . 
Giran River. . . 

Heirinbi 

Taishu 

Taishu 



5 

4-75 

4-75 

4.71 

4.42 

4.20 

4-33 
4.41 



4-50 
4.16 
4-25 
3-84 
4.20 
4.20 
4.64 
4.42 



3.8 
3.8 
3.8 
3.8 
3.8 
3.8 
3.8 
3.8 



3. 5 
3. 5 
3. 5 
3. 5 
3. 5 
3. 5 
3. 5 
3. 5 



6I-46-6I230 
7 -47-6 1 83 
65-47-61200 
7 -47-6:255 
7 -47-6,102 
6 -47-6 1 80 
6 -4 7-6 1 78 
6 -47-6' 90 



Genus Hemibarbus Bleeker. 

1861. Hemibarbus Bleeker, Prodr. Cyprin., p. 281. (Tj-pe Gobio barbiis Tem- 

minck & SchlegeL) 
1869. Gobiobarbus Dybowski, Verb. Zool.-Bot. Gesell. Wien, XIX, p. 951. (Type 

Cyprinus labeo Pallas.) 

Body elongate, rather slender, and compressed. Head elongate, 
somewhat pointed, and wath many mucous cavities about the eyes 
and along the edge of the pre-operculum; snout long, blunt at the tip; 
eye rather large, high; mouth inferior, the maxillar}- not reaching eye; 
lip fleshy; each maxillary with a barbel as long as eye; teeth 5, 3, i-i, 
3, 5. Gill-rakers short; intestine short. Peritoneum silvery; scales 
cycloid, about forty-nine. Dorsal inserted nearer tip of snout than 
base of caudal, and armed with a slender, sharp, strong, and smooth 



The Fresh Water Fishes of the Island of Formosa. 211 

spine; anal inserted far behind tip of compressed dorsal; caudal deeply 
emarginate, the lobes pointed; ventrals inserted behind origin of the 
dorsal. Lateral line slightly deciirved and continuous. (Jordan & 
Fowler.) 

Distribution: Formosa; China; Corea; Amur Province; Japan. 

21. Hemibarbus labeo (Pallas). 
Migoi (Japan); Tekotau (Formosa). 

1776. Cyprinus labeo Pallas, Reise III, p. 207, 703; Onon. — N. Acta Acad. 

Petrpol., I, 1787, p. 355. T. XI, figs. 8, 9; Onon; Ingoda; Schilka. — Zoogr. 

Ross. Asiat., Ill, 1811, p. 305; Dauria; Ingoda; Onon; Schilka. 
1842. Gobio barbus Schlegel, Fauna Japonica, Poiss., p. 198, PI. XCIX, Fig. i; 

Nagasaki. 
i860. Hemibarbus barbus Bleeker, Prods. Cyprin., p. 281. — Jordan & E\'er- 

MANN, Proc. U. S. Nat. Mus., XXV, 1902, p. 322; Formosa. — ^Jordan & 

Fowler, Proc. U. S. Nat. Mus., XXVI, 1903, p. 824; Japan (Yodo River; 

Lake Janzabrobata; Aomori; Chikugo River; Tokyo). 

1868. Barbus schlegeli GiJnther, Cat. Fish., VII, p. 135; Formosa; Japan. 

1869. Gobiobarbus labeo Dybowski, Verh. Zool.-Bot. Gesell. Wien, XIX, p. 951, 
T. XV, Fig. 3; Onon; Ingoda. 

1892. Barbus schlegeli Steindachner, Denkschr. Akad. Wien, LIX, p. 370; 

Seoul, Korea. 
1896. Acanlhogobio giinlheri GtJNTHER, Ann. Mus. Zool. St. Petersb., I, p. 215; 

Hui-hsien; Huang-ho; Sinin River, China. 
1904. Acanlhogobio oxyrhynchus Nilolsky, Ann. Mus. Zool. St.. Petersb., VIII, p. 

358; Ussuri. 
1907. Barbus labeo Berg, Ann. Mus. Zool. St. Peterb., XII, p. 3; Corea. 
1909. Hemibarbus labeo Berg, Ichthyol. Amus., p. 75. — Jordan & Metz, Mem. 

Carneg. Mus., VI, no. 2, 1914, p. 15; Corea. 

Head 3.44 in length; depth 4.83; D. Ill, 6; A. Ill, 6; P. 20; V. 9: 
width of head 2 in its length; interorbital space 3.75 in head; eye 5; 
snout 2.15; pectoral 1.38; ventral 1.91; forty-nine scales in the lateral 
line, seven scales in an oblique series between origin of dorsal and lateral 
line, six scales from the latter to the middle of belly, four scales be- 
tween lateral line and the root of ventral; pharyngeal teeth 5, 3, i-i, 
3, 5; gill-rakers 7 + 10. 

Body elongate and compressed; head elongate, pointed, its top 
compressed, with many mucous cavities around the eye and below 
and behind the pre-operculum; snout long, pointed and produced; 
eyes large, superior and slightly posterior; mouth inferior, with fleshy- 
lips, its angle not reaching the orbit; upper jaw protractile, longer 
than the lower; barbels two, maxillary, slender, as long as the diameter 



212 



Annals of the Carnegie Museum. 



of eye; nostrils together, in front of the eye, much nearer the eye 
than the tip of snout, anterior nostril in a short tube; interorbital 
space broad and flattened; gill-openings large; gill-rakers short and 
fleshy. 

Dorsal fin inserted nearer tip of snout than base of caudal, rays 
straight and strong, longer in front, when depressed first ray reaches 
beyond the others, spine straight, smooth, and long; pectorals reach 
three-fourths the distance to ventrals; origin of ventral behind that 
of dorsal, about midway between tip of snout and base of caudal; the 
anal entirely behind the dorsal, inserted midw^ay between origin of 
ventral and base of caudal; caudal fin deeply emarginate, the tip of 
each lobe pointed; caudal peduncle elongate, its depth 3 in head. 

Scales large, of more or less uniform size, cycloid; pectoral with 
scaly flap, ventral flap moderate, pointed; lateral line continuous, 
anterior part slightly decurved. 

Color in formalin grayish brown above, paler below; sides and 
lower parts silvery, with no dark spots; dorsal and caudal fins pale 
gray, the rest of flns whitish. 

Total length 290 mm. 

Described from a specimen from Tamusui Ri\er, collected by T. 
Aoki in December, 1915. 

Habitat: Tamusui River near Shintcn and Heirinbi; Rigyokutsu, 
Nanto. 

Remarks: The present species is very closely related to Ilcniibarbiis 
maculatiis from China, differing from it only in color. Color of H. 
maculatiis in alcohol (Stanford Collections No. 8414; Pei-ho, China, 

Measurements of Hemibarbus laben. 



Locality. 



o — 

c'-2 



So 2 



Tamusui River. 
Tamusui River. 
Rigyokutsu . . . . 
Heirinbi 



3-80 
3-50 
3-50 
3-31 



4.83 III, 7 III, 6 20 

5 {III, 7 III, 6! 20 

4.33111, 7 III, 6 20 

4.66'III, 7lin, 61 20 



2.50 
2.33 



3-75 
3-40 
4 
3-66 



2.1S 
2. II 
2.25 
2.42 



5 7-49-6290 
4.50J 7-49-6 260 
5 7-48-6 265 
3.87I7-48-6I128 



290 mm. long; described by Jordan and Starks under the name; 
Uemibarbiis joiteni) pinkish yellow, with a longitudinal series of eight 
large spots above the lateral line; smaller spots irregularly placed 
on back and sides; dorsal and caudal with similar black spots; other 
fins without markings. Although faint dark sjiols are present in the 



The Fresh Water Fishes of the Island of Formosa. 213 

young specimen of Ilemibarbns labeo, they are not j^ermanent; the 
color of the adult is always uniformly grayish brown. 

Genus BARBODES Blacker. 

i860. Barbodes Bleeker, Nat. Tijdschr. Ned. Ind., XX, p. 431. (Type Barbodes 
belinka Bleeker.) 

Body strongly compressed, more or less elevated, with the profile of 
the back arched; head of moderate size, its length being one-fourth 
of the length without caudal; snout shorter than the postorbital part. 
Mouth subinferior, arched, without inner fold. Barbels four,' two 
maxillary and two rostral, rather short. Scales of moderate size, 
about forty in the lateral line. Lateral line continuous, running along 
the middle of the sides. Dorsal fin with three osseous spines, the 
third the longest, inner border of which is coarsely serrated or smooth; 
with not more than nine branched rays, inserted above or a little 
behind the origin of the ventral. Pharyngeal teeth 4, 3, 2 — 2, 3, 4. 

Distribution: Philippine Islands; Malay Archipelago; India to 
South China; Formosa. 

22. Barbodes paradoxus (Giinther). 

1868. Barbus paradoxus Gijnther, Cat. Fish., VII, p. 97. Formosa. 

Head 4 in length; depth 3.5; D. Ill, 9; A. 2. 6; P. 15; V. 9; width of 
head 1.86 in its length; interorbital space 2.66 in head; snout 3; eye 
3; pectoral 1.18; ventral 1.20; thirty-nine scales in the lateral line, 
six scales in an oblique series between origin of dorsal and lateral line, 
six scales between the latter and the middle of the belly, three scales 
from lateral line to the root of the ventral; pharyngeal teeth, 4, 3, 2 — ■ 
2, 3, 4; gill-rakers 4-1-9; branchiostegals 3. 

Body oblong, slightly compressed; abdomen rounded; head mod- 
erate, its top convex; snout shorter than postorbital part, obtusely 
pointed anteriorly, its dorsal profile rounded; mouth subinferior, 
arched, its angle not reaching the orbit; lips fleshy, lower lip distinct 
only near the angle of mouth; lower jaw shorter than the upper, its 
tip naked; four barbels, two maxillary and two rostral; eyes moderate, 
slightly superior and anterior; nostrils close together, in front of eye, 
the anterior in a short tube. 

Origin of dorsal midway between the tip of snout and base of 
caudal, opposite the root of ventral; inner border of third spine not 
serrated, its length 1.71 in head; pectorals not reaching the ventral; 



214 Annals of the Carnegie Museum. 

ventral beneath the dorsal, not reaching the anal; anal rather short, 
anterior ray the longest; caudal fin emarginate, tip of each lobe pointed. 

Scales moderate, cycloid; lateral line nearly straight, extending 
along the middle of sides, very slightly decurved. 

Color in formalin dark gray above, lower parts of sides dusky yellow; 
belly whitish; sides with seven dark brown cross-bars, of which the third 
and fifth reach the back, the others shorter; membrane of the dorsal 
with a black streak between each ray; caudal fin dusky; other fins 
whitish. 

Total length 6i mm. 

The present description is from a specimen from Taiko River, 
collected by T. Aoki in December, 1916. 

Habitat: Taiko River (a single specimen, young fish). 

Remarks: In Giinther's original description there is no statement 
about the dark cross-bars on the sides. Probably they had disap- 
peared, as his specimens from Formosa were adults, measuring from 
eight to eight and one-half inches in length. 

Genus Capoeta Cuv. & Val. 

1842. Capoeta Cuv. & Val., Hist. Nat. Poiss., XVI, p. 278. (Tjpe Capoeta 
fundulus Cuv. & Val.) 

Body elongate, compressed, rather deep; head moderate; snout 
somewhat pointed, nearly as long as the eye. Mouth narrow, sub- 
inferior; upper jaw slightly overlapping the lower; barbels two, maxil- 
lary, slender. Scales large, about twenty-five in the lateral line. 
Dorsal fin with three osseous spines, the third strongly serrated behind; 
eight branched rays, fin inserted in advance or opposite to the origin 
of the ventral; anal fin rather short. Pharyngeal teeth 5, 3, 2 — 2, 3, 5; 
lateral line running along the' middle of the sides. 

Distribution: Malay Archipelago, India to South China. 

23. Capoeta semifasciolata (Giinther). 

Anbakutai (Formosa). 

1868. Barbus fasdolalus GDnther, Cat. Fish., VII, p. 140; China (not of page 108). 

1868. Barbiis semifasciolalus Gxjnther, Cat. Fish., VII, p. 484; China (substitute 

for B. fasciolatus of page 140; B. fasciolattis pre-occupied). 
1 87 1. Puntius {Capoeta) giintheri Bleeker, Mem. Cyprin. Chine, p. 9 (substitute 
for B. fasciolatus Giinther of page 140, /. c). 

Head 3.66 in length; depth 2.8; D. Ill, 8; A. 2, 6; P. 13; \'. 9; width 
of head 1. 71 in its length; eye 3 in head; interorbital space 2.33; 



The Fresh Water Fishes of the Island of Formosa. 215 

snout 3; pectoral 1.375; ventral 1.33; twenty-five scales in the lateral 
line; four scales in an oblique series from origin of dorsal to lateral line, 
two between the latter and the root of ventral, four scales between 
lateral line and the middle of belly; pharyngeal teeth 5, 3, 2 — 2, 3, 5; 
gill-rakers rudimentary. 

Body elongate, compressed, rather deep; head moderate, top more 
or less convex; snout short, obtusely rounded anteriorly; mouth subin- 
ferior, arched, its angle not reaching the orbit; lip fleshy; lower jaw 
slightly shorter than the upper; barbels two, maxillary, slender; eyes 
moderate, anterior; nostrils close together, in front of eye, the anterior 
in a short tube. 

Origin of dorsal midway between the tip of snout and base of caudal, 
nearly opposite that of the ventral, first spine minute, third spine 
strongly serrated behind, its length 1.66 in head; pectoral fin not 
reaching the ventral; anal short, anterior ray the longest; caudal fin 
emarginate, the tip of each lobe pointed. 

Scales rather large; base of the dorsal and anal provided with scaly 
sheath; ventral flap present, scaly; no pectoral flap; lateral line con- 
tinuous, extending along the middle of the sides, slightly decurved. 

Color in formalin pale gray above, paler below; belly whitish; 
sides with about seven short black cross-bars, none of them reaching 

Measurements of Capoeta semifasciolata. 



Locality. 



c'2 






Ako 
Ako 
Ako 



3-66 
3-33 
3-25 



2.80 III, 8 
2.86JIII. 8 
3 IlII. 8 



2,6 
2.5 
2. S 



1. 71 
1. 71 
1. 71 



2.33 
2.40 
2.60 



3.00 
3-00 
3-25 



4-25-4 
4-24-4 



3.254-24-4 



the back or belly; a round black spot on the end of lateral line; back 
with numerous small black spots; a large dark spot on the top of the 
head; an obscure black longitudinal streak along the median dorsal 
line; dorsal and caudal fins dusky, the rest of fins whitish. 

Total length 51 mm. 

Described from a specimen from Ako, collected by T. Aoki in 
December, 1916. 

Habitat : Ako (nine specimens). 



216 Annals of the Carnegie Museum. 

Genus Puntius Hamilton. 
1822. Puntius Hamilton, Fishes of the Ganges, p. 388. (Type Cyprinus punlio 
Hamilton.) 

Body oblong, rather high, slightly compressed. Mouth anterior 
and oblique, with no labial fold; lower lips continuous, without horny 
covering. Barbels none. The osseous dorsal spine of moderate 
strength, smooth or serrated behind, the fin inserted in advance of- 
or a little behind, the origin of the ventral; anal with five or six divided 
rays. Scales large, less than thirty in the lateral line. Lateral line 
continuous, extending along the middle of the tail. Pharyngeal teeth 
in three series, 5, 3, 2 — 2, 3, 5. 

Distribution: Malay Archipelago; British India. 

Remarks: Dr. Bleeker restricted the present genus to include a 
group of fishes, the tyi)e of which is Cyp-rinus sophore Hamilton, but 
the type of the genus Puntius has no barbel, thus being distinguished 
from C. sophore which has four barbels. 

24. Puntius snyderi sp. nov. (Plate L, Fig. 2). 
Anbakutai (Formosa). 

Head 3.58 in length; depth 3; D. IV, 9; A. 2, 6; P. 13; V. 9; width 
of head 1.7 in its length; eye 3 in head; interorbital space 2.66; snout 
3; pectoral 1.42; ventral 1.42; twenty-four scales in the lateral line, 
four scales in an oblicjue series between origin of the dorsal and lateral 
line, four scales between the latter and the middle of belly, three 
scales between lateral line and the root of the ventral; pharyngeal 
teeth 5, 3, 2 — 2, 3, 5; gill-rakers i + 4- 

Body oblong, slightly compressed, abdomen rounded; head moder- 
ate, its top more or less convex, profile on the nape slightly concave; 
snout rather short, anterior margin obtusely rounded; mouth anterior 
and oblique, its angle not reaching the orbit; lower jaw slighth- shorter 
than the upper; lips fleshy; no barbel; nostrils close together, nearer 
than orbit the tip of snout, the anterior nostril in a short tube; eyes 
moderate, slightly anterior and superior; pharyngeal teeth sharp and 
hooked; gili-rakers short, rudimentary. 

Origin of the dorsal midway between tip of snout and base of caudal, 
slightly behind that of the ventral, first and second spines minute, 
third s[)ine about one-third as long as the fourth, which is the strongest 
and is serrated behind; tiie pectoral slender, not reaching the ventral; 
ventral fin inserted in front of origin of the dorsal; the anal short, 



TiiF, Fresh Water Fishes of the Island of Formosa. 217 



entirely behind the dorsal; caudal fin emarginate, each lube sharply 
pointed. 

Scales rather large, cycloid; base of the dorsal fin covered with 
scaly sheath; lateral line continuous, very slightly decurved, extending 
along the middle of the tail. 

Color in formalin dark gray above, paler below; sides with three 
black cross-bars, a black spot near the base of the caudal; above the 
gill-opening a short dark brown streak; a brown semilunar spot on 
the occiput; all the fins uniformly dusky white. 

Total length 77 mm. 

The present description is from a specimen from Rigyokutsu, 
Nanto, collected by T. Aoki in December, 1916. 

Habitat: Rigyokutsu, Nanto; Maruyama near Taihoku; Daito 

River. 

Measurements of Puntius snyderi. 



Locality. 



Rigyokutsu. 
Rigyokutsu. 
Maruj-ama. . 
Daito River. 
Daito River. 
Daito River. 







S 


Q 


3.58 


3 


3.60 


3 


3-50 


3 


3.16 


3 


3-50 


3 


3-50 


3 



IV, 9 2, 6 
IV, 8' 2, 6 
IV, 8' 2, 5 
IV, 8: 2, 5 
IV, 8 2, 5 
IV, 8 2, 5 



oi 


> 


13 


9 


13 


9 


13 


9 


13 


9 


13 


9 


13 


9 






1.70 2.66 
1.87! 2.66 
1.83 2.50 
2 2.60 

i.86| 2.80 
1.83 2.60 



3 

3-33 

3 

3-5 

3-4 

3-25 



3 

3-33 

3 

3-25 

3-4 

3-75 



So S 



4 -24-4 77 
3I-24-4 66 
3 1-2 4-4 43 
32-23-4 48 
3^23-4 1 58 
4-24-45' 50 



Genus Spinibarbus gen. nov. 

Body elongate, more or less compressed, abdomen rounded. Dorsal 
fin short, slightl}' in advance of the origin of ventral, with eight 
branched rays and three unserrated osseous spines, the third spine 
stronger than the others. Anal fin rather short, with five branched 
rays. Mouth somewhat inferior, with the margin of the lower jaw 
obtuse; lips without inner fold. Four barbels the maxillary much 
longer than the rostral. A recumbent spine in front of the dorsal, 
pointing forward, its basal portion hidden by the scales. Scales 
large; lateral line running along the middle of the tail. Gill-openings 
extend to below the angle of operculum. Gill-rakers short and fleshy. 
Pharyngeal teeth 5, 3, 2 — 2, 3, 5. 

Distribution: Formosa. 

Remarks: Although the type of the present genus agrees very weU 
with Gunther's Mystacolencus (Gunther, Cat. Fish. Brit. Mus., \TI, 
1868, p. 206), it differs remarkably from it in having unserrated dorsal 



218 Annals of the Carnegie Museum. 

spines, a less number of divided anal rays, pharyngeal teeth of 5, 3, 2 — 

2, 3, 5 instead of 4, 3, 2 — 2, 3, 4, and large scales. 

25. Spinibarbus hoUandi sp. nov. (Plate L, Fig. 3; Plate LI, Fig. i). 

Head 3.8 in length; depth 4; D. Ill, 8; A. II, 5; P. 16; V. 9; width 
of head 1.85 in its length; eye 7 in head; interorbital space 3, snout 3; 
pectoral 1.25; ventral 1.33; twenty-six scales in the lateral line, four 
in an oblique series between origin of dorsal and lateral line, four 
between the latter and the middle of belly, three scales between lateral 
line and the root of the \'entral; pharyngeal teeth 5, 3, 2 — 2, 3, 5; 
gill-rakers 4-1-9. 

Body elongate, more or less compressed; head elongate, pointed, 
top compressed, tith many mucous cavities around the orbit; snout 
long, pointed; eyes rather small, superior and anterior; nostrils close 
together, in front of eye above; mouth inferior, its angle not reaching 
the orbit; lips flesh}', rather thin; upper jaw very slightly longer than 
the lower; barbels four, the rostral reaching the orbit, much slenderer 
and shorter than the maxillary which is nearly as long as the snout; 
interorbital space broad and flattened; gill-openings moderate; gill- 
rakers short and flesh}-. 

Dorsal fin inserted nearer tip of snout than the base of caudal, 
spines smooth, not serrated, anterior dorsal ray the longest; a recum- 
bent spine in front of the dorsal, pointing forwards, its basal portion 
hidden by the scales; the pectoral reaching beyond two-thirds the 
distance to ventral; origin of ventral one scale behind that of the dor- 
sal; anal fin entirely behind the dorsal, inserted midway between 
origin of ventral and base of caudal; caudal fin deeply forked, the tip 
of each lobe pointed; caudal peduncle elongate, its depth 2.5 in head. 

Body covered with large cycloid scales; ventral fin with scaly flaps; 
lateral line continuous, slightly decurved anteriorly, extending along 
the middle of sides. 

Color in formalin grayish brown above, paler below; sides and 
lower parts silvery; edge of the dorsal fin with a series of dark streaks, 
the rest of the fins dusky. 

Total length 340 mm. 

Described from a specimen from Sobun River near Tabani, col- 
lected by T. Aoki in December, 191 6. 

Habitat: Sobun River near Tabani (four specimens). 

Remarks: Named for Dr. \V. J. Holland, Director of the Carnegie 
Institute, Pittsburgh, U. S. A. 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate LI. 




^ 







^^^iS''^ 



^^ 





Recumbent spine of Spinibarbus hoUandi. 
Gnathopogon iijimce Oshima, sp. nov. 
Phoxiscus kikuchii Oshima, sp. nov. 



The Fresh Water Fishes of the Island of Formosa. 219 



Measurements of Spinibarbus hollanei. 



Locality. 


•T3 
V 


J3 
S. 

Q 


Q 


< 


CL," 


> 


o 


S3 


3 
O 

c 










3-8o 
3-85 


4 


III. 8 
III. 8 


2, 5 
2. 5 


i6 
i6 


9 
9 


1.85 
1.66 


3 
2.66 


3 
3 


7 
6 


4-26-4 
4-2 7-4 


340 




2l<, 











Genus Gnathopogon Bleeker. 
i860. Gnathopogon Bleeker, Ichth. Archipel. Indie. Prodr., II, p. 434. (Type 

Capoela elongata Temminck & SchlegeL) 
1872. Sqnalidus Dybowski, Verb. ZooL-Bot. Ges. Wien, XXII, p. 215. (Type 

SqtiaUdus chankxnesis Dybowski.) 
1896. Leiicogobio Gunther, Ann. Ac. Sci. Petersb., p. 212. (Type Leucogobio 

hersensteini Giinther.) 
Body elongate, compressed; abdomen not carinated. Scales of 
moderate size; lateral line continuous, running along the middle of the 
tail. Mouth anterior and oblique, with a minute maxillary barbel at 
the corner; both jaws with simple, narrow lips. Dorsal fin short, 
without spine, inserted in front of, or behind, that of the ventral. 
Anal fin short, with not more than six branched rays. Gill-rakers 
rudimentary; pharyngeal teeth 5, 3 or 2 or i — i or 3, 5, slightly hooked. 
Distribution: Formosa; China; Corea; Japan; Amur Provinces. 

26. Gnathopogon iijimae sp. nov. (Plate LI, Fig. 2). 

Head 3.70 in length; depth 4.85; D, 3, 7; A. 2, 6; P. 16; V. 7; width 
of head twice in its length; eye three times in head; interorbital space 
3.6; snout 3; pectoral 1.3, thirty-three scales in the lateral line, four 
scales in an oblique series between origin of dorsal and lateral line, 
four scales between the latter and the middle of belly; pharyngeal 
teeth 5, 3 — 3, 5; gill-rakers rudimentary; five branchiostegals. 

Body elongate, compressed, abdomen not carinated; head mod- 
erate, its top more or less convex; snout pointed anteriorly, its tip 
swollen, interorbital space rather flat, with bony ridge along superior 
margin of the orbit; mouth oblique, with very thin lips; upper jaw 
longer than the lower; barbels two, maxillary, minute, about half as 
long as the diameter of eye; eyes large, superior and slightly anterior; 
nostrils close together, nearer to eye than to tip of snout. 

Origin of the dorsal slightly nearer the tip of snout than the base of 
caudal, first simple ray very short, the second about half as long as the 
third, anterior divided ray the longest, nearly as long as the head: 



220 Annals of the Carnegie Museum. 

pectoral scarcely reaching the base of ventral; \entral fin opposite the 
dorsal, inserted beneath the first branched dorsal ray; anal entirely 
behind the dorsal, anterior ray the longest; caudal fin deeply emargi- 
nate, tip of each lobe sharply pointed; the depth of caudal peduncle 
2.83. 

Body covered with thin cycloid scales; lateral line continuous, 
extending along the middle of the tail, slightly decurved. 

Color in alcohol pale yellowish gray above, lower parts whitish; 
most of the scales speckled with black; top of the head rather dark; 
dorsal fin white, each ray with black spots; caudal fin dusky, spotted 
with black; other fins whitish. 

Total length 79 mm. 

Described from a specimen from Tozen River, collected by T. 
Aoki in December, 1916. 

Habitat: Tozen River (a single specimen). 

Remarks: Named for Prof. Isao lijima of the Science College, 
Imperial University of Tokyo. 

This species resembles Gnathopogon coreanus (Berg)^ from Corea, 
but has shorter barbels and thirty-three scales instead of thirty-five 
in the lateral line. 

Genus PsEUDOGOBio Bleeker. 

1863. Pseudogobio Bleeker, Atlas Ichthyol., Cyprin., p. 29. (Type Gobio 
esocinns Temminck & Schlegel.) 

Body elongate, rather slender and tapering behind. Head elon- 
gate, pointed; snout long, concave above, and slightly produced, with 
its tip bluntly rounded; eye small, nearer posterior edge of opercle 
than tip of snout; mouth small, protractile downwards, inferior, the 
maxillary not reaching nostrils: lips broad, fleshy, and covered with 
well-developed papillae; a rather short, thick maxillary barbel; teeth 
small, 6 or 5, 2-2, 5 or 6; interorbital space broad and concave. Intes- 
tine short. Peritoneum silvery. Scales moderate, cycloid, about 
forty-two. Origin of the dorsal nearer tip of snout than base of 
caudal; origin of anal far behind tip of depressed ventral; caudal 
emarginate; ventrals inserted well behind origin or dorsal. Lateral 
line almost straight or very slightly decur\ed and continuous. Dorsal 
and caudal with distinct narrow blackish cross-bands. (Jordan and 
Fowler.) 

Distribution: Formosa; China; Jai)an. 

^ Leiicogobio coreanus Berg, Ann. Mag. Nat. Hist. (7), 1906, p. 394; River Sam- 
bau, Kyong-sang-do, Corea. 



The Fresh Water Fishes of the Island of Formosa. 221 
27. Pseudogobio brevirostris (".iiniher. 

1868. Pseudogobio brevirostris Gunthkr, Cat. Fish., VII, p. 174; Formosa. 

Head 4.44 in length; depth 5; I). 2, 7; A. 2, 6; P. 13; V. 8; width of 
head 1.5 in its length; eye 4 in head; interorbital space 3; snout 2; 
ventral 1.28; pectoral slightly longer than the head; thirty-eight 
scales in the lateral line, four and one-half scales in an oblique series be- 
tween origin of dorsal and lateral line, five scales between the latter 
and the middle of belly; between lateral line and the root of the ventral 
two scales; pharyngeal teeth 5-5; twenty-seven gill-rakers on first 
arch. 

Body elongate, not compressed, dorsal profile nearly straight, 
abdomen rounded; head squarish, its top flat; snout pointed anteriorly, 
suddenly depressed in front of the nostrils, tip slightly swollen; 
mouth inferior, suctorial; upper lip distinctly fringed, lower lip 
densely tuberculated; each jaw with a transverse, sharp horny edge; 
eyes moderate, superior, and a little posterior; nostrils close together, 
anterior nostril in a short tube; interorbital space flat; pharyngeal 
teeth hooked; gill-rakers quite short, set very closely. 

Origin of the dorsal much nearer the tip of snout than the base of 
caudal, in advance of that of the ventral, its longest ray nearly as long 
a§ the head; pectoral fin horizontal, extending beyond the origin of 
dorsal, middle rays longer; the ventral inserted beneath the middle 
of base of dorsal; anal fin moderate, second branched ray the longest; 
caudal fin emarginate, tip of each lobe obtusely pointed. 

Scales rather large, thin, with radiated striae and concentric rings; 
lateral line continuous, nearly straight, extending along the middle 
of the sides. 

Color in alcohol pale grayish brown above; lower part of sides and 
belly whitish, with a longitudinal black band along the lateral line, 
all the fins whitish, roughly speckled with brown; top of head uniformly 
gray. 

Total length 92 mm. 

The present description is from a specimen from Tamusui River 
near Shinten, collected by T. Aoki in December, 1916. 

Habitat: Tamusui River (four specimens). 

Remarks: The nearest relative of the present species is Pseudogobio 
sinensis'' from China. It differs, however, in having fewer scales in 

* Tylognathus sinensis Kner, Novara Fisch., Ill, 1865, p. 354; Shanghai, China. 
Pseudogobio sinensis Giinther, Cat. Fish., VII, 1868, p. 175; Shanghai (after 
Kner). 



222 



Annals of the Carnegie Museum. 



the lateral line (36-37) and six scales between origin of dorsal and 

lateral line. 

AIeasurements of Pseudogobio breviroslris. 



Locality. 


13 

s 


0. 


Q 


< 


0.' 


>" 


Width 
of Head. 

Interor- 
bital. 


3 


c 




Scales. 

Length, 
Mm. 


Tamusui River 

Tamusui River 


4.44 

5 


5 
4.64 


2, 7 
2, 7 


2, 6 
2,6 


13 
13 


8 
8 


1-50 3 
1.42 3 


2 
2 


4 :4|-38-5 92 
4 I4I-39-5 94 



Genus PsEUDORASBORA Bleeker. 

i860. Pselidorasbora Bleeker, Act. Soc. Indo-Neerl., \'I, p. 97. (Type Leuciscus 
parvus Temminck & Schlegel.) 

Body elongate. Head pointed, compressed; snout bluntly pointed; 
eye rather large; mouth terminal, above, oblique, and mandible pro- 
jecting and the maxillary not reaching nostril; no barbels; teeth 5-5: 
interorbital space broad and flat. Inside of gill-openings with a 
notch below. Intestine short. Peritoneum silvery. Scales large, 
cycloid, and about thirty-eight in lateral line; breast scaled. Origin 
of dorsal nearer tip of snout than base of caudal; origin of anal begins 
below origin of the depressed dorsal; caudal emarginate; ventrals 
inserted below origin of dorsal. Lateral line slightly decurved and 
continuous. Breeding males with the snout and sides of the head 
with horny tubercles (Jordan and Fowler). 

Distribution: Formosa; China; Amur Province; Corea; Japan. 

28. Pseudorasbora parva (Schlegel). 
Moroko or Haya (Japan); Chasui or Bohoe (Formosa). 

1846. Leuciscus parvus Schlegel, Fauna Japonica, Poiss., p. 215, PL CI I, Fig. 3; 
streams near Nagasaki. 

1846. Leuciscus pusillus Schlegel, /. c, p. 216, PL CII, Fig. 4. 

1867. Pseudorasbora parva Kner, Novara, Fisch., Ill, p. 355, PI. XVII, Fig. 2; 
Shanghai. — Gunther, Cat. Fish., VII, 1868, p. 186; Japan; China. — 
Bleeker, Mem. Cyprin. Chine, 1871, p. 11; Slianghai; Tji-kiang. — GiiN- 
ther, Ann. Mag. Nat. Hist., Sept., 1873, p. 247; China. — Ishik.a.wa, 
Z06I. Mag. Tokyo, Vll. 1875, P- 128; Otsu; Maebara; Matsubara. — Peters, 
Monatsb. Ak. Berlin, 1880, p. 925; China. — Jordan & Snyder, Proc. U. S. 
Nat. Mus., XXIII, 1900, p. 344; Lake Biwa. — Jord.4.n & Fowler, Proc. 
U. S. Nat. Mus., XXVI, 1903, p. 840; Tsuchiura; Nagoya; Lake Yogo; 
Lake Biwa; Iwai River; Chikugo River; Yodo River. — Smith & Pope, 
Proc. U. S. Nat. Mus., XXXI, 1905, p. 461; Japan. — Jordan & Metz, 
Mem. Carncg. Mus., VI, No. 2, 1913, p. 16; Suigen, Corea. — Berg. IchthyoL 
Amur, ]:>. 94; Amur Province 



The Fresh Water Fishes of the Island of Formosa. 223 



Head 4 in length; depth 4; I). 3, 7; A. 2, 6; P. 14; V. 8; width of 
head 1.78 in its length; eye 3.66 in head; interorbital space 2.5; snout 
3; pectoral 1.25; ventral 1.25; thirty-eight scales in the lateral line, 
five scales in an oblique series between origin of dorsal and lateral line, 
five scales between the latter and the middle of belly, three scales 
between lateral line and the root of ventral; pharyngeal teeth 5-5; gill- 
rakers rudimentary, merely fleshy rudiments. 

Body elongate, compressed; head triangular, pointed, compressed, 
upper surface more or less compressed, interorbital space rather 
broad; snout truncated in front, anterior part very slightly swollen; 
mouth anterior, transverse, its angle not reaching the nostril; upper 
lip rather thick and fleshy; lower jaw projecting, with horny anterior 
edge; eyes large, slightly anterior and superior; nostrils together, in 
front of eye above;' no barbel; gill-openings moderate; peritoneum 
silvery. 

Origin of the dorsal nearer tip of snout than base of caudal, one 
Scale in advance of that of the ventral, first single ray minute, the 
second about half as long as the third, anterior divided ray the longest; 
pectorals elongate; but not reaching the root of the ventral; ventrals 
beneath the dorsal; anal fin entirely behind the dorsal, its origin much 

Measurements of Pseudorasbora parva. 



Locality. 



Taihoku 4 

Taihoku 4 

Taihoku 4 

Taihoku 4 

Taihoku 4.28 

Taihoku 4 

Raupi 3.6 

Raupi 3.4 

Rigyokutsu 4.25 

Shori 4.5 

Shori 4.5 

Ako I 4.29 

Shinchiku j 4.5 

Bokusekikaku I 4 



4 

4 

4 

4 

4 

3-87 

4 

4 

4.14 

3-64 
3-5 I 3 
4-ii;3 
3-29 3 
3-6o 3 



2,6 
2, 6 
2. 6 
2,6 
2,6 
2,6 
2, 6 
2, 6 
2, 6 
2, 6 
2, 6 
2, 6 
2,6 
2,6 



i-77| 
2 

I.75J 
1.77 

1-75 
1-75 



1.82 
1.64 

1-54 
1.80 
1.89 

1-75 



2.50 3 
2.50 3 
2.33 3 
2.28 3 
2.33 3 
2.33 3 
3 3 
3 3 
2.50 2 
2.25 3 
2.13 3 



86 



2.37 2.71 
2.37 3 
2.36 2.66 



3.665 
3-75 ,5 



3-50 

3-66 

3-50 

3.80 

4 

3-50 

4 

3-8o 

3-66 

3-66 

3-66 

4 



-38-5 
-36-5 
-36-5 
-37-5 
-37-5 
-36-5 
-36-5 
-36-5 
-36-5 
-38-5 
-36-5 
-35-5 
-37-5 
-38-6 



M E 

-1 



77 
75 
70 
77 
71 
69 
47 
38 
100 
96 
93 



75 



nearer to that of the ventral than base of the caudal; caudal fin bifur- 
cate, tip of each lobe sharply pointed. 

Body covered with rather large cycloid scales; lateral line very 
slightly decurved, extending along the middle of the sides. 



224 Annals of the Carnegie Museum. 

Color in alcohol dark grayish above, sides beneath lateral line and 
lower surface silvery; a slaty gray lateral band along the middle of 
sides; most of the scales with black edges; all the fins grayish. 

Total length 77 mm. 

Described from a specimen from Taihokii, collected by Oshima in 
December, 1916. 

Habitat: The present species is very abundant in pools and rivers in 
Formosa. I have a number of specimens from the Tamusui River; 
Taihoku; Raupi, Giran; Tozen River; Nanto; Rigyokutsu, Nanto; 
Shinchiku; Ako; Shori, Toyen; Bokusekikaku. 

Remarks: The length of head and the depth of body are variable. 

Genus Pararasbora Regan. 

1908. Pararasbora Regan, Ann. Mag. Nat. Hist. (8), II, p. 360. (Type Pararas- 
bora nioltrechli Regan.) 

Scales large, 5.5-6 scales in an oblique series between origin of 
dorsal and lateral line, 2 scales between the latter and the root of 
ventral. Lateral line decurved, running along the lower part of 
the tail. Dorsal fin with seven branched rays, inserted behind the 
ventral; the anal entirely behind the dorsal, with seven branched rays. 
Mouth oblique, its angle extending to the anterior border of the orbit; 
lower jaw slightly shorter than the upper, with no prominence in 
front, upper jaw entire, with no emargination. Barbels none. Gill- 
rakers rudimentary. Pharyngeal teeth in two series, 4, 4 — 4, 4. 

Distribution: Formosa. 

29. Pararasbora moltrechti Regan. 
Baahii (Formosa). 

1908. Pararasbora moltrechti Reg.\n, Ann. Mag. Nat. Hist. (8), II, p. 360; Lake 
Candidius, Formosa. 

1909. Jordan & Richardson, Mem. Carneg. Miis., IV, No. 4, 1909, p. 170; Lake 
Candidius (after Regan.) 

Head 4. 11 in length; depth 4; D. 3. 7; A. 2, 7; P. 14; V. 7; width of 
head 1.75 in its length; eye 3.25 in head; intcrorbital space 2.1; snout 
3.25; i^ectoral 1.2; ventral 1.4; thirty-six scales in the lateral line, six 
scales in an oblique' series between origin of dorsal and lateral line, 
two scales between the latter and the root of ventral, four scales 
between lateral line and the middle of belly; pharyngeal teeth 4, 
4 — 4, 4; gill-rakers 2 + 6, rudimentary. 

Body elongate, compressed; dorsal and \entral profiles ecpiallj' 



The Fresh Water Fishes of the Island of Formosa. 225 



arched; head moderate, pointed anteriorly, upper surface more or less 
depressed; snout bluntly pointed, interorbital space broad and flat- 
tened; mouth oblique, its angle reaching beneath the anterior border 
of orbit; upper lip thin; lower jaw slightly shorter than the upper, 
with a sharp edge; five branchiostegals; nostrils close together, in front 
of the eye above. 

Origin of dorsal nearer to base of caudal than tip of snout, inserted 
above the space between the ventral and anal, anterior ray longest; 
pectoral fin not reaching the ventral, origin of ventral much in advance 
of that of dorsal; anal fin entirely behind the dorsal, its origin nearer 
to that of ventral than base of caudal, its base covered with a series of 
scales; caudal fin emarginate, tip of each lobe obtusely pointed; 
caudal peduncle elongate, its depth about twice in the length of head. 

Body covered with rather large cycloid scales; lateral line much 
decurved, running along the lower half of the tail. 

Color in formalin dark gray above, lower parts yellowish white ; 
top of head black; a dark brown stripe from occiput to base of the 
caudal above, running along the dorsal median line; side with a broad 
dark brown lateral band, distinct posteriorly; all the fins grayish. 

Total length 83 mm. 

The present description is drawn from a specimen from Jitsuget- 
sutan, collected by T. Aoki in August, 1916. 

Habitat: Restricted to Jitsugetsutan (Lake Candidius). Two 
specimens. 

Measurements of Pararasbora moUrechti. 



Locality. 


Head. 
Depth. 


P < 

1 


cu 


> 


Width 
of Head. 

Interor- 
bital. 


3 

a 
to 


>• 

W 






Jitsugetsutan 

Jitsugetsutan 


4.11, 4 
4-I3I 4 


3. 7 , 2, 7 
3. 7 ! 2. 7 


14 
14 


7 
7 


1.75 2. II 
1.88 2.25 


3.25 
3.20 


3-25 

4 


6-36-4 
6-35-4 


83 
81 



Phoxiscus gen. nov. 
Body elongate, compressed; postventral part keeled. Head 
moderate, its top rather fiat; snout obtusely rounded. Mouth 
oblique; lower jaw slightly longer than the upper; lips thin, normal. 
Barbels none. Teeth hooked, in double series, 5, 3 — 4, 4. Lateral 
line incomplete, visible only in the anterior part of the body. Scales 
large, about thirty in a lateral series. Dorsal tin short, without osseous 
spine, inserted behind the origin of the ventral; anal fin of moderate 
16 — DEC. 17, 1919. 



226 Annals of the Carnegie Museum. 

length, with seven branched rays, not extending forwards to below the 
dorsal. 

Distrih ution : Formosa . 

Remarks: The present genus is closely related to Ilemitreruia Cope. 
It differs, however, in having three and four teeth in the second row, 
instead of two as in the latter, as also in the carinate belly. 

30. Phoxiscus kikuchii sp. nov. (Plate LI, Fig. 3). 

Head 3.5 in length; depth 3.5; D. 2, 7; A. 2, 7; P. 13; V. 7; width 
of head 1.75 in its length; eye four in head; interorbital space 2.33; 
snout 3; pectoral 1.5; ventral 1.75; thirty scales in a lateral series, 
five scales in an oblique series between origin of dorsal and lateral line, 
six scales between the latter and the middle of belly, two scales be- 
tween lateral line and the root of ventral; pharyngeal teeth 5, 3 — 4, 4; 
gill-rakers 2 + 5. 

Body elongate, compressed, post\'entral part weakly keeled; head 
moderate, its top rather flat; snout obtusely rounded anteriorly, 
interorbital space broad, more or less convex; mouth oblique, with 
thin lips; lower jaw slightly longer than the upper; no barbel; maxil- 
lary scarcely reaching a vertical through anterior border of orbit; eye 
moderate, superior and anterior; nostrils close together, the anterior 
in a short tube; pharyngeal teeth slender and hooked; gill-rakers on 
the first arch short and separated. 

Dorsal fin short, nearer the base of caudal than the tip of snout, 
anterior ray the longest, its height 1.4 in length of head; pectoral fin 
not reaching the ventral, with a small fleshy flap; the ventral inserted in 
front of the origin of dorsal, rather slender; anal fin entirely behind the 
dorsal, rather short, anterior ray the longest, its height 1.66 in head; 
caudal fin cmarginate, tip of each lobe pointed. 

Body covered with large imbricated scales; lateral line incomplete, 
decurved, reaching posterior third of the ventral. 

Color in alcohol brownish gray above, lower parts yellowish; top 
of head and dorsal median line purplish; sides with a bluish gray 
longitudinal band; all the fins uniformly cream-colored. 

Total length 60 mm. 

Described from a specimen from Bokusekikaku, collected by Yone- 
taro Kikuchi of the Taihoku Museum. 

Habitat: Bokusekikaku (a single specimen). . 

Remarks: Named for Vonetaro Kikuchi, collector of the Taihoku 
Museum. 



The Fresh Water Fishes of the Island of Formosa. 227 
Genus Distcechodon Peters. 

1880. Dislxchodon Peters, Monatsb. Konigl. Ak. Wiss. Berlin, p. 924. (Type 
DisUvchodon tumiroslris Peters.) 

Scales large; lateral line decurved, running along the middle of the 
tail. Dorsal fin short, with seven branched rays and two smooth 
spines, opposite the ventral. Anal fin of moderate length, with nine 
branched rays, not extending forwards to below the dorsal. Abdomen 
not carinated. Snout swollen, produced anteriorly, tip of its skin 
overlapping the upper lip. Mouth subinferior, transverse. Upper 
lip with a transverse inner fold; anterior border of the lower jaw rather 
sharp. Barbels none. Gill-rakers setiform, set very closely; pharyn- 
geal teeth 7, 3 — 3, 7, compressed, with grinding surface. 

Distribution: China; Formosa. 

31. Distcechodon tumirostris Peters. 
Gonhii (Formosa). 
1880. Distcechodon tumiroslris Peters, Monatsb. Konigl. Ak. Wiss. Berlin, p. 925; 
Ningpo, China. 

Head 4.64 in length; depth 3.92; D. II, 7; h. 3, 9; P. 17; V. 9; width 
of head 1.80 in its length; eye four in head; interorbital space 2.5; 
snout 3; pectoral 1.28; ventral 1.5; seventy-two scales in the lateral 
line, thirteen scales in an oblique series between origin of dorsal and 
lateral line, ten scales between the latter and the middle of belly, six 
scales between lateral line and the root of ventral; pharyngeal teeth 
7, 3 — 3, 7; gill-rakers on the first arch 75. 

Body elongate, compressed, dorsal profile slightly depressed at the 
occiput, abdomen rounded, not carinated; head rather small, triangular, 
its top more or less convex, profile on the nape concave; interorbital 
space broad, rather flat; snout obtusely rounded anteriorly, the tip 
swollen, the end of its skm overlapping the upper lip; no barbel; 
mouth subinferior, transverse; upper lip fleshy, not fringed, with a 
thin inner fold, lower jaw nearly as long as the upper, with a sharp 
anterior edge, lower lip not continuous; a deep oblique fissure, crossing 
the angle of mouth and extending downwards and backwards to the 
vertical from the nostril; eye large, anterior; nostrils close together, 
in front of eye above; pharyngeal teeth on the outer row very strong, 
with grinding surface and sharply gointed tip, strongly compressed 
laterally, those on the inner row exceedingly small and slender, with 
grinding surface; gill-rakers minute, setiform, set very closely. 



228 Annals of the Carnegie Museum. 

Origin of dorsal midway between tip of snout and base of caudal, 
armed with two strong osseous smooth spines, opposite the ventral; 
pectorals reaching midway of the distance to origin of ventral, with 
a fleshy flap; ventral rather slender, with a scaly flap; anal fin entirely 
behind the dorsal, short, anterior ray the longest; caudal fin bifurcate, 
tip of each lobe pointed; caudal peduncle rather long, its depth 2.25 
in length of head. 

Body covered with uniform cycloid scales; lateral line slightly 
decurved, extending along the middle of the tail. 

Color in formalin uniform dark gray above, paler below; lower parts 
silvery; pectoral, dorsal, and caudal fins dusky, other fins whitish. 

Total length 230 mm. 

Described from a specimen from Giran. 

Habitat: Taihasho, Giran. (Two specimens). 

Genus Ischikauia Jordan & Snyder. 
1901. Ischikauia Jordan & Snyder, Proc. U. S. Nat. Mus., XXIII, p. 346. 
(Type Opsariichlhiis sieenackeri Sauvage.) 

Body compressed; caudal peduncle deep. Mouth obliciue; lower 
jaw slightly projecting; maxillary freely protractile, not extending to 
edge of orbit; no barbels. Teeth, all slightly hooked, with a narrow 
grinding surface; in three rows; three or four on first, or outer row, 
five on second, two on third, or inner row. Pseudobranchiae present. 
Gill-rakers on first arch 13+4; low, pointed. Alimentary canal 
twice as long as body. Air bladder in two divisions, extending pos- 
teriorly to vent. Peritoneum with black pigment. Scales of moder- 
ate size, about sixty-five in lateral line; thirteen from lateral line to 
insertion of dorsal. Lateral line sharply decurved anteriorly, gradu- 
ally curving upward and extending posteriorly along middle of caudal 
peduncle. Dorsal inserted a little behind origin of ventrals, composed 
of nine rays; first ray short, and closely adnate to the next; second ray, 
spine-like, strong; the other rays branched. Seventeen anal rays; 
the first two spine-like, weak. Caudal forked, the tips sharp. Pec- 
torals pointed. (Jordan & Snyder.) 

Distribution: Formosa; Indo-China; Japan. 

32. Ischikauia macrolepis Regan. 

1908. Ischikauia macrolepis Recan, Ann. Mag. Nat. Hist. (8), I, i). 150; Kagi, 
Formosa. •' 

Depth of body 3.33 in the length, length of head 4. Snout a little 
shorter than eye, the diameter of which is 3.25 to 3.50 in the length 



The Fresh Water Fishes of the Island of Formosa. 229 

of head and less than the interorbital width. IMouth ()l)li(iue. Dorsal 
rays ten, seven branched, its origin behind the vcntrals and nearer 
to the base of caudal than to the end of snout. Anal rays sixteen to 
seventeen, thirteen or fourteen branched. Pectoral extending to the 
ventrals. Thirty-eight to forty scales in a longitudinal series, seven 
or eight in a transverse series from origin of dorsal to lateral line, three 
between lateral line and base of ventral. 

Three small specimens, the largest 60 mm. in total length, from 
Kagi, Formosa, collected by Herr Sauter. This species is very similar 
to the Japanese /. steenackeri (Sauvage), which, however, has much 
smaller scales. (Regan.) 

Habitat: Kagi. (Regan.) 

Remarks: Not seen. 

Genus Ctenopharyngodon Steindachner. 

1866. Ctenopharyngodon Steindachner, Verh. Zool.-Bot. Ges. Wien, p. 782. 
(T^-pe LeuciscHS idella Cuv. & Val.) 

Body oblong. Scales of moderate size; lateral line complete, 
running nearly into the middle of the side of tail. Dorsal fin short, 
without spine, opposite the ventral, anal fin short. Mouth of moder- 
ate width, anterior, with upper jaw somewhat longer. Both jaws 
with simple lips, the lower distinct at the angle of the mouth only. 
Upper jaw slightly protractile. Barbels none. Gill-rakers rather 
short, lanceolate, rather widely set. Pseudobranchiae present. 
The attachment of the branchial membrane to the isthmus takes 
place behind a vertical from the orbit. Pharyngeal teeth 5, 2 — 2, 5, 
those of the outer series very strong, strongly compressed, with the 
outer layer deeply folded. (Giinther.) 

Distribution: Formosa; China; Amur Province. 

33. Ctenopharyngodon idellus (Cuv. & Val.). 
Tsauhii (Formosa). 

1844. Leuciscus idella Cuv. & Val., Hist. Nat. Poiss., XVII, p. 362. — Richard- 
son, Ichthyol. China, 1846, p. 297; Canton. — Bleeker, Mem. Cyprin. 
Chine., 1871, p. 47; Canton; Yang-tze-kiang. 

1855. Leuciscus tschiliensis B.\silewsky, Nouv. Mem. Soc. Nat. Mosc., X, p. 233; 
Northern China. 

1866. Ctenopharyngodon laticeps Steindachner, Verh. Zool.-Bot. Ges. Wien, 
p. 782, Taf. XVIII, Figs. 1-5; Hongkong. 

1868. Ctenopharyngodon idellus Gijnther, Cat. Fish. VII, p. 261, China. — Ann. 
Mag. Nat. Hist., Sept., 1873, Shanghai.— Peters, Monatsb. Konigl. Ak. 



230 Annals of the Carnegie Museum. 

Berlin, 1880, p. 926; China. — -Jordan & Evermann, Proc. U. S. Nat. Mus., 
XXV, 1903, p. 322; Taihoku, Formosa. ^ — Jordan & Richardson, Mem. 
Carneg. Mus., IV, 1909, No. 4, p. 169; Formosa. — Berg, Ichthyol. Amur., 
1909, p. 120; Amur. 

Head 4.09 in length; depth 3.4; D. 3, 7; xA.. 3, 7; P. 18; \'. 9; width of 
head 1.33 in its length, eye 6 in head; interorbital space 1.8; snout 
2.5; pectoral 1.33; ventral 1.71; forty scales in the lateral line, seven 
scales in an oblique series between origin of dorsal and lateral line, 
five scales between the latter and the root of ventral, eight between 
lateral line and the middle of belly; pharyngeal teeth 5, 2 — 2, 5; gill; 
rakers 6 + 10. 

Body stout, dorsal and ventral profiles about equally arched, tail 
compressed, head small, flattened; snout broad and short, anterior 
margin obtusely rounded; mouth subinferior, its angle reaching a 
vertical through anterior border of the anterior nostril; upper jaw 
somewhat longer than the lower; upper lip rather thin, thicker in front; 
lower lip distinct at the angle of the mouth only; eyes moderate, 
superior and anterior; interorbital space very broad; nostrils together, 
in front of eye above; gill-membranes united, scarcely reaching to 
isthmus behind a vertical through the anterior border of orbit: gill- 
rakers long and slender, widely set; pharyngeal teeth of the outer row 
strong, laterally compressed, apical half comb-shaped, with a series of 
folds on both sides; those of the inner row very small, laterally com- 
pressed, the apical part comb-shaped. 

Origin of dorsal slightly nearer tip of snout than base of caudal, 
anterior rays longer, pectorals reaching beyond half or the distance to 
ventrals; ventral rather small, inserted one scale behind the origin of 
dorsal; anal fin entirely behind the dorsal; its origin nearer to base of 
caudal than that of the ventral. 

Body covered with large scales with radiating striae and concentric 
rings; lateral line continuous, slightly decurved, extending along the 
middle of the sides. 

Color in formalin uniformly gray above, paler below; lower parts 
and ventral fins whitish; upper surface of the pectoral dusky; other 
fins pale gray. 

Total length 360 mm. 

Described from a specimen from Shori, TuNen. 

Habitat: Bred in ponds throughout the island. 

Remarks: The present species is one of the inijiortant fresh-water 



The Fresh Water Fishes of the Island of P'ormosa. 2.31 

food-fishes among the Chinese people, though it is not a native of the 
Island. Every year larv;c of this fish are imported from Southern 
China and are bred artificialh' by Chinese farmers. 

Genus Acheiloc.nathus Bleeker. 
i860. Acheilogtialhus Bleeker, Ichth. Archipel. Indie. Prodr., II, p. 228. (Type 
AcheilognalltHs melanogasler Bleeker.) 

Body more or less deep and compressed; head short; eye more or 
less large; snout rather short and blunt; mouth small, the maxillary 
not reaching the eye; maxillaries each with a barbel; teeth 5-5, 
smooth, with a narrow grinding surface; interorbital space rather 
broad. Intestine long. Peritoneum black. Scales large, some of 
those on the sides imbricated, 36-39. Origin of the dorsal about 
midway in the length of the body without caudal, base of fin moder- 
ate, with eight to ten developed rays; caudal deeply emarginate; 
ventrals generally inserted a little before origin of dorsal. Lateral 
line slightly decurved, and continuous. (Jordan and Fowler.) 

Distribution: Formosa; China; Corea; Amur Province; Japan. 

34. Acheilognathus himantegus Giinther. 

1868. Acheilognathus himantegus Gunther, Cat. Fish., VII, p. 277; Formosa. 

Head 4.58 in length; depth 2.89; D. 2, 8; A. 2, 11 ; P. 12; V. 7; width 
of head 1.66 in its length; eye 2.66 in head; interorbital space 2; snout 
3; pectoral 1.2; ventral 1.33; thirty-four scales in the lateral line, six 
scales in an oblique series between origin of dorsal and lateral line, 
four scales between the latter and the middle of belly, three scales 
between lateral line and the root of the ventral; pharyngeal teeth 5-5; 
gill-rakers 3 + 9. 

Body compressed, rather deep; head small, its top convex, with a 
median horny ridge; snout short, obtusely rounded anteriorly, its 
tip with a bony prominence, tip of its skin overlapping the upper lip; 
interorbital space broad, interspace between nostrils swollen; mouth 
subinferior and oblique, its angle reaching the nostrils below; lips thin; 
lower jaw slightly shorter than the upper; two maxillary barbels; eyes 
moderate, slightly anterior and superior; nostrils close together, in 
front of eye above. 

Origin of dorsal midway between tip of snout and base of caudal, 
much behind that of the ventral, anterior rays longer; pectoral fin not 
reaching the ventral; ventral slender, scarcely reaching the root of 
anal; anal fin inserted below thje middle of dorsal, elongate, its origin 



232 



Annals of the Carnegie Museum. 



nearer to base of pectoral than the base of caudal; caudal fin emar- 
ginate, the tip of each lobe sharply pointed; caudal peduncle rather 
long, its depth 1.6 in length of head. 

Scales moderate, with a black marking at the tip; lateral line 
strongly decurved on the trunk, extending along the lower part of 
the tail. 

Color dark gray above, paler below; lower part of the sides sky-green ; 
a short yellow lateral band on the nape; a black lateral band runs along 
the middle of the tail, terminating in a black band-like spot between 
the middle caudal rays; interspace between each dorsal ray with a 
black streak; dorsal and anal fins pinkish, the other fins dusky; no 
black spot on the shoulder. 

Total length 67 mm. 

The present description is from a specimen from Taihoku, col- 
lected by Oshima in September, 1916. 

Habitat: Taihoku; Wodensho, Taichu; Shimotamusui River. 

Remarks: The nearest relative of the present species is Acheilog- 
nathus cyanostigma Jordan «& Fowler'' from Japan. It differs from 
A. himantegus in having shorter barbel, smaller number of anal rays, 
and thirty-nine scales instead of thirty-four in lateral line. 

Measurements of Acheilognathiis himantegus. 



Locality. 



Taihoku 4.58 

Taihoku 4.33 

Taihoku 4-58 

Taihoku 4.45 

Taihoku 4.27 

Taihoku ' 4-55 

Shimotamusui River. 4.30 

Shimotamusui River. 4.20 

Shimotamusui River. 4 

Wodensho 4 



J3 












a 


Q 


< 


Ph 


> 


Q 












2.89 


2, 


8 


2, II 


12 


7 


2.74 


2, 


8 


2, II 


12 


7 


2.84 


2, 


8 


2, 12 


12 


7 


2.72 


2, 


8 


2, 11 


12 


7 


2.87 


2, 


8 


2, 10 


12 


7 


3 


2, 


8 


2, II 


12 


7 


2.58 


2, 


8 


2, II 


12 


7 


3 


2, 


8 


2, II 


12 


7 


3 


2, 


8 


2, II 


12 


7 


3 


2, 


8 


2, II 


12 


7 






c3 


3 

C 


1.66 


2 


3 


1. 71 


2.16 


3 


1. 71 


2.16 


3 


1. 71 


2 


3-33 


1. 71 


2.40 


3 


1.80 


2.25 


3 


1.83 
1.83 


2.20 
2.25 


3 
3 


1-75 
1.60 


2.50 
2.33 


3 
3 



■J 



2.66 6-34-4 67 



2.66 6-34-4 
3 J6-34-4 
2.66 6-34-4 



[6-33-4 58 

j 6-3 4-4 1 53 
6-34-4 i 57 
6-34-4! 53 

16-34-4 35 
6-34-41 39 



Genus Rhodeus Agassiz. 
1835. Rhodeus Agassiz, Mem. Soc. Hist. Neuchat., I, p. 37. (Type Cyprinus 
aniarus Bloch.) 
Scales of moderate size; lateral line incomplete, only on the anterior 
part of the trunk. Dorsal fin with from nine to twelve branched rays, 
''Acheilognathiis cyanostigma Jordan & Fowler, Proc. U. S. Nat. Mus., XXVI, 
1903, p. 820; Lake Biwa; Lake Yogo. 



The Fresh Water Fishes of the Island of Formosa. 233 

extending from the ventrals to beyond the origin of the anal. Anal 
fin rather elongate, with about twelve rays. Mouth subinferior, 
small, arched; lower jaw without labial fold. Barbels none. Gill- 
rakers very short; pseudobranchia\ Pharyngeal teeth 5-5, com- 
pressed, not denticulated, the bevelled surface with a simple groove. 

Male, during the spawning season, with tubercles on the snout, 
and the female with a long external urogenital tube. (Giinther). 

Distribution: Europe; Caucasus; China; Formosa; liasin of Amur; 
Corea; Japan. 

35. Rhodeus ocellatus (Kner). 

1859. Pseudoperilampus (?) ocellatus Kner, Novara, Fisch., Ill, p. 365, Taf. 15, 

Fig. 6; Shanghai. 
1868. Rhodeus ocellatus Gijnther, Cat. Fish., VII, p. 280; China. — Bleeker, 

Mem. Cyprin. Chine., 1871, p. 34, PI. VI, Fig. 3; Yang-tze-kiang. — Gijnther, 

Ann. Mag. Nat. Hist., Sept., 1873, p. 249; Shanghai. — Jordan & Seale. 

Proc. U. S. Nat. Mus., XXIX, 1905, p. 518; Shanghai. — Jordan & Metz, 

Mem. Carneg. Mus., VI, 1913, p. 20; Suigen, Corea. 

Head 4.23 in length; depth 2.28; D. 2, 12; A. 2, 13; P. 1 1 ; V. 6 ; 
width of head 2 in its length; eye 3 in head; interorbital space 2.5; 
snout 3; pectoral 1.25; ventral 1.33; thirty-four scales in a longitudinal 
series, thirteen scales in an oblique series between origin of dorsal and 
the middle of belly; pharyngeal teeth 5-5; gill-rakers 2 + 10. 

Body much compressed, deep and rhomboidal; head small, its 
dorsal profile slightly concave above the eyes; snout as long as the 
diameter of eye, tip obtusely rounded, with a bony oval swelling, which 
is provided with a number of minute tubercles, more or less over- 
lapping the upper lip; interorbital space broad, rather fiat; mouth 
subinferior, rather transverse, its angle reaching the anterior border 
of nostril below; lower jaw slightly shorter than the upper; no barbel; 
nostrils close together, approximated to eyes, posterior nostril widely 
opened; pharyngeal teeth compressed, the sides not serrated, with a 
grinding surface. 

Dorsal fin elongate, its origin nearer tip of snout than base of caudal, 
inserted behind the origin of ventral, base of the fin one and one-third 
times as long as the head; pectoral fin reaching to within a short dis- 
tance to ventral; the ventral slender, scarcely reaching the origin of 
anal; anal fin elongate, inserted beneath the middle of the base of 
dorsal, anterior ray longest; caudal fin deeply emarginate, the tip of 
each lobe pointed. 



234 



Annals of the Carnegie Museum. 



Scales moderiite, imbricated; lateral line incomplete, visible only 
on four or five scales near gill-openings. 

Color dark gray above, paler below; lower parts of the sides bluish 
anteriorly; scales on the back and sides with black edges; a black 
lateral band runs along the middle of the posterior half of the side, 
commencing near the origin of the dorsal below; a black spot above the 
gill-opening and a dusky cross-bar on the shoulder; dorsal and caudal 
fins dusky, the pectoral and ventral whitish, anal fin pinkish. 

Total length 68 mm. 

Described from a specimen from Taihoku, collected by Oshima 
in September, 1916. 

Habitat: Abundant in the ponds and rivulets near Taihoku. My 
specimens came from Taihoku and Nanto. 

Measurements of Rhodeus ocellatus. 



Locality. 



< 


oj 


>■ 





S2 


3 


B 




u 

0! 


2, 13 


II 


6 


2 


2.50 


3 


3 


34-13 


2, II 


10 


7 


2 


2.50 


3 


3 


34-11 


2, 12 


II 


6 


2 


2.50 


3-50 


3 


34-11 


2, 12 


II 


7 


2 


2.33 


3 


3 


34-11 


2, 12 


II 


6 


2 


2.33 


3-66 


3 


34-11 


2, 12 


II 


6 


2 


2.50 


3-50 


3 


34-11 


2, II 


10 


7 


2 


2.33 


3 


3 


32-11 






Taihoku 
Taihoku 
Taihoku 
Taihoku 
Taihoku 
Taihoku 
Nanto. . 



4.23 2.28 
3.91J 2.60 
3.90 2.29 
3.80 2.37 
4 I 2.52 
3.80 2.22 
4- 40' 2.25 



68 
55 
50 
48 
46 
50 
55 



Genus Z.\CCO Jordan & Evermann. 

1902. Zaffo Jordan & Evermann, Proc. LI. S. Nat. Mus., XXV, p. 322. (Type 
Leucisciis platypus Temminck & Schlegel.) 

Body moderately elongate and compressed; head compressed; 
guout conical, pointed; eyes moderate; mouth oblique, not notched; 
no barbels; teeth 5 or 4, 4, and 2 or i — ^i of- 2, 4 and 4 or 5; interorbital 
space convex. Intestine short. Peritoneum black. Scales cycloid, 
narrowly imbricated, forty to sixty in the lateral line. Dorsal nearer 
tip of snout than base of caudal, or midway between, its developed 
rays seven; anal inserted below, or a trifle before, tip of depressed 
dorsal; its basis long, and composed of nine or ten developed rays; 
caudal cmarginate; pectorals sometimes reaching ventrals; ventrals 
inserted a little before or below the origin of dorsal. Lateral line 
continuous and decur\-ed. Breeding males have the head, the lower 
surface of the caudal peduncle, and the anal fin furnished with horny 



The Fkesh Water Fishes of the Island of Formosa. 235 

tubercles, not as numerous as those in Opsariichthys, and larger in 
proportion. The anal fin also has the developed rays elongated and 
with adipose expansions. (Jordan & Fowler.) 
Dislribiition: Japan; Corea; Formosa. 

Synopsis of the Formosan Species. 
a. Scales in the lateral line 42-44; 8 scales between lateral line and the origin of 

dorsal; 2 scales between lateral line and the root of ventrals platypus. 

aa. Scales in the lateral line 49-52; 9-10 scales between lateral line and the origin 

of dorsal; 3 scales between lateral line and the root of ventrals. . . Aernmincki. 

aaa. Scales in the lateral line 53-55; 12 scales between lateral line and the origin 

of dorsal; 4 scales between lateral line and the root of ventrals . . pachyccphaliis. 

36. Zacco platypus (Temminck & Schlegel). 
Oikawa, Haya or Ilae (Japan); Chopien or Anoye (Formosa). 

1846. Leiiciscus platypus Temminck & Schlegel, Fauna Japonica, Poiss., p. 297, 

PI. CI, Fig. i; streams near Nagasaki. 
1863. Opsariichthys platypus Gunther, Cat. Fish., VII, p. 296; Japan & Formosa. 

— ISHIKAWA, Zool. Mag. Tokyo, 1895, P- 121; Hakone; Matsubara on Lake 

Biwa. — Prel. Cat., 1897, p. 11; Tega Lake; Tokyo; Chichibu; Lake Suwa; 

Lake Biwa; Kyoto; Tsuyama. — Boulenger, Proc. Zool. Soc. London, 1899, 

p. 961; Hainan. 
1900. Barilius platypus Jord.\n & Snyder, Proc. U. S. Nat. Mus., XXII, p. 344; 

Lake Biwa. 
1903. Zacco platypus Jordan & Fowler, Proc. U. S. Nat. Mus., XXVI, p. 851; 

Tsuchiura; Kinu River; Yodo River; Chikugo River; Yabe River; Tama 

River; Nagoya; Kawatana; Lake Biwa. — Smith & Cope, Proc. U. S. Nat, 

Mus., XXXI, 1908, p. 462. — Jordan & Thompson, Mem. Carneg. Mus., 

VI, no. 4, 1914, p. 232; Lake Biwa; Okayama. — Snyder, Proc. U. S. Nat. 

Museum, XLII, p. 404; Takamatsu River. — Jordan, Snyder, & Tanaka. 

Journ. Coll. Sci. Tokyo, XXXIII, 1913, p. 75; Japan. 
1846. Leuciscus macropus Temminck & Schlegel, Fauna Japonica, Poiss., p. 209, 

PI. CI, Fig. 2; Nagasaki. 
1846. Leuciscus minor Temminck & Schlegel, Fauna Japonica, Poiss., p. 210, 

PI. CI, Fig. 3; Nagasaki.' 
1903. Zacco evolans Jordan & Evermann, Proc. U. S. Nat. Mus., XXV, p. 323; 

Taihoku, Formosa. 

Head 4 in length; depth 3.71 ; D. II, 7; A. 3, 9; P. 15; V. 9; width of 
head 1.92 in its length; eye 4 in head; interorbital space 2.78; snout 3; 
ventral 1.32; pectoral a little longer than the head; forty-three scales 
in the lateral line, eight scales in an oblique series between origin of 
the dorsal and lateral line, four scales between the latter and the middle 
of belly, two scales between lateral line and the root of the \'entral; 
pharyngeal teeth 5, 4, i — i, 4, 5; gill-rakers 2 + 8. 



236 Annals of the Carnegie Museum. 

Body elongate, compressed; dorsal profile more or less arched, head 
moderate, its top very slightly arched; snout bluntly pointed anteriorly; 
mouth oblique, its angle extending to a vertical through anterior 
border of orbit; lower jaw slightly shorter than the upper; eye moder- 
ate, superior; nostrils close together, in front of eye, anterior nostril 
in a short tube; lower part of pre-operculum and sides of the snout 
with a number of colorless tubercles; extremity of the snout, upper lip, 
and outer part of lower jaw with a series of minute tubercles. 

Origin of dorsal nearer tip of snout than base of caudal, opposite 
the ventral, short, rather high, but not exceeding the length of head; 
pectoral elongate, extending beyond the root of ventral; ventral fin 
rather short, reaching the root of anal; anal fin elongate, its middle 
ray much longer than the head, when depressed reaching beyond the 
root of caudal, two or three of the anal rays provided with tubercles; 
caudal fin strongly bifurcated; depth of caudal peduncle 2.4 in length of 
head. 

Scales thin and cycloid; lateral line decurvcd, extending along the 
lower half of the tail. 

Color in alcohol brownish gray above, paler below; lower parts 
silvery; sides with about twelve dark cross-bars; membrane of dorsal 
and anal fins with a series of black streaks; dorsal and caudal fins 
dusky; the pectoral and ventral whitish. 

Total length 120 mm. 

Described from a specimen from Choso River near Koshiryo, 
collected by T. Aoki in i\ugust, 1917. 

Habitat: Tamusui River (Shinten and Heirinbi); Shinchiku; Choso 
River. 

Remarks: In the year 1903, Jordan and Evermann described a 
species of the genus Zacco from Formosa, giving the name Zacco 
evolans. According to their statement it agrees fairly well with the 
Japanese species, Zacco platypus, except in the much greater length of 
the pectorals. All the characters, especially the length of the pectoral 
fin, of the above described specimen agree quite well with those of the 
type of Zacco evolans in the Stanford collections (No. 7129; Taihoku, 
Formosa). But, after a close examination of a vast number of speci- 
mens of the present species from the same locality, I found that the 
length of the pectoral fin is variable. Even the co-type of Z. evolans 
in the Stanford collections (No. 7333; Taihoku) is proxided with 
shorter pectoral fins which scarcely reach to the root of the ventral. 



The Fresh Water Fishes of the Island of Formosa. 237 



Such being the case, it is inadvisable to specilically separate the For- 
mosan form from the Japanese Z. platypus. 

Description of the Type Specimen of Zacco evolans, Jordan & 

Evermann. 
(No. 7129; Stanford collection; Taihoku; Tada coll.) 

Head 4.22 in length; depth 3.42; D. II, 7; A. 3, 9; P. 15; V. 9; width 
of head 2 in its length; eye 3.33 in head; interorbital space 3; snout 3; 
forty-four scales in the lateral line, eight scales in an oblique series 
between origin of the dorsal and lateral line, four scales between the 
latter and the middle of belly, two scales between lateral line and the 
root of the ventral; ventral 1.20 in head; pectoral considerably longer 
than the head. 

Body elongate, compressed; dorsal profile convex; head moderate, 
its top very slightly arched; snout pointed anteriorly; mouth oblique, 
its angle extending to a vertical through the anterior border of orbit; 
jaws subequal; eyes moderate, superior; nostrils close together, in front 

Measurements of Zacco platypus. 



Locality. 



Choso River 

Tamusui River . . 
Tamusui River. . 
Tamusui River. . 
Tamusui River. . 

Shinchiku 

Taihoku 

(Type Z. evolans) 

Tsuchiura 

(S. No. 7340) 

Tsuchiura 

(S. No. 7340) 

Tsuchiura 

(S. No. 7340).. . . 

Lake Biwa 

(S. No. 22623).. • 

Lake Biwa 

(S. No. 22623). • • 

Lake Biwa 

(S. No. 22623).. . 



4 3-71 
4 4-66 
3.864.61 
4 5 
4 |4-25 
4 14-35 



4.22 3.42 IL 7 
3-8o|n, 7 

7 
7 
7 
7 

• 4 '4-71 n. 7 



4 

4-30 
4.24 
4 



3-64 
4-50 
4-75 



4-5015 



< 


Pk' 


>' 


25 

1^ 


3 


a 


3.9 


15 


9 2.77 


3 


3.9 


15 


9 [3 


3 


3.9 


15 


9 3 


3 


3.9 


16 


9 3 


3 


3.9 


15 


9 3 


3 


3.9 


13 


9 


3 


3 


3.9 


15 


9 


3 


3 


3.9 


IS 


9 


3 


3 


3.9 


15 


9 


3 


3 


3-9 


15 


9 3 


3 


3.9 


15 


9 3 


3 


3.9 


15 


9 '3 


3 


3.9 


15 


9 


3-25 


3-25 



4 8-43-4 Beyond V. 

3-33 9-44-4 Reaching V. 

3-33 1 8-44-4 Reaching V. 

3-33 j 8-43-4 Reaching V. 

3.40 8-42-4 Reaching V. 

3 8-44-41 Not to V. 

3-338-44-4 Beyond V. 



5iS 



4 

4-50 

3-66 

4 

4 

4 



8-44-4 Reaching V 
8-44-4 Reaching V. 

8-44-4 Not to V. 

I 
8-44-4 j Not to V. 

8-44-4 Not to V. 

8-42-4 1 Not to V. 



120 

115 
100 
108 
107 
92 



135 
150 

148 
112 
138 
120 



of eye; lower part of pre-operculum, sides of the snout, and outer part 
of the lower jaw with a number of round tubercles. 

Origin of the dorsal nearer tip of snout than base of caudal, opposite 



238 Annals of the Carnegie Museum. 

the ventral, the fin very high, rays longer than the head; pectoral 
fins elongate, reaching to the root of anal; anal fin \ery large, its rays 
nearly twice as long as head; caudal fin deeply bifurcate, the tip of 
each lobe pointed. 

Scales thin and cycloid; lateral line decurved, low, extending along 
the lower half of the tail. 

Color in alcohol brownish gray above, silvery; sides with about 
eleven dark cross-bars; membrane of dorsal and anal fins with a 
series of dark streaks; other fins pale. 

Total length 98 mm. 

37. Zacco temmincki (Schlegel). 
Chopien (Formosa). 

1846. LenciscHs lemminckii Schlegel, Fauna Japonica, Poiss., p. 210, PI. CI, 

Fig. 4; Nagasaki. 
1868. Opsariichthys lemminckii Gunther, Cat. Fish., VII, p. 295; Japan. — ISHI- 

KAWA, Zool. Mag. Tokyo, 1895, p. 121; Hikone; Matsubara. 
1901. Barilius lemminckii Jordan & Snyder, Check-list, Fish. Japan, p. 47; 

Lake Biwa. 
1903. Zacco lemminckii Jordan & Fowler, Proc. U. S. Nat. Miis., XXVI, p. 852; 

Kawatana; Mogi River. — Snyder, Proc. U. S. Nat. Mus., XLII, 1912, 

p. 404; Yamaguchi. — Jordan & Metz, Mem. Carneg. Mus., \'I, no. 2, 

1913, p. 21; Fusan; Pung-tung, Corea.- 
1901. Opsariichthys acanthogenys Boulenger, Proc. Zool. Soc. London, p. 269;. 

Ningpo. 
1903. Zacco pachycephalus Jordan & Everm.^nn, Proc. U. S. Nat. Mus., XXV, 

p. 322; Taihoku, Suwata, Formosa (not of Gunther). 

Head 4 in length; depth 4.5; D. II, 7; A. 3, 9; P. 15; V. 9; width of 
head 2 in its length; eye 4 in head; interorbital space 2.75; snout 3; 
\entral 1.5; pectoral as long as head; fifty scales in the lateral line, 
nine scales in an oblicjue scries between origin of the dorsal and lateral 
line, six scales between the latter and the middle of belly, three scales 
between lateral line and the root of the ventral; pharyngeal teeth 
4, 4, i-l, 4, 5; gill-rakers 3 + 7. 

Body elongate, compressed, deeper anteriorly, postventral edge 
rather sharp, but not carinated ; head moderate, its top more or less 
convex; snout bluntly pointed anteriorly; mouth oblique, its angle 
reaching beyond the anterior border of orbit; upper jaw normal, 
lower jaw not protruding; eyes superior and anterior; nostrils close 
together, in front of eye; sides of snout, cheeks, and lower jaws with a 
number of tubercles. 



The Fresh Water Fishes of the Island of Formosa. 239 



Origin of dorsal at a point midway beween lij) of snout and base 
of caudal, slightly behind that of ventral, fin short and high, its ray 
not exceeding the length of head; pectoral elongate, scarcely reaching 
the root of ventral; ventral fins reaching the origin of anal; anal fin 
rather long, its middle rays considerably elongate, when depressed 
reaching beyond the root of caudal, furnished with horny tubercles; 
caudal fin forked, rather long; depth of caudal peduncle 2.5 in length 
of head. 

Scales thin and cycloid; lateral line strongly decurvcd, extending 
along lower half of the tail. 

Color in formalin dark gray above, paler below; belly and lower 
part of sides silvery; sides with ten black cross-bars and two irregular 

Measurements of Zacco temminkii. 



Locality. 





„. 




j> 


' Cll 

in 






tn 


9-50-6 


3 


I 0-49-6 


3 


I 1-53-6 


2 


10-52-6 


3 


I 0-49-6 


3 


I 0-5 1-6 


3 


I 0-49-6 


3 


I 0-49-6 


3 


1 0-49-6 


3 


1 0-49-6 


3 


9-52-6 


3 


1 0-49-6 


3 


I 0-49-6 


3 


I 0-5 1-6 


3 


I 1-50-6 


3 


10-50-6 


3 


10-48-6 


3 



Ij c 



Ako 

Ako 

Tamusui River. . . 

Giran River 

Heirinbi 

Inzanpo 

Inzanpo 

Tensonpi 

Suwo 

Shimotamusui R. . 

Shimotamusui R . 

Sobun River 

Rigyokutsu 

Lakusui River . . . 

Lakusui River . . . 

Taihoku 

(Z. pachytephalus) 

S. 12232 (7720) . 

Suwata 

(Z. pachycephalus) 

S. 12076 (7720) . 



4 

3.80 

3-30 

3-87 

3-33 

4 

3-6o 

3-50 

3-72 

4 

4 
3-82 

3-75 
3-33 
3-31 

3.60 

3-6o 



4-50 

3-8o 

4-25 

3-31 

3.88 

3-71 

4- 

4.66 

4.14 

4 

3.89 

4 

3-75 



2-75 

3 

3 

2.66 

3 

3 

3 

3-25 

3 

3 

2.75 

3 

3 

3 

3-25 



3 

3 

3.18 

3 

3 

3 

3 

3-25 

3 

3 

3-33 

3 

3-25 

3 

3 



n, 7 3.9 15 



3-75|n. 7|3.9 15 



3-25 3 



3-75 
4-SO 
5 



4. 4. I— 
1.4. 5 

4.3. I— 
1.4.5 

4.4. I— 
1,4.4 

4. 4. I— 
1.4.5 

4-50 

4 

5 

4-33 

4-75 

3-5 [4.4. I- 
1.4. 5 

3.66I4, 4, I- 

i 1.4.5 
4. 4. I— 

1.4.5 
4. 4. I— 

1.4.5 
4.4. I— 

1.4.4 



4-50] 



96 

92 

125 

145 

117 
108 
130 
no 
no 
86 

80 

97 

100 

100 



240 Annals of the Carnegie Museum. 

dark spots on the tail; nape just behind the gill-opening dark; dorsal 
fin with a series of black streaks; dorsal, anal, and caudal fins dusky; 
other fins whitish. 

Total length 96 mm. 

Described from a specimen from Ako, collected by T. Aoki in 
December, 1916. 

Habitat: One of the most abundant of the Formosan Cyprinidcs. 
My specimens came from Tamusui River; Daiko River; Daito River; 
Shinchiku; Dakusui River; Rigyokutsu, Xanto; Sobun River; Shimo- 
tamusui Ri\er; Ako; Heirinbi; Inzanpo; Tensonpi; Suwo; Giran. 

Remarks: Jordan and Evermann described a species of the genus 
Zacco from Suwata and Taihoku, Formosa, under the name Zacco 
pachycephalus (Giinther) (Proc. U. S. Nat. Mus., XXV, 1903, p. 322). 
As shown in the following table, the two specimens which are contained 
in the Stanford collections have ten scales between origin of dorsal 
and lateral line, three scales between the latter and the root of ventral, 
and forty-eight to fifty scales in lateral line, instead of 12-55-4 ^^ ^^ 
the type of Giinther's Z. (Opsariichthys) pachycephalus. Moreover, 
the other characters of these two specimens agree quite well with 
those of Zacco temmiucki. It is reasonable, therefore, to transfer 
Jordan & Evermann's Zacco pachycephalus to the present species. 

38. Zacco pachycephalus Giinther. 

1868. Opsariichthys pachycephalus Gunther, Cat. Fish., VII, p. 297; Formosa. 

Head 3.63 in length; depth 4; D. II, 8;.A. 3, 9; P. 14; V. 9; width of 
head 2 in its length; eye 3.6 in head; interorbital space 3.14; snout 3; 
pectoral 1.22; ventral 1.66; fifty- three scales in the lateral line, twelve 
scales in an oblique series between origin of dorsal and lateral line, 
nine scales between the latter and the middle of belly, four scales 
between lateral line and the root of the ventral; pharyngeal teeth 
5, 3, I — I, 3, 4; gill-rakers 2 + 8. 

Body oblong, compressed; head moderate, its top more or less 
flattened, cheek with traces of tubercles; snout pointed anteriorly, 
its dorsal profile curved; mouth oblique, its angle extending beyond 
the vertical through the anterior margin of the orbit; lower jaw very 
slightly shorter than the upper; eye superior; nostrils close together, 
supra-lateral, in front of the eye. 

Origin of dorsal at a point midway between tip of snout and base of 
caudal, o])posite to that of the ventral; pectoral fin not extending to 



The Fresh \\'ater Fishes of the Island of Formosa. 241 

the root of \enlral; \enlral reaching vent; four anterior branched 
rays of the anal elongate, reaching beyond the root of caudal; caudal 
fin deeply emarginatc, tip of each lobe pointed. 

Scales minute, thin; lateral line decur%ed, extending along the 
lower half of the tail. 

Color in formalin dark gray abo\e, lower half of tlie sides and belly 
silvery; sides with about twelve indistinct dark cross-bars; a black 
longitudinal streak runs along the middle of the tail; membrane of 
the dorsal with a series of dark streaks; dorsal and caudal fins dusky, 
other fins whitish. 

Total length 102 mm. 

The present description is from a specimen from the Tamusui River 
near Shinten, collected by T. Aoki in December, 1916. 

Habitat: Tamusui River (a single specimen). 

Genus Metzia Jordan & Thompson. 

1914. Metzia Jordan & Thompson, Mem. Carneg. Mus., VI, no. 4, p. 227. 
(Type Acheilognalhus mesetnbrinum Jordan & Evermann.) 

Body short, deep, very greatly compressed; head small and pointed; 
mouth moderate, somewhat oblique; the jaws subequal, maxillary 
reaching anterior edge of orbit; pharyngeal teeth in three rows, 
4, 4, 2 — 2, 4, 4, with brown tip. Scales large and well imbricated; 
lateral line complete, decurved. Origin of the dorsal slightly nearer 
tip of snout than tip of caudal fin, anal inserted behind last dorsal ray; 
ventrals and pectorals moderate, the latter falcate; caudal fin lunate. 
Peritoneum black; intestine elongate. No barbel. 

Distribution: Botel Tobago Island (Near Formosa). 

39. Metzia mesembrina (Jordan & Evermann). 

1903. Acheilognalhus mesetnbrinum Jordan & Evermann, Proc. U. S. Nat. Mus., 
XXV, p. 323; Kotosho, Formosa. 

Head 4 in length; depth 3.18; D. 2, 7; A. 3, 14; P. 15; V. 8; width 
of head 2 in its length; eye 3.6 in head, interorbital space 2.57; snout 
3.6; thirty-six scales in the lateral line, eight scales in an oblique 
series between origin or dorsal and lateral line, five scales between the 
latter and the middle of belly, three scales between lateral line and 
the root of the ventral; pharyngeal teeth 4, 4, 2 — 2, 4, 4. 

Body deep, very greatly compressed; head small and pointed; 
snout short, truncated in front; mouth oblique, its angle reaching 
beyond the anterior border of orbit; the jaws subequal, with thin lips; 
17 — DEC. 19, 1919. 



242 



Annals of the Carnegie Museum. 



eyes large, anterior; nostrils superior, close together, in front of eye 
above; tip of the pharyngeal teeth brown. 

Origin of the dorsal much nearer base of caudal than tip of snout, 
inserted opposite the interspace between the ventral and anal, the 
fin short and high, its anterior ray slightly shorter than the he^d; 
pectoral fin reaching the ventral; ventral inserted much in advance 
of the dorsal, not reaching the vent; the anal elongate, entirely behind 
the dorsal, its base oblique, free edge somewhat concave, anterior ray 
the longest; caudal fin emarginate, the tip of each lobe pointed; 
caudal peduncle rather short, its depth 2.33 in length of head. 

Scales large, well imbricated; lateral line decurved, extending along 
the lower half of tail. 

Color in alcohol gra)'ish above; belly and lower part of sides silvery, 
with no markings; all the fins whitish. 

Total length 84 mm. 

Described from a specimen from Kotosho (Botel Tobago Island) , 
collected by T. Tada. (Cotype; No. 71 51, Stanford collections). 

Habitat: Botel Tobago Island (Kotosho). 

Measurements of Metzia mesembrina. 



Locality. 


•a 
u 


J3 

S. 
u 

Q 


d 


< 


cC 


V. 

Width 
of Head. 


Interor- 
bital. 

Snout. 




Scales. 

Length, 
Mm. 


Kotosho (Type; 
S. U. No. 7131 
Kotosho (Cotype; 

S. U. No. 7151) 

Kotosho (Cotype; 
S. U. No. 715O 


4 

3-89 

4 


2.8 

3 

3-18 


8 

2, 7 
2. 7 


15 
3. 13 
3. 14 


15 
15 


8 
8 


2 
2 


2.50 
2.85 
2.57 


4 

3-8o 

3.60 


3-50 
3-40 
3.60 


8-38-5 
7-36-5 
8-36-S 


83 
90 

84 



Genus Candidia Jordan & Richardson. 

1909. Candidia Jordan & Richardson, Mem. Carneg. Mus., I\', No. 4, p. 169. 
(Type Opsariichthys barbalus Regan.) 

Body covered with small scales. Lateral line slightly decur\ed, 
running along the lower part of the tail. Dorsal fin short, with seven 
branched rays, inserted opposite the root of ventrals, midway between 
tip of snout and base of caudal. Anal fin with nine branched rays. 
Barbels two, minute, maxillary. The angle of mouth extends beyond 
anterior margin of orbit. Gill-rakers very short, conical. Pharyngeal 
teeth 5, 4, I — I, 4, 5. 

Distribution: Formosa. 



The Fresh Water Fishes of the Island of Formosa. 243 

40. Candidia barbata (Regan). 
Koeko or Gogahii (Formosa). 

1908. Opsariichlhys barbaius Regan, Ann. Mag. Nat. Hist. (8), II, p. 359; Lake 
Candidius, Formosa. 

1909. Candidia barbata Jord.-vn & Richardson, Mem. Carneg. Mus., IV, no. 4, 
p. 169; Lake Candidius (after Regan). 

Head 3.6 in length; depth 3.66; D. 3, 7; A. 3, 9; P. 15; \\ 9; width of 
head 2 in its length; snout 3 in head; interorbital space 2.7; eye 5; 
pectoral 1.5; ventral 1.75; fifty-six scales in the lateral line, three 
between the latter and the root of ventral, eight between lateral line 
and middle of belly; gill-rakers 3 + 8; pharyngeal teeth 5, 4, i — i, 4, 5. 

Body elongate, compressed, curvature of the dorsal profile equal to 
that of the ventral; head rather long, lower parts of operculum and 
suborbicular parts with a number of tubercles; snout truncated in 
front, partially covering the upper lip, its sides provided with conical 
tubercles, of which the anterior ones are larger; a notch in front of eye; 
mouth oblique, extending beyond the anterior border of orbit; upper 
lip thicker than the lower, the proximal half of which is provided with 
a series of large conical tubercles; upper jaw slightly protruding; two 
very short maxillary barbels; eyes anterior and superior; nostrils close 
together, superior, in front of eye. 

Origin of dorsal midway between tip of snout and base of caudal, 
inserted above the origin of ventral, anterior rays longer; pectorals not 
reaching ventrals; anal entirely behind the dorsal, middle rays pro- 
longed, each ray with traces of tubercles; caudal fin forked, each lobe 
sharply pointed; depth of caudal peduncle 2.33 in head. 

Body covered with small scales; lateral line continuous, slightly 
decurved, extending along the lower part of the tail. 

Color in formalin yellowish gray above, paler below, belly whitish; 
basal two-thirds of the membrane of the dorsal black; caudal fin gray- 
ish; the rest of the fin dusky white, with faint black mottlings; a black 
longitudinal band from the nape to the base of caudal. 

Total length 120 mm. 

Described from a specimen from Jitsugetsutan, collected by T. Aoki 
in August, 1916. 

Habitat: Jitsugetsutan (Lake Candidius); Shito, Giran. 

Remarks: The tubercles on the head and anal fin are very distinct 
in male specimens, while the anal of the female is nearly smooth. 
Lower parts of caudal peduncle of male sometimes tuberculated. 



244 



Annals of the Carnegie Museum. 



Measurements of Candidia harbala. 



Locality. 







U-, 










. 










■5'S 


o_' 


3 


i^ 


>' 


T^ OJ 










^■s 


G^-a 


tfl 


15 


9 


2 


2.70 


3 


15 


9 


1.77 


2.71 


2.71 


15 


9 


1.92 


2.66 


3 


14 


9 


2 


2.70 


2.89 


15 


9 


1.8s 


2.77 


3 



Jitsugetsutan 
Jitsugctsutan 
Jitsugetsutan 
Jitsugetsutan 
Shito 



3-6o 
3-48 
3-64 
3-66 
3-54 



3-66 3, 7 3. 9 
3-73|3. 7 3. 9 
3-83' 3. 7 I 3. 9 
3-66|3, 7 I 3. 9 
4 I 3. 7 3. 9 



5 

5-33 

4.80 

5 
4-50 



12-56- 

11-57- 
12-54- 
11-54- 
12-57- 



120 

|i59 
J109 
|ii7 
100 



Genus Hypophthalmichthys Bleeker. 
i860. Hypophthalmichthys Bleeker, Prodr. Cypr., p. 405. (Type Leiicisciis 

molitrix Cuv. & Val.) 
1869. Abramocephalus Steindachner, Wien, Sitzungsb., LX, p. 383. (Type 

Ahramocephalus microlepis Steindachner.) 
1872. Onychodon Dybowsky, Verb. ZooL-Bot. Ges. Wien, XXII, p. 211. (Tjpe 

Cephalus manischiiriciis Basilewsky.) 

Body stout, compressed, back rounded, abdomen strongly com- 
pressed, with a sharp keel from throat to vent. Head rather small; 
mouth anterior, broader than deep, its angle not reaching the orbit; 
lips thin; barbel none. Eye situated in the lower half of the head, 
its lower margin being below the level of the angle of mouth. Gill- 
rakers continuous, forming a broad, cresccntic, horny membrane, its 
basal portion perforated. Dorsal fin short, inserted behind the 
origin of ventral; anal fin triangular, entirely behind the dorsal. 
Gill-membranes united, forming a broad bridge across the isthmus. 
Scales small, about 115 in the lateral line. Pharyngeal teeth in one 
row, 4-4, compressed on the longitudinal axis of the bone. Lateral 
line decurved, running along the middle of the tail. 

Distribution: China; Indo-China; Formosa; Amur Province. 

41. Hypophthalmichthys moUtrix (Cuv. & Val.). 
I-lcnhii (F^ormosa). 

1844. Leiiciscus molitrix Cuv. & Val., Hist. Nat. Poiss., XVII, p. 360. — Richard- 
son, Ichtbyol. China, 1846, p. 259; Canton, China. 

1844. Leiiciscus hypophthalmiis (Gray) Rich.\rdson, Iclithyol. Voy. Sulph., 
p. 139, PI. 63, Fig. i; Canton. 

1855. Cephalus mantschuricus Basilewsky, Mem. Sec. Nat. Mosc, X, p. 235, 
T. VH, Fig. 3; Manchiuia. 

1860. Hypophthalmichthys molitrix Bleeker, Ichth. Arch. Ind. Prods. II, Cyprin._ 
p. 288. 

1863. Hypophthalmichthys molitrix Bleeker, Atlas Cyprin., Ill, p. 28. — ^Gijnther, 
Cat. Fish. VII, 1868, p. 298; Cliina. — Bleeker, Mem. Cyprin. China, 1871, 



The Fresh Water Fishes of the Island of Formosa. 245 

p. 83, PI. XII, Fig. i; Vang-tzc-kiang.- — Gunther, Ann. Mag. Nat. Hist., 

Sept., 1889, p. 223; Ichang.^Ann. Mag. Nat. Hist. (7), I, 1898, p. 362; 

Newchang. — Berg, Ichthyol. Amur., 1909, p. 154; Amur Provinces. 
1872. Onychodon mantschuricus Dybowski, Verh. Zool.-Bot. Ges. Wien, XXII, 

p. 211; Ussuri. 
1878. Hypophlhalmichthys dabryi Bleeker, Versl. en Medcdcl. Konin. Akad. 

Wetensch. Amst. (2), XII, p. 210. 

Head 3.58 in length; depth 3.25; D. 3, 7; A. 3, 12; P. I, 17; V. 8; 
width of head 1.48 in its length; eye 6 in head; interorbital space 2; 
snout 3.16; pectoral 1.30; ventral 1.48; one and fifteen scales in the 
lateral line, twenty-eight scales in an oblique series between origin of 
the dorsal and lateral line, twenty-one scales between the latter and 
the middle of belly, fifteen scales between lateral line and the root of 
ventral; pharyngeal teeth 4-4. 

Body stout, compressed, back rounded, abdomen strongly com- 
pressed, with a sharp keel from throat to vent; head moderate, smooth, 
postoperculum with radiated striae; snout blunt, obtusely rounded 
anteriorly; mouth anterior, broader than deep, its angle not extending 
to the orbit; lips rather thin; lower jaw slightly longer than the upper; 
eyes rather small, anterior and inferior; nostrils close together, superior; 
pharyngeal teeth stout, high, laterally compressed, inner surface with 
a large oval concavity of brown color; gill-membranes strongly extend- 
ing beyond the gill-covers united across the isthmus; gill-rakers con- 
tinuous, forming a broad, crescentic, horny membrane, its basal portion 
perforated. 

Origin of dorsal midway between tip of snout and base of caudal, 
very short, anterior ray the longest, when depressed its tip reaching 
beyond all other rays; pectoral fin armed with a smooth spine, reaching 
the root of ventral; origin of ventral much in advance of that of dorsal, 
rather slender, not reaching the vent; anal fin triangular, entirely 
behind the dorsal, its origin nearer to that of ventral than base of 
caudal; caudal fin deeply emarginate, tip of each lobe sharply pointed; 
caudal peduncle elongate, its depth 1.5 in the length of head. 

Scales minute, cycloid; lateral line decurved, continuous, extending 
along the middle of tail. 

Color in formalin grayish above, sides and belly silvery; dorsal and 
caudal fins pale gray; upper surface of the pectoral speckled with fine 
black spots, lower surface whitish; other fins whitish. 

Total length 370 mm. 

Described from a specimen from Shori, Toyen. 



246 



Annals of the Carnegie Museum. 



Habitat: Bred in ponds throughout the island. 

Remarks: The present species is one of the important food-fishes, 
though it is not a native of the island. In the spring, young of H. 
molitrix are imported from Southern China and are bred artificially 
like other Chinese fishes. 



Measurements of Hypophthalmichthys molitrix. 



Locality. 


X 


s. 

v 


d 


<i 


Oh* 


> 


'0 . 


2«' 


3 

s 


lU* 

W 


15 


Kb S 




3.58 
3.14 


3-25 
3.33 


3. 7 
3. 7 


3. II 
3. II 


I, 17 
1,17 


8 
8 


I.41 

1-75 


2 
2.20 


3.16 
3 


6 
6 


2 8-1 I 5-2 I 
2 8-1 19-2 I 


370 
330 


Giran 



Aristichthys gen. nov. 
Type Leiiciscus nohilis (Gray) Richardson. 

Body stout, compressed, rather high in front; abdomen rounded, 
with a keel only in the postventral part. Head Jarge; snout rather 
short, obtusely rounded. Mouth oblique, anterior, its angle reaching 
below the center of eye; lips thin; barbel none. Eye inferior and much 
anterior. Gill-rakers separated, slender and long, set very closely, 
with many membranous septa. Dorsal fin short, inserted behind the 
origin of ventral; anal fin triangular, entirely behind the dorsal. 
Gill-membranes united, forming a broad bridge across the isthmus. 
Scales small, about 115 in the lateral line. Pharyngeal teeth in one 
series, 4-4, strongly compressed laterally. Lateral line strongly 
decurved anteriorly, running along the middle of the tail. 

Distribution: Formosa; South China. 

Remarks: The present genus is a near relative of Ilypophthalniichthys. 
It differs distinctly from the latter in having clearly separated gill- 
rakers, large head, and rounded abdomen, which has no keel in front 
of the ventral. 

42. Aristichthys nobilis (Richardson). 

Chikuyoren (Formosa). 

1844. Leuciscus nohilis (Gray) Rich.^rdson, Ichthyol. \'oy. Sulph., p. 140, PL 63, 
Fig. 3; Canton, China. 

1866. Cephalus Jiypophlhalmus Steindachner, Verb. Zool.-Bot. Gesell. Wien, p. 
383; Hongkong. 

1867. H y po phthalmichlhys manchiiricus Kner, Novara Fisch, III, p. 350; Shang- 
hai. 

1868. HypoplUhahnichthys nohilis GtJNTHER, Cat. Fish., VII, p. 299; Amoy. — 



The Fresh Water Fishes of the Island of Formosa. 247 

Bleeker, Mem. Cyprin, Chine, 1871, p. 85; Yang-tze-kiang. — Gunther, 
Ann. Mag. Nat. Hist., Sept., 1873, P- 249; China. — Peters, Monatsb. 
Ak. Berlin, 1880, p. 926. — Sauvage, Bull. Soc. Philom., 1881, p. 7; Swatow. 
— Gunther, Ann. Mag. Nat. Hist. (6), IV, 1889, p. 228; Yang-tze-kiang. — 
RUTTER, Proc. Acad. Nat. Sc. Philad., 1897, p. 60; Swatow. 

Head 2.8 in length: depth 3.28; D. 3, 7; A. 3, ii; P. I, 19; V. I, 7; 
width of head 1.76 in its length; eye 7.75 in head; interorbital space 
2.11; snout 2.5; pectoral 1.28; ventral 1.83; one hundreJ and fifteen 
scales in the lateral line, twenty-five scales in an oblique series 
between origin of dorsal and lateral line, twenty-five scales between 
the latter and the middle of belly, seventeen scales between lateral 
line and root of ventral; pharyngeal teeth 4-4. 

Body compressed, rather high in front; abdomen rounded, post- 
ventral part with a keel; head large, postoperculum with radiated 
striae; snout rather short, broad, obtusely rounded anteriorly; mouth 
oblique, anterior, its angle reaching below the center of eye; lower 
jaw more or less protruding; middle part of the upper lip thick; eye 
inferior, much anterior; nostrils close together, superior, in front of 
eye above; gill-openings very large, with broad gill-membranes which 
are united on the throat and not attached to isthmus; pharyngeal 
teeth very high, strongly compressed laterally, inner surface with a 
large oval concavity; gill-rakers slender and long, set very closely, 
with many membranous septa. 

Origin of dorsal nearer to base of caudal than tip of snout, rather 
short, anterior ray the longest, the depressed tip of the ray reaching 
beyond the others; pectoral large, reaching beyond the root of ventral, 
armed with a smooth osseous ray; ventral slender, the tip reaching 
vent, its origin in advance of that of the dorsal; anal fin entirely behind 
the dorsal, triangular, external margin more or less concave, inserted 
nearer origin of ventral than base of caudal; caudal fin deeply emar- 
ginate, tip of each lobe pointed; caudal peduncle elongate, its depth 3 
in the length of head. 

Body covered with small cycloid scales; lateral line strongly de- 
curved in front; extending along the middle of the tail. 

Color in formalin grayish above, paler below; sides and lower parts 
silvery; dorsal, anal, and caudal fins grayish, speckled with minute 
black spots; lower surface of pectorals and ventrals white, upper 
surface grayish and finely spotted with black. 

Total length 455 mm. 



248 Annals of the Carnegie Museum. 

Described from a specimen from Shori, Toyen. 

Habitat: The present species is not a native of Formosa. Propagated 
artificially throughout the island. 

Remarks: Aristichtyhs nobilis is a native of Southern China. Its 
young which are collected in the rivers near Swatow, Amoy, or Foo- 
chow are imported to Formosa and are bred in ponds, mingling with 
Hypophthalmichthys molitrix. Sometimes it reaches an enormous 
size, though it never spawns in Formosa. 

Genus Chanodichthys Bleeker. 

i860. Chanodichthys Bleeker, Nat. Tijdschr. Ned. Ind., XX, p. 432. (Type 

Leplocephaliis mongolicus Basilewsky.) 
1865. Parahramis Bleeker, Nedrl. Tijdschr. Dierkunde, II, p. 21. (Tj-pe 

Abramis pekinensis Basilewsky.) 

Body oblong, very greatly compressed; scales large. Snout con^■ex; 
profile of the nape convex; mouth small, lower jaw not protruding, 
upper jaw more or less overlapping the former. Eye very large. 
Gill-openings not extending as far as the orbit below. Lateral line 
slightly curved. Dorsal fin short, inserted somewhat nearer to tip of 
snout than base of caudal, armed with two smooth, strong spines. 
The anal elongate, with numerous rays. Pharyngeal teeth 4, 4, 2 — 
2, 4, 4. 

Distribution: Formosa; China; Amur Province. 

43. Chanodichthys macrops Giinther. 
Toabakon (Formosa). 
1868. Chanodichthys macrops Gunther, Cat. Fish., VII, p. 326; Formosa. 

Head 4.31 in length; depth 3.375; D. II, 7; A. 3, 23; P. 16; V. 9; 
width of head 1. 81 in its length; eye 3 in head; intcrorbital space 3.25; 
snout 3; pectoral 1.14; ventral 1.28; sixty scales in the lateral line, 
eleven scales in an obliciue series between origin of dorsal and lateral 
line, eight scales between the latter and the middle of belly, five scales 
between lateral line and the root of \entral; pharyngeal teeth 4, 4, 2— 
2, 4, 4; gill-rakers 3 + 9. 

Body strongly compressed, rather deep, postventral part carinate, 
dorsal profile abruptly arched behind the occiput; head rather small; 
snout obtusely pointed, as long as the diameter of e\e; mouth subin- 
ferior and oblique, its angle reaching beneath the hind margin of 
nostril; lips thin; lower jaw shorter than the upper; eye \cr\ large, 
anterior and lateral; nostrils close together, large, in front of eye. 



The Fresh Water Fishes of the Island of Formosa. 249 



Origin of dorsal nearer tij) of snout than base of caudal, armed with 
two smooth spines, of wliirh the .second is very strong, anterior ray 
longest; pectoral fin long, with a fleshy Hap, almost reaching the root 
of ventral; ventral fin inserted in advance of that of dorsal, with a 
scaly flap; anal fin elongate, rays numerous, entirely behind the dorsal, 
posterior rays very low; caudal fin strongly emarginate, tip of each 
lobe sharply pointed; caudal peduncle rather short, its depth 2.25 in 
length of head. 

Scales moderate, cycloid; lateral line more or less decurved, ex- 
tending along lower half of the tail. 

Color in formalin dark gray above, lower half of sides and belly 
silvery; all the fins except ventrals dusky. 

Total length 207 mm. 

Described from a specimen from Tamusui River near Shinten, 
collected by T. Aoki in December, 1916. 

Habitat: Tamusui River (Shinten and Heirinbi). 

Remarks: The present species is very closely allied wdth Chanodicli- 
thys stenzi from Kiautschau, China, difi"ering from it in having a 
smaller number of anal rays. 



Measurements 


OF ChanoL 


iicht 


hys macrops. 






Locality. 


T3 

X 


0. 

Q 


Q 


■i. 


cu 


> 


-si 

I; 




u 
— 

c3 


a 

c 

in 




Scales. 

Length, 
Mm. 


Shinten 


4-31 
4.10 
4.16 


3-37 
3-40 
3-42 


n, 7 


3. 23 


16 
16 
15 


9 
9 
9 


1. 81 

2 

2.09 


3-25 
3-45 
3 


3 
3 
3 


3 1 1 1-60-8 207 


Shinten 

Heirinbi 


II. 7 
II. 7 


3. 22 
3.23 


3 

2.75 


I 1-58-8 185 
IO-61-8 122 



Genus Culter Basilewsky. 
1855. Culler B.\siLE\vsKY, Nouv. Mem. Soc. Nat. Moscou, X, p. 236. (Type 
Culler alburnus Basilewsky as restricted by Giinther.) 

Body oblong, much compressed, the entire or postventral abdominal 
edge being trenchant. Scales of moderate or small size; lateral line 
without conspicuous curvature. Mouth directed upwards; barbels 
none. Dorsal fin short, with strong smooth spines, inserted above 
the interspace between ventral and anal; anal fin long, many-rayed; 
caudal fin forked, pectorals elongate. Gill-openings very wide; gill- 
rakers long, setiform. Pseudo-branchise present. Pharyngeal teeth 
in a triple series, slender and hooked. Intestinal tract short. Air- 
bladder tripartite. (Giinther). 

Distribution: China; Formosa; Corea; Amur Province. 



250 Annals of the Carnegie Museum. 

Synopsis of the Formosan Species. 
a. Postventral edge only carinate; scales about 85 in the lateral line; 18 scales 
between lateral line and the origin of dorsal; 10 scales between lateral line and 

the root of ventrals aokii. 

aa. Abdomen entirely carinate; scales about 64 in the lateral line; 11-12 scales 
between lateral line and the origin of dorsal; 6 scales between lateral line and 
the root of ventrals brevicauda. 

44. Culter aokii sp. nov. (Plate LII, Fig. i). 
Kyauyo (Formosa). 

Head 4.05 in length; depth 4.6; D. Ill, 7; A. Ill, 23; P. 15; V. 9; 
width of head 2.6 in its length; eye 4 in head; interorbital space 6; 
snout 3.75; eighty-six scales in the lateral line, eighteen scales in an 
oblique series between origin of dorsal and lateral line, ten scales 
between the latter and the middle of belly, six scales between lateral 
line and the root of ventral; pharyngeal teeth 5, 4, 2 — 2, 4, 5; gill- 
rakers 5 + 22. 

Body oblong, much compressed, postventral abdominal edge 
carinate, dorsal profile convex; head rather long, laterally compressed, 
the top osseous, with two bony ridges between the eyes; interorbital 
space very narrow, slightly convex; many mucous cavities below and 
behind the orbit; snout bony, tip swollen, a slight depression in front 
of eye above, tip of its skin not overlapping the upper lip; mouth 
anterior, subvertical, its angle not reaching the vertical through 
anterior margin of orbit; lips thin; lower jaw more or less protruding; 
mentum provided with two strong osseous ridges w^hich are united 
in front, extending backward to the operculum; isthmus entirely 
hidden beneath those ridges; eyes large, anterior; nostrils close to- 
gether, more or less superior; gill-openings very large, gill-membranes 
entirely separated; gill-rakers slender and long. 

Origin of dorsal in a point midway between tip of snout and base 
of caudal, with three smooth spines, the first spine very short, hidden 
beneath the skin, second shorter than half the length of the third, 
anterior ray the longest; pectoral armed with an osseous spiny ray, 
reaching the base of ventral; origin of ventral in advance of that of 
dorsal; anal fin entirely behind the dorsal, elongate, length of its base 
1. 1 7 in head; caudal fin emarginate, the tip of each lobe sharply 
pointed; caudal peduncle elongate, its«depth 3 in length of head. 

Body covered with thin cycloid scales; lateral line continuous, very 
slightly decurved, extending along near the middle of tail. 



The Fresh Water Fishes of the Island of Formosa. 251 

Color in alcohol pale gra\- above, bell\- and lower parts of the sides 
sih'ery; caudal fin grayish, other fins dusky white. 

Total length 280 mm. 

Described from a specimen from Jitsugetsutan, collected by T. 
Aoki in August, 1916. 

Habitat: Restricted to Jitsugetsutan (Lake Candidius). 

Remarks: This species is most nearly allied to Culler sieboldi Dy- 
bowski* from Amur Province. The differences are as follows: a lesser 
number of scales in the lateral line, a lesser number of scales between 
lateral line and the middle of belly, as well as small size of the body. 
The above described species is one of the largest forms which is found 
in Lake Candidius. 

Named for Mr. Takeo Aoki of the Bureau of Fisheries, Government 

of Formosa. 

Measurements of Culler aokii. 



Locality. 



Jitsugetsutan.. 4.06 4.60 III, 7 III, 23 I, 15 9 2.60 6 I 3-75 4 18-86-10 
Jitsugetsutan.. 3.88 4.33I1II, 7 III, 22 I, 14I 9 









2.60 6 3-75 4 18-86-10 
2.60 5-89! 3-71 3-8518-85-10 



280 

254 



45. Culter brevicauda Giinther. 

1868. Culler brevicauda Gunther, Cat. Fish., VII, p. 329; Formosa. — Bleeker 
Mem. Cyprin. Chine, 1871, p. 69, Tab. XI, fig. 3; Yang-tze-kiang. — GtJN- 
ther, Ann. Mag. Nat. Hist., Sept., 1873, P- 250; Shanghai. 

Head 4.38 in length; depth 3.79; D. II, 7; A. 3, 27; P. 16; V. 9; 
width, of head 2.5 in its length; eye 4 in head; interorbital space 4; 
sixty-four scales in the lateral line, eleven scales in an oblique series 
between origin of dorsal and lateral line, eight scales between the 
latter and the middle of belly, six scales between lateral line and the 
root of the ventral; pharyngeal teeth 4, 4, 2 — 2, 4, 4; gill-rakers 
6 + 23. 

Body much compressed, abdominal edge entirely carinate, dorsal 
profile broadly convex, top of the head more or less fallen from the 
back, ventral profile undulating at the base of the ventral; head 
rather small, narrow, pointed, its top more or less convex; snout as 
long as the diameter of eye, truncated in front; mouth anterior and 

^Culler sieboldi Dybowski, Verh. Zo6].-Bot. Gesell. Wien, XXII, 214; Middle 
Amur; Ussuri; Sungari; Chanka. 



252 Annals of the Carnegie Museum. 

oblique, its angle reaching the nostril below; upper lip thin; lower jaw 
slightly protruding; eyes moderate, anterior; nostrils close together, 
in front of the eye above. 

Origin of dorsal nearer base of caudal than tip of snout, opposite 
the interspace between ventral and anal, armed with two smooth 
spiny rays, the second spine is very strong; pectoral elongate, reaching 
beyond the base of ventral; ventrals rather slender, inserted much in 
advance of that of dorsal; anal fin very long, anterior ray the longest; 
caudal fin strongly forked, the tip of each lobe sharply pointed; 
caudal peduncle rather short, its depth 2.33 in the length of head. 

Body covered with moderate cycloid scales; lateral line slightly 
decurved, extending along the lower half of the tail. 

Color in alcohol grayish above, belly and lower half of the sides 
silvery; fins dusky white. 

Described from a specimen from Kagi, collected by Y. Kikuchi. 

Habitat: Kagi (a single specimen). 

Remarks: This species is closely related to Ciilter recitrviceps (Rich- 
ardson)^ from which it differs in having an entirely carinate abdominal 
edge and a lesser number of scales in the lateral line. 

CULTRICULUS gen. nov. 
Type Ciilter lencisculus Kner (not of Basilewsky) = IlemicuUer kneri 

Kreyenberg. 

Body oblong, much compressed, abdominal edge entirely carinate. 
Scales of moderate size; lateral line continuous, abruptly bending 
downward above the pectoral, scarcely reaching the tip of the fin, 
thence advancing backward, ascending gradually, and running along 
the middle of the sides of the tail. Mouth oblique; jaws subequal; 
barbels none. Dorsal fin short, witji no smooth spines, inserted 
behind the origin of ventral; pectoral moderate, not reaching the 
ventral; anal fin rather short, with fifteen to seventeen rays. Gill- 
rakers slender and long. Pharyngeal teeth 5, 4, 2 — 2, 4, 5. 

Distribiilio)i: Formosa; China; Indo-China. 

Remarks: Bleeker's IlemicuUer is the genus most closely related 
to Cidtricidus. It differs from the present genus in having the 
abdomen non-carinate or only partially carinate (postventral part 
only). 

^ Leuciscus rcciirviceps Richardson, Ichthyol. China, 1845, p. 259; Canton, China. 
Culliir recurviceps Giinthcr. Cat. Fish., VII, 1868, p. 328, China. 



The Fresh Water Fishes of the Island of Formosa. 253 

46. Cultriculus kneri (Kreyenbcrg). 
Unahii or Kirara (Formosa). 

1867. Culler leiicisctilus Kner, Novara Fisch., Ill, p. 362; Shanghai. 

1868. Chanodichlhys leucisciiliis Gunther, Cat. Fish., VII, p. 327; Shanghai 
(after Kner.) 

1873. Hemiculler leucisculus Gunther, Ann. Mag. Nat. Hist. Sept., p. 249; 

Shanghai. — Ann. Mag. Nat. Hist., 1888, p. 433; Yang-tzc-kiang. — Berg, 

Ichthyol. Amur., 1909, p. 146. 
1908. Hemiculler kneri Kreyenberg, Berhn Sitzb. Ges. Natt. Freunde, p. 105 

(nom. nor. for Culler leucisculus Kner). 

Head 4.56 in length; depth 4.56; D. II, 7; A. 2, 11 ; P. 15; V. 9; 
width of head 2.17 in its length; eye 4 in head; interorbital space 3.5; 
snout 3.25; pectoral i.ii; ventral 1.5; fifty-two scales in the lateral 
line, eight scales in an oblique series between origin of dorsal and lateral 
line, 3.5 scales between the latter and the middle of belly, 2 scales 
between lateral line and the root of the ventral; phar}-ngeal teeth 
5, 4, 2—2, 4, 5; gill-rakers 4 + 17. 

Body elongate, much compressed; entire abdominal edge carinate; 
head moderate, narrow; snout pointed anteriorly, its tip slightly 
swollen; mouth anterior and oblique, its angle scarcely reaching the 
anterior margin of the nostril below; lower jaw slightly shorter than 
the upper, with rather sharp edge; eyes moderate, anterior and supe- 
rior; nostrils close together, in front of eye above. 

Origin of dorsal about midway between tip of snout and base of 
■caudal, inserted behind that of the ventral, with two smooth spines, 
of which the second is stronger, anterior ray the longest; pectoral iin 
moderate, not reaching the ventral; ventrals slender, inserted in front 
of the origin of dorsal; anal entirely behind the dorsal, rather short, 
triangular, anterior ray the longest; caudal fin bifurcate, tip of each 
lobe sharply pointed; caudal peduncle elongate, its depth 2.66 in 
length of head. 

Body covered with thin cycloid scales; lateral line continuous, 
abruptly bending downward above the pectoral; thence passing back- 
ward, making a weak curve, extending along the middle of the sides 
of the tail. 

Color in formalin oli\'e-gray above, lower half of the body white; 
dorsal and caudal fins grayish, other fins white. 

Total length 182 mm. 

Described from a specimen from Jitsugetsutan, collected by 
T. Aoki in August, 1916. 



254 Annals of the Carnegie Museum. 

Habitat: Jitsugetsutan (Lake Candidius) ; Shimotamusui River. 
Remarks: According to the description by Kner the type of the 
present species has 4, 3, 2 — 2, 3, 4 pharyngeal teeth instead of 5, 4, 2 — 

2, 4> 5- 

The Chinese people who live near Lake Candidius use two ver- 
nacular names for the present species, namely, " Unahii" and " Kirara." 
They treat these two as different fishes. But there is no doubt that 
"Kirara" is the young form of "Unahii," for no morphological dif- 
ferences exist between them. 

Description of "Kirara." 

Head 4.5 in length; depth 5; D. H, 7; A. 2, II ; P. 15; V. 9; width- of 
head 2.33 in its length; eye 3 in head; interorbital space 3.5; snout 3.5; 
ventral 1.5; pectoral as long as the head; fifty-two scales in the lateral 
line, eight scales between origin of dorsal and lateral line, four scales 
between the latter and the middle of belly, two and one-half scales 
between lateral line and the root of the ventral; pharyngeal teeth 
5, 4, 2—2, 4, 5; gill-rakers 5 + 16. 

Body slender, elongate, compressed, entire abdominal edge carinate; 
head moderate, narrow, its top very slightly convex; snout pointed 
anteriorly, upper surface flat, tip swollen; mouth anterior and oblique, 
its angle scarcely reaching a vertical through anterior margin of 
nostril; lower jaw slightly shorter than the upper, with rather sharp 
anterior edge; eyes large, anterior; nostrils close together, supralateral, 
the anterior nostril in a short tube. 

Origin of the dorsal midway between tip of snout and base of caudal, 
inserted behind the origin of the ventral, with two smooth spines; 
height of the fin ecjual to the length of head; pectoral elongate, scarcely 
reaching the root of ventral; ventral fin small, inserted in advance of 
the dorsal; anal fin entirely behind dorsal, triangular, anterior ray 
the longest; caudal fin slender, elongate, deeply emarginate, tip of 
each lobe sharply pointed; caudal peduncle elongate, strongly com- 
pressed laterally, its depth 2.75 in the length of head. 

Body covered with thin cycloid scales; lateral line continuous, 
abruptly bending downward from the nape to the tip of the pectoral, 
thence passing backward and ascending gradualh', running along the 
middle of the sides of tail. 

Color in formalin pale olive-gray, lower j^arts whitish; top of head 
brown; a dark brown longitudinal band runs from nape to tiie base 



The Fresh Water Fishes of the Island of P'ormosa. 255 



of the caudal; median dorsal line with a pale brown hand; all the fins 
whitish. 

Length of body 64 mm. 

Described from a specimen from Jitsugetsutan, collected by T. Aoki 

in August, 1916. 

Measurements of CuUriculus kneri 



Locality. 



Jitsugetsutan 

Jitsugetsutan 

Jitsugetsutan 

Jitsugetsutan 

Shimotamusui River 
Shimotamusui River 



4-56 
4-34 
4-45 
4-50 
4-50 
4-35 



4.56 II, 7 
4.86 II, 7 
4.90 II, 7 

5 jn. 7 

4-50 II. 7 
5 In, 7 



2, II 
2, 12 

2, 13 

3. II 
2, 12 
2, 12 



2.17 
2.20 
2.44 
2.33 
2.33 
2.66 



o_ 

(U 2 

c3 



3-50 
330 
3.66 
3-50 
3.38 
3-6o 



3-25 
3-56 
3-66 
3-50 
3-71 
3-6o 



4 

3-56 

3-66 

3 

3-71 

3.60 



8-52-3I 

8-51-4 

8-52-4 

8-52-4 

8-50-4 

8-49-3 



S06 



182 
170 
117 

64 
145 

90 



Doubtful Species. 
In the year 1903, Jordan and Evermann mentioned two species 
of cyprinoid fishes from Formosa under the name Cirrhina sp. and 
Dillonia sp. (Proc. U. S. Nat. Mus., XXV, p. 322 and 324). For the 
sake of completeness I give the original descriptions by those authors. 
I have no specimens which belong to the general Cirrhina and Dillonia; 
therefore nothing more can be said of them at present. 

Cirrhina sp. Jordan & Evermann. 
"Closely allied to Cirrhina chinensis Giinther. D. 15; A. 7; scales 
37; teeth 5, 4, 2. (No. 837, Formosa; Imperial Fisheries Institute, 
Japan.)" 

Dillonia sp. Jordan & Evermann. 

"Allied to Dillonia aculeata Cuvier and Valenciennes. Head shaped 
as Scaphiodon. Mandibles with barbels; D. 10; A. 11 ; scales 39." 

Family PCECILIID.^. 

Genus Oryzias Jordan & Snyder. 

1906. Oryzias Jordan & Snyder, Proc. U. S. Nat. Mus., XXXI, p. 289. (Type 
Pcecilia latipes Temminck & SchlegeL) 

Body elliptical in form, compressed, covered with large scales; 
mouth small, with two rows of small, simple, pointed teeth; no teeth 
on vomer; gill-opening not restricted above; intestinal canal short, 
about as long as body; peritoneum black. Dorsal fin short, inserted 
above middle of anal; anal very long, having from seventeen to twenty 



256 Annals of the Carnegie Museum. 

rays; caudal fin truncate. Sexes similar, except in color; anal fin not 
modified in the male. (Jordan & Snyder.) 
Distribution: Japan; Corea; Formosa. 

47. Oryzias latipes (Temminck & Schlegel). 
Medaka (Japan); Tamhii (Formosa). 

1846. Pcecilia talipes Temminck & Schlegel, Fauna Japonica, Poiss., p. 224, PL 

CII, Fig. 5; Nagasaki. 
1866. Haptocheilus talipes Gijnther, Cat. Fish., \T, p. 311, Nagasaki. 
1901. Aploclieilus talipes Jord.\n & Snyder, Proc. U. S. Nat. Mus., XXIII, 

P- 350. 
1906. Oryzias talipes Jordan & Snyder, Proc. U. S. Nat. Mus., XXXI, p. 289; 

Japan. — Snyder, Proc. U. S. Nat. Mus., XLII, 191 2, p. 407; Shiogama; 

Yamaguchi; Akune; Nanao; Dogo Island. — Jordan & Metz, Mem. Carneg. 
■ Mus., VI, no. 2, 1913, p. 24; Fusan; Suigen, Corea. — Jordan, Snyder, & 

Tanaka, Journ. Coll. Sci. Tokyo, XXXIII, 1913, p. 91; Japan. 

Head 4 in length; depth 4.5; depth of caudal peduncle 9.5; eye 2.5 
in head ; interorbital space 2 ; snout 4 ; D. 6 ; A. 1 8 ; P. 9 ; V. 5 : thirty-one 
scales in a lateral series; five branchiostegals. 

Posterior half of the body compressed, becoming broader anteriorly, 
highest in front of the anal; head flattened; interorbital space broad; 
snout shorter than the diameter of eye, broadly rounded anteriorly; 
mouth anterior, transverse; lower jaw slightly projecting, each jaw 
with two rows of minute pointed teeth, those on the posterior row 
smaller; vomer smooth; thirteen short, pointed gill-rakers on the first 
arch; eyes very large, anterior and superior. 

Dorsal fin short, on the posterior half of body, its origin above the 
posterior two-thirds of anal, its height ec}ual to the distance between 
tip of snout and posterior margin of orbit; pectoral inserted on the 
median line of body, its length contained 5.5 in the length of body; 
the ventral small, reaching vent; base of the anal very long, its pos- 
terior end opposite to that of the dorsal, anterior ray longest; tip of the 
caudal fin rounded. 

Top and sides of head, throat, and chin naked; body covered with 
thin cycloid scales; lateral line absent. 

Color in formalin pale gray above, lower parts silvery; a black 
longitudinal streak from the nape to the origin of the dorsal; sides of 
body with a faint dusky stripe along the middle line, top of head 
dark; the edges of scales dusky; fin-rays of the ventral and anal dotted 
with minute black spots; all the fins whitish; jieritoneum black. 

Length of body 28 mm. 



i 



The Fresh Water Fishes of the Island of F'ormosa. 257 



The present description is from a specimen from Shori, collected by 
T. Aoki in February, 1917. 

Habitat: The present species is very common in rice-fields and pools 
on the island. My specimens came from Shori; Ako; and Giran 
(Kizanto and Taiko). 

Measurements of Oryzias latipes. 



Locality. 


•a 


.5 
0. 

V 




a 
T3 
3 a 


Q 


<i 


a," 


>■ 


" 


1 


V 

>> 



t/3 


J3 . 


Shori 


4 

3-33 

4 

3-66 

3-88 

3-50 

3-50 


4-50 

4 

4 

4-33 

4.40 

4-50 

4 


9-50 

9-50 

9 

9 

8.66 

9 

8.66 


6 
6 
6 
6 
6 
6 
6 


18 
18 
18 
18 
17 
17 
17 


9 
9 
9 
9 
9 
8 
9 


5 

5 
5 
5 

5 
5 
5 


2 

233 
2.33 
2.50 

2.33 

2 

2 


4 

3-33 

3 

3 

3 

3 

3 


2.50 
2.33 
2.33 
2.33 
2.33 
2.50 
2.33 


31 
31 
31 
31 
30 
30 
29 


-^8 


Kizanto 


32 

40 
27 
35 
21 
20 


Kizanto 


Kizanto 

Kizanto 

Taiko 

Taiko 



Genus Gambusia Poey. 
1855. Gambusia Poey, Mem. Cub., I, p. 382. (Type Gambusia punctata Poey.) 

Body moderately elongate, becoming deep in the adult female. 
Mouth moderate, the lower jaw projecting, the bones well joined; 
both jaws with a band of pointed teeth which are not movable; snout 
not produced. Eyes normal, not divided. Scales large. Gill- 
openings not restricted. Dorsal and anal fins both rather short and 
small, the anal more or less in advance of the dorsal; anal fin of the 
male much advanced and modified into a long intromittent organ, 
which is about as long as head. Intestinal canal short. Six branchios- 
tegals. X'ertebrae about thirty-two. (Jordan & Evermann.) 

Distribution: Mexico; Cuba; Southern States of North America. 

48. Gambusia affinis (Baird & Girard). 
Top-minnow. 

1853. Heterandria affinis Baird & Girard, Proc. Ac. Nat. Sci. Philad., p. 390; 

Texas. 
1859. Gambusia speciosa Girard, Proc. Ac. Nat. Sci. Philad., p. 121; Rio San 

Diego, New Ulm, Mexico. 
1859. Gambusia gracilis Girard, Proc. Ac. Nat. Sci. Philad., p. 121; Mexico. 
1866. Gambusia humilis GOnther, Cat. Fish., VI, p. 334. 
1866. Gambusia affinis Gunther, Cat. Fish., VI, p. 336; Texas. — Evermann & 

Kendall, Bull. U. S. Fish. Comm., XII, 1892, p. 107, PI. 25; Fig. 2. — 

Jordan & Evermann, Bull. U. S. Nat. Mus., 47, 1896, p. 680 (in part). 

18 — dec. 19, 19 19. 



258 Annals of the Carnegie Museum. 

Head 3.8 in length; depth 4.33; D. 7; A. 10; P. 12; V. 6; eye 3 in 
head; interorbital space 2; snout 2.66; pectoral 1.33; ventral 2.5; 
thirty-one scales in a lateral series. 

Posterior part of body compressed, anterior half high arid broad, 
curvature of the dorsal profile stronger than that of the ventral; head 
moderate, broad, depressed; snout short, obtusely rounded anteriorly, 
upper surface flattened; mouth anterior, transverse; lower jaw longer 
than the upper; jaws with a broad band of villiform teeth, palatine 
with a narrow transverse band of villiform teeth; eye moderate, 
superior and slightly anterior; five branchiostegals; thirteen gill-rakers 
on the first arch ; gill-openings large. 

Origin of the dorsal slightly in advance of the end of the base of 
anal, distance to the base of the caudal about half of that to tip of 
snout, first ray shortest; pectorals inserted just below the middle of 
sides, their tips reaching beyond the origin of the ventral; ventral 
fins rather slender; the anal higher than the dorsal, anterior rays 
shorter, middle rays prolonged; caudal fin broad, tip obtusely rounded; 
caudal peduncle rather deep, its depth twice in the length of head. 

Body covered with large scales; snout and cheek smooth; no lateral 
line. 

Color in formalin dusky above; sides and belly whitish; the edges 
of scales dusky; top of head dark; a dark longitudinal stripe from 
occiput to the origin of the dorsal; a very narrow dark band along the 
middle of the sides; no dark marking below the eye; dorsal fin-rays 
speckled with black; all the fins whitish. 

Length of body 44 mm. 

Described from a specimen from the Government Hatchery at 
Shori. (Female). 

Habitat: The present species is not a native of the island. About 
five years ago it was imported from Hawaii in order to exterminate 
anopheline mosquitoes and has been propagated artificially. It was 
introduced from Galveston, Texas, into Hawaii, and later into the 
Philippines by Mr. Alvin Seale, for the purpose of destroying mos- 
cjuitoes. 

Remarks: The male fish is very small; the anal modified into a 
sword-like intromittent organ. Measurements of a male specimen 
are as follows: Head 3.66 in length; depth 4.5; D. 7; A. 8 (third ray 
prolonged); P. 12; V. 6; snout 3 in head; interorbital space 2; eye 2.5; 
total length 24 mm. 



The Fresh Water Fishes of the Island of Formosa. 259 



Measurements of Gambusia affinis. 



Locality. 



-a 


S. 


Q 


<i 


3-80 


4-33 


7 


10 


350 


4 


7 


10 


3-66 


4-13 


7 


10 


3-33 


4.14 


7 


9 


3-66 


4-50 


7 


8 


3-17 


4-25 


7 


8 


3-50 


4.40 


7 


8 





o_: 


a 




V 


> 


c3 




a 


w 


in 


6 


2 


2.66 


3 


31 


6 


2 


2.50 


3 • 


30 


6 


1.8 


3 


3-33 


30 


6 


2 


2.66 


3-33 


31 


6 


2 


3 


2.50 


30 


6 


3 


4 


4 


30 


6 


2 


3 


3 


30 



Shori ( 9 ) . 
Shori ( 9 ) . 
Shori ( 9 ) . 
Shori ( 9 ) . 
Shori (cf). 
Shori (dl). 
Shori (cf). 



44 
41 
40 
37 
24 
23 



Family MONOPTERID^. 
Genus Fluta Bloch & Schneider. 



1798. 



Monopterus Lacepede, Hist. Nat. Poiss., II, p. 139. (Type Monoplerus 

javanensis Lacepede). Not Monopterus of Volta, 1796, a genus of fossil 

fishes. 
1801. Fluta Bloch & Schneider, Ichth., p. 565. (Type Monopterus javanensis 

Lacepede.) 
1845. Ophicardia McClelland, Calcutta Journ. Nat. Hist., V, p. 191. (Type 

Ophicardia pharyriana McClelland.) 
1855. Apterigia Basilewsky, Nouv. Mem. Soc. Nat. Mosc, X, p. 247. (Type 

Apterigia saccogularis Basilewsky.) 

Body elongate, naked; tail short, tapering to a point; no barbels; 
margin of the upper jaw formed by the premaxillaries, the maxillaries 
well developed, lying behind them and parallel with them; lips thick; 
palatine teeth small, in a narrow band; gill-opening confluent into a 
ventral slit, the membrane united to the isthmus; gill-arches three, 
with gill-fringes rudimentary, and with moderate slit between them; 
no accessory breathing sac; lateral line present; no pectoral or ventral 
fins; dorsal and anal reduced to low folds'; ribs present; no air-bladder; 
stomach without caecal sac or pyloric appendages. Ovaries with 
oviducts. (Jordan & Snyder). 



49. Fluta alba (Zuiew). 
Taunagi (Japan); Senhii (Formosa). 

Murcena alba Ziew, Nov. Act. Sci. Petropol, p. 229, PI. VII, Fig. 2. 

Monopterus javanois'LAci.VE.Ti^, Hist. Nat. Poiss., II, p. 139; Java. 

Monopterus javanensis Bloch & Schneider, Syst. Ichth., p. 565, after 
Lacepede. — Cantor, Malayan Fishes, 1850, p. 339, Pi. V, Figs. 6-8. — 
Bleeker, Atlas Ichth. Mur., 1864, p. 118, PI. XL VII, Fig. i; Java; Sumatra; 
Banka; Bintang, Borneo; Celebes.— Gunther, Cat. Fish., VIII, 1870, 
p. 14; Batavia; Borneo; Siam; Formosa; China; Japan. — Day, Fish. Brit. 



1793- 
1798. 
1801. 



260 Annals of the Carnegie Museum. 

India, 1876, p. 656, PI. CLXIX, Fig. i; Burma; China. — Gunther, Ann. 

Mus. St. Petersburg, 1896, p. 219; Kansu; Sze-chuen. — Regan, Ann. Mag. 

Nat. Hist. (7), XIII, 1904, p. 194; Yunnan. 
1803. Unibranchapertiira Icevis Lacepede, Hist. Nat. Poiss., V, p. 658, PI. XVII, 

Fig. 3- 
1846. Monopterus lends Richardson, Voy. Sulph., Ichth., p. 116; Hongkong. 

1845. Ophicardia pharyriana McClelland, Calcutta Journ. Nat. Hist., Y, pp. 
191, 218, PI. XII, Fig. i; River Ganges. 

1846. Monopterus cinereus Richardson, Voy. Sulph. Ichth., p. 117, PI. LII, Figs. 
1-6; Chusan; Woosung. 

1846. Monopterus (?) xanthognathns Richardson, Voy. Sulph. Ichth., p. 118, PI. 

LII, Fig. 7; Canton. 
1846. Monopterus marmoralus Rich.\rdson, Ichthyol. China, p. 315; Chusan. 
1846. Monopterus helvolus Richardson, Ichthyol. China, p. 316; Canton. 
1855. Aptcrigia saccogularis Basilewsky, Nouv. Mem. See. Nat. Mosc, X, p. 247, 

PI. II, Fig. 2; Tschili. 
1855. Apterigia Jtigromaculata B.\sile\vsky, I.e., p. 248, PI. II, Fig. 2; Peking. 
1855. Apterigia immaculata Basilewsky, I.e., p. 248; Peking. 
1897. Monopterus albus Rutter, Proc. Acad. Nat. Sc. Philad., p. 61; Swatow. — - 

Jordan & Snyder, Proc. U. S. Nat. Mus., XXIII, p. 838; Okinawa; Amami- 

Oshima. — Jordan & Everm.^nn, Proc. U. S. Nat. Mus., XXV, 1903, p. 324; 

Hokota, Formosa. — Jordan & Richardson, Mem. Carneg. Mus., IV, 1909, 

p. 171; Hokoto. — Jordan & Metz, Mem. Carneg. Mus., VI, 1913; No. 2. 

p. 24; Suigen, Corea. — Jordan, Snyder & T.\n.\ka, Journ. Coll. Sci. Tokyo, 

XXXIII, 1913, p. 76; Corea; Japan; Riukiu. 

Head 10.17 i" total length, its depth greater than that of the body, 
1.66 in its length; depth 20.33 in total length; tail 3.5 in the length of 
body; snout 5 in head; interorbital space 6. 

Body elongate, compressed, tapering towards the tip of tail, greatest 
diameter at the occiput, dorsal profile ascending suddenly at the nape 
and descending slowly to the tip of snout; cross-section of the body 
oval; head sw^ollen; throat pouch-like; snout rather short, compressed, 
pointed anteriorly; lips broad and fleshy; maxillaries contained twice 
in head; teeth on both jaws granular, forming a band tapering towards 
the angle of mouth; palatine and vomerine teeth granular, forming two 
bands parallel to the former, lower jaw with corresponding bands of 
granular teeth along the inner side of the outer teeth; eyes very small, 
eight times in head and covered by thin skin; nostrils very small, 
separated, one in front of eye above, the other on the extremity of the 
snout; gill-openings inferior, confluent into a \'entral slit; branchial 
arches three. 

Dorsal fin very low, membranous, commencing abo\'e the \ent, 
hind part reaching to tip of tail; anal fin indistinct, about half as long 
as the dorsal; no pectorals and ventrals; tail short and jiointed. 



The Fresh Water Fishes of the Island of Formosa. 261 

Body naked, smooth; lateral line continuous, slightly depressed, 
running along the middle of the sides. 

Color in formalin l)ro\vnish gray above, mottled with darker spots, 
with traces of j)aler and darker streaks; lower surface, lips, and throat 
whitish. 

Length of body 305 mm. 

Described from a specimen from Shokwa, collected by T. Aoki in 
December, igi6. 

Habitat: The present fish is very common in streamlets and canals. 
My specimens came from Jitsugetsutan (Lake Candidius) ; Shokwa; 
Taihoku; Kiburan, Giran. 

Remarks: The color of the lower parts is variable. In two specimens 
from Taihoku the belly is mottled with brown, while another from the 
same locality has a nearly white belly with indistinct brown markings. 

Fluta alba is capable of living a considerable time out of water. 

Measurements of Fluta alba. 



Locality. 


Head. 


Depth. 


Depth of 
Head. 


Tail in 
Trunk. 


Interor- 
bital. 


Snout. 


Length, 
Mm. 


Shokwa 

Taihoku 

Taihoku 

Taihoku 

Kiburan 


10.17 
10.83 
9.71 
11-33 
12.50 


20.33 
21.66 
22.33 
24.28 
23-44 


1.66 
1.66 
1.66 
1.66 
1.68 


3-50 
3-33 
2-26 
2.24 
3-17 


6 

6.50 

6.33 

6.33 

6 


5 

5-40 

5-40 

4-83 

5 


305 
318 
332 
332 
390 



Family ANGUILLID^. 
Genus Anguilla Shaw\ 

1804. Anguilla Shaw, General Zoology, IV, p. 15. (Type Murcena anguilla 
Linnaeus.) 

Body elongate, compressed behind, covered with imbedded scales 
which are linear in form and placed obliquely, some of them at right 
angles to others. Lateral line well-developed. Head long, conical, 
moderately pointed, the rather small eye well forward and over the 
angle of mouth. Teeth small, subequal, in bands on each jaw and a 
long patch on the vomer. Tongue free at tip. Lips rather full, with 
a free margin behind, attached by a frenum in front. Lower jaw 
projecting. Gill-openings rather small, slit-like, about as wide as 
base of pectorals and partly below them. Nostrils superior, well- 
separated, the anterior with a slight tube. Vent close in front of anal. 
Dorsal inserted at some distance from the head, confluent with the 



262 Annals of the Carnegie Museum. 

anal around the tail. Pectorals well-developed. (Jordan & Snyder.) 
Distribution: Cosmopolitan, but not extending into the arctic regions. 

Synopsis of the Formosan Species. 

A. Mandibulary band of teeth is longitudinally divided by a groove, the outer 

strip containing a series of somewhat larger teeth. 
a. Length of head less than the distance between the origins of dorsal and 

anal maiiriliana. 

aa. Length of head longer than the distance between the origins of dorsal and 
anal japonica. 

B. Mandibulary teeth in narrow bands, without longitudinal groove. 

a. Angle of mouth below the posterior margin of eye\ lips thick, .sineyisis. 

50. Anguilla mauritiana Bennett. 
Ounagi (Japan); Roma (Formosa). 

1831. Anguilla mauritiana Bennett, Proc. Comm. Zool. Soc, p. 128. — Gunther, 
Cat. Fish. VIH, 1870, p. 25; East Indian Ocean and Archipelago; Formosa, 
Amboyna; Almorah; Cejion; Philippine Islands; Islands of Johanna. — 
Jordan & Evermann, Proc. U. S. Nat. Mus., XXV, 1903, p. 325; Kotosho, 
Formosa. — Evermann & Seale, Bull. U. S. Bur. Fish., XX\'I, 1907, p. 56; 
Tarlac. — Jordan & Richardson, Bull. U. S. Bur. Fish., XXVII, 1908, 
p. 238; Calayan, Mindoro. — Seale & Bean, Proc. U. S. Nat. Mus., XXXIII, 
1907, p. 239, Zamboanga. — Ishikawa, Journ. Coll. Agric. Tokyo, IV, 1914, 
p. 427; Japan proper; Bonin Island. 

1864. Murana maculata Bleeker, Ned. Tydschr. Dierk. I, p. 237. 

1864. Murana manilensis Bleeker, Atl. Ichthyol. IV, p. 10; Manila. — Jordan & 
Everm,\nn, Proc. U. S. Nat. Mus., XXV, 1903, p. 325; Kotosho, Formosa. 
— Jord.^n & RiCH.\RDSON, Mem. Carneg. Mus., IV, no. 4, 1909, p. 171, 
Kotosho. 

1867. Anguilla (Murcena) marmovala Kner, Novara Fisch., Ill, p. 369; Hongkong. 

Length of head 6.05 in the total length, .89 in the distance of the 
gill-opening from the origin of the dorsal, 1.68 in its distance from the 
vent; distance between origin of dorsal and anal slightly longer than 
head, length of head contained 1.03 in the former; distance from tip of 
snout to the origin of dorsal 3.03 in total length; length of the pectoral 
3.64 in head; snout 4.77; length of upper jaw 2.7; diameter of eye 2.6 
in snout, 2.4 in interorbital space; height of bod\' in front of anus 
15.62 in total length; length of pre-anal part 1.03 in post-anal part. 

Body stout; angle of mouth extending far beyond the posterior 
margin of orbit; lips well developed, fleshy; jaws subequal; teeth on 
both jaws, maxillary and mandibular teeth divided by a longitudinal 
groove into two strips, teeth on both rows in a single series, those on 
the outer row larger; palate with a band of sub-equal, villiform teeth, 



The Fresh Water Fishes of the Island of Formosa. 263 



which is slightly broader and shorter than that of the maxillary and 
tapers posteriorly; nostrils separated, the anterior in a long tube, 
hanging over antero-lateral margin of snout, the posterior minute, in 
front of eye. 

Body covered with rudimentary imbedded scales, linear in form, 
arranged in small groups and placed obliciuely, at right angles to those 
of neighboring groups. 

Color in formalin brownish gray above, mottled with dark brown; 
lower parts whitish; vertical fins grayish, mottled with dark; the 
pectoral dark gray with white outer margin. 

Length of body 375 mm. 

Described from a specimen from Giran. 

Habitat: Very common on the island, often attaining a large size. 
My specimens came from Giran and Jitsugetsutan (Lake Candidius). 



Measurements of Anguilla mauritiana. 

Locality. Giran. 

Total length 370 mm. 

Length of head 62 

Distance of gill-opening from origin of dorsal 55 

Distance of gill-opening from the vent no 

Distance between origin of dorsal and anal 64 

Distance between tip of snout and origin of dorsal. . 113 

Length of the pectoral 17 

Length of snout 13 

Length of upper jaw 20 

Diameter of eye 5 

Interorbital space 12 

Height of body in front of vent 24 

Length of pre-anal part 170 

Length of post-anal part 205 



Giran. 
445 mm. 

67 

60 

125 



14 

23 

6 

13 

27 
190 
25s 



1847- 



1855. 



51. Anguilla japonica Temminck & Schlegel. 
Unagi (Japan); Pehmoa (Formosa). 

Anguilla japonica Temminck & Schlegel, Fauna Japonica, Poiss., p. 258, 
PI. CXin, Fig. 2; Nagasaki.— Bleeker, Verh. Bot. Gen., XXV, p. 51; 
Japan. — Kner, Novara Fisch., IH, 1867, p. 370; Shanghai. — ^Jordan & 
Snyder, Proc. U. S. Nat. Mus., XXIIL 1901, p. 348; Yokohama. — Snyder, 
Proc. U. S. Nat. Mus., XLH, 1912, p. 406; Mororan; Shiogama; Tokyo, 
Misaki. — ^Jordan & Metz, Mem. Carneg. Mus., VI, no. 2, 1913, p. 24, 
Fusan; Suigen, Corea. — Jordan, Sn^'der, & Tanaka, Journ. Coll. Sci., 
Tokyo, XXXIII, 1913, p. 76; Hakodate to Nagasaki. — Ishikawa, Journ. 
Coll. Agric. Tokyo, IV, 1914, p. 417; Japan; Formosa; Corea. 
Murana pekinensis Baksilewsy, Nouv. Mem. Soc. Nat. Mosc, X, p. 246, 
PI. Ill, Fig. 2; Peking. 



264 Annals of the Carnegie Museum. 

1870. Angiiilla boslonensis Gunther {part.). Cat. Fish., VIII, p. 31; Japan; 

Formosa; China. — Gunther, Ann. Mag. Nat. Hist. (7), 1898, I, p. 263; 

Newchang. 
1903. Anguilla remlfera Jordan & Evermann, Proc. U. S. Nat. Mus., XXV, p. 

325; Holoto, Formosa. 
1909. Anguilla sinensis Jordan & Richardson (pari.), Mem. Carneg. Mus., IV, 

no. 4, p. 171; Takao, Formosa. 

Head 7.77 in the total length, 1.25 in the distance between gill- 
opening and origin of the dorsal; 1.94 in the distance between tip of 
snout and origin of the dorsal; distance between origins of dorsal and 
anal shorter than head, contained 1.24 in the latter; distance between 
tip of snout and origin of the dorsal 3.51 in total length; snout 5.09 
in head; maxillary 3.1 1; pectoral 2.8; eye 2.2 in snout, 1.8 in inter- 
orbital space; length of body in front of anus 2.23 in total length; 
pre-anal part 1.72 in postanal part. 

Angle of mouth extending to the posterior margin of the orbit; 
lips thick and fleshy; lower jaw slightly longer than the upper; teeth 
on both jaws and palatines in villiform bands, palatine band slightly 
longer and broader than that of maxillary, tapering posteriorly; 
maxillary and mandibulary bands of teeth divided into two strips by a 
distinct deep groove, outer strip with two rows of teeth, inner strip 
somewhat broader than the outer, with two rows of teeth, in both 
strips number of rows increases anteriorly and their arrangement 
becomes irregular; nostrils separated, the anterior in a long tube, 
situated near antero-lateral extremity of the snout, the pos.terior 
nostril in front of eye, slit-like. 

Pectoral fins longer than broad, marginal end acutely rounded; 
tip of the tail rounded. 

Body covered with rudimentary imbedded scales, linear in form, 
arranged in small groups and placed obliquely at right angles to those 
of neighboring groups; lateral line continuous, running along the middle 
of the sides. 

Color in formalin brownish gray above, lower parts whitish; dorsal 
and caudal fins disky; pectoral and anal fins whitish. 

Length of body 445 mm. 

Described from a specimen from Taihoku, collected by Oshima, 
in February, 1917. 

Habitat: Very common in the fresh waters of Formosa. 

Remarks: After examining forty-seven individuals of the common 
Formosan eel. Dr. Ishikawa expressed his belief that it is only a local 



The Fresh Water Fishes of the Island of Formosa. 265 

variety of Augiiilla japonica, tliough minor differences exist between 
these two forms, as shown in the following table of relative proportions: 

Formosan 
A. "Japonica. Eel. 

Total length loo.oo loo.oo 

Distance from gill-opening to the origin of dorsal. . 18.18 18.41 

Length of head 12.35 12.70 

Distance from gill-opening to vent 27.20 28.45 

Distance from tip of snout to origin of dorsal 30.54 3i-ii 

Distance between origins of dorsal and anal fins. . . 9.76 10.31 

Length of pectoral 4-0i 4-77 

Length of upper jaw 3-24 3-20 

Length of snout 2.16 2.14 

Diameter of eye 1.02 i.ii 

Interorbital space 1.92 2.06 

Height of body in front of anus 521 5.40 

Ratio of pre-anal and post-anal parts i :i-53 i ' i-47 

Number of vertebras iiS-65 ii5-57 

Moreover, he has regarded Anguilla remifera from Hokoto which 
was described by Jordan & Evermann {Proc. U. S. Nat. Miis., XXV, 
1903, p. 325) as being also a variety of A. japonica, because the dis- 
tinctive characters of that species, that is, longer and rather pointed 
pectoral fins (2.17 in head) and the distance from front of dorsal to 
front of anal slightly more than length of head, are the points which 
are to be seen in extreme variations among examples of A. japonica 
{Joiirn. Coll. Agriciil., IV , 1914, p- 426). 

Four specimens of an eel in the Stanford collections (No. 21 181; 
Takao, Formosa) which had been described by Jordan and Richardson 
under the name Anguilla sinensis were examined and compared with 
my specimens of the common Formosan eel. It appears that in the 
former the distance from gill-opening to origin of dorsal, distance 
from gill-opening to vent, distance from tip of snout to origin of dorsal, 
diameter of eye, and height of body in front of anus (average length) 
are somewhat greater, while the length of head, distance between 
origins of dorsal and anal, length of pectoral, length of upper jaw, 
length of snout, interorbital space are smaller. As shown in the fol- 
lowing tables (III, V), however, these characters are variable and 
have no specific value. 




266 



Annals of the Carnegie Museum. 



I. The Average Length of the Parts of Body of Formosan Eel. 

A. 
A . japonica. sinensis. 

Total length loo.oo loo.oo 

Distance from gill-opening to origin of dorsal 17.01 18.43 

Length of head 12.42 12.12 

Distance from gill-opening to vent 26.55 ■' 26.72 

Distance from tip of snout to origin of dorsal 28.95 29.40 



Distance between origins of dorsal and anal 10.72 , 



Length of pectoral 

Length of upper jaw 

Length of snout 

Diameter of eye 

Interorbital space 

Height of body in front of anus 

Ratio of pre-anal and postanal parts. 



4-I5- 
4.04. 
2.51. 
1. 13. 
2.19 . 
4-51 ■ 
1:58. 



9.64 
4.09 
3-56 
2.26 
1.38 
1-73 
4-73 
1 : 60 



n. Measurements of Anguilla japonica. 
Actual Length in Mm. 



Locality. 



Taihoku. Taihoku 



Total length ! 310 mm. 357 mm. 

Head | 41 

Gill-opening to dorsal. . j 57 
Gill-opening to vent. . . 87 

Dorsal to anal 33 

Upper jaw 12 

Pectoral 13 

Snout 8 

Eye 3 

Interorbital space .... 7 

Height of body 18 

Snout to vent I 124 

Vent to tip of caudal. . . J 186 
Snout to dorsal 95 



46 
61 
98 
40 

15 
16 

9 
5 
8 

19 

142 

215 
107 



Taihoku. 



445 mm. 

56 

70 
no 

45 
18 
20 
II 

5 

9 
22 

163 

282 
124 



Taihoku. 



355 mm. 
47 
70 
96 

37 
12 

14 
8 

4 

7 

17 

141 

214 

no 



Taihoku. 



Giran. 



355 mm- 
46 
60 
95 
39 
14 
15 
10 

4 

9 

19 

138 

217 

105 



182 mm. 
22 
27 
46 
20 

7 

6.5 

4-5 

2 

4 

9 

68 

114 

50 



in. Length in Percent of the Total Length. 



Locality. 

Total length 

Head 

Gill-opening to dorsal 
Gill-opening to vent . 

Dorsal to anal 

Upper jaw 

Pectoral 

Snout 

Eye 

Interorbital space . . . 

Height of bod J' 

Snout to vent 

Vent to tip of caudal. 
Snout to dorsal 



Taihoku. Taihoku. Taihoku. Taihoku. Taihoku. Giran. 



100.00 


100.00 


13.22 


12.85 


18.38 


17.08 


28.06 


27-45 


10.64 


11.20 


3-87 


4.20 


4.19 


4-47 


2.58 


2.52 


0.96 


1.40 


2.25 


2.24 


5-8o 


5-32 


40.00 


39-80 


60.00 


60.22 


30.64 


30.00 



100.00 


100.00 


12.58 


13-23 


15-73 


19.71 


24.71 


27.04 


1 0.1 1 


10.42 


4-04 


3.38 


4.49 


3-97 


2.47 


2.25 


I. II 


1. 12 


2.02 


1.97 


4-94 


4-78 


36.62 


39-71 


63-37 


60.28 


27.86 


30.98 



100.00 

12.95 

16.90 
26.76 
10.98 

3-94 

4.22 

2.81 

1. 12 

2-53 

5-35 

38-87 

61.12 

26.76 



The Fresh Water Fishes of the Island of Formosa. 267 



IV. Measurements of Anguilla sinensis, Jordan & Richardson. 
(No. 21181, Stanford Collection, Takao.) 
Actual Length in Mm. 



Locality, 

Total length 

Head 

Gill-openings to dorsal 
Gill-openings to vent. . 

Dorsal to anal 

Upper jaw 

Pectoral 

Snout 

Eye 

Interorbital space 

Height of body 

Snout to vent 

Vent to tip of caudal. . 
Snout to dorsal 



Taihoku. 


Taihoku. 


405 mm. 


425 mm. 


SO " 


S4 ' 




80 " 


74 ' 




120 " 


no ' 




43 " 


41 ' 




IS " 


IS ' 




18 " 


15 ' 




ID " 


10 ' 




4 " 


4-5 ' 




7 " 


7 ' 




23 " 


19 ' 




169 " 


160 ' 




236 " • 


265 ' 




123 " 


124 ' 





Giran. 



375 mm. 

44 " 

68 " 

94 " 

33 " 

12 " 

15 " 
8.5" 
3-5 " 
7 " 

16 " 
139 " 
236 " 
no " 



435 mm. 

51 

79 

114 

41 

13 

19 

9 

4 

7-, 

20 

165 

270 

125 



V. Length in Percent of the Total Length. 



Locality 

Total length 

Head 

Gill-opening to dorsal . 
GiII-op)ening to vent . . 

Dorsal to anal 

Upper jaw 

Pectoral 

Snout 

Eye 

Interorbital space .... 

Height of body 

Snout to vent 

Vent to tip of caudal. . 
Snout to dorsal 



Taihoku. 



Taihoku. 



Taihoku. 



Giran. 



100.00 
12.34 

I9-7S 

29.87 

10.64 

3-70 

4.44 

2.44 

0.98 

1.72 

5-68 

41.72 

58.27 

30.37 



100.00 

12.70 

17.70 

25.88 

9.64 

3-S2 

352 

2.29 

1. 00 

1.64 

4-47 

36.64 

62.35 

29.17 



100.00 

11.73 

18.13 

25.06 

8.99 

3.20 

4.06 

2.26 

0.93 

1.85 

4.27 

37.06 

62.93 

29-33 



100.00 

11.72 

18.16 

26.20 

9.42 

2.98 

4-36 

2.06 

0.91 

1.72 

4-59 

37-93 

62.06 

28.73 



52. Anguilla sinensis McClelland. 

Anguilla sinensis McClelland, Calc. Journ., IV, p. 406, Tab. 25, Fig. 2; 
China. — Jordan & Everiviann, Proc. U. S. Nat. Mus., XXV, 1903, p. 325; 
Taihoku, Formosa. — Jordan & Richardson, Mem. Carneg. Mus., IV, no. 4, 
1909, p. 171; Taihoku. 

Anguilla latiroslris Gunther, Cat. Fish., VIII, p. 32; China {part). 

Anguilla sinensis (?) Ishikawa, Journ. Coll. Agric. Tokyo, p. 428; Tokj^o. 

A small specimen, Xo. 6447, from Taihoku. It agrees with Giin- 
ther's account of A. latiroslris, but that species was originally described 
from Nice. The long head, greater than the distance from front of 
dorsal to front of anal, is characteristic of this species. (Jordan & 
Evermann.) 



1844. 



1870 
1914 



268 Annals of the Carnegie Museum. 

Habitat: Taihoku (Jordan & Evermann). 

Remarks: This specimen may not be different from the common 
eel, Anguilla japonica. As I have no specimens to examine, nothing^ 
more can be said of it at present. 

Family MUGILID.E. 
Artificial Key to the Formosan Genera of Mugilid^. 

A. An adipose eye-lid well developed, covering at least a third of the iris posteriorly. 

Mugil. 

B. Adipose eye-lid not developed Liza. 

Genus Mugil (Artedi) Linnaeus. 

1758. Mugil (Aptedi) Linn.^us, Syst. Nat., Ed. X, p. 316. (Type Mugil 
cephalus Linnaeus.) 

Body more or less oblong and compressed, covered with cycloid 
scales of moderate size; no lateral line. Mouth more or less trans- 
verse; anterior margin of the mandible sharp, sometimes ciliated. 
No true teeth in the jaws. Gill-openings wide; gills four. Eyes 
lateral, with adipose eyelids. Two dorsal fins, the first consisting of 
four stiff spines; anal slightly longer than the second dorsal; ventrals 
abdominal, with one spine and five rays. Branchiostegals from four 
to six; pseudobranchiaj present. 

Distribution: Migratory fishes of all the temperate and tropical 

regions. 

Synopsis of the Formosan Species. 

A. Median dorsal line not carinate. 

a. Mandibular angle obtuse; cleft of mouth contained two times in the dis- 
tance between the angles of mouth cephalus. 

aa. Mandibular angle a right angle; cleft of mouth contained less than two 
times in the distance between the angles of mouth oeur. 

B. Median dorsal line carinate m front and back of the spinous dorsal. 

a. Mandibulary angle obtuse; cleft of mouth contained less than two times in 
the distance between the angles of mouth carinalus. 

53. Mugil cephalus Linnaeus. 
Bora (Japan); Oahii (Formosa). 

1758. Mugil cephalus Linn.^us, Syst. Nat., Ed. X, p. 316; Europe. — Cuv. & 
Val., Hist. Nat. Poiss., XI, 1830, p. 307. — Gunther, Cat. Fish., Ill, 1861, 
p. 417; Mediterranean; Coast of Madeira; Nile; fresh-water lakes of Tunis; 
West coast of Africa. — Jordan & Starkes, Proc. U. S. Nat. Mus., XXXI, 
1906, p. 516; Port Arthur. — Jordan & Seale, Proc. U. S. Nat. Mus., 
XXIX, 1906, p. 521; Hongkong; Shanghai. — Jordan & Richardson, 



i 



The Fresh W'ateu Fishes of the Island of Formosa. 269 

Bull. IT. S. Bur. Fish., XXVI 1, 1908, p. 244; Calayan, p. i. — Snyder, 
Proc. U. S. Nat. Mus., XLII, 1912, p. 416; Misaki; Tokyo.— Proc. U. S 
Nat. Mus., XLII, 1912, p. 459; Okinawa. — Jordan, Snyder, & Tanaka, 
Journ. Coll. Sci. Tokyo, XXXIII, 1913, p. 113; Hawaii; Tahiti; Atlantic 
coast of the United States; Panama; Southern California; Red Sea; Mediter- 
ranean; New Guinea; Nukahiva; Solomon Island. — Jordan & Metz, Mem. 
Carneg. Mus., VI, no. 2, 1913, p. 26; Fusan, Corea. — Jordan & Thompson, 
Mem. Mem. Carneg. Mus., VI, no. 4, 1914, p. 239; Matsushima; Osaka. 
1855. Mugil soiuy Basilevvsky, Ichthyol. China, p. 226, PI. IV, Fig. 3; China. 

Head 3.86 in length; depth 4.42; D. IV, i, 8; A. Ill, 8; P. 17; V. I, 5; 
width of head 1.55 in its length; eye 4.23 in head; interorbital space 2; 
snout 3.5; forty-two scales in a lateral series, fifteen scales in an oblique 
series between origin of dorsal and middle of belly, thirteen scales 
between origins of dorsal and ventral. 

Body rather robust, elongate, somewhat compressed, dorsal profile 
nearly straight, ventral profile broadly rounded; head rather small, 
broad, its top flattened; snout short and obtuse, broadly rounded 
anteriorly, interorbital space very broad, rather flat; mouth subin- 
ferior, slightly oblique, its angle reaching a vertical through posterior 
nostril; cleft of mouth half as deep as broad (between the angles of 
mouth); lips thin; the angle between two mandibulary bones obtuse; 
lower jaw shorter than the upper, outer edge rather sharp, with an 
obtuse short median keel at the anterior part which fits into the cor- 
responding concavity on the roof of upper jaw; teeth along the outer 
edges of both jaws minute, scarcely visible without lens; eyes hidden 
anteriorly and posteriorly by a broad adipose membrane; nostrils 
separated, anterior nostril in a very short tube, posterior nostril slit- 
like, in front of eye above. 

Dorsal fins well separated, origin of the spinous dorsal midway 
between tip of snout and base of caudal, with a pointed scaly flap at 
the base, length of the anterior spine 2.2 in head; soft dorsal inserted 
behind the origin of anal, anterior ray longest; pectoral reaching 
beyond the origin of ventral, not reaching the spinous dorsal, its 
base above the middle of body; ventral a little nearer the root of 
pectoral than the spinous dorsal; anal fin opposite the soft dorsal, 
inserted in advance of the origin of the latter; caudal fin bifurcate; 
depth of the caudal peduncle 2.66 in the length of head. 

Head and body covered with large cycloid scales, those on top of 
head slightly enlarged; soft dorsal, anal, and pectoral with \ery few 
scales; base of caudal covered with large scales. 



270 



Annals of the Carnegie Museum. 



Color in alcohol grayish above, lower parts silvery; sides with dark 
longitudinal stripes along the rows of scales; pectorals, dorsals, and 
caudal fin dusky; ventrals and anal whitish. 

Length of body 255 mm. 

Described from a specimen from Daitotei Fish Market, Taihoku, 
collected by Oshima in March, 1917. 

Habitat: Taihoku (?) . A very common species in Formosa. 

Remarks: All the characters of the present species agree ciuite well 
with those of M. cephaliis from Italy in the Stanford University Col- 
lections. 

Measurements of Mugil cephaliis. 



Locality. 


T3 
a 

V 


.a 

D. 

Q 


Q 


< 


Oh' 


> 


Width 
of Head. 

Interor- 
bital. 


3 


c 




^ 






Taihoku 

Taihoku 

Venice, Italy. 
(No. 1479. S. U.).. 
Naples, Italy. 
(No. 1469, S. U.).. 


3-86 
3-95 

3-94 

3-69 


4.42 
4-79 

4.14 

4.09 


IV, I, 8 
IV, I, 8 

IV, I, 8 

IV, I, 8 


111,8 
111,8 

111,8 

111,8 


17 
16 

16 

16 


1.5 
1.5 

1.5 

1.5 


1-55 2 
1.502.38 

1.45 2.80 

1 
1.502.52 


3-50 
3-II 

3-20 

3.65 


4-23 
4-58 

4-73 

4-85 


2 '42-15 
2 40-15 

2 42-15 

2 41-15 


255 
284 

260 

220 



1775- 



54. Mugil oeur Forskal. 

Mugil ocitr Forskal, p. XIV, No. 109. — Rutter, Proc. Acad. Nat. Sc. 
Philad., 1897, p. 70; Swatow, China. — ^Jordan & Snyder, Proc. U. S. Nat. 
Mus., XXIII, 1901, p. 744; Yokohama. — Jordan & Evermann, Proc. 
U. S. Nat. Mus., XXV, 1903, p. 332; Taihoku, Formosa. — Jordan & Rich- 
ardson, Mem. Carneg. Mus., IV, No. 4, 1909, p. 176; Giran; Keelung; 
Taihoku, Formosa. 

Mugil cephalotus Cuv. & Val., Hist. Nat. Poiss., XI, p. no; India. — 
Gunther, Cat. Fish., Ill, 1861, p. 419; Red Sea; Coast of Pondicherry, 
Chinese and Japanese Seas. — -Kner, Novara Fisch., II, 1865, p. 224; Manila. 

Mugil japonicus Schlegel, Fauna Japonica, Poiss., p. 134, PI. 72, Fig. i. — 
Richardson, Ichthyol. China, 1846, p. 247; China. — Bleeker, Verh. 
Batav. Genootsch., XXV, 1853, p. 41 

Mugil macrolepidotus Richardson, Ichthyol. China, p. 249; China. 

Mugil cephalus Jordan & Seale (not of Linnaeus), Proc. Davenport Acad. 
Sc, X, p. 4; Hongkong. 

Head 3.6 in length; depth 4; D. IV, i. 8; A. Ill, 8; P. 16; V. I, 5; 
width of head 1.58 in its length; eye 4 in head; interorbital space 2.64; 
snout 3.75; thirty-nine scales in a lateral series, fifteen scales between 
origin of the spinous dorsal and ventral. 

Body oblong, compressed posteriorly, dorsal and ^'entral profiles 
ecjually arched; head rather small, its top very slightly convex; snout 



1836. 



1846. 



1846. 
1905. 



i 



The Fresh Water Fishes of the Island ok Formosa. 271 

short, moderately broad, truncate in front, interorbital space rather 
flat; mouth subinferior, slightly oblique, its angle reaching'a vertical 
through anterior margin of orbit, cleft of mouth 1.66 in the transverse 
distance between angles of mouth; lips rather thin, upper lip much 
thicker than the lower; the angle between two mandibulary bones a 
right angle; lower jaw shorter than the upper, its outer edge rather 
sharp, with a short median keel at the tip which fits into the corre- 
sponding concavity on the roof of the upper jaw; teeth on the upper jaw 
minute, scarcely visible without a lens, closely set; eyes hidden 
anteriorly and posteriorly by a broad adipose membrane; nostrils 
separated, anterior nostril in a very short tube, posterior nostril slit- 
like, in front of eye above. 

Dorsal fins well separated; origin of the spinous dorsal midway 
between tip of snout and base of caudal, with pointed scaly flaps on 
both sides of the base, length of the anterior spine twice in head; soft 
dorsal inserted behind the origin of anal, anterior ray^ longest; pectoral 
not reaching spinous dorsal, its base above the middle of body; 
ventral inserted nearer the origin of the spinous dorsal than that of 
the pectoral, with a scaly, pointed flap near the base; anal fin opposite 
the soft dorsal, inserted in front of the origin of the latter, caudal fin 
forked; depth of caudal peduncle 2.66 in head. 

Head and body covered with large cycloid scales; all the fins except 
the spinous dorsal with very few scales; base of the caudal covered 
with large scales. 

Color in alcohol dark gray above, belly and lower half of the sides 
silver}'; sides with dark longitudinal stripes along the rows of scales; 
pectorals, dorsals, and caudal fin dusky, other fins whitish; a black 
spot near the base of pectoral. 

Length of body 132 mm. 

Described from a specimen from Inzanpo, Giran, collected by T. 
Aoki in August, 1917. 

Habitat: Inzanpo and Ritakukan, Giran. Giran; Keeking; Taihoku. 
(Jordan and Evermann.) 

Remarks: The present species is very closely related to Mugil 
cephalus Linnaeus, differing mainly in the angle between the two 
mandibulary bones. In the latter it is always obtuse instead of being 
a right angle, and the cleft of the mouth is contained twice in the 
distance between the angles of mouth. 



272 



Annals of the Carnegie Museum. 



Measurements ok Miigil oeiir. 



Locality. 


■a 

V 


.a 


Q 


<i 


(li 


> 


. 


It 


3 





■a 3 
V 


a 


■2 




X 


Q 












■^^ 


cI5 


'^■ 


W 


uS 


m 


J^ 


Inzanpo . . .' 


3.6 
3-5 


4 

4 


IV, I, 8 


I'll. 8 


16 




5 
5 


1.58 
1.58 


2.64 
2.50 


3-75 
3-75 


4 
4.66 


1.66 


39-15 
39-15 


132 
133 


Ritakukan 


IV, 1,8 


111,8 


16 


I, 


1.50 


China 


3-91 


4.37 


IV, 1,8 


111,8 


16 




5 


1.50 


2.58 


3.92 


1.77 


1.77 


40-15 


226 


(No. 1606, S. U.) 
























Hongkong 






























(No. 9884, S. U.) . 


3.89 


4-58 


IV, 1,8 


111,8 


16 


I, 


5 1.61 


2.47 


3-70 


3-70 


1-57 


38-iS 


166 


Hongkong 






























(No. 9884, S. U.) . 


3-87 


4-38 


IV, I, 8 


111,8 


16 


I, 


5 


I-S7 


2.62 


3-55 


3-55 


1.50 


40-15 


152 


Hongkong 






























(No. 9884. S. U.) . 


3-42 


4-50 


IV, I, 8 


111,8 


16 


I, 


5 


1.61 


2.75 


3-66 


3.66 


1-57 


41-16 


150 


Hilo, Hawaii. 






























(No. 7852, S. U.).. 


4. 1 1 


4-25 


IV, 1,8 


111,8 


16 


I, 


5 


1.48 


2.14 


3-75 


3-75 


1-75 


38-15 


236 


Hilo, Hawaii. 




























(No. 7852, S.U.).. 


396 


4.10 


IV, I, 8' III, 8 16 


I, 


5 1-47 


2.18 


3-64 


4 


1.89 


40-15 


260 


Hilo, Hawaii. 
























(No. 7852, S. U.) . 


3.85 


4.12 


IV, I, 8 III, 8 17 


I. 


5 1-24 


2.47 


3-6o 


4 


1. 71 


41-15 


176 



55. Mugil carinatus (Khrenberg) Cuv. & \'al. 

1830. Mugil carinaliis (Ehrenberg) Cuv. & Val., Hist. Nat. Poiss., XI, p. 148; 
Red Sea.— Day, Fish. Brit. India. 1888, Suppl., p. 800; Sea of India. 

Head 4 in length; depth 4.19; D. lY, 2, 7; A. Til, 9; P. 15; \'. I, 5 
width of head 1.58 in its lengtli; e>'e 3.7 in head; interorbital space 
2.79; snout 4; thirty-nine scales in a lateral series, thirteen scales in a 
transverse series between origin of the dorsal and the middle of belly, 
ele\en scales between origins of the dorsal and ventral; i:iectoral 1. 5 
in head; ventral 1.73. 

Body elongate, compressed i)osteriorly, curvature of the dorsal 
profile weaker than that of the ventral, median dorsal line keeled in 
front and behind the spinous dorsal; head ratlier small, top more or 
less convex; snout short, truncate in front; mouth subinferior, its 
angle reaching a vertical through the posterior nostril, depth of the 
mouth 2.66 in its width; upper li]) rather thick; lower jaw shorter than 
the upi)er; the angle between the mandibulary bones obtuse; upper 
jaw with a series of minute uniform teeth along the outer border, 
lower jaw with a series of lleshy tubercles along its inner margin; 
pre-orl)ital not scaled, wa\\', lower margin finely serrated; extremity 
of maxillary visible; eyes moderate, anterior, with narrow adipose 
eyelids; posterior nostril slit-like, nearer to tlie anterior nostril than 
the orbit. 

Dorsal lins well si'parated; sjjinous dorsal inserted much nearer tip 



The Fresh Water 1'ishes of the Island ok Formosa. 273 



of snout than base of caudal, higher than the soft dorsal, height of the 
first spine equal to the width of head; tenth and twenty-eighth scales 
of lateral series below the origins of the two dorsals, soft dorsal in- 
serted behind the origin of the anal; pectoral fin extending beyond the 
root of the ventral, tip reaching to tenth scale of lateral series; vcntrals 
inserted midway between pectorals and spinous dorsal; anal fin oppo- 
site the soft dorsal, anterior ray longest; caudal fin forked; depth of 
caudal peduncle 2.22 in the length of head. 

Bod}' covered with large cycloid scales; head entirely scaled; soft 
dorsal, anal, pectorals, \-entrals, and base of the caudal covered with 
small scales; spinous dorsal and ventrals with pointed scaly flaps. 

Color in formalin grayish above, lower parts silvery; pectorals, 
dorsals and caudal dusky, the rest of the fins whitish. 

Length of body 185 mm. 

Described from a specimen from Shinio-Tamusui River, collected 
by T. Aoki in December, 191 6. 

Habitat: Shimo-Tamusui River, Taihoku (collected in the fish- 
market). 

Measurements of Mugil carinatus. 



Locality. 


i 


0. 


Q 


•si 


cJ 


> 


■51 


h" 


3 




>> 


V 


J3 . 

5se 




M 


Q 










^-K 


£3 


fii 




in 




Shimo-Tamusui R. . . 


4 


4.19 


IV, 2, 7 III, 9 


15 I. 5 


1.58 


2.78 


4 


3-7I39-I3 185 


Taihoku 


4-13 


4-38 


IV, 2, vim, 9:16 1,5 


i-.S? 


2.94 


4 


4 


39-13I220 


Taihoku 


4.26 


4-85 


IV, 2, 7,111, 9 16 I, 5 


1.74 


2.61 


3-5i4 

1 


40-131277 



Genus Liza Jordan & Swain. 

1884. Liza Jordan & Swain, Proc. U. S. Nat. Mus., VII, p. 261. (Type Mugil 
capilo Cuvier.) 

Body robust, more or less oblong and compressed, head and body 
covered with large cycloid scales; lateral line none. Mouth subin- 
ferior, more or less transverse; anterior margin of the mandible thin 
and sharp. Gill-openings wide. Eyes lateral and anterior, adipose 
eyelid obsolete. Two dorsal fins, the first consisting of four stiff 
spines; anal opposite the soft dorsal, slightly longer than the latter; 
ventrals abdominal, with one spine and five rays. 

Distribution: British and Scandinavian coast; Canary Islands; 
Mediterranean; Nile; Freshwater lakes of Tunis; From Red Sea 
through Indian Ocean and Archipelago to the coasts of Australia and 
19 — DEC. 19, 1919. 



274 Annals of the Carnegie Museum. 

to Polynesia; India; Ceylon; Philippine Islands; Indo-China; China; 
Formosa; Japan; Riukiu Islands. 

56. Liza troscheli (Bleeker). 

1858. Mugil troscheli Bleeker, Nat. Tijdschr. Ned. Ind., XVI, p. 277. — Act. 
Soc. Sc. Indo-Neerl., VIII, i860, p. 80; Sumatra. — Gunther, Cat. Fish., 
Ill, 1861, p. 448; Coast of Java, Borneo, and Ceylon. — Day, Fish. Brit. 
India, 1878, p. 358; Indian Sea to Malay Archipelago. — Rutter, Proc. 
Acad. Nat. Sc. Philad., 1897, p. 70; Swatow. 

1903. Liza troscheli Jordan & Evermann, Proc. U. S. Nat. Mus., XXV, p. 332; 
Hokoto, Formosa. — Jordan & Seale, Bull. U. S. Bur. Fish., XXVI, 1906, 
p. 11; Cavite, P. I. — Jordan & Richardson, Bull. U. S. Bur. Fish., XXVII 
1908, p. 244; Iloilo. — ^Smith & Seale, Proc. Biol. Soc. Wash., XIX, 1906, 
p. 76; Mindanao. — Seale & Bean, Proc. U. S. Nat. Mus., XXXIII, 1907, 
p. 240; Zamboanga. — Jordan & Richardson, Mem. Carneg. Mus., IV, 
No. 4, 1909, p. 176; Takao; Hokoto. — Snyder, Proc. U. S. Nat. Mus., 
XXXII, 1912, p. 495; Okinawa. — Jordan & St.a.rks, Ann. Carneg. Mus., 
XI, Nos. 3 and 4, 1917, p. 439; Ceylon. 

Head 4.3 in length; depth 4; D. lY, 2, 7; A. Ill, 9; P. 14; V. I, 5; 
width of head 1.46 in its length; eye 4 in head; interorbital space 2.28; 
snout 3.64; pectoral 1.31; ventral 1.42; thirty-one scales in lateral 
series, eleven scales in an oblique series between origin of dorsal and 
the middle of belly, nine scales between origins of dorsal and ventral. 

Body robust, compressed posteriorly, the ventral profile much more 
curved than the dorsal, deepest in front of the vent; head rather small, 
top slightly conve.x; snout nearly as long as the diameter of eye, 
obtusely rounded anteriorly; mouth subinferior, its cleft one-third as 
deep as the distance between the angles of mouth; upper lip thick, 
no lower lip; lower jaw slightly shorter than the upper, angle between 
two mandibulary bones obtuse, outer edge of mandible thin and sharp; 
upper jaw with no teeth, lower jaw with a series of minute fleshy 
tubercles along the inner margin; pre-orbital edge more or less wavy, 
lower part indistinctly denticulated; eyes moderate, anterior, adipose 
eyelid not well developed; nostrils separated, the anterior in a very 
short tube, the posterior slit-like. 

Dorsal fins well separated; origin of the spinous dorsal nearer to 
base of caudal than tip of snout, as high as the soft dorsal, length of 
the first spine shorter than the width of head; soft dorsal inserted 
behind the origin of anal, eleventh and twentieth scales of lateral 
series below the origins of the dorsals; pectorals a little above the middle 
of body, reaching eighth scale of lateral series; the \entral inserted mid- 



The I<^kksii Water Fishes of the Island of Formosa. 275 

way beUveen origin of pectoral and that of spinous dorsal; the anal 
inserted in advance of the origin of soft dorsal; caudal tin emarginate; 
depth of caudal peduncle twice in the length of head. 

Head and bod>- covered with large cycloid scales; soft dorsal, anal, 
and base of ventral and caudal covered with small scales; spinous dor- 
sal and ventral with pointed scaly flaps. 

Color in formalin grayish above, lower parts silvery; sides with 
longitudinal stripes along the rows of scales; pectorals, dorsals, and 
caudal dusky, the rest of the tins whitish. 

Length of body 215 mm. 

Described from a specimen from Sobun River near Tabani, col- 
lected by T. Aoki in December, 1916. 

Habitat: Sobun River near Tabani (two specimens). Hokoto 
(Jordan and Evermann); Takao (Jordan and Richardson). 



Measurements of Liza troscheli. 



Locality. 



■2 rt 



em E 



Sobun River. 
Sobun River. 



4-3 
4-3 



4 
3-58 



IV, 2, 7 III, 9 14 I, 5 
IV, I, 8 III, 9I15I1. 5 



1.46 2.28 3.64 
1.42 2.13 3.40 



4 31-11 
4.1731-11 



215 
272 



Family LABYRINTHICI. 

Artificial Key to the Formosan gener.a. 

a. Teeth fixed in the jaws; none on the palate; ventrals well-developed; 13-20 
dorsal or anal spines; dorsal and anal rays not filamentous; caudal fin rounded. 

Poly acanthus. 

aa. Teeth fixed in the jaws; none on the palate; ventrals well-developed; 13-18 

dorsal or anal spines; middle rays of the dorsal and anal filamentous; caudal 

fin forked or ending in a bundle of long filaments Macropodus. 

Genus PoLYACANTHUS (Kuhl) Cuvier. 
1829. Polyacanlhus (Kuhl) Cuvier, Regne Animal, Ed. II, Vol. II, p. 227. 
(Type Chcetodon chinensis Bloch =Labrus operculalus Gmelin, as restricted by 
Cuvier & Valenciennes.) 

Body compressed, oblong; operculum without spine or serrature; 
cleft of mouth small, more or less oblique, not extending beyond a ver- 
tical from orbit, and little protractile. Small fixed teeth in the jaws, 
none on the palate. Dorsal and anal spines numerous (13-20); 
ventral fins composed of one spine and five well-developed rays; the 



276 Annals of the Carnegie Museum. 

soft dorsal and anal, the caudal and ventral more or less elongate in 
the mature specimens, but not filamentous; caudal rounded. Lateral 
line interrupted or absent. Branchial arches with toothed tubercles. 
Air-bladder simple; pseudobranchia^ none (glandular). Pyloric appen- 
dages in small number; intestine with many circumvolutions. 

Distribution: Java; Sumatra; Borneo; Ceylon; Malabar; China; 
Formosa. 

57. Polyacanthus operculatus (Linnaeus). 

Taiwan-kingyo (Formosa). 

1735. Lahrus opercularis Linn.«us, Amoen. Acad., IV, p. 428 (not binomial). 

1789. Labrus operculatus Gmelin, Syst. Nat., p. 1286; Asia (after Linnaeus). 

1785. Chcetodon chinensis Block, Ichthyologie, VII, p. 3, Taf. 218, Fig. i; China. 

1831. Polyacanthus chinensis Cuv. & Val., Hist. Nat. Poiss., VII, p. 357. — Rich- 
ardson, Ichthyol. China, 1846, p. 250; Canton. 

1842. Macropodus ocellalus Cantor, Ann. Mag. Nat. Hist., IX, p. 484; Chusan. 

1846. Polyacanthus} opercularis Richardson, Ichthyol. China, p. 250; China. 

1846. Polyacanthus ? paludosus Richardson, Ichthyol. China, p. 250; Canton. 

1861. Polyacajilhus opercularis Gunther, Cat. Fish, III, p. 379; Chusan; Hong- 
kong; China. — Ann. Mag. Nat. Hist., Sept., 1873, p. 243; Shanghai. — 
Pkters, Monatpb. Ak. Berlin, 1880, p. 923. — Abbott, Proc. U. S. Nat. Mus., 
XXIII, 1901, p. 490; Tien-tsin. 

Head 3 in length; depth 2.75; D. 14, 6; A. 20, 13; P. 10; V. 2, 4; 
width of head 1.66 in its length; eye 3.66 in head; interorbital space 
2.8; snout 4; twenty-eight scales in a lateral scries, six scales between 
orbit and posterior margin of operculum, twelve scales in an oblique 
series between origin of dorsal and the root of ventral. 

Body oblong, compressed, dorsal and ventral profiles equally 
arched; head moderate, triangular, entirely covered with scales; 
operculum with no spine or serrature; snout pointed anteriorly, 
interorbital space more or less convex; mouth small, terminal and 
subvertical, its angle not reaching the orbit; lower jaw slightly pro- 
truding; both jaws with small, fixed, villiform teeth; eyes rather large, 
anterior; nostrils separated, the anterior approximating the upper lip, 
the posterior in contact with eye. 

Origin of the dorsal much nearer tip of snout than base of caudal, 
inserted behind that of anal, soft rays much longer than the spinous, 
the anterior longest, reaching middle of caudal; pectoral fins thoracic, 
the tips reaching beyond the anterior third of anal; anal fin very long, 
its base longer than that of dorsal, soft rays elongate, extending 
beyond the middle of caudal; caudal fin oblong, rounded at the tip; 



The Fresh Water F'ishes of the Island of Formosa. 277 



caudal peduncle very short, its depth less than twice in the length of 
head. 

Scales large, ctenoid; lateral line indistinct, \'isible only on the 
anterior nine scales. 

Color greenish gray, with ten brown cross-bars on the sides, one of 
which is on the nape; a dark brown, round spot on the extremity of 
gill-cover, a somewhat paler short streak between the orbit and that 
spot; caudal fin with two brown cross-bars near the base; dorsal and 
caudal fins speckled with brown; base of the anal brownish; pectoral 
and ventral fins whitish. 

Length of body 58 mm. 

Described from a specimen from Taihoku, collected by Oshima in 
February, 1917. 

Habitat: Abundant in the ditches and stagnant pools throughout 
the Island. My specimens came from Taihoku; Wodensho, Taichu. 

Measurements of Poly acanthus operculalus. 



Locality. 



T3 

u 


g. 


d 


<i 


a," 


> 


"0 . 




3 


c 


3 


2.75 


14, 6 


20.13 


10 


2,4 


1.66 


2.80 


4 


2.83 


2.83 


14, 6 


19. 13 


10 


2,4 


1. 71 


3 


4 . 


2.91 


2.91 


14, 6 


19.13 


10 


2,4 


1.83 


2.75 


4 


2.90 


2.90 


13.7 


19. 13 


10 


2,4 


1.83 


2.75 


3-50 


2.77 


2-77 


13.7 


18. 13 


10 


2,4 


'2 


3 


3-33 



w 






Taihoku 
Taihoku 
Taihoku 
Taihoku 
Taihoku 



3.66 28-12 
3.33J28-12 
3.50128-12 
3 28-12 
3.66'28-i2 



Genus Macropodus Lacepede. 

1802. Macropodus Lacepede, Hist. Nat. Poiss., Ill, p. 416. (Type Macropodus 

viridiaiiralus Lacepede). 
1861. Macropus GiJnther, Cat. Fish., Ill, p. 381. (Type Macropodus viridiaur- 

atus Lacepede.) 

Body compressed, oblong; operculum without spine or serrature; 
cleft of mouth rather small, not extending beyond a vertical from the 
orbit. Small fixed teeth in the jaws, none on the palate. Dorsal 
and anal spines numerous (thirteen to eighteen) ; ventral fins composed 
of one spine and five rays, which are w^ell-developed; soft dorsal, anal 
and caudal with long filamentous rays. Caudal fin forked or ending 
in a bundle of long filaments. Lateral line absent. Branchial arches 
with toothed tubercles. Pyloric appendages in small number; 
intestine of moderate length. 

Distribution: Fresh waters of China, Cochin-China, and Botel 
Tobago. 



278 Annals of the Carnegie Museum. 

58. Macropodus filamentosus sp. nov. (Plate LII, Fig. 2). 

Head 3.2 in length; depth 3; D. 13, 8; A. 18, 15; P. 10; V. I, 5; 
width of head 1.66 in its length; eye 3.5 in head, interorbital space 
2.75; snout 3.5; twenty-eight scales in a lateral series, twelve scales in 
an oblique series between origin of dorsal and the root of the ventral, 
six scales between orbit and the posterior end of operculum. 

Body oblong, compressed, tapering posteriorly, dorsal and ventral 
profiles equally arched; tail slightly turned downward (probably 
distorted) ; head moderate, triangular, entirely covered with scales, 
top convex; snout short, pointed anteriorly, interorbital space rather 
broad, slightly convex; operculum with no spine; minute denticulations 
at the pre-orbital and the angle of the pre-operculum; mouth anterior, 
oblique, its angle not reaching the orbit; both jaws with a band of 
fixed, villiform teeth, the lower jaw slightly protruding; nostrils separ- 
ated, the anterior in a short tube, situated nearer the lip than orbit, 
the posterior nearly in contact with eye. 

Dorsal fin elongate, inserted midway between tip of snout and base 
of caudal, soft rays much longer than the spines, fifth ray filamentous, 
reaching the middle of caudal; ventral with one spine and five rays, 
first ray filamentous, reaching anterior two-thirds of the base of anal; 
anal fin very long, its origin in advance of that of dorsal, posterior end 
of the base nearly in contact with the base of caudal, soft rays longer 
than the spines, middle rays longest, seventh to ninth rays filamen- 
tous, reaching beyond the end of dorsal filament; caudal fin gradually 
tapering posteriorly, middle rays elongated, forming a bundle of 
delicate filaments, not forked. 

Scales moderate, ctenoid; with no lateral line; head provided with 
many mucous pores, four along the posterior margin of the upper lip, 
one on each side behind the posterior nostril, one in contact with the 
hind border of the orbit, three on the posterior part of pre-operculuni; 
bases of the dorsal and anal covered with scales. 

Color in alcohol pale grayish white, with about eleven brownish 
cross-bars on the sides and one of the same color on the base of the 
caudal; a dark brown round spot at the extremity of the operculum; 
postorbital space with traces of two brown streaks; membrane of the 
soft dorsal and upper half of the caudal speckled with l)lack; base of 
the anal brownish, the rest of the fin whitish. 

Length of the body excluding the caudal 43 mm. (caudal fin about 
33 mm. long). 



The Fresh Water Fishes of the Island of Formosa. 279 

Descrilx'd from a specimen from Kotosho (Hotel Tobago Island), 
collected by Yonetaro Kikuchi. 

Habitat: Kotosho (Botel Tobago Island). A single specimen. 

Remarks: Macropodiis viridi-aiiratits Lacepfede'" = Macropodus ven- 
ustiis Cuv. & V'al., may be identical with the present species. Accord- 
ing to the description and figure by Cuvier and Valenciennes it is 
provided witii a distinctly forked caudal, instead of forming a bundle 
of long filaments as in M. filamentosiis. 

Family KUHLIID.E. 
Genus Kuhlia Gill. 

1861. Kuhlia Gill, Proc. Ac. Nat. Sci. Pliilad., p. 48. (Type Perca ciliata 

Cuv. & Val.) 
1863. Moronopsis Gill, Proc. Ac. Nat. Sci. Philad., p. 82. (Type Dules mar- 

ginala Cuv. & Val.) 
1872. Paradules Bleeker, Nederl. Tijdschr. v. Dierk. I, p. 257. (Type Dules 

marginaia Cuv. & VaL) 

Body elongate, fusiform. Head conical. Six branchiostegals. 
All the teeth villiform, without canines; teeth on the palatine bones. 
Anterior dorsal fin sustained by nine spines, the posterior with a spine 
and about eleven articulated rays; anal fin with three spines. Oper- 
culum with two or three flat spines; pre-operculum serrated. Scales 
moderate, minutely serrated. 

Distribution: Rivers of the intertropical regions (Java, Philippines, 
India, Ceylon, Formosa, Japan). 

59. Kuhlia marginata (Cuv. & Val.). 
Dokugyo (Japan). 

1829. Dules marginata Cuv. & Val., Hist. Nat. Poiss., Ill, p. 116, PI. LII; Java. — 
GiJNTHER, Cat. Fish., I, 1859, p. 268; Seas of Java, Amboyna, Vanicolo, 
Fiji Islands.— Day, Fish. Brit. India, 1876, p. 67, PL XVIII, Fig. i; Seas 
of India, Malay Archipelago. 

1903. Kuhlia marginata Jordan & Evermann, Proc. U. S. Nat. Mus., XXV, 
p. 340; Kotosho, Formosa. — Jordan & Richardson, Bull. U. S. Bur. Fish., 
XXVIII, 1907, p. 254; Mindoro Island. — Jordan & Richardson, Mem. 
Carneg. Mus., IV, no. 4, 1909, p. 183; Formosa. — Jord.an, SnIi'der & 
Tan.ak.\, Journ. Coll. .Sci., Tokyo, XXXIII, 1913, p. 146; Japan. 

Head 3.19 in length; depth 2.69; D. X, 11 ; A. Ill, 10; P. 14; V. I, 5; 
width of head 1.86 in its length; eye 4.5 in head; interorbital space 3; 

^'^ Macropodus viridi-auratus Lacepede, Hist. Nat. Poiss., Ill, 1802, p. 417, PL 
16, Fig. I. 

Macropodus venuslus Cuv. & Val., Hist. Nat. Poiss., VII, 1831, p. 375, PL 197. 



280 Annals of the Carnegie Museum. 

snout 3.4; pectoral 1.69; ventral 1.64; forty- two scales in the lateral 
line, four and one-half scales in an oblique series between origin of 
dorsal and lateral line, twelve scales between the latter and the middle 
of belly, nine scales between lateral line and the ventral; gill-rakers 
5 + 18; six branchiostegals. 

Body oblong, compressed, rather deep, dorsal and ventral profiles 
strongly convex; head moderate, with two sharp flat spines on the 
posterior margin of operculum, pre-opercles finely serrated; ventral 
surface of head convex and smooth; snout smooth, pointed anteriorly, 
interorbital space more or less convex; mouth oblique, with thin 
fleshy lips, its angles extending backward to the middle of the orbit 
below; lower jaw protruding; teeth on both jaws in villiform bands, 
vomer and palatines with a band of villiform teeth, no canine teeth; 
nostrils approximated, supralateral, in front of eye above, anterior 
nostril in a short tube. 

Origin of the dorsal opposite that of ventral, much nearer tip of 
snout than base of caudal, elongate, fourth and fifth spines longest, 
soft rays higher than spines, the base covered with a scaly sheath; 
pectoral lateral, reaching beyond the middle of ventral; ventrals close 
ogether, each armed with a short strong spine, the end of the fin- 
membrane adnate to the belly, tip not reaching the anal; anal fin rather 
long, with three strong spines, second strongest but shorter than the 
third, inserted in front of the origin of first dorsal ray, the base of the 
fin covered with a scaly sheath; caudal fin rather broad, divergent, 
very slightly emarginate; caudal peduncle rather long, its depth 2.33 
in length of head. 

Body covered with moderate ctenoid scales; cheeks and operculum 
covered with scales; lateral line continuous, slightly upcurved, running 
along the middle of the tail. 

Color in formalin dark grey above, belly and lower parts of sides 
yellowish; posterior margin of scales of the sides spotted with dark, 
their anterior margins finely speckled with black; dorsal fin dark 
gray, soft rays fuscous; pectorals dusky; ventrals white; membrane 
of the anal finely speckled with black; caudal fin dark brown. 

Length of body 205 mm. 

Described from a specimen from Tamusui River, collected by T. 
Aoki in December, 1916. 

Habitat: Tamusui River near Shinten; Choso River at Koshiryo; 
Giran River near Inzanpo; Bokusekikaku. 



The Fresh Water Fishes of the Island of Formosa. 281 

Rc))iarks: The present species is one (if llic brackish water lishes 
which enter the streams. 



Measurements of Kuhlia marginata. 



Locality. 



T3 


a 

Q 


P 


< 


zl 


> 


319 


2.69 


X, II 


III, 10 


14 I. 5 


.3 


2.72 


X, II 


III, 10 


14 I. 5 


3 


2.63 


X, II 


III, 10 


14 I. 5 


3-12 


2.75 


X, II 


III, 10 i4'I, 5 


3.28 


2.66 


X, II 


III, 10 


14 


1.5 






Taniusui River 
Tamusui River 
Tamusui River 

Inzanpo 

Koshiryo 



1.863 
1. 81 3.27 
1.76 3.18 

2 [3 
2 3 



3-40 
27 
25 



40 



4-50 
4.18 
3-80 
4 



503-25 



4^-42-12 
5 -40-13 
5 -40-13 
4 -40-12 
4I-40-12 



205 
181 
132 
132 
104 



Family OPHICEPHALID.T:. 
Artificial Key to the Formosan Genera of Ophicephalid.«. 

A. Ventral fins present Ophicephalus. 

B. Ventral fins absent Channa. 



Genus Ophicephalus Bloch. 

1794. ophicephalus Block, Naturg. Ausl. Fische, VIII, p. 137. (Type Ophiceph- 
alus punctalus Bloch.) 

Body elongate, subcj'lindrical anteriorly; head depressed, covered 
with plate-like scales. Eyes lateral. Gill-openings wide, the mem- 
branes of the two sides connected beneath the isthmus; four gills; 
pseudobranchiae none. A ca\Tty accessory to the gill-cavity, for the 
purpose of retaining water; no suprabranchial organ developed. 
Teeth in the jaws, vomer, and palate. One long dorsal and anal fin, 
without spines; ventral fins thoracic, composed of four to six rays, 
the outer of which is not branched. Lateral line interrupted. Air- 
bladder present. 

Distribution: British India; Ceylon; Borneo; Sumatra; Mindanao; 
Luzon; Bengal; Siam; Hindostan; Indo-China; Formosa; China; 
Amur Province. 

Synopsis of the Formosan Species. 

a. Depth of body less than 6 in the length; nine scales between orbit and the tip 
of operculum; eleven scales between lateral line and the root of ventrals. 

tadianus. 

aa. Depth of body more than 6 in the length; twelve scales between orbit and the 

tip of operculum; fourteen scales between lateral line and the root of ventrals. 

maculatus. 



282 Annals of the Carnegie Museum. 

60. Ophicephalus tadianus Jordan & Evermann. 
Raihii (Formosa). 

1903. Ophicephalus tadianus Jordan & Evermann, Proc. U. S. Nat. Mus., XXV, 
P- 330; Formosa. — Jordan & Richardson, Mem. Carneg. Mus., IV, no. 4, 
1909, p. 193; Formosa. — Vaillant, Bull. Mus. Hist. Nat., VI, 1904, p. 298; 
Tongkiiig. 

Head 3 in length; depth 5.33; D. 44; A. 28; \'. 4; P. 16; width of 
head 2 in its length; eye S in head; interorbital space 5; snout 6; 
maxillary 3; mandible 2.25; pectoral 2.2; ventral 3; fifty-seven scales 
in a longitudinal series, four scales between origin of dorsal and lateral 
line, eleven scales between the latter and the root of ventral, sixteen 
scales between lateral line and the middle of belly; nine scales between 
orbit and the angle of pre-operculum. 

Body torpedo-shaped, posterior part compressed, anterior part 
depressed; head rather elongate, depressed, its dorsal profile gradually 
inclined anteriorly; snout more or less produced, anterior margin 
obtusely rounded, interorbital space flat; eyes relatively small, superior 
and exceedingly anterior; nostrils separated, superior, the anterior in 
a short tube, in contact with the upper lip, the posterior in front of 
eye above; mouth oblique, large, its angle extending beyond the 
posterior margin of orbit; a band of small teeth on outer edge of jaws, 
a band of large, wide-set, caniniform teeth on palatine and inner 
side of lower jaw; upper jaw protractile, slightly shorter than the lower. 

Dorsal fin very long, with numerous spine-like rays, low anteriorly, 
gradually lengthening posteriorly, when depressed the tip of fin 
reaching beyond root of caudal; the pectoral obtusely rounded, 
nearly twice as long as broad, reaching beyond the middle of ventral; 
ventral fins small, not reaching vent; anal similar to the dorsal, 
inserted beneath anterior third of the base of dorsal, when depressed 
reaching beyond the root of caudal; caudal fin squarish, with rounded 
tip; caudal peduncle very short, deep, strongly compressed laterally. 

Body covered with rather large cycloid scales with irregular con- 
centric rings and radiated strise; top of head and cheeks covered with 
large scales; lateral line discontinuous, undulating, running along the 
center of body from base of caudal to just over third anal ray, thence 
upward for two rows of scales, forward sixteen rows, downward one 
row, and then forward to edge of gill-opening. 

Color in formalin oli\'aceous brown above, paler below; a row of 
twelve dark gray blotches along the base of dorsal, ])artly on the fin, 



The Fresh W'aticr Fishes of the Island of T'ormosa. 283 

parth' on both'; below those, hut above tlie lateral line, another row of 
nine similar, hut lari;er, hlotches, extending anleriorh' to the upper 
posterior border of eye, running below the upward curve of lateral 
line and forming a continuous longitudinal band; below this another 
row of about se\enteen similarly colored, irregular blotches, extending 
from the base of pectoral posteriorly along lower edge of lateral line 
to base of caudal; below this another row of irregular and lighter- 
colored blotches; head olixaceous brown, streaked with irregular lines 
of black above; a band of dark brown as wide as pupil, but becoming 
wider posteriorly, running from postero-inferior edge of eye to the 
middle of base of pectoral; lower part of head paler; dorsal fin marbled 
with dark; membrane of caudal fin dark, its rays dusky brown, two 
vertical stripes of gray color near the base; anal fin with seven dark 
blotches at the base of posterior half, its membrane dusky; the ventral 
pale; pectorals grayish, with no marking. 

Length of body 220 mm. 

Described from a specimen from Taihoku, collected by Oshima in 
October, 1916. 

Habitat: Widely distributed throughout the Island. My specimens 
came from Taihoku, Raupi, Giran; Tozen River; Nanshisho, Giran. 



Measurements of Ophicephalus ladianus. 



Locality. 













^^ 










T3 

u 


-a 


<i 


cu 


>■ 


. 




3 







K 


Q 






^S 


C-Q 


m 




t/3 


3 


5-33 44 


28 


16 


4 


2 


s 


6 


8 


4-5 7-14 


2.83 


S-50 41 


25 


IS 


4 


2.2 


4.66 


6.33 


8.25 


4-S4-I4 


3 


S.iil 44 


27 


16 


4 


2 


5-30 


S.60 


7-33 


5-56-14 






Taihoku . 
Taihoku . 
Raupi. . 



220 

215 

196 



61. Ophicephalus maculatus Lacepede. 

1802. Boslrichus maculatus Lacepede, Hist. Nat. Poiss., Ill, p. 140, 143. 

1831. Ophicephalus maculatus Cuv. & Val., Hist. Nat. Poiss., VH, p. 437. — 
Kner, Novara, Fisch., H, 1865, p. 234; Hongkong. — Bleeker, Naturk. 
Verh. A. K. Amst., XIX, 1879, p. 50; Philippine Islands. — Rutter, Proc. 
Acad. Nat. Sc. Philad., 1897, p. 69; Swatow. — Jordan & Evermann, Proc. 
U. S. Nat. Mus., XXV, 1903, p. 330; Formosa. — Jordan & Richardson, 
Mem. Carneg. Mus., IV, No. 4, 1909, p. 193; Formosa. — -Jordan & Richard- 
son, Check-List Philip. Fish., 1910, p. 34; Manila. 

i86i. Ophiocephalus maculatus Gijnther, Cat. Fish.. Ill, p. 480; China. 

Head 3.66 in length; depth 6.18; D. 40; A. 28; P. 17; V. 6; width of 
head 1.8 in its length; eye 8 in head; interorbital space 5; snout 6; 



284 Annals of the Carnegie Museum. 

pectoral 2.19; ventral 3.45; fifty-six scales in a lateral series, five scales 
between origin of dorsal and lateral line, fourteen scales between the 
latter and the root of ventral, sixteen scales between lateral line and 
the middle of belly, twelve scales between orbit and the end of pre- 
operculum. 

Body spindle-shaped, posterior part compressed; head broad, 
depressed; snout rather short, obtusely rounded anteriorly, inter- 
orbital space flat; eyes small, superior, and exceedingly anterior; 
nostrils separated, the anterior in a short tube, on upper edge of maxil- 
lary, the posterior in front of eye above; mouth oblique, large, its 
angle extending bei'ond the posterior margin of orbit; maxillary 3 in 
head; mandible 2.66; lower jaw slightly longer than the upper; a row 
of large caniniform teeth on palatine and inner side of lower jaw, a band 
of small teeth on outer edge of both jaws. 

The dorsal very long, with numerous spine-like rays, low anteriorly, 
when depressed reaching beyond the root of caudal; pectoral oval, 
outer margin rounded; ventrals rather small and slender, not reaching 
the vent; the anal similar to the dorsal, but not so long, inserted 
beneath the anterior third of the base of dorsal; caudal fin squarish, 
with rounded tip; caudal peduncle short and deep, greatly compressed. 

Body covered with large cycloid scales with irregular concentric 
rings and radiating striae; head and cheek covered with plate-like 
scales; lateral line broken anteriorly, running along the middle of 
sides from the base of caudal to just above the second anal ray, thence 
upward for two rows of scales, forward thirteen rows, downward one 
row, then forward to the upper edge of gill-opening. 

Color in formalin dark bluish gray above, paler below; belly dusky; 
eight large dark gray blotches above the lateral line, extending ante- 
riorly to the middle of and beneath the upward curve of the lateral 
line, forming a more or less continued longitudinal band; below this 
another row of about fourteen similarly colored, large,' irregular 
blotches, extending from the base of pectoral posteriorly to the base 
of caudal, running below the lateral line; interspace between the upper 
row of markings and the base of dorsal irregularly mottled with dark: 
upper surface of the head uniforml}- bluish gray, lower parts paler; a 
dark brown band from the upper posterior corner of the orbit run- 
ning backward, entering the upper row of markings of the sides; below 
this is an irregular, more or less undulating streak of the same color 
from eye to middle of base of the pectoral; upper part of the dorsal 



The Fresh \A'ater Fishes of the Island of Formosa. .285 

grayish, palrr below, its membrane with a series of dark spots along 
the base; caudal tin uniformly gray, with a faint strii)e near the base; 
anal fin gray, with indications of dark blotches at the posterior half of 
the base; ventral fins whitish, rays dusky; the pectoral gray. 

Length of bod}' 285 mm. 

Described from a specimen from W'odensho, Taichu, collected by 
T. Aoki in December, 1916. 

Habitat: Wodensho, Taichu (a single specimen). 

Remarks: The present species is easily distinguished from Ophi- 
cephalns tadianiis by its lower body; the greater number of scales 
between orbit and posterior angle of pre-operculum; and the greater 
number of scales in an oblicjue series between the lateral line and the 
root of ventral. 

Genus Ch.anna Gronow. 

1763. Channa Gronow, Zoophyl., p. 135. (Type Channa orienlalis Bloch & 
Schneider.) 

Body elongate, subcylindrical anteriorly, compressed posteriorly; 
head slightly depressed, covered with plate-like scales. Eyes lateral 
and anterior. Gill-openings wide, membranes of the two sides con- 
nected beneath the isthmus. Fine teeth in the jaws, on the vomer, 
and the palatine bones, intermixed with larger ones in the lower jaw. 
Pyloric appendages none. One long dorsal and anal fin, without 
spine; ventral fins none. Lateral line broken anteriorly. 

Distribution: Ceylon; China; Formosa. 

62. Channa formosana Jordan & Evermann. 
Kotai (Formosa). 

1903. Channa formosana Jordan & Evermann, Proc. U. S. Nat. Mas., XXV, p. 
330; Formosa. — Jordan & Richardson, Mem. Cameg. Mas., IV, No. 4. 
19C9, p. 193; Suwata (after Jordan & Evermann). 

Head 3.4 in length; depth 5; D. 44; A. 27; P. 14; width of head 
1.6 in its length; eye 7 in head; interorbital space 3.5; snout 5; maxil- 
lary 2.5; mandible 2.2; pectoral 1.6; fifty-four scales in a lateral 
series, five scales between origin of dorsal and lateral line, thirteen 
scales between the latter and the middle of belly, about six scales 
between orbit and the angle of pre-operculum. 

Body elongate, anterior part subcylindrical, compressed posteriorly; 
head rather broad, its top depressed; snout short, obtusely rounded 
anteriorly; mouth large, oblique, its angle extending beyond the 



286 



Annals of the Carnegie Museum. 



posterior margin of orbit; lower jaw slightly longer than upper; teeth 
in both jaws and on vomer, cardiform; eyes moderate, superior, and 
much anterior; nostrils widely separated, the anterior in a long tube, 
just behind the upper lip, the posterior in front of eye. 

Dorsal fin elongate, beginning over the base of pectoral, the rays 
quite uniform in length, when depressed reaching beyond the root of 
caudal; anal fin similar to dorsal but shorter, beginning beneath the 
fifteenth dorsal ray and ending under the forty-second; rays of the 
dorsal and anal all unbranched; pectoral fin broad, not reaching the 
vent; no ventral fins; caudal fin broad and rounded; caudal peduncle 
short and deep, greatly compressed. 

Lateral line broken anteriorly, running along the middle of the 
sides from the base of caudal to just above second anal ray, thence 
upward one row for one scale, again upward for one row, extending 
forward on seven scales, thence dropping one row, reaching to the 
upper extremity of gill-opening; body covered with large cycloid scales 
with irregular concentric lines and radiating striae; head and cheeks 
covered with plate-like scales. 

Color in formalin yellowish brown above, paler below; the sides 
with about nine V-shaped dark cross-bars, the apex pointing forward, 
these markings clearer posteriorly and more or less broken and irregu- 
lar in front; a large round black spot, bordered by white, on caudal 

Measurements of Channa formosana. 



Locality. 


Head. . 
Depth. 


Q 


< 


Ph" 


Width 
of Head. 

Interor- 
bital. 


3 

a 


0) 

W 


Scales. 

Length, 
Mm. 


Taihoku 

Taihoku 


3-40 5 
3-64! 5-5 


44 
45 


27 
28 


14 
15 


1.60 3-5 
1-57 3-5 


5 

4-5 


7 
6.86 


5-54-13 210 
5-53-13200 



peduncle near the base of caudal fin; sides of head with two broad, 
dark streaks from eye to the posterior edge of operculum, more or less 
undulating; dorsal and anal fins uniformly dusk\' gray, their edge 
somewhat darker; other fins grayish white. 

Length of body 210 mm. 

The present description is from a specimen from Taihoku, collected 
by Oshima in September, 1916. 

Habitat: Taihoku; Shori, Toyen; Tamusui Ri\er; Jitsugetsutan 
(Lake Candidius). One of the commonest fishes in ponds and stag- 
nant pools. 



The Fresh \\'ater Fishes of the Island fo Formosa. 287 

Rctuarks: This species is very closely allied to CIkiiuki occUata^^ Iroin 
China. It differs in having no teeth on palatines and a greater num- 
ber of scales in a transverse series. 

Family GOBIID.^. 
Artificial Key to the Formosan Genera. 

A. Ventral fins entirely separated; pectorals normal; eyes not erectile. 

a. Pre-opercle with a concealed, hook-like spine; scales moderate, ctenoid; 

dorsal spines low; interorbital space without ridge Eleotris. 

aa. Pre-opercle without spine; scales moderate, ctenoid; dorsal spine low; 
interorbital space with prominent ridges Bulls. 

B. \'entrals joined at least at base. 

a. Ventrals adherent to the belly; body scaly; teeth of the upper jaw mov- 
able, in a single series Sicyoplerus. 

aa. Ventrals not adherent to the belly; body scaly; teeth conical, fixed, those 

of the upper jaw in several series. 

b. Soft dorsal and anal short, each composed of nine to twelve soft rays. 

c. Tongue truncate or rounded or pointed at tip; gill-openings 

chiefly confined to the sides; pectorals without silk-like rays 

above; dorsal spines rather weak, some of them often elongate. 

Rhinogobiiis. 

cc. Tongue emarginate at tip; gill-openings extending forward 

below; pectorals without silk-like rays above; scales rather large, 

about fifty Glossogobius . 

bb. Soft dorsal and anal long, the former composed of fourteen to thirty 

rays, dorsal spines seven to nine. 

c. Scales moderate; soft dorsal composed of fourteen or fifteen rays; 

cheeks scaly at least above; pectorals without free silk- like rays 

above Acanthogobius . 

Genus Eleotris (Gronow) Schneider. 

1763. Eleotris Gronow, Zoophyl., p. 83 (non binomial). 

1801. Eleotris Schneider, Syst. Ichth., p. 65. (Type Gobius pisonis Gmelin). 

1874. Culius Bleeker, Archiv. Neerl., IX, p. 303 (Poecia fusca Schneider). 

Body long and low, compressed behind. Head long, low, Battened 
above, without spines or crests, almost everywhere scaly. Mouth 
large, oblique, lower jaw projecting. Lower pharyngeals rather broad, 
the teeth small, bluntish. Pre-opercle with a small concealed spine 
below, its tip hooked forward. Branchiostegals unarmed. Tongue 
broad, rounded. Posttemporal bones very strongly divergent, their 
insertions close together; top of skull somewhat elevated and declivous; 
interorbital area slightly convex transversely; dorsal fins well apart, 

1' Channa ocellala Peters, Monatsb. Acad. Wiss. Berlin, 1864, p. 384. China. 



1288 Annals of the Carnegie Museum. 

the first composed of five or six low, flexible spines; ventrals separate. 
Scales moderate, ctenoid, forty-five to seventy in a longitudinal 
series; vertebrae (pisonis) 11-15. Tropical seas, entering fresh waters 
(Jordan & Snyder). 

Distribution: Freshwater fishes of the tropics, some of the species 
entering the sea; cosmopolitan. 

Synopsis of the Formosan Species. 

a. Scales of moderate size, about 50 in a lateral series, 15 between origins of the 

second dorsal and anal oxycephala. 

aa. Scales small, about 60 in a lateral series, 16 between origins of the second 
dorsal and anal fusca. 

63. Eleotris oxycephala (Schlegel). 
Doman (Lake Biwa, Japan); Onkora (Formosa). 

1845. Eleolris oxycephala Schlegel, Fauna Japonica, Poiss., p. 150, PI. LXXVII, 
Fig. 4, 5; Nagasaki. — GOnther, Cat. Fish., Ill, 1861, p. 116; China. — Kner 
Novara, Fisch., II, 1865, p. 185; China. — Jordan & Snyder, Proc. U. S. 
Nat. Mas., XXII, 1900, p. 371; Lake Biwa.— Proc. U. S. Nat. Mus., XXIV, 
1901, p. 46; Haneda; Wakayama. — Jordan & Richardson, Mem. Carneg. 
Mus., IV, No. 4, 1909, p. 200; Takao, Formosa. 

1846. Eleolris canthrinus Richardson, Ichthyol. China, p. 209; Macao. 

1905. Eleotris balia Jordan & Seale, Proc. U. S. Nat. Mus., XXIX, p. 526; 
Hongkong. 

Head 3.33 in length; depth 5; D. VI, 8; A. 9; P. 18; V. 5; depth of 
caudal peduncle 2.33 in head; eye 6; interorbital space 3; snout 4; 
fifty scales in a lateral series, fifteen scales between origins of second 
dorsal and anal. 

Body deep, caudal peduncle strongly compressed; head long, much 
depressed, considerably broader than the body, a longitudinal groove 
along the median line of the top; snout rather short, broadly rounded 
anteriorly, interorbital space flat; mouth large, terminal, slightly 
oblique, its angle extending to a vertical through the anterior liorder 
of orbit; lips thick; lower jaw projecting beyond the upper; teeth on 
both jaws, simple, in rather broad bands, inner ones larger; palatines 
and vomers without teeth; eyes supsrior and anterior; tongue oblong, 
anterior edge rounded; nostrils separated, very small, anterior nostril 
in a short tube; gill-openings lateral, not extending very far forwards; 
posterior border of the pre-opercle with a hidden spine, which projects 
downward and forward; gill-rakers 3 +9, short and covered with 
<ielicale setai. 



The Fresh Water Fishes of the Island of Formosa. 289 

Two pairs of dorsal fins; first dorsal rather small, when depressed 
reaching the insertion of the second dorsal, posterior ray of the latter 
longest; pectorals acutely rounded, reaching to the posterior end of the 
base of the first dorsal; ventrals separated, rather small; the anal 
inserted beneath the third soft ray, when depressed reaching so far as 
the end of second dorsal; caudal fin rounded; anal flap developed. 

Head with minute scales except snout, chin, and throat; occiput and 
cheeks with small cycloid scales; body covered with uniform ctenoid 
scales, those on belly, breast, and nape are minute and cycloid. 

Color in formalin uniformly dark brown, ventral parts whitish, 
densely speckled with minute brown spots; a wide black stripe, made 
up of small black dots, extending from opercle to caudal; a distinct 
narrow black line through eye from snout to origin of dark stripe on 
bod>'; another short black line from posterior margin of orbit to pos- 
terior margin of pre-^opercle; first dorsal with two brown bands which 
run horizontally; second dorsal and anal with numerous dark spots; 
caudal and pectorals with small brow'nish spots arranged in vertical 
rows. 

Length of body 112 mm. 

Described from a specimen from Tozen River, collected by T. Aoki 
in December, 1916. 

Habitat .-Tamusui River near Maruyama; Tozen River; Giran River 
near Inzanpo and Kiburan; Buroko River near Suwo. 



Measurements of Eleotris oxycephala. 



Locality. 





13 


a 


3-33 


3 


26 


3 


35 


3 


15 


3 


28 


3 


3-45 


3 


3 


26 



< 


cu 


> 


Depth of 

Caudal 

Peduncle. 


c 2 
S2 


9 


18 


5 


2-33 


3 


9 


17 


5 


2.33 


3-37 


9 


18 


5 


2.13 


3 


9 


19 


5 


2.22 


3-25 


9 


19 


5 


2.40 


350 


9 


17 


5 


2.50 


3-44 


9 


19 


5 


2.50 


3-40 


9 


18 


5 


2.50 


3-75 


8 


19 


5 


2.25 


3 





m 


>> 




U 







W 



tag 



Tozen River. . . 
Tamusui River. 
Tamusui River. 

Inzanpo 

Kiburan 

Kiburan 

Kiburan 

Kiburan 

Giran 



5 VI. 8 
5.22iVI, 9 
4.81 VI, 9 

4.25 VI, 9 
4.50JVI, 8 
4 jVI, 9 
5-26JVI, 9 
4.60JVI, 9 
4.66 VI, 9 



4 

3-83 
3.66 
390 

4 

4 

3-77 

4 

3-44 



6 50-15 
6.5o!48-i5 

7 150-16 
7 48-15 

6 [48-15 
5.40 46-15 
6.66148-15 
5.66J48-15 

7 I50-15 



112 
109 
120 
150 
105 
113 
130 
no 
165 



64. Eleotris fusca (Schneider). 

1801. Pcecia fusca Schneider, Bloch. Syst., p. 453. 

1861. Eleotris fusca Gunther, Cat. Fish., Ill, p. 125; Ganges; Calcutta; Bengal; 
Amboyna; Aneitum; Oualan; Wanderer Bay; Ceylon; Canton. — Ishikawa, 
20 — JAN. 12, 19 19. 



290 Annals of the Carnegie Museum. 

Prel. Cat. Fishes, 1897, p. 31; Riukiu Islands. — Day, Proc. Zool. Soc. 

London, 1869, p. 519. — Kner, Novara Fisch., II, 1865, p. 186; Ceylon; 

Nicobar; Tahiti. — Day, Fish. Brit. India, p. 313. — Peters, Monatsb. 

Konig. Akad. Wiss. Bed., 1868, p. 268; Luzon; Samar. — Boulenger, Ann. 

Mag. Nat. Hist. (6), XV, 1895, p. .186; Palawan. — Jordan & Snyder, 

Proc. U. S. Nat. Mus., XXIV, 1902, p. 45; Honolulu. — Jordan & Evermann, 

Proc. U. S. Nat. Mus., XXV, 1903, p. 361; Suwata, Formosa.- — Jordan & 

Richardson, Mem. Carneg. Mus., IV, no. 4, 1909, p. 200; Formosa. — Bull. 

U. S. Bur. Fish., XXVII, 1908, p. 274; Sibuj-an; Apari. — Check-list Philip. 

Fish., 1910, p. 45. 
1822. Cheilodipterus culius Buchanan-Hamilton, Fish. Ganges, p. 55, PL V, 

Fig. 16; Ganges. 
1824. Eleotris nigra Quoy & Gaimard, Zool. Voy. Freycinet, p. 259, PI. LX, 

Fig. 2; Guam, Waigiou.^ — Cuv. & Val., Hist. Nat. Poiss., XII, 1837, p. 235; 

Isle de France; Ganges; Malabar; India; Java; Otaiti; Borabora; Society 

Island; Madagascar. 
1831. Eleotris mauriiianus Bennett, Proc. Comm. Zool. Soc, I, p. 166; Mauritius. 

Head 3.39 in length; depth 5.33; depth of caudal peduncle 2 in 
head; eye 5; interorbital space 3.66; snout 4; maxillary 2.4; D. VI, 9; 
A. 9; P. 18; V. I, 5; sixty-four scales in a lateral series, sixteen scales 
between origins of the second dorsal and anal. 

Body rather depressed, posterior part compressed; head long, 
depressed, broader than body, a shallow longitudinal groove along the 
median line of the top; snout rather short, broadly rounded anter- 
iorly, interorbital space more or less concave; mouth large, terminal, 
slightly oblique, its angle reaching a vertical through the anterior 
border of orbit; lips thick; lower jaw protruding; teeth in villiform 
bands, those on the outer series somewhat enlarged; palatines with no 
teeth; eyes superior and anterior; tongue oblong, with rounded anterior 
margin; nostrils separated, minute, the anterior located just behind 
the upper lip, the posterior approximated to eye above; gill-openings 
lateral, not extending very far forward; a spine at the angle of pre- 
operculum, projecting downward and forward. 

Dorsal fin separated; first dorsal rather small, when depressed 
scarcely reaching the root of second dorsal; second dorsal without 
filamentous rays, posterior ray the longest; pectoral fins acutely 
rounded, reaching beyond the posterior end of base of first dorsal; 
ventrals separated, rather short; the anal inserted beneath third dorsal 
ray, when depressed reaching end of second dorsal; caudal fin rounded; 
anal flap developed. 

Body covered with minute ctenoid scales; scales on the upper 
surface of the head small and cycloid, extending forward to the inter- 



The Fresh Water Fishes of the Island of Formosa. 291 



orbital space; cheeks covered with minute scales; those on the belly 
minute and cycloid. 

Color in formalin dark gray above; lower surface pale brown, with 
minute brown spots; pectorals and anal with a number of rows of 
faint brown spots; dorsal and caudal fins with several series of brown 
spots; ventral fins indistinctly spotted with dark. 

Length of body 94 mm. 

Described from a specimen from Buroko River near Suwo, col- 
lected by T. Aoki in August, 191 7. 

Habitat: Buroko River near Suwo (a single specimen). 

Remarks: The head of the above-described specimen is' a little 
longer than that of the specimen described by Giinther or Day. But 
as the specimens of Eleotris fusca from Sumatra, which are contained 
in the Stanford collection have longer heads, there is no doubt with 
reference to the identity of the present species. 

Measurements of Eleotris fusca. 



Locality. 


•a" 


H. 



Q 


< 


fc 


>■ 




2-5 

c3 


3 

C 
72 


>> 

W 


"3 

Hi 


SbS 


Bliroko River 


3.39 tJ-TJ 


VI, 9 
VI, 9 
VI, 9 


9 
9 
9 


18 
18 
18 


i.S 
1.5 
1.5 


2 


3-66 

3-55 
3.22 


4 
4 
4 


t; ,fi/i— ifii o/i 


Sumatra (No. 8009; S. U.). . . 
Sumatra (No. 8009; S. U.). . . 


3-37 
3-43 


4.71 
5-39 


2.21 
2.23 


6 
6.5 


60-16 
60-16 


128 
116 



Genus BuTis Bleeker. 

1874. Biitis Bleeker, Arch. Neer. Sc. Nat., IX, p. 543. (Type Cheilodipterus 
bulls Hamilton.) 

Body elongate, posterior part compressed, somewhat cylindrical 
in front. Head elongate, strongly depressed, its upper surface 
gradually inclining anteriorly; snout produced; interorbital space 
provided with a pair of bony ridges along the superior margin of the 
orbit. Cheeks rather rough, sometimes with striae. Teeth in many 
series, set close together, minute and subequal, those of the outer row 
somewhat larger; tongue and palate edentulous. Pre-operculum 
without a spine. Scales large, about thirty in a longitudinal series; 
scales on occiput extending to the interorbital space; cheek covered 
with scales. 

Distribution: Philippine Islands; Malayan Archipelago; India to 
China; Formosa. 



292 Annals of the Carnegie Museum. 

65. Butis butis (Hamilton). 
1822. Cheilodipterus bulis Hamilton, Fish. Ganges, pp. 57, 367. 
1849. Eleotris butis Cantor, Catal. Malay. Fish., p. 196 — Giinther, Cat. Fish., 

Ill, 1861, p. 116; China; Philippine Islands; Borneo; Amboyna; Penang; 

India. — Day, Fish. Brit. India, p. 315; PI. LXVIII, Fig. 3; Seas and estuaries 

of India to the Malay Archipelago. — Sauvage, Bull. See. Philom., 1883, p. 8; 

Siam. 
1861. Butis bulis Bleeker, Versl. Akad. Amst., XII, p. 77; Penang. — Boulenger, 

Ann. Mag. Nat. Hist., Ser. 6, XV, p. 186; Palawan. — Jordan & Rich.\rdson, 

Bull. U. S. Bur. Fish., XXVII, 1908, p. 125; Manila; Iloilo. 
1849. Butis melano stigma Bleeker, Verh. Bat. Gen., XXII, p. 23; Madras Straits. 

— Bleeker, Arch. Neerl. Sc. Nat., X, 1875, P- 68. 
1905. Butis leucurus Jordan & .Seale, Proc. U. S. Nat. Mus., XX\'III, p. 794, 

Fig. 13; Negros. — ^Jordan & Seale, Bull. U. S. Bur. Fish., XXVI, 1906; 

p. 40; Negros. — Evermann & Seale; Bull. U. S. Bur. Fish., XXVI, 1906 

(1907), p. 104; Bacon. 

Head 3 in length; depth 4.33; depth of caudal peduncle 3 in head; 
eye 5; interorbital space 3.83; snout 2.5; D. VI, 9; A. 9; P. 20; V. 1,5; 
thirty scales in a lateral series, nine scales between origins of second 
dorsal and anal; gill-rakers 2+6. 

Body elongate, compressed, deepest in front of the first dorsal; 
head long, flattened, its upper surface and cheeks covered with minute 
scales; pre-operculum with no spine; snout depressjed and produced, 
tip slightly swollen, anterior margin rounded; interorbital space more 
or less flat, covered with minute scales; mouth large, oblique, its angle 
not reaching the vertical through anterior margin of orbit; lower jaw 
protruding; teeth in villiform bands; palatines and vomers with no 
teeth; eyes separated, superior, slightly nearer to tip of snout than end 
of operculum; nostrils separated, the anterior in a very short tube; a 
finely serrated ridge along the posterior and superior edge of the orbit, 
interspace between the ridge and orbit covered with two series of small 
scales. 

Dorsal fin separated; spinous dorsal rather small, when depressed 
reaching to the origin of soft dorsal; height of soft dorsal rays subequal; 
pectorals lateral, middle ray elongate; ventrals separated, inserted in 
advance of the origin of pectoral; anal fin opposite the soft dorsal, 
inserted beneath the third ray, posterior rays longer; caudal fin rather 
short, squarish, tip rounded. 

Body covered with large ctenoid scales, those on the head are minute 
and cycloid. 

Color in alcohol uniformh' dark gray, with lighter longitudinal 



The Fresh Water Fishes of the Island of Formosa. 293 

stripes along the sides; a deep brown spot on the lower half of the 
base of the pectoral; spinous dorsal nearly black; soft dorsal somewhat 
paler, with dark spots; caudal and anal fms dusk>' with darker spots; 
pectoral tins pale gray. 

Length of body 93 mm. 

Described from a specimen from Daitotei fish-market, Taihoku. 

Habitat: Taihoku (a single specimen). 

Genus Sicyopterus Gill. 

i860. Sicyopterus Gill, Proc. Acad. Nat. Sci. Philad., p. loi. (Type Sicydium 
slimpsoni Gill.) 

Body subcylindrical, covered with ctenoid scales of rather small 
size; head as broad as high; mouth nearly horizontal, with the upper 
jaw prominent; lips very thick and fleshy. The upper jaw with a 
series of numerous small teeth, implanted in the gum and movable; 
lower jaw with a series of conical, widely set teeth. Eyes of moderate 
size. Two dorsal fins, the anterior with six flexible spines, of which 
the third is filiform; caudal fin cjuite free; ventral fins united to a small 
semicircular disc, more or less adherent to belly. Gill-openings of 
moderate width; four branchiostegals; no air-bladder. 

Distribution: Hawaii; Japan; Philippine Islands; Formosa. 

66. Sicyopterus japonicus (Tanaka). 
Bozuhaze (Japan); Fushunhii (Formosa). 

1909. Sicydium japonicum Tanaka, Journ. Coll. Sci., Tokyo, XX\TI, p. 22; 

Tosa, Japan. 
1913. Sicyopterus japonicus Jordan, Sn\'der, & Tanaka, Journ. Coll. Sci., 

XXXIII, p. 431; Tosa; Kii; Kinokawa. — Tanaka, Fishes of Japan, XI, 

p. 203; Kii; Hiuga. 

Head 4.21 in length; depth 5.88; D. \T, 11 ; A. 10; P. 19; eye 6.5 in. 
head; interorbital space 2.66; snout 2.18; depth of caudal peduncle 
1. 71; fifty-nine scales in a lateral series; sixteen scales in a transverse 
series. 

Body elongate, subcylindrical, slightly compressed; head rather 
large, round, its top slightly depressed; snout broadly rounded ante- 
riorh-; mouth horizontal, its angle reaching the vertical through an- 
terior margin of orbit; upper jaw prominent; lips thick and fleshy, a 
series of numerous minute teeth in upper jaw, implanted in the gum 
and movable; lower jaw with a series of conical widely-set teeth; eyes 
small, superior, and in the middle of the head; nostrils widely sepa- 
rated, the anterior in a small round hole, the posterior slit-like and 



294 



Annals of the Carnegie Museum. 



minute; gill-openings lateral, in front of the base of the pectoral; gill- 
membranes broadly united to isthmus. 

Two dorsal fins, the first dorsal inserted above the posterior half of 
pectoral, nearer to the origin of second dorsal than to the tip of snout, 
with six flexible spines, of which the third is filiform, reaching the 
anterior third of the base of second dorsal, fin-membrane united to the 
back behind the last spine; second dorsal opposite to the ventral, 
elongate, its rays subequal in height; pectoral fins lateral, leaf-shaped, 
with rather sharp tip, reaching midway between origin of ventral and 
that of the anal; ventrals united, forming a small semicircular disk, 
more or less adherent to the belly; origin of the anal slightly behind 
that of the second dorsal, similar in form to the latter; caudal fin 
rounded. 

Body covered with ctenoid scales, those on the anterior parts of 
body and the base of caudal cycloid; head entirely naked; no lateral 
line. 

Color in formalin brownish gray, paler below; dorsal surface with 
about ten black cross-bands, some of which extend downwards, first 
band in front of the base of pectoral, second one in front of the origin 
of first dorsal; head uniformly dark gray; all the fins dark gray; first 
dorsal sparingly spotted with dark, each ray of the second dorsal with 
a series of pale brown spots; proximal part of caudal fin darker. 

Length of body 120 mm. 

Described from a specimen from Tamusui Ri\er near Shinten, 
collected by T. Aoki in December, 1916. 

Habitat: Tamusui River; Tozen River; Koranronsho, Taichu; 
Shinchiku; Bokusekikaku; Choso River (Koshiryo); Taiko. 

Measurements of Sicyopterus japoniciis. 



Locality. 



Tamusui River 
Tamusui RivxT 
Tamusui River 
Tamusui River 
Tozen River. . . 
Tozen River. . . 
Choso River. . . 

Heirinbi 

Taiko 



-a 


s. 


e 




<i 


a," 


4.21 


5-85 


VI. 




10 


19 


4-13 


6.33 


VI. 




10 


18 


4-32 


6.25 


VI, 




II 


18 


4-32 


6.33 


VI. 




II 


19 


4-38 


5-88 


VI. 




II 


19 


4.40 


5-65 


VI. 




II 


19 


4-47 


5.80 


VI. 




10 


19 


4.18 


4.42 


VI. 




10 19 


4-34 


5-41 


VI. 


II 


io|i9 



2.66 

2.44 
2.44 

2.40 

2.7s 
2.50 
2.50 
2.40 
2.31 



2.18 



2.10 
2.27 

2.33 

2 

2.31 




1. 12 60-15 
1-69 59-16 

I. S3 160-16 



hJ 



120 
116 
120 
IIS 
113 
140 
105 
112 
166 



The Fresh Water Fishes of the Island of Formosa. 295 

Genus Rhinogobius Gill. 

1859. Rhinogobius Gill, Proc. Ac. Nat. Sci. Philad., p. 145. (Type Rhinogobius 
siinilis Gill.) 

Body oblong, compressed. Head oblong, not much compressed. 
Eyes high, anterior, close together; opercles unarmed. Mouth 
moderate, the lower jaw unusually short. Teeth on jaws only; conical, 
ill few or several series, those in the outer row enlarged; no large 
canines; tongue usually truncate. Isthmus broad. Shoulder-girdle 
without fleshy "flaps or papill?e. Skull depressed, abruptly widened 
behind the eyes and without distinct median keel. Scales moderate 
or large, ctenoid, permanently covering the body; cheeks naked; 
opercles naked, or scaled above only; belly generally scaly. Dorsal 
with six rather weak spines; pectoral well-developed, the upper rays 
without free or silk-like tips; ventrals completely united, not adnate 
to the belly; caudal lin usually obtuse. (Jordan & Snyder.) 

Distribution: Philippine Islands; Formosa; China; Corea; Japan; 

North America. 

Synopsis of the Formosan Species. 

I. Opercles entirely naked. 

a. Nape with a naked area. 

b. Dorsal spines filamentous. 

c. Scales about thirty-six; head with many raucous pores. taiwanus. 
bb. Dorsal spines not filamentous. 

c. Scales thirty-six to thirty-eight; head without mucous pores. 

candidius. 

cc. Scales twenty-nine; head without mucous pores formosaniis. 

aa. Nape closely scaled. 

b. Scales about twenty-nine; dorsal spines not filamentous; teeth in two 

series giurinus. 

bb. Scales about twenty-eight; dorsal spines not filamentous, teeth in 
villiform band, lower jaw with two canine teeth on each side. caniniis. 

67. Rhinogobius candidius (Regan). 

1908. Ctenogobiiis candidius Regan, Ann. Mag. Nat. Hist. (8), I, p. 153; Lake 
Candidius, Formosa. 

Head 3.77 in length; depth 5.5; caudal peduncle 2. 11 in head; eye 5; 
interorbital space 4.75; snout 2.71; D. VI, 9; A. 9; P. 20; thirty-six 
scales in a lateral series, ten scales between origins of the second dorsal 
and anal; gill-rakers 3 -|- 6. 

Body robust, a little deeper than wide; caudal peduncle compressed; 
head slightly depressed; snout rather long and sharp; mouth terminal, 
a little oblique, its angle not reaching a vertical through the anterior 



296 



Annals of the Carnegie Museum. 



margin of orbit; jaws subcqual; lips thick; teeth on both jaws in 
villiform bands, those on the outer series conical and somewhat larger; 
tongue broad, anterior border rounded; nostrils separated, the anterior 
in a short tube; eyes high, upper margin more or less projecting; 
interorbital space concave. 

Dorsal fin separated, none of the spines elongated, when depressed 
not reaching the base of second dorsal, soft rays a little longer than 
the spines, rather high; pectorals large, not reaching the vent; ventrals 
united, forming a concave disk which is not adnates to* the belly; anal 
fin inserted below the third dorsal ray, when depressed extending 
beyond the end of second dorsal, sixth and seventh rays the longest; 
caudal fin quadrate, tip rounded. 

Head and nape entirely naked; scales in front of first dorsal as well 
as those on the breast and belly minute and cycloid, rest of the 
scales ctenoid. 

Color in alcohol pale olive-gray, sides with seven brownish broad 
cross-bars; back mottled with dark; a black elongated spot on the 
base of the caudal; lower surface whitish; first dorsal dusky, second 
dorsal spotted with black; caudal fin dusky; other fins whitish. 

Length of body 84 mm. 

Described from a specimen from Bokusekikaku. collected by Yone- 
taro Kikuchi. 

Habitat: Bokusekikaku; Heirinbi; Tozen River; Shinchiku; Daiko 
River; Taito River. Lake Candidius (Regan). 

Remarks: The nearest relative of the present species is Rhinogobius 
bedfordi (Regan)^- from Corea. It differs in having the second spine 
of the first dorsal produced into a filament. 

Measurements of Rhinogobius candidius. 



Locality. 



Bokusekikaku . 
Bokusekikaku . 

Heirinbi 

Tozen River. . 
Shinchiku. . . . 
Daiko River. . 
Daiko River. . 
Daito River. . 



1^1 



77 


5-50 


VI, 9 


9 


20 


52 


5-35 


VI, 9 


9,20 


63 


6.22 


VI, 9 


9 |I9 


58JS-2S 


VI, 9 


9 


16 


33 5-66 


VI, 9 


9 


19 


75|5-50 


VI, 9 


9 


19 


25i 5-42 


VI, 9 


9 


17 


62 


5-33 


VI, 9 


9 


18 



2.66 

2 

2.28 
2.16 
2.16 



4-75 
4.66 

5 
5 
5 
5 
6 
4.66 



2.71JS I36-10 
2.63 5.5038-10 
2.42I4.66I37- 9 
2.42, 4.5035-10 
2.50! 4 35-10 
2.60 4 136-10 
2.60] 4.50J35-10 
2.75 4-50135- 



90 
70 
72 
61 

55 
48 
57 



^'^ Ctenogobius bedfordi Regan, Proc. Zool. Soc. London, 1908, p. 62; Chong-ju, 
Corea. 



The Fresh Water Fishes of the Island of Formosa. 297 
68. Rhinogobius giurinus (Rutter). 

1897. Gohius giurinus Rutter, Proc. Acad. Nat. Sc. Philad., p. 86; Swatow, 

China. 
1901. Gobius giurus Abbott. Proc. U. S. Nat. Mus., XXIII, p. 491; Tiensin. 
1903. Ctenogohius plalycephahis Jord.\n & Ev^ermann, Proc. U. S. Nat. Mus., 

XXV, p. 362; Taihoku. 
1909. Rhinogobius giurinus Jordan & Rich.\rdson, Mem. Carneg. Mus., I\', 

no. 4, p. 200; Kotosho; Taihoku, Formosa. 

Head 3.47 in length; depth 4.33; depth of caudal peduncle 2.5 in 
head; eye 4; interorbital space 7.66, snout 2.5; D. VI, 9; A. 10; P. 19; 
twenty-nine scales in a lateral series, nine scales between origins of 
the second dorsal and anal. 

Body elongate, a little deeper than wide, posterior part compressed; 
head broader than deep; snout rather long, sharp, anterior margin 
acutely rounded; interorbital space concave, narrow; mouth moderate, 
somewhat oblique, its angle scarcely reaching a vertical through 
anterior margin of orbit; lips thick; jaws subequal; teeth in two series, 
outer ones small, canine-like, second series in upper jaw minute, no 
large caniniform teeth; tongue broad, the tip truncate; eye high in 
head, upper margin projecting, situated midway between tip of snout 
and posterior edge of opercle; gill-openings not extending far forwards; 
gill-rakers 2 + 8 on first arch; no barbels on lower jaw. 

Dorsal fin separated, rather short, spinous rays not filamentous, 
when depressed not reaching the second dorsal; soft rays a little longer 
than the spines: pectorals rounded, extending to the vent above; 
ventra'.s united, not adnate to the belly, the tips not reaching the 
vent; the anal inserted below second ray of second dorsal, when 
depressed extending as far posteriorly as the second dorsal; caudal iin 
rounded. 

Head, except the occiput, naked; body covered with large, regular 
ctenoid scales; those on the occiput smooth and cycloid. 

Color in formalin pale olive-gray above, lower parts whitish; head 
spotted with dark, cheek with oblique, rather wavy bars; sides with a 
number of irregular black spots; back mottled with dark; occiput with 
closely crowded dark blotches; dorsal fins with dusky spots arranged 
in longitudinal rows; caudal fin fuscous, with indistinct wavy cross- 
band near the base; pectoral fins dark; ventrals and anal whitish. 

Length of body 65 mm. 

Described from a specimen from Shimo-Tamusui River near Ako, 
collected by T. Aoki in December, 1916. 



298 



Annals of the Carnegie Museum, 



Habitat: Shimo-Tamusui River near Ako; Tozen River; Wodensho, 
Taichu; Daito River; Tamusui River; Shori, Toyen; Hyoko and Maru- 
yama, Giran. 

Remarks: In the year 1903, Jordan and Evermann described the 
present species from Kotosho and Taihoku under the name Cleno- 
gohiiis platycephalus (Richardson), though there were some doubts 
with reference to its identity. In Jordan and Richardson's paper, 
however, the name has been changed to Rhinogohins giiirinus (Rutter). 

The present species is very closely allied to Rhinogobius hadropterus^^ 

from Japan. But it differs in having a greater number of scales in the 

lateral series. 

Measurements of Rhinogobius giurinus. 



Locality. 



-a 
0. 

Q 




d 


< 


Dh' 


^^4 

•5-0 c 


4-33 


V] 


[. 9 


10 


19 


2.50 


5-17 


VJ 


, 10 


10 


20 


3 


5-70 


V] 


[. 9 


9 


20 


2.83 


6 


V 


. 9 


9 


20 


2.83 


•S 


V] 


. 9 


9 


20 


2.80 


S-36 


V 


[, 10 


10 


19 


2.63 


5.66 


V] 


. 9 


9 


19 


3-14 


5-10 


V] 


. 9 


9 


20 


2.66 


5-21 


VJ 


. 9 


9 


19 


2.75 


S-75 


VJ 


, 10 


10 


19 


2.71 


5.66 


V 


. 9 


9I19 


3 


5 


VJ 


. 9 


9 19 


2.33 


6 


VJ 


. 9 


9 


20 


2.57 



cl5 



V 


m 


>. 




W 


u 




W 


4 


29-9 


4-33 


29-9 


4 


30-9 


4-25 


30-8 


4.66 


30-8 



►J 



Shimo-Tamusui River 
Shimo-Tamusui River 
Shimo-Tamusui River 

Tamusui River 

Shori 

Sliori 

Daito River 

Wodenslao 

Tozen River 

Heirinbi 

Hyoko 

Maruyama 

Swatow (Type) 



7.661 2.50 



7.66 

6 

6.66 

7-50 

7 

7-50 

6.66 

6.50 

5-66 

5 

5 

7-33 



2.71 
2.71 
2.63 
2.50 



2.50 


4.60 


29-8 


2.551 5 


31-8 


2.80 4.25 


29-8 


2.44 


5-40 


30-9 


2.50 


4.66 


30-8 


3 


4 


3»-8 


2.50 


4 


30-9 


2.57 


4 


30-9 



65 
78 
74 
73 
60 
89 
83 
63 
90 
82 
50 
59 
73 



69. Rhinogobius taiwanus sp. nov. (Plate LI II, Fig. i.) 

Head 3.35 in length; depth 5.6; depth of caudal peduncle 2.25 in 
head; eye 5; interorbital space 7; snout 2.38; D. VI, 9; A. 9; P. 17; 
V. I, 5; thirty-six scales in a lateral series, twelve scales in a transverse 
series between origin of second dorsal and anal; gill-rakers 2 + 7. 

Head and body of nearly the same depth throughout, caudal peduncle 
slightly low and compressed; head long and depressed, its width 
contained 1.64 in its length, cheeks slightly bulged out; snout rather 
long, sharply pointed anteriorly; mouth terminal, slightly oblique, its 
angle not reaching to a vertical through the anterior margin of orbit; 

^^ Clenogobius hadropterus Jordan & Snyder, Proc. U. S. Nat. Mus., 1901, XXIV, 
p. 60; Nagasaki; Kurume; Tsuruga; Kawatana. 

Rhinogobius hadropterus, Jordan, Snyder, & Tanalca, Journ. Coll. Sci. Tokyo, 
XXXIII, 1913. p. 343- 



ANNALS CARNEGIE MUSEUM, Vol. XI 



Plate LI 1 1 






Rhinogobius laiwaniis Oshima, sp. nov. 
Rhinogobius formosaniis Oshima, sp. nov. 
Glossogobius parvus Oshima, sp. nov. 



The Fresh Water Fishes of the Island of Formosa. 299 



upper jaw more or less projecting; lips fleshy; teeth on both jaws in 
villiform bands, those on outer row enlarged; eyes superior, situated 
midway between tip of snout and posterior angle of opercles; inter- 
orbital space concave and narrow; anterior nostril in a short tube. 

Dorsal fins well separated; spinous dorsal greatly elevated, second 
spine the longest, filamentous, its height 3.8 in total length without 
caudal, the third a little shorter than the second, the last spine about 
one-third as long as the second; soft dorsal high, posterior ray the 
longest, when depressed scarcely reaching the root of caudal; pectorals 
extending beyond the posterior end of base of spinous dorsal; ventrals 
united, forming a round, concave cup, the tip of which reaches middle 
of pectoral, free from the belly; anal fin inserted below second ray of 
soft dorsal, posterior ray the longest, when depressed reaching end of 
soft dorsal; caudal fin rounded. 

Body covered with large ctenoid scales; small cycloid scales near the 
base of spinous dorsal and oh the belh'; head and occiput entirely 
naked. 

Head with a number of mucous pores nearly as large as the nostril, 
one on each side above and before the eyes, one on the posterior part of 

Measurements of Rhinogobius taiwanus. 



Locality. 





0— ■ 

CJ3 


2.25 


7 


2.33 


7 


2.42 
2.63 


7 
7 


2.42 


6 


2.10 


6 


2.14 


6 


2.40 


6 


2.14 


6.5 


2-57 
2.66 


5 

7 






Shinchiku 

Shinchiku 

Shinchiku 

Shinten 

Sobun River. . 
Jitsugetsutan. . 
Dakusui River 
Dakusui River 
Dainansho. . . . 
Bokusekikaku. 
Inzanpo 



S-40 

5-44 

6.25 

6.27 

6 

5-23 

6 

6.09 

5.20 

5-80 

5-88 



VI, 9 
VI, 9 
VI, 9 
VI, 9 
VI, 9 
VI, 9 
VI, 10 
VI. 10 
VI, 10 
VI, 9 
VI, 9 



2.38 
2.66 
2.42 

2.33 
2.66 
2.25 
2.42 
2.50 
2.50 
2.50 
2.33 



5 36- 
S [36- 
4-5035- 
4-6036- 

5 35- 



5 
5 
5 
5 
5 
4-50 



interorbital space, one behind each eye, two along the posterior edge 
of pre-operculum. 

Color in alcohol pale brown, somewhat darker anteriorly; all the 
fins pale brown, soft dorsal and anal bordered with white. 

Length of body 69 mm. 

Described from a specimen from Shinchiku, collected by T. Aoki 
in December, 1916. 



300 Annals of the Carnegie Museum. 

Habitat: Shinchiku; Taniusui River near Shinten; Dakusui River; 
Sobun River; Jitsugetsutan (Lake Candidius) ; Daiiiansho, Nanto; 
Bokusekikaku; Inzanpo, Giran. 

Remarks: This species is near Rhiiwgobiiis similis'^^ from Japan, but 
differs distinctly in the smaller number of mucous pores and in the 
greater number of scales in the lateral series. 

Rhinogohius candidius is another closely related species. It differs, 
however, in having no mucous pores and the spines of the first dorsal 
which are not filamentous. 

70. Rhinogobius formosanus sp. nov. (Plate LIII, Fig. 2). 

Head 3.53 in length; depth 4.25; depth of caudal peduncle 2 in head; 
eye 5; interorbital space 5.5; snout 2.8; D. VI, 9; A. 9; P. 19; twenty- 
nine scales in lateral series, nine scales in an oblique series between 
origins of second dorsal and anal. 

Body rather robust, highest in front of first dorsal; tail slightly 
compressed; head elongate, entirely naked; snout pointed anteriorly; 
mouth moderate, more or less oblique, its angle reaching a vertical 
through anterior margin of orbit; jaws subequal; lips fleshy, upper 
lip rather wide; tongue broad, with rounded tip; teeth on the upper 
jaw in a single series, conical and minute, those on the lower jaw in a 
narrow villiform band, outer ones pointed and in a single row; eyes 
high in head, upper margin projecting above the contour of head, 
situated nearer to tip of snout than the posterior edge of opercle; 
interorbital space convex; nostrils separated, anterior nostril in a short 
tube. 

Dorsal fins well separated, none of the spines elongated, when 
depressed not reaching the base of second dorsal; second dorsal rather 
short, middle ray longest; pectorals rhomboidal, pointed at the tip, 
not reaching the vent; ventrals united, forming a concave disk, not 
adnate to the belly; anal fin inserted below the third ray of second dor- 
sal, when depressed extending to end of second dorsal, posterior ray 
longest; caudal fin rounded. 

Body covered with large ctenoid scales, those on the belly smaller 
and cycloid; cheeks, operculum, and tip of head entirely naked. 

Color in alcohol pale olive-gray, back mottled with dark; sides with 
about seven squarish dark cross-bars; lower surface whitish; top of 

^* Rhinogobius similis Gill, Proc. Ac. Nat. Sci., Philad., 1859, p. 145; near Shi- 
moda, Japan. 



The Fresh Water Fishes of the Island of Formosa. 301 

head, cheeks, and snout with a number of wavy, ])ale-brown, longi- 
tudinal streaks; membrane of the dorsal fins grayish, spines fuscous, 
soft rays spotted with dark; ventral fin whitish; pectoral and anal 
dusky, the latter bordered with white; caudal fin with wavy vertical 
cross-bands. 

Length of body 65 mm. 

Described from a specimen from Shinchiku, collected by T. Aoki 
in December, 1916. 

Habitat: Shinchiku (a single specimen). 

Remarks: The present species much resemble Rhinogobiiis liadrop- 
tcnis^'" from Japan. It differs, however, in having an entirely naked 
head. 

71. Rhinogobius caninus (Cuv. & Val.). 

1837. Gobiiis caninus Cuv. & V.al., Hist. Nat. Poiss., XII, p. 86. — Bleeker, 
Verh. Batav. Gen., XXII, 1849, p. 27. — Gltnther, Cat. Fish., Ill, 1861, 
p. 38; China; East Indian Archipelago. 

1905. Rhinogobius caninus Jordan & Seale, Proc. U. S. Nat. Mus., XXVIII, 
p. 796; Negros. — Jordan & Richardson, Bull. U. S. Bur. Fish., XXVII, 
1908, p. 259; Iloilo; Lubang. — ^Jordan & Richardson, Mem. Carneg. 
Mus., IV, no. 4, 1909, p. 200; Takao, Formosa. 

1912. Ctenogohius caninus Snyder, Proc. U. S. Nat. Mus., LXII, p. 514; Okinawa. 

Head 3.45 in length; depth 5.42; caudal peduncle 2.44 in head; eye 
4.8; interorbital space 8; snout 3.28; D. VI, i, 9; h. 9; P. 19; twenty- 
eight scales in a lateral series, ten scales between the origin of soft 
dorsal and that of the anal. 

Body oblong, compressed, depth nearly subequal throughout the 
body, gradually tapering posteriorly; head rather high, not depressed; 
snout rounded anteriorly, its profile slightly arched, interorbital 
space narrow; mouth oblique, its angle reaching beyond a vertical 
through anterior margin of orbit; lower jaw longer than the upper, 
with two large canine teeth on each side; teeth on both jaws in a villi- 
form band, outer ones larger; tongue rather short, truncated in front; 
eyes high in head, much nearer tip of snout than the end of opercle; 
nostrils separated, in front of e^-e, the posterior slightly larger than 
the anterior. 

Dorsal fins well separated, none of the spines elongated, when 

^^ Ctenogohius hadroplerus Jordan & Snyder, Proc. U. S. Nat. Mus., XXIV, 1901, 
p. 60. 

Rhinogobius hadroplerus Jordan, Snyder, & Tanaka, Journ. Coll. Sci. Tokyo. 
XXXIII, 1913. p. 343. 



302 Annals of the Carnegie Museum. 

depressed scarcely reaching the origin of soft dorsal; soft dorsal rather 
long, armed with a short, slender undivided ray, posterior ray the 
longest, when depressed reaching root of caudal; pectorals quadrate, 
reaching a vertical through the vent; ventrals united, forming a shallow 
concave disc, not adnate to the belly; anal fin inserted below the third 
dorsal ray, posterior ray the longest, when depressed reaching the end 
of soft dorsal; caudal fin obtusely rounded. 

Body covered with large ctenoid scales, those on the belly and occi- 
put small and cycloid; cheek naked; occiput covered with minute 
scales. 

Color in alcohol pale yellowish brown, sides with large irregular 
brown spots disposed in two longitudinal series; a dark round spot 
above the base of pectorals; head with no markings; spinous dorsal 
and pectorals dusky; membrane of the soft dorsal with longitudinal 
series of dark spots; outer margin of the anal black; caudal fin uni- 
formly dusky. 

Length of body loo mm. 

Described from a specimen from Takao, collected by Hans Sauter 
(Stanford Collections, No. 20995). 

Habitat: Takao. 

Genus Glossogobius Gill. 

Body rather elongate, tail slightly compressed; head broad and 
depressed anteriorly, naked; interorbital space flat. Mouth very 
large, terminal and oblique; teeth moderate, in broad bands, the inner 
teeth depressible; tongue emarginate at tip. Sides of head naked; 
no barbels; eyes well separated; isthmus very narrow; the gill-openings 
ending forward below; pseudobranchiae well-developed; no fleshy flaps 
on shoulder-girdle. Dorsal fins both short, the first composed of six 
slender spines; pectorals without silk-like rays above; ventrals united, 
not adnate to the belly; anal with nine soft rays; caudal free from the 
dorsal and anal. Body covered with rather large scales, which number 
about forty in the lateral series. 

Distribution: Formosa; Philippine Islands; China; Japan; Malay 
Archipelago; Cejlon. 

vSynopsis of the Formosan Species. 

a. Dorsal fin \'I, 10; anal composed of 9 rays; scales 31-32 in a lateral series, 

9-10 in an oblique series between origins of the soft dorsal and anal . brunneus. 

aa. Dorsal fin VI, 11; anal composed of 11 rays; scales 54 in a lateral series, 13-14 

in an oblique series between origins of the soft dorsal and anal . gramme pontus. 



The Fresh \\'ater Fishes. of the Island of Formosa. 303 

aaa. Dorsal fin \'I, 9; anal composed of 9 rays; scales 41 in a lateral series, 13 in an 

oblique series between origins of the soft dorsal and anal parvus. 

aaaa. Dorsal fin VI, 10; anal composed of 10 rays; scales 28-30 in a lateral series. 

abacopus. 

72. Glossogobius brunneus (Schlcgel). 
Urohaze (Japan); Kaugam (Giran, Formosa). 

1847. Gobius brunneus Schlegel, Fauna Japonica, Poiss., p. 142, PI. LXIV, 
Fig. 2; Nagasaki. — Gunther, Cat. Fish., Ill, i86i, p. 65; after S(*hlegel. — 
ISHiKAWA, Cat. Fish., 1897, p. 39; Tokyo; Boshu. 

1847. Gobius olivaceus Schlegel, Fauna Japonica, Poiss., p. 143, PI. LXXIV, 
Fig. 3; Nagasaki. 

1901. Glossogobius brunneus Jordan & Snyder, Proc. U. S. Nat. Mus., XXV, p. 74; 
Wakanoura; Onomichi; Hakodate; Kurume; Nagasaki. — Jordan & Ever- 
MANN, Proc. U. S. Nat. Mus., XXV, 1903, p. 361; Kotosho; Keelung, 
Formosa. — Jordan & Richardson, Mem. Carneg. Mus., IV, no. 4, 1909, 
p. 200; after Jordan & Evermann. — Jordan, Snyder, & Tanaka, Journ. 
Coll. Sci. Tokyo, XXXIII, 1913, p. 350; Hakodate to Nagasaki. 

1846. Gobius platycephalus Richardson, Ichthyol. China, p. 204; Macao. 

1846. Gobius fasciato-punctalus Rxchardson, Ichthyol. China, p. 204; Canton. 

1897. Gobius giurus Rutter, Proc. Acad. Sci. Philad., Jan., p. 85; Swatow. 

Head 3.16 in length; depth 4.88; depth of caudal peduncle 2.84; 
D. VI, 10; A. 9; P. 20; V. I, 5; eye 5.71 in head, interorbital space 
6.83; snout 3; maxillary 2.44; thirty-one scales in a lateral series, ten 
scales in a transverse series; gill-rakers 3 -|- 10. 

Body robust, elongate, posterior part slightly compressed, highest 
near the insertion of spinous dorsal; head very large, depressed, broader 
than body, with a deep longitudinal groove on the top; snout more 
or less pointed, interorbital space rather flat; eyes supralateral, 
diameter nearly as long as the interorbital space; mouth large, ter- 
minal and oblique, its angle extending to a vertical through anterior 
third of the orbit; lower jaw protruding beyond the upper; lips broad; 
teeth simple, in two series, inner one depressible and somewhat larger; 
tongue broad, deeply notched at the tip; gill-openings lateral, running 
far forward below; width of isthmus nearly equal to the interorbital 
space; gill-rakers on the first arch short, reduced to mere elevations 
near the end of arch; nostrils separated, the anterior in a short tube; 
chin smooth. 

Body covered with large ctenoid scales; head naked; no lateral line. 

Dorsal fins separated; anterior dorsal inserted behind the base of 
the pectoral, second spine the longest, when depressed, reaching be- 
yond the origin of second dorsal; fin-rays of second dorsal subequal in 



304 



Annals of the Carnegie Museum. 



height, a little shorter than first dorsal; pectoral fins lateral, large, 
rounded; ventrals united, forming an oval disc, not adnate to belly; 
anal inserted below third dorsal ray, posterior rays longer, reaching 
as far backward as do those of the dorsal, both not reaching the root of 
caudal; caudal fin rounded. 

Color in formalin dark gray above, lower parts white; upper parts of 
sides mottled with black; dorsal fins grayish, with small dusky spots 
in more or less definite longitudinal series; pectorals and caudal with 
vertical rows of small dark spots; ventral fins whitish; the anal dusky. 

Length of body 145 mm. 

Described from a specimen from Taihoku, collected by T. Aoki 
in September, 1916. 

Habitat: Taihoku; Tamusui River; Ritakukan, Giran. 

Measurements of Glossogobius brunnetis. 



Locality. 



Taihoku . . 
Taihoku. . . 
Taihoku. . . 
Taihoku. . . 
Ritakukan 
Ritakukan 
Ritakukan 



-a 


0. 



-CXI c 


3.16 


4.88 


2.85 


3-07 


5-34 


3 


3-21 


5-89 


2.91 


3 


5-iS 


3 


3-18 


5-15 


2.73 


3-25 


5-6o 


2.87 


3-42 


5-76 


2-73 



VI, 10 
VI, 10 
VI, 10 
VI, 10 
VI, 10 
VI, 10 
VI, 10 



20 1,5 
201,5 
2o|i,5 
20 1,5 
20 I, 5 
2o'i, 5 
20 I, 5 



6.83 

7 

6.60 

7 

6.25 
6 
6.66 



3 j 5-71 31-10 
3-25t 6.33!32-io 



3-09 

3 

3 

3 

2.93 



5.6032-10 
5.66 32- 9 
6.33 32- 9 
6.18I32- 9 
6.33 32- 9 






145 
150 
132 
127 
173 
180 
177 



73. Glossogobius grammepomus (Bleeker). 

1849. Gobius grammepomus Bleeker, Verh. Batav. Gen., XXII, p. 34. — Gunther, 

Cat. Fish., Ill, 1861, p. 64; Malay Archipelago. 
1849. Gobius melanocephalus Bleeker, Verh. Batav. Gen., XXII, p. 34. — ■ 

Naturk. Tydschr. Nederl. Ind., 1851, I, Fig. 4.— D.\y, Fish, Brit. India, 

p. 292, PI. LXIII, Fig. 6; Seas of India to Malay Archipelago. 
1861. Gobius lilluratus Steind.^chner, Sitzsb. VVien, Acad., XLII, p. 289, Figs. 

4, 5; Philippine Islands. 

Head 3.17 in length; depth 4.5; D. VI, 1 1 ; A. 1 1 ; P. 17; V. I, 5: 
width of head 1.42; eye 5.5 in head; interorbital space 4.42; snout 
2.39; maxillary 2; depth of caudal peduncle 2.58; fifty four scales in a 
lateral series, fourteen scales between origin of sscond dorsal and 
that of the anal; gill-rakers 2+4. 

Body thick, dorsal profile arcuate, ventral profile nearly straight, 
deepest in front of the spinous dorsal, tail compressed; head very 
large, broader than the body; cheek more or less bulged out; snout 



The Fresh Water Fishes of the Island of Formosa. 305 



rather long, broadly rounded anleriorly; eyes moderate, superior, but 
directed laterally; interorbital space broad, more or less concave; 
mouth slightly oblique, with very thick lips, its angle reaching a 
vertical through center of the orbit; upper jaw a little longer than the 
lower; teeth in a broad villiform band, no canine teeth, those in outer 
row somewhat larger; tongue broad, deeply notched in front; gill- 
openings run forward below; inner edge of shoulder-girdle without 
papillae; gill-rakers very short; nostrils separated, the anterior in a 
short tube. 

Head naked, except the occiput; body covered with large ctenoid 
scales, those on the nape and belly small and cycloid. 

Dorsal fins separated; spines not filamentous, third spine longest, 
contained 1.82 in head, soft rays a little shorter than the spines, 
anterior longest; pectorals large, reaching the posterior end of the 
base of spinous dorsal; ventrals united, free from belly; anal flap well 
developed; anal fin inserted below second dorsal ray, when depressed 
reaching to the root of caudal; caudal fin rounded. 

Color in formalin dark brown above, paler below, sides and upper 
part of the body with a number of vermiculated black spots; mem- 
branes of the dorsal fins dusky, spines and rays with a series of dark 
spots; caudal fin with a number of vertical cross-bars; the rest of the 
fin grayish. 

Length of the body 118 mm. 

Described from a specimen from Inzampo, Giran, collected by 
T. Aoki in August, 191 7. 

Habitat: Inzampo, Giran (two specimens). 

Measurements of Glossogobius grainmepomus. 



Locality. 


•0 


.5" 
p. 

u 

Q 


Q 


< 


D-' 


>■ 





u . 

c2 


3 

a 
in 




C/5 


CUD C 


Inzanpo 

Inzanpo 


3-17 
3-39 


4-50 
4-75 


VI, II 
VI, II 


II 
II 


17 
17 


1.5 
1.5 


1.42 
1.44 


4.42 

5 


2.39 
2.40 


5-5 
5-5 


54-14 
54-13 


118 

97 



. 74. -Glossogobius parvus sp. nov. (Plate LI II, Fig. 3). 
Head 3.18 in length; depth 4; D. VI, 9; A. 9; P. 16; V. I, 5; width 
of head 1.5 in its length; eye 4 in head; snout 3; interorbital space 
2.5; depth of caudal peduncle 2.25; maxillary 2.2; scales about forty- 
one in a lateral series, thirteen scales between origins of the soft 
dorsal and anal; pectoral 1.5 in head; ventral 2. 
21 — JAN. 12, 1919. 



306 Annals of the Carnegie Museum. 

Body rather slender, laterally compressed; head more or less 
depressed, broader than body, cheek fleshy, somewhat bulged out 
laterally; snout flat, broadly rounded anteriorly, interorbital space 
rather flat; eyes superior, directed laterally; mouth oblique, its angle 
reaching a vertical through centre of orbit; lips fleshy; upper jaw 
slightly longer than the lower; teeth in villiform bands, no caniniform 
teeth; tongue quadrate, emarginate at the tip; nostrils widely separated, 
the anterior in a short tube, just behind the upper lip; gill-openings 
extending downward and a little forward; gill-rakers very short. 

Dorsal fin separated; spinous dorsal rather small, triangular, the 
anterior spine longest, no filamentous spines, when depressed not 
reaching the origin of soft dorsal; soft dorsal rather low, the posterior 
ray longest, its height contained twice in the length of head; pectorals 
rounded, reaching beyond the middle of base of spinous dorsal; 
ventrals united, not adnate to the belly, cup round and deep; anal 
flap well developed, elongate; the anal inserted below the second ray 
of soft dorsal, as high as the soft dorsal; caudal fin rounded. 

Body covered with thin minute ctenoid scales; head, except occiput, 
naked; scales on the occiput and belly minute and cycloid; base of the 
pectoral fleshy and scaly. 

Color in formalin brownish gray, sides with a number of irregular 
dark cross-bars; cheek with two longitudinal brown stripes originating 
at the orbit; dorsal fins brownish gray; tip of the spinous dorsal black, 
interspace between each ray with a black streak; caudal fin dusky, 
with three dark cross-bars; the rest of the fins uniformly dusky. 

Length of body 44 mm. 

Described from a specimen from Kizanto, Giran, collected by T. 
Aoki in August, 191 7. 

Habitat: Kizanto, Giran (a single specimen). 

Remarks: Kizanto is a small island near Giran; the present species 
is probably a marine fish. 

75. Glossogobius abacopus Jordan & Richardson. 

1909. Glossogobius abacopus Jordan &; Richardson, Mem. Carneg. Mus., IV, 
no. 4, p. 200; Takao, Formosa. 

Head to tip of lower jaw three times in length; depth 5.5; eye 4 in 
head; dorsal VI, 10; anal 10; scales 28-30; snout 3.4 in head; maxillary 
2.4; interorbital space slightly greater than width of juipil. Body 
elongate, rather depressed in front, tapering gradualh' backward to 



The Fresh \\'ater Fishes of the Island of Formosa. 307 

the depressed caudal peduncle; depth of caudal peduncle 3.4 in head; 
head pointed; lower jaw projecting width of pupil; mouth large, 
maxillary reaching vertical from back of pupil; jaws, vomers and 
palatines with rows of fine, sharp-pointed teeth; tongue deeply notched; 
origin of spinous dorsal nearly an eye-length behind insertion of 
ventrals, its base 2.4 in head; base of soft dorsal 1.5 in head; longest 
dorsal spine 2.25; longest ray twice in head; origin of anal under 
second ray of soft dorsal, equidistant between base of caudal and 
back of eye; pectoral 1.3 in head; ventrals 1.4; depth of membranous 
cup of united ventrals two-thirds of length of eye; caudal 1.3 in head, 
rounded; a large anal papilla with notch behind. 

Color in spirits brownish-olive, back and caudal peduncle crossed 
obliquely by four broad saddle-like bands of dark color; membranes 
of dorsals and of anal chiefly blackish, with some small spot-like 
intervals of paler on these and on rays; caudal and ventrals barred 
or checkered with dark; pectorals lightly speckled with dusky, with 
darker and denser specks below at base; under parts unevenly punct- 
ulated, the dots forming indistinct bars on chin and lower jaws; tip of 
lower jaw blackish. 

This species is near Glossogobius vaisiganis from Samoa, but differs 
in the details of coloration, notably in the sharply checkered ventral 
fin. (Jordan & Richardson). 

Habitat: Takao. (Jordan & Richardson). Not seen. 

Genus Acanthogobius Gill. 
1859. Acanthogobius Gill, Proc. Ac. Nat. Sci. Philad., p. 145. (Type Gobius 

flavimanus Temminck & Schlegel.) 
1863. Synechogobius Gill, Proc. Ac. Nat. Sci. Philad., p. 266. (Type Gobius 

hasta Temminck & Schlegel.) 

Body oblong, little compressed, covered with medium-sized roughish 
scales; cheeks with small scales; snout rather long, the head rounded 
in profile; mouth moderate, oblique, the jaws about equal, the teeth 
moderate; tongue truncate or very slightly notched; isthmus rather 
broad, the gill-openings slightly continued forward below; no flaps on 
shoulder-girdle. Dorsal fins rather long, the first composed of seven 
to nine slender spines, the second of fourteen or fifteen soft rays; anal 
of twelve to thirteen rays. (Jordan & Snyder). 

76. Acanthogobius ommaturus (Richardson). 
1846. Gobius ommaturus Rich.^rdson, Voy. Sulph. Fish., p. 146, PI. LV, Figs. 1-4; 
Woosung, Yang-tze-kiang. — Gunther, Cat. Fish., Ill, 1861, p. 77; Amoy. — • 
RuTTER, Proc. Nat. Sc. Philad., Jan., 1897, p. 85; Swatow. 



308 Annals of the Carnegie Museum. 

1905. Acanlhogobius ojnmatiiriis Jordan & Seale, Proc. U. S. Nat. Mus., XXIV, 
p. 528; Shanghai. 

Head 3.56 in length; depth 7.5; D. IX, 18; A. 15; P. 21; V. i, 4; 
width of head 1.81; eye 5.33 in head; interorbital space 7.5; snout 
2.75; depth of caudal peduncle 3.75; seventy scales in a longitudinal 
series, eighteen scales between origins of second dorsal and anal; 
gill-rakers 3 + 8. 

Body elongate, slender, anterior part somewhat cylindrical, tail 
compressed; head large, slightly depressed, its top scaly, cheek 
covered with minute scales; snout somewhat produced, acutely 
rounded anteriorly, tip slightly swollen; mouth large, inferior, its angle 
not reaching a vertical through anterior border of orbit; lips thick and 
fleshy; upper jaw slightly longer than the lower; teeth conical and 
fixed, in several rows, those of the outer series somewhat larger; eyes 
high up, interorbital space concave; nostrils separated; tongue broad, 
truncated in front; gill-openings not extending far forward; isthmus 
rather broad. 

Dorsal fin separated; spinous dorsal slender, anterior spine longest, 
when depressed not reaching second dorsal; soft dorsal elongate, 
with many rays, length of each ray subequal; pectorals rather large, 
without free silk-like rays above, their bases smooth and muscular; 
ventrals completely united, forming a concave round disk which is not 
adnate to belly; anal inserted below fourth dorsal ray, when depressed 
reaching posteriorly as far as the dorsal, both not extending to the 
root of caudal; caudal fin rhomboidal, pointed at middle; caudal 
peduncle elongate. 

Body covered with thin ctenoid scales; scales on head small and 
cycloid. 

Color in alcohol pale gray above, lower half of the sides and belly 
whitish; rays of the dorsal fins spotted with black, caudal fin yellowish 
olive, with a number of dark markings near the base; the rest of the 
fins whitish. 

Length of body 140 mm. 

Described from a specimen from Taihoku, collected by T. Aoki. 

Habitat: Taihoku (two specimens). 

Summary. 
Family SALMON I D.-E. 
I. Plecoglossus Temminck & Schlegel. 
I. altivelis Temminck & Schlegel; Tamusui River. 



The Fresh Water Fishes of the Island of Formosa. 309 

Family SALANGID/E. 
2. Parasalanx Regan. 

2. acuticeps (Regan); not seen. 

3. ariakensis (Kishinouye) ; Tamusui River. 

Family SILURID.^. 
3. Parasilurus Bleeker. 

4. asotiis (Linnanis) ; Jitsugetsutan; Tamusui River; Inzampo; 

Ritakukan. 

4. PsEUDOBAGRUS Bleeker. 

5. hrevianalis (Regan); Jitsugetsutan; Dainansho. 

6. taiiconcnsis Oshima; Tozen River; Daito River; Shinchiku. 

7. adiposalis Oshima; Tamusui River; Daito River; Sobun River. 

5. LiOBAGRUS Hilgendorf. 

8. nanto'cnsis Oshima; Dainansho. 

9. formosaniis Regan; not seen. 

6. Clarias Gronovius. 

10. fiisciis (Lacepede) ; Jitsugetsutan; Taihoku; Tamusui River; 

Maruyama, Giran. 

Family COBITID^. 
7. MiSGURNUS Lacepede. 

11. anguillicaudatus (Cantor); Tamusui River; Maruyama; Giran; 

Rato; Raupi; Jitsugetsutan. 

12. decemcirrosus (Basilewsky) ; Taihoku; Taichu. 

8. CoBiTis Linnaeus. 

13. tccnia Linnaeus; Shinchiku; Jitsugetsutan; Rigyokutsu; Maruyama, 

Giran. 

Family HOMALOPTERID.E. 

9. FoRMOSANiA Oshima. 

14. gilbert i Oshima; Tamusui River. 

10. Hemimyzon Regan. 

15. formosaniis (Boulenger); Taiko River. 

Family CYPRINID^. 
I. Carassius Nilsson. 

16. auratus (Linnaeus); Taihoku; Giran. 



310 Annals of the Carnegie Museum. 

12. Cyprinus (Artedi) Linneeus. 

17. carpio Linnaeus; Taihoku; Temsonpi; Inzampo; Maruyama. 

13. Labeo Cuvier. 

18. jordani Oshima; Shori. 

14. AcRossocHEiLUS Oshima. 

19. formosanns (Regan) Tamusui River; Tozen Rivor; Shinchiku; 

Horisha; Jitsugetsutan. 

15. ScAPHESTHES Oshima. 

20. tamusuicnsis Oshima; Tamusui River; Choso River; Giran River. 

16. Hemibarbus Bleeker. 

21. lubeo (Pallas); Tamusui River; Rigyokutsu. 

17. Barbodes Bleeker. 

22. paradoxus (Giinther); Taiko River. 

18. Capoeta Cuvier & Valenciennes. 

23. semifasciolata (Giinther); Ako. 

19. PuNTius Hamilton. 

24. snyderi Oshima; Rigyokutsu; Maruyama; Daito River, 

20. Spinibavhiis Oshima. 

25. hoUandi Oshima; Sobun River. 

21. Gnathopogon Bleeker. 

26. iijimcc Oshima; Tozon River.. 

22. PsEUDOGOBio Bleeker. 

27. hrevirostris Giinther; Tamusui River. 

23. Pseudorasbora Bleeker. 

28. parva (Schlegel) ; Tamusui River; Taihoku; Raupi; Tozon River; 

Nanto; Rigyokutsu; Shinchiku; Ako; Shori; Bokusekikaku. 

24. Pararasbora Regan. 

29. mollrcchti Regan; Jitsugetsutan. 

25. Phoxiscus Oshima. 

30. kikuchii Oshima; Bokusekikaku. 



The Fresh Water Fishes of the Island of Formosa. 311 

26. Distgechodon Peters. 

31. tnmirostris Peters; Taihasho, Giran. 

27. Ischikauia Jordan & Snyder, 

32. niacrolcpis Regan; not seen. 

28. Ctenopharyngodon Steindachner. 

33. ideUns (Cuvier & Valenciennes); Shori. 

29. Acheilognathus Bleeker. 

34. himantegus Giinther; Taihoku; Wodensho; Taichu; Shimo- 

Tamusui River. 

30. Rhodeus Agassiz. 

35. oceUatus (Kner); Tailioku; Nanto. 

31. Zacco Jordan & Snyder. 

36. platypus (Schlegel); Tamusui River; Shinchiku; Choso River. 

37. temmincki (Schlegel); Tamusui River; Daiko River; Daito River; 

Shinchiku; Dakusui River; Rigyokutsu; Sobun River; Shimo- 
Tamusui River; Ako; Heirinbi; Inzampo; Tensonpi; Suwo; 
Giran. 

38. pachycephahis Giinther; Tamusui River. 

32. Metzia Jordan & Thompson. 

39. mesembrina (Jordan & Evermann) ; Kotosho. 

33. Candidia Jordan & Richardson. 

40. harhata (Regan); Jitsugetsutan; Shito, Giran. 

34. Hypophthalmichthys Bleeker. 

41. molitrix (Cuvier & Valienciennes) ; Shori. 

35. Aristichthys Oshima. 

42. nobilis (Gray) (Richardson); Shori. 

36. Chanodichthys Bleeker. 

43. macro ps Giinther; Tamusui River. 

37. Culter Basilewsky. 

44. aokii Oshima; Jitsugetsutan. 

45. brevicauda Giinther; Kagi. 



312 Annals of the Carnegie Museum. 

38. CuLTRicuLUS Oshima. 

46. kneri (Kreyenberg) ; Jitsugetsutan; Shimo-Tamusui River. 

Family PCECILIID.'E. 
39. Oryzias Jordan & Snyder. 

47. latipes (Temniinck & Schlegel) ; Shori; Ako; Giran. 

40. Gambusia Poey. 

48. affinis (Baird & Girard) ; Shori. 

Family MONOPTERID.F:. 
41. Fluta Bloch & Schneider. 

49. alha (Zuiew); Jitsugetsutan; Shokwa; Taihoku; Kiburan, Giran. 

Family ANGUILLID.E. 

42. Anguilla Shaw. 

50. mauritiana Bennett; Giran; Jitsugetsutan. 

51. japonica Temminck & Schlegel; Taihoku; Giran. 

52. sinensis McClelland; not seen. 

Family MUGILID.F:. 
43. MuGiL (Artedi) Linnccus. 

53. ceplialus Linna?us; Taihoku. 

54. oeiir Forskal; Inzampo; Ritakukan. 

55. carinatus (Ehrenberg) Cuvier & Valenciennes; Taihoku; Shimo- 

Tamusui River. 

44. Liza Jordan & Swain. 

56. troscheli (Bleeker); Sobun River. 

Family LABYRINTHICI. 
45. PoLYACANTHUS (Kuhl) Cuvier. 

57. opcrcidztns (Linnaeus); Taihoku; Wodensho. 

46. Macropodus Lacepede. 

58. fikunentosiis Oshima; Kotosho. 

Family KUHLIID.F:. 
47. KUHLIA Gill. 

59. marginata (Cuvier & Valienciennes) ; Tamusui River; Chose 

River; Giran Ri\er; Bokusekikaku. 



The Fresh Water Fishes of the Island of Formosa. 313 

Family OPHlCEPHALID.^i. 
48. Ophicephalus Bloch. 

60. tadianus Jordan & Evermann; Taihoku; Raupi; Tozen River; 

Nanshisho. 

61. maculatiis Lacepede; Wodensho. 

49. Channa Gronow. 

62. formosana Jordan & Evermann; Taihoku; Toyen; Tamusui 

River; Jitsugetsutan. 

Family GOBIID.^. 
50. Eleotris (Gronow) Schneider. 

63. oxycephala (Schlegel) ; Tamusui River; Tozen River; Giran River; 

Buroko River. 

64. fusca (Schneider) ; Buroko River. 

51. BuTis Bleeker. 

65. biitis (Buchanan-Hamilton) ; Taihoku. 

52. Sicyopterus Gill. 

66. japoniciis (Tanaka) ; Tamusui River; Tozen River; Koanronsho; 

Shinchiku; Bokusekikaku; Choso River; Raoko. 

53. Rhinogobius Gill. 

67. candidiiis (Regan); Bokusekikaku; Heirinbi; Tozen River; 

Shinchiku; Daiko River; Daito River. 

68. giiirinus (Rutter); Shimo-Tamusui River; Tozen River: Woden- 

sho; Daito River; Tamusui River; Shori; Hyoko; Maruyama, 
Giran. 

69. tahcaniis Oshima; Tamusui River; Dakusui River; Sobun River; 

Shinchiku; Jitsugetsutan; Dainansho; Bokusekikaku; Inzampo. 

70. formosanns Oshima; Shinchiku. 

71. caniniis (Cuvier & Valenciennes); Takao (not seen). 

54. Glossogobius Gill. 

72. brunneiis (Schlegel); Taichu; Tamusui River; Ritakukan. 

73. granimepomus (Bleeker); Inzampo. 

74. parvus Oshima; Kizanto. 

75. abacopus Jordan & Richardson; not seen. 

55. Acanthogobius Gill. 

76. ommaturus (Richardson) ; Taihoku. 



314 



Annals of the Carnegie Museum. 



Of the seventy-six species above enumerated the following twenty- 
nine are artificially introduced species or semi-marine fishes or species 
which have been collected in an outlying island, and therefore have no 
bearing on the problems of the geographical distribution of the 
Formosan fresh -water fishes. 



Introduced Species. 

4. Aristichthys nobilis. 



1. Labeo jordani. 

2. Ctenopharyngodon idellus. 

3. Hypophthalm ichthys moUtvix. 



5. Gambiisia affinis. 



Semi-marine Species. 
6. PJecoglossiis altivelis. 17. KuJilia marginata. 



7. Parasalanx aciiticeps. 

8. Parasalanx ariakensis. 

9. Fliita alba. 

10. Angiiilla mauritiana. 

11. An guilla japon ica . 

12. A 71 guilla sinensis. 

13. Mugil ccpliahis. 

14. Mugil oeur. 

15. Mugil carinatus. 

16. Liza troscheli. 



18. Eleotris oxycephala. 

19. Eleotris fusca . 

20. Butis butis. 

21. Rliinogobius giurinus. 

22. Rliinogobius caninus. 

23. Glossogobius brnnneus. 

24. Glossogobius abacopus. 

25. Glossogobius parvus. 

26. Glossogobius grammepomus. 

27. Acantlwgobius ommaturus. 



The Species from Botel-Tobago. 
28. Metzia mesembrina. 29. Macropodus filamentosus. 

Of the the remaining forty-seven species twenty six (55%) are 
peculiar to the island. 

Peculiar Species. 

1. Pseudobagrus brevianalis. 11. Spinibarbus hollandi. 

2. Pseudobagrus taiwanensis. 12. Acrossocheilus formosanus. 



3. Pseudobagrus adiposalis. 

4. Liobagrus nantoensis. 

5. Liobagrus formosanus. 

6. Formosania gilberti. 

7. Ilemimyzon formosanus. 

8. Scaphesthes tamusuiensis . 

9. Barbodes paradoxus. 
10. Puntius snvderi. 



13. Gnathopogon iijimce. 

14. Pseudogobio brevirostris. 

15. Pararasbora moltrechti. 

16. Phoxiscus kikuchii. 

17. Iscliikauia m,acrolepis. 

18. Acheilognathus himantegus. 

19. Zacco pachycephalus. 

20. Candidia barbata. 



The Fresh Water Fishes of the Island of Formosa. 315 

21. Chanodichthys tnacrops. 24. Rhinogobius candidius. 

22. Culler aokii. 25. Rhinogobius taiwanus. 

23. Channa formosana. 26. Rhinogobius forniosaniis. 

At present the percentage of peculiar species is extraordinarily 
high. But it is quite possible that some of them may be found in 
adjacent regions, probably in Southern China, and sooner or later a 
slight reduction may have to made. 

Of the above-mentioned twenty-six species, the relationship of five 
is somewhat dubious, namely, Ilemimyzon formosanus, Spinibarbus 
hollandi, Pararasbora moUrechti, Candidia barbata, and Phoxiscus 
kiknchii. One is very closely related to a species from the interior 
of Hainan, namely Scaphesthes tamusuiensis. Five have their nearest 
relatives in China, four- in Corea, one in Amur Province, and eight in 
Japan. There is no record regarding near relatives of the remaining 
two, namely, Barbodes paradoxus and Capoela snyderi. But the 
majority of the fishes which belong to these genera are distributed in 
British India, Indo-China, and China. Therefore, there is no doubt 
with reference to their relationship with the continental forms. 

Formosan Species. Nearest Relatives. 

1. Scaphesthes tamusuiensis. Scaphesthes leptiirus from Hainan. 

2. Formosania gilberti. Formosania stenosoma from China. 

3. Acrossocheilus formosanus. Acrossocheilus kreyenhergi from 

China. 

4. Chanodichthys macro ps. Chanodichthys stenzi from China. 

5. Pseudogobio brevirostris. Pseudogobio sinensis from China. 

6. Channa formosana. Cha7ina ocellata horn China. 

7. Liobagrus nanto'ensis. Liobagrus andersoni from Corea. 

8. Liobagrus formosanus. Liobagrus andersoni irom Corea. 

9. Gnathopogon iijimce. Gnathopogon coreanus from Corea. 

10. Rhinogobius candidius. Rhinogobius bedfordi from Corea. 

11. Culler aokii. Culler sieboldi irom Amur. 

12. Pseudobagrus taiwanensis. . Pseudobagrus aurantiacus from 

Japan. 

13. Pseudobagrus adiposalis. Pseudobagrus aurantiacus from 

Japan. 

14. Pseudobagrus brevianalis. Pseudobagrus aurantiacus from 

Japan. 

15. Ischikauia macrolepis. Iscliikauia steenackeri from Japan. 



316 Annals of the Carnegie Museum. 

1 6. Zacco pachycephalus. Zacco temmincki from Japan. 

17. Acheilogiuithus himantegiis. Acheilognathus cyanostigma from 

Japan. 

18. Rhinogobiiis formosaniis. Rhinogohius hadropterus from Japan. 

Not only such a relationship is manifested by the peculiar species, 
but by the rest of the fresh-water fishes as well. Of the twenty-one 
species, which occur outside of Formosa, Clarias fuscus, Capoeta 
semifasciolata, DistcecJiodoii tumirostris, Culter hrevicauda, Cultriculus 
kneri, Ophicephalus tadianus, and Ophicephalus maculatns are species 
of southern affinities, because they are distributed in Indo-China and 
South China, but not to the north of the Yang-tze-kiang. Polyacan- 
tJiiis is a genus of the Indo-Malayan type, extending into the Malay 
Archipelago, but not occuring in eastern Asia. Such being the case, 
Polyacanthus operculatus may be included in this category, though it 
has been recorded from Tien-tsin, North China. 

Eleven species are of more or less general distribution, extending 
from South China to Corea and Japan proper, and one occurs only 
upon the Chinese mainland. Finally, Sicyoptcnis japonicus is one 
of the peculiar species of Japan, and its relationship is somewhat 
dubious, as it is not known to occur on the Asiatic continent. 

It will thus be seen that all the Formosan fresh- water fishes which 
have Chinese affinities ditTerentiate into more or less distinct species, 
while those of southern affinities have remained unchanged. More- 
over, eight per cent of the non-peculiar species have been recorded 
from the Chinese mainland. These two facts explain very clearly 
that the island had been preoccupied by the fresh-water fishes of the 
Chinese fauna when those of the southern affinities appeared through 
South China. 

Next to the prevalence of Chinese affinities, the total absence of 
any indication of affinity to the fresh-water fish fauna of the Philippine 
Islands and Malay Archipelago is a very striking fact. As shown in 
the table, only one species is recorded from the Philippines, namely 
Ophicephalus maculatns. However, as it also occurs in South China, 
its way of dispersal is clearly indicated, though there is no record of it 
in India and the Malay Acrhipelago. Cyprinus carpio is another 
species which has been recorded from Java. But it is evident that 
the carp is not a native of Java, but an introduced species. 

According to Leonhard Stejneger, there exists the same relationship 
between Formosa and the Philippine Islands with regard to the her- 



The Fresh Water Fishes of the Island of Formosa. 317 
Geographical Distribution of the Formosan Fresh-Water Fishes. 



• 


c 


B 

03 


S 


3 
C 

'S 

>> 


s 
'x. 
U 

o 

XI 

n 


c 
a 
a 

X 


rt 


B 

U 



c 



J3 
3 


m 


rt 
c 






rt' 




y 

'> 



£ 

u 

s 
B 

< 


c 

rt 

a 

a 


i 

B 
a 


rt 

> 
s 
—I 


u 




V 

a 

u 



m 


T3 
C 
_« 

4) 

c 
'S. 

.9" 

15 

Oh 


Family SILURID.E: 

Parasiluriis asotus 

Pseudobagrus brevianalis 

Pseudobagrus taiwanensis 

Pseudobagrus adiposalis 

Liobagrus nantoensis 

Liobagrus formosanus 

Clarias fuscus 

Family c'OBITID^: 

Misgurnus- anguilUcaudalus . . . . 

Misgurnus decemcirrosus 

Cobitis iTnia 

Family HOMALOPTERID.E: 

Formosajiia gilberti 

Heminiyzon formosanus 

Family CVPRINID^: 

Carassius auratus 

Cyprinus carpio 

Acrossocheilus formosanus 

Scapheslhcs tamusuiensis 

Hemibarbus labeo 

Barbodes paradoxus 

Capoeta semifasciolata 

Punlius snyderi 

Spinibarbus hollandi 

Gnathopogon iijimx 

Pseudogobio brevirostris 

Pseudorasbora parva 

Pararasbora mollrechti 

Phoxiscus kikuchii 

Distoechodon tumirostris 

Ischikauia macrolepis 

Acheilognalhus himanlegus 

Rhodeus ocellatus 

Zacco platypus 

Zacco lemmincki 

Zacco pachycephalus 

Candidia barbata 

Parabramis macrops 

Culler aokii 

Culler brevicauda 

Cullriculus kneri 

Family PCECILIID^: 

Oryzias laticeps 

Family LABYRINTHICI: 

Polyacanthus opercidaris 


- 


— 


- 


~ 


- 

X 

X 
X 


X 


X 
X 
X 
X 
X 
X 
X 

X 
X 
X 

X 
X 

X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 
X 

X 

X 


X 

X 

X 
X 
X 

X 
X 

X 

X 

X 

X 
X 

X 
X 

X 


X 

X 
X 
X 

X 
X 

X 

X 


X 

X 

X 
X 

X 

X 

X 
X 

X 


X 

X 
X 

X 
X 

X 
X 


X 

X 
X 

X 
X 

~ 
X 

X 

X 
X 

X 


- 


X 




- 


- 



318 



Annals of the Carnegie Museum. 



Geographical Distribution of the Formosan Fresh-Water Fishes.- 

Continued. 







B 

3 

pa 




J5 

S 
'a 

>< 


u 
6 

T3 


C 


ci! 

O 

B 

U 


u 

.a 

3 
O 


u 

O 


V 

o 
U 


c 

■> 
o 

u 

0, 
u 
3 

E 


c 
a 
>— 1 


E 

3 


> 

1—1 


V 

u 


o 

u 

c 

i- 

o 

M 


•a 
a 
_ca 

V 

a 
'S, 

.2* 

CL, 


Family OPHICEPHALID^: 

Ophicephalus ladianus 

Ophicephalns maciilatus 

Channa formosana 


— 


— 


- 


- 


X 


- X 

- X 

- V 


X 


- 




- 


X 


- 


- 


- 


- 


X 


Family GOBIID^: 

Sicyopleriis japonicus 

Rhinogobiiis candidius 

Rhinogobins taiwanus 

Rhinogohius formosanus 


- 






X 
X 
X 
X 


- 



petological fauna. He states that "A number of wide-ranging species 
of southern origin occur in both faunas, but as these also occur in 
southern China, on the mainland opposite Formosa, their way of 
dispersal is clearly indicated. There are only two species of this 
category which have not yet been collected in Chinese territories, 
namely, Dasia snmragdina, of wide distribution, and which may owe 
its occurence in Formosa to introduction by human agency, the other 
being a snake, Psammodynastes pulvernlentiis, the discovery of which 
within the limits of China would not cause surprise, as its known dis- 
tribution includes Sikkim, Assam., and the Shan states." (Proc. 
U. S. Nat. Mus., XXXVIII, 191 1, pp. 93-94.) Finally he has ex- 
pressed his belief that there has been no direct land connection be- 
tween Formosa and the Philippine Islands since Formosa received its 
batrachians and reptiles, because of the total absence of the Formosan 
herpetological fauna in the latter. The case of the fresh-water fishes is 
quite the same. Therefore it is reasonable to support his view with 
reference to the relationship between Formosa and the Philippine 
Islands. 

On the contrary, the occurence of all Japanese species in the main- 
land opposite to Japan is another interesting fact. There seems to be 
good reason for asserting the prehistoric land connection between 
Japan proper and the Asiatic continent, though the relationship 
between Japan and Formosa is somewhat dubious on account of the 
total absence of fresh-water fishes in the Riu Kiu Islands which cover 
the interspace between the two. 



The Fresh Water Fishes of the Island of Formosa. 319 



Bibliography. 
Abbott, James Francis. 

1901. List of Fishes collected in the River Pei-Ho, at Tien-Tsin, China, by- 
Noah Fields Drake, with Descriptions of Seven New Species. Proc. 
U. S. Nat. Mils., XXIII, pp. 483-491. 7 figs. 
Basilewsky, Stephanus. 

1855. Ichthyographia Chinae Borealis. Noiiv. Mein. Soc. Nat. Moscow, X, 
pp. 215-264. 9 pis. 
Berg, Lew Semenowitch. 

1906. Description of a New .Species of Leucogobio from Korea. Ann. Mag. 

Nat. Hist. (7), XVILpp. 394-395- 

1907. Description of a New Cyprinoid Fish, Acheilognathus signifer, from 

Korea, with a Synopsis of All the Known Rhodinae. Attn. Mag. 

Nat. Hist. (7), XIX, pp. 159-163. 
1907. Description of a New Cyprinoid Fish, Paraleucogohio nolacanthiis, 

from Northern China. Ann. Mag. Nat. Hist. (7), XIX, pp. 163-164. 
1907. Notes sur Quelques Especes Palearctiques du Genre Phoxinus. Ann. 

Mus. Zool. Acad. Sci. St. Petersb., XI, pp. 196-213. 
1907.- Notice sur Gobio rivularts Basilewsky. Ann. Mus. Zool. Acad. Sci. 

St. Petersb., XI, p. 2. 
1907. Notice sur le Groupe des Rhodeinae (fam. Cyprinidae) du Bassin de 

I 'Amour. Ann. Mus. Zool. Acad. Sci. St. Petersb., XI, pp. 11-14. 
1907. Les Poissons du bassin de 1 'Amour. Ichthyologia Amurensis. Bull. 

Acad. Sci. St. Petersb. (6), I, pp. 326-327. 
1907. A Review of the Cobitoid Fishes of the Basin of the Amur. Proc. 

U. S. Nat. Mus., XXXII, pp. 435-438. 
1907. Revision des Poissons d'Eau Douce de la Coree. Ann. Mus. Zool. 

Acad. Sci. St. Petersb., XII, pp. 1-12. 
1909. Ichthyologia Amurensis. Mein. Acad. Sci. St. Petersb. (8), XXIV, 

pp. 1-270. 3 pis. 
1909. Ueber die Gattungen Acanthogobio Herzenstein und Hemibarbus 

Bleeker. Ann. Mus. Zool. Acad. Sci. St. Petersb., XIV, pp. 103-106. 
1912. Ueber die Zusammensetzung und Herkunft der Fischfauna des Amur- 

Flusses niit Bezug auf die Frage von den Zoogeographischen Re- 

gionen fiir die Susswasserfische. Zool. Jahrb. (Abt. Syst.), XXXII, 

PP- 475-520. 
Bleeker, Pieter. 

1853. Diagnostische Beschrijvingen van Nieuwe of Wenig Bekende Visch- 

oorten van Sumatra. Nat. Tijdschr. Ned. Ind., IV., pp 243-302. 

1858. Conspectus Specierum Mugilis Archipelagi Indici Analyticus. Nat. 

Tijdschr. Ned. Ind., XVI, pp. 275-280. 

1859. Derde Bijdrage tot de Kennis der Vischfauna van Sombawa. Nat. 

Tijdschr. Ned. Ind., XIX, pp. 434-440. 
i860. Conspectus Systematis Cyprinorum. Nat. Tijdschr. Ned. Ind., 

XX, pp. 421-441. 
i860. Zesde Bijdrage tot de Kennis der Vischfauna van Japan. Act. Soc. 

Sc. Indo-Neerl., VIII, pp. 1-104. 



320 Annals of the Carnegie Museum. 

1862 Atlas Ichthyologique des Indes Orientales Neerlandaise, Publie sous 

to les Auspices du Gouvernement Colonial Neerlandais. I-IX. 

1877. 
1863. Notice sur les Noms de Quelques Genres de la Famille des Cyprinoides. 

Versl. Akad. Amsterdam, XV, pp. 261-264. 
1863. Sur une Nouvelle Espece de Poisson du Japon Appartenant a un 

Nouveau Genre. Versl. Akad. Amsterdam, XV, pp. 257-260. 
1865. Notices sur Quelques Genres et Especes de Cyprinoides de Chine. 

Ned. Tijdschr. Dierk., II, pp. 18-29. 

1869. Neuvieme Notice sur la Faune Ichthyologique de Japon. Versl 

Akad. Amsterdam (2), III, pp. 237-252. 

1870. Description d'une Espece Inedite de Botia de Chine et Figures du 

Bolia elongata et du Botia modesta. Versl. Akad. Amsterdam (2), 
IV, pp. 254-256. 
1870. Description et Figure d'une Espece Inedite de Hemibargus de Chine. 
Versl. Akad. Amsterdam (2), IV, pp. 257-258. 

1870. Mededeeling Omtrent Eenige Nieuwe Vischoorten van China. Versl. 

Akad. Amsterdam (2), IV, pp. 251-253. 

1871. Memoire sur les Cyprinoides de Chine. Verh. Akad. Arnslerdam, 

XII, pp. 1-91. 

1872. Notice sur les Peintures Chinoises de Cyprinoides Deposees au Museum 

de I'Universite de Groninque par M. J. Senn van Basel. Versl. Akad. 
Amsterdam (2), VI, pp. 117-121. 

1873. Sur les Especes du Genre Culler Basilewsky. Ned. Tijdschr. Dierk., 

IV, pp. 89-91. 
1875. Notice sur VElopichthys dahuricus. Proc. Zool. Soc. London, pp. 
534-536. 

1878. Sur les Especes du genre Hypophthalmichthys Bleeker. Versl. 

Akad. Amsterdam (2), XII, pp. 209-218. 

1879. Memoire sur les Poissons a Pharyngiens Labyrinthiformes de I'lnde 

Archipelagique. Verh. Akad. Amsterdam, XIX, pp. 1-56. 
1879. Sur Quelques Especes Inedites ou Peu Connues de Poissons de Chine 
appartenant au Museum de Hambourg. Verh. Akad. Amsterdam, 
XVIII, pp. 1-17. 
Block, M.\rc Elieser. 

1785. Ichthyologie, ou Histoire Naturelle, Generale et Particuliere des 
Poissons. 12 vols. 
Boulenger, George Albert. 

1894. Descriptions of a New Lizard and a New Fish Obtained in Formosa. 

Ann. Mag. Nat. Hist. (6), pp. 462-463. 

1895. List of Freshwater Fishes Collected by Mr. A. Everett on Palawan 

and Balabac. Ann. Mag. Nat. Hist. (6), XV, pp. 185-187. 
1899. On the Reptiles, Batrachians, and Fishes Collected by the late Mr. 

John Whitehead in the Interior of Hainan. Proc. Zool. Soc. London, 

pp. 956-962. 
1901. Descriptions of New Freshwater Fishes Discovered by Mr. J. W. 

Styan at Ningpo, China. Proc. Zool. Soc. London, March, p. 268-271. 



The Fresh Water Fishes of the Island of Formosa. 321 

Chaudiu^ri, Banavvari Lal. 

19 1 2. Contributions to the Fauna of Yunnan Based on Collections made by 
J. Coggin Brown. Part II, Fishes. Rec. Indian Mus., Calcutta, 
VI, pp. 13-24. 
CuviER, Georges, et Valenciennes, Achille. 

1828-1846. Histoire Naturelle des Poissons. 22 vols. 
Day, Francis. 

1867. On the Fishes of the Ncilgherry Hills and Rivers Around their Bases. 
Proc. Zool. Soc. London, pp. 281-302. 

1869. On the Fishes of Orissa. Proc. Zool. Soc. London, pp. 296-310; 

369-387.' 

1870. Description of Five New Species of Fishes from Burmah. Proc. 

Zool. Soc. London, pp. 99—101. 
1875. The Fishes of India. London. 
1888. Supplement to the Fishes of India. London. 
DvBOWSKi, Benedikt Ivan. 

1869. \'orlaufige Mittheilungen uber die Fischfauna des Ononfiusses und 

des Ingoda in Transbaikalien. Verh. Zool.-Bot. Ver., Wien, XIX, 

pp. 945-958. 
1872. Zur Kenntniss der Fischfauna des Amurgebietes. Verh. Zool.-Bot. 

Ver., Wien, XXII, pp. 209-222. 
Evermann, Barton Warren, and Se.'^le, Alvin. 

1906. Fishes of the Philippine Islands. Bull. U. S. Bur. Fish., XXVI, pp. 

49-110. 
1906. Fishes Collected in the Philippine Islands by Maj. Edgar A. Mearns, 

Surgeon, U. S. Army. Proc. U. S. Nat. Mus., XXXI, pp. 505-512. 
Fowler, Henry Weed. 

1899. Notes on a Small Collection of Chinese Fishes. Proc. Acad. Nat. 

Set. Philad., pp. 482-896. 
1910. Description of Four New Cyprinoids (Rhodeinae). Proc. Acad. Nat. 

Sci. Philad., pp. 476-486. 
Franz, Viktor. 

191 0. Die Japanischen Knochenfische der Sammlungen Haberer und Doflein. 

Abh. Bayer. Akad. Wiss., 2. CL, Suppl. IV, No. i, pp. i-i35- 
Gill, Theodore Nicholas. 

1858. Prodromus Descriptionis Subfamilije Gobinarum Squamis Cycloidi. 

Piscium, a W. Stimpsono in mare Pacifico Acquisitorum. Anns 

Lyceum Nat. Hist., N. Y., VII, pp. 12-16. 
i860. Notes on a Collection of Japanese Fishes, Made by Dr. J. Morrow. 

Proc. Acad. Nat. Sci. Philad., pp. 144-150. 
1 86 1. Conspectus Piscium in Expeditione ad Oceanum Pacificum Septen- 

trionalem, C. Ringoldio et J. Rodgersio Ducibus, a Gulielmo Stimp- 
sono Collectorum. Sicydianae. Proc. Acad. Nat. Sci. Philad., 

i860, pp. 100-102. 
GiNSBURG, Isaac. 

191 7. On Two New Species of Fishes from Yalu River, China. Proc. U. S. 

Nat. Mus., LIV, 191 7, pp. 99-101. 
22 — JAN. 12, 1920. 



322 Annals of the Carnegie Museum. 

GiJNTHER, Albert. 

1859-70. Catalogue of the Fishes of the British Museum. 8 vols. 

1872. Report on the Deep-Sea Fishes Collected by H. M. S. "Challenger" 

during the years 1873-76. Ann. Mag. Nat. Hist., pp. 397-399- 

1873. On a Collection of Fishes from Chefoo, North China. Ann. Mag. 

Nat. Hist. (4), XII, pp. 377-380. 

1873. Report on a Collection of Fishes from China. Ann. Mag. Nat. Hist. 

(4), XII, pp. 239-250. 

1874. Third Notice of a Collection of Fishes made by Mr. Swinhoe in China. 

Ann. Mag. Nat. Hist. (4), XIII, pp. 154-159. 
1880. Report on the Shore Fishes Procured During the Voyage of H. M. S. 
"Challenger" in the years 1873-1876. Voyage of H. M. S. "Chal- 
lenger," Zoology, I, Pt. VI, 82 pp. 

1887. Report on the Deep-sea Fishes Collected by H. M. S. "Challenger" 

during the years 1873-76. Voyage of H. M. S. "Challenger," 
Zoology, Vol. XXII, Pt. 57. PP- 1-268. 

1888. Contribution to our Knowledge of the Fishes of the Yangtsze-Kiang, 

Ann. Mag. Nat. Hist. (6), I, pp. 429-435. 

1889. Third Contribution to our Knowledge of Reptiles and Fishes from the 

Upper Yangtsze-Kiang. Ami. Mag. Nat. Hist. (6), IV, pp. 218-229. 
1892. List of the Species of Reptiles and Fishes Collected by Mr. A. E. Pratt 

on the Upper Yangtsze-Kiang and in the Province Sze-Chuen, with 

Descriptions of New Species. Pratt; "Snow of Tibet," 1892, pp. 

238-250. 
1896. Report on the Collection of Reptiles, Batrachians, and Fishes made by 

Messrs. Potanin and Berezowski in the Chinese Provinces Kansu 

and Sze-Chuen. Ann. Acad. Nat. Set. Petersb., pp. 199—219. 
1898. Report on a Collection of Fishes from Newchang, North China. Ann. 

Mag. Nat. Hist. (7), I, pp. 257-263. 
Herzenstein, Salomon Markovich. 

1892. Ichthyologische Bemerkungen aus dem Zoologischen Museum der 

Kaiserlichen Akademie Wissenschaften. Melanges Biol. Bull. 

Acad. Sci. St. Petersb., XIII, pp. 219-235. 
Herzenstein; Salomon Markovich & V.^crpakovskii, N. 

1887. Notizen iiber die Fischfauna des Amur-Beckens und der angrenzenden 

Gebiete. Trudi St. Petersb. Nat., XVIII, pp. 1-58. 

HjlLGENDORF, FrANZ MaRTIN. 

1878. Ueber Einige Neue Japanische Fischgattungen. Sitzber. Ges. Naturf. 

Freunde Berlin, pp. 155-157. 
Ishikawa, Chiyomatsu. 

1895. A Preliminary Note on the Fishes of Lake Biwa. Zool. Mag. Tokyo 

VII, No. 82, p. 120. 
1898. On the Variations of the Proportional Length of the Head, etc., as to 

the Total Length in our Common Eel. Annot. Zool. Jap., Tokyo, 

II, pp. 125-126. 
1901. Notes on Two New Species of Fishes from the Lake Biwa. Annot. 

Zool. Jap., Ill, pp. 161-164. 



The Fresh Water Fishes of the Island of Formosa. 323 

1904. Notes on Some New or Little-known Fishes of Japan. Proc. Nat. 

Hist. Tokyo Museum, I, pp. 1-17. 
IsHiKAWA, Chiyomatsu & Takahashi, Nisuke. 

1914. Note on the Eels of Japanese, Corean, Formosan, and Adjacent Waters. 

Xotirn. Coll. Agricult. Tokyo, IV, No. 7, pp. 415-433. 
Jordan, David Starr. 

1902. The Salmon and Trout of Japan. Annot. Zool. Jap., IV, pt. 2, pp. 

69-75- 

1905. Note on the Salmon and Trout of Japan. Annot. Zool. Jap., Y, pp. 

161-162. 
1913. Description of Anguilla manabei, a New Eel from Japan. Proc. U. S. 

Nat. Mus., XXXIV, pp. 359-360. 
Jordan, Da\id Starr & Evermann, Barton Warren. 

1896. The Fishes of North and Middle America. Bull. U. S. Nat. Mus., 

XXXVII, 1896, Pt. 1; 1898, Pt. 2; Pt. 3; 1900, Pt. 4. 

1903. Notes on a Collection of Fishes from the Island of Formosa. Proc. 

U. S. Nat. Mus., XXV, pp. 315-368. 
Jordan, David Starr & Fowler, Henry W. 

1902. Notes on Little-known Japanese Fishes, with Description of a New 

Species of Aboma. Proc. U. S. Nat. Mus., XXV, pp. 573-576. 

1903. A Review of the Cobitidae, or Loaches, of the Rivers of Japan. Proc. 

U. S. Nat. Mus., XXVI, pp. 765-774. 
1903. A Review of the Cj^prinoid Fishes of Japan. Proc. U. S. Nat. Mus., 

XXVI, pp. 811-862. 
1903. A Review of the Siluroid Fishes or Catfishes of Japan. Proc. U. S. Nat. 
Mus., XXVI, pp. 897-911. 
Jordan, David Starr & Metz, Charles William. 

1913. A Catalog of the Fishes Known from the Waters of Korea. Mem. 
Carneg. Mus., VI, No. 2, pp. 1-65. 
Jordan, David Starr & Richardson, Robert Earl. 

1907. On a Collection of Fishes from Echigo, Japan. Proc. U. S. Nat. Mus., 

XXXIII, pp. 263-266. 

1908. List of Fishes Collected in the Philippines by Richard Crittenden 

McGregor. Bull. U. S. Bur. Fish., XXVII, pp. 233-287. 

1909. A Catalog of the Fishes of the Island of Formosa. Mem. Carneg. 

Mus., IV, No. 4, pp. 159-204. 

1910. Check-list of the Species of Fishes Known from the Philippine Archi- 

pelago. Manila, 1910, 78 p. 
Jordan, David Starr & Seale, Alvin. 

1905. List of Fishes Collected by Dr. Bashford Dean on the Island of Negros, 
Philippines. Proc. U. S. Nat. Mus., XXVIII, pp. 769-803. 

1905. List of Fishes Collected in 1882-83 by Pierre Louis Jouy at Shanghai 

and Hongkong, China. Proc. U. S. Nat. Mus., XXIX, pp. 517-529. 

1906. Descriptions of Six New Species of Fishes from Japan. Proc. U. S. 

Nat. Mus., XXX, pp. 143-148. 

1907. Fishes of Luzon and Panay. Bull. Bur. Fish. U. S., XXVI, pp. 1-48. 



324 Annals of the Carnegie Museum. 

1907. List of Fishes Collected at Hongkong by Captain William Finch, with 
Descriptions of Five New Species. Proc. Davenport Acad. Sci., X, 
pp. 1-17. 
Jordan, David Starr & Snyder, John Ottkrbein. 

1901. List of Fishes Collected in 1883 and 1885 by Pierre Louis Jouy and 
Preserved in the LTnited States National Museum, with Descriptions 
of Six New Species, rmc. U. S. Nat. Miis., XXIIL pp. 739-769. 

1 90 1. A List of Fishes Collected in Japan by Keinoske Otaki, and by the 
United States Fish Commission Steamer "Albatross," with Descrip- 
tions of Fourteen New Species. Proc. U. S. Nat. Miis., XXIH, pp. 
335-380. 

1901. A Preliminary Check-list of the Fishes of Japan. Atniol. Zool. Jap., 
Ill, pp. 1-159- 

1901. A Review of the Apodal Fishes or Eels of Japan, w-ith Descriptions of 
Nineteen New Species. Proc. U. S. Nat. Mus., XXIH, pp. 837-890. 

1901. A Review of the Ciobioid Fishes of Japan. Proc. U. S. Nat. Mus., 

XXIV, pp. 33-132. 

1902. A Review of the Salmonoid Fishes of Japan. Proc. U. S. Nat. Mus., 

XXIV, pp. 567-593- 

1906. On a Species of Loach, Alisgurnus dccemcirrosus (Basilewsk\) from 
Northern China. Proc. U. S. Nat. Mus., XXX, pp. 833-834. 

1906. A Review of the Poeciliidse or Killifishes of Japan. Proc. U. S. Nat. 
Mus., XXXL pp. 287-290. 
JORD.\N, David Starr, S.nyder, John Otterbein, & T.\n.a.k.\, Shigeho. 

1913. A Catalogue of the Fishes of Japan. Journ. Coll. Sci. Tokyo, XXXIH, 

pp. 1-497- 
J0RD.A.N, D.wiD Starr & Starks, Edwin Ch.a.pin. 

1903. Description of a New Cyprinoid Fish, Hemibarbus joileni, from the 

Pci-Ho, Tientsin, China. Smithson. Misc. Coll., 45, pp. 241-242. 

1905. On a Collection of Fishes made in Korea by Pierre Louis Jouy, with 

Descriptions of New Species. Proc. U. S. Nat. Mus., XXVI H, pp. 
193-211. 

1906. List of Fishes Collected on Tanega and Yaku, Offshore Islands of 

Southern Japan, by Robert Van Vleck Anderss, with Descriptions 
of Seven New Species. Proc. U. S. Nat. Mus., XXX, pp. 695-706. 

1907. List of Fishes Recorded from Okinawa or the Riukiu Islands of Japan. 

Proc. U. S. Nat. Mus., XXXII, pp. 491-504. 
1917. Notes on a Collection of Fishes from Ceylon, with Descriptions of 
New Species. Ann. Carneg. Mus., XI, Nos. 3 & 4, pp. 430-460. 
Jordan, David Starr & S\v.a.in, Joseph. 

1885. A Review of the American Species of Marine Mugilid;e. Proc. U. S. 
Nat. Mus., VII, pp. 261-275. 
Jordan, David Starr & Thompson, William Francis. 

1914. Record of the Fishes Obtained in Japan in 191 1. Mem. Carneg. Mus., 

VI, No. 4, pp. 205-313. 
Kner, Rudolf. 

1865. Reise dcr Osterreichischen Fregatte "Novara" um die Erde in den 



I 



The I"f<esh Water I'ishes (jf the Iseamj of Formosa. 325 

Jahren 1857-1859, untcr den Bcfehlen dcs Commodore B. von 
Wiillerstorf-lJrbair. Zoi'dogischer TheiL, Fische, 1-3 Abth. 
Kreyenberg, M. 

1910. Ein Beitrag zur Kenntniss der Fische der Yangtze und seiner Zuflusse. 

Abh. Mus. Nalur-Heimalk. Magdeberg, II, pp. 1-24. 

191 1. Eine Neue Cobitinen-Gattung (Gobiobotia pappenheimi n. sp.) aus 

China. Zool. Anz., XXXVIII, pp. 417-419. 
Kreyenberg, M. & Pappe.nheim, P. 

1909. Ein Beitrag zur Kenntniss der Fische der Yangtze und seiner Zuflusse. 
Sizber. Ges. Nalurf. Freunde Berlin, pp. 95-109. 
Lacepede, Bernhard Germain Etienne. 

1798-1803. Histore Naturelle Des Poissons. Tomes 5. 
Pellegrin, Jacques. 

1907. Mission Permanente Francaise en Indo-Chine. Poissons du Tonkien. 

Bull. Mus. Hist. Nat. Paris, XIII, pp. 499-503. 

1908. Poissons d'Eau Douce de Formosa. Bull. Museum Paris, XIV, 

pp. 262-265. 
Peters, Wilhelm Carl Hartwig. 

1864. Ueber Einige Neue Siiugethiere, Amphibicn, und Fische. Monalsb. 

Akad. Wiss. Berlin, pp. 381-399. 
1868. Ueber die von Herrn. Dr. F. Jagor in dem Ostindischen Archipel 

Gesammelten [Fische. Monalsb. Akad. Wiss. Berlin, pp. 254-281; 

460-461. 
1880. Ueber eine Sammlung von Fischen, welche Dr. Gerlach in Hongkong 

gesendt hat. Monalsb. Akad. Wiss. Berlin, pi). 1029-1037. 
1880. Ueber die von der Chinesischen Regierung zu der Internationalen 

Fischerei-Ausstellung Gesandte Fischsammlung aus Ningpo. Mon- 
alsb. Akad. Wiss. Berlin, pp. 921-927. 
PoPTA, Canna M. L. 

1907. Einige Fischarten aus Chine, Xenocypris lamperli und Chanodichlhys 

stenzii n. sp. Zool. Anz., XXXII, pp. 243-251. 
Regan, Charles Tate. 

1904. Descriptions of Two New Cyprinid Fishes from Yunnan Fu. Ann. 

Mag. Nat. Hist. (7), XIV, pp. 416-417. 

1904. On a Collection of Fishes made by Mr. John Graham at Yunnan Fu. 

Ann. Mag. Nat. Hist. (7), XIII, pp. 190-194. 

1905. Descriptions of Five New Cyprinid Fishes from Lhasa, Tibet, collected 

by Captain H. J. Walton. Ann. Mag. Nat. Hist. (7), XV, pp. 185- 

188. 
1905. Descriptions of Three New Fishes from Japan, Collected by Mr. R. 

Gordon Smith. Ann. Mag. Nat. Hist. (7), XVI, pp. 363-365. 
1905. Descriptions of Two New Cyprinid Fishes from Tibet. Ann. Mag. Nat. 

Hist. (7), XV, pp. 300-301. 

1905. On a Collection of Fishes from the Inland Sea of Japan made by Mr. 

R. Gordon Smith. Ann. Mag. Nat. Hist. (7), XV, pp. 17-26. 

1906. Descriptions of Two New Cyprinid Fishes from Yunnan Fu, Collected 

by Mr. John Graham. Ann. Mag. Nat. Hist. (7), XVII, pp. 332-333. 



326 Annals of the Carnegie Museum. 

1907. Descriptions of Three New Fishes from Yunnan, Collected by Mr. J. 

Graham. Ann. Mag. Nat. Hist. (7), XIX, pp. 63-64. 

1908. Descriptions of New Freshwater Fishes from China and Japan. Ann. 

Mag. Nat. Hist. (8), I, pp. 149-153. 
1908. Descriptions of Three New Cyprinid Fishes from Yunnan, Collected 

by Mr. John Graham. Ann. Mag. Nat. Hist. (8),- II, pp. 356-357. 
1908. Descriptions of New Fishes from Lake Candidius, Formosa, Collected 

by Dr. A. Moltrecht. Ann. Mag. Nat. Hist. (8), II, pp. 358-360. 
1908. Descriptions of Three New Freshwater Fishes from China. Ann. 

Mag. Nat. Hist. (8), I, pp. 109-111. 
1908. A Synopsis of the Fishes of the Subfamily .Salanginae. Anti. Mag. Nat. 

Hist. (8), II, pp. 444-446. 
1908. A Collection of Freshwater Fishes from Corea. Proc. Zool. Soc. 

London, pp. 59-63. 
1914. Fishes from Yunnan, Collected by Mr. John Graham, with Descriptions 

of a New species of Barilius. Ann. Mag. Nat. Hist. (8), XIII, 

pp. 260-261. 
Richardson, John. 

1844. Ichthyology. (In The Zoology of the Voyage of H. M. S. "Sulphur," 

under the Command of Captain Sir Edward Belcher, during the 
years 1836-42.) 

1845. Report on the Ichthyology of the Seas of China and Japan. Rept. 

Brit. Assoc. Adv. Sci., Fifteenth Meeting, 1845, pp. 187-320. 
1845. Ueber die Ichthyologie Chinas. Neue Notizen (Froroep), XXXV, 
No. 756, pp. 117-118. 

RUTTER, ClAUDSLEY M. 

1897. A Collection of Fishes Obtained in Swatow, China, by Miss Adele M. 
Fielde. Proc. Acad. Nat. Sci. Philad., pp. 56-90. 
Sauv.\ge, Henri Emile. 

1877. Considerations sur la Faune Ichthyologique des Eaux Douces de I'Asie 

et en Particulier de I'lndo-Chine. C. R. Assoc. Franc. Avanc. Sci., 
le 6me Session, pp. 615-620. 

1878. Note sur Quelques Cyprinidse et Cobitidinte d'Especes Inedites, 

Provenant des Eaux Douces de la Chine. Bull. Soc. Philom. Paris 
(7), II, pp. 86-90. 

1880. Notice sur Quelques Poissons de ITle Campbell et de I'lndo-Chine. 

Bull. Soc. Philom. Paris (7), IV, pp. 228-233. 

1 88 1. Sur une Collection de Poisson de Swatow. Bull. Soc. Philom. Paris 

(7). V, pp. 104-108. 
1881. Recherches sur la Faune Ichthyologique de I'Asie et Description d'Es- 
peces Nouvelles de I'lndo-Chine. Nouv. Arch. Mus. Hist. Nat. 
Paris, IV, pp. 123-194. 

1883. Sur une Collectioi;! de Poisson Requeillie dans Lac Biwako (Japon), 

par M. F. Steenackers. Bull. Soc. Philom. Paris (7), \'II, pp. 
144-150. 

1884. Contribution a la Faune Ichthyologique du Tonkin. Bull. Soc. Zool. 

France, IX, pp. 209-215. 



The Fresh Water Fishes of the Island of Formosa. 327 

Seale, Alvin. 

1905. Report on the Introduction of Top-minnows into the Hawaiian Islands; 

Hawaiian Forest Agric, II, Pt. 2, 364-367. 
1909. New Species of Philippine Fishes. Philippine Journ. Set., Sec. A, 

IV, pp. 491-543. 
1917. The Mosquito-fish, Gambusia affinis (Baird & Girard), in the Philippine 

Islands. Philip. Journ. Sci., XII, Sec. D, No. 3, pp. 177-187. 
Se.\le, Alvin & Bean, Barton A. 

1907. On a Collection of Fishes from the Philippine Islands, Made by Maj. 

Edgar A. Mearns, Surgeon, U. S. Army, with Descriptions of Seven 

New Species. Proc. U. S. Nat. Mus., XXXIII, pp. 229-248. 

SlEBOLD, PHILIPP FRANZ VON. *• 

1842. Fauna Japonica. Pisces. 
Smith, Hugh M. 

1901. Notes on Five Food-fishes of Lake Buhi, Luzon, Philippine Islands. 
Bull. U. S. Fish. Comm., XXI, pp. 167-171. 
Smith, Hugh M. & Pope, E. B. 

1905. List of Fishes Collected in Japan in 1903, with Descriptions of New 

Genera and Species. Proc. U. S. Nat. Mus., XXXI, pp. 459-499. 
Smith, Hugh M. & Seale, Alvin. 

1906. Notes on a Collection of Fishes from the Island of Mindanao, Philippine 

Archipelago, with Descriptions of New Genera and Species. Proc. 

Biol. Soc. Wash., XIX, pp. 73-82. 
Snyder, John Otterbein. 

1912. Japanese Shore-fishes Collected by the United States Bureau of 

Fisheries Steamer "Albatross" Expedition of 1906. Proc. U. S. 

Nat. Mus., XLII, pp. 399-450. 
1912. The fishes of Okinawa, one of the Riu Kiu Islands. Proc. U. S. Nat. 

Mus., LXII, 487-519.- 
Steindachner, Franz. 

1866. Ichthyologische Mittheilungen. IX. Ueber ein Neues Cyprinoiden- 

Geschlecht von Hongkong. Verh. Zool.-Bot. Ges. Wien., XM, 

pp. 782-784. 

1866. Ueber Cephalus hypophthalmus (Gray) Richardson. Verh. Zool.- 

Bot. Ges. Wien, XVI, pp. 383-384. 

1867. Ichthyologische Notizen (VI), Sitzber. Akad. Wiss. Wien, LVI, pp. 

307-376. 
1869. Ichthyologische Notizen (IX). Ueber Eine Neue Gattung und Art 

der Cyprinoiden aus China. Sitzber. Akad. Wiss. Wien, LX, pp. 

302-304. 
1881. Ichthyologische Beitrage (X). Beschreibungen von Zwei und Dreizig 

Fischarten aus Japan. Sitzber. Akad. Wiss. Wien, LXXXIII, Abt. 

I, pp. 179-219. 
1883. Ichthyologische Beitrage (XIII). Macrones chinensis n. sp. Sitzber. 

Akad. Wiss. Wien, LXXXVIII, p. 47-48. 
1892. Ueber Einige Neue und Seltene Fischarten aus der Ichthyologischen 

Sammlung des K. K. Naturhistorischen Hofmuseums. Denkschr. 

Akad. Wiss. Wien, LIX, i Abth., pp. 357-384. 



328 Annals of the Carnegie Museum. 

1896. Bericht iiber die wahrend der Reise Sr. Maj. Schiff "Aurora" von Dr. 
C. Ritter v. Microszewski in den Jahren 1859 und 1896, Gesammelten 
Fische. Ann. Natiirh. Hofmns. Wien, XI, pp. 197-230. 
Stewart, F. H. 

1911. Notes on Cyprinidae from Tibet and the Chumbi Valley, with a De- 
scription of a New Species of Gynocypris. Rec. Indian Mus. Calcutta, 

VI, pp. 73-92. 

T.\N.AKA, SHIGOHO. 

1908. Description of Eight New Species of Fishes from Japan. Annot. 
Zool. Jap., VII, pp. 27-47. 

1908. On Some Fishes from Lake Biwa with Description of One New Species 
'■ (Acheilognathus shhnazui) and a List of All the Fish Species Hitherto 

Known from that Locality. Ann. Zool. Jap., VII, pp. 1-15. 

1909. Notes on the Freshwater Fishes from the Province of Shinano. An7i. 

Zool. Jap., Vll, pp. 125-138. 
1909. Descriptions of One New Genus and Ten New Species of Japanese 

fishes. Joiirn. Coll. Sci. Tokyo, XXVII, art. 8, pp. 1-27. 
191 1. Figures and Descriptions of the Fishes of Japan. Tokyo. 
Vaillant, Leon Louis. 

1892. Remarques sur Quelques Poissons de Haut Tonkin. C. R. Acad. Sci. 

Paris, 114, pp. 1028-1029. 

1893. Sur les Poissons Provenant du Voyage de M. Bonvalot et du Prince 

Henri d'Orleans. Bull. Soc. Philotn. Paris (5), VIII, pp. 197-204. 
1904. Poissons Recueillis par M. A. Pavie en Indo-Chine. In Mission Pavie 

Indo-Chine, 1879-1895, Vol. Ill, pp. 459-470. 
1904. Quelques Reptiles, Batrachiens, et Poissons du Haut-Tonkin. Bull. 

Mus. Hist. Nat. Paris. X, pp. 297-301. 
Weber, Max & Beaufort, L. F. 

191 1. The Fishes of the Indo- Australian Archipelago. 3 vols. 



IV. ON ELEPHENOR, A NEW GENUS OF FISHES FROM 

JAPAN. 

By David Starr Jordan. 

(Plates LIV-LVIII.) 

In the Atti Soc. Nat. Italiana of Milan in 1903, p. 137, fig. 6, Dr. 
Cristoforo Bellotti described a peculiar fish from near Yokohama, 
Japan, under the name of Pteraclis macropus. 

This fish, extremely fragile in structure, and with an exceedingly 
high dorsal fin, is obviously not a true Pteraclis, as in that genus the 
anal fin is about as long as the dorsal, and the scales are firm and hard, 
while in P. macropus, the base of the anal is only about half as long as 
that of the dorsal, and the scales are small, thin, and caducous. 

In 1905, Gill and Smith described from the Japanese island of 
Shikoku, a species bearing a striking resemblance to Bellotti's fish, 
under the name of Caristms japonicus (Proc. Biol. Soc. Washington, 
XVIII, 1905, p. 249). In Proc. U. S. Nat. Mus. XXXI, 1906, p. 490, 
fig. 10, Smith & Pope published a figure of this species which we here 
reproduce (Plate LIV, upper figure). In the American Naturalist. 
December, 1912, p. 748, I printed a notice of the work of Dr. R. H. 
Shufeldt on the genus Pterycombus, a relative of Pteraclis, remarking 
that "the singular Caristius, lately described from Japan by Dr. 
Smith, is an ally of Pterycombus and belongs to the same family". 
This view, I repeated in a memoir by Jordan & Thompson, "Record 
of the Fishes obtained in Japan in 191 1" (Memoirs Carnegie Mus., 
VI, September, 1914, p. 245). In this paper we reprinted the plate 
of Caristius japonicus Smith & Pope with a new plate of Bellotti's 
fish, which we called Caristius macropus. Of the genus Caristius we 
said, "Its afifinities seem obviously to be wuth the scombroid forms, 
especially with Pteraclis, the genus in which Bellotti placed it." 

This opinion was based on a specimen of Pteraclis macropus from 
the Kuro Shiwo, or Gulf Stream of Japan, obtained by me for the 
Carnegie Museum from Mr. Alan Owston in 191 1. I regarded this 
as a Caristius and I called it Caristius macropus. Meanwhile Dr. 
Erich Zugmayer of Munich (Result. Camp. Sci. Monaco, XXXV, 

329 



330 Annals of the Carnegie Museum. 

191 1, p. loi, PI. V. fig. 5) had described a deep sea berycoid fish under 
the name of Platyheryx opalescens (Plate LV, upper figure). This 
fish strongly resembles Caristiiis japonicus . Mr. C. Tate Regan 
(Ann. Mag. Nat. Hist. (8), Vol. X, Dec, 191 2, p. 637, compared 
Platyberyx to Caristius and expressed the belief that both are berycoid 
fishes, "probably congeneric and perhaps not specifically distinct from 
each other." 

I now think Mr. Regan is right in aligning Caristius with the 
berycoid fishes. In Caristius as in Platyberyx, there are numerous 
short, stiff rays, or fulcra, at the base of the caudal'fin both above and 
below. There are only about sixteen developed rays in the caudal 
fin, and the presence of but I, 5 rays in the ventral fins in both, is not 
decisive, as some of the varied genera referred to the berycoid group, 
notably Bathyclupea, show the same number. 

I was mistaken, however, in placing Pteraclis niacropus in the genus 
Caristius, and in taking my idea of Caristius from that species. I was 
quite right, I think, in rejecting the idea of berycoid afifinities for 
P. niacropus. In spite of superficial resemblance, P. fnacropus has 
really very little in common with Caristius or Platyberyx. It has no 
true fulcra on the caudal fin and the slender rays of that fin are more 
than twenty in number. Moreover all the rays of the dorsal fins are 
slender, unjointed spines. The species is, in fact, allied to the Pter- 
aclidcc and forms the type of a new genus which I may call Elepheyior} 
But its divergence from the known genera of Pteraclidcc is very strong, 
entitling it, I think, to distinction, as representing a separate family. 
For the present at least, we consider it as belonging to a new family, 
Elephenoridce, distinguished by its small, weak scales, and its relatively 
short anal fin. 

In the type of Caristius japonicus the dorsal and anal rays are 
nearly all broken off short, and it is impossible from the plate to know 
whether any or all of them were spines. The dorsal fin is elevated in 
front and the numbers of the rays are about as in Elephcnor. 

In Platyberyx the dorsal and anal fins are scarcely elevated, and 
the rays are nearly all jointed and branched, the numbers being 
D. II. 28; A. II, 16. In Platyberyx, as in Caristius and Elephenor the 
ventral fins are well developed, their rays I, 5. The difference in the 
form of the dorsal fin would seem to indicate that Platyberyx is generi- 
cally distinct from Caristius. 

1 EXe0i7J'wp, chief of the fleet Abantes of EubcEa, "with long hair flowing 
behind," captain of one of the ships of Achilles. 



Jordan: Elephenor, A New Genus of F"ishes from Japan. 331 

Analysis of Genera Related to Pteraclis. 
(Dorsal and anal fins composed exclusively of spines.) 
.4. Dorsal and anal spines very slender, close set, most of them almost hair-like. 
B. Pteraclid-^. Scales rather large, firm and hard, more or less lobate; 
anal fin beginning far forward, before the pectorals, its base 
almost as long as that of dorsal; dorsal fin very high, continuous, 
its spines very slender and flexible, highest anteriorly, dorsal 
spines about fifty-three, anal spines about forty-three; both 
dorsal and anal depressible in a basal sheath of large scales; 
ventral fins very small, weak, I, 5 (or perhaps sometimes I, 3) 
jugular, inserted below (or before) the eye; form of body ovate- 
elliptical, much compressed; nape not greatly elevated; forehead 
rounded, not subvertical. 
C. Dorsal and anal fins each with one of the anterior spines much 
thickened; dorsal fin beginning on tip of snout, and beginning 
just behind eye, the long rays of dorsal and anal nearly reaching 
caudal. Dorsal with about eight graduated spines in front, the 
eighth longest; ventrals very small, the rays I, 5; vent directly 

below ej^es Bentenia.'' 

CC. Dorsal fin beginning on the snout before the eye, with two to four 
graduated spines, none of them notably thickened; anal fin 
similar, both very high; longest dorsal spine when depressed 
reaching past middle of fin; ventral rays I, 5 or I, 3.. Pteraclis. ' 
CCC. Dorsal fin beginning behind the eye with nine to thirteen graduated 
spines in front, none of them enlarged; ventral rays I, 5. 
D. Anal fin high, its first ray reaching bej'ond middle of fin; gradu- 
ated spines of dorsal 10 to 13 Centropholis.^ 

* CenlrophoUs Hilgendorf, Sitzungsbcrichte Naturforschende Freunde, Berl., 
1878, p. i; type CenlrophoUs pelersi Hilgendorf; off Tokyo. 
DD. Anal fin of moderate height, its first long ray reaching only to 
middle of fin; dorsal beginning over preopercle, with nine 

graduated spines Pterycombus.^ 

BB. Elephenorid.«; Scales very small, thin, caducous; dorsal sheath probably 
caducous;^ ventral fins large, i, 5, inserted just before pectorals. 

- Bentetiia Jordan & Snyder, Journ. College Sci. Imp. Univ. Tokyo, XV; 1901, 
306; type Bentenia cesticola Jordan & Snyder from the Kuro Shiwo (Japan Current) 
off Misaki, Japan (C/. Plate LVI of this Article.) 

3 Pteraclis Gronow; Zoophylaceum, 1763, p. 136 and also in Acta Helvetica, VH, 
1772, p. 44. Type Coryphcena velifera Pallas; Indian Ocean. Pteridimn Scopoli 
(1777) and Oligopodus Lacepede (1800) are based on the same species. (C/. PI. LV, 
lower figure, of this articles.) 

'" Plerycombus Fries, Kongl. Vet. Akad. Handl. Stockholm, 1837, pp. 14, 22, 
PL II; type Pierycombiis brama Fries; Arctic Coast of Norvvay. (C/. Plate LVII, 
upper figure of this article.) ♦ 

^ Bellotti's figure shows plainly a dorsal and anal sheath of scales. There is no 
trace of these in our specimen. 



332 Annals of the Carnegie Museum. 

Anal fin beginning notably behind insertion of pectorals, the length 

of its base only about half that of the dorsal; dorsal fin exceedingly 

high, its first two rays short, graduated, the third longest, reaching 

past base of caudal. Anal moderate, its longest ray about reaching 

middle of fin; upper and lower rays of caudal with fine recurved hooks; 

front of head straight, subvertical, the nape elevated. Dorsal spines 

34, anal spines 22. Elephenor. 

A A. D1ANID.E. Dorsal and anal spines somewhat pungent, wide set, not specially 

elevated (about twenty-two in the dorsal, eighteen in the anal); anal 

shorter than dorsal beginning just behind base of pectoral, the spines rather 

low, the longest behind middle of fin; ventral very small, jugular, under 

posterior part of gill-opening; nape greatly elevated, the dorsal beginning 

just behind eye; caudal peduncle very slender; scales reduced to small, 

star-like tubercles; no sheath at base of dorsal or anal Diana." 

We may note that Pteraclis papilio Lowe, from Madeira, is said to 
have one of the anterior spines of dorsal and anal notably enlarged. 
It probably belongs to Bentenia. Pteraclis carolinus Cuv. & Val. 
from off South Carolina and Pteraclis trichipterns Cuv. & Val. (Plate 
LVII, lower figure) from unknown locality (but doubtless in the 
Pacific, having been collected by Quoy & Gaimard) are said to have 
the third or fourth dorsal spine somewhat enlarged. Nothing is said 
as to the position of the ventral fin, but it may be assumed that this is 
jugular as in Pteraclis ocellatus. Until we have further information 
we may leave these in Pteraclis. The figure of Pteraclis ocellatus 
(Plate LVIII, upper figure) does not show any spine to be enlarged, 
but the description states that the third spine is thick and easily 
divides itself into two halves, left and right. The number of rays in 
the ventral fins is also uncertain. Pteraclis papilio is said to have the 
ventrals I, 5 as in Bentenia. P. ocellatus was thought to have these 
rays i, 3, but Valenciennes regards this count as doubtful. In 
P. velifera but one ray is indicated, the soft rays being probably all 
broken off. As these fins are extremely fragile, it is likely that I, 5 is 
the normal number in all the species. 

The figure of the type of Pteraclis, (P. velifera) as copied by Bonna- 
terre from Pallas, shows the third spine apparently slightly enlarged 
and nearly as long as the spines succeeding. The copy of this plate 
given by Schneider is erroneous in several respects. 

The species called Pteraclis carolinus by Goode & Bean (Oceanic 
Ichthyology, p. 212, pi. LIX, fig. 218) has none of the dorsal spines 

''Diana Risso, Europe Meridionale, III, p. 267, 1826; type Diana semilunata 
Risso, from off Nice. Aslrodermus giiUatns Bonelli 1829, is the same species. 



Jordan: Elepiienor, A New Genus of Fishes from Japan. 333 

enlarged, while the first ten are graduated. {Cf. Plate LVIII, of 
this article, lower figure.) The dorsal fin begins behind the eye. It 
cannot be the same as the original P. carolinus. It is plainly a 
Centropholis and it may stand as a new species, Centropholis goodei 
Jordan. The ventral rays arc I, 5, as in Centropholis. 

The tj'pe-specimen of Bentenia cesticola is in the Imperial Museum 
of Tokyo. In connection with the description of Jordan and Snyder, 
Dr. Mitsukuri published a photograph of this example. A drawing 
made from this photograph, with the torn vertical fins restored is 
given on Plate LVI. In this photograph, an appendage resembling 
the ventral fins appears under the lower jaw. Mr. Shigeho Tanaka, 
ichthyologist of the Imperial University, has reexamined the type, 
and finds this appendage to be a detached piece of the left branchio- 
stegal membrane, torn loose by accident. The ventral fins were 
described as "I, 5" and "jugular" by Jordan and Snyder. They do 
not appear in this photograph and Mr. Tanaka says that they are 
not present on the actual specimen, and he finds no scar where the 
fins might have been torn off. Being very fragile, they have probably 
been broken off in manipulation. 

If one of the dorsal spines is really enlarged in the type species of 
Pteraclis, the genus Bentenia must be very close to it. The tip of the 
shoulder-girdle, the presumable insertion of the ventral fins, is a little 
farther forward in Bentenia. 

It is a curious fact that nearly all of the species noted in this paper 
are each known from only a single specimen, and the others from very 
few. As surface fishes of the ocean currents they are very rarely 
taken, and are doubtless nowhere abundant. 

The PteradidcE, Pterocyidce, and Dianidce differ from the related 
families in having all the dorsal rays simple, not jointed, nor branched. 
In the Bramidcc, also closely related, the rays of dorsal and anal are 
nearly all soft and articulate. 

The Velijeridce of Japan, resemble superficially the Pteraclidce, but 
have a very different mouth-structure, and the posterior part of the 
dorsal and anal is made up of soft rays, as in related families. 

The species here discussed may be recapitulated. 



334 Annals of the Carnegie Museum. 

Family PTERACLID^. 

Genus Bentenia Jordan & Snyder. 

1. Bentenia crsticola Jordan & Snyder; Japan Current. 

2. Bentenia papilio (Lowe); off Madeira. 

Genus Pteraclis Gronow. 

1. Pteraclis trichiptenis (Cuv. & Val.); South Seas (?). 

2. Pteraclis carolinus (Cuv. & Val.); Gulf Stream. 

3. Pteraclis velifera (Pallas); Indian Ocean. 

4. Pteraclis ocellatiis (Cuv. & Val.); off Mozambique. 

Genus Centropholis Hilgendorf 

1. Centropholis petcrsi Hilgendorf; Japan Current. 

2. Centropholis goodei Jordan; Gulf Stream. 

Genus Pterycombus Fries. 
7. Pterycombus brama Fries; Arctic seas of Europe. 

Family ELEPHENORID.E. 
Genus Elephenor Jordan. 
7. Elephenor mucropus (Bcllotti) ; Japan Current. 

Family DIANID.E. 
Genus Diana Risso. 
7. Diana seniilnnata Risso; South Coast of France. 

Family CARISTIID.^E. 
Genus Caristius Gill & Smith. 
7. Caristius japoniciis Gill & Smith; Shikoku, Japan. 

Genus Platybervx Zugmayer 
7. Platyberyx opalesccns Zugmayer; off Gibraltar. 



336 Annals of the Carnegie Museum. 



Explanation of Plate LIV. 

Upper Figure: Caristiits japonicus Gill & Smith. After Smith & Pope, Proc. 
U. S. N. M., Vol. XXXI, p. 491. (Reduced.) 

Lower Figure: Elephenor macropiis (Bellotti) Jordan. Reproduced from Figure 
of Caristius macropiis (Bellotti) Memoirs Carnegie Museum, Vol. \"I, PI. XX\'III. 
(Reduced.) 



ANNALS CARNEGIE MUSEUM, Vol. XII 



Plate LIV. 








u.^ 



^:^^- 



y-i it 



p: 



\x 



vv 




(For explanation see opposite page.) 



338 Annals of the Carnegie Museum. 



Explanation of Plate LV. 

Upper Figure: Platyberyx opalescens Zugmayer. Reproduced from Result. 
Campagnes Scient. Albert I, Prince de Monaco, Fasc. XXXV, PL V, Fig. 5. (Re- 
duced.) 

Lower Figure: Pteraclis velifera (Pallas) . After Bonnaterre, Encyc. Methodique, 
a copy of the figure given by Pallas, Spicil., viii, p. 19, PI. 3, fig. 7. (Reduced.) 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate LV. 




(For explanation see opposite page.) 




^ ^ 









340 Annals of the Carnegie Museum. 



Explanation of Plate LVII. 

Upper Figure: Pterycombus brama Fries. Reproduced from Proc. Biol. Soc. 
Washington, Vol. XXV. PI. II. (Reduced.) 

Lower Figure: Pteraclis trichipterus Cuv. & Val. After figure in Cuvier's Animal 
Kingdom (translated by Griffith), Vol. X, 1834, PI. 32, fig. 2. (Reduced.) 



i 



1 



ANNALS CARNEGIE MUSEUM, Voi XII. 



Plate LVII 




(For explanation see opposite page.) 



342 



Annals of the Carnegie Museum. 



Explanation of Plate LVIII. 

Upper Figure: Pteraclis ocellatus Cuv. & Val. After Cuvier & Valenciennes, 
Hist. Nat. Poiss., Vol. IX, p. 363, PI. 271. (Reduced.) 

Lower Figure: Centropholis goodei Jordan, nom. nov. Not = Pteraclis carolinus 
Cuv. & Val., Hist. Nat. Poiss., Vol. IX, p. 368. Reproduced from "Oceanic 
Ichthyology," by Good & Bean, PI. LIX, fig. 218. 



i 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate LVIII 



V6C 



■'////////■y^^y 



"^s^^?- 



vyyy 



^t^^r 




(For explanation see opposite page) 




V. A DESCRIPTION OF CYPRIPEDIUM PASSERIXUM 
By Otto E. Jennings. 

(Plate LIX.) 

Cypripedium passeriniim was originally described by Richardson 
from specimens collected in northern Canada by the Sir John Franklin 
Expedition, 1819 to 1822. (Narrative of a Journey to the Shores of 
the Polar Sea, in the Years 1819, 20, 21, and 22, by John Franklin, 
p. 762, second edition.) 

A number of good herbarium specimens of this rare Cypripedium 
were collected in flower near Moose Factory, northwestern Ontario, 
July 2-3, igo8, by Mr. M. A. Carriker, Jr., and in view of the fact 
that this species is as yet but sparsely represented in American her- 
baria, and that the published descriptions and figures are somewhat 
at variance with respect to certain characters, it seems desirable to 
here present a detailed description, with figures. That our material 
is quite typical of the species as described by Richardson is evidenced 
by the closeness with which the original description applies to our 
specimens and it may be further mentioned that the Sir John Franklin 
Expedition made collections at Moose Factory, the locality at which 
our specimens were obtained. 

Description of Species. 
Cypripedium passerinum Richardson. Plants 2-3.5 dm. high, 
erect from an ascending base, which often consists of the basal portion 
of the stems of the two previous years. Stems leafy, rather densely 
puberulent, bearing from 3 to 5 leaves and a leaf-like floral bract of 
variable size, the lower part of the stem being enveloped in from 2 to 
4 sheathing, more or less bract-like, leaves. Leaves 7-15 cm. long, 
2.5-4 cm. wide, lanceolate-ovate, with 5 to 7 prominent nerves, leaf- 
surfaces sparsely puberulent, apex acute, base gradually narrowed 
and completely sheathing the stem. Floral bract similar to the 
leaves, varying from 3-8 cm. long, lanceolate-ovate, acute to acumi- 
nate, narrowed to a sheathing base. Flower one, terminal, erect, 
general color pale yellow with the lip lined and spotted with violet- 

343 
I 



\ 



344 Annals of the Carnegie Museum. 

purple. Upper sepal orbicular, apiculate, with 3 to 5 prominent 
veins, about 1.5 x 1.3 cm. Lateral sepals united into a doubly api- 
culate structure, the apiculations being about 2.5 mm. apart, each 
half of the double sepal having a prominent mid-vein with forking 
and often anastomosing lateral veins, the double sepal orbicular, 
about 1.2 cm. each diameter. Lateral petals slightly longer than 
the sepals, about 1.6 cm. long, 3 mm. wide, linear elliptic, rounded at 
the apex, the base rounded on the upper side. Lip 15 cm. long by 
I cm. vertical diameter, general profile obovate, veins variously 
forking and anastomosing, puberulent, and more or less glandular 
below, exteriorly, and also laterally at the base inside. Midway on 
each side of the orifice a flap about 3 mm. long and widely obtuse is 
folded into the cavity of the lip; the orifice is circular, about i.i cm. 
in diameter. The ovary is about 2.5 cm. long, elliptic, tapering to a 
peduncle below, obtusely angled, minutely glandular-puberulent. 
Stigma with an orbicular and somewhat concave disk about 3 mm. 
in diameter; staminodium expanded into a flat ovate structure about 
6 mm. long and 4 mm. wide, arching or ascending over the stigma, 
obtuse or sub-cordate at the base and retuse or with a recurved tip 
at the apex. Stamens with flat, obtusely ovate appendages extending 
dorsally over the anthers, free from them, but of about the same length. 
Capsule erect, 2.5 cm. long, 1.5 cm. thick, obtuse at the apex, tapering 
into a slender peduncle 1.2 cm. in length, sometimes remaining at the 
apex of the defoliated stem during the next season. 

Carnegie Museum, 
April 17, 1919. 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate LIX. 




C'ypriprdiuiii pas.-^i-'riiiiiiii. 



A Description of Cypripedium Passerinum, 345 



Explanation of Plate LIX. 

A. Cypripedium passerinum Richardson. One half natural size. 

B. Upper sepal. Natural size. 

C. Double sepal. Natural size. 

D. Lateral petals. Natural size. 

E. Lip, split open lengthwise, and flap, f, bent upward. Natural size. 

F. Staminodium, ventral view. Twice natural size. 

G. Stigma, S, anthers, a. Figure at the left represents a dorsal view; at the 

right a ventral view, x, free tips of stamens. Twice nat. size. 
H. Ovary from which flower has been removed. Natural size. 
/. Capsule. Natural size. 



CHARLES ROCHESTER EASTMAN. 

Born January 5, 1868; Died September 27, 1918. 

(Plate LX.) 

Although not connected with the Carnegie Museum at the time 
of his death, Charles Rochester Eastman held more or less intimate 
relations with the institution for more than sixteen years, during 
three of which he was a member of its staff. 

Professor Eastman was born at Cedar Rapids, Iowa, on January 
5, 1868, the son of Austin V. and Mary (m.n. Scoville) Eastman. 
He was graduated B.A. at Harvard in 1890, and in the following 
year received the degree of M.A. In 1893 he married Caroline A. 
Clark, daughter of the late Alvan G. Clark, the famous maker of 
telescopes, and with his bride repaired to Munich to prosecute his 
studies under Professor Karl von Zittel, receiving his degree of 
Ph.D. in 1894. He then returned to Cambridge, Massachusetts, 
and was given a post in the Museum of Comparative Zoology. 
He there continued his studies upon the fossil fishes begun at 
Munich, and subsequently almost continuously devoted himself 
to this, his chosen field of research. Later he studied at Johns 
Hopkins, and also in Europe, whither he occasionally went. His 
studies as a specialist were utilized in researches upon material 
submitted to him by the Geological Survey of the United States, 
and of several of the individual States, and he taught geology and 
palaeontology in Harvard and Radcliffe Colleges. 

The writer of these lines first became personally acquainted with 
Dr. Eastman on the occasion of a visit which the latter paid to 
Pittsburgh for the purpose of examining our collections and urging 
upon Professor J. B. Hatcher the importance of endeavoring to make 
a collection of Palaeozoic fishes. At this time the fact that the 
collection belonging to Baron Ernst Bayet of Brussels was on the 
market was mentioned, and Dr. Eastman volunteered to obtain 
more information as to the collection than was in the possession 
of Professor J. B. Hatcher and myself at that time. It was finalh' 



ANNALS CARNEGIE MUSEUM, Vol. XII 



Plate LX. 




N^^^Uxv^C^CXV^t^ 




d>'6/C%^.<>4y^ ^ 



Obituary Notes. 347 

decided by Mr. Carnegie on cunfereiice with me lu acquire the 
Bayet Collection for the Carnegie Museum. Inasmuch as Pro- 
fessor J. B. Hatcher,* our Curator of Palaeontology, in the spring 
of 1903, was unable to go to Europe, the task of arranging for the 
safe transfer of the collection fell upon my shoulders, and at the 
urgent entreaty of Dr. Eastman I took him along to assist in the 
work. He was especially interested in the fossil fashes, and re- 
ceived permission to take a certain number of specimens with him 
to Paris for comparison with the types of the elder Agassiz, which 
are preserved in the National Museum of France, and later he 
returned them to the Carnegie Museum. It had been arranged 
by the writer and Professor Hatcher that the fossil fishes of the 
Bayet Collection should be assigned to Dr. Eastman for study and 
description. However, an unfortunate disagreement between 
Professor Hatcher and Dr. Eastman led to the abandonment of 
this plan, and it was not until a number of years after Professor 
Hatcher's death that the writer again decided to employ Dr. 
Eastman for the accomplishment of the task, for which he was 
eminently fitted. He came to Pittsburgh in the early summer of 
1 910, and thereafter for three years was allnost constantly occupied 
with the work of classifying and arranging the fossil fishes in our 
collection, at the same time holding a minor position in the Univer- 
sity of Pittsburgh. The results of his labors upon the material 
in this Museum are contained in the following papers from his pen: 

In the Memoirs — 

Vol. II, No. 3. "Fossil Avian Remains from Armissan," 8 pp., 

4 pis. 
" IV, No. 7. "Catalog of the Eocene Fishes from Monte 

Bolca in the Carnegie Museum," 66 pp., 

12 pis. 
" VI, No. 5. "Supplement to the Catalog of the Fishes from 

the Upper Eocene of Monte Bolca," 34 pp., 

6 pis. 
" " No. 6. " Catalog of Fossil Fishes from the Lithographic 

Stone of Cerin, France," 40 pp., 9 pis. 
" " No. 7. "Catalog of Fossil Fishes from the Lithographic 

Stone of Solenhofen, Bavaria," 35 pp., 

17 pis. 



348 Annals of the Carnegie IMuseum. 

In the Annals — 

Vol. \'. "A New Species of Helodus," 2 pp. 
" VIII. "Jurassic Saurian Remains Ingested in Fish," 6 pp., 

2 pis. 
" " "Tertiary Fish Remains from Spanish Guinea, 
West Africa," 9 pp., 2 pis. 
IX. "Notes on Triassic Fishes belonging to the Families 
Catopteridse and Semionotidee," 10 pp., 3 pis. 
" Dipterus Remains from the Upper Devonian of 
Colorado," 5 pp. 

Having completed the work for which he was employed, he 
repaired to Washington, where he was for a short time engaged in 
scientific work at the United States National Museum, and then 
went to the American IMuseum of Natural History in New York, 
undertaking the preparation of a Bibliography of Fishes, which 
that Museum was engaged in getting out, the first two volumes of 
which have already appeared. For this task he was extremely well 
qualified, because of his erudition and linguistic attainments. 

Professor Eastman's contributions to the literature of science 
number about one hundred titles, a number of these being notes of a 
bibliographic or semi-historical character. He will best be remem- 
bered by his translation in three volumes of Karl von Zittel's 
Text-Book of Paleontology, published by the Macmillans, by the 
Bibliography of the Fishes, issued by the American Museum of 
Natural History, and by his Catalogues of the Fishes of Monte 
Bolca and Solenhofen and other papers published by the Carnegie 
Museum. His systematic contributions to his favorite science 
consist of the definition of three new families, twelve new genera, 
and one hundred and fifteen new species of fossil fishes, many of 
which were first published by the Carnegie Museum. 

Professor Eastman early devoted himself in his studies to the 
investigation of the placoderm fishes of the Devonian, which have 
proved a puzzle to students ever since their existence became known. 
Probably no individual ever devoted more time to the study of 
these interesting forms than he. Wc have in the Carnegie Museum 
an interesting series of models and a mounted skull of one of the 
hugest of these fishes, to the preparation of which he devoted a great 
deal of time. 



Obituary Notes. 349 

In liis chosen field of research he was easily llie fureniosl anion^ 
his American contemporaries, and his great erudition and perfect 
familiarity with the literature of his subject caused him to be con- 
stanth- consulted b>- others who had not devoted themselves, as he 
had done, to the intricacies of that branch of science which was his 
special delight. 

Professor Eastman came to his end under tragic circumstances. 
He had been for some time employed by the War Trade Board in 
Washington, where his linguistic attainments made his services 
valuable. He suffered an attack of inlluenza, and, when convales- 
cent, repaired to Long Beach, N. J. On the evening of his arrival, 
September 27, 1918, although suffering, he went out to take the 
air upon the boardwalk. He apparently wandered far in the 
darkness, to a place where the walk had become dilapidated, and 
either fell from the walk or stumbled through an opening, was 
stunned, washed out by the tide and drowned, his body being 
found the next day. 

His premature end robbed science of one of its most industrious 
and indefatigable workers. W. J. H 

(For the portrait of Dr. Eastman accompanying this article we 
are indebted to the Editor and Secretary of the Geological Scoiety 
of America, who first published it in the Bulletin of that Society, 
March, 1919.) — Editor. 



350 Annals of the Carnegie Museum. 

ANDREW ARNOLD LAMBING. ■ 

Born February i, 1842; Died December 24, 1918. 

(Plate LXJ.) 

Among the names of those who were originally appointed by 
Mr. Carnegie as members of the Board of Trustees of the Carnegie 
Institute was that of the Reverend Father Andrew Arnold Lambing, 
and he continued to hold an honored place on the Board from the 
time of his appointment until the day of his death. During a 
portion of this time he served as a member of the Committee upon 
the Museum, and during almost the whole of the time he was the 
Honorary Curator of the Historical Collections of the Museum. 
Andrew Arnold Lambing was born at Manorville, Armstrong 
County, Pennsylvania, on February i, 1842. He was a descendent 
of Christopher Lambing, who migrated from Germany, and settled 
in Bucks County, Pennsylvania, in 1749, dying there in 18 17, at 
the age of ninety-nine years. His son Matthew married and 
settled in New Oxford, Adams Co., Pa., where the father of Andrew 
Arnold Lambing, Michael A. Lambing, was born, in 1806. In 1823 
the family came across the mountains to Armstrong County, 
where Michael A. Lambing married Annie Shields on December i, 
1837. There were born of this union five sons and four daughters, 
of whom the subject of this brief sketch was the third son and child. 
In his boyhood and early youth he labored on his father's farm, 
and also was employed in a brickyard. He worked for a while as a 
day-laborer helping in the grading of the Pennsylvania Railroad. 
In the winter months he attended the country schools. He 
acquired a taste for reading, and became profoundly interested 
in matters relating to the local history of the region in which he 
lived. His native ability was presently recognized by those who 
knew him, and ways and means were found for obtaining an educa- 
tion, qualifying him for the priesthood in the church of his fathers. 
After his ordination he held a number of appointments, the first 
being as a teacher in St. Francis' College, Loretto. He then 
served as the Priest of St. Patrick's Church, Cameron Bottom, 
Indiana Count}' ; then ministered at St. Mary's Church, Kittan 



ANNALS CARNEGIE MUSEUM, Vol. XI 



Plate LXI. 




MoNSiGNOR Andrew Arnold Lambing, D.D. 



Obituary Notes. 351 

ning, and the church at Freeport, which were jointly under his care. 
He later became chaplain at St. Paul's Orphan Asylum, Pittsburgh, 
and then on January 7, 1874 took charge of St. Mary's of Mercy 
congregation, Pittsburgh. He remained tlierc for some time, 
when he was transferred to St. Jaines Church in Wilkinsburg, 
where he served for more than a third of a century, the congre- 
gation growing under his care to be one of the largest in western 
Pennsylvania. 

In 1884 he undertook the publication of a quarterly magazine 
called Catholic Historical Researches, the first of its kind devoted 
to the history of the Catholic Church in this country. It is now 
continued in Philadelphia as a monthly. For some years after- 
ward he was an active member of "The Old Settlers' Association," 
a society out of which grew "The Historical Society of Western 
Pennsylvania," of which for some years he was the President. 
In 1915 he was honored by the Pope with the title of " Monsignor." 

Father Lambing published a number of important historical 
papers, one of the most excellent of which is the "Baptismal 
Register of Fort Duquesne," by Frere Denys Baron, to which 
he prefixed an historical account, and which he copiously annotated. 
He also published "A History of the Catholic Church in the 
rJioceses of Pittsburgh and Allegheny." He was a constant con- 
tributor to the pages of church periodicals, and was greatly in 
demand on anniversary occasions, where he frequently made 
addresses replete with historical information. 

The writer of these lines recalls with gratitude Father Lambing's 
faithful discharge of his duties as a member of the Committee 
upon the Museum, and later as Honorary Curator of the Historical 
Collections. In quite recent years, owing to age and growing 
infirmities, he was not able to do much, but still faithfully attended 
the meetings of the Board of Trustees, and whenever information 
of an historic character was called for, he was ready to put the stores 
of his knowledge at the service of his associates in the Museum. 

Personally Father Lambing was a most delightful companion, 
abounding in good nature and wit. He was a man of very vener- 
able appearance, and one who attracted the immediate attention 
of those who saw him. His tales of the early vicissitudes through 
which he passed when western Pennslyvania was still regarded in a 



352 



Annals of the Carnegie Museum. 



measure as on the frontier, were replete with interest and with 
humor. He was greatly beloved, not only by his parishioners, but 
by a wide circle of friends outside of his own immediate ecclesias- 
tical relationship. In his death the church to which he belonged 
lost one of its most beloved parish-priests and ablest men, and the 
entire city of Pittsburgh experienced a sense of genuine bereave- 
ment. 

He died on December 24, 1918, in the seventy-seventh year of 
his age. W. J. H. 



ANNALS CARNEGIE MUSEUM, Vol, XII 



Plate LXIi. 




Herbert Huntingdon Smith, 

(OBIIT a. D. 1918, ANNO /ETATIS SU/E LXVII.) 



Ohituary Notes. 353 

HERBERT HUNTINGTON SMITH.i 

Born January 21, 1S51; died March 22, 1919. 

(Plate LXII.) 

The wide circle of his friends and acquaintances were shocked to 
read in the daily journals that on March 22 Mr. Herbert Hunting- 
ton Smith, the curator of the Alabama Museum, had been killed 
by being run over by a freight train. In recent years he had be- 
come very deaf, and it was owing to this infirmity that he came to 
his untimely end. Once before, in the city of Pittsburgh, he had 
been struck by an electric car, the approach of which he had not 
observed, but fortunately escaped at that time with only a few 
bruises. 

A number of years ago Lord Walsingham in an address before 
the Entomological Society of London in speaking of the work of 
field naturalists and the additions made by them to the sum of 
human knowledge, made the statement that the two ablest collectors 
were Americans, one of them the late William H. Doherty, the 
other Herbert Huntington Smith. With both of these men the 
writer of these lines was intimately associated, both of them having 
made extensive collections for him in foreign parts, and both came 
to their end under tragic circumstances. Doherty died in Uganda, 
as the result of nervous prostration brought about partly by ex- 
posure, partly by the fact that his camp was haunted by man-eating 
lions, which had killed several of his assistants. Smith passed 
away in the midst of important activities, as the result of a horrible 
accident. 

My acquaintance with Mr. Herbert Huntington Smith, which 
has covered nearly thirty years of his life, enables me to speak of him 
with an appreciation founded upon intimate knowledge. 

He was born at Manlius, New York, on January 21, 1851. 
He studied at Cornell University from 1868 to 1872. In 1870 he 
accompanied his friend and teacher, the late Professor C. F. Hartt, 

' This article is a reproduction of the substance of biographies by the writer 
pubh'shed in Science, N.S., Vol. XLIX, No. 1273, pp. 481-483, and in the 
Enlotnological News, Vol. XXX, pp. 21 1-2 14. 



354 Annals of the Carnegie Museum. 

on an excursion to the Amazons. He thus caught his first glimpse 
of tropical life, which wove about him a spell, which always there- 
after bound him. 

In 1874 he returned to Brazil for the purpose of collecting and 
studying the fauna of the Amazonian regions. Two years were 
spent in the neighborhood of Santarem, and subsequently he passed 
a year in explorations upon the northern tributaries of the Amazons 
and the Tapajos, after which he stayed about four nionths in Rio 
de Janeiro. Returning to the United States he was commissioned 
by the Messrs. Scribner to write a series of articles upon Brazil for 
their magazine, and accordingly made two more trips to that 
country, studying -the industries, social and political conditions, 
and investigating the famine district in Ceara. On one of these 
journeys he was accompanied by Mr. J. Wells Champney, who 
was employed to prepare illustrations for his articles. One of the 
results of these journeys was the volume entitled "Brazil, the 
Amazons, and the Coast," which was issued by Charles Scribner's 
Sons in 1879. On October 5, 1880, Mr. Smith married Miss 
Amelia Woolworth Smith, of Brooklyn, New York. She entered 
with zest into his labors, and in all the years which followed was 
his devoted and most capable assistant. There was a remarkable 
accord in their tastes and Mrs. Smith developed unusual skill and 
efficiency in the manipulative processes involved in collecting 
specimens of natural history. Indeed, it is no exaggeration to say 
that her learned husband would not have been able to accomplish 
the vast amount of work, which was achieved in later years, had 
it not been for her facile fingers. She became an accomplished 
taxidermist, and was able to prepare the skins of birds and preserve 
insects, in the most approved manner. Mr. Smith and his wife 
spent the years from 1881 to 1886 in Brazil. He made his general 
headquarters in Rio de Janeiro, where he received miich encourage- 
ment from the Emperor, Dom Pedro II., who was deeply interested 
in scientific research. During these years he traveled extensively 
and spent a long time in exploring the then little known territory 
along the upper waters of the Rio Paraguay and the Rio Guapore 
on the western confines of Brazil, in the vicinit}- of Matto Grosso 
and Chapada. The extensive series of specimens which he gathered 
during these years of fruitful collecting were accpiired partly by the 



Obituary Notes. 



355 



Xalional Museum in Rio dc Janeiro, parll\- by Mr. F. D. (Jodin.iu 
of London, and i)artl\- In- the writer of these lines, who siibsequenth 
purchased most of the lepidoi)tera, and, at a later date by the 
Carnegie Museum, which secured most of the vast collection of 
other insects, which Mr. Smith had made, numbering; a])proximalely 
thirty thousand species and not far from two hundred thousand 
specimens. 

In 1886 there appeared in Portuguese from his pen an account 
of some of his travels, entitled " De Rio de Janeiro a C.uyaba." 
Mr. F. D. , Godman, whose monumental work, the " Biologia 
Centrali-Americana," called for an intensive study of the fauna of 
Mexico, commissioned Mr. Smith to make collections for him in 
that country, and he labored there during the year 1889. He spent 
much of his time in the years 1890-1895 in the employment of the 
\\'est Indian Commission of the Royal Society in making collections 
in Trinidad and the Windward Islands, and in reporting upon the 
same. These collections are in the British Museum. During the 
same years he was actively engaged as one of the staff of writers 
employed in the preparation of the "Century Dictionary," the 
"Century Cyclopedia of Names," and "Johnson's Cyclopedia." 
In these works almost everything relating to South and Central 
America and the fauna and flora of these lands is from his pen. 

When plans were being formed for the development of the Car- 
negie Museum, Mr. Smith took occasion, not only in letters, but 
by personal visits to the writer, to urge the desirability of selecting 
as one of the major objects of the new institution, a biological 
survey of South America. While it was not at that time possible 
to fully accept his proposals, one of the results of his visits to Pitts- 
burgh, was his employment by the infant museum to act in a 
curatorial capacity, devoting himself to the formation of collections 
illustrating the natural resources of the region of which Pittsburgh 
is the metropolis. Assisted by his wife and various volunteers 
he made extensive collections representing the flora and fauna of 
western Pennsylvania and West Virginia. These collections num- 
ber many tens of thousands of insects, shells, and plants, as well as 
fishes, reptiles, birds, and small mammalia. When not in the field, 
he devoted his time to the arrangement of collections which began 
to rapidly come into the possession of the museum. 






356 Annals of the Carnegie Museum. 

He was not, however, entirely happy in the confinement of the 
walls of a museum. He constantly heard "the call of the wild," 
and his heart longed for the life of the tropics, in w^hich he had 
passed so many happy years. He proposed to the authorities of 
the Carnegie Museum that he should be allowed to go to the United 
States of Colombia to make collections. The writer agreed himself 
to become the purchaser of the collections of lepidoptera which 
might be made, the Carnegie Museum agreed to purchase the birds, 
a set of the mammals, the ethnological material which might be 
gathered, and to take one or more sets of the botanical specimens 
collected. Accompanied by his wife and young son he set out for 
Colombia to begin his work in the Province of Santa Marta. One 
of the chronic revolutions of that period developed and he encoun- 
tered much difficulty. The period from the fall of 1898 to the 
spring of 1902 was spent in this work. It was a period of trial and 
hardship. Mr. Smith finally fell ill and it was feared that he would 
not recover. When at last he was pronounced to be out of danger, 
the party hastened to return to the United States, and thenceforth 
all thought of further investigations in the tropics was abandoned. 
The collections made in the face of hardship and disease were 
nevertheless large and valuable and contained many species wholly 
new to science. 

Mr. Smith and his wife on their return resumed their employ- 
ment at the Carnegie Museum, devoting themselves to the arrange- 
ment of the Colombian material and to the classification of the 
large and increasing collections of moUusca belonging to the 
museum. One of the results- of this period is the "Catalog of the 
Genus Partula," which was published in 1902. After about a year 
in Pittsburgh Mr. and Mrs. Smith felt the need of a change and 
resolved upon removal to Wetumpka, Ala., where they began the 
systematic collection of fresh-water shells, belonging to the family 
Strepomatidcr, which abound in the Coosa and other rivers of 
that region. They were supported in their work by four ardent 
conchologists: Mr. George H. Clapp, of Pittsburgh, Messrs. 
John B. Henderson and T. H. Aldrich, of Washington, D. C, and 
Mr. Bryant Walker, of Detroit, Mich., who formed a "syndicate" 
to enable the work to be done. When Mr. Aldrich dropped out of 
their number. Professor H. A. Pilsbry, of the Academy of Natural 



Obituary Notes. 357 

Sciences in Philadelphia, took the vacant place for such time as he 
was able to command the necessary funds. In 1910 Dr. Eugene 
A. Smith, of the Geological Survey of Alabama, induced Mr. and 
Mrs. H. H. Smith to take charge of the museum at the University 
of Alabama, and here they have been engaged for nearly a decade 
in arranging and caring for the collections which have been accumu- 
lated principally by the Geological Survey of Alabama. For the 
past two or three years the Alabama Museum and the Carnegie 
Museum have been working conjointly in the exploration of the 
Tertiary deposits of Alabama, under the oversight of Mr. Smith, 
and the result has been the discovery of a number of new and rich 
deposits of Tertiary mollusca. Vast series of specimens had been 
gathered by our indefatigable friends, and the last letter received 
by the writer contained a request for a fresh supply of labels. It 
was written only a day or two before the lamented death of the 
sender. 

The work done by Mr. Smith in the field of entomology was 
particularly great. The entomological collections made by him 
are mainly contained in the National Museum at Rio de Janeiro, 
in the British Museum (derived from the gift of the collections of 
F. D. Godman) and in the Carnegie Museum, though parts of his 
collections are scattered widely in other museums. There are in 
the Carnegie Museum in the neighborhood of 25,000 species of 
Brazilian Coleoptera assembled by him and many thousands of 
species of insects in other orders. A memorandum recently re- 
ceived by the writer from Mrs. Smith states that the Arthropoda 
collected during the years of Mr. Smith's journeys in Brazil up to 
May, 1886, aggregated approximately 40,000 species, distributed 
as follows: 

Hymenoptera 5,000 

Diptera 2,500 

Lepidoptera 2,600 

Coleoptera 23,000 

Hemiptera 3.300 

Orthoptera 600 

Neuroptera 300 

Arachnida 2,000 

Crustacea 250 

Total 39.550 species 



358 Annals of the Carnegie Museum. 

The collections contained an aggregate of at least half a million of 
individual specimens. Portions of the collections have been care- 
tully studied and reported upon. Ashmead, Cresson, and others 
ha^•e in part worked over the H>inenoptera. Williston did some- 
thing with part of the Diptera. The Lepidoptera so far as they 
represented the species of Middle-America were studied by God- 
man and by Herbert Druce. Champion wrote up a part of the 
Coleoptera, but the beetles of Brazil as a whole remain for the most 
part to be studied; P. R. Uhler has described man}- of the Hemip- 
tera, as did also W. L. Distant. The Orthoptera have been 
studied by Dr. Lawrence Bruner and the Odonata by Dr. P. P. 
Calvert. 

Mr. Smith was not a mere collector of natural history specimens. 
He was a naturalist in the true sense of that much abused word. 
He had a wide and accurate knowledge of the major divisions of 
the animal kingdom and keen powers of discrimination. He was 
especially well versed in conchology, though he wrote and published 
but little. He was a systematist of far more than ordinary ability, 
whose opinions were received with great respect by those who 
employed him. He was an accomplished linguist. He was 
familiar with the Greek and Latin classics, spoke Spanish readily 
and used Portuguese as if it were his mother tongue. He also had a 
good knowledge of French and German, sul^cient to enable him to 
consult works in those languages. He was one of the survivors of a 
group of naturalist explorers and investigators to whom we are 
indebted for much of our knowledge of the fauna and flora of trop- 
ical America. He belonged to an illustrious company which, 
beginning with Humboldt and Bonpland, included in its ranks 
such men as Alfred Russel Wallace, Henry W. Bates, J. N. Natterer, 
J. J. Tschudi, J. B. Hatcher and J. D. Haseman, who courageously 
faced dangers in the wilderness in order to secure information at 
first hand as to the fauna and flora of the great continent where 
they labored. \V. J. H. 



ANNALS CARNEGIE MUSEUM, Vol. XII. 



Plate LXII 




Ht NRY JOHN Heinz, 
(Born Oct. ii, 1844; Died May u, 1919.) 



Obituary Notes. 359 

HENRY JOHN HEINZ. 

Born October ii, 1844; Died May 14, 1919. 

(Plate LXIII.) 

The death of Mr. Henry John Heinz which occurred on May 14 
1919, robbed Pittsburgh of one of its most successful men of affairs, 
who had won an international reputation, not only in the broad 
fields of manufacture and commerce, but as a lover of men and as a 
doer of good deeds. 

Like the vast majority of Americans who have achieved fortune 
and fame, he began his career under circumstances which the 
thoughtless often speak of as "adverse," but which for earnest 
souls provide the best stimulus for endeavor. Mr. Heinz was born 
in Pittsburgh, October li, 1844, the eldest child of Henry and 
Anna Margaretta {m.n. Schmitt) Heinz. His parents had recently 
come from Germany. His father was a man of industry, who did 
not always find the battle of life easy. His mother possessed much 
force, and was an earnest and faithful Christian woman, who at- 
tended well in the sphere of the household to her domestic duties. 
From their infancy onward she endeavored to fill the minds of her 
children with devotion to those things which are "pure" and 
"lovely" and of "good report." She and her husband had dedi- 
cated their son to the Christian ministry, and with this end in view 
they exerted themselves to secure for him a good education. He 
made excellent progress in the schools to which he was sent; but 
there came a time in his youth when it became plain to the family 
that their united efforts would be required to meet the battle of 
life in the field where they found themselves, and the lad, full of 
devotion to his parents, laid aside the ambitions which they had 
fostered in him, and manfully addressed himself to the task of 
helping to win the daily bread which was needed. 

When he was but a little child the family had removed to Sharps- 
burg, a suburb of Pittsburgh. Sharpsburg was then but a strag- 
gling village. The father there continued his occupation of manu- 
facturing building-brick and added to it the erection of houses. 
In this enterprise the son, when he became older, assisted him. 



360 Annals of the Carnegie Museum. 

The family home was situated on a plot of ground containing some 
three acres of fertile soil. The boy, in his teens, resolved to culti- 
vate a portion of this land as a garden, and finding that it produced 
more than was necessary for the consumption of the family, con- 
ceived the idea of preserving the surplus and marketing it in the 
neighborhood. His mother, who was a mistress of the culinary 
arts, aided him, and the products of the garden, converted into 
sauces, pickles, and condiments, put up in neat form, and most 
savory, found ready purchasers. The business begun under the 
family roof was so successful that the young manufacturer resolved 
with the profits of his early sales to enlarge the enterprise, and after 
attaining his majority associated with himself a partner, and 
removing to Pittsburgh, began the conservation of pure food pro- 
ducts on a liberal scale. He experienced many difficulties, but, 
nothing daunted, he went forward, until to-day the establishment 
which bears his name has come to be one of the greatest, if not the 
greatest, in the entire world. The manufacturing plant in Pitts- 
burgh at the time of his death covered thirty acres, to which are 
to be added other manufacturing plants in the United States, 
Canada, Great Britain, and Spain, covering an area of seventy 
acres of buildings equipped with all modern appliances; forty thou- 
sand acres of land under cultivation; tens of thousands of people 
in his employment, with four hundred traveling salesmen going 
forth from forty-five distributing centers located in the leading 
cities of the world. The viands prepared by skilful hands in his 
vast establishments are found on the tables of the poor and the 
rich alike in every clime. His motto from the beginning of his 
career to its end, as he often told the writer, was "to do common 
things uncommonly well." 

But it is not as a preeminently successful manufacturer and 
merchant-prince that Mr. Heinz claims our chief attention. He 
won and held the regard and friendship of men not so much because 
of his success in business affairs, as because of his loving kindness 
to all about him and his enthusiasm for the higher things which 
adorn life. In his youth, though the dream of entering the ministrx' 
was not to be fulfilled, he found a sphere of kindred usefulness 
in the Sunday-school of the little church which he and his parents 
attended, and where he became the Superintendent. His interest 



Obituary Notes. 30 1 

in tlic work of Sunday schools lasted throughout life. For many 
years he was the President of the Pennsylvania Sabbath School 
Association, and Chairman of the Executive Committee of the 
World's Sunday School Association. In 1916 he was elected 
\'ice-President of the Sunday School Union of London, the oldest 
organization of its kind in existence. He was looking forward at 
the time of his death with great expectancy to the International 
Convention of the Friends of Sunday-Schools, which is to take 
place in 1920 in the city of Tokyo, and was, deeply occupied. in 
making preliminary plans for this great gathering. No more faith- 
ful and helpful friend to the cause of Sabbath-schools has lived in 
our Commonwealth. 

He was also intensely interested in the cause of higher education. 
He was largely instrumental in founding, and was one of the chief 
supporters of the University of Kansas City, and was the President 
of the Board of Trustees of that institution. He was a faithful 
friend of the University of Pittsburgh and did much for it during 
his lifetime, and by his last will and testament gave the University 
two hundred and fifty thousand dollars for the purpose of providing 
a building for the social and religious activities of the student body. 
He was active in almost all movements looking towards the welfare 
of his native city, and participated in all benevolent and public- 
spirited enterprises which arose during the last four decades. 
No good cause ever appealed to him in vain for help. 

Mr. Heinz traveled very extensively in the later years of his life, 
not only in America, but also in Europe, Asia, and North Africa. 
He made it a habit in the early days of his wanderings to and fro to 
bring back with him souvenirs of the places which he had visited, 
selecting them for their beauty, or for some association with the spot. 
He thus was led' gradually to become a collector. The first number 
of the present volume of the Annals of the Carnegie Museum is a 
catalog of the collection of watches which he made. He had become 
interested in the story of the evolution of this form of timepiece, 
and the beautiful assemblage of specimens which he made was placed 
by him in the Carnegie Museum, that others might enjoy what he 
had found so much pleasure in gathering together and studying. 
He became interested in the art of the ivory-carver, and the ivory- 
carvings, principally Japanese and older Chinese, which he accumu- 



362 Annals of the Carnegie Museum. 

lated, is one of the most extensive collections of its kind in existence. 
In recent years he took up the study of antique Chinese jades, and 
amassed a very great and beautiful collection, which is in his 
palatial home, the latter having during the last two decades 
become a storehouse of beautiful and interesting objects which he 
had gathered. He not only had the instincts of a collector, but 
he desired that others should share with him the pleasure which 
he derived from his pursuits, and he converted a group of buildings 
in the rear of his home into a museum, to which access was freely 
given at stated times to the entire community. The beautiful 
conservatory which stood nearby was also thrown open to the 
public, that his neighbors and the people in the vicinity might enjoy 
the sight of the plants and flowers which grew and bloomed there. 
For a number of years Mr. Heinz proved himself deeply inter- 
ested in the work of the Carnegie Museum, established by his 
fellow townsman and friend as a department of the Carnegie 
Institute. He accepted the position of Honorary Curator of 
Ivory-Carvings and of Textiles in the Museum, and spent as much 
time as he could command in the midst of his multifarious occupa- 
tions, in arranging and looking after the collections which he had 
either donated to the Museum or deposited there as loans. Many 
delightful hours were spent by the writer in his company. Though 
loaded down with cares and responsibilities, he always was filled 
with a certain spirit of gaiety and mirth, which made him a most 
attractive companion. His removal from our midst has subtracted 
from the sunshine of life, and yet, 'though dead he still speaks,' 
through his works, which follow him, and which are destined for 
years to come to be a blessing and source of instruction, of mental 
and spiritual advancement to his fellow-men. All honor to his 
memory. W. J. H. 



INDEX 



abacopus, Glossogobius, 171, 303, 306, 

313 
aberrans, Parahyus, 82 
Acanthogobius, 307, 313 

ommaturus, 307, 313 
Achaenodon insolens, 79,80 
robustus, 80, 81 
uintense, 80, 81, 82 
Achsenodontidae, 79 
Acheilognathus, 198, 231, 311 

himantegus, 231, 311 
Acrossocheilus gen. nov., 170, 197, 206, 
310 
formosanus, 206, 310 
acuticeps, Parasalanx, 171, 173, 174, 

309 
adiposalis, Pseudobagrus, sp. nov., 

170, 178, 181, 309 
advenum, Amynodon, 131 
aesticola, Bentenia, 329-334 
affinis, Gambiisia, 170, 257, 312 
agilis, Tritemnodon, 47 
agrarius, Hyraqhyus, 128, 130 
agrestis, Colonoceras, 128 
Agriochoeridae, 82 
Agriochoerus, 77 

pumilus, 85 
alba, Fluta, 259, 312 
altivelis, Plecoglossus, 172 
amarorum, Eomoropus, 139, 140 
Amnion, Mrs. Samuel A.: collection of 

coins and medals, 37 
Amynodon, 130 

advenum, 131 

intermedium, 131 
Amynodontinse, 130 
Anguilla, 261, 312 

japonica, 262, 263, 312 

mauritiana, 262, 312 

sinensis, 171, 262, 267, 312 
Anguillidae, 261, 312 
anguillicaudatus, Misgurnus, 187, 309 
annectens, Eomoropus, 139, 140 

Hylomer^^x, 68 

Isectolophus, 103, 116, 117, 120 

Leptolophiodon, 116, 126 

Protagriochaerus, 76, 86, 87 

Protylopus, 91, 92, 99 
Anoplotheriidae, 76 
Aoki, Takeo, 170 



aokii, Culter, sp. nov., 170, 250 
ariakensis, Parasalanx, 170, 173, 174 
Aristichthys gen. nov., 170, 198, 246, 
311 
nobilis, 170, 246, 311 
Artiodactyla, 66 
asotus, Parasilurus, 176 
auratus, Carassius, 199, 309 

barbata, Candidia, 243,311 
Barbodes, 197, 213, 310 
paradoxus, 213, 310 
Battersea enamel, 19 
Bentenia, 329-334 
aesticola, 329-334 
papilio, I.e. 
Bibliography' of Formosan Fresh- 
water Fishes, 319-328 
boops, Helaletes, 103, 104, 105, 109, 

III 
Bovidae, 90 

brama, Ptcrycombus, 329-334 
Breguet, Abraham Louis, French watch- 
maker, 30 
brevianalis, Pseudobagrus, 178, 309 
brevicauda, Culter, 251 
brevirostris, Pseudogobio, 221, 310 
brunneus, Glossogobius, 302, 303, 313 
BunomerjTc elegans, 67, 73, 74 

montanus, 72 
Butis, 170, 287, 291, 313 
butis, 170, 29^, 313 
butis, Butis, 170, 292, 313 

Caenopus mite, 128 

platycephalum, 128 
Camelidae, 88 
Camelomeryx, 98 
Candidia, 198, 242, 311 

barbata, 243, 311 
candidius, Rhinogobius, 295, 313 
caninus, Rhinogobius, 295, 301, 313 
Capoeta, 197, 214, 310 

semifasciolata, 170, 214, 310 
Carassius, 197, 199, 309 

auratus, 199, 309 
carinatus, Mugil, 170, 268, 273, 312 
Caristiidae, 329-334 
Caristius, 329-334 

japonicus, 329-334 



363 



364 



Index. 



carolinus, Pteraclis, 332, 334 
Catablepas gnu, 90 

Catalog of the Collection of Watches 
Belonging to Mr. H. J. Heinz of 
Pittsburgh and Deposited by him 
in the Carnegie Museum. By Doug- 
las Stewart, W. J. Holland, and 
A. S. Coggeshall, 4-32 
Centropholis, 329-334 

goodei, 329-334 

petersi, 329-334 
cephalus. Mugil, 170, 268, 312 
Cercoleptes, 49, 50 
Chalicotheroidea, 139 
Channa, 285, 313 

formosana, 285, 313 
Chanodichthys, 199, 248, 311 

macrops, 248, 311 
Chinese art, Objects of, 2 
Cirrhina sp.? 255 
Clarias, 175, 184, 309 

fuscus, 185, 309 
clavis, Crocodilus, 41 
Cobitidae, 187, 309 
Cobitis, 191, 309 

taenia, 170, 192, 309 
Collections added to the Carnegie 

Museum, 35-38 
Colodon longipes, 115 
Colonoceras, 113 

agrestis, 128 
compressidens, Paramys, 60 
Crocodilia of the Eocene of the Uinta 

Basin, 41 
Ctenopharyngodon, 198, 229, 311 

idellus, 229,311 
cubitalis, Triplopus, 132, 137 
culbertsoni, Merycoidodon, 86, 87 
Culter, 199, 249, 311 

aokii sp. nov., 170, 250, 311 

brevicauda, 251, 311 
Cultriculus^en. nov., 170, 199, 252, 312 

kneri, 170, 253, 312 
cuspidens, Schizolophodon, 122 
Cyprinidae, 197, 309 
Cyprinus, 197, 201, 310 

carpio, 201, 310 
Cypripedium passerinum, A Descrip- 
tion of, By O. E. Jennings, 343-345 

Darlington, Wm. M., 37 
decemcirrosus, Misgurnus, 189, 309 
delicatus, Paramys, 61 
Desmatotherium guyotii, 103, 127 
Diana, 329-334 

semilunata, 329-334 
Dianida?, 329-334 
Dillonia sp.?, 255 



Dilophodon minusculus, 103, 104, 

113 
Diplobune quercyi, 76, 77, 78 
Diplobunops matthewi, 76 
Diplobunopsinae, 79 
Distoechodon, 198, 227, 311 

tumirostris, 170, 227, 311 
douglassi, Limnocyon, 45 
DuPuy, Herbert, Collection of, 3 
dysclerus, Oxyaenodon, 42, 43, 44 
dysodus, Oxyaenodon, 42, 43 

Eastman, Charles Rochester, Death, 

35; Obituary, 347-350 
Editorial Notes. By W. J. Holland, 

1-3; 33-39 
elegans, Bunomer>Tc, 67, 73. 
Eleotris, 287, 313 

fusca, 288, 289, 313 

oxycephala, 288, 313 
Elephenor, a New Genus of Fishes from 
Japan. By D. Starr Jordan, 
329-342 

macropus, 329-334 
Elephenoridae, 329-334 
Eomoropus amarorum, 139, 140 

annectens, 139, 140 
Eotylopus, 92 
Epihippus gracilis, loi 

parvus, 102 

uintensis, loi 
Equidae, iii 
evansi, Leptomeryx, 94, 95 

Ferae, 57 

filamentosus, Macropodus, sp. nov., 

170, 278, 312 
Fishes and reptiles of the Uinta Eocene, 

40 
Fishes of Formosa, 169-328 
Fluta, 259, 312 

alba, 259, 312 
Formosa, Island of, 169 
formosana, Channa, 285, 313 
Formosania gen. nov., 170, 193, 194, 

309 
gilberti sp. nov., 170, 194, 309 
formosanus, Acrossocheilus, 206 
Hemimyzon, 196, 309 
Liobagrus, 171, 184, 309 
Rhinogobius sp. nov., 170, 295, 
300, 313 
fusca, Eleotris, 288, 289 
fuscus, Clarias, 185, 309 

Gambusia, 257, 312 

afifinis, 170, 257, 312 
gidleyi, Ischyrotomus, 63 



Index. 



365 



gilberti, Formosania, sp. nov., 170, 194, 

309 
Gilmore, C. W., 40 
giurinus, Rhinogobius, 295, 297, 313 
Glossogobius, 170, 171, 302, 313 

abacopus, 171, 303, 306, 313 

brunneus, 302, 303, 313 

grammepomus, 170, 302, 304, 313 

parvus, sp. nov., 170, 303, 305, 313 
Gnathopogon, 198, 219, 310 

iijimae sp. nov., 170, 219, 310 
gnu, Catablepas, 90 
Gobiidse, 287, 313 
goodei, Centropholis, 329-334 
gracilis, Epihippus, loi 
Graham, George: English watchmaker, 

14 
grammepomus, Glossogobius, 170, 302, 

304. 313 
grande, Hyrachyus, 130 
grangeri, Mesomeryx, 73 
Gray, Benjamin: watchmaker to George 

II, 17 
Guyasuta Indian Mound: Exploration 

of, 36 
guyotii, Desmatotherium, 103, 127 

Halecomorphi of the family Amiidee, 41 

Hall, Robert C: Collection of American 

Indian objects, 37 

hargeri, Lycarion (Vulpavus), 52 

Harpagolestes uintensis, 41, 42 

Healey, General: D'Orsay (carriage) 
presented by, 38 

Hearn, Mrs. George A., 28 

Heasle3% Mrs. Jessie Porterfield: pot- 
tery and glassware presented by, 38 

Heinz, H. J., Collection of Watches, 
2. 3. 3-42; Death of, 34; Obituary 
by W. J. Holland, pp. 360-363 

Helaletes boops, 103, 104, 109, 11 1 
nanus, 104, 109, 112 

Helaletidae, 104 

Hemibarbus, 197, 210, 310 
labeo, 211, 310 

Hemimyzon, 193, 195, 309 
formosanus, 196, 309 

Heptodon, 104, iii 

herpestoides, Oodectes, 54 

Heteromeryx, 94 

himantegus, Acheilognathus, 231, 311 

Holland, W. J., Editorial Notes by, 

1-3; 33-39 
Obituary of Eastman, Charles R., 

347-350 
Obituary of Heinz, H. J., 360-363 
Obituary of Lambing, Andrew 

Arnold, 351-353 



Holland, W. J., Obituary of Smith, 

Herbert Huntingdon, 354-359 
hollandi, Spinibarbus, sp. nov., 170, 

218, 310 
Homacodon, 73, 74, 75 
Homacodontinse, 66 
Homalopteridse, 193, 309 
Homogalax protapirinum, 118 
Homogalaxinse, 116 
Hylomer>Tc, 67, 72, 74 

annectens, 68 
Hypertragulidae, 93 
Hypertragalus, 94, 97 
Hypophthalmichthys, 198, 244, 311 

molitrix, 170, 244, 311 
Hyrachyinae, 127 
Hyrachyus agrarius, 128, 129, 130 

grande, 130 

modestus, 122 

nanus, 112, 113 

princeps, 130 
Hyracodontidae, 127 
Hyracodontinae, 131 
Hyracotheriinse, 10 1 

idellus, Ctenopharyngodon, 229 
iijimae, Gnathopogon, sp. nov., 170, 

219, 310 

insolens, Achaenodon, 79, 80 
intermedium, Amynodon, 131 
Ischyrotomus gidleyi, 63 
Ischyromidse, 60 
Isectolophidae, 115 
Isectolophinse, 116 

Isectolophus annectens, 103, 116, 117, 
120 

latidens, 103 

scotti, 120 
Ishikauia, 199, 228, 311 

macrolepis, 171, 228, 311 
isselensis, Lophiodon, 114 
japonica, Anguilla, 262, 263, 312 

japonicus, Caristius, 329-334 

Sicyopterus, 170, 293, 313 
Jennings, O. E., Description of Cypri- 
pedium passeriniitn Richardson, 343 
Jones, Henry: English watch and 

clock-maker, 12 
Jordan, D. Starr, Article by, 329- 

342 
jordani, Labeo, sp. nov., 170 

kikuchii, Phoxiscus, sp. nov., 170, 226, 

310 
"Kirara," Description of, 254 
Klages, Samuel M., 2 
kneri, Cultriculus, 170, 253 



366 



Index. 



Kuhlia, 279, 312 

marginata, 279, 312 
Kuhliidae, 279, 312 

Labeo, 197, 203, 310 

jordani, 170, 204, 310 
labeo, Hemibarbus, 211, 310 
Labyrinthici, 275, 312 
Lambing, Monsignor Andrew Arnold, 
death of, 34; Obituary of, by W. J. 
Holland, 351-353 
latidens, Isectolophus, 103 

Parisectolophus, 121 
latipes, Oryzias, 170, 256, 312 
Leptaceratherium trigonodum, 128 
Leptalophiodon annectens, 126 
Leptolophiodon (Isectolophus) annec- 
tens, 115 
Leptomerj'x evansi, 94, 95 

transmontanus, 95, 96, 97 
Leptoreodon marshi, 98 
Leptotragulus, 96 

medius, 94 

proavus, 93, 95 

profectus, 97 
Le Roy, Julien: watchmaker, 14 
Letters from men of science, 38-39 
Limnocyon douglassi, 45 

potens, 46 

verus, 44, 45, 47 
Link, John, in Officers' Training Camp, 

34 
Liobagrus, 175, 182, 309 

formosanus, 171, 184, 309 

nantoensis, 170, 183, 309' 
Liza, 268, 273, 312 

troscheli, 274, 312 
longipes, Colodon, 115 

Mimocyon, 48, 52 
Lophiodon isselensis, 114 

nanus, 112 
Lophiodontidae, 104 
Lycarion (Vulpavus) hargeri, 52 

macrolepis, Ishikauia, 171, 228 
Macropodus, 170, 277, 312 

filamentosus sp. nov., 170, 278, 312 
macrops, Chanodichthys, 248, 311 
macropus, Elephenor, 329-334 
maculatus, Ophicephalus, 281, 283 
magnus, Pleurocyon, 52 
Mammalia of the Eocene of the Uinta 

Basin, 41 
marginata, Kuhlia, 279, 312 
marshi, Leptoreodon, 98 
matthewi, Diplobunops, 76 
mauritiana, Anguilla, 262, 312 
meadi, Palaearctonyx, 56 



medius, Leptotragulus, 94 

Paramys, 61 

Pleurocyon, 59 

Protoreodon, 82, 84 
Merycoidodon, 84, 85, 86 

culbertsoni, 86, 87 
mesembrina, Metzia, 241, 311 
Mesohippus, 103 
Mesomeryx grangeri, 73 
Mesonychidse, 41 
Metriotherium, 75 
Metzia, 198, 241, 311 

mesembrina, 241, 311 
Miacidae, 48 
Miacis parvivorus, 49 

vulpinus, 48, 54 
Miller, Paul C, 86 
milleri, Pareumys, 66 
Mimocyon longipes, 48, 52 
minor, Protoreodon, 83 
Misgurnus, 187, 309 

anguillicaudatus, 187, 309 

decemcirrosus, 189, 309 
minus, Paloplotherium, iio-iii 
minusculus, Dilophodon, 113 
mite, Caenopus, 128 
modestus, Hyrachyus, 122 
molitrix, Hypophthalmichthys, 170, 

244. 311 
moltrechti, Pararasbora, 224, 310 
Monopteridae, 259, 312 
montanus, Bunomeryx, 72 
Mouillacitherium, 75 
Mugil carinatus, 170, 268, 272, 312 

cephalus, 170, 268, 312 

oeur, 268, 270, 312 
Mugilidae, 268, 312 
Muridae, 66 
Murie, O. J., in aviation service, 34 

nantoensis, Liobagrus, sp. nov., 170 
nanus, Helaletes, 103, 104 
Nelson's (Lord) Watch, 3 
nitidens, Sciuravus, 64 
nobilis, Aristichthys, 170, 246, 311 

obliquidens, Protapirus, 125 

Prothyracodon, 131, 132, 133, 137 
ocellatus, Pteraclis, 332, 334 

Rhodeus, 170, 232, 311 
ceur, Mugil, 268, 270, 312 
ommaturus, Acanthogobius, 307, 313 
opalescens, Platyberyx, 329-334 
operculatus, Polyacanthus, 170, 276, 

312 
Ophicephalidae, 281, 313 
Ophicephalus, 281, 313 

maculatus, 281, 283, 313 



Index. 



367 



Ophicephalus tadianus, 281, 282, 313 
Oodectes herpestoides, 54 

proximus, 49 
Orohippus progressus, 102 
Oromeryx plicatus, 98 
Oryzias, 255. 312 

latipes, 170, 256, 312 
Oshima, Masumitsu, "Contributions 
to the Study of the Fresh-water 
Fishes of the Island of Formosa," 
169 
ovatus, Vulpavus, 60 
Oxacron-Caenotlierium, 75 
Oxyaenodon dysclerus, 42, 43, 44 

dysodus, 42, 43 
Oxyaenidae, 42 
oxycephala, Eleotris, 288 

pachycephalus, Zacco, 235, 240, 311 
Palaearctonyx meadi, 56 
Palaeomeryx, 74 
Palaeotherium, no 
Paloplotherium minus, iio-iii 
papilio, Bentenia, 329-334 
Pappichthys plicatus, 41 
paradoxicus, Protoreodon, 83 
paradoxus, Barbodes, 213, 310 
Parahyus aberrans, 82 

vagus, 81, 82 
Parameryx sulcatus, 99 
Paramys compressidens, 60 

delicatus, 61 

(Ischyrotomus) petersoni, 61 

medius, 61 

robustus, 61, 62 

sciuroides, 61 

uintensis, 61 
Pararasbora, 198, 224, 310 

moltrechti, 224, 310 
Parasalanx, 173 

acuticeps, 171, 173, 174, 309 

ariakensis, 170, 173, 174, 309 
Parasilurus, 175, 309 

asotus, 176, 309 
Pareumys milleri, 66 
Parisectolophus latidens, 118, 121 
parva, Pseudorasbora, 170, 222, 310 
parvivorus, Miacis, 49 
parvus, Epihippus, 102 

Glossogobius, sp. nov., 170, 303, 
305, 313 

Protoreodon, 83 
passerinum, Cypripedium, 343 
Perissodactyla, loi 
petersi, Centropholis, 329-334 
Peterson, O. A. Report upon the 
material discovered in the Upper 
Eocene of the Uinta Basin by Earl 



Douglas in the years 1908-1909, and 
by O. A. Peterson in 191 2, 40-165 
Petersoni, Paramys, 61 

Protylopus, 88, 92 
Phoxiscus gen. nov., 170, 198, 225, 310 

kikuchii sp. nov., 170, 226, 310 
Pinchbeck, Charles, 20 
Pinchbeck, composition of, 20 
Platyberyx, 329^334 

opalescens, 329-334 
platycephalum, Caenopus, 128 
platypus, Zacco, 235, 311 
Plecoglossus, 171, 308 

altivelis, 172, 308 
Pleurocyon magnus, 52 

medius, 59 
plicatus, Oromeryx, 98 

Pappichthys, 41 
Poebrotherium, 89, 90, 96 
Poeciliidae, 255, 312 
Polyacanthus, 170, 275, 312 

operculatiis, 170, 276, 312 
potens, Limnocj'on, 46 
Prentice, Sydney C, 40 
prince ps, Hyrachyus, 130 
priscum, Schizotherium, 139 
proavus, Leptotragulus, 93, 95 
Procynodictis vulpiceps, 51, 52 
Prodaphaenus (Miacis) uintensis, 51 

(?) robustus, 50, 54 

scotti, 51 
profectus, Vulpavus, 54, 58, 60 
progressus, Orohippus, 102 
Prosciurus (?) robustus, 65 
Protagriochaerus annectens, 76, 77, 86, 

87 
protapirinum, Homogalax, 118 
Protapirus obliquidens, 125 

simplex, 125 

validus, 125 
Prothyracodon obliquidens, 131, 132, 
136, 137 

uintense, 134, 135, 137, 138 
Protoreodon medius, 82, 84 

minor, 83 

paradoxicus, 83 

parvus, 83 

pumilus, 83 
Protylopus, 90, 92 

annectens, 91, 92, 99 

petersoni, 88, 92 
proximus, Oodectes, 49 
Pseudobagrus, 170, 175, 178, 309 

adiposalis sp. nov., 170, 178, 181, 
309 

brevianalis, 178, 309 

taiwanensis sp. nov., 170, 178, 180, 
309 



368 



Index. 



Pseudogobio, 198, 220, 310 

brevirostris, 221, 310 
Pseudorasbora, 198, 222, 310 

parva, 170, 222, 310 
Pseudotapirs of the North American 

Eocene, 103 
Pteraclidae, 331, 334 
Pteraclis, 329, 331, 334 

carolinus, 332, 334 

ocellatus, 332, 334 

trichipterus, 332, 334 

velifera, 331, 332, 334 
Pterycombus, 329-334 

brama, 329-334 
pumilus, Agriochoerus, 85 

Protoreodon, 83 
Puntius, 197, 216, 310 

snyderi sp. nov., 170, 216, 310 

quadricuspis, SphenomerjTC, 71 
Quare: English watchmaker, 11 
quercyi, Diplobune, 76, 77, 78 

Reptilia of the Eocene of the Uinta 

Basin, 41 
Rhinocerotoidea, 127 
Rhinogobius, 170, 295, 313 

candidius, 295, 313 

caninus, 295, 301, 313 

formosanus, 170, 295, 300, 313 

giurinus, 295, 297, 313 

taiwanus, 170, 295, 298, 313 
Rhodeus. 198, 232, 311 

ocellatus, 170, 232, 311 
robustus, Achaenodon, 80, 81 

Paramys, 61, 62 

Prodaphcenus, 50, 54 

Prosciurus, 65 
Rodentia, 60 

Salangidae, 173, 309 
Salmonidae, 171, 303 
Scaphesthes gen. nov., 170, 197, 208, 
310 

tamusuiensis sp. nov., 170, 209, 
310 
Schizolophodon cuspidens, 122 
Schizotheriinse, 139 
Schizotherium priscum, 139 
Sciuravus altidens, 64 

nitidus, 64 
sciuroides, Paramys, 61 
scotti, Prodaphaenus, 51 
semifasciolata, Capoeta, 170, 214, 310 
semilunata, Diana, 329-334 
Shagreen, manner of manufacture, 18 
Sicyopterus, 170, 287, 293, 313 

japonicus, 170, 293, 313 



Siluridae, 175, 309 

simplex, Protapirus, 125 

sinensis, Anguilla, 171, 262, 267, 312 

Smith, Herbert Huntingdon, death of, 

35; Obituary Account of, by W. J. 

Holland, 354-359 
snj^deri, Puntius sp. nov., 170, 216, 310 
Sphenomeryx quadricuspis, 71 
Spinibarbus gen. nov., 170, 197, 217, 
310 
hoUandi sp. nov., 170, 218, 310 
Stewart, Douglas, with American Red 

Cross at Washington, 34 
sulcatus, Parameryx, 99 
"Systemodon"Homogalax, 103 

« 
tadianus, Ophicephalus, 281, 282 
taenia, Cobitis, 170, 192, 309 
taiwanensis, Pseudobagrus, sp. nov., 

170, 178, 180, 309 
taiwanus, Rhinogobius, sp. nov., 170, 

295. 298, 313 
tamusuiensis, Scaphesthes, sp. nov., 

170, 209, 310 
Tapiroidea, 104 
Tapirs (Pseudo) of the North American 

Eocene, 103 
temmincki, Zacco, 170, 235, 238, 311 
Tompion: "Father of English watch- 
making," II 
Torricellian tube constructed by Henr}' 

Jones, 12 
transmontanus, Leptomer\TC, 95 
trichipterus, Pteraclis, 332, 334 
trigonodum, Leptaceratherium, 128 
Triplopus cubitalis, 132, 137 
Tritemnodon agilis, 47 
troscheli, Liza, 274 

tumirostris, Disteechodon, 170, 227, 311 
Tylopoda, 89 
Types of Formosan Fishes in Carnegie 

Museum, 171 

Uinta Basin, Report upon the Material 
discovered in the Upper Eocene by 
Earl Douglas in the years 1908-1909, 
and by O. A. Peterson in 1912, 40 
uintense, Achaenodon, 80, 81, 82 

Prothyracodon, 134, 135, 137, 138 
uintensis, Epihippus, loi 
Harpagolestes, 41, 42 
Paramys, 61 
Prodaphcenus (Miacis), 51 

vagus, Parahyus, 81, 82 
validens, Protapirus, 125 
validus, Protapirus, 125 
velifera, Pteraclis, 331, 332, 334 



Index. 



369 



vcrus, Limnocyon, 44, 45, 47 
\'iverra zibetha, 44 
\'iilliainy, Justin; watchmaker, 17 
\'ulpavus ovatus, 60 

profectus, 54, 58, 60 
viilpiceps, Procynodictis, 51, 52 
viilpiniis, Miacis, 48, 54 
Watclics: Collection of H. J. Heinz in 
Carnegie Museum. By Douglas Ste- 



wart, VV. J. Holland, A. S.Coggeshall, 
4-32 

Zacco, 198, 234, 311 

pachycephalus, 235, 240, 311 

platypus, 235, 311 

temmincki, 170, 235, 238, 311 
zibetha, Viverra, 44 



Publications of the Carnegie Museum Serial No. 97 



ANNALS 



CARNEGIE MUSEUM 



Vol. XII. No. I. 



November, 1917. 



For sale by Messrs. Win. Wesley & Sons, 28 Essex St. Strand, London. England; 
Maruzen Company, Ltd., 11— 16, Nihonbashi Tori Sanchome, Tokyo, Japan; and al 
the Carnegie Museum. Schenley Park, Pittsburgh, Pa., U. S. A. 



CONTENTS 
Editorial Notes 



I. A Catalog of the Collection of Watches Belonging to Mr. 
H. J. Heinz of Pittsburgh and Deposited by him in 
the Carnegie Museum. By Douglas Stewart, W. J. 
Holland and A. S. Coggeshall 4 



Publications of the Carnegie Museum Serial No. loo 



ANNALS 



CARNEGIE MUSEUM 



Vol. XII. Nos. 2-4. 



October, iQig* 



For sale by Messrs. Wm. Wesley & Sons, 28 Essex St. Strand, London. England; 
Maruzen Company, Ltd., 11-16, Nihonbashi Tori Sanchome, Tokyo, Japan; and at 
the Carnegie Museum, Schenley Park, Pittsburgh, Pa., U. S. A. 



CONTENTS 

Editorial Notes 

33 

II. Report Upon the Material Discovered in the Upper 
Eocene of the Uinta Basin by Earl Douglas in the 
Years 1908-1909, and by O. A. Peterson in 1912. 

By O. A. Peterson . . 

40 

III. Contributions to the Study of the Fresh Water Fishes 
of the Island of Formosa. By Masamitsu 

OSHIMA . . 

169 

On Elephenor, A New Genus of Fishes from Japan. 

By David Starr Jordan 329 

A Description of Cypripedium passerinum. By Otto 
E. Jennings 

Obituary Notes 

Charles Rochester Eastman 

350 



IV. 



346 

Andrew Arnold Lambing 

Herbert Huntington Smith 3^3 

Henry John Heinz 

Index to Volume XII . . 

363 



MBL WHOl Library - Serials 



5 WHSE 04434 






i 



!