HARVARD UNIVERSITY.

LIBRARY

OF THE

MUSEUM OF COMPARATIVE ZOOLOGY

iS'Vio

^^Ia^uloMj 9, 1 1 1S' %n^A^iJU / ?, ' ? ^ <

MEMOIRS

OF THE

CARNEGIE MUSEUM

VOL. VII

4920)-

W. J. HOLLAND, Editor

PITTSBURGH

PnaUBHED BY THE AUTHOEITY OF THE BoABD OF TkUSTEES OF THE)

CARNEGIE INSTITUTE

PRESS OF

IHE NEW ERA PRINTING COMPANY

LANCASTER, PA.

PREFATORY NOTE

The six Memoirs included in this vohime appeared at various dates as follows:
Memoir No. 1, December, 1915;
No. 2, November, 1916;
No. 3, November, 1916;
" No. 4, April, 1917;

No. 5, September, 1918;
No. 6, July, 1920.

Owing to circumstances beyond the control of the Director and Editor of the
publications of the Carnegie Museum a partial suspension of the publishing activi-
ties of the Museum took place during the period in which the United States of
North America were at war with the so-called "Central Powers" of Europe.
This cessation of our issue of scientific literature was only temporary. At the time
these lines are being written a resumption of our activities along the lines alluded
to has taken place. The issue of Memoir No. 6 from the pen of Mr. 0. A. Peterson
upon "The American Diceratheres " concludes the Seventh Volume of the Memoirs.
The Eighth Volume of the Memoirs is well under way, Memoir No. 1 having already
appeared and No. 2 being in the hands of the printers. The publication of the

"Annals" has also been resumed.

W. J. Holland.

Carnegie Museum,
July, 1920.

in

TABLE OF CONTENTS

PAGE

Prefatory Note iii

Table of Contents v

List of Figures in Text vii

List of Plates xi

List of Species New to Science Described in this Volume xv

Errata, Corrigenda, and Amplifications by the Author of Me-
moir on Cheirodontin^ xix

Memoir No. 1. The Cheirodontinse, a Subfamily of Minute Characid

Fishes of South America. By Carl H. Eigenmann . . 1-100
Memoir No. 2. The Fossil Turtles of the Uinta Formation. By Charles

W. Gilmore 101-162

Memoir No. 3. A Catalog of the Ophidia from South America at
Present (June, 1916) contained in the Carnegie Mu-
seum, with Descriptions of Some New Species. By

Lawrence Edmonds Griffin 163-228

Memoir No. 4. Pimelodella and Tj'phlobagrus. By Carl H. Eigen-
mann 229-258

Memoir No. 5. The Pygidiidae, a Family of South American Catfishes.

By Carl H. Eigenmann 259-398

Memoir No. 6. The American Diceratheres. By 0. A. Peterson 399-476

Index 477-488

LIST OF FIGURES IN TEXT.
Memoir No. 1.

FIGTJEE ^A«^

1. Types of teeth and their variations in the Cheirodontinw 9

2. Phylogenic Arrangement of the Genera of the Cheirodontiim 13

3. Arrangement of teeth in Grundulus bogotensis (Humboldt) 18

4. Arrangement of teeth in Spintherobulus papilliferus Eigenmann 20

6. Outhne of head and teeth in Probolodus heterostonms Eigenmann 22

6. Outhne of head and teeth in Aphyocharax dentatus Eigenmann 26

7. Outline of head and teeth in Aphyocharax anisitsi Eigenmann 30

8. Aphyocharax avary Fowler. 31

9. Aphyocharax rathbuni Eigenmann. Greatly enlarged 32

10. Dentition of Aphyocharax paraguayensis Eigenmann 33

11. Outline of head and dentition of Prionobrama filigera Eigenmann 40

12. Outline of head and dentition of Parecbasis cyclolepis Eigenmann 45

13. Dentition of Macropsobrycon Uruguayans Eigenmann 49

14. Outline of head and dentition of Megalamphodus melanopltrus Eigenmann . . 51

15. Dentition of Megalamphodus heteresthes (Ulrey) 54

16. Outline of head and dentition of Megalamphodus vdcrupkrus Eigenmann. . 55

17. Dentition of Oligobrycon microstomus Eigenmann 57

18. Outline of head and dentition of Aphyocheirodon hemigrammus Eigenmann . 59

19. Outline of head and dentition of Compsura heterura Eigenmann 61

20. Dentition of Microbrycon ribeiroi Eigenmann 63

21. Dentition of Cheirodon annce McAtee "8

22. Interhsemal spines of Cheirodon annoe, cf 69

23. Dentition and enlarged view of interhsemal spines of Cheirodon annce 69

24. Interhsemals in Cheirodon interruptus Jenyns 72

25. Dentition of Cheirodon monodon Cope ' ^

26. Cheirodon notomelas Eigenmann '5

7Q

27. Cheirodon piaba Liitken ' ^

28. Cheirodon piaba Liitken °^

29. Cheirodon microdon Eigenmann t 81

CO

30. Cheirodon stenodon Eigenmann

31. Holesthes pequira (Steindachner)

32. Holesthes heterodon Eigenmann ^^

33. Odontostilbe hastata Eigenmann ^^

vii

Vlll MEMOIRS OF THE CARNEGIE MUSEUM.

34. Odontostilbe drepanon Fowler 93

35. Odontostilbe microcephala Eigenmann 94

36. Odontostilbe madeirce Fowler 97

Memoir No. 2.

1. Carapace of Baena arenosa Leidy 108

2. Carapace of Baena inflata Gilmore 113

3. Plastron of Baena inflata Gilmore 114

4. Anterior lobes of Baena inflata Gilmore 116

5. Carapace of Baena gigantea Gilmore 117

6. Plastron of Baena gigantea Gilmore 118

7. Lateral View of Carapace and Plastron of Baena gigantea 119

8. Plastron of Baena platyplastra Gilmore 120

9. Carapace of Echmatemys callopyge Hay 124

10. Plastron of Echmatemys callopyge Hay 125

11. Carapace of Echmatemys douglassi Gilmore 129

12. Plastron of Echmatemys douglassi Gilmore ' 130

13. Carapace of Echmatemys hollandi Gilmore 133

14. Carapace of Echmatemys obscura Gilmore 137

15. Plastron of Echmatemys obscura Gilmore 138

16. Carapace of Echmatemys depressa Gilmore 140

17. Portions of the Carapace and Plastron of Echmatemys pusilla? Hay 142

18. Plastron of Hadrianus corsoni (Leidy) 145

19. Anterior half of plastron of Hadrianus robustus Gilmore 147

20. Plastron of Hadrianus utahensis Gilmore 149

21. Carapace of Testudo uintensis Gilmore 152

22. Plastron of Testudo uintensis Gilmore 153

Memoir No. 4.

1. Pimelodella buckleyi (Boulenger) 241

2. Pimelodella laticeps Eigenmann 243

Memoir No. 5.

1. Phylogenetic tree of the Pygidiida? 277

2. Details of structure of skulls of Hatcheria patagoniensis Eigenmann and

Scleronema operculatus Eigenmann 282

3. Hatcheria patagoniensis Eigenmann 283

4. Hatcheria areolata (Cuvier & Valenciennes) 286

5. Hatcheria burmeisteri (Berg.) 286

6. Hatcheria macrcei (Girard) 287

7. Pygidium tenue (Weyenbergh) 293

LIST OF FIGURES IN TEXT. IX

8. Pygidium corduvense (Weyenbergh) 293

9. Pygidium barbouri Eigenmann 303

10. Pygidium oroyce Eigenmann & Eigenmann 304

11. Pygidium quechorum Steindachner 305

12. Pygidium laticeps (Kner & Steindachner) 308

13. Pygidium chapmani Eigenmann 309

14. Pygidium tcenium (Kner) 310

15. Pygidium caliense Eigenmann 311

16. Pygidium proops (Ribeiro) 332

17. Pygidium brasiliense (Reinhardt) 338

18. Pygidium brasiliense (Reinhardt) 338

19. Pygidium itatiayoe (Ribeiro) 339

20. Pygidium minutum (Boulenger) 340

21. Outline of head, &c., of Pareiodon microps Kner 344

22. Henoneinus macrops (Steindachner) 346

23. Henonemus pundatus (Boulenger) 347

24. Henonemus anatomical details 347

25. Henonemus taxistigmus 348

26. HomodicBhis maculatus (Steindachner) 352

27. Stegophilus i7isidiosus Reinhardt 354

28. Acanthopoma annectens Liitken 355

29. Vandellia cirrhosa Cuvier & Valenciennes 361

30. Vandellia plazai Castelnau 362

31. Vandellia wieneri Pellegrin 363

32. Vandellia wieneri Pellegrin 363

33. Vandellia hasemani Eigenmann 364

34. Vandellia hasemani, anatomical details 364

35. Vandellia hasemani, anatomical details 365

36. Vandellia sanguinea Eigenmann 365

37. Vandellia, left premaxillary 366

38. Branchioica bertomi

39. Phreatobius cisternarum Goeldi 372

Memoir No. 6.

1. Diceratherium pleuroceros (Duvernoy) 403

2. Diceratherium minutum (Cuvier) 404

■ 404

3. Diceratherium doumllei

4. Upper dentition of Diceratherium cooki 409

5. Rhinoceros {? Diceratherium) pacificus Leidy 410

6. Rhinoceros {? Diceratherium) hesperius Leidy 411

X

MEMOIRS OF THE CARNEGIE MUSEUM.

7. Rhinoceros {? Diceratherium) oregonensis Marsh 412

8. Diceratherium iruquianum (Cope) 412

9. Diceratherium petersoni Loomis 413

10. Diceratherium armatum Marsh 414

11. Diceratherium annectens (Marsh) 418

11a. Diceratherium. annectens (Marsh) 420

12. Diceratherium gregorii Peterson 423

13. Diceratherium niobrarense Peterson 426

14. Diceratherium niobrarense Peterson 427

15. Diceratherium niobrarense Peterson, atlas 428

16. Diceratherium cooki Peterson, outhne of skull 433

17. Diceratherium, atlas 436

18. Diceratherium, axis 436

19. Diceratherium, fourth cervical 437

20. Diceratherium, sixth cervical 437

21. Diceratherium, seventh cervical 438

22. Diceratherium, first dorsal 438

23. Diceratherium, tenth dorsal 439

24. Diceratherium, eighteenth dorsal 439

25. Diceratherium, second lumbar 443

26. Diceratherium, fourth lumbar 440

27. Diceratherium, fifth lumbar 441

28. Diceratherium, sacrum 441

29. Diceratherium, sternum 442

30. Diceratherium, scapula 443

31. Diceratherium, humerus 444

32. Diceratherium, radius and ulna 445

33. Diceratherium, radius and ulna 445

34. Diceratherium, pelvis 446

35. Diceratherium, femur 447

36. Diceratherium, tibia and hbula 447

37. Diceratherium, patella 448

LIST OF PLATES

PLATE

I. Map Showing Route of John D. Haseman of the
Carnegie Museum Expedition to South America.

II. Fig. L Grundulus bogotensis CRumholdt) .

Fig. 2. Aphyocharax paraguayensis Eigenmann.

III. Figs. 1-4. Spintherobolus papilliferus Eigenmann.
Fig. 5. Aphyocharax raihhuni Eigenmann.

Fig. 6. Aphyocharax anisitsi Eigenmann & Kennedy.

IV. Fig. 1. Aphyocharax dentatus Eigenmann & Kennedy.
Fig. 2. Paragoniates alburnus Steindachner.

Fig. 3. Leptagoniates steindachneri Boulenger.

Fig. 4. Prionobrama filiferus (Cope) .

Fig. 5. Prionobrama paraguayensis (Eigenmann).

V. Fig. 1. Phanagoniales unlsoni Eigenmann.

Fig. 2. Parecbasis cyclolepis Eigenmann.

VI. Fig. 1. Leptobrycon jatuaranoe 'Eigenmarm.

Fig. 2. Macropsobrycon uruguayance Eigenmann.

VII. Megalamphodus megalopterus Eigenmann.

VIII. Fig. 1. Megalamphodus micropterus Eigenmann.

Fig. 2. Microschemobrycon guaporensis Eigenmann.

IX. Fig. 1. Oligobrycon viicrostomus YAgenmsimi.

Fig. 2. Aphyocheirodon hemigrammus Eigenmann.

X. Fig. 1. Compsura heterura Eigenmann.

Fig. 2. Misobrycon ribeiroi Eigenmann.

XI. Fig. L Cheirodon amiw McAtee.

Fig. 2. Cheirodon parahybce Eigenmann.

XII. Fig. 1. Cheirodon interruptus J enyna.

Fig. 2. Cheirodon notomelas Eigenmann.

XIII. Fig. L Cheirodon madeirce Eigenmann.
Fig. 2. Cheirodon piaba Liitken.

XIV. Fig. 1. Cheirodon microdon Eigenmann.
Fig. 2. Cheirodon stenodon Eigenmann.

XV. Fig. 1. Holesthes pequira (Steindachner).

Fig. 2. Holesthes heterodon Eigenmann.
XVI. Fig. 1. Odontostilbe hastata Eigenmann.

Fig. 2. Odontostilbe paraguayensis Eigenmann & Kennedy.

XI

xu

PLATE

XVII.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5-6

XVIII.

Fig. 1.

Fig. 2.

XIX.

Fig. 1.

Fig. 2.

XX.

Fig. 1.

Fig. 2.

XXI.

Fig. 1.

Fig. 2.

XXII.

Fig. 1.

Fig. 2.

XXIII.

Fig. 1.

Fig. 2.

XXIV.

Fig. 1.

Fig. 2.

XXV.

Fig. 1.

Fig. 2.

XXVI.

Fig. 1.

Fig. 2.

XXVII.

Fig. 1.

Fig. 2.

iCXVIII.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Fig. 8.

Fig. 9.

Fig. 10.

Fig. 11.

Fig. 12.

XXIX.

Fig. 1.

Fig. 2.

Fig. 3.

MEMOIRS OF THE CARNEGIE MUSEUM.

Odontostilbe pulchra (Gill).

Cheirodon insignis Steindachner.

Odontostilbe paraguayensis Eigenmann & Kennedy.

Cheirodon pisciculus Girard.

Cheirodon piaba Liitken.

Carapace of Baena arenosa Leidy.

Plastron of Baena platyplastra Gilmore.

Carapace of Baena inflata Gilmore.

Plastron of Do.

Carapace of Baena gigantea Gilmoie.

Plastron of Do.

Carapace of Echmatcmys callopyge Hay.

Plastron of Do.

Carapace of Echmatemys douglassi Gilmore.

Plastron of Do.

Carapace of Echmatemys hollandi Gilmore.

Carapace of Echmatejrnjs depressa Gilmore.

Carapace of Echmatcmys obscura Gilmore.

Plastron of Do.

Plastron of Hadrianus corsoni (Leidy)

Plastron of Hadrianus robustus Gilmore.

Plastron of Hadrianus utahensis Gilmore.

Carapace of Amy da scxdumantiquum (Cope)

Carapace of Testudo uintensis Gilmore.

Plastron of Do.

Atractus toeniatus Griffin.

Dorsal view of head of Do.

Ventral view of head of Do.

Tropidodipsas spilogaster Griffin.

Dorsal view of head of Do.

Ventral view of head of Do.

Clelia euprepia Griffin.

Dorsal view of head of Do.

Ventral view of head of Do.

Elaps hollandi Griffin.

Dorsal view of head of Do.

Ventral view of head of Do.

Pimelodella serrata Eigenmann.

Pimelodella cristata (Miiller & Troschel).

Pimelodella avanhandavce Eigenmann.

LIST OF PLATES.

Xlll

PLATE

XXX.

Fig.

1,

Fig.

2,

Fig.

3

XXXI.

Fig.

1,

Fig.

2

Fig.

3

XXXII.

Fig.

1

Fig.

2

Fig.

3

XXXIII.

Fig.

1

Fig.

2,

Fig.

3,

XXXIV.

Fig.

1.

Fig.

2.

XXXV.

XXXVI-XXXIX.

XL.

XLI.

Figs. 1-2,

Figs. 3-4

XLII.

Figs. 1-2

Figs. 3-5

XLIII.

Figs. 1-2.

Figs. 3-5

XLIV.

XLV.

XLVI.

XLVII.

XLVIII.

XLIX.

L.

LI.

LII.

LIII.

LIV.

Fig. 1.

Fig. 2.

Pimelodella hasemani Eigenmann.

Pimelodella laticeps Eigenmann.

Pimelodella notomelas Eigenmann.

Pimelodella metce Eigenmann.

Piynelodella holiviana Eigenmann.

Pimelodella itapicuruensis Eigenmann.

Pimelodella meeki Eigenmann.

Pimelodella meeki.

Pimelodella griffini Eigenmann.

Pimelodella mucosa Eigenmann.

Pimelodella grisea Eigenmann.

Pimelodella chagresi (Steindachner).

Pimelodella euicenia Regan.

Typhlobagrus kronei Ribeiro.

Pectoral spines of species of Pimelodella viewed from

above and greatly magnified.
Maps showing the geographical distribution of the

PygidiidcE.
Anatomical details of the structure, principally of the

heads, of Eremophilus mutisii Humboldt and

Henonemus punctatus Boulenger.
Eremophilus mutisii Humboldt.
Paracetopsis occidentalis (Steindachner).
Nematogenys inermis (Guichenet) .
Hatcher ia maculata (Cuvier & Valenciennes).
Tridens melanops Eigenmann & Eigenmann.
Branchioica bertonii Eigenmann.
Scleronema, Hatcheria, Pygidium, and Pseudostego-

philus.
Pygidium.
Pygidium.
Pygidium,.
Pygidium.
Pygidium.
Pygidium.
Pygidium.
Pygidium .
Vandellia.

Eremophilus mutisii Humboldt.
Eremophilus mutisii Humboldt, juv.

XIV

MEMOIRS OF THE CARNEGIE MUSEUM.

PL.\TE

Fig. 3.

LV.

Figs. 1-

-5.

LVI.

FlGS. 1,

2

Figs. 3,

5,

LVII.

LVIII.

LIX.

LX.

LXI.

LXII.

LXIII.

LXIV.

LXV.

LXVI.

Pareiodon microps Kner.

Ochmacanthus.

Phreatobhts cisternarum Goeldi.

Homodicetus anisitsi Eigenmann & Ward.

Diceratherium armaium Marsh.

Type of D. 7ianum Marsh and D. Cooki Peterson.

Diceratherium gregorii Peterson.

Diceratherium cooki Peterson and Diceratherium nio-

brarense Peterson.

Do. Do.

Do. Do.

Diceratherium cooki Peterson and D. annectens

(Marsh).
Diceratherium cooki Peterson.
Diceratherium cooki Peterson and D. annectens

(Marsh).
Diceratherium cooki Peterson and D. annectens

(Marsh).

LIST OF GENERA AND SPECIES NEW TO SCIENCE DESCRIBED IN

THIS VOLUME

Class MAMMALIA.

Order UNGULATA.

Superfamily RHINOCEROTOIDEA.

Family Rhinocerotid^.

Genus Diceratherium Marsh.

(Fossil)
D. gregorii Peterson p. 421, PL LIX, text-fig., p. 423.

Genus Ccenopus Cope.
Ccenopus dakotensis Peterson, noni. nov. for Aceraiherium (Ccenopus Osborn) mite Cope, pp. 402, 435

Class REPTILIA.

Order TESTUDINES.

Family Baenid^b.

Genus Baena Leidy.

(Fossil)

, Baena inflata Gilmore p. 112, PI. XIX; text-figs. 2 and 3.

Baena gigantea Gilmore p. 116, PI. XX, figs. 1, 2; text-figs. 5, 6, 7.

Baena platyplastra Gilmore p. 120, PI. XVIII, fig. 2; text-fig. 8.

Family EmydidjE.
Genus Echmatemys Hay.

(Fossil)

Echmatemys douglassi (iilmore p. 128, PL XXII; text-figs. 11 and 12.

Echmatemys hollandi Gilmore p. 133, PI. XXIII, fig. 1; text-fig. 13.

Echmatemys obscura Gilmore p. 135, PL XXIV; text-figs. 14 and 15.

Echmatemys depressa Gilmore p. 139, PL XXIII, fig. 2; text-fig. IG.

Family Testudinid^.

Genus Hadrianus Cope.
(Fossil)

Hadrianus robustus Gilmore p. 146, PL XXV, fig. 2; text-fig. 19.

Hadrianus utahensis Gilmore p. 148, PL XXVI, fig. 1 ; text-fig. 20.

Genus Testudo Linnseus.
(Fossil)

Testudo uintensis Gilmore p. 150, PL XXVII; texts-figs. 21, 22.

XV

XVI LIST OF NEW GENERA AND SPECIES.

Order SQUAMATA.
Suborder SAURIA.

Family Anguid^.
Genus Glyptosaurus Auct.
Glyptosaiirus sp. indet p. 159.

Sul)ordcr SERPENTES.
Family Typhlophid^.
Genus Helminthophis Peters.
Helminthophis bondensis Griffin p. 165.

Family Colubrid^.

Aporophis mdanocephalus Griffin p. 171.

Atractus tceniatus Griffin p. 173, PI. XXVIII, figs. 1-3.

Liophis elceoides Griffin p. 187.

Bhadina'a orina Griffin p. 195.

Tropidodipsas spilogaster Griffin p. 197, PI. XXVIII, figs. 4-6.

Clelia euprepa Griffin p. 203, PI. XXVIII, figs. 7-9.

Clelia peruviana Griffin p. 204.

Elaps columbianus Griffin p. 216.

Flaps hoUandi Griffin p. 218, PI. XXVIII, figs. 10-12.

Class PISCES.

Family Characid^.

Subfamily Cheirodontin^.

Aphyocharax paraguayensis Eigcnmann p. 33, PI. II, fig. 2.

Leptobrycon Eigenmann, gen. nov p. 46.

Leptobrycon jatuarance Eigenmann p. 46, PI. VI, fig. 1.

Macrupsobrycon Eigenmann, gen. nov p. 48.

Macropsobrycon uruguayance Eigenmann p. 48, PI. VI, fig. 2.

Megalamphodus Eigenmann, gen. nov p. 49.

Meyulamphodus rnegalopterus Eigenmann p. 50, PI. VII.

Megalamphodus micropterus Eigenmann p. 54, PI. VIII, fig. 1.

Microschemobrycon Eigenmann, gen. nov p. 56.

Microschemobrycan guaparensis Eigenmann p. 56, PI. VIII. fig. 2.

Oligobrycon Eigenmann, gen. nov p. 56.

Oligobrycon microstomus Eigenmann p. 57, PI. IX, fig. 1.

Aphyocheirodon Eigenmaiui, gen. nov p. 58.

Aphyocheirodoji heynigramynus Eigenmann p. 59, PI. IX, fig. 2.

Compsura Eigenmann, gen. nov p. 60.

Compsura heterura Eigenmann p. 61, PI. X, fig. 1.

Mixobrycon Eigenmann, gen, nov p. 62.

LIST OF NEW GENERA AND SPECIES. XVll

Cheirodon parahybce Eigenmann p. 70, PI. XI, fig. 2.

Cheirodon notonielas Eigenmann p. 74, PI. XII, fig. 2.

Cheirodon ynadeirw Eigenmann p. 76, PI. XIII, fig. 1.

Cheirodon microdon Eigenmann p. 80, PI. XIV, fig. 1.

Cheirodon stenodon Eigenmann p. 82, PI. XIV, fig. 2.

Holesthes heterodon Eigenmann p. 87, PL XV, fig. 2.

Megalamphodus ecuadorensis Eigenmann p. 99.

Familj' Silurid^.

Pimelodella serrata Eigenmann p. 235, PI. XXIX, fig. 1.

Pimelodella steindachneri Eigenmann p. 237.

Pimelodella avanhandava: Eigenmann p. 240, PL XXIX, fig. 3.

Pimelodella hasemani Eigenmann p. 241, PL XXX, fig. 7.

Pimelodella laticeps Eigenmann p. 243, PL XXX, fig. 2.

Pimelodella laticeps austraUs Eigenmann p. 243.

Pimelodella notomelas Eigenmann. . . , p. 244, PL XXX, fig. 3.

Pimelodella metce Eigenmann p. 244, PL XXXI, fig. 1.

Pimelodella boliviana Eigenmann p. 254, PL XXXI, fig. 2.

Pimelodella itapicuruensis Eigenmaiui p. 247, PL XXXI, fig. 3.

Pimelodella griffini Eigenmann p. 250, PL XXXII, fig. 3.

Family Pygidiid^.

Scleronema operculatum Eigenmann p. 280, PL XLIV, fig. 1.

Pygidium zonalum Eigenmann p. 330, PL LI, fig. 1.

Pygidium proops parahybce Eigenmann p. 332.

Pygidium triguitatum Eigenmann p. 339, PL LII, fig. 4.

Pygidium santw-ritce Eigenmann p. 341, PL LII, fig. 5.

ERRATA AND CORRIGENDA.

PAGE

13. In fig. 2 above "f " for " Aphiocharax " read Aphyocharax.

16. In fig. 2 above "o" for "Gompsoura" read Compsura.

27. Tenth line from bottom for " Macrosobrycon " read Macropsobrycon.

40. Ninth fine from top for "paraguayense" read paraguayensis.

69. Figs. 22 and 23 for " Cheiwdon annce" read Cheirodon insignis. (N.B.
This change is made according to oral instructions reported to have been
given by the author of the paper to one of the ladies in my front office.
Editor.)

80. Third fine from bottom for " Aphiocheirodon" read Aphyocheirodon.

99. Eleventh line from bottom for "having" read leaving.

99. Tenth line from bottom for "ridge" read wedge.

99. Second line from bottom for "small" read mostly lost.

99. Bottom line after the word "spot" add no dorsal spot.
241. Fourth line from top for " Eigenmanniorum " and "Eigenmanni" read

eigenmanniorum and eigenmanni.
260. Nineteenth line from top for "Myoglanis" read Miuroglanis.
267. Fifteenth line from top for " Pseudolatystomus " read Pseudoplatystomus.
327. Seventh line from bottom for "braziliense" read brasiliense.
332. Fig. 16, for "Ribiero" read Ribeiro.

345. Ninth line from bottom for " Coblitiganis " read Cobitiglanis.
392. Fig. 4. for " reinhardtii " read reinhardti.
401. Thirteenth line from bottom delete ", both" and insert "both of " before "which."

PLATES

V. For "wilsoni Eigenmann. Type" read macrolepis Meek & Hildebrand.
Type of wilsoni.
IX. For "Jacquara" read Jag? uara.
X. For "Itapicucru" read Itapicuru.

XIX

ADDENDA AND CORRIGENDA CHEIRODONTIN.E

(Note. More than a year after Memoir No. 1 upon the Cheirodontince had
been pubhshed the Author sent the following paragraphs to one of the stenographers
in the Museum, requesting her to include them in the corrigenda. This lady very
properly turned the paper over to the Editor, who presumes that the Author intends
that they shall be published, and accordingly inserts them here as submitted.)

P. 43. Instead of "19. Phanagoniates wilsoni Eigenmann" read:

"19. Phanagoniates 7nacrolepis (Meek & Hildebrand).
Roeboides macrolepis Meek & Hildebrand, Field Mus. Publ. Zool. Ser., vol. X, 1913,

p. 84 (Rio Cupe, Boca de Cupe, Rio Tuyra).
P. 69. Meek & Hildebrand reporting on the wealth of material collected by them
in Panama (Field Mus. Publ. Zool. Ser., vol. X, 1916, pp. 273-276), find that
the specimens reported as Cheirodon insignis by Evermann and Goldsborough
are representatives of a new genus and species, Pseudocheirodon affinis,
Meek & Hildebrand, and that Cheirodon gorgonae Evermann & Goldsborough,
placed in the synonymy of Cheirodon insignis, is a member of the genus Comp-
sura.
P. 99, add: Mimagoniates Regan.
Mimagoniates Regan, Ann. & Mag. Nat. Hist., (7), XX, 1907, p. 402.

This genus is related to Prionobrama but lacks maxillary teeth and the anal
is but slightly emarginate. The original description by Regan follows.

"Body strongly elongate, compressed; abdomen keeled, but not strongly
compressed to an edge. Mouth small; teeth tricuspid, in a single series; no maxil-
lary teeth: palate toothless. Nostrils close together. Gill-membranes not united,
free from the isthmus. Scales cycloid, of moderate size; lateral Ime incomplete.
Dorsal fin short, posterior in position; adipose fin present; anal fin elongate.

"Intermediate between Chirodon, Girard, and Leptagoniates, Blgr."

Mimagoniates Barberi Regan.

"Depth of body 3-3.66 in the length, length of head 4-4.4. Snout much

shorter than eye, the diameter of which is 2.5-2.75 in the length of the head and a

Uttle less than the interorbital width. Cleft of mouth nearly vertical; maxillary

not extending to below the eye. 42-45 scales in a longitudinal series; lateral line

xxi

XXU ADDENDA AND CORRIGENDA CHEIRONDONTIN^E.

on 4-8 scales only. Dorsal 10; origin equidistant from gill-opening and base of
caudal, above the anterior part of the anal. Anal 34-38; origin equidistant from
anterior part of eye and base of caudal; anterior rays the longest, about .6 the
length of head; free edge straight or slightly concave. Pectoral extending to or a
little beyond the base of ventral. Caudal forked. A lateral band (blackish in
preserved specimens) from the lower part of eye to the lower lobe of caudal. An
oblique dark stripe on the dorsal; anal with a dark margin.

"Habitat, Arroyo Yaca, Estacion Cabellero, Paraguay.

"Several specimens, the largest 40 mm. in total length, collected by Dr. A.
Barbero."

Chirodon arnoldi Boulenger, Ann. and Mag. Nat. Hist. (6), Chirodon arnoldi
Boulenger, Ann. and Mag. Nat. Hist. (8), IV, p. 497.

This species described as a Cheirodon is probably a Hemigrammus. If it is a
Cheirodon and if it comes from the northern part of the Isthmus of Tehuantepec,
it extends the known range of the genus Cheirodon northward for more than ten
degrees latitude.

Head 4; depth 3. D. II. 9: A. III. 19; scales 32, 4 or 5 with pores, 11 scales in a
transverse series.

Strongly compressed, snout shorter than eye, eye 2.66 in head, equals inter-
orbital; maxillary not reaching anterior border of eye; lower jaw scarcely projecting.
Origin of dorsal just behind base of ventrals, equidistant from snout and caudal
longest ray of dorsal as long as head.

Yellowish above, finely speckled with black, silvery white beneath; a large
round black spot on caudal peduncle, extending on base of middle rays of caudal;
dorsal, ventrals, and caudal tinged with orange.

Length 33 mm. Said to have been imported by the Aquariest Arnold of Ham-
burg from Puerto Mexico, on the north coast of the Isthmus of Tehuantepec.

Publications oj the Carnegie Museum, Serial No. 87.

MEMOIES

OF THE

OAENEaiE MUSEUM.

VOL. VII. NO. 1.

W. J. HOLLAND, Editok.

THE CHEIEODONTIN^, A SUBFAMILY Of MINUTE
CHAEACID FISHES OF SOUTH AMEEICA

By GAEL H. EIGENMANN.

PITTSBUEGH.

Pttblished by the Atjthoeity of the Boaed op Trustees of the

CARNEGIE INSTITUTE.

Deoember, 1915.

PRICE LIST OF PUBLICATIONS OF THE CARNEGIE MUSEUM

ANNUAL REPORTS OF THE DIRECTOR
1898, 30e. (scarce); 1899, 25c.; 1900, 30e. (scarce); 1901-14, 25c. each.

REPORTS OF PROCEEDINGS OF FOUNDER'S DAY

1898-1914, 35c. each.

ANNALS OF THE CARNEGIE MUSEUM

The Annals are supplied to those who subscribe in advance in parts (paper-bound), as published, @ $3.50 per
volume; Vols. I-IX, 1901-1914, bound in green cloth @ $4.00; bound ia 1/2 Morocco @ $4.50.

MEMOIRS OF THE CARNEGIE MUSEUM
The Memoirs are supplied to those who subscribe in advance in parts (paper-bound), as published, at $10
per volume; bound in green buckram at $10.75, and in half -morocco at $11.50.

VOL. L (1901-3)

No. 1. Diplodocus, Its Osteology, Taxonomy, and No. 3. The Osteology of the Steganopodes. By

Probable Habits, with a Restoration of R. W. Shufkldt $2.76

the Skeleton. By J. B. Hatcheb. {Very " 4. Classification of the Chalcid Flies. By

scarce.) $3.00 W. H. Ashmead 6.50

" 2. Oligocene Canidae. By J. B. Hatcheb... 1.50

VOL. n. (1904-6)

No. 1. Osteology of Haplocanthosaurus and Ad- No. 5. The Tertiary of Montana. {Scarce.) By

ditional Remarks upon Diplodocus. By „ „ _,?^i' Pouglass .^ ... . 1.00

T T5 TT riT \ 4.0 nn " 6. The Osteology of Diplodocus. By W. J.

J. B. Hatcher. {Very scarce.) $2.00 Holland J.E3

« 2. Osteology of Baptanodon. By C. W. Gil- No. 7. The Osteology of Protostega. By G. R.

MORE. {Scarce.) 1.50 „ „ Wieland $ .76

„„„.,.. ^ . , . . "8. New Suillme Remains from the Miocene

« 3. Fossil Avian Remams from Armissan. of Nebraska. By 0. A. Peterson 76

By C. R. Eastman 1.00 No. 9. Notes on the Osteology of Baptanodon,

XT A tl • f « 10 _ T> J 1. J T»- vnth a Description of a New Species.

No. 4. Description of New Rodents and Discus- By C W Gilmore $100

sion of the Origin of Dsemonelix. By 0. « 10. A Monograph upon the Crawfishes of

A. Peterson. {Scarce.) $1.75 Pennsylvania. By A. E. Ortmann. ... 4.00

VOL. in.

No. 1. Archasological Investigations in Costa. No. 2. Osteology of Moropus. By W. J. Hoii-

Rica. By C. V. Haetman $6.00 land and 0. A. Peterson $6.00

VOL. IV.

No. 1. Early Chinese Writing. By Frank H. No. 5. New Carnivores from the Miocene of

Chalpant. Pp. 36, 50 Plates $3.00 Western Nebraska. By C. A. Peterson.

No. 2. Formosan Fishes. By D. Starr Jordan. Pp- 73, 12 Plates 1.50

Pp. 4, 3 Plates 25 No. 6. Monograph of the Najades of Pennsyl-

No. 3. Revision of Entel'odontid».' By"'o.'*A. ' '^ania. Pts. I and II. By A. E. Get-

r>„ „^«„ T) 1 1 c n Di i «o en MANN. Pp. 69, 4 Plates $2.60

Peterson. Pp. 116, 9 Plates $2.50 -kt « « ^ , .„ L- , „ ,,..«,

xr„ A no4.oi„,r^f fT,»T'j.i,«»«<'T<^ .,„o T^„T,.vT * No. 7. Catalog EoccHe Fishcs from Mouto Bolca

No. 4. Catalogcf the Fishes of Formosa. Jordan .^ Carnegie Museum. By Charles R.

and Richardson. Pp. 46, 12 Plates.... 1.25 Eastman $3.00

VOL. V.

The Fresh Water Fishes of British Guiana. By C. H. Eigenmann. Pp. i-xx -(- 578, frontispiece, and 103
plates. $10 unbound; $10.75 cloth; $12.00 % morocco.

VOL. VI.

No. 1. A Catalog of the Fishes known from the No. 4. Record of the Fishes obtained in Japan,

Waters of Korea. Jordan and Metz. ign. D. Stare Jordan and William

Pp. 66, 10 Plates $1.50 Francis Thompson Pp. 109, 19 Plates

■vToim-Ti. ijv rr r^ and 87 text-figures $3.50

No. 2. The Lantern-fishes of Japan. Charles °

H. Gilbert. Pp. 46, 4 Plates $1.00 j^^^ 5,7 Catalog of the FossU Fishes in the

No. 3. The Gymnotid Eels of Tropical America. Carnegie Museum. Parts II-IV. By C.

Max Mapes Ellis. Pp. 88, 9 Plates $1.50 R. Eastman $5.00

FEB 8 19ie

MEMOIRS

OF THE

OAENEGIE MUSEUM.

VOL. VII. NO. 1.

THE CHEIRODONTIN^, A SUBFAMILY OF MINUTE CHARACID
FISHES OF SOUTH AMERICA.^

By Carl H. Eigenmann.

Introductory.

The greater part of the work of preparing this monograph was done between
January and May, 19L5, while enjoying the hospitaUty of Mr. and Mrs. Carl G.
Fisher on their estate at Miami, Florida. President W. L. Bryan and the Trustees
of Indiana University appointed me Research Professor for the collegiate year
1914-1915, and the Director of the Carnegie Museum relieved me of resident
curatorial duties at the Museum. I thus gained the opportunity under ideal
conditions to give my undivided attention to this exceedingly difficult group of
fishes. I am indebted, as in former articles, to Dr. W. J. Holland, Director of the
Carnegie Museum, for assistance in arranging the figures in the text and on the
plates and for his editorial revision of the manuscript. The drawings given on the
plates were executed by Mr. Clarence Kennedy of Leland Stanford Jr. University.
The drawings given in the text are from camera lucida sketches made by the author.

This paper would naturally form a chapter in my Monograph of the Characidse,
to be pubfished by the Museum of Comparative Zoology at Cambridge, Mass.,
but, as the publication of the first volume of the monograph has long been delayed,
it is deemed best to publish this article at once.

The material on which this paper is based- consists of (a) the collections of

' Contributions from the Zoological Laboratory of Indiana University, No. 150.

' In enumerating the specimens at my disposal I have cited (a) the current numbers in the various
museums; (6) the letters a-x, indicating the number of specimens in a given series in the Carnegie Museum;
(c) the number of specimens in the particular lot under examination; (d) the size of the largest and sometimes

1

2 MEMOIRS OF THE CARNEGIE MUSEUM.

Harvard University, made chiefly by the Thayer expedition; (b) the collections of
Indiana University, made by H. von Ihering in Rio Grande do Sul, by J. D. Anisits
in Paraguay, and by Charles Wilson during the Landon-Fisher expedition to
Colombia; (c) the collections made under the joint auspices of the Indiana Uni-
versity and the Carnegie Museum on the occasion of the author's expedition to
British Guiana, and the reconnaissance undertaken by him in Colombia; and (d)
the collections made by Mr. John D. Haseman during the expedition of the Carnegie
Museum to central South America. The collections made by Mr. Haseman are
by far the largest both in the number of specimens and species. An account of
Haseman's travels, together with a list of his localities, was published in the Annals
of the Carnegie Museum, Vol. VII, p. 287. A map showing his route accompanies
the present paper.

There are twenty-one genera and fifty-six species and varieties of Cheirodontinse
now known. ^ In the present paper seven genera and seventeen species for the first
time are described. In all I have at one time or another described fourteen genera
and thirty-three species. Nineteen of the genera and thirty-nine species are repre-
sented in the collections of the Carnegie Museum. In the other museums, so far
as known, the species are represented as foUows:

Paris (Mus6e National) 1

Vienna (K. K. Hofmuseum) 8

London (British Museum) 9

Genoa (Museo Civico) 2

Copenhagen (Zoologisches Museum) 1

Cambridge (Mus. Comp. Zoology) 4

New York (Am. Mus. Nat. History) 1

Washington (U. S. National Museum) 2

PUladelphia (Acad. Nat. Sciences) 7

Bloomington (Indiana University Museum) 35

Ithaca, N. Y. (Cornell University) 1

I have examined practically all of the known species except Cheirodon pisciculus
from western Chili and Odontostilbe pulchra from Trinidad. However, while pre-

of the smallest specimen; (e) the locality; and (/) frequently the date of collecting and name of the collector.
Where the entire series is reserved for the Carnegie Museum the letters after the current number and the
number agree. When specimens have been destroyed by dissection or otherwise, or where there are numerous
duplicates, the letters and numbers do not necessarily agree.

' I have placed the genus Psalidodon in the Tetragonopterince, although it has the single row of notched
teeth characteristic of the Cheirodontinse. Psalidodon and Henochilus, in the latter of which there is a double
row of teeth in the upper jaw, form a little group bridging the gap between the Tetragonopterinaj and Cheir-
dontina;; or, on account of the absence of lips, they may be regarded as forming a little group distinct from either
of the above. Megalamphodus ecuadorensis, sp. nov. is described in the Appendix.

eigenmann: the cheirodontin^. o

paring this revision, I did not have access to specimens of Aphyocharax avanj
Fowler, Cheirodon eques Steindachner, Cheirodon agassizi Steindachner, Cheirodon
pulcher = nattereri Steindachner, Odontostilbe drepanon Fowler, Odontostilbe madeirce
Fowler, and Leptagoniates steindachneri Boulenger.

It is quite possible that several of these are here described under other names.
It is possible that C. agassizi is the male of Aphyocharax pulcher, and that Odon-
tostilbe drepanon is Holesthes pequira.

The Cheirodontin^.

The subfamily Cheirodontinse (Aphyocharacinse auctorum) belongs to the large
family Characida;. All the species are small or even minute. The giants of the
subfamily are only about 90 mm. long at their best. Parqgoniates alburnus reaches
a length of 90 mm. The largest recorded Grundulus is 80 mm. long, the largest
Probolodus 81, the largest Parecbasis 80, the largest Odontostilbe microcephala 80.
Then follow Aphyocharax dentatus with a maximum length of 72 mm., Cheirodon
interruptus QO mm., Aphyocharax alburnus and pusillus 58 mm., Holesthes pequira
56 mm., H. heterodon 50 mm., Prionobrama paraguayensis 50 mm., and P. filigerus
60 mm. The rest are all under 50 mm. in length.

Generalized type of the subfamily.— A composite of all the known species will
give us an idea of the ancestor of these species, assuming for the moment that they
had a common ancestor, which is open to some doubt. However, even if a few of
the genera included do not belong to this an otherwise homogeneous group, they
are so nearly like them that their inclusion will scarcely impair the full value of
the generalized type.

The generahzed type is a fash rather under fifty millimeters, or two inches, in
length; compressed, oval, with symmetric dorsal and ventral outlines. Its depth
at the origin of the dorsal is about one-third of the length from the tip of the snout
to the end of the median series of scales. The head is about equal to one-fourth
of this length. The eye is large, about one-third as long as the head. The mouth
is terminal and the maxillary reaches to about the origin of the eye. The cheeks
and postorbital portion of the head are protected by the well-developed chain of
suborbital bones, of which the third is in contact with the lower hmb of the
preopercle, there being a naked wedge between it and the vertical hmb of the pre-
opercle. The teeth are in a single series, comparatively few in number, and with
lateral notches. They occur along the entire edge of the premaxillary, at the
upper angle of the maxillary, and along the front and sides of the lower jaw. The

4 MEMOIRS OF THE CARNEGIE MUSEUM.

teeth of the maxillary are similar to those of the premaxillary and the lateral teeth
of the mandible are always smaller than those near the front. A frontal and a
parietal fontanel are present, the latter being the larger, truncated in front; the
former is triangular, the base of the triangle being caudad. The occipital crest is
narrowly triangular and divides the scales of the two sides for a distance of about
one-fourth of the length from its base to the dorsal. The dorsal is short, pointed,
consisting of one rudimentary and ten developed rays; having its origin midway
between the tip of the snout and the base of the middle caudal rays. The adipose
is a small free lobe as in the greater number of all the characid fishes, and is placed
over the end of the anal. The caudal is deeply forked. The anal is emarginate,
having its origin under the last dorsal ray, and it consists of twenty-five rays. The
ventrals are placed slightly in front of the origin of the dorsal and they do not quite
reach the anal. The pectorals are lanceolate and do not quite reach the origin
of the ventrals. The scales are thin, very regularly arranged. The fins are naked
except for a few scales along the base of the anterior anal rays. There is a well-
developed axillary scale over the ventrals. There are thirty-five scales along a
median series, eight of which have lateral line pores. There is a dark spot on the
sides from about the third to the fifth scales of the lateral line, and another larger
spot on the end of the caudal peduncle and the base of the caudal. In a triangular
area over the sides of the anterior air-bladder the wall of the body consists of skin
and peritoneum only.

Minor deviations from the generalized type. — The deviations from this general
type are numerous, but not very great. The greatest deviations are found in the
size of the frontal fontanel; the armature of the cheek, especially the postorbital
portion of it; the length of the anal; the degree of the development of the pseudo-
tympanum; and especially the size of the mouth and its parts and the style of the
teeth. Leaving these to be considered last, we find some of the species (Aphyo-
charax) are much less compressed than others, and in these the depth is frequently
less than one-third of the length, the minimum depth being contained about four
times in the length in a number of species of Aphyocharax. In the deepest the
depth is contained but 2.4 times in the length {Megalamphodus megalopterus) .
The head varies from 3.3-4.66 in the length in different species. It is comparatively
shortest in Paragoniates paixiguayensis and comparatively longest in Spintherobolus.
The eye is always large. In different species it is contained from 2.3-4.33 in the
length of the head, but in only two forms, Grundulus and Spintherobolus is it con-
tained as many as 3.75 times. In the great majority of cases it is contained 2.5-3
times. The adipose fin is absent in Grundulus and Spintherobolus, which otherwise

eigenmann: the cheirodontin.^. 5

do not differ greatly from the other species. The caudal lobes may be a little
longer or shorter, a little more pointed or rounded, and there may be more or less
difference between the upper and lower lobes, but there is no striking deviation
from the type. The ventrals and pectorals may be a little longer or shorter, but
here again there is no great divergence from the central type. The scales differ
materially. The lateral line may be developed on but two scales, or it may be
complete. It is complete in Probolodus, Parecbasis, Holesthes, and Odontostilbe.
It is almost complete in Microschemobrycon. In the other genera it is developed
on less than fifteen scales, the exact number varying with the species. In Grundulus
the predorsal scales have disappeared; in Aphyodite the caudal has become mostly
covered with small adherent scales, and in Compsura and Odontostilbe hastata
the male s provided with a few enlarged scales on the caudal which recall the
Glandulocaudinse. The degree of the development of the pseudotympanum differs
greatly, the humeral region being apparently normal in a number of species. It
is most highly developed in Holesthes, Odontostilbe, Megalamphodus, and the deeper
species of Cheirodon. In color (alcoholic, which means the distribution of melano-
phores only) the species of this subfamily do not differ greatly from species of
Heniigrammus and Hyphessobrycon of the Tetragonopterinse. In many species
of the Tetragonopterinse and other subfamilies some sort of a spot occurs on the
sides, a little behind the origin of the lateral line. This spot is found in over half
of the species with tricuspid teeth. In the species with multicuspid teeth it occurs
only in Mixobrycon. A caudal spot at the end of the caudal peduncle and on the
base of the caudal has an even wider distribution among the Characins in general.
Among the Cheirodontinse it is all but uniformly found in the species with multi-
cuspid teeth, i. e., in those species in which the humeral spot is not developed.
Other markings are some sort of a spot on the dorsal, which occurs in six species
belonging to four genera. It is therefore not a sign of relationship. Similar spots
occur on some of the smallest Tetragonopterids. Another marking which occurs
sporadically in several species is a dark band along the tips of the short anal rays
and across the lobe of the anal.

Deviations from the generalized type on which the genera are based. — The frontal
fontanel may be small, having the form of an equilateral triangle, and may form
only a wedge between the posterior part of the frontals (it has nearly reached the
vanishing point in Aphyocheirodon) or it may, as in Megalamphodus, be of nearly
uniform width and entirely separate the frontals.

In all but three genera, Grundulus, Spintherobolus, and Mixobrycon, the second
suborbital is in contact with the preopercle below. In both Grundulus and

6 MEMOIRS OF THE CARNEGIE MUSEUM.

Spintherobolus the suborbitals are feeble, very different from those in the other
genera. In Mixobrycon there is a naked area around the entire distal edge of the
third suborbital, similar to that in most of the Tetragonopterinse. The greatest
difference in the armature of the cheeks occurs in the postorbital region. In
Aphyocharax there are two postorbitals. Of these the upper is minute and neglig-
ible, the lower is large, convex, similar to the third suborbital, and covers the entire
postorbital area. Prionobrama has a similar arrangement. In other genera there
are three or more postorbitals and there is a wider or narrower naked area between
them and the vertical limb of the preopercle. In one genus, Aphyocheirodon,
there is considerable individual variation in the number and size of the postorbitals.
Anal fin. — The anal varies. In Cheirodon annce the base is very short, entirely
behind the dorsal, the margin is rounded, and the highest rays extend beyond the
tip of the last. In Paragoniates, Leptagoniates, and Phanagoniates the base is very
long, having its origin below the first dorsal ray or far in front of it. In Prionobrama
its margin is extremelj'^ falcate. Between these extremes there are various modi-
fications, the rays varying from twelve in Spintherobolus, fourteen in Leptobrycon,
Cheirodon annw, and Cheirodon pisciculus, to fifty in Paragoniates alburnus, and
seventy in Leptagoniates. In the great majority of species the number of rays
ranges between twenty and twenty-six. In the genus Cheirodon, the species of
which fall into two groups, the greatest range is from twelve to twenty-seven.
In C. pisciculus and C. annce the number of rays ranges from twelve to fifteen;
and in the nine other species from seventeen to twenty-seven. In seventy-nine
specimens of one species, Cheirodon interruptus, the number of anal rays is as follows:

Number of rays: 17, 18, 19, 20, 21, 22, 23, 2^

Number of individuals: 2 6 28 16 12 11 3 1

These are from several distinct localities, and the extremes have not been
observed in specimens from one locality.

Dorsal fin. — The variation of the dorsal is not nearly as great as that of the
anal, and therefore of less taxonomic importance. Usually its origin is a little
behind or in front of the middle of the body, the distance in either direction being
negligible, but in Grundulus and in Prionobrama, Paragoniates, Leptagoniates, and
Phanagoniates it is distinctly behind the middle. Its outline may be rounded, ob-
liquely truncate, or it may be distinctly falcate as in Parecbasis. It reaches its
extreme development in Megalamphodus megaloptertis.

Mouth. — The greatest evolution in this group, as in the rest of the Characins,
has taken place in the mouth and teeth. The mouth ranges in size from such a
minute affair as is found in Oligobrycon microstomus, Compsura heterura, Cheirodon

EIGENMANN: the CHEIRODONTINiE. <

notomelas, and C. piaba, to the clupeoid openings in Leptohrycon jatuarance, Macro-
psohrycon urugumjanm, Megalamphodus megalopterus, and M. micropterus. But the
size of the mouth in itself is of no great importance, for there is great variation
within such genera as Aphyocharax and Cheirodon. There is very great difference
in the size and shape of the premaxillary and maxillary, as the outhnes in the
text-figures show.

Teeth. — The greatest interest centers in the teeth. In Grundulus all of the
teeth are peg-like, conical, single-pointed. Such teeth frequently appear on the
sides of the lower jaw, when the teeth are otherwise very different. They are also
frequently found on the distal part of the maxillary, when the number of teeth on
this bone are considerable, and they are also found in the premaxillary when the
teeth are feeble or numerous. The next degree of complexity is found in the three-
pointed teeth in the genera Spintherobolus, Probolodus, Aphyocharax, Macropso-
brycon, Mwroschemobrycon, and Oligobrycon. In Aphyocheirodon there are three-
to five-pointed teeth in the upper jaw and five-pointed teeth in the lower jaw. In
the remaining genera the teeth have typically five or more points. Frequently
the teeth in the sides of the lower jaw are not only smaller, but belong to a lower
order, i. e., they have fewer points than the others in the same mouth. The same
may be true, but to a less extent, of the teeth on the outer part of the premaxillary
and on the distal part of the maxillary. Usually the number of teeth in any bone
differs inversely as the number of points to each tooth, though this is not always the
case. In the species with many-pointed teeth the number of teeth is usually very
limited, none to three in the maxillary of Cheirodon, none to four in Odontostilbe,
but in one species of that genus ranging from four to seven. However, several of
the genera with tricuspid teeth have no teeth on the maxillary, others have as
many as twenty or more.

To say that the teeth are unicuspid, tricuspid, or multicuspid, does not tell
the whole story. There have evidently been divergent radiations within each of
these groups both in the shape of the individual teeth and in their arrangement.
These teeth are so difficult to observe, even with the aid of the modern binocular
microscope and a spot-light, that in all cases where I had material the individual
bones were dissected out and mounted in balsam. Camera-lucida sketches were
then made. Under all the circumstances I think it will be best to entirely ignore
statements about the teeth in the older descriptions, whether made by myself or
by others. Statements that the entire edge of the maxillary is denticulate and
that there are no teeth in the maxillary are especially to be doubted.

Reverting to the modifications of the three types of teeth, unicuspid, tricuspid,

8 MEMOIRS OF THE CARNEGIE MUSEUM.

and multicuspid, it may be observed that the conic tooth may be a simple cone, but
is much more likely to be recurved. All the teeth of Grundulus are of this type.
Part of the teeth on the maxillaries of the genera with tricuspid teeth are also uni-
cuspid, and probably by degeneration some of those on the premaxillary. The
tricuspid type varies from a slender conic tooth with a minute notch on each side
to a tooth in which the three points are of about equal size, nearly coterminous, and
arranged in a line, to a heavy tooth with a blunt central point and two minute
lateral points, so arranged that the three points mark the angles of a triangle
(Probolodus) . The five- to nine-pointed teeth may have a large central cusp and
two graduate cusps on the sides of the tooth,the line connecting the five (or more
points) forming parts of an ellipse, or the points may be of nearly equal value and
nearly coterminous. Between these there are many shades, several variations
not infequently occurring in different parts of the same jaw. A very distinct type
of tooth as well as arrangement is found in the lower jaw of Aphyocheirodon. The
teeth in this jaw are usually five-pointed. The three middle points are of about
equal size and subtruncate, so that their tips form chisels rather than points.
The outer cusps are very minute and so far withdrawn from the level of the rest
that they are easily overlooked. That this surprising shape is not the result of
wear is shown by the relay-teeth which have the same shape as the rest. With all
these modifications the sides of the multicuspid teeth may be parallel or very
much contracted basally. The teeth are usually quite flat, or rather thin, but in
Mixobrycon the teeth are heavy and approach the shape of the teeth of the Tetra-
gonopterinse. In all but one species the teeth are strictly uniserial. In Megalam-
phodus micropterus one of the teeth of the premaxillary is sometimes out of line
with the rest, a little further forward, and forms either an incipient or a reminiscent
anterior series.

The number of teeth as well as the shape of the teeth described above indicate
that the dentition of this group of the Characins is highly speciahzed. In this
character of high specialization they are not unique among the Characins, for it is
in the shape, number, and arrangement of the teeth that the greatest divergence
has taken place.

The number of teeth on the premaxillary and the frequency of the appear-
ance of any given number is indicated in the following table:

Number of teeth in premaxillary: 3, 4, 5 , 6 , 7 , 8, 9, 10, 11, 12, IS, 14

Number of species having the given number of teeth: 171110 14 96 5 2 2 1 2

In this table seventeen species occur in more than one count; to be exact, five
species occur in two counts, ten in three, one in four, and one in five. In other

eigenmann: the cheirodontin^.

9

words, in five species there may be a deviation of one tooth from the normal, in ten
species there may be a deviation of as many as three teeth (Aphyocharax seven to
ten), and in one (Odontostilbe melandeta) there may be a deviation of as many as

Fig. 1. Types of teeth and their variations in the Cheirodontinae. a, mandibular tooth of Grundulus;
b and &', premaxillary teeth of Macropsobrycon; c-c-, premaxillary teeth of Aphyocharax anisitsi, c, the sym-
physeal tooth, c\ the second, and c^, the third tooth from it; d, mandibular tooth of Mcgalamphodus megalop-
terus; e, mandibular tooth of Spintherobolus; f and /', a mandibular and premaxillary tooth of Parecbasis; g
and g', a maxillary and premaxillary tooth of Prionobrama; h-K , entire set of mandibular teeth of an Oligo-
brycon microstomus; i-i", entire set of premaxillary teeth of a Megalamphndus microptcnis, i and i', the usual
type, i*, with incipient cusps on the sides of the median cusp; j-/, entire set of premaxillary teeth of an Aphyo-
cheirodon; f and /, active mandibular teeth of an Aphyocheirodon; /", relay tooth, wliich has not pierced the
gum of an Aphyocheirodon; k, k\ k°, k^, a mandibular tooth, a premaxillary tooth, and two maxillary teeth of
Compsoura heterura; 1-V-, a premaxillary and a maxillary tooth of a Cheirodon piaba.

10 MEMOIRS OF THE CARNEGIE MUSEUM.

four teeth. But of this latter species I have but few specimens, unsatisfactorily
preserved, and the result is doubtful. The larger number of the species have seven
premaxillary teeth. The number of species having six or eight teeth are nearly
equally matched. It must be borne in mind that the number of specimens examined
has not been so great that we can be sure that all variations have been observed.
The maxillary teeth recorded are as follows :

Number of teeth: 0,1, 2, 3 , 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20
Number of species: 47 19 10 75. 5 2113244322 1 1 1 3

Four of the species occur in two counts, six occur in three, four in four, one in
five, one in seven, and one in twelve (from nine to twenty in Aphyocharax dentatus) .
The prevailing number is two teeth, which are found in nineteen species, as shown
by the table; which also shows that the number of species having from twelve to
fourteen and twenty teeth is relatively greater than those having from seven to
eleven, and from fifteen to nineteen teeth.

The Relationship op the Cheirodontinje.

Are the Cheirodontinse a homogeneous group with a common ancestry, or
are they dwarfs of various other subfamilies? The most of them form a homo-
geneous group, divisible, however, into a number of minor groups. Doubt arises
as to Grundulus, which has only conic teeth, Paragoniates, Leptagoniates, and
Phanagoniates with a posterior dorsal, and Mixobrycon, which has tetragonopterid
teeth and cheeks. Certain other characters, notably the peculiar scaling in the
caudal of the males of Compsura and Odontostilbe hastata, also suggest relationship
to another subfamily, the Glandulocaudinse.

The unicuspid teeth of Grundulus suggest relationship with the Characinse,
as some of the Characinse with partially tricuspid teeth, Oligosargus exodon, and
Bramocharax suggest relationship with the Cheirodontinse. The general shape and
backward position of the dorsal of Paragoniates as well as the peculiar scales in the
tail of the male Compsura and Odontostilbe recall the Glandulocaudinse. The
heavy teeth and armature of the cheeks of Mixobrycon suggest Hyphessobrycon of
the Tetragonopterinse and so does the tooth out of line with the rest in Megalam-
phodus micropterus. However, a double row of teeth has several timesbeen evolved
in the Characinse from a single row; or a single row from a double row. I have
pointed out such cases in Indiana University Studies No. 20, and will have occasion
to point out others in the monographs on the Chalcininse and Gasteropelicinse.
In fact if it were not for other considerations, the single series of teeth in the Cheiro-
dontinse would be no more sufficient to segregate them from the TetragonoiDterinse

eigenmann: the cheirodontin^.

11

than would the completeness or incompleteness of the lateral hne justify the col-
location of the genera having these characters in subfamihes.

In the peculiar armature of the cheeks Prionobrama shows such similarity to
Aphyocharax that Cope placed it in the latter genus. In spite of its general appear-
ance and the backward position of the dorsal, Prionohrama is probably more nearly
related to Aphyocharax than to any member of the Glandulocaudinae.

The pecuhar scaling of the caudal in two of the species recalls the similar
character in the Glandulocaudinae and in Argopleura of the Tetragonopterinse.
It is possible that this is also an independently acquired character in a number of
remotely related genera.

While it is very probable that we are dealing with a natural group, it is certain
that different members point to three distinct subfamilies from which they may
have been derived, or to which they may have given rise. The Cheirodontinse
are certainly near the generahzed type of all the Characins.

Secondary Sexual Differences.

There are no conspicuous secondary sexual differences in the group under con-
sideration, unless there should be a striking difference in the lipochromes, which
have been dissolved in alcohol.

The differences when present consist in the length of the fin-rays, in the fila-
mentous termination of the fins, in the development of hooks on the anal, and less
frequently on the caudals and ventrals of the male, in the development of scaly
pockets on the caudal of the males, in the color of the dorsal, and, what is unique
for this group, in the high development of the interhsemals in the male in Cheirodon.
The details are given under the respective species.

Distribution.

The Cheirodontinae apparently reach their maximum development in the
middle Amazon and the upper La Plata basins. It is quite possible that this
greater abundance in the middle Amazon and in the Paraguay basin is apparent
rather than real. Little collecting has been done with fine meshed nets in the
Orinoco and in the upper waters of the Amazon.

The group as a whole has a very wide distribution. The genus Cheirodon has
a range all but coextensive with that of the subfamily. The species, with the
exception of Cheirodon piaba Liitken, are confined to rather limited ranges. Only
two species are found both in the Paraguay and in the Rio Guapore. A number of
species in the Madeira and Paraguay are evidently very closely related.

12

MEMOIRS OF THE CARNEGIE MUSEUM.

Table of the Distribution of the

Species

OF THE

CHEIRODONTIN.E.

a

o
'o

o c

<

.3

o
O
.a

ca

'd
a

■c

O

s

ca

a

0^

a
s

1

s

5

^C3

2

o

o
P3

.to

So
6

(

o

s

t 'S
c

Ph

aj
a

1

1—1

a
es
02

C8J2

-a a

Oh

X

6

s
s

1

1. Grundulus bogolcnsis (Humboldt)

2. Spintherobolus papilliferus Eigenmann

X

3. Probolodus heterostomus Eigenmann

X

4. Aphyocharax dentatus Eigenmann & Kennedy. . .

X

5. Aphyocharax alhurniis (Giinther)

X

X

6. Aphyocharax erythrurus Eigenmann

X

7. Aphyocharax pusillus Giinther

X

X

8. Aphyocharax anisitsi Eigenmann & Kennedy. . .

X

X

9. Aphyocharax avary Fowler

X

10. Aphyocharax rathbuni Eigenmann

X
X

?

11. Aphyocharax paragtiayensis Eigenmann !

12. Aphyocharax nattereri (Steindachner)

X
X

f

13. Aphyocharax agassizi (Steindachner)

14. Aphyocharax maxillaris XJlrey

15. Paragoniates alhurnus Steindachner

X

16. Prionobrama paraguayensis (Eigenmann)

X

17. Prionobravia filigerus (Cope)

X
X

X

18. Leptagoniates steindachneri Boulenger

19. Phanagoniates wilsoni Eigenmann

20. Parecbasis cyclolepis Eigenmann

X
X

X

21. Leptobrycon jatuarancB Eigenmann

22. Aphyodile grammica Eigenmann

X

23. Macropsobrycon uruguayanw Eigenmann

X

24. Megalamphodus megaloptenis Eigenmann

X

25. Megalamphodus eqiies (Steindachner)

X

26. Megalamphodus melanotus Eigenmann

X

27. Megalamphodus heteresthes (Ulrey)

X

28. Megalamphodus micropterus Eigenmann

X

X

29. Microschemobrycon guaporensis Eigenmann

30. Oligobrycon microstomus Eigenmann

X

31. Aphyocheirodon hemigrammus Eigenmann

X

32. Compsura heterura Eigenmann

X

X

33. Mixobrycon ribeiroi (Eigenmann)

X

34. Cheirodon piscicidus Girard

V

35. Cheirodon anna McAtee.'

36. Cheirodon insignis Steindachner

37. Cheirodon parahybce Eigenmann

X

38. Cheirodon interruptus (Jenyns)

X

39. Cheirodon monodon Cope

X

40. Cheirodon ibicuhyensis Eigenmann

X

41. Cheirodon notom.elas Eigenmann

X

42. Cheirodon madeirm Eigenmann

X
X

43. Cheirodon piaba Liitken

X
X

X

X

X

44. Cheirodon microdon Eigenmann

45. Cheirodon stenodon Eigenmann

X

46. Holesthes pequira (Natterer)

X

X

47. Holesthes helerodon Eigenmann

X

X

X

X

48. Odontostilbe haslata Eigenmann

X

49. Odoidoslilbe drepanon Fowler

X
X

50. Odontostilbe fuqitiva Cope

X

51. Odontostilbe microcephala Eigenmann

X

X

.52. Odimtostilbc pulchra (Gill)

X

53. Od<mtoKtilbc paragriayensis'Eigen. &Kennedy . . . .

X

54. Odontostilbe madeirm Fo-w\eT

X

55. Odontostilbe melandeta Eigenmann

X

3

4

1

13

12

13

6

5

2

4

4

2

1

1

' Habitat unknown. For Megalamphodus ecuadorensis sp. nov. Cf. Appendix.

eigenmann: the cheirodontin^.

13

The paired species are as follows :

Rio Madeira and its tributaries

Aphyocharax alburnus
Aphyocharaz pusillus
Prionobrama filigerus

Rio Paraguay

Aphyocharax dentatus
Aphyocharax anisitsi
Prionobrama paraguayensis

Cheirodon pisciculus, which ranges in western Chile at least from Santiago to
Puerto Montt, is one of two species of this group found on the Pacific slope. This
species also has the distinction of being the only member of the family of Characins

Fig. 2. Phylogenic arrangement of the genera of the CheirodontiniE. In tliis diagram the letters cor-
respond to and have the significance of the same letters in the following key. The genus Paragoniates, in
the diagram, includes the genus Prionobrama, now recognized as distinct.

14 MEMOIRS OF THE CARNEGIE MUSEUM.

which is found on the Pacific slope south of the dry area of central South America.
Members of the genera Cheirodon and Odontostilbe are found in the Chagres and
Atrato rivers in Panama and Colombia. Megalamphodus ecuadorensis is known
from one specimen taken in western Ecuador.

The highest altitude is reached by Grundidus, which swarms in the streams of
the plains of Bogota at an elevation of about 9,000 feet. Several species are
found in the San Francisco basin at a considerable elevation. The rest are, as
far as known, confined to the low lands.

Phylum PISCES Artedi.

Class TELEOSTOMI Bonaparte.

SuPERORDER OSTARIOPHYSI Sagemehl.

Order PLECTOSPONDYLI Cope.

Suborder HETEROGNATHI Cope.

Family CHARACIDiE Gill.

Subfamily Cheirodontin^ Eigenmann.
Teeth well-developed, in a single series, arranged vertically in the premaxillary,
dentary, and usually the maxillary. Usually some of the teeth with one or more
notches or cusps symmetrically arranged on either side of the median cusp. Ventral
surface rounded or slightly compressed, without serrations; body scaled; a parietal
and a frontal fontanel; dorsal short, with eleven or fewer rays; gill-membranes
free from the isthmus and from each other; nares close together, separated by a
flap only; adipose fin usually present.

Key to the Genera op the Cheirodontin^.
a. No adipose fin; mouth small; armature of cheeks very weak. Teeth coaic or triscupid in both jaws; lateral
line incomplete.
h. Predorsal area partly naked; tactile papillae normal; teeth all conic in a regular series.

1. Grundulus Valenciennes.
66. Predorsal area entirely scaled; tactile papilte excessively developed; teeth mostly tricuspid or conical.

2. Spintherobolus Eigenmann.
aa. Adipose fin well developed, except in No. 7, Phanagoniales, which see.
c. Teeth tricuspid or conic.

d. Teeth few, very heavy, tricuspid, five or six in each dentary, of which the first is directed upward
and outward, the third laterad, the fourth and later ones upward, at nearly a right angle to the

eigenmann: the cheirodontin^. 15

third; premaxillary with three or four teeth directed downward and forward, maxillary with
three to five similar teeth; third suborbital covering entire cheek in adult; three postorbitals
with only a very narrow naked border; scales small, 45-53; lateral Une complete; anal long;

interhsemals normal 3. Probolodus Eigenmann.

dd. Teeth slender, tricuspid, or conical in both jaws. Interhismals normal.^

e. Tliird suborbital covering the entire cheek, two postorbitals, of wliich the upper is minute,

the lower very large, similar to the third suborbital, and covering the entire area from

eye to pre-opercle; lateral line incomplete. Anal with seventeen to twenty-seven rays

in "/" and twenty-nine to nearly forty in "ff."

f. Anal short, with a blunt lobe or none; origin of anal behind the vertical from the middle

of dorsal; mouth variable, maxillary with from two to twenty teeth; dorsal and ventral

profiles similar in front of the vertical from the dorsal; caudal naked.

4. AphyocharaxGiinther.

ff. Anal long, the anterior rays long, filiform; origin of anal about on the vertical from

the origin of the dorsal; mouth large, maxillary with teeth along its entire margin

directed downward and backward, largest near the middle; ventral profile much more

arched than the dorsal 5. Prionobrama Fowler.

ee. Third suborbital large, the fourth in one genus at least smaller than the fifth or second post-
orbital. Anal with more than forty rays, its origin in advance of that of the dorsal.
x^. Adipose dorsal small; lateral line incomplete; anal not falcate; teeth along entire
maxillary; origin of dorsal but little behind vertical from origin of anal, which is

equidistant from snout and base of middle caudal rays.

6. Paragoniates Steindachner.

xx^. Adipose dorsal small; lateral line complete; origin of anal equidistant from snout and

the fourth fifth of its base 7. Leptagoniates Boulenger.

xxx^. Noadipose; lateralline short; origin of anal equidistant from the snout and the pos-
terior fourth of its base 8. Phanagoniates Eigenmann & Wilson.

eee. Tliird suborbital variable; three postorbitals, of wliichthe uppermost is minute; all of them
sometimes feeble. Anal moderate, its origin beliind that of the dorsal.
g. Teeth minute, very slender, many conical, except in No. 17.

h. Lateral line complete; caudal lobes partly scaled; no pseudotympanum; third sub-
orbital covering entire cheek; postorbitals large, but feeble; maxillary without
teeth, its upper margin thickened; sides of lower jaw raised, without teeth; origin

of dorsal in front of the middle 9. Parecbasis Eigenmann.

hh. Lateral line incomplete.

i. Anal short, of fourteen rays, the liighest extending beyond the tip of the last, ita
base equal to length of caudal peduncle; no teeth on the sides of the lower jaw,
or on the maxillary; sides of the lower jaw high, as in Parecbasis; maxillary
large, its outer margin very convex; lateral line very short.

10. Leptobrycon Eigenmann.

ii. Anal long, twenty-three to thirty rays, the highest ray not extending to the base

of the last ray.

* Dorsal 9-11, Cheirodontinae; the CrenuchinaB, which are similar to the foregoing, have dorsal 15-18.

^The letters x, xx and xxx are used here out of order because text-figure 2, in which the letters were

made to agree with those in the key, was drawn before genera 6 to 8, which are quite aberrant, were

included in this monograph.

16 MEMOIRS OF THE CARNEGIE MUSEUM.

j. Lateral line with but few pores; cheeks behind the eye mostly naked.

k. Caudal partly scaled H. Aphyodite^ Eigenmann.

kk. Caudal naked.

I. Pseudotympanum conspicuous; no teeth on maxillary; interorbital

very convex; frontal fontanel moderate; premaxillary with five,
to seven mostly conical teeth, similar to Parecbasis; the lateral
Une incomplete 12. Macrosobrycon Eigenmann.

II. No pseudotympanum; maxillary with two to twenty teeth; four

to eight tricuspid teeth in front part of mandible, minute teeth on
side; fontanels very large.. . . 13. Megalamphodus Eigenmann.
jj. Lateral Une extending to within four scales of the caudal. Twenty-five
or more teeth on each ramus of the mandible; twelve teeth on the maxil-
lary; frontal fontanel moderate; cheeks behind as well as below the eye
covered; mouth moderate; subfusiform.

14. Microschemobrycon Eigenmann.
gg. Teeth heavy, four on the premaxillary; mouth minute; maxillary not reaching to below
eye, with one or two teeth; origin of dorsal equidistant from snout and caudal; no
pseudotympanum; tlu'ee postorbitals covering nearly the entire postorbital area.

15. Oligobrycon Eigenmann.
cc. Teeth in part at least with five or more lobes.

m. Sides of teeth nearly parallel, those of the upper jaw narrow, with a prominent median cusp and
one or two minute cusps on each side, about ten teeth in each premaxillary; mandible with
tliree- to five-lobed teeth, much expanded at tip, the tln-ee middle lobes equal in size, blunt,
those of successive teeth in contact, forming a continuous cutting edge, the lateral lobes
excessively minute; mouth large, cheeks partly naked; frontal fontanel very short; no pseudo-
tympanum; interhfemals normal; lateral line incomplete. .. 16. Aphyocheirodon Eigenmann.
mm. Teeth with five or more points, tops of those of the lower jaw not forming a continuous cutting
edge,
n. Lateral Une incomplete.

0. Male with a lobe of large scales extending along the base of the middle caudal rays; all
but three or four last anal rays of the male with hooks; interhaemals of the caudal
peduncle weak; mouth minute; maxillary barely reaching to below eye; second sub-
orbital covering the entire cheek; postorbitals very weak; no pseudotympanum;

frontal fontanel moderate 17. Compsura Eigeimiarm.

00. Caudal of male without a median lobe of scales.

p. InterhiEmals of the caudal peduncle of the male, next the caudal, feeble; no pseudo-
tympanum; teeth heavy, one on the maxillary; third suborbital with a naked

border 18. Mixobrycon Eigenmann.

pp. Interha;mals of the caudal peduncle, especially in the male, prominent, spine-like,
protruding; a conspicuous pseudotympanum; one to three teeth on the maxillary;
teeth of upper and lower jaw similar, with five or more points.

19. Cheirodon Girard.
nn. Lateral line complete; interhsemals feeble, not projecting; maxillary with a few broad-tipped
teeth; second suborbital in contact with the preopercle below.
' Aphyodite, Macropsobrycon, and some of the species of Megalamphodus are very similar.

eigenmann: the cheirodontin^. 17

q. Premaxillary teeth but little expanded toward the tips, pointed, the median point a little
the larger; mandibulary teeth much expanded at the tip, with a small basal notch on
each side and three median points of about the same size; third suborbital in contact

with the preopercle below; fontanel short 20. Holesthes Eigenmann.

qq. Teeth of the upper and lower jaws similar; maxillary with a few broad-tipped teeth;
third suborbital in contact with the preopercle below. Caudal in the male naked or
with a row of scales reaching to the tip of the rays just below the shortest ones at the
middle 21. Odontostilbe Cope.

Genus Grundulus'' Valenciennes.
Grundulus Valenciennes, Hist. Nat. Poiss., Vol. XVIII, 1846, p. 216.
Ctenocharax Regan, Ann. & Mag. Nat. Hist. (7), Vol. XX, 1907, p. 402.

Type: Poecilia bogotensis Humboldt.

Cyprinodontoid fishes, reaching a length of about 80 mm., found on the
eastern highlands of Colombia.

Teeth all recurved, conical, in a single series on greater part of maxillar}^, on
premaxillary and mandible; third suborbital in the largest specimens a little wider
than the naked area around its convex border, narrower than the naked area in
the young; postorbital area naked; mouth small, maxillary just about reaching
the orbit; eye comparatively small, placed high; a wide naked strip from dorsal
to head, and another along middle of belly; dorsal placed behind the middle; origin
of ventrals near the middle; anal short; no adipose fin; pectorals not nearly
reaching ventrals; lateral line short; scales with very many faint, diverging radial
strise; no pseudo tympanum. Gill-rakers about 8 + 10, very short, about one-third
as long as eye; a few feeble interhaemals appearing as caudal fulcra; a larger num-
ber of interneurals, but equally feeble.

1. Grundulus bogotensis (Humboldt). (Plate II, fig. 1.)
Native name "Guapuche."
Poecilia bogotensis Humboldt, Recherches sur les Poissons Fluviatiles, in Rec.
d'Observations, Zoologie et Anat. Comp., Vol. II, 1821, pp. 154 and 159,
pi. XLV, fig. 1.
Ctenocharax bogotensis Regan, Ann. and Mag. Nat. Hist. (7), Vol. XX, 1907, p.
403 (Bogota).
Range : The rivers of the plains of Bogota and northward in the streams of
Santander.

5084, C. M., 12843, I. U. M., over one thousand specimens, largest 80 mm.
North of Bogota at Puente de Suba. Eigenmann.

' Probably from, Grundules, an appellation of the Lares.

18

MEMOIRS OF THE CARNEGIE MUSEUM.

5085, C. M., 12844, I. U. M., eighty. Esperanza. Eigenmann.

5086, C. M., 12845, I. U. M., two. Electric liglit plant at edge of plains of
Bogota. Eigenmann.

5087, C. M., 1285G, I. U. M., one hundred and fifty. Madrid. Eigenmann.

5491, C. M., 13182, 1. U. M., nine, largest 70 mm. Boyaca, Rio Chiquinquiseto.
Gonzales.

5492, C. M., 13183, 1. U. M., three. Quebrada Zuaita, Santander. Gonzales.
Head 3.5-3.7; depth 3-3.3; D. 11, rarely 10; A. 13-16; scales in a median

series 32-34 of which 4 to 6 bear pores; eye 4-4.45 in the head, equal or nearly equal

Fig. 3. Grundulus bogotensis (Humboldt), a, maxillary; b, premaxillary; c, mandible.

to the snout, 1-1.33 in the interorbital. Compressed but heavy, dorsal and ventral
outlines equally curved; the head rather blunt. Preventral area rounded, naked
in the middle below the pectorals, the scales bordering the naked area rounded,
isolated from each other; area along base of dorsal narrowly naked, a naked
area in front of the dorsal as wide as the eye, growing wider toward the head and
extending on both sides of the occipital process; the scales bordering the area sub-
circular, small, isolated from each other; occipital process reaching about one
seventh to the dorsal; interorbital broad; frontal fontanel large, entirely separating
the frontals in the young, in contact for a variable length in the adult, frontal
fontanel being nearly as long or only half as long as the parietal. IMouth moderate,

eigenmann: the cheirodontin^. 19

the maxillary oblique, about equal to length of eye; seven to nine teeth in the pre-
maxillary, nine or ten teeth in the maxillary, about thirteen to sixteen teeth on
the dentary; tooth-bearing portion of dentary over half the length of the lower jaw.

Origin of dorsal about equidistant from preopercle and base of middle caudal
rays; margin of dorsal rounded, the fifth ray highest, httle if any longer than post-
orbital portion of head; caudal but feebly forked, its lobes short, about equal to
'ength of head less snout; origin of anal under posterior part of dorsal; base of
anal half as long as head, or a little 'onger; marg'n of anal nearly straight or
rounded in front, without lobe, the fifth ray highest; ventrals nearly reaching
anal, or a little beyond the base of the first ray; pectorals broad, reaching half-way
to middle of ventrals; origin of ventrals in middle, or a httle behind the middle of
the distance from snout to caudal.

Scales of middle of s'des large, decreasing rapidly towards the back and belly.
No scales on base of anal or caudal.

Sides reticulated with dark, basal angles of scales very dark; upper parts of
head and naked predorsal area blackish, an obscure humeral spot over and some-
times on the second and third scales of the lateral line; no caudal spot, a faint dusky

lateral band.

Genus II. Spintherobolus Eigenmann.^

Spintherobol'us Eigenmann, Ann.- Carnegie Mus., Vol. VIII, 1911, p. 167, PL V,
figs. 1-4.

Type: Spintherobolus papilliferus 'Eigenmann.

Teeth tricuspid, in a single series in dentary, premaxillary, and upper part of
maxillary; no adipose fin; lateral line on one or two scales; caudal naked; anal
very short, naked; tactile papillae excessively developed; suborbitals very feeble;
predorsal area fully scaled.

2. Spintherobolus papilliferus Eigenmann. (Plate III, figs. 1-4.)
Spintherobolus papilliferus Eigenmann, I. c.
Range : Upper Tiete basin.
3882, C. M., type 41 mm. 3883, C. M., paratypes, four. 25-39 mm. Alto

da Serra, Tiete basin, Sao Paulo, Aug. 4, 1908. Haseman.
Head 3.33; depth 3; D. 11; A. 12; scales 35, 13 between dorsal and ventral;
eye 4.3; interorbital 3.5 in the head.

Cyprinodontiform; profile sloping rapidly to above the ventrals; caudal
peduncle a trifle less than half the greatest depth, length of the peduncle twice its

8 (Tiri!'9ijpo/3oXos, emitting sparks, referring to the appearance of the yellow tactile organs of the head.

20

MEMOIRS OF THE CARNEGIE MUSEUM.

height; predorsal area with thirteen scales, preventral area short, rounded, without
a distinct median series of scales; occipital process about six times in the distance

from its base to the dorsal; frontal fontanel a very nar-
row sht; cheeks entirely naked; the suborbitals very
narrow, concealed;' mouth small, terminal; about six
arrow-shaped teeth on the maxillary; seven similar
more distinctly three-lobed teeth on the premaxillary;
eight similar and two conical teeth in each dentary;
organs of lateral line excessively developed about the
head, each papilla orange; no gill-rakers.

Origin of dorsal near middle of body, its highest ray
four and one -half times in the length; caudal lobes four
to four and one-half times in the length; anal very
short, its origin equidistant from preopercle and caudal;
ventrals reaching to the anal, as long as the base of the
latter; pectorals reaching beyond the origin of the ven-
trals. Pores of lateral hne on only two scales. Scales regularly imbricate, no inter-
polated rows; caudal and anal naked; axillary scale minute.

Scales of sides outlined in dark, a dusky spot over the scales with developed
pores.

This species, which resembles Grundulus in many ways, is known to occur only
in small rills leading into the Rio Tiete and thence into the Parana, at Alto da Serra
near the coast at Santos, Sao Paulo, Brazil. The related genera Grundulus and
Spintherobolus are thus known from nearly opposite parts of South America.

Fig. 4. Spintherobolus papil-
liferus Eigenmann. a, premaxil-
lary; b, maxillary; c, mandible.

Genus III. Probolodus^ Eigenmann.
Probolodus Eigenmann, Ann. Carnegie Mus., Vol. VIII, 1911, p. 164, fig. 1; PL
IV, fig. 1.

Type : Probolodus heterostomus Eigenmann.

Premaxillary with three teeth somewhat directed outward, each with three
points in the angles of a nearly isosceles triangle, the middle point, which is also
the anterior one, much heavier; maxillary with three to five teeth, the first two or
three of which are directed outward; each ramus of the mandible with four larger
teeth, the first three directed outward, the fourth and one or more smaller ones
following it, directed upward; the larger teeth of the lower jaw heavy, conical,
with a minute cusp on each side. Lateral line complete; caudal naked. Adipose

' 7rpo/3oXij = a putting forward; 65o6s = a tooth.

eigenmann: the cheirodontin^. 21

well-developed. Mouth large; lower jaw heavy, its sides vertical; third suborbital
in contact with preopercle below and behind in the larger, with a narrow naked
border in the smaller specimens; three postorbitals covering the entire postorbital
area; no pseudo tympanum ; frontal fontanel very long; the frontals not in contact.

3. Probolodus heterostomus Eigenmann.
Probolodus heterostomus Eigenmann, Ann. Carnegie Mus., Vol. VIII, 1911, p. 164,

fig. 1; PL IV, fig. 1. (Campos on the Rio Parahyba; Iporanga on the Ribeira.)

Range: Southeastern Brazil in coastal streams from the Rio Doce to the
Ribeira da Iguape.

2973, C. M., type, 63 mm.; 2974, C. M., paratypes, three, 48-64 mm. Campos
on the Parahyba. June 13, 1908. Haseman.

2975, C. M., paratypes, two, 78-81 mm. Iporanga on the Ribeira, Dec. 1, 1908.
Haseman.

20910, M. C. Z., one, 77 mm. Rio Doce, between Linhares and Porto Souza,
Hartt and Copeland. Thayer Expedition.

6882a, C. M., 45 mm. Sao Joao do Barra, Rio Parahyba. Haseman.

6879a-b, C. M., two, 37 and 44 mm. Jacarehy, Rio Parahyba, July 15, 1908.
Haseman.

6880a-g, C. M., seven, largest 35 .mm. Jacarehy, July 4, 1908. Haseman.

Head 4-4.33; depth 2.66; D. 11; A 26-31; scales 8 or 9-45 to 53-7 to ven-
trals; eye 3, interorbital 2.5, snout 3.5 in the head in the type, 2.5, 3.2, 3.5 respec-
tively in the paratypes.

Dorsal and ventral outlines nearly equally curved, without distinct humps or
depressions; ventral region rounded, the postventral area narrowly so. No regular
series of median scales in front of the ventrals. Predorsal area narrowly rounded,
with a regular median series of about twelve scales. Occipital process about five
times in the distance from its base to the dorsal, bordered by four or five scales
on each side; interorbital rounded, the frontal fontanel little more than half as
long as the parietal without the occipital groove. Snout sharp, the lower jaw
entering the profile when the mouth is closed; a distinct angle between maxillary
and premaxillary; maxillary not quite equal to length of eye; third suborbital
leaving a naked border around its entire lower edge. Gill-rakers 6 + 12.

Scales everywhere regularly imbricate, except over the anal musculature; caudal
naked, a weak anal sheath of one series of scales along the base of the anterior rays;
each scale of the side, with several (maximum about eight) radial striae; axillary
scale of the ventrals well -developed.

22 MEMOIRS OF THE CARNEGIE MUSEUM.

Dorsal small, about four times in the length; adipose well -developed; caudal
forked, the lobes three and one-half to four times in the length; anal slightly
emarginate, the highest ray reaching tip of thirteenth ray. Ventrals not quite
reaching anal, pectorals to, or a little beyond, origin of ventrals.

Fig. 5. Probolodus heleroslomus Eigenmann. a, outlines of side of head; b, top of head showing frontal
(/) and occipital (o) fontanels; c, dentition. C. M. 6881.

A large vertical humeral spot, chiefly above the third and fifth scale of the
lateral Hne; a silvery lateral band. A spot on caudal peduncle, in the young
definitely continued to the end of the middle rays.

Genus IV. Aphyocharax^" Glinther.
Aphyocharax Gtinther, Proc. Zool. Soc. London, 1868, p. 245 (pusillus).
Holoprion Eigenmann, Smithsonian Misc. Coll., Quarterly, Vol. XLV, 1903, p. 145
{agassizii).
Type : Aphyocharax pusillus Gtinther.

Elongate, slender fishes, reaching a maximum length of about 80 mm., most of
them much shorter.

'" a<t>bv, = a small fish, anchovj', sardine, or Motella; xapaf = a pointed stick, or paUsade; Charax, a
genus of characid fishes with pointed teeth.

eigenmann: the cheirodontin^. 23

Head pointed, skull convex above, with parietal and frontal fontanel, the
latter very short, not extending beyond middle of eye in adult; mouth terminal,
greatly variable in size in different species and with age in the same species (c/. A.
dentatus); teeth all in a single series, conical, and usually with a cusp on each
margin; those of upper and lower jaw alternating and interlocking when the mouth
is closed; those of the side of the lower jaw notably smaller than the anterior
teeth; cheeks entirely covered by the suborbital; nares close together; gill-mem-
branes free from each other; gill-rakers setiform; tongue slender, free; scales
regular, of about the same size, well imbricated, cycloid, firm; caudal naked; lateral
line incomplete; adipose well-developed; origin of dorsal near middle of body; no
pseudotympanum ; a long pore on the base of the middle caudal ray.

The genus Aphyocharax is well marked by the very strong armature of the
cheeks, the third suborbital and the single postorbital being of the same strength
and texture, and leaving but little or any of the cheek naked; the scales are firmer
than in any other genus of the subfamily and they have peculiar sculpturing.
There are two sub-parallel radial striae on the exposed part of the scales; the
circuli are well-marked near the base of the exposed portion of the scale and parallel
with the radial striae at this point, or converge toward a median line.

A. dentatus, alburnus, erythrurus, pusillus, and anisitsi form a series from the
large-mouthed dentatus with many maxillary teeth to the small-mouthed anisitsi
with few maxillary teeth. Of these A. alburnus and A. erythrurus are scarcely
distinct. A. dentatus is the Paraguayan representative, alburnus the Amazonian,
and A. erythrurus the Guianian. A. anisitsi similarly is the Paraguayan repre-
sentative, A. pusillus the Amazonian. A. melanotus probably belongs to another
genus. A. avary is probably a synonym of pusillus.

At the time A. agassizi was selected as the type of the new genus, Holoprion,
I had not seen that species. The genus was erected in view of Steindachner's
statement, "oberer Theil des Oberkiefers am ganzen vorderen Rande deutlich
gezahnt." While in Vienna, I was able to examine the types. It is certain that
they have several teeth along the upper anterior margin, but I was unable with
a hand-lens to detect any teeth along the distal portion of the maxillary. It
seems possible that the type species of the genus Holoprion lacks the character
assigned to the genus. However, there certainly are species with this character,
viz. : Paraguay ensis, maxillaris, and possibly nattereri, which may be referred to
Holoprion, provisionally retained as a subgenus. If a final microscopic examina-
tion of agassizi shows that it actually lacks teeth on the distal portion of the
maxillary the name Holoprion becomes an exact synonym of Aphyocharax, and a

24 MEMOIRS OF THE CARNEGIE MUSEUM.

new generic or subgeneric name will have to be given to the above named
species.

Nothing is said in the original description of the teeth on the maxillary of
Cheirodon pulcher Steindachncr. The description reads "Kieferzahne einreihig,
sehr klein, schlank und zahlreich." This does not apply to the genus Cheirodon,
in which genus it is placed by Steindachner, but applies very well to A. paraguayen-
sis, a species evidently very closely related to pulcher, but which has teeth along
the entire maxillary and belongs to the genus Holoprion as originally defined. It is
more than probable that pulcher also belongs here, and it is so ranked.

Range: Guiana, Amazon, and Paraguay, rare in the Uruguay basin.

Key to the Species op Aphyocharax.
a. Teeth not along the entire margin of the maxillary. (Aphyocharax.)
b. Dorsal without black.

c. Maxillary with many teeth; hooks in the male confined to the lobe of the anal; a humeral spot.

d. Mouth very large, the maxillary in fully grown specimens reaching to the third suborbital,

2.5 in the head; snout in the adult 3.5-3.8 in the head, but very little shorter than eye;

maxillary-premaxillary border 2.1-2.33 inhead; depth 3. 5-3.75; middle caudal rays pale.

4. dentatus Eigenmann & Kennedy.
dd. Mouth moderate, the maxillary not reaching the second suborbital.

e. Maxillary with 9-20 teeth extending over more than half the length of the maxillary;
scales 37-40.
/. Premaxillary with six to eight teeth, mandible with 16-20; middle caudal rays dusky;
a well-developed humeral spot; A. 17-19; depth 4.25-4.5; snout over 4 in the head;

maxillary 3.25 in head 5. alburnus (Giinther).

ff. Premaxillary with six teeth, mandible with about tliirteen; A. 17; middle caudal

rays pale, caudal red in life 6. erythrurus Eigenmann.

fff. Maxillary with teeth along about one-tliird to two-fifths its length; premaxillary
with seven teeth, mandible with nine; A. 18; scales 36, 8 with tubes; head 4.5;

depth 4.5; middle caudal rays black 7. pusillus Giinther.

cc. Maxillary with two to fourteeth, mouth small; depth 3-3.75; scales 5 or 5.5-30 to 35-4.5 to 6;
mandible with nine to ten teeth; hooks in the male on nearly all anal rays; no humeral spot.

8. anisitsi Eigenmann & Kennedy.
ccc. Maxillary with four teeth; mouth moderate, maxiOary reaching to below anterior margin of pupil;

depth 4.25; scales 38; A. 17 9. avary Fowler.

66. Dorsal, at least in male, with black. Mouth minute; posterior part of anal margined with black, the
black extending obliquely across the lobe; hooks in the male on nearly all anal rays; depth 3; A. 19
or 20; maxillary very short, convex, not reaching eye, with two teeth or none; premaxillary with

five to seven teeth; six to nine on the mandible 10. rathbuni Eigenmann.

aa. Teeth along the entire margin of the maxillary {Holu prion).

g. Dorsal plain; a black border or band from the tip of the last anal ray to the lobe, thence obliquely
to the anterior rays and forward to near the ventrals; no humeral spot; origin of dorsal behind
the middle; a black caudal spot.

eigenmann: the cheirodontin^.

25

h. Scales34;caudalspotnotcontmuedtotheendof the middle rays. .11. paraguayensis Eigenmann.
hh. Scales 30; caudal spot continued to the end of the middle rays. A. 23. 12. nattereri Steindaclmer.
gg. Dorsal with a black spot; no caudal spot.

i. A black band along the tips of the shorter anal rays and across the middle of the elongate rays ;

A. 27; scales 30 13. agassizi Steindaclmer.

ii. A small, black spot near the tips of the first few anal rays; A. 22-23; scales 30.

14. maxillaris Ulrey.

4. Aphyocharax dentatus Eigenmann & Kennedy. (Plate IV, fig. 1.)
? Aphyocharax {Chirodon) alburnus Perugia, Ann. Mus. Civ. Storia Nat. Genova,

(2), Vol. XVIII, 1897, p. 25 (Rio Beni; Missioni Mosetenes).
Aphyocharax dentatus Eigenmann and Kennedy, Proc. Acad. Nat. Sci. Phila., 1903,

p. 576. (Asuncion; Arroyo Trementina); Eigenmann, Ann. Carnegie Mus.,

Vol. IV, 1907, p. 126 (Corumba, Puerto Max; Rio Negro; Rio Pilcomayo);

Reports Princeton Univ. Exped. Patagonia, Vol. Ill, 1910, p. 429.

Range : Paraguay basin.

10032, I. U. M.,type, 71 mm. and 10033, 10036, 10037, 10038, I. U. M., ten.
Asuncion. Anisits.

10034 and 10035, I. U. M.; 932a & b, C. M. Aguadas near AxToyo Tremen-
tina. Anisits.

10290 and 10187, I. U. M., nine. R. Paraguay, Corumba. Anisits.

10190, I. U. M., one. Rio Paraguay, Puerto Max. Anisits.

10230, I. U. M., ten. Tributary of the Rio Negro emptying into Rio
Pilcomayo. Anisits.

6911a, C. M., one, 37 mm. Asuncion, Paraguay. Haseman.

6912a-q, C. M., seventeen, 23-72 mm. R. Paraguay, Caceres, May 26, 1909.
Haseman.

1913a-j, C. M., eleven, all females, largest 52 mm. Corumba, May 9, 1909.
Haseman.

6914a-e, C. M., five (one male) 56-64 mm. Corumba, April 27, 1909. Hase-
man.

7313a-j, C. M., ten, largest about 72 mm. Villa Hays, April 13, 1909. Hase-
man.

?7314a-c, C. M., three, largest about 25 mm. Santa Rita, June 12, 1909.

Haseman.

Head 4-3.8; depth 3.6-3.75 (3.4-4 in extremes); D. 11; A. 18-22; scales 5.5
to 7-36 to 42-4.5 to 6; 7 to 14 scales with pores; eye 3.5-4 in the head, 1 in snout;
interorbital 3-3.33 in the head.

26

MEMOIES OF THE CARNEGIE MUSEUM.

Elongate, compressed-fusiform ; dorsal and ventral profiles regularly and about
equally arched; preventral area broad, rounded, with a median series of about
seventeen scales; predorsal area rounded, with a median series of about fourteen
scales, the regularity of the series broken near the middle; occipital process reaching
about one-eighth to one-ninth to the dorsal, bordered by two to three scales; skull
smooth, slightly convex; fontanels narrow, the frontal fontanel reaching to above

Fig. 6. Aphijocharax dentatus Eigemnann & Kennedy, a, outline of head, 9 mm. long, showing size of
maxillary in a specimen 68 mm. long; 1, 2, 3, suborbitals, 4, 5, postorbitals. b, outline of top of head showing
frontal (/), and occipital (o) fontanels, c, premaxillary, d, maxillary, e, mandible, much enlarged, from a
specimen 28 mm. long, 10036, I. U. M.

the middle of the eye, its length about two and one-half times in the length of
the parietal fontanel without the groove; snout pointed, the mouth large, ter-
minal, the jaws equal; maxillary in adult reaching to the third suborbital; max-
illary-premaxillary border two and one-third times in the head, maxillary relatively
much shorter in the young; teeth recurved, slender, acutely pointed, those in the
front of the jaws with a minute cusp on each side; premaxillary with seven to ten
teeth; maxillary with nine to twenty on more than half of its edge; mandible with
thirteen to twenty-one teeth; teeth of the premaxillary and of the front of the
mandible all of about the same size; third suborbital in contact with the pre-
opercle below and behind; a single postorbital, similar to the third suborbital, its
face convex, its posterior margin in contact with the preopercle; upper part of the
opcrcle similar to, but smaller than the postorbital. Gill-rakers 7 + 9.

eigenmann: the cheirodontin^. 27

Origin of dorsal equidistant from tip of snout and base of middle caudal rays
or a little nearer the latter; dorsal truncate, tip of third to seventh ray about equal
when the fin is half closed; highest ray a little shorter than head; adipose fin small,
but well developed; caudal lobes about equal to length of head; origin of anal
under end of dorsal, last twelve rays of anal of about the same height, the third to
eighth forming a lobe two and a half times as high as the posterior rays, very
little longer than snout and eye ; ventrals not reaching anal by three or four scales,
their origin far in advance of the dorsal; pectorals not reaching ventrals by two to
three scales.

Scales everywhere very regularly imbricate, pores developed on seven to
fourteen scales and on the last scale of the series above the lateral line series, a
long tube on the base of the middle caudal ray; sheath of scales covering the
basal fifth of the caudal lobes, the last scale on each lobe largest ; anal with not over
four scales forming a sheath for the anterior rays ; axillary scale large ; scales usually
with two sub-parallel radials, the circuli on at least the basal part of the exposed
portion of the scale prominent above and below.

A diffuse, but quite evident, shoulder-spot; margin of caudal dusky, a silvery
lateral band; first dorsal ray dark, a few chromatophores on the membranes from
the middle of the first to the tip of the penultimate, forming a very faint, oblique
band; bases of some of the rays dusky.

Male with hooks on the third to the eighth anal ray; i. e., on the anal lobe,
which is equal to the snout and eye.

This is the largest species of the genus and is abundant in the Paraguay to
San Luiz de Caceres.

5. Aphyocharax alburnus (Glinther).

Chirodon alburnus Glinther,^' Proc. Zool. Soc. London, 1869, p. 424, fig. 2 (Peruvian

Amazons) .
Aphyrocharax alburnus Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., Vol.

XIV, 1891, p. 292; Ulrey, Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 292;

Eigenmann, Reports Princeton Univ. Exped. Patagonia, Vol. Ill, 1910, p. 429.

" Giinther's original description is very brief and reads:
"D. 10. A. 20. L. lat. 37. L. transv. 11.

Tlie hciglit of the body is a little more tlian the length of the head, and one-fourtii of the total (without
caudal) . Upper profile of the head not concave. The pectoral does not extend to the ventral. Teeth scarcely
compressed, pointed, with a minute (microscopical) lobe on each side; there are about twelve in the upper and
eighteen in the lower jaw. Sides with an ill-defined silvery longitudinal band; the middle caudal rays blackish.

Two and a half inches long."

28 MEMOIRS OF THE CARNEGIE MUSEUM.

Range : Amazon basin.

11087, I. U. M., one, 58 mm. Rio Jurua. From British Museum.

6915a-i, C. M., four females, 47-55 mm., six males, 34-43 mm. Santarem,
Dec. 6, 9, and 15, 1909. Haseman.

20713a-b, M. C. Z., two, 37 mm., Villa Bella. L. Agassiz. Thayer Expedition.

20813a-f, M. C. Z., six, largest 51 mm. Iga. WiUiam James. Thayer Expe-
dition.

20813a-d, C. M., four, largest 39 mm. San Joaquin, Bolivia, Sept. 4, 1909.
Haseman.

Head 4.2-4.5; depth 4.2-4.5; D. 11; A. ^, ^, ^; scales 39-40, of which 10-12
with pores; eye 3 in head, snout 4 in head, interorbital slightly larger than eye.

Maxillary always falling short of the third suborbital; maxillary-premaxillary
border three times in the head, in the largest specimen; premaxillary with six to
eight teeth; maxillary with ten to sixteen teeth on over half the length of the max-
illary; mandible with eight large teeth and eight small ones on the side.

Margin of caudal and the middle rays dusky. In all other respects the
description of dentatus applies to this species.

6. Aphyocharax erythrurus Eigenmann.
Aphyocharax erythrurus Eigenmann, Mem. Carnegie Mus., Vol. V, 1912, p. 313,

PI. XLIV, fig. 4 (Rockstone, Crab Falls, and Maripicru Creek, British Guiana.)

1879a, C. M., type. 1880a-e, C. M.; 12161, I. U. M., paratypes, 28-58 mm.
Rockstone. Eigenmann.

1881a, C. M., paratype, 29 mm. Maripicru Creek. Grant.

2494a, C. M., paratype, 35 mm. Crab Falls. Eigenmann.

Head about 4; depth 3.66-4; D. 10 or 11; A. 17 or 18; scales 5-34 to 37^^-3;
nine to eleven scales with pores; eye a little longer than snout; 3.5 in the head;
interorbital 3 in the head.

Maxillary-premaxillary border 2.5-2.75 in the head; six teeth in the pre-
maxillary, twelve to fourteen along the greater part of the maxillary, about thirteen
in the dentary. Middle caudal rays pale.

This species is almost identical with alburnus, but can readily be distinguished
by the pale middle caudal rays and the slightly longer maxillary.

7. Aphyocharax pusillus Giinther.
Aphyocharax pusillus Giinther, Proc. Zool. Soc. Lond., 1868, p. 245 (Huallaga;
Xeberos); Cope, Proc. Acad. Nat. Sci. Phila., 1871, p. 260 (Ambyiacu); Cope,

" And a few on the caudal.

eigenmann: the cheirodontinjE. 29

Proc. Am. Philos. Soc, Vol. XVII, 1878, p. 689 (Peruvian Amazon) ; Eigen-
mann & Eigenmann, Proc. U. S. Nat. Mus., Vol. XIV, 1891, p. 55; Ulrey,
Ann. N. Y. Acad. Sci., Vol. VIII, 1885, p. 292; ? Boulenger, Ann. & Mag.
Nat. Hist. (7) Vol. II, p. 478 (Jurua)i^; Fowler, Proc. Acad. Nat. Sci. Phila.,
1906, 333, fig. 22 (Ambyiacu, Peruvian Amazon.)
Range: Marailon basin; Madeira basin.
6917a, C. M. 50 mm. Palo Grande Fall, Rio Mamore, Sept. 30, 1909. Hase-

man.
6918a-b, C. M. two, larger 58, Berlin, Bolivia, Sept. 15, 1909. Haseman.
6919a, C. M. one, 43 mm. Maciel, Rio Guapore, July 2, 1909. Haseman.
Head 4.33-4.8; depth 4-4.3; D. 11 or 12; A. 18 or 19; eye 3-3.5 in the head,

1 in snout; interorbital 2.6-3 in the length of the head.

Maxillary-premaxillary border three times in the head, maxillary about four

times; maxillary with five to eight teeth on one-third to one-fourth of the margin;

premaxillary with seven teeth; mandible with ten to sixteen. Middle caudal rays

dark.

This species differs from alburnus in the number of maxillary teeth.

8. Aphyocharax anisitsi Eigenmann & Kennedy. (Plate III, fig. 6.)
Aphyocharax anisitsi Eigenmann & Kennedy, Proc. Acad. Nat. Sci. Phila., 1903,
p. 517 (Asuncion; Campo Grande; Arroyos Trementina and Chagalalina) ;
Eigenmann, Reports Princeton Univ. Exp. Patagonia, Vol. Ill, 1910, p. 429.
Range: Paraguay basin; Cacequy, Uruguay basin.

10024, 10026, I. U. M.; 936a-b, C. M., fifteen specimens, Campo Grande.
Anisits.

6920a-k, C. M., twelve (four males), largest 35 mm., Asuncion, March 28, 1909.

Haseman.
10028, I. U. M., type, 41 mm.; 10027 & 10029, I. U. M., three, Asuncion.
Anisits.

6921a, C. M., one, 37 mm. Puerto Suarez, Bolivia. Steinbach.
10031, I. U. M., one, Arroyo Trementina. Anisits.

10025, I. U. M., one, Arroyo Chagalalina. Anisits.

6222a-c, C. M., three, largest 32 mm. to base of caudal. Cacequy. R. Ibicuhy,

into Uruguay. Haseman.
Head 3.75-4.2; depth 3-3.75; D. 10 or 11; A. ^i-, ^, ¥, ¥, ¥; scales 30-35,

13 1 have specimens of alburnus from the Jurua sent by the British Museum. It is possible that these
are some of the specimens identified as pusillus by Boulenger.

30

MEMOIRS OF THE CARNEGIE MUSEUM.

6 to 9 with pores; eye 2.75-3.33 in the head, .5-.75 in the snout; interorbital 2.4
in the length of the head; interorbital a very little greater than eye.

About thirteen preventral scales, an equal number of predorsal scales; maxillary-
premaxillary border three times in the head, maxillary about two-thirds as long as the
eye. Premaxillary with six to eight teeth; maxillary with two to four, mandible
with nine to ten; mouth small, the maxillary scarcely reaching to eye; pectorals
reaching ventrals; origin of dorsal nearer caudal than to tip of snout.

Anal and caudal sometimes margined with dark; chromatophores on the
dorsal rays; outer pectoral ray sometimes dark; no humeral spot.

Anal in the male with four, rarely more, hooks on all the rays but the first

Fig. 7. Aphjocharax anisitsi Eigenmann. a, side of head; h, c, d, premaxillary, maxillary, and portion
of mandible of a specimen, 10029, I. U. M. 28 mm. to base of caudal.

two and the last, the hooks strongest on the posterior rays and on the middle third
of the rays.

While quite similar in general appearance to dentatus, this species differs in
many ways. It is smaller, deeper, has fewer scales, much fewer maxillary teeth,
a smaller mouth and different color. The anal in the male with hooks on practically
all the developed rays, is quite different from that of A. dentatus, which has hooks
on only the rays of the lobe.

9. Aphyocharax avary Fowler.
ApJiyocharax avary Fowler, Proc. Acad. Nat. Sci., Phila., 1913, p. 532.

Known only from the type, 54 mm. long, from the Madeira River about two
hundred miles east of Long. 62° 20', Brazil.

Head 3.87; depth 4.25; D. 13; A. 17; scales forty-two, of which eleven are

eigenmann: the cheirodontin^. 31

with tubes; eleven scales between dorsal and ventrals; nineteen predorsal scales;
depth of caudal peduncle 2.4 in the length of the head; height of dorsal 1.25;
height of anal 1.66; length of pectoral 1.33; ventral 1.6.

Elongate, slender, fusiform; maxillary reaching to below anterior margin of
pupil; teeth conic, each usually with a very small or obsolete pointed basal cusp;
maxillary with about four conic teeth.

Origin of dorsal midway between posterior nostril and base of caudal; origin

Fig. 8. Aphyocharax avary Fowler. (After Fowler, Proc. Acad. Nat. Sc, Phila., 1913, p. 532.)

of anal behind the vertical from the base of the last dorsal rays; origin of ventrals
well before the dorsal.

A silvery lateral band; humeral spot crossing third, fourth, and fifth scales of
the lateral line; dorsal with a transverse median streak. Anal broadly whitish in
front; rest of fin sprinkled with dusky dots, the edge of the fin dark, the dark con-
tinued across the anal lobe.

10. Aphyocharax rathbuni Eigenmann. (Plate III, fig. 5.)
Aphyocharax alburnus Eigenmann & Kennedy {non Giinther), Proc. Acad. Nat.

Sci., Phila., 1903, p. 517.
Aphyocharax anisitsi (part) Eigenmann and Kennedy, Proc. Acad. Nat. Sci.,

Phila., 1903, p. 517.
Aphyocharax rathbuni Eigenmann, Proc. U. S. Nat. Mus., Vol. XXXIII, 1907, p.

10. (Arroyo Chagalalina, Paraguay) ; Reports Princeton Univ. Exped. Pata-
gonia, Vol. Ill, 1910, p. 429.
Aphyocharax stramineus Eigenmann, Proc. U. S. Nat. Mus., Vol. XXXIII, 1907,

p. 11 (Arroyo Trementina); Eigenmann, Reports Princeton Univ. Exped.

Patagonia, Vol. Ill, 1910, p. 429.

Range : Paraguay basin.

32 MEMOIRS OF THE CARNEGIE MUSEUM.'

10025, I. U. M., one, 26 mm. to base of caudal. Arroyo Chagalalina, Para-
guay basin. J. D. Anisits. Type of rathbuni.

10030, I. U. M., one, 25 mm. to base of caudal. Arroyo Trementina. J.
D. Anisits. Type of stramineus.

A close comparison of rathbuni and stramineus in the light of the study of the
group of the Cheirodontince makes it quite probable that stramineus is the female
of rathbuni.

No additional material was collected by Haseman.

Head 4; depth 3; D. 11; A. 19 or 20; scales 35, 9 between dorsal and ventrals;
eye 2.6-3 in the head, equal to the interorbital ; snout about half length of eye.

Elongate, compressed; preventral area rounded, without a distinct median
series of scales, of which there are eleven between ventrals and the isthmus; pre-
dorsal area rounded, with about fourteen scales, of which about nine in front of
the dorsal are in a median series; occipital process very short, reaching about

Fig. 9. Aphyocharax rathbuni Eigenmann. Greatly enlarged.

one-seventh to the dorsal; frontal fontanel about half as long as the parietal with
its groove; second suborbital in contact with the lower limb of the preopercle
along its entire margin; a naked wedge behind it; postorbital strong, its margin
more convex in the male than in the female, having a naked area around its entire
distal margin; lateral line tubes on the postorbital and on the third suborbital
strong; mouth very small, the maxillary very short, its margin very convex, not
reaching the anterior margin of the eye; premaxillary with about five teeth, maxil-
lary with one to three; mandible with six to nine teeth, the first three rapidly
graduate.

Origin of dorsal a Uttle nearer base of caudal than tip of snout; origin of anal
under dorsal; anal with a slight lobe, the end of the fin rounded; most of the anal

eigenmann: the cheirodontin^.

33

rays of the male with a few hooks; ventrals reaching to or nearly to the anal;
pectorals not reaching to ventrals.

Seven or eight scales with pores; scales firm, mostly with two slightly di-
vergent radial striae; caudal naked, the scales extending but Uttle on the base of
the lobes, the last scale on each lobe large; anal naked.

No humeral or caudal spots; back thickly dusted; dorsal dusted like the back;
quite black in the male, especially along its base and posterior part; margin of
caudal dusky; margin of anal forward to the lobe black or dusky, a narrow dusky
band extending obliquely across the lobe to the basal third; opercle with a few
chromatophores, largest near its anterior margin; the color everywhere more
intense in the male. The color of the anal is evidently much like that of A. agassizi.

Closely allied to A. anisitsi, the mouth still smaller.

11. Aphyocharax paraguayensis sp. nov. (Plate II, fig. 2.)
6906a, C. M. Type, 25 mm. 6907, C. M., paratype, 20 mm. Rio Paraguay,

Caceres, May 24, 1909. Haseman.
Head 4; depth 3.5-3.8; D. 10 or 11; A. 22; scales 5 + 29; eye about three
in the head, a trifle less than interorbital.

Elongate, dorsal and ventral profiles alike; preventral area rounded, without
a distinct median series of scales, about fourteen rows; predorsal area rounded,

Fig. 10. Dentition of Aphyocharax paraguayensis Eigenmann. 6906, C. M.

with a perfect median series of thirteen scales; occipital process nearly equilateral,
reaching about one-seventh to the dorsal, bordered by one and one-half scales;
frontal fontanel reaching to near anterior margin of eye, but httle, if any, shorter
than the parietal without the groove; third suborbital in contact with the pre-

34 MEMOIRS OF THE CARNEGIE MUSEUM.

opercle in front and behind; a single postorbital, its lower half in contact with
the vertical limb of the preopercle, a naked margin behind its upper half; no tubes
on the suborbital or postorbital; mouth large, very oblique, the chin entering the
profile; the maxillary-premaxillary border a gentle curve, longer than eye; maxillary
reaching beyond suture between second and third suborbital; maxillary
with thirteen conical teeth along practically its entire margin; premaxillary with
six slightly graduate teeth with a small notch on each side near the tip of each
tooth; mandible with twelve graduate teeth along the greater part of its margin,
similar to those of the premaxillary, the lateral ones simply conic. Gill-rakers
4 + 7.

Origin of dorsal equidistant from tip of snout and near end of middle
caudal rays; the highest ray a little shorter than head; origin of anal under
anterior part of dorsal; ventrals not quite reaching anal, pectorals a little beyond
origin of ventrals.

Each scale with two sub-parallel radial striae dividing the scale into three
fields of about equal height; caudal naked; pores on but few scales.

A sub-triangular caudal spot, its base across the entire caudal peduncle, its
tip on the basal third of the middle caudal rays; dorsal nearly evenly peppered;
sides of head and abdomen silvery; chin, upper lip, and sides peppered, the pepper-
ing densest along the back and along the base of the anal, the spots becoming larger
along the base of the anterior anal rays, and continued forward to the ventrals;
a dark band beginning at the base of the first to fourth anal rays, extends
obliquely to the tip of the sixth and along the margin of the succeeding rays;
middle caudal rays without chromatophores, the lobes faintly peppered.

12. Aphyocharax nattereri Steindachner.
Tetragono'pterus diaphanus Cope (part), Proc. Am. Philos. Soc, Vol. XVII, 1878,

p. 691 (Peruvian Amazon).
Cheirodon pidcher Steindachner, Flussf. Slidam., Vol. IV, 1882, p. 39 (Villa Bella);

Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., Vol. XIV, 1891, p. 54;

Fowler, Proc. Acad. Nat. Sci. Phila., 1906, p. 332, fig. 21 (based on Cope's

specimens mentioned above).
Cheirodon nattereri Steindachner, Anz. Akad. Wiss., Wien, 1882, p. 180 (Villa

Bella); Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., Vol. XIV, 1891,

p. 54; Ulrey (part), Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 291 (Para);

?Boulenger, Boll. Mus. Univ. Torino, Vol. XIV, 1900 (Urucum, Matto

Grosso); Eigenmann, Reports Princeton Univ. Exped. Patagonia, Vol. Ill,

1910, p. 429.

eigenmann: the cheirodontin^. 35

Cheirodon steindachneri Eigenmann & Eigenmann, Proc. U. S. N. M., Vol. XIV,

1891, p. 54.

Range : Amazons from Para to Peru.

The names C. pulcher and C. nattereri were proposed for the same species, the
latter being substituted by Steindachner for the name pulcher, which was pre-
occupied. Without knowing that the specific name nattereri had been substituted
by Steindachner, the name steindachneri was also proposed by the writer for the
l^re-occupied pulcher. It is a pure synonym.

It is quite certain that some at least of Ulrey's specimens belong to the genus
Astyanax. It also seems doubtful whether the specimens figured by Fowler are
A. nattereri, since they distinctly differ in color. I have no specimens at hand; the
description of Steindachner may be in part reproduced :

[P. 39.] "Korpcrformsehrgestreckt, ^Zburrt^s-artig. Rlicken- und Bauchlinie
gleichformig, ausserst schwach gebogen.

"Seitenlinie unvoUstandig. Dorsale mit ihrem ersten Strahle eben so weit
von der Caudale wie vom hinteren Augenrande entfernt, somit nicht unbetrachtlich
weit hinter der Mitte der Korperlange beginnend. Ventrale vor der Mitte der
Korperlange eingelenkt.

"Grosste Korperhdhe c. 33^-32^ mal, Kopfliinge c. 3^-3^ mal in der
Korperlange enthalten," und der Schnauzenlange bis zur Kinnspitze gemessen wie
der Stirnbreite gleich.

"Kopf nach vorne zugespitzt. Mundspalte sehr schrage gestellt, Unterkiefer
nach vorne vorspringend. Kieferzahne einreihig, sehr klein, schlank und zahlreich.
Knochen des Augenringes die niedrige Wangengegend voUkommen deckend.

"Dorsale nach oben zugespitzt, an Hohe etwas der Kopflange nachstehend.
Pectorale bis zur Basis der Ventralen zurtickreichend, an Lange ein wenig geringer
als die Hohe der Rlickenflosse.

[P. 40.] "Ventrale mit ihrer Spitze den Beginn der Anale nahezu erreichend.

"Anale in ihrem vorderen Theile massig lappenformig erhoht.

" Schuppen klein, ziemlich festsitzend. Die Seitenlinie durchbohrt nur 4-6
Schuppen am Vorderrumpfe.

" Rumpf goldgelb. Humeralfieck ausnahmslos fehlend. Ein intensiv schwarz-
violetter, haufig rhombenformiger Fleck an und vor der Basis der Caudale, nach
hinten iiber die mittleren Caudalstrahlen bis zu deren hinterem Rande sich
fortsetzend.

" Evidently there is an omission here. Probably in the manuscript Dr. Steindachner had stated that
the eye is contained so many times in the head, but the statement was omitted by the printer.

36 MEMOIRS OF THE CARNEGIE MUSEUM.

"Eiii hellgelber Fleck am oberen und unteren Caudallappen unmittelbar
hinter dem Caudalfleck. Ein gleichfalls intensiv violetter Streif am Bauch ein
wenig hinter der Insertionsstelle der Ventralen beginnend und sich Ijings der
ganzen Basis der Anale hinziehend. Ein Nebenast dieses Streifens zieht, ein
wenig an Breite zunehmend (daher bindenahnlich) , von der Basis der 3-4 ersten
Analstrahlen schrage nach hinten und unten zum unteren Rande des 6. und 7.
Analstrahles und bildet hierauf einen schmalen Saum am freien Rande der folgenden
Analstrahlen.

"D. 9-10. A. 23. L. lat. c. 30. L. tr. 4/1/3.

"Zahlreiche Exemplare, nur bis zu 25-26 Mm. in der Totallange, von Villa
B ella ( Amazonenstrom) . ' '

13. Aphyocharax agassizi (Steindachner) .
Cheirodon agassizii Steindachner, Flussf. Slidam., Vol. IV, 1882, p. 38 (Jatuarana);

Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., Vol. XIV, 1891, p. 54.
Aphyocharax agassizii Ulrey, Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 293.
Holoprion agassizii Eigenmann, Smiths. Misc. Quarterly, Vol. XLV, 1903, p. 145;

Reports Princeton Univ. Exped. Patagonia, Vol. Ill, 1910, p. 429.

Range : Jatuarana, Amazon basin.

I have recognized no specimens of this species in the collections at my disposal
and add the original description:

" Korperform sehr gestreckt. Seitenlinie unvoUstandig; ein braunlichvioletter
Fleck am vorderen Theile der oberen Hohenhalfte der Dorsale, hoher als lang.

"Riicken- und Bauchlinie sehr schwach gebogen, erstere ein wenig rascher
zur Dorsale ansteigend, als letztere bis zur Ventrale sich senkt. Dorsale in der
Mitte der Korperlilnge und nur wenig hinter der Basis der Ventralen in verticaler
Richtung beginnend. Anale im vorderen Theile erhoht, lappenformig tiber den
Rest der Flosse vorragend. Humeralfleck sehr undeutlich; Caudalfleck fehlend,
Kopflange mehr als 33/3mal, grosste Rumpfhohe 3mal in der Korperlange.

" Augendiameter ctwas weniger als 3mal, Stirnbreite 33;^mal, Schnauzen-
lange gleichfalls 33/211^^^ "^ der Kopflange enthalten. Kieferzahne zahlreich,
schlank, verhaltnissmassig sehr klein, spitz, mit kurzen Nebenzacken, im Zwischen-
kiefer einreihig.

"Oberer Theil des Oberkiefers am ganzen vorderen Rande deutlich gezahnt.

"Obere Profillinie des Kopfes grade, nur wenig nach hinten ansteigend.

"Pectorale und Ventrale nach hinten zugespitzt; letstere tiberragt mit ihrer
Spitze den Beginn der Anale bei einem Exemplare nicht unbedeutend, erstere
erreicht nur die Basis der Ventralen.

eigenman: the cheirodontin^. 37

"Dorsale an Hohe einer Kopfliinge gleich, Ventrale um die Lange der Schnauze
kiirzer als der Kopf. Die Seitenlinie durchbohrt 7-8 Schuppen am Rumpfe.

"Der untere Rand der kurzen Analstrahlen ist dunkelviolett gesaumt, und
diese Farbung setzt sich strichformig horizontal nach vorn fort, so dass der vordere
erhohte Theil der Anale durch diesen violetten Streif der Hohe nach halbirt
erscheint. Der vordere lange Randstrahl der Anale (der dritte der ganzen Flosse)
zeigt eine milchweisse Farbung.

" Rumpf seiten goldgelb, silbergraue Seitenbinde nicht scharf abgegrenzt.

"D. 11. A. 27. P. 13 (14). V. 8. L. lat. 30 (bis zur Caud.) L. tr. 5/1/3.

"Zwei Exemplare, jedes c. 40 mm. lang, von Jatuarana vmd ein Geschenk des
Herrn Prof. L. Agassiz, dessen Andenken ich diese interessante Art widme."

14. Aphyocharax maxillaris Ulrey.
ApJiyocharax maxillaris Ulrey, Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 293

(Brazil).
Holoprion maxillaris Eigenmann, Reports Princeton Univ. Exped. Patagonia, Vol.

Ill, 1910, p. 429.

Range: Brazil, definite locality not known, probably from the lower Amazon.

This species is known only from the types.

Head 3.5; depth 3-3.5; D. 11 ; A. 22-23, scales 6 + 24. Eye 2.33 in the head.

Snout very short, the maxillary extending beyond anterior margin of eye;
premaxillary with about ten teeth, the median four three-pointed; mandible with a
few conical teeth.

Origin of dorsal equidistant from tip of snoutand base of caudal; pectorals
reaching beyond origin of ventrals ; the latter to the anal.

A small circular humeral spot, sometimes reduced to two or three color-cells.
A large black spot on the upper half of the first dorsal rays, the tips of these rays
white; a small black spot near tip of first few anal rays.

Genus V. Prionobrama^^ Fowler.
Prionobrama Fowler, Proc. Acad. Nat. Sci. Phila., 1913, p. 534 (madeirce).
Bleptonema Eigenmann, Indiana University Studies, No. 20, 1914, p. 44 {paraguay-

ensis) .

Type: Prionobrama madeirce Fowler = Aphijocharax filigerus Cope.

General appearance of Gephyrocharax. Teeth tricuspid and conical in a single
series on mandible, premaxillary, and along entire edge of maxillary; origin of

^^irpiuv = a saw; ^pa/ia = a bream.

38 MEMOIRS OF THE CARNEGIE MUSEUM.

dorsal behind the middle, about over the vertical from the origin of the anal;
adipose fin well -developed; anal falcate, with fewer than forty rays, its origin
nearer caudal than snout, first and second developed rays filigerous, curved;
pectorals placed low, long and falcate, their margins nearly along the edge of the
compressed belly when the fin is closed; caudal naked; lateral Une incomplete;
mouth very oblique; profile from tip of snout to near dorsal straight; two post-
orbitals, similar to those in ApJnjocharax, the lower large, the upper minute, some-
times a minute triangular wedge between the third suborbital and the postorbital,
representing the lower postorbital of other genera; third suborbital covering the
entire cheek; no well-defined pseudo tympanum.

Key to the Species op Phionobram.-v.

a. Lateral line 35-38; eight or nine scales between ventrals and dorsal; anal rays 29-34.

15. paraguayensis (Eigenmann).
aa. Lateral line 38-41; nine or ten scales between ventrals and dorsal; anal rays usually 32.

16. fiUgera (Cope).

15. Prionobrama paraguayensis (Eigenmann). (Plate IV, fig. 5.)
Bleptonema paraguayensis Eigenmann, Indiana University Studies, No. 20, 1914,

p. 44 (Corumba).

Range : Uruguay and Paraguay basins.

6884a-b, C. M., two, largest about 50 mm. Corumba, May 9, 1909. Haseman.

6885a, C. M., one, about 42 mm. Puerto Suarez, May 6, 1909. Haseman.

6886a & b, C. M., two, largest 45 mm. Villa Hays, April 11, 1909. Haseman.

6887a-c, C. M., three, largest 42 mm. Asuncion, March 29, 1909. Haseman.

6888a-j, C. M., ten, largest 49 mm. Arequa, April 8, 1909. Haseman.

847, M. C. Z., four, largest 35 mm. Uruguay River. Wyman.

5499, C. M., type, 40 mm. to base of caudal; 5499, C. M., paratypes, six,
largest over 50 mm. Corumba, April 27, 1909. Haseman.

Head 4.66-5; depth 3.33; D. 10, rarely 11; A. 29-34; lateral line 8 to 11 + 26
to 28 = 35 to 38; eight or nine scales between ventrals and dorsal; eye 3 in the
head, about equal to interorbital.

Elongate, compressed, breast with a series of large median scales; belly
between pectorals and ventrals trenchant, the margins of the scales of one side
bent over the middle line, no median series of scales; predorsal with a median
series of about sixteen scales, the series less regular near the occipital process,
which extends about one -eighth of the way to the dorsal; skull smooth, convex,
frontal fontanel large, triangular, a little more than half the length of the parietal;

eigenmann: the cheirodontin^. 39

second suborbital in contact with both the posterior and the lower limb of the pre-
opercle; mouth very oblique, maxillary-premaxillary border a little more than
orbital length; sixteen to twenty teeth on the maxillary, those on the posterior
half larger, pointing backward and outward; seven premaxillary teeth; mandible
with six or seven tricuspid teeth, the first and last distinctly larger than those
between ; in addition to these six or seven larger teeth several minute teeth appear
posteriorly.

Origin of dorsal a little nearer to base of middle caudal rays than to the eye;
origin of anal in front of or under the dorsal; first developed anal ray heavy and
much prolonged, the second less so; outer ventral raj^ filiform; extending beyond
the origin of the anal, pectorals to near middle of the ventrals; pectorals placed
low, their base oblique, their shortest ray about one-half of the outer ray, which is
similar to the outer ventral and first developed anal ray.

Scales everywhere regularly imbricate, with few radial striae; a series of very
small scales along the base of anal, caudal with a few scales at the base of the
lobes; axillary scales small.

No definite markings, margin of anal in the male dusky.

16. Prionobrama filigera (Cope). (Plate IV, fig. 4.)

Aphyocharax filigerus Cope, Proc. Am. Philos. Soc, 1870, p. 564 (Pebas); Eigen-
mann & Eigenmann, Proc. U. S. Nat. Mus., Vol. XIV, 1891, p. 55; Fowler,
Proc. Acad. Nat. Sci. Phila., 1906, p. 334, fig. 23 (Pebas); Eigenmann, Reports
Princeton Univ. Exped. Patagonia, Vol. Ill, 1910, p. 429.

Paragoniates ynulleri Steindachner, Ichthyol. Beitr., Vol. V, 1876, p. 72 (Obidos).

Prionohrama viadeirm Fowler, Proc. Acad. Nat. Sci. Phila., 1913, p. 534, fig. 9.
(Tributary of Rio Madeira near Porto Velho, Brazil.)

Bleptonema amazoni Eigenmann, Indiana University Studies, No. 20, 1914, p. 44
(Santarem) .

Aphyocharax analis Nichols, Bull. Am. Mus. Nat. Hist., Vol. XXXIV, 1915, p. 127
(Manaos).

Range : Amazons from Pebas to Villa Bella, Madeira basin.
5499a, C. M. Type of amazoni, 54 mm. 5599a-c, C. M., paratypes, 50-55 mm.

Santarem. Dec. 9, 1909. Haseman.
21188, M. C. Z.,"' about 58 mm. Villa Bella, Jan. 1866. L. Agassiz.
21182, M. C. Z., twelve, largest 53 mm. Villa Bella, Jan. 1866. L. Agassiz.
21248, M. C. Z., seven, largest 51 mm. I^a. Jan. 1866. William James.
^^ Anal filament reaching to base of last anal ray. A. 30; l.I. 39.

40

MEMOIRS OF THE CARNEGIE MUSEUM.

21231, M. C. Z., two, larger 50 mm. Hyavary. D. Bourget.
21234, M. C. Z., twenty -one, largest 55 mm. Tabatinga. Bourget.
6889a, C. M., one, 50 mm. Santarem, Dec. 9, 1909. Haseman.
6890a-q, C. M., seventeen, largest 50 mm. Santarem, Dec. 15, 1909. Haseman.
6891a-n, C. M., fourteen, largest 60 mm. Villa Bella, Oct. 5, 1909. Haseman.
6892a-c, C. M., one, 54 mm. San Joaquin, Sept. 6, 1909. Haseman.
6923a, C. M., one, 50 mm. San Antonio de Rio Madeira, Nov. 3, 1909. Hase-
man.
This species is very similar to paraguayense.

/IM^ZJA

Fig. 11. Prionohrama filigcra (Cope), a, side of head; h, top of head; c, d, e, the premaxillary, maxillary
and portion of mandible; c', premaxillary of the other side; e', outline of entire mandible showing the extent of
dentigerous portion, /, another mandible.

The scales are eight to fourteen (usually eleven) + twenty-six to thirty =
thirty-eight to forty-one, nine or ten between ventrals and dorsal; the origin of the
dorsal is equidistant from the base of the middle caudal rays and the anterior margin
of the eye, above the origin of the anal.

Number of anal rays: 31,
Number of individuals : 3

Sj^, 35, 36,
2 1 1

37
1

Genus VI. Paragoniates^'^ Steindachner.
Paragoniates Steindachner, Ichthyol. Beitr., Vol. V, 1876, p. 69.

Type : Paragoniates alburnus Steindachner.

Teeth mostly tricuspid, the lateral cusps minute, very much smaller than
the median cusp; teeth all in a single series, those of the premaxillary anteriorly

" irapa = besides; dTwi-idrT;?, o = a nervous person; Agoniates, a genus of fishes.

eigenmann: the cheirodontin^. 41

crowded, the second, and sometimes the fourth, more or less crowded out of Hne
with the rest, suggesting an incipient second series; teeth along the entire edge
of the maxillary. Origin of dorsal behind the middle; origin of anal about in the
middle; adipose fin well -developed; anal fin without a lobe, its margin nearly
str ight; pectorals placed low, foliate, their margins nearly along edge of belly,
when the fin is closed; caudal naked. Lateral line incomplete. Mouth large,
oblique. Three postorbitals, covering entire postorbital area, the middle one the
largest.

17. Paragoniates alburnus Steindachner. (Plate IV, fig. 2.)
Paragoniates alburnus Steindachner, Sb. Akad. Wiss. Wien, LXXIV, Ichthyol.

Beitr., Vol. V, 1876, p. 69, pi. VIII, fig. 3 (Teffe); Boulenger, Proc. Zool.

Soc, 1887, p. 281 (Canelos); Eigenmann, Report Princeton Univ. Exped.

Patagonia, Vol. Ill, 1910, p. 441.

Range : Amazon basin above Teffe.

7315a-b, C. M., two, 70 arid 86 mm. Villa Bella, Oct. 5, 1909. Haseman.

7316a, C. M., one, 73 mm. Santarem, Dec. 15, 1909. Haseman.

Steindachner's diagnosis of this species reads:

"Mundspalte sehr lang, hinteres Ende des voUstandig bezahnten Oberkiefers
bis hinter die Augenmitte bei geschlossenem Munde in verticaler Richtung sich
erstreckend. Wangen voUstiindig von den Knochen des Suborbitalringes liber-
deckt. Korperhohe 2-3/4mal, Kopfliinge mehr als 4-l/3mal in der Korperliinge,
Schnauzenlange circa 3-2/3mal, Stirnbreite circa 3mal, Augendiameter circa 3mal,
Liinge der Mundspalte zwischen l-3/4-l-4/5mal in der Kopflange enthalten.
Pectorale lang, iiber die Insertionsstelle der Ventrale, letztere iiber den Beginn der
Anale zurlickreichend. Fettflosse sehr klein; Dorsale in verticaler Richtung circa
iiber dem 7. oder 8. Analstrahle beginnend. Seitenlinie noch vor dem Beginn
der Anale endigend. Kiemenspalte lang; Verbindungshaute der Kiemenstrahlen
mit dem Isthmus nicht verwachsen und unter der Kehle noch gespalten. Rechen-
zahne der Kiemenbogen schlank, locker gestellt. Ein grosser, nicht scharf ausge-
priigter, braunlicher Fleck am Schwanze."

Head 4.75; depth 3; D. 11; A. 44-48; scales 6 or 7-46-4 or 5, thirteen to
seventeen with pores; eye a little greater than snout, 3 in the head, equal to, or a
little less than, the interorbital; caudal peduncle deeper than long.

Compressed; the ventral profile nearly regularly arched from chin to end of
anal; dorsal profile more gently and less regularly arched; preventral area com-
pressed, trenchant, the scales of the two sides narrowly bent over the mid-ventral
ridge; about twenty-nine predorsal scales; a regular median series extending from

42 MEMOIRS OF THE CARNEGIE MUSEUM.

the dorsal some distance forward, but the scales irregular anteriorly; occipital
process short, only one-eighth of the distance from its base to the dorsal, bordered
by three scales; frontal fontanel reaching at least to above the anterior margin of
the pupil, entirely separating the frontals in the smaller specimen; snout pointed;
the mouth large; maxillary slender, extending beyond the suture between the
second and third suborbitals; premaxiUary with seven teeth; maxillary with about
twenty teeth along its entire edge, tricuspid in the largest specimen, in the smaller
the upper ones similar to those of the premaxiUary, the lower ones conical; dentary
with seven to nine tricuspid teeth and about fourteen minute ones on the sides,
conical in the smallest specimen, mainly tricuspid in the largest. Gill-rakers 7 + 11.

Origin of dorsal equidistant from tip of snout and a point beyond the tip of
the middle caudal rays; height of dorsal about four times in the length; adipose
fin small; caudal forked, the lobes 3.5-4 in the length; origin of anal equidistant
from the base of the last ray and some portion of the eye; anal without indication
of a lobe, its margin slightly convex or straight; ventrals reaching beyond origin
of anal; pectorals beyond base of ventrals.

Lateral line developed on thirteen to seventeen scales; scales with one to four

radial striae; scales of the sides continued downward to form an anal sheath of

from one to three series of scales free from the fin; scales on the end of the caudal

peduncle irregularly arranged; caudal naked; a round or oval spot on the end of

the caudal peduncle, not continued on the sides or on the fin. No other definite

markings.

Genus VII. Leptagoniates^^ Boulenger.

Leptagoniates Boulenger, Proc. Zool. Soc, 1887, p. 281.

Type : Leptagoniates steindachneri Boulenger.

PremaxiUary, maxillary, and mandible with a single series of tricuspid teeth;
origin of anal far in advance of origin of dorsal, equidistant from tip of snout and
base of last dorsal ray; lateral line complete; adipose fin small.

I am in doubt whether in the final analysis this genus will remain associated
with the Cheirodontinse.

18. Leptagoniates steindachneri Boulenger. (Plate IV, fig. 3.)
Leptagoniates steindachneri Boulenger, I. c, p. 281, pi. XXIII, fig. 3 (Sarayacu,

Peru); Eigenmann, Reports Princeton Univ. Exped. Patagonia, Vol. Ill,

1910, p. 441.

This species is known only from the type, 95 mm. long, in the British Museum.
I append the description given by Boulenger, I. c. :

18 XexTos = thin; Agoniates, a genus of fishes.

eigenmann: the cheirodontin^e. 43

"D. 10; A. 70; V. 8; P. 12; scales 7-47-7.

"The depth of the body is one-fourth of the total length (without caudal),
the length of the head one-sixth. Mandible strongly projecting beyond the
mouth; maxillary not reaching below the anterior border of the eye; premaxillarj^
teeth 15, maxillary (on each side) 11, mandibular 14; mandibular teeth largest,
maxillary smallest. The diameter of the eye equals nearly two-fifths the length
of the head, and exceeds the width of the interorbital space. The pectoral fins
reach nearly the extremity of the ventrals, which are small; the dorsal originates
above the 23d anal ray. Colourless; sides of head and a lateral band above the
lateral line silvery."

Genus VIII. Phanagoniates Eigenmann and Wilson. ^^
Phenagoniates Eigenmann & Wilson, Indiana University Studies, No. 19, 1914, p. 2.

Type : Phanagoniates wilsoni Eigenmann.

Mouth minute, teeth in a single series in each jaw, tricuspid, except in posterior
part of maxillary, where they are conical, gill-opening wide, much compressed; chest
not trenchant; pectorals large, reaching to middle of ventrals; anal very long, its
origin far in advance of the dorsal; dorsal a little behind middle of body. No adi-
pose fin, lateral line incomplete.

19. Phanagoniates wilsoni Eigenmann. (Plate V, fig. 1.)
Phenagoniates wilsoni Eigenmann, Indiana University Studies, No. 19, 1914, p. 2.

5354, C. M., type, 41 mm.; paratypes, 13030, 1. U. M. 7, 21-38 mm. Manigru.
Charles Wilson.

5355a, C. M., paratype, 30 mm. Certegui. Charles Wilson.

5356a, C. M., 13031, I. U. M., paratypes, 30 and 37 mm. Rio Truando.
Charles Wilson.

Head 4.6; depth 3.33; D. 9; A. 53-55. Scales 7-7 + 34-7 (9 + 35 in one),
Eye 2.2 in the head.

Much compressed, dorsal profile highest at origin of dorsal, ventral profile
deepest at origin of anal; preventral area rounded, without distinct median series
of scales; occipital process about as broad as long; occipital fontanel much wider
and twice as long as the parietal, cheeks narrow and long, entirely covered by the
second suborbital; the mouth very small, the maxillary not reaching to the eye;
lower jaw with nine teeth on each side, premaxillary with six, maxillary with eight,

" 'Pami = bright. The name should have been printed Phanagoniates in my paper, Indiana University
Studies, No. 19, and I herewith correct the spelling.

44 MEMOIRS OF THE CARNEGIE MUSEUM.

the first four tricuspid, the last four conical. The first four forming a continuous
series with those of the premaxillary the four conical ones on the distally curved
portion of the bone. Gill-rakers about 5 + 8.

Scales thin, the margins obscure, a single row of scales along the base of the
anal rays; a few scales along the base of the caudal lobes.

Origin of dorsal slightly behind the middle of body; height of dorsal equal
to length of head; caudal lobes about equal to the height of the dorsal; origin
of anal about equidistant from the snout and origin of its last third; ventrals small,
reaching the anal; pectorals equal to the head less snout, reaching to middle of
ventrals.

Translucent, a median dusky band along middle of caudal peduncle, fading
out forward and continued narrowly on middle caudal rays; scales of back faintly
marked with chromatophores, which become restricted to the margin of the scales
on the upper part of the back. Chin and maxillary black, sometimes a dark streak
back from upper part of gill-opening.

Genus IX. Parecbasis-" Eigenmann.
Parecbasis Eigenmann, Indiana University Studies, No. 20, 1914, p. 45.

Type : Parecbasis cydolepis Eigenmann.

Teeth tricuspid, in a single series on anterior part of mandible and premaxillary,
none on the maxillary or sides of the mandible; upper margin of the maxillary
for its entire length heavy, then passing abruptly into a broad, thin, blade-like
expanded portion extending to the convex free margin; sides of mandible raised;
adipose fin well-developed; caudal partly scaled; lateral line complete. Origin
of the dorsal in front of the middle. Cheeks entirely covered by the third sub-
orbital, no apparent pseudo tympanum ; frontal fontanel very short.

20. Parecbasis cyclolepis Eigenmann. (Plate V, fig. 2.)
Parecbasis cyclolepis Eigenmann, Indiana University Studies, No. 20, 1914, p. 45.
Range : Madeira and Amazons.

5495, C. M., type, 74 mm.; 5496, C. M., paratype, 80 mm. San Antonio de

Rio Madeira, Nov. 3, 1909. Haseman. (Two other specimens from the

same place taken by the same collector at the same time have been placed

in exchange in the Museum of the Indiana University.)

6893a, C. M. 50 mm. to end of lateral line. Santarem, Dec. 9, 1909. Haseman.

6894a-c, C. M., 3, largest 78 mm., San Joaquin, Sept. 6, 1909. Haseman.

^^ Tapkaffaats, v = a, goiiig out aside from.

eigenmann: the cheirodontin^.

45

Head 4; depth 2.75-3; D. 11; A. 24-26; scales G or 7-38 to 40-5; eye 2.75-3 +
in the head, equal to interorbital.

Compressed, fusiform in outline, the dorsal and ventral outlines equally
symmetrically curved; preventral area rounded, with a nearly complete median
series of fourteen scales, predorsal area narrowly rounded with a median series of
ten scales; occipital process bordered by four scales, extending one-fourth to dorsal;
skull convex in cross-section; frontal fontanel 2.25 in the parietal; third suborbital
in contact with both the vertical and horizontal limb of the preopercle, but leaving

Fig. 12. Parecbasis cyclolepis Eigenmann. a, outline of head; b, outline of top of head, showing frontal
(J) and occipital (o) fontanels; c, premaxillary; d, dentary, teeth greatly enlarged; e,f, g, premaxillary, maxillary,
and mandible in outline, moderately enlarged.

a narrow angle below its anterior edge ; maxillary reaching to just below eye or not
quite so far; mouth clupeoid, the premaxillary transverse, without an antero-
posterior extent; teeth minute, confined to the premaxillary and but little more than
the portion of the mandible in contact with it when the mouth is closed.

Origin of dorsal equidistant from tip of snout and end of adipose or a little
nearer the latter; dorsal falcate, its highest ray exceeding length of head; anal

46 MEMOIRS OF THE CARNEGIE MUSEUM.

emarginate, its origin below some part of the last dorsal ray, its base greater than
length of head; origin of ventral below origin of dorsal, just reaching anal, or a
little shorter; pectorals short,. just about reaching ventrals.

Scales thin, the margins convex, with many radial striae; lateral line but little
decurved; anal naked; caudal lobes scaled for one-fourth to one-third of their
length.

A small, but conspicuous humeral spot, about equal to the size of the pupil,
over the fourth scale of the lateral line; middle caudal rays faintly peppered, or a
faint dusky streak parallel with the margin in the middle and the upper lobe of
the caudal, the rays beyond them dotted.

Genus X. Leptobrycon^^ gen. nov.

Type : Leptobrycon jatuaranoe Eigenmann.

Very similar to Parecbasis. Anal short, the highest ray, the fourth, extending
beyond the tip of the last; lateral line short; mouth very large, the premaxillary
very feeble, the maxillary very large, the upper margin thickened; lower jaw
scoop-shaped, the sides raised; teeth numerous (fourteen in the premaxillary),
feeble, conical, none on the maxillary, or on the raised part of the mandible; cheeks
partly naked; postorbitals three, covering most of the postorbital area; no pseudo-
tympanum; both fontanels large; adipose well-developed.

21. Leptobrycon jatuaranae Eigenmann sp. nov. (Plate VI, fig. 1.)
20952, M. C. Z., type, 29 mm. to base of caudal. Jatuarana. Navez.
Head 3.75; depth 3.5; D. 11; A. 14; scales ?. Eye 2.5 in the head con-
siderably larger than the interorbital; occipital reaching about one-sixth to the
dorsal; frontal fontanel much narrower than the parietal, its length about 1.25
in the length of the parietal; interorbital nearly flat; maxillary very large, slightly
longer than the eye, reaching to below the pupil, its front margin quite convex, its
proximal margin slightly concave and thickened; premaxillary altogether transverse,
each premaxillary with about fourteen minute, apparently conical teeth, without
lateral notches. Mandib!e with minute teeth on its anterior edge, the side of the
jaw upturned. Third suborbital narrowly in contact with the preopercle below.
A small naked angle below its anterior edge. A narrow naked strip behind it.
Gill-rakers 8 -1- 17, long and slender, the longest more than half length of eye.

Origin of dorsal equidistant from tip of snout and base of caudal; origin of
anal behind the vertical from last dorsal ray; base of anal not much longer than eye,

■'■ XeiTTos, small, or delicate; Drycon, a related genus of the Characidse, from /Spi/tw, to eat greedily.

eigenmann: the cheirodontin^. 47

its highest ray reaching beyond tip of last ray; ventrals reaching anal; pectorals
small, not reaching ventrals, whose origin is below origin of dorsal. Scales with
numerous parallel radial stria?; caudal naked, lateral line incomplete. Scales
mostly lost.

Color uniform; a silvery lateral stripe.

This species is readily distinguished from uruguayance by its numerous teeth
and the very short anal.

This specimen is very small and in bad repair, but its characters are so well
marked that there will be no difficulty in recognizing the species.

Genus XI. Aphyodite^^ Eigenmann.

Type : Aphyodite grammica Eigenmann, Mem. Carn. Mus., Vol. V, 1912, p. 314.

Teeth minute, conical, or but feebly notched, seven on the premaxillary,
fifteen or more on the mandible, none on the maxillary; premaxillary feeble, maxil-
lary considerably larger, its outer margin convex; its inner margin concave and
slightly thickened; sides of mandible much raised; adipose fin well developed;
caudal lobes scaled to near their tips; cheeks covered by the third suborbital;
postorbitals in three or four pieces, covering about half the width of the postorbital
area, the tubes prominent; frontal fontanel about half as long as the parietal; no
pseudotympanum ; anal long.

Distinguished from the related genera, Macropsobrycon, Megalamphodus, and
in fact from all the other genera of the subfamily by the scaled caudal.

22. Aphyodite grammica Eigenmann.
Aphyodite grammica Eigenmann, Mem. Carnegie Mus., Vol. V, 1912, p. 314, PL

XLIV, fig. 5 (Konawaruk).

Range : British Guiana.

1882, C. M., type, 32 mm. ; 12162, I. U. M., paratypes, two, 30-32 mm. Kona-
waruk, Middle Essequibo, British Guiana.

Head 4.5; depth 3.33; D. 11; A. 22; scales 4-7 + 23-3. Eye twice as
long as the snout, 2.5 in the head; interorbital a little less than the eye.

Compressed, slender. Head short, compressed, mouth small, oblique; the
maxillary not reaching to below the eye, about two-thirds as long as eye; cheeks
small, entirely covered by the suborbital below, a naked angle below its anterior
margin and a naked border behind it; maxillary margin convex, with three scarcely
perceptible teeth, or none. Anal emarginate; ventrals not reaching anal; pectorals

^' a4>vri, a small fish; SItt) with the force of dite in Aphi-odite, i. e., born of, descended from.

48 MEMOIRS OF THE CARNEGIE MUSEUM.

not to ventrals. Scales of the back margined with dark; a black median line;
some black at base of ventrals and base of anal.

Genus XII. Macropsobrycon-^ gen. nov.

Type: M acropsobry con Uruguay ance mgenmsinn.

Related to Parecbasis, but having even fainter dentition, and an incomplete
lateral line; related to Aphyodite, but with a naked caudal and a well developed
pseudo tympanum; also to Leptobrycon, but with a long anal. It is possible that
some of the species of Megalamphodus should be placed in this genus.

Teeth minute, conical, or but few of them with a lateral notch; six to eight in
the premaxillary, five or six in the front part of the mandible, none on the side of
the lower jaws, none on the maxillary, premaxillary feeble, maxillary several times
as large, nearly as long as eye, its outer margin very convex, inner margin concave,
not thickened; sides of mandible much raised; adipose fin well-developed; caudal
apparently naked; cheeks entirely covered by the third suborbital; postorbitals
covering abovit half the width of the postorbital space; frontal fontanel short or
medium; a well marked pseudotympanum.

23. Macropsobrycon uruguayanae, sp. nov. (Plate VI, fig. 2.)

6895a, C. M., type, 46 mm., paratypes 6896a-d, C. M., four, about 45 mm.
Feb. 1, 1909. Haseman.

6897a, C. M., 26 mm. Uruguayana. Haseman.

Head 4.5; depth 3; D. 11; A. 23-25; scales about 30-33, of which about five
are with pores. Eye 3 in head, greater than the very convex interorbital.

Dorsal and ventral profiles both regular, without humps or depressions, the
ventral profile a little more arched than the dorsal; preventral area with about
thirteen scales; predorsal area narrowly rounded. with a median series of about
thirteen scales. Occipital process reaching about one-sixth to one-seventh to the
dorsal; frontal fontanel as broad, and less than half as long, as the parietal; the
skull very convex, the third suborbital in contact with lower limb of the pre-
opercle; without a naked angle under its anterior margin, leaving a wide naked
wedge behind its posterior margin; postorbitals narrow and feeble, leaving about
half of the postorbital area naked. Gill-rakers 7 + 15 to 17, very long and slender,
the longest more than one-half the eye.

Premaxillary very feeble, one specimen examined has six slightly graduate
conical teeth, of which only one has a slight notch on one side; another has seven

'^^ liaKpo^ii with a long face, i. e., maxillary.

eigenmann: the cheirodontin^. 49

teeth, none of which is notched, and still another has either five or six, none notched;
mandibular teeth similar to the maxillary teeth and nearly of the same size.

Origin of dorsal very little nearer base of middle caudal rays than tip of snout;
origin of anal under some part of the base of the dorsal, its base longer than head;
ventrals just about reaching anal or not quite to anal; pectorals reach to the
ventrals, which are inserted in advance of the vertical from front of dorsal.

Fig. 13. Macropsohijcon uruguayance Eigenmann. 6896 CM., a, b, c, outlines of premaxillary,
maxillary, and mandible.

Scales have been mostly lost, best preserved in smallest specimen; very reg-
ularly arranged, and apparently absent from caudal. Interhsemals of caudal
peduncle few and feeble.

Color uniform, an ovate dark spot on middle caudal peduncle.

This species has the most feeble dentition of any of the members of the
Cheirodontinse; its conical teeth would place it outside this subfamily, but its
relationships are unmistakable.

Genus XIII. Megalamphodus-* Eigenmann, gen. nov.

Type : Megalamphodus megalopterus Eigenmann.

Mouth large, teeth in part notched, in part conical, in a single series, seven to
eleven on the premaxillary, two to twenty on the maxillary; four to eight tricuspid
teeth along front of mandible, minute ones on the sides; all the teeth narrow and
pointed, those of front of mandibles, most of those of premaxillary, and usually
the upper ones of the maxillary, each with a minute notch on the sides, the rest
conical; maxillary teeth few or along almost the entire edge; fontanels both very

^* ij.eya\aiJi<i>o5os = with spacious ways.

50 MEMOIRS OF THE CARNEGIE MUSEUM.

large, cheeks below the eye entirely covered, the part behind the eye mostly naked.
Form compressed; fins large, origin of anal under dorsal; adipose fin well-developed,
anal truncate or with a narrow lobe; caudal naked; lateral line incomplete; scales
with a few diverging strise.

It is quite possible that the species here included in one genus should be dis-
tributed to other genera. It is certain that the species selected as the type is
generically different from any of the other genera recognized in this paper. As
stated above melanotus may belong to Macropsobrycon, with which the present
genus is very closely related.*

Key to the Species of Megalamphodus.

a. Anal with a narrow lobe, 25-27; depth 2.2-2.4; fourth dorsal reacliing caudal; a very conspicuous humeral

bar; dorsal dark, without a distinct spot; about twenty maxillary teeth ... 24. megalopterus Eigenmann.

aa. Anal truncate, the rays graduate; dorsal not falcate in the specimens examined. Maxillary with two to six

teeth.

b. A. 30; depth 2.5; a conspicuous humeral band; most of the dorsal black; anal margined with black.

25. eques Steindachner.

66. A. 25-28; depth 3.8; premaxillary with about twelve teeth; maxillary with about four teeth, its

length equal to that of the eye; dorsal with a dark spot; a faint humeral spot and a faint caudal spot.

26. melanotus Eigenmann.
666. A. 27-30; depth 3; no humeral spot; upper half of the first five developed rays of the dorsal black; six

maxillary teeth. 27. heteresthes Ulrey.

6666. A. 24-26; depth 2.4; sides uniformly dusted or with a narrow humeral bar; all but the first and the
last three or four dorsal rays black; caudal and anal margined with black; maxillary with two to six
teeth 28. micropterus Eigenmann.

24. Megalamphodus megalopterus Eigenmann sp. nov. (Plate VII.)

6806, C. M., type, 35 mm. 6807a-b, C. M., paratypes, 3, largest 34 mm.
Caceres, May 23, 1909. Haseman.

Head 3.5; depth 2.2-2.4; D. 11; A. 25-27; lateral line 32-35, 4 to 6 scales
with pores; eye 2.5, slightly greater than interorbital.

Greatly compressed, ventral outline regularly curved; dorsal outline steep
to the dorsal, with a slight depression over the eye, base of dorsal very oblique;
depth of caudal peduncle equal to its length; predorsal area narrow, with a regular
series of nine or ten scales; preventral area broad, covered with two series of scales
overlapping along the median line with an occasional scale at their angles. Fon-
tanels both very large, anterior end of frontal fontanel equidistant from tip of
snout and its posterior end; parietal fontanel much wider than the frontal; occipital
process extending more than one-fourth to the dorsal. Mouth very oblique,
maxillary as long as eye, reaching to suture between second and third suborbitals;

* For additional species see Appendix to this article.

eigenmann: the cheirodontin^.

51

premaxillary and maxillary teeth of nearly the same size, about eleven on the
premaxillary, about twenty on the maxillary, anterior mandibulary teeth con-
siderably larger; eight in front and minute ones on the side, third suborbital with
a spur running up behind the eye; a wedge-shaped naked area behind it; post-
orbitals obsolete; opercle emarginate above. Gill-rakers 7 + 13, the longest about
equal to the pupil.

Origin of dorsal a little nearer to tip of snout than base of caudal, the fourth
ray highest, reaching to the middle of the caudal in the type; caudal forked, the
lobes longer than head; origin of anal equidistant from _base of middle caudal
rays and the origin of the dorsal, below middle of dorsal as measured from tip of

Fig. 14. Megalaniphodus melanopterus Eigenmann. 6807, C. M., a, side of liead; b, top of head, showing
frontal (/) and occipital (o) fontanels; c, d, e, premaxillary, maxillary, and mandible, much enlarged. The relay
teeth in the lower jaw are so numerous that they give the impression of a double row.

snout; anal with a narrow anterior lobe equal to length of head, the rays then
of about equal length or decreasing but little to near the end where the fin is
rounded; pectorals reaching beyond entire base of ventrals, ventrals to about the
seventh anal ray.

Scales thin with few strise and the rest of the sculpturing very weak; caudal
and anal naked, a single series of scales along the base of the anal; pores developed
on a very few scales; eleven scales between ventrals and dorsals.

A very conspicuous, large, humeral bar; dorsal and caudal of varying degrees
of blackness. Anal dusky.

52 MEMOIRS OF THE CARNEGIE MUSEUM.

25. Megalamphodus eques (Steindachner).
Cheirodon eques Steindachner, Flussf. Slidam., Vol. IV, 1882, p. 37 (Villa Bella;

Obidos); Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., Vol. XIV, 1891,

p. 54; Ulrey, Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 293; Eigenmann,

Reports Princeton Univ. Exped. Patagonia, Vol. Ill, 1910, p. 429.

Range: Amazon.

The following is the original description of this species by Steindachner:

"Seitenlinie unvollstilndig, nur 5-8 Schuppen im vorderen Theile des Rumpfes
durchbohrend. Ein qyerbindenahnlicher, intensiv braunlichschwarzer Fleck in
der Humeralgegend. Ein eben so gefarbter grosser Fleck fast liber die ganze
Dorsale sich ausbreitend. Anale am ganzen unteren Rande braunlich punktirt,
wie braun gesaumt. Caudalfleck fehlend.

"Die RiickenUnie erhebt sich viel rascher zur Dorsale, als die Bauchlinie sich
bis zur Ventrale senkt, und ist bei grosseren Exemplaren auch etwas starker gebogen
als die Bauchlinie. Hinter der Dorsale senkt sie sich minder rasch als die Bauch-
linie langs der Analflossenbasis ansteigt.

" Die Dorsale beginnt in der Mitte der Korperlange, hinter der Einlenkungstelle
der Ventralen in verticaler Richtung.

"Die grosste Rumpfhohe ist 2J/2inal, die Kopflange 3mal in der Korper-
lange, der Augendiameter 2}/2rQ.2i\, die Breite der queriiber massig gerundeten Stirne
etwas mehr als 3mal in der Kopflange enthalten und der Schnauzenlange nachste-
hend.

" Der obere Theil des vorderen Oberkieferrandes ist, unter der Loupe betrachtet,
fein gezahnt. Zwischenkieferzahne einreihig.

"Die Spitze der Ventralen reicht liber den Beginn der Anale betrachtlich
hinaus, und die der Pectoralen liberragt gleichfalls ziemlich bedeutend die Inser-
tionsstelle der Ventralen. Vom 4. oder 5. hochsten Strahle der Anale angefangen
nehmen die folgenden Strahlen nur allmalig an Hohe ab, so dass diese Flosse im
vorderen Theile nach unten keinen lappenformigen Vorsprung zeigt.

"Die Hohe der Dorsale gleicht der Kopflange mit Ausschluss der Schnauze,
die Lange der Ventrale steht der Hohe der Dorsale circa um eine halbe Augenlange
nach.

"Rumpf goldgelb, mit zahllosen violetten Plinktchen libersaet, die jedoch erst
unter der Loupe deutlich unterschieden werden konnen. (P. 38.) Der Humeral-
fleck ist schrag gesteUt, nach unten und vom geneigt stets schmal, doch an Breite
ein wenig variabel und zuweilen von einer heUen Zone nach vorn und hinten umge-
ben, scharf abgegrenzt und ausnahmslos tief schwarzbraun. Eine gleich intensive

eigenmann: the cheirodontin^. 53

Farbung zeigt der grosse runde Fleck auf der Dorsale. Langs der mittleren
horizontalen Schuppenreihe des Rumpfes liefen bis zum Beginn der Caudale 33
Schuppen.

"D. 11. A. 30. L. lat. 33. L. tr. 6/1/3M.

"Das grosste der von uns untersuchten Exemplare ist 30 mm. lang (mit
EinscWuss der Caudale).

"Fundort: Amazonenstrom bei Villa Bella und Obidos."

The lengths of the pectoral and ventrals, the shape of the anal, and the color
make it very probable that eques, which I have not been able to examine in the new
light of these studies, belongs to this genus.

26. Megalamphodus melanotus (Eigenmann) .
Aphyocharax melanotus Eigenmann, Mem. Carnegie Mus., Vol. V, 1912, p. 312

(Rockstone on the Essequibo River, British Guiana).

This species is known only from the specimens originally described.

1877a, C. M. Type, 43 mm. Rockstone sand-bank. Eigenmann.

Head 4; depth 3.8; D. 10; A. 25; scales 5-33-2, six with pores. Eye 2.75 in
head, interorbital 3.75.

Compressed, preventral and predorsal areas rounded, the latter with a median
series of ten scales.

Frontal fontanel not entirely separating the f rentals; second suborbital in
contact with the preopercle below, a very narrow naked area behind it.

Mouth large, the antero-posterior extent of the premaxillary very short; the
maxillary large, with a curved anterior margin, its length about equal to that of
the eye; about twelve teeth in each premaxillary; maxillary with about four
similar teeth; about twenty teeth on each side of the lower jaw.

Origin of dorsal a little nearer snout than caudal; origin of anal under end of
dorsal; ventrals not quite reaching anal, pectorals not quite to ventrals. Scales
with a few divergent striae. Pseudotympanum faintly evident on one side. Four
chromatophores, tip of anterior dorsal rays dark.

27. Megalamphodus heteresthes (Ulrey).
Aphyocharax heteresthes Ulrey, Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 293

(Brazil); Eigenmann, Report Princeton Univ. Exped. Patagonia, Vol. Ill,

1910, p. 429.

Range : Brazil, definite localities not known.

Head 3.33; depth 3; D. 11; A. 27-30; scales about thirty-one. Eye twice
the length of the snout, 3.5 in the head.

54 MEMOIRS OF THE CARNEGIE MUSEUM.

Maxillary teeth six or seven; premaxillary with six to eight conical teeth and
two to four with lateral cusps ; mandible with ten conical teeth and four with lateral
cusps; maxillary extending considerably beyond the anterior margin of the eye.

Origin of dorsal midway between tip of snout and base of caudal. Pectorals

Fig. 15. Megalamphodus heteresthes (Ulrey), a, top of head, showing frontal (/) and occipital (o) fon-
tanels; b, maxiUary; c, premaxillary.

extending beyond tips of the axillary scale; ventrals reaching anal; anal rays
graduate.

No humeral or caudal spots, the upper half of the first developed rays of the
dorsal black.

28. Megalamphodus micropterus sp. nov. (Plate VIII, fig. 1.)

6900a, C. M., t3rpe, 30 mm. 6901a-q, C. M., paratypes, seventeen, largest

about 28 mm. Lagoa do Porto, Dec. 24, 1907. Haseman.
6904a, C. M., paratype, one, 27 mm. Rio Salitre. Haseman.
6902a-i, C. M., paratypes, nine, largest 30 mm. Santa Rita, Jan. 24, 1908.

Haseman.
6903, C. M., paratypes, thirty-five, largest 32 mm. Pirapora, Dec. 14, 1907.

Haseman.
6905a-d, CM., four in bad state. Boqueirao near mouth of Rio Preto.
This species from the basin of the Rio San Francisco is similar to eques, from
which it differs at least in the number of anal rays.

Head 3.4; depth 2.4-2.75; D. 11; A. 24-26; scales 5 to 7 + 24 to 27. Eye
2.5, much larger than the interorbital.

Similar to M. megalopterus in shape, the back not quite so elevated; predorsal
area with nine or ten scales.

Mouth large, the maxillary not quite as long as the eye, not reaching to suture
between the second and third suborbitals; premaxillary with seven to nine teeth,
of which one or two may be conical, the rest tricuspid; maxillary with two to six

eigenmann: the cheirodontin^.

55

teeth, of which one or more may be conical; mandible with seven or eight tricuspid
and five to nine conical teeth; occasionally one of the premaxillary teeth is placed
a little in front of the line of the others, suggesting Hyphessohrycon; second sub-
orbital leaving only a narrow naked wedge behind it; postorbitals very feeble.
Gill-rakers 5 + 13.

Origin of dorsal equidistant from tip of snout and base of caudal; dorsal

Fig. 16. Megalamphodus micropterus Eigenmann. a, side of head; 6, top of head, showing frontal (/)
and occipital (o) fontanels; c, d, e, premaxillary, maxillary, and mandible of a specimen, 6903, C. M., greatly
enlarged; /, outline of entire mandible; g, portion of a maxillary; h, i, j, k, premaxillaries showing the teeth. In
fig. h there is at x a tooth out of hne, forming an incipient second series.

pointed, its highest rays, second to fourth, not quite equal to the head in length;
caudal lobes not quite equal to the length of the head; origin of anal under middle
of dorsal; margin of anal truncate, the rays slightlj^ graduate, no lobe, the highest
ray equal to snout and eye; ventrals reaching nearly to, or a little beyond, origin of
anal, pectorals to beyond base of ventrals; males with hooks on the third to six-
teenth anal rays.

Scales thin with a few divergent radial striic; caudal and anal naked.

Sides (in males?) uniformly dusted, or (in females?) with a distinct humeral
band ; all but first and last three or four dorsal rays black ; caudal and anal margined
with black; ventrals and pectorals sometimes tipped with dark.

56 MEMOIRS OF THE CARNEGIE MUSEUM.

Genus XIV. Microschemobrycon^^ gen. nov.

Type : MicroscJieniobrycon guaporensis Eigenmann.

General appearance of Aphyocharax, the lateral line complete to within four
scales of the caudal; three postorbitals, of which the middle one is largest, covering
most of the postorbital area; cheeks covered by the third suborbital; maxillary
slender, the teeth much crowded, and on less than one-third of its length; teeth
narrow, minute, crowded; no pseudo tympanum ; frontal fontanel moderate; adi-
pose fin moderate; anal base long.

29. Microschemobrycon guaporensis, sp. nov. (Plate VIII, fig. 2.)

6910a, C. M., type, about 37 mm. (30.5 to base of caudal) . Maciel, Rio Guapore,
July 23, 1909. Haseman.

Head 4; depth 4; D. 11; A. 22; scales 4-33-3, twenty-nine or thirty scales
with pores; eye 2.75 in the head, snout 3.4, interorbital 3.

About twelve preventral scales, about nine predorsal scales, occipital process
reaching about one-sixth to the dorsal; frontal fontanel not quite 2 in the parietal,
its anterior end a little in front of the middle of the eye; mouth large, maxillary-
premaxillary border 2.75 in the head; maxillarj^ not nearly reaching third sub-
orbital; teeth all minute, those on the mandible conical, with a minute lateral
notch, those on the premaxillary broader; twelve teeth on the premaxillary, twelve
on the maxillary on less than one-third its length; twenty-five or more on the
mandible. Postorbitals in three pieces, the upper and lower ones minute, the
middle one high, less strongly developed than in Aphyocharax.

Gill-rakers 6 -f- 11, long, 2.5 in the eye.

Origin of dorsal a little nearer snout than base of caudal ; origin of anal under
end of dorsal; ventrals almost reaching anal, pectorals not quite to anal.

Scales well imbricate, firm, without radial striae; caudal and anal naked.

A series of dots from between the ventrals to the base of the anal and along
the root of the anal; some chromatophores on the bases of the caudal rays.

Genus XV. Oligobrycon^^ gen. nov.

Type : Oligobrycon microstomus Eigenmann.

Teeth heavy, tricuspid, in a single series, very few in number (four on the
premaxillary); the maxillary not reaching anterior margin of orbit; the eight teeth
of the two premaxillaries in a very shallow crescent, maxillary with one to two teeth;

^^ fuKpoaxvt^s = small of stature.

'^^ oXiyos = small; Brycon, a genus of fishes.

eigenmann: the cheirodontin^.

57

cheeks with a very narrow naked margin; three postorbitals ; lateral hne incom-
plete; scales with few divergent radial striae; caudal naked; adipose well-developed.
Interhsemals not evident; no pseudotympanum.

This genus differs from Aphyocharax and from Mixobrycon ribeiroi in dentition,
the size of the mouth, the armature of the cheeks, etc.

30. Oligobrycon microstomus sp. nov. (Plate IX, fig. 1.)
6898a, C. M., type, 39 mm.; 6899a, C. M., paratype, 27 mm. Jacarehy, Rio

Parahyba, July 15, 1908. Haseman.
Very similar in shape, size of mouth, and general appearance to Cheirodon
interrwptus.

Head 3.5; depth 2.3-2.6; D. 11; A. 24 or 25; scales 7 or 8 + 25, twelve scales

Fig. 17. Oligohrycon microstomus 'EAgsnms.rai.. 6699, C. M. a, outline of top of head; 6, c, d, premaxil-
lary, maxillary, and mandible enlarged; e, dentition of specimen 6898, C. M. ; /, suborbitals and postorbitals of
6699, CM.

between dorsal and ventrals ; eye three times in the head, a little less than interorbital ;
caudal peduncle as deep as long; base of anal a little longer than head.

Very deep and compressed; dorsal and ventral profiles nearly equally arched;
preventral area rounded, without a definite median series of scales; predorsal area
with a blunt keel; about eight scales in a median series from the dorsal forward

58 MEMOIRS OF THE CARNEGIE MUSEUM.

and three overlapping rows of scales between these and the occipital process;
occipital process extending about one-fifth to dorsal, bordered by three scales;
frontal fontanel long, pointed, about one-half as long as the parietal with its groove;
interorbital nearly flat, very broad; snout blunt, face of maxillary nearly vertical;
margin of second suborbital convex, leaving a narrow naked margin of nearly
equal width around its entire margin; postorbitals strong, leaving a narrow naked
margin; mouth minute; maxillary very broad and short, not reaching to below eye,
a little more than half as long as eye; maxillary-premaxillary border about three-
fourths as long as eye. Maxillary with one or two minute teeth; premaxillary
with four graduate teeth; mandible with six graduate teeth. Gill-rakers 5 + 8,
short.

Origin of dorsal equidistant from tip of snout and base of caudal or a little
nearer the latter, its height equal to length of head to upper angle of gill-opening;
origin of anal under middle or behind middle of dorsal; ventrals just reaching anal;
pectorals reaching to ventrals.

Scales very regularly imbricate, those of the belly with many sub-parallel
radial striae, those of the sides with divergent striae.

A vertical humeral bar crossing the third and fourth scales of the lateral line;
an oval spot on the end of the caudal peduncle; region between nape, dorsal, and
anal nearly evenly peppered; region forward of a line joining base of ventrals
and the humeral spot silvery.

Genus XV. Aphyocheirodon" gen. nov.

Type : Aphyocheirodon hemigrammus Eigenmann.

Teeth notched, in a single series, those in the maxillary and premaxillary
narrow, with nearly parallel edges, and with three or five cusps, the median one
prominent, the lateral ones small; premaxillary teeth about eight to ten in number;
maxillary teeth four or five on the upper fourth of the edge of the maxillary. Mandi-
bulary teeth much expanded at tip, the margins of neighboring teeth in contact, the
tip chisel-shaped, divided into three lobes of equal size, a minute cusp on each side.
Mouth large; cheeks partly naked; frontal fontanel very short; adipose fin well
developed; caudal naked; lateral line incomplete; scales with a few diverging radial
striae; no pseudotympanum.

It is very probable that there is but little difference between the sexes, the
caudal fulcra being normal. Only one species is thus far known.

'^ i.4>mi = a small fish; Cheirodon, a genus of this subfamily of fishes.

eigenmann: the cheirodontinje.

59

31. Aphyocheirodon hemigrammus sp. nov. (Plate IX, fig. 2.)
6802, C. M., type, 45 mm. 6803, C. M., paratypes, twenty-six, largest 48 mm.

Jaquara, Aug. 18, 1908. Haseman.
6804a-d, C. M., paratypes, four; largest 45 mm., Mogy Guassii, Aug. 25, 1908.

Haseman.
6805a, C. M., paratype, one, 39 mm. Riberao Azul, 12 miles from Tiete, Oct. 7,

1908. Haseman.
Range : Tiete basin, Jaquara.

Head 4; depth about 3; D. 11 rarely 10; A. 23-25 (27 in one); scales usually
34, more rarely 36 or 37; eye 2.5-3 in the head, about equal to the interorbital.

Fig. 18. Aphyocheirodon hemigrammus Eigenmann. a, outline of head; 6, c, variations in the post-
orbitals; d, top of head, to show form and location of the fontanels; e, premaxillary;/, maxillary; g, mandible,
showing relay teeth at r and p; h, outline of entire mandible, showing teeth-bearing portion at .r. (All figures
much enlarged.)

Compressed, elongate, dorsal and ventral profiles nearly equally arched; pre-
ventral area rounded; sometimes a nearly regular median series of twelve scales,
sometimes the scales irregular; postventral area rounded; predorsal area rounded,
with a median series of about twelve scales; occipital process bordered by two or
three scales reaching one-seventh to the dorsal; skull convex above; frontal fon-
tanel very small, the f rentals sometimes mesially in contact with the bridge;

60 MEMOIRS OF THE CARNEGIE MUSEUM.

parietal fontanel very large; mouth large, the maxillary equal to the eye, not quite
reaching the suture between the second and third suborbitals. Third suborbital
leaving a naked area around its entire convex free margin; a naked area about
half as wide as the postorbital behind the postorbitals. The postorbitals in two
or three parts, very variable. Gill-rakers 4 + 10, longest about half the length
of the eye.

Premaxillary teeth eight to ten, of nearly uniform size, their margins sub-
parallel, the median cusp more prominent than the lateral; maxillary teeth similar,
but smaller than the premaxillary teeth, and with three cusps. Mandibulary
teeth graduated on the side of the jaw, the anterior six or seven of about equal size,
their tips broad, the bases narrowed, the three median points alike, forming a con-
tinuous cutting edge, the expanded tips of the teeth being in contact, a minute
cusp on each side.

Origin of dorsal equidistant from tip of snout and end of middle series of scales,
the dorsal pointed, its height about equal to the length of the head; adipose fin
well developed; caudal lobes about equal to length of the head; origin of anal
below middle or posterior part of dorsal, its margin distinctly emarginate, the
height of the lobe less than the length of the head; ventrals reaching anal; pectorals
reaching ventrals.

Scales with a varying number of radiating strise, everywhere regularly imbri-
cate; caudal naked; no distinct row of scales along the base of the anal; axillary
scale very small; eight to ten scales with pores; ten scales between dorsal and
ventrals.

A conspicuous black spot occupies the entire width of the caudal peduncle and
half the length of the middle caudal rays. A median dusky line associated with a
narrow silvery line on posterior half of body. No other markings.

Male scarcely distinguishable from the female, lower caudal fulcra not modified.

Genus XVI. Compsura^* gen. no v.

Type : Compsura heterura Eigenmann.

Closely allied to Cheirodon, but differing much in the structure of the male.

Teeth few; multicuspid incisors in a single series; mouth minute; second sub-
orbital in contact with the preopercle below and partly behind ; postorbital leaving
a naked area behind about half as wide as the bone; adipose fin well -developed;
caudal fulcra in both sexes covered by scales; a lobe of large scales in the male
extending along the base of the middle caudal rays; lateral line incomplete; scales

'^ KOfixpos = well dressed; 6vpa = tail.

eigenmann: the cheirodontin^.

61

with a few diverging stria?. All but the three or four last of the divided anal rays
of the male with hooks.

32. Compsura heterura sp. nov. (Plate X, fig. 1.)
6808, C. M., type cf , 36 mm. 6809a-k, C. M., paratypes, five males and eight

females, largest 37 mm. Queimadas, Rio Itapicuru, March 2, 1908. Hase-

man.
6810a, C. M., cf, 31 mm. Barreiras, Lagoaof Rio Grande of Rio San Francisco,

Jan. 3 and 4, 1907. Haseman.
6811a-c, C. M., c?, 32 mm. Santa Rita, Jan. 24, 1908. Haseman.
Range : Rio San Francisco and Rio Itapicuru.
This species with the general appearance of a Cheirodon can readily be dis-

FiG 19 Compsura heterura Eigenmann. a, outUne of side of head; b, top of head, showing frontal
(/) and parietal, or occipital (o) fontanels; c, c', right and left premaxillaries; d, maxiUary, e, mandible of an
individual, 6809, C. M., greatly enlarged;/, premaxiUary of another individual; g, scaUng of the tail of a male,
6808, C. M. (All figures greatly enlarged.)

tinguished by the black upper half of the first dorsal membrane together with the
black tip of the dorsal.

62 MEMOIRS OF THE CARNEGIE MUSEUM.

Head 4-4.25; depth 2.5-2.75; D. 10 or 11; A. most frequently 19, ranging
from 17-20; scales 34-35, rarely 32, eight or nine scales with pores, ten scales
between dorsal and ventral; eye 2.6-3 in head, about equal to the interorbital ;
depth of caudal peduncle about equal to its length.

Compressed, dorsal and ventral outlines evenly and equally curved. Pre-
ventral area rounded with a median series of eleven scales; predorsal area rounded
with a median series of nine scales, extending to within two scales of the occipital
process; occipital process broad and short, reaching about one-seventh to the dorsal,
bordered by two and one-half scales on each side; skull convex, frontal fontanel
only about one-fourth as long as the parietal; mouth very short and small, maxillary
very little, if any, more than half as long as the eye; teeth broad-tipped, five- or
seven-pointed, median point largest, projecting; four teeth on the premaxillary,
two on the maxillary, eight or nine on the mandible, those on the side of the man-
dible graduate, the last one may be minute and single pointed, all the rest of the
teeth similar in size and shape; third suborbital strong, in contact with the pre-
opercle below and partly behind; postorbitals thin, ill defined, leaving a considerable
naked area. Gill-rakers 4 + 8, very short, only about one-fifth as long as eye.

Origin of dorsal a little nearer tip of snout than base of caudal, the fin pointed,
the highest ray about equal to the head; adipose fin small; caudal forked, the
lobes about equal to the length of the head; origin of anal behind the vertical
from the last dorsal ray; anal fin but slightly emarginate, its base little shorter
than the head; all but the last three or four of the divided anal rays of the male
with seven or eight strong recurved hooks. Pectorals of the male reaching the
ventrals; ventrals of the male truncate, reaching the anal, ventrals of the female
pointed, not reaching the anal; the pectorals in the female not reaching the ventrals.

Scales thin, regularly imbricate, very few diverging strise; anal with about
three scales forming a sheath at the base of the anal; base of caudal in the male
scaled, the scales covering the caudal fulcra and forming a lobe along the middle
of the fin.

Tip of dorsal and upper part of membrane between the rudimentary, and first
full ray black; tip of anal in the male black; a conspicuous triangular caudal spot
not quite extending to the end of the middle rays; a black band extending forward
to below the dorsal.

Genus XVIII. Mixobrycon^^ gen. nov.

Type: Mixobrycon ribeiroi (Eigenmann).

Closely resembling Hyphessobrycon; teeth heavy, few, in a single series; third

" pi^is = a mixing; Brycon, a genus of Characins. Name chosen because tlie teeth show some of the
characters of Hyphessobrycon.

eigenmann: the cheirodontin^. 63

suborbital with a wide naked area around its entire border. Postorbital covering
nearly the entire postorbital space. Mouth moderate, the maxillary not reaching
to the end of the second suborbital; adipose fin well developed; caudal naked;
lateral line short; frontal fontanel large, entirely separating the frontals; no pseudo-
tympanum; no prominent interhsemals.

33. Mixobrycon ribeiroi (Eigenmann). (Plate X, fig. 2.)
Cheirodon ribeiroi Eigenmann, Proc. U. S. Nat. Mus., Vol. XXXIII, 1907, p. 9;

Reports Princeton Univ. Exp. Patagonia, Vol. Ill, 1910, p. 429.

10229, 1. U. M., one, type, 35 mm. to base of caudal. Puerto Max, Paraguay,
J. D. Anisits.

This species is known only from the type.

Head 3.4; depth 3; D. 11; A. 26; scales 5-33-4, seven scales with pores.
Eye 2.5 in the head, greater than interorbital.

Compressed, elongate; preventral area rounded; predorsal area rounded, with
a median series of eleven scales; occipital process elongate, pointed, reaching one-

FiG. 20. Dentition of Mixobrycon ribeiroi Eigenmann.

fifth to dorsal, bordered bj^ three scales on each side; frontal fontanel long, pointed,
1.5 in the parietal; maxillary long, slender, nearly three-fourths as long as eye;
premaxillary with four large, heavy teeth; the second tooth largest, with seven
points, the middle point of which is the largest, the base of the tooth much nar-
rower than the tip; the second to fourth teeth graduate; no teeth on the maxillary;
mandible with four, heavy, five-pointed, graduate teeth; teeth of side of jaw lost
or absent; second suborbital with a wide naked area around its entire margin.
Gill-rakers 7 -|- 11, the longest about one-third as long as eye.

Origin of dorsal very nearly equidistant from tip of snout and caudal; origin
of anal under back part of dorsal; only one interhsemal (or none?) on the caudal
peduncle; pectorals reaching ventrals, the latter not to anal; adipose fin well-
developed.

64 MEMOIRS OF THE CARNEGIE MUSEUM.

A large humeral spot over third to fifth scales of lateral line; a dark line from
upper part of humeral spot to middle of caudal; caudal spot extending across the
entire caudal peduncle, and on middle caudal rays.

Named in honor of Dr. Alipio de Miranda Ribeiro of Rio de Janeiro.

Genus XIX. Cheirodon^" Girard.
Cheirodon Girard, Proc. Acad. Nat. Sci. Phila., Vol. VII, 1854, p. 199.

Type : Cheirodon pisciculus Girard.

Minute fishes ranging from 25 to 60 mm. in length; teeth of the upper and
lower jaws similar, with five or more points; usually the tips of the teeth are
expanded, more rarely the margins of the teeth are nearly parallel, arranged in a
single series; one to three teeth on the maxillary, four to nine teeth on the premaxil-
lary (most frequently five) ; cheeks usually nearly completely covered by the third
suborbital, but with a naked area around its entire border in C. annce; the origin
of the dorsal very nearly equidistant from tip of snout and base of middle caudal
rays; adipose fin well-developed; origin of anal about under the last dorsal ray; a
variable number of the anal rays and sometimes the ventral rays in the males with
numerous hooks, the base of the hook-bearing portion of the anal not infrequently
at an angle with the base of the normal portion; interhsemal spines of the caudal
peduncle variable, those of the male strong, protruding, sometimes ankylosed and
sometimes with broad wing-like lateral processes. Scales thin, regularly imbricate,
with a variable number of radial strise. Caudal naked, anal with a few scales in a
single series at the base of the anterior rays; scales moderate, between thirty to
thirty-six in the lateral series, of which not more than twelve have pores. No
humeral spot, the covering of the anterior air-bladder very thin, in a triangular
pseudotympanum ; usually a conspicuous caudal spot, more rarely a dorsal spot.

Range: Amazons, Panama, eastern slope of Colombia, south to the Rio
San Francisco, Rio Parahyba, Rio Grande do Sul, and the La Plata basin; western
slope of Chili.

No specimens of Cheirodon pisciculus are at hand. It is possible that the
specimens of Cheirodon annce, listed below, the origin of which is unknown, are in
reality the types of Cheirodon pisciculus. If so, then the species listed after C.
annce are representatives of a genus distinct from Cheirodon. If, however, Cheirodon
pisciculus is distinct from C. annce and agrees with the species which here succeed
it, then C. annce should be made the type of a distinct genus. C. annce certainly is
not congeneric with the other species of Cheirodon here figured. C. microdon and

'"xe'Pi V = hand; dSuv, 6 = tooth.

eigenmann: the cheirodontin^. 65

C. stenodon are also quite distinct from the rest of the species. This is the only
genus of Characins reaching the Pacific slope of Chili, where at Puerto Montt it
also attains the southernmost latitude recorded for the Characins.

Key to the Species of Cheirodon.
a. A naked area about the entire distal margin of the tliird suborbital, very wide beliind it; elongate, depth 3 or
more in the length; mouth moderate; teeth broad-tipped with narrow bases; anal short, 12-15, the
tip of the first developed ray extending beyond the tip of the last; caudal peduncle slender, about
twice as long as deep. (Not examined in C. Tpisciculus.)

b. Maxillary with two teeth ■ [M. pisciculus Girard.

66. Maxillary with one tooth 35. annae McAtee.

aa. Second suborbital in contact with the preopercle at least below; anal emarginate 17-27.

c. Mouth minute or moderate; teeth broad-tipped.

d. Fifteen to twenty interhaemals,'''^ extending from near base of last anal ray to caudal, with broad,
wing-hke lateral processes in the male; a naked area along the entire posterior edge of the
tliird suborbital.
e. Three, more rarely two, maxillary teeth; maxillary very little more than half the length of
the eye, its margin straight; premaxiUary with five teeth; A. 18-19; eighteen to twenty-
three interhsemals from near base or beliind tip of last anal to caudal; base of anal equals

length of caudal peduncle; largest about 25 mm 36. insignis Steindachner.

ee. Two maxillary teeth; maxillary about half the length of the eye, its margin but little curved;
premaxiUary with four teeth; A. 19-24; fifteen to twenty interhsemals, extending from near
base of last anal ray to caudal ; base of anal much longer than caudal peduncle. Length of

largest recorded specimen 39 mm 37. parahybae Eigenmann.

eee. One maxillary tooth; maxillary about half the length of the eye, its margin strongly convex;
premaxillary with four or five teeth; A.17-24; twenty or more interhaemals; extending from
near base of last anal ray to caudal. Largest recorded specimen, 60 mm.

38. interruptus Jenyns.

39. monodon Cope.

40. ibicuhiensis Eigenmann.
dd. Eight to thirteen interhtemals extending from near tip of last anal ray to caudal ; base of anal equals

caudal peduncle and middle caudal rays; maxillary scarcely reaching eye; cheeks with a
naked wedge behind the second suborbital.
/. Dorsal with black spot along the base of the anteriorrays; A. 20-22; scales 32-34; premaxillary

with four teeth, maxillary with two; a spot across the entire caudal peduncle.

41. liolomelas Eigenmann.

jf. Dorsal without a distinct black spot.

g. No caudal spot or other definite markings; premaxillary with five teeth; maxillary with

two; about nine weak interhaemals. A. 23; scales 36 42. madeirae Eigenmann.

gg. A large conspicuous caudal spot sometimes continued to the end of the middle rays.
h. Anal plain; scales 31-36; premaxillary with four or five teeth, maxillary usually with
two; nine to thirteen interhaemals, occupying half the distance between the anal

and caudal; A. 19-27 43. piaba Liitken.

cc. Mouth moderate; teeth in premaxillary narrow, five to nine in number; maxillary teeth two-tliirds to
'' See also insignis, in which there may be only eight.

66 MEMOIRS OF THE CARNEGIE MUSEUM.

three-fourths as long as eye; dorsal with some black; depth 2.8-3; interhaemals of caudal peduncle
feeble; third suborbital leaving only a naked wedge behind it.

i. Three median points of teeth of mandible equal in size; seven to nine teeth in premaxillary; depth
2.8; A. 23-35; maxillary three-fourths as long as eye; seven to nine teeth in premaxillary, two in
maxillary. Scales 34-36 44. microdon Eigenmann.

n. Teeth of lower jaw similar to those of upper, but a little wider, one large median point and two small
points on each side; five to seven teeth in premaxillary, two teeth in maxillary; depth 3; A. 18-
22; maxillary two-thirds in eye; scales tliirty-six or thirty-seven 45. stenodon Eigenmann.

34. Cheirodon pisciculus Girard. (Plate XVII, fig. 4.)
Cheirodon pisciculus Girard, Proc. Acad. Nat. Sci. Phila., 1854, p. 199; U. S. Nav.

Astronom. Exped., Fishes, 1855, p. 249, pi. 34, figs. 4 and 7 (Santiago, Chili);

Eigenmann and Eigenmann, U. S. Nat. Mus., Vol. XIV, 1891, p. 54; Ulrey,

Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 290; Eigenmann, Reports Princeton

Univ. Exped. Patagonia, Vol. Ill, 1910, p. 429.
Chirodon pisciculus Giinther, Cat. Fish. Brit. Mus., Vol. V, 1864, p. 332; Stein-

dachner, Zool. Jahrbuch., Suppl. IV, 1898, p. 328 (Llanquihue, near Puerto

Montt, Chili).

Range : Southern Chili.

I have seen no specimens of this species. The following is an abridgment of
Girard's description:

Head 4; depth 3-4; D. 10; A. 14. Eye 3 in the length of the head; depth of
caudal peduncle 2.5 in the greatest depth.

Snout short and rounded; eye rather large. Maxillary teeth very small and
few. Dorsal fin higher than long. Caudal forked. Anal nearly as deep as long.
Ventrals and pectorals slender. Scales proportionately very large, higher than
long. A silvery band along the middle of the flanks, margined above with black.
Fins unicolor, olivaceous.

A small fish of a rather short appearance, in spite of the slenderness of the
peduncle of the tail. Dorsal and ventral fines equally arched. Body very much
compressed; mouth small and slightly oblique; maxillary extending to a vertical
line immediately in advance of the anterior rim of eye when the mouth is closed.
Dentary teeth much larger than those on the intermaxillaries. Their form is
flattened, dilated towards their upper edges, which are provided generally with
five subconical points, the middle one being the longest, giving them the appearance
of digits.

Origin of dorsal nearer base of caudal than tip of snout; dorsal much higher
than long, its upper edge rounded or subconvex. The adipose is slender, nearer
to the base of the caudal than to the posterior edge of the dorsal and consequently

eigenmann: the cheirodontinjE. 67

situated behind the anal. The anal longer than the dorsal, and nearly as long as
deep; its exterior edge convex anteriorly, and subconvex posteriorly. Its anterior
margin is situated backwards of the posterior edge of the dorsal.

The caudal fin, which constitutes about one-fifth of the total length, is deeply
forked posteriorly; its lobes are rather round and only acute at their extremities.
The insertion of the ventrals is on the middle of the abdomen, somewhat in advance
of the anterior margin of the dorsal. These fins are rather slender, with their tips
acute and reaching the vent. The origin of the pectorals is situated near the
inferior region of the thoracic belt. These fins are longer and slenderer than the
ventrals, their tips almost reaching the origin of the latter fins. Their anterior ray
is simple; the central rays are but once bifurcated, and only towards the last third
of their length.

The scales are of moderate development, higher than long, subelliptical in
shape, sometimes very irregularly so. Ten or eleven longitudintal rows on the
line of the greatest depth, and six or seven rows on the peduncle of the tail. The
lateral line is not to be seen.

Olivaceous brown; a silver band along the middle of the flanks, extending
from the upper angle of the opercular apparatus to the base of the caudal fin. The
cheeks, the opercles, and branchiostegal apparatus are silvery. A blackish stripe
along the upper edge of the silvery band of the sides. The dorsal region is minutely
dotted with blackish, the dots being more particularly crowded upon the outline
of the scales. These dots extend to the upper surface of the head, and sparingly
to the upper region of the thoracic and abdominal regions; also to the inferior
half of the peduncle of the tail. The dorsal, caudal, and anal fins are almost greyish
through the accumulation of the above-mentioned dots. The ventrals are unicolor;
the pectorals greyish upon their external margin. The abdominal region sometimes
exhibits an argentine reflection.

Inhabits the lagoons in the vicinity of Santiago, Chile.

35. Cheirodon annae McAtee. (Plate XI, fig. 1.)
Cheirodon annce McAtee, Proc. Acad. Nat. Sci. Phila., 1903, p. 515 (South America) ;

Eigenmann, Reports Princeton Univ. Patagonia, Vol. Ill, 1910, p. 429.

Habitat : Some unknown locality in South America.

The specimens of this species were received as an exchange from the U. S.
National Museum. Their origin is in doubt. It is possible that they are the types
of C pisciculus.

68

MEMOIRS OF THE CARNEGIE MUSEUM.

4301, I. U. M. 927a-b, C. M. Type and paratypes, fifteen. Length of type,
43 mm. South America.

The specimens are soft and have lost their scales. The characters are other-
wise well-preserved. If these specimens came from Chile, there is no doubt that
they represent C. pisciculus, with which they agree in their extremely short anal
and their elongate form.

Head 4.2; depth 3.6-4.2; D. 9-12; A. 12-15; scales 32-36, eleven scales
between ventrals and dorsal, seven to nine scales with pores; eye 2.8-3.2; depth of
caudal peduncle about 2 in its length.

Elongate, little compressed; predorsal and preventral areas rounded, with
about (?) fifteen scales; occipital process short, reaching one-eighth to the dorsal;

Fig. 21. Cheirodon annce McAtee, a, maxillary; h, premaxillary; c, portion of mandible; d, dentition
seen from in front. 4307, C. M.

frontal fontanel half as long as the parietal; third suborbital very small, leaving a
naked area, which is much the widest behind, about its entire distal margin;
postorbitals minute, not covering more than one-fourth of the width of the cheek
behind the eye; snout blunt, mouth small, maxillary a little over half as long as eye;
teeth five- to seven-pointed, the middle point a little prominent, the bases of the
teeth much contracted; four or five teeth in the premaxillary, one tooth (absent in
two) on the maxillary, mandible with five or six graduated teeth. Gill-rakers
8 -|- 12, short, about one-third as long as eye.

eigenmann: the cheirodontin^.

69

Origin of dorsal a little nearer caudal than tip of snout, its height a little less
than the length of head; adipose fin well-developed; caudal lobes about as long
as head; base of anal about equal to snovit and eye, considerably' less than the length
of the caudal peduncle; origin of anal below, or a little behind, the base of the
last dorsal rays, first developed ray of the anal extending beyond the tip of the

Fig. 22. Cheirodon annw McAtee. Interhaemal spines of &.

Fig. 23. Cheirodon annce McAtee. a, premaxillary ; b, maxillary; c, portion of mandible; d, portion of
the interhffimals of c?.

last ray; ventrals extending to, or a little short of, origin of anal, pectorals to, or a
little short of, ventrals.

Thirteen interhsemals on the caudal peduncle of the female, extending four-
tenths to base of last anal ray. The spines very strong, with broad lateral processes
in the male, extending a little further toward the anal. Scales mostly removed. A
distinctly silvery lateral band.

36. Cheirodon insignis Steindachner. (Plate XVII, fig. 2.)
Cheirodon insignis Steindachner, Fisch-Fauna Cauca & Fliisse bei Guayaquil, 1880,

p. 22, pi. VI, fig. 3 (Cauca); Eigenmann and Eigenmann, Proc. U. S. Nat.

Mus., Vol. XIV, 1891, p. 54; Eigenmann, Reports Princeton Univ. Exped.

Patagonia, Vol. Ill, 1910, p. 429; Evermann & Goldsborough, Proc. Biol.

Soc. Washington, Vol. XXII, p. 98 (Tabernilla, Atlantic slope of Panama

Canal Zone).
Cheirodon gorgonce Evermann & Goldsborough, I. c. p. 99. Below the dam at Gorgona,

Canal Zone.^-

^ This species is said to differ from insignis "in the larger eye, the fewer anal rays and the shghtly shorter
dorsal rays" as weU as in the teeth. The differences found may be tabulated as shown at foot of p. 70:

70 MEMOIRS OF THE CARNEGIE MUSEUM.

Range : Magdalena and Atrato basin, Atlantic slope of Panama.

5367, C. M., 13042, 1. U. M., many, largest about 25 mm. Truando. Wilson.
726, Universitj' of Michigan Museum, two, largest about 28 mm. Marsh at
Fundacion, Colombia.

Head 4; depth 2.5-3; D. 10 or 11; A. 17-20; scales 28-32, of which about
six have pores; eye 3 in head, about equal to interorbital.

Compressed, breast slightly flattened, with a median series of nine or ten
scales; predorsal area with about nine scales; occipital process short, reaching
about one-sixth to the dorsal ; frontal fontanel an equilateral triangle less than one-
half as long as the parietal fontanel without the groove; maxillary reaching to below
anterior margin of eye; third suborbital covering the entire cheek; premaxillary
with five teeth; maxillary with two or three teeth.

Dorsal falcate, sometimes reaching to adipose, its origin in middle of body;
anal emarginate, its lobe in the male reaching tip of last ray, not quite tip of last
ray in female; ventrals reaching beyond origin of anal in male, the pectorals to the
middle third of the ventrals; both fins shorter in the female.

Scales regularly imbricate, largest just above pectorals and ventrals; caudal
naked; anal with a single series of scales along its base; ten scales between dorsal
and ventrals.

Interhsemals extending from near base or tip of last anal ray to caudal, eight to
twenty-three in number, those of the males antrorse and with lateral wings, those
of the female slender, their tips not exposed, pointing downward and backward.

A conspicuous caudal spot, surrounded by an unpigmented area; back and
sides nearly uniformly punctate; margin of anal and dorsal dark.

37. Cheirodon parahybae sp. nov. (Plate XI, fig. 2.)
6841a, C. M., type, 38 mm. 6841b-f, C. M., paratypes, six, largest 39 mm.

Campos, June 14, 1908. Haseman.
Head 3.75-4; depth 2.5-2.7; D. usually 11, rarely 10; A. 19-24; scales 34-36,
eight or nine with pores; eye 2.5 in the head.

Compressed, preventral area and predorsal areas rounded, the somewhat

C. insignis C. gorgonce

Head 3.6-3.8 3.5-3.8

Depth 2.8-3 3-3.2

Eye in the head 2.5-2.6 2.2-2.45

Longest dorsal ray in liead 8 1

Anal 21-22 17-10

Longest anal ray in head 1.4-1.8 1.4-1.0

eigenmann: the cheirodontinjE. 71

irregular median series consisting of about eleven scales; occipital process short,
extending one-seventh to dorsal, bordered by three scales on each side; frontal
fontanel half as long as the parietal; second suborbital in contact with the pre-
opercle below, a broad naked area along its entire posterior margin; postorbitals
feeble, not covering more than half the width of the postorbital area; maxillary
little, if any, more than half as long as eye, slender, its margin not curved; teeth
small, slender, contracted at the base, with five points, the middle one being largest;
four teeth in the premaxillary, two in the maxillary, seven or eight graduated teeth
on the mandible. Gill-rakers about 7 + 10, very short, not quite one-fourth as
long as eye.

Origin of dorsal a little nearer tip of snout than base of middle caudal rays or
the reverse. Dorsal pointed, its height equal to the length of the head, or a httle
longer; adipose fin well-developed; caudal lobes much longer than head; origin of
anal below the posterior part of the dorsal or behind the vertical from the last ray;
fifteen to twenty interhsemal spines, extending from near base of last anal ray to
caudal, with broad wings in the male; ventrals about reaching anal, pectorals a
little beyond origin of ventrals. Scales thin, regularly imbricate, with few radial
striae; caudal naked, anal with a few scales on the bases of the anterior rays.

A large spot extending across the entire caudal peduncle, not extending on
the middle caudal rays.

This species is very similar to C. piaba and C. interruptus, differing in the size
of the naked area of the cheek, the number of interhaemal spines, the length of the
maxiUary, and the number of maxillary teeth.

38. Cheirodon interruptus (Jenyns). (Plate XII, fig. 1.)

Tetragonopterus interruptus Jenyns, Voy. Beagle: Fishes, 1842, p. 127, tab. 23,
fig. 4 (Maldonado).

Chirodon interruptus Giinther, Cat. Fishes Brit. Mus., Vol. V, 1864, p. 332; Perugia,
Ann. Mus. Civ. Storia Nat. Genova, (2), Vol. X, p. 45, 1891 (Rio Plata).

Cheirodon interruptus Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., Vol. XIV,
1891, p. 54; Ulrey, Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 290; ? Boulenger,
Boll. Mus. Univ. Torino, Vol. XII, 1897 (Tala); Eigenmann, Reports Prince-
ton Univ. Exped. Patagonia, Vol. Ill, 1910, p. 429.

Cheirodon monodon Cope, Proc. Am. Philos. Soc, Vol. XXXIII, 1894, p. 91 (Rio
Grande do Sul); Ulrey, Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 290; Fowler,
Proc. Acad. Nat. Sci. Phila., 1906, p. 332 (Rio Grande do Sul); Eigenmann
& Ogle, Proc. U. S. Nat. Mus., Vol. XXXIII, p. 9 (Rio Grande do Sul).

72 MEMOIRS OF THE CARNEGIE MUSEUM.

Eigenmann, Reports Princeton Univ. Exped. Patagonia, Vol. Ill, 1910, p. 429.

Range: Maldonado to Rio Grande do Sul.

The types of interruptus came from Maldonado and the specimens collected
by Carey at the same place (No. 20698, M. C. Z.) are more than probably inter-
niptus. These specimens are specifically distinct from the Paraguayan specimens,
heretofore referred to C. interruptus. A careful comparison with the excellent
fresh specimens from the Paraguay river and others from various places in the
San Francisco basin, collected by Haseman, shows that they are so similar, that
the Paraguayan form hitherto referred to interruptus must be referred to C. piaba.

Fig. 24. Interhsemals in Cheirodon interruplm Jenyns. 6818, C. M.

Specimens from Rio Grande do Sul, the type locality of C. monodon, are very

similar to C. interruptus, and monodon may be considered a variety of interruptus.

6815a-k, C. M., two males, eleven females, largest 45 mm. Rio Guahyba,

Porto Alegre, Rio Grande do Sul. Jan. 21, 1909. Haseman.
6817a-d, C. M., fourteen, largest 40 mm. Santa Maria, Rio Vaccacahy-

Mirim, secondary tributary of the Rio Jacuhy, Rio Grande do Sul. Hase-
man.
6818a-x, C. M., three males, largest 43 mm., twenty-.six females, largest 49 mm.

Cachoeira, Rio Jacuhy, Rio Grande do Sul. Jan. 26, 1909. Haseman.
6852a, C. M., one female, 47 mm. Cacequy. Haseman.
846 & 847, M. C. Z., twelve, poor, largest a male about 35 mm. long. Uruguay

river. Wyman.
20698, M. C. Z., forty, largest about 60 mm. and 20699, M. C. Z., one, 46 mm.

Maldonado. T. G. Carey.
6816a-j, C. M., one male, 42 mm., nine females, largest 52 mm. Cacequy, Rio

Ibicuhy, into Rio Uruguay. Jan. 31 & Feb. 1, 1909. Haseman.
This species can readily be distinguished by the naked area behind the third
suborbital and by the numerous interhaemals. They extend from the anal to the

eigenmann: the cheirodontin^. 73

caudal and number seventeen to twenty-seven, all or only half of which may pro-
trude in the male. The maxillary has but one tooth.

Head 4.5-4.66; depth 2.4-3; D. 11, very rarely 12; A. 17-24; scales ¥, ¥,
^, %-,^, in those examined; seven to twelve scales with pores; eye 2.5-3, about
equal to interorbital ; depth of caudal peduncle 1-1.5 in its length.

Compressed, dorsal and ventral profiles equally arched; preventral area
flattened, with a median series of about thirteen scales; predorsal area rounded,
with thirteen scales; frontal fontanel from half to one-third as long as the parietal;
a broad naked area behind the third suborbital, sometimes extending forward a
little at the angle of the preopercle, postorbital not half as wide as the naked
area behind it; mouth small, maxillary little, if any, longer than half the eye,
shortest in specimens from Cacequj^ and Maldonado, in which its free margin is
more convex. Premaxillary with four or five teeth (six in a few premaxillaries) ;
maxillary uniformly with a single tooth (except in one maxillary, which in addition
has a minute tooth); five, six, or seven graduate teeth in mandible, the teeth
five- to seven-pointed, the base narrower than the tip. GiU-rakers 7 + 12, to
9 + 13, short, the longest not quite a third as long as the eye.

Origin of the dorsal equidistant from tip of snout and base of mid-caudal rays.

Fig. 25. Cheirodon monodon Cope, a and b, premajdllary and maxillary with their teeth.

Adipose fin well-developed. Origin of anal about equidistant from snout with
last dorsal ray. Pectorals reaching ventrals, the ventrals not quite to the anal,
the fourth or fifth to the eighth to fourteenth anal rays of the male with hooks;
the ventrals in adult males also with hooks; seventeen or more interhsemals on
caudal peduncle, those of male contiguous, with broad lateral processes, especially
the anterior ones, their spines projecting from near base of last anal ray.

Scales thin, regularly imbricate, with few to many diverging radial striae;
caudal naked; anal with a few scales in a single series at base of anterior rays.

A silvery lateral stripe; chromatophores variously developed, some speci-
mens from Cacequy are almost free from pigment, except a faint caudal spot;
in others the pigment is well-developed, the specimens appearing quite dark, the
caudal spot being well defined or more diffuse in outline, not extending upon mid-

74

MEMOIRS OF THE CARNEGIE MUSEUM.

caudal rays. The base of the anal in specimens from Porto Alegre, Cachoeira,
and Maldonado is but little, if any, longer than the caudal peduncle. In all but
one of the Cacequy specimens it is equal to the caudal peduncle and middle caudal
rays, and usually contains twenty-two rays, more rarely twenty-one, twenty-three,
or twenty-four.

There may be two, possibly three, varieties in the material at hand. If so,
they may be distinguished by the following characters :

a. Base of anal less than caudal peduncle and middle caudal rays.

b. A. 18-23, most frequently 19-20; most frequently 5-5 teeth in the premaxillaries, less frequently, 4-4
or 4-5 or 5-6; base of anal equal to the length of the caudal peduncle. Caudal spot usually well

defined. Nos. 6815, 6817, 6818 (Rio Grande do Sul) 39. monodon Cope.

bb. A. 17-22, most frequently 19; usually 5-5 teeth in the premaxillary; base of anal equal to the length
of the caudal peduncle or a little longer. Nos. 846, 847, 20698 and 20699, M. C Z. (Uruguay

river basin) 38. interruptus Jenyns.

aa. A. 21-24, most frequently 22; 4-4, or 4-5 teeth in the premaxillary; base of anal equal to length of caudal
peduncle and middle caudal rays. Caudal spot diffuse. No. 6816 (Cacequy).

40. ibicuhiensis Eigenmann var. nov.

Table Showing in Detail the Number of Specimens with the Indicated Number of Premaxillary

Teeth and of Anal Rats.

Premaxillary Teeth.

Anal Rays.

4-4.

4-5.

5-5.

5-6.

6-6.

17.

18.

19.

20.

21.

22.

23.

24.

No. 6818, CM

No. 6815, CM

No. 6817, CM

No. 20698, M. C Z. . . .
No. 6816, CM

9

1

1
5

4
4
1
2
4

13

8

1

21

1

1

1

4

1
1

2
2

2

10

4

14

3

2

2
8

2
1
1
6
2

3

6

7

2

1
1

1

41. Cheirodon notomelas, sp. nov. (Plate XII, fig. 2.)
6812, C. M., type, 9 , 35 mm. ; 6813a-q, C. M., paratypes, nineteen, largest 40 mm.
collected in a lake, four miles from Miguel Calmone, Tiete basin. Oct. 11,
1908. Haseman.
6814, C. M., one, 30 mm. Piperao Azul, lake twelve miles from Tiete. Oct. 7,

1908. Haseman.
Head 4; depth 2.5-2.66; D. 10 or 11; A. 20-22; scales usually 33 (32-34),
six to eight with pores; eye about 2.5 in the head, equal to the interorbital; depth
of caudal peduncle about equal to its length.

Compressed, dorsal and ventral profiles equally curved; preventral area rather
flat with a regular median series of eleven scales ; predorsal area narrowly rounded
with a median series of ten or eleven scales; occipital process broad, reaching one-
seventh to dorsal, bordered by two and one-half scales on each side; skull convex;

eigenmann: the cheirodontin^.

75

frontal fontanel an equilateral triangle, not more than one-third as long as the
parietal; third suborbital less than half as wide as eye, in contact with the pre-
opercle along the lower limb and the angle of the preopercle, a narrow naked area
behind it, and a narrow naked strip behind the postorbitals. Mouth small, the
maxillary half as long as eye. Teeth broad-tipped, seven-pointed, the median
point prominent, especially in the premaxillary. Four teeth in the premaxillary,
two in the maxillary, and seven or eight in the mandible. Gill-rakers 5 + 7,
about one-fourth as long as the eye.

Origin of dorsal equidistant from tip of snout and base of caudal. Dorsal
pointed, its height about equal to the length of the head; adipose fin well-developed,
caudal lobes a little longer than head; origin of anal a little behind the vertical

Fig. 26. Cheirodon notomclas Eigenmann. a, anal and interhsemals of a 9, 6813, C. M.; b, details of
the arrangement of the scales at the base of the caudal in a cf , 6S13a, C. M.

from the last dorsal ray; tip of highest ray reaching to the base of the last fourth
of the base; pectorals reaching ventrals, ventrals not quite to anal.

Scales regularly imbricate, a few scales in a single series along base of anterior
anal rays, ten scales between dorsal and ventral, caudal lobes naked.

First dorsal rays and bases of the rest black; a sub-rhomboidal black spot
across the entire caudal peduncle, the spot bordered by unpigmented areas in front
and behind, the spot not extending to the end of the middle rays; anal dusky, the
first rays sometimes black. General color darker than usual in the genus.

Base of the anterior half of the anal of the male as usual for this genus, much
more oblique than the base of the rest of the fin; fourth to ninth anal rays of the
male much thicker than the rest, with many retrorse hooks along the posterior edges
of the middle part of the'rays.

76 MEMOIRS OP THE CARNEGIE MUSEUM.

42. Cheirodon madeirae, sp. nov. (Plate XIII, fig. 1.)

6847, C. M., one, 34 mm. San Joaquin, Bolivia. Sept. 4, 1909. Haseman.

A small-mouthed species, without color-markings.

Head 4; depth 3; D. 11; A. 23; scales 10 + 24; ten scales between dorsal
and ventrals; eye 2.4 in the head; depth of caudal peduncle nearly equal to its
length.

Moderately compressed; preventral area rounded, with a median series of
twelve scales; predorsal area rounded, with a median series of eleven scales;
occipital process extending about one-seventh to the dorsal; skull convex; frontal
fontanel a little longer than broad, less than half the length of the parietal; mouth
minute, maxillary about half the length of the eye; premaxillary with five teeth,
maxillary with two; mandible with broad, seven-pointed teeth, the median point
being a little the longer; cheeks covered by the third suborbital, leaving a narrow
naked wedge behind; postorbitals nearly covering the entire postorbital area.
Gill-rakers 7 + 11.

Origin of dorsal equidistant from tip of snout and base of caudal; origin of
anal under posterior part of dorsal. Caudal peduncle with about nine weak
interhsemals ; ventrals not reaching anal; the pectorals reaching beyond origin of
ventrals. Scales as in other species of the genus. No color-markings.

43. Cheirodon piaba Lutken. (Plate XIII, fig. 2; Plate XVII, figs. 5-6.)
Cheirodon piaba Lutken, Oevers. Dan. Selsk. No. 3, 1874, p. 134 (Rio das Velhas);

Velhas-Flodens Fiske, 1875, p. xiv and p. 219, fig. on p. 221 (Rio das Velhas);

Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., Vol. XIV, 1891, p. 54; XJlrey,

Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 291; Eigenmann, Reports Princeton

Univ. Exped. Patagonia, Vol. Ill, 1910, p. 429.
Tetragonopterus bellottii Ulrey {non Steindachner) : in part, Ann. N. Y. Acad. Sci.,

Vol. VIII, 1895, p. 286 (Santarem).
Cheirodon insignis Ulrey {non Steindachner) : in part, I. c. 291 (Brazil) ; Eigenmann

& Kennedy, Proc. Acad. Nat. Sci. Phila., 1903, p. 515 (Arroyo Trementina &

Arroyo Pypucu).
Cheirodon calliurus Boulenger, Boll. Mus. Torino, Vol. XV, 1900, p. 370 (Caranda-

sinho near Corumba. San Lorenzo, Prov. Jujuy, Argentina); Eigenmann,

Reports Princeton Univ. Exped. Patagonia, Vol. Ill, 1910, p. 429.
Cheirodon interruptus Eigenmann & Kennedy (non Jenyns), Proc. Acad. Nat. Sci.

Phila., 1903, p. 514 (C'ampo Grande; brook near Arroyo Trementina);

Ann. Carnegie Mus., Vol. IV, 1907, p. 126 (Puerto Max); Eigenmann & Ogle,

Proc. U. S. Nat. Mus., Vol. XXXIII, 1907, p. 9.

eigenmann: the cheirodontin^e. 77

Cheirodon micropterus Eigenmann, Proc. U. S. Nat. Mus., Vol. XXXIII, 1907, p. 9

(Santarem); Reports Princeton Univ. Exped. Patagonia, Vol. Ill, p. 429.

Range: Abundant in Rio Paranahyba, Rio San Francisco, upper Parana,
and Paraguay; rare south of these points, and in Amazons.

The very large number of specimens recorded below, most of them in excellent
condition, together with the specimens recorded by me in the papers quoted above,
enables me to revise the synonymy and bibliography of this species. As stated
by Eigenmann & Ogle a comparison of one of the types of calliurus with specimens
recorded by me as interruptus showed that these specimens belong to the same
species. At the same time the opinion was expressed that these specimens were
probably distinct from interruplus. The material at hand shows that the Para-
guayan specimens are specifically identical with C. piaba, originally recorded from
the Rio das Velhas and found abundantly throughout the Rio San Francisco.

The type of micropterus is poor, but is very probably also a C. piaba.

Aside from the specimens previously recorded, some of which are again
enumerated, I have examined many specimens collected by Mr. J. D. Haseman.

6821a-b, C. M., two males, largest 33 mm. Aregua, Paraguay. April 8, 1909.
Haseman.

6822a-g, C. M., seven, two males, 33 and 39 mm., two females, 36 and 39 mm.
three females, the largest 30 mm.^^ Corumba, Paraguay. April 27, 1909.
Haseman.

6824a-p, C. M., seven males, largest 38 mm.; nine females, largest 38 mm.
Puerto Suarez. May 6, 1909. Haseman.

9984, I. U. M. 933a-b, C. M., ten, largest 42 mm. to base of caudal. Campo
Grande. Anisits.

9997 and 10122, I. U. M.; 926a-b, C. M., sixteen, largest 31 mm. to base of

caudal. Arroyo Pypucu. Anisits.
6825a-h, C. M., nine, largest 35 mm. Asuncion, Paraguay. March 28, 1909.

Haseman.
10289, I. U. M., Colonia Gonzales. Anisits.
6826a-b, C. M., two, 2V' and 44 mm. Caceres. May 27, 1909. Haseman.

9985, I. U. M. 953a-b, C. M., thirteen, largest 40 mm. to base of caudal.
Arroyo Trementina. Anisits.

6908a-f, C. M., six, 15^^-20 mm. Caceres. May 24, 1909. Haseman.

^^ In these smaller specimens and in the two males, the caudal spot extends further on the caudal than
in the two larger females.

^* The smaller has only one tooth on the maxillary.

'' In the smallest the color is most profuse and the caudal spot extends to near tip of the middle rays.

78 MEMOIRS OF THE CARNEGIE MUSEUM.

6823a-u, C. M., twenty-one, largest 39 mm. Lagoa de Parnagua. Jan. 17, 1908.

Haseman.
6827a-k, C. M., one male, twelve females, largest 43 mm. Rio das Velhas,

tributary of Rio San Francisco. May 13, 1908. Haseman.
6828a-x, C. M., twenty-one males, largest 38 mm.; thirty -two females, largest

38 mm. Pirapora, Rio San Francisco. Dec. 15, 1907. Haseman.
6829a-b, C. M., two. Lagoa Pereira, San Francisco basin. Dec. 23, 1907.

Haseman.
6830a-b, C. M., two. Lagoa de Porto, San Francisco basin. Dec. 24, 1907.

Haseman.
6831a-x, C. M., seven males,^'' largest 37 mm., twenty -seven females, largest

39 mm. Barreiras, Rio San Francisco. Jan. 4, 1908. Haseman.
6832a-g, C. M., seven, largest 31 mm. Boqueirao, near mouth of Rio Preto,

San Francisco basin. Jan. 6, 1908. Haseman.
6833a-x, C. M., twenty-eight, largest 38 mm. Santa Rita, San Francisco basin.

Jan 26, 1908. Haseman.
6834a-e, C. M., five,^' largest 39 mm. Penedo at mouth of Rio San Francisco.

March 22, 1908. Haseman.
6835a-c, C. M., three. Rio Coite, San Francisco basin. Nov. 6, 1907.

Haseman.
6846a, CM., one, 40 mm. Lagoa Salgado, San Francisco basin. Nov. 10, 1907.

Haseman.
6838a-d, C. M., four, largest 35 mm. Rio Itapicuru, Fazenda de Amaratu, 6

miles north of Bom Fin. Nov. 21, 1907. Haseman.
6836a-f, C. M., six, largest 33 mm. Rio de Jacobina, tributary of Rio Itapicuru.

A. 19-21. Haseman.
6837a-b, C. M., two, largest 39 mm. Queimadas, Rio Itapicuru. March 2, 1908.

Haseman.
6839a-h, C. M., eight, 39 mm. Alagoinhas, Rio Catu. March 4, 1908.

Haseman.
6840a, C. M., one, 35 mm. Cachoeira, Rio Paraguassu. April 14, 1908. Hase-
man.
6842a-x, C. M., twenty -five, largest 42 mm. Bebedouro, near Rio Grande and

Rio Parana. Sept. 1, 1908. Haseman.
6844a, C. M., one male, 25 mm. Santa Maria, Rio Vaccacahy-Mirim, tributary

of the Jacuhy, Rio Grande do Sul. Jan. 29, 1909. Haseman.

'^ Color prominent, spot extending to end of middle rays in some. Depth 2.33-2.8.
" With ripe eggs.

eigenmann: the cheirodontin^b. 79

6843a-x, C. M., thirty-five, largest 38 mm. Jaquara, Rio Grande, into Rio
Parana. Aug. 18 and 19, 1908. Haseman.

6845a-e, C. M., five, largest 41 mm. Cacequy, secondary tributary of the Rio
Uruguay. Jan. 31, 1909. Haseman.

Head 3.6-4.5; depth 2.25-3; D. 11; A. 19-27 most frequently 22 or 23; scales
31-36, most frequently 33 or 34, 9-12 with spines; eye equal to interorbital,
about 2.5 in the head; depth of caudal peduncle 1.25 in its length; base of anal
much longer than caudal peduncle.

Compressed, depth very variable, the dorsal and ventral profiles symmetric,
nearly equally arched; preventral area flat, with well-marked lateral edges, eleven
to thirteen scales in a median series; predorsal area keeled, with ten scales in a
median series, which is regular to near the occipital process; occipital process short,
extending one-sixth to one-seventh to the dorsal, bordered by two or three scales;

Fig. 27. Cheirodon piaba Liitken. a, premaxillary; b, maxillary.

frontal fontanel an equilateral triangle, 2.5-3.5 in the length of the parietal fon-
tanel; interorbital convex; third suborbital covering the entire cheek, or a very
narrow naked wedge behind it; lower one of the postorbitals as wide as the third
suborbital at its tip, sometimes covering the entire width to the preopercle; the
upper postorbital much narrower, leaving a wider naked strip; mouth minute,
maxillary not, or barely, reaching the eye, nearly vertical, very little more than
half the length of the eye; maxillary with one tooth in three specimens, two teeth
in sixty-two specimens, and three teeth in seven specimens examined; premaxillary
with five teeth; mandible with about seven teeth; teeth sometimes black-tipped.
Gill-rakers about 6 + 11, not over one-fourth as long as eye.

Origin of dorsal equidistant from tip of snout and base of caudal, the highest
ray a little over length of head; origin of anal a little behind vertical from base of
last dorsal ray; pectorals reaching to, or beyond, origin of ventrals, ventrals not
to anal ; eight to thirteen rays of the anal (beginning with the second or third) of the
male with hooks, base of the hook-bearing portion of the anal, making an angle with
the base of the rest of the fin; all but the outer two ventral rays with similar hooks;

80

MEMOIRS OF THE CARNEGIE MUSEUM.

base of anal about equal to length of caudal peduncle plus the middle caudal rays;
the interhsemals, nine to thirteen in number, occupy four-tenths to six-tenths of
the distance between the base of the last anal ray and the caudal, beginning a
httle behind the tip of the last anal ray. Those of the male without lateral proc-
esses, sometimes with a low ridge or keel running along the sides of the anterior
ones and ending in a knob in front of them; lower caudal fulcra prominent, con-
tinuous with the interhsemals. Scales normal.

There is a caudal spot of varying size and intensity, sometimes extending

a ¥

Fig. 28. Cheirodon piaba Liitken. a, region between anal and caudal in 9 ; 6, Do. in (f. Lagoa
Paranagua, 6823, C. M. ; c, anal fin and caudal fulcra and interhsemals of (f , 6828, CM.; d, skeleton of fulcra and
interhsemals of cf , 6843, C. M. (All figures greatly enlarged.)

entirely across the end of the caudal peduncle, sometimes extending on the middle
caudal rays, rarely to their tips; a faint black line overlaid with silvery along the
middle of the sides; anterior margin of dorsal dark; general color very pale or
quite dark, depending on the nature of the locality from which the specimens came.
It is possible that we should refer to this species a small specimen, 6909, C. M.,
15 mm. long, taken in a tributary of the Guapore about forty miles south of Villa
de Matto Grosso.

44. Cheirodon microdon sp. nov. (Plate XIV, fig. 1.)
This species is allied to Cheirodon stenodon and to Aphiocheirodon hemigrammus.
6850, C. M. Type, 42 mm. 6851a-b, C. M., paratypes two, the larger 41 mm.,
Caceres, Upper Paraguay, May 24, 1909. Haseman.

eigenmann: the cheirodontin^.

81

Head 4-5; depth 3; D. 11; A. 23-25; scales in a median line are 34-36, of
which 10-11 are with pores; ten and one-half scales between dorsal and ventrals.
Eye 2.75 in head, very little greater than interorbital.

Compressed, slender; dorsal and ventral profiles nearly equally curved; pre-
ventral area flat, with rather well-defined lateral angles, and with a nearly regular
median series of twelve scales; predorsal area with a median series of eleven scales;
occipital process extending one-sixth to one-seventh to the dorsal, bordered by
three scales on each side; frontal fontanel triangular, the sides of the triangle but
little longer than the base, about 2.5 in the length of the parietal fontanel; mouth
large, maxillary three-fourths as long as eye; teeth five-pointed; the median points

Fig. 29. Cheirodon microdon Eigenmann. a, outline of side of head; h, top of iiead, sliowing fontanels;
c, premaxillary; d, maxillary; e, maxillary; g, dentition; h, details of interneurals and interhaemals, 6850, C. M.

of the premaxillary projecting considerably beyond the lateral points, the three
median points of the teeth of the lower jaw about equal, the extreme lateral points
minute; seven to nine teeth in the premaxillary, two in the maxillary; teeth in
front of the lower jaw rapidly graduated from the third to the sixth; sides of
lower jaw upturned, with several (about six) minute, conical teeth; third suborbital
leaving but a very narrow wedge-shaped naked area behind; postorbitals covering

82

MEMOIRS OF THE CARNEGIE MUSEUM.

half to two-thirds of the postorbital area. Gill-rakers 7 + 13, the longest about
one-third as long as eye.

Origin of the dorsal very little nearer tip of snout than to middle caudal rays,
the highest ray a little greater than the head; adipose fin well developed; origin
of anal below the posterior part of dorsal. Ventrals not reaching anal, the pectorals
about to the origin of the ventrals.

Scales thin, regularly imbricate, with few diverging radial strife. Caudal
naked; anal with a few scales in a single series along base of anterior rays.

Interhaemal spines of the caudal peduncle feeble, about eight to ten in number.

A faint caudal spot, not directly continued forward as a dark band; upper
part of the anterior dorsal rays dark. A silvery lateral stripe.

The three specimens here described appear to be females. They are probably
from an area in which the color-cells do not reach their fullest pigmentation. The
species distinctly differs from C. stenodon in the character of the teeth.

45. Cheirodon stenodon, sp. nov. (Plate XIV, fig. 2.)
6848a, C. M., type, 33 mm., 6849a-x, C. M., paratypes, over thirty, largest 34

mm. Bebedouro, near Rio Grande and Rio Parana. Sept. 1-5, 1908.

Haseman.
A long-jawed, small-toothed, slender species, with feeble interhsemals.

^(^0

Fig. 30. Cheirodon stenodon Eigenmann. a, b, c, premaxillary, maxillary, and mandible; d, dentition;
e-f, symphyseal line.

Head 4; depth a little more than 3; D. 10; A. usually 20 or 21, rarely 18, 19,
or 22; scales in a median series 36 or 37, rarely 32; six to eleven scales with pores;
eight to ten scales between dorsal and ventrals; eye 2.75 in head, equal to the inter-
orbital; depth of caudal peduncle about 1.33 in its length.

eigenmann: the cheirodontin^. 83

Compressed, slender; dorsal and ventral profiles nearly equally curved; pre-
ventral area rounded, with a nearly regular median series of twelve to thirteen
scales; predorsal area rounded, with a perfect median series of nine to ten scales;
occipital process extending one-sixth to one-seventh the distance to the dorsal,
bordered by three scales on each side; frontal fontanel equilateral, less than half
as long as the parietal; skull convex, smooth between the eyes.

Maxillary comparatively long, at least two-thirds as long as eye, its free margin
a little convex; teeth narrow, the sides nearly parallel; a large median point and
two small points on each side, those of the front of the lower jaw shghtly broader
than those of upper jaw; premaxillary teeth five to seven; maxillary teeth two;
mandibulary teeth six to nine, the first four of nearly equal size, those on sides
rapidly graduated, the last ones conical; third suborbital covering the cheek below,
leaving only a narrow naked wedge behind it. Postorbitals covering about two-
thirds of the width of the postorbital area. Gill-rakers 6 -|- 12.

Origin of the dorsal equidistant from snout and middle caudal rays, the dorsal
pointed, its highest ray about equal to length of head. Adipose fin well-developed.
Origin of anal under the vertical from posterior part of the dorsal. The interhsemals
of the caudal peduncle weak, very few, about five, if at all developed, not projecting
in any of these specimens. Ventral not reaching anal. Pectorals long, reaching
to or nearly to the ventrals. None of the anal rays have hooks; it is probable
therefore that the specimens are all females.

Scales regularly imbricate, with a few diverging radial striae; caudal naked.
Anal with a few scales along the base of the anterior rays.

Straw-colored, the scales of the back margined with a row of chromatophores;
a large triangular caudal spot not extending to the end of the middle rays, continued
forward along the middle of the sides as a narrow band to in front of dorsal. Tip
of dorsal and membranes between first and second and upper half of second and
third dark.

Genus XX. Holesthes^^ Eigenmann.
Holoshesthes Eigenmann, Smiths. Misc. Coll. Quarterly, Vol. XLV, 1903, p. 144.

Type : Cheirodon pequira Steindachner.

Very similar to Aphyocheirodon and Odontostilbe.

Teeth notched, in a single series, six or seven in the premaxillary, rather
narrow, but little expanded toward tip, more or less ovate, with five to seven
notches, of which the median one is a little the larger, the rest being lateral; mandib-

^ oXos = complete; I<r9ijs, 17 = clothing, in allusion to the complete dentition of the maxillary. I here
take opportunity to amend the spelling of this generic name.

84 MEMOIRS OF THK CARNEGIE MUSEUM.

ulary teeth much expanded at the tip, with a small basal notch on each side and
three median points of about the same size and extent. Maxillary with few teeth,
broad-tipped, the points nearly alike. Mouth moderate; third suborbital in con-
tact with the preoperclc below; frontal fontanel short; adipose fin well-developed;
origin of anal below end of dorsal; caudal naked; lateral line complete; scales with
very few radial striae; interhsemals of the caudal peduncle not projecting.
Minute fishes of southern and eastern Brazil and Paraguay.

Key to the Species of Holesthes.
a. Dorsal with a large, oblique, black wedge extending from the upper part of the anterior ray toward the
middle of the sixth ray; male with the outer ventral and first developed dorsal ray fiUform.

46. pequira (Steindachner).

aa. Tip of dorsal faintly dusky; interhajmals of the caudal peduncle of the male much stronger than the inter-

neurals; outer ventral ray and first dorsal rays not filiform 47. heterodon Eigemnann.

46. Holesthes pequira (Steindachner). (Plate XV, fig. 1.)
? Salmo pequira Natterer, MS.
Cheirodon pequira Steindachner, Anz. Ak. Wiss. Wien, 1882, p. 180 (Rio Guapore);

Flussf. Slidam., Vol. IV, 1882, p. 38 (Cujaba); Eigenmann & Eigenmann,

Proc. U. S. Nat. Mus., Vol. IV, 1891, p. 54; Ulrey, Ann. N. Y. Acad. Sci.,

Vol. VIII, 1895, p. 290; Boulenger, Boll. Mus. Univ. Torino, Vol. XII, 1897

(Caiza; Mission de San Francisco).
Holoshesthes pequira Eigenmann, Smiths. Misc. Coll. Quarterly, Vol. XLV, 1903,

p. 144; Reports Princeton Univ. Exp. Patagonia, Vol. Ill, 1910, p. 429.
Odontostilbe trementince Eigenmann and Kennedy, Proc. Acad. Nat. Sci. Phila.,

1903, p. 513 (Arroyo Trementina); Eigenmann, Ann. Carnegie Mus., Vol. IV,

1907, p. 125 (Puerto Max) ; Reports Princeton Univ. Exped. Patagonia, Vol.

Ill, 1910, p. 429.

Range : Upper Madeira, Paraguay.

In the original description of pequira Steindachner states that the entire

Fig. 31. Hukdhes pequira (Natterer). a, maxillary; b, premaxillary.

anterior edge of the maxillary is finelj- dentate and that the caudal spot is very
small. If Steindachner's statement is correct then trementince is a valid species

EIGENMANN: the CHEIRODONTINiE. 85

and the type of a distinct genus. But the statement was probably made without
microscopic preparations. Furtliermore the specimens had been in alcoliol for
fifty-eight years when the description was prepared and it is possible that they
were faded and the caudal spot small in consequence. Except for these two char-
acters the description of pequira applies verj^ well to specimens of trementince and
indeed in two of the specimens enumerated below the caudal spot is quite small.
There is no other species known from the upper Paraguay related to these, which
has an oblique bar on the upper part of the dorsal.

Specimens sent me from the British Museum as pequira are Odontostilbe
microcephala.

7317o, C. M.; 9987, 1. U. M., nine, type and paratypes of trementince. Arroyo
Trementina. Anisits.

9986 and 9987, I. U. M. Brook near Arroyo Trementina.

6857a-l, C. M., twelve, largest 43 mm. Villa Hays, Paraguay. April 13,
1909. Haseman.

10187, I. U. M., one, Puerto Max. Anisits.

6858a-e, C. M., five,^^ largest 31 mm. Asuncion, Paraguay. March 28, 1909.
Haseman.

6859a, C. M., one, 35 mm. Caceres, Paraguay. May 27, 1909. Haseman.

6860a-b, C. M., two, largest 43 mm. Corumba, Paraguay. April 27, 1909.
Haseman.

6861a-h, C. M., eight, largest 45 mm. Sapucay, Paraguay. April 2, 1909.
Haseman.

6862a-e, C. M., five, largest 56 mm. Cacequy, Rio Ibicuhy. Feb. 1, 1909.
Haseman.

6877a-c, C. M., three, largest 55 mm. Santa Maria, Rio Maccacahy Mirim.
Jan. 29, 1909. Haseman.

Head 4.5; depth 3-3.5; D. 11; A. ^, ¥, ¥, ^^ ¥; scales 6-¥,¥,¥, ¥, ¥
—4; eye 2.75-3 in the head, equal to, or a little greater than, interorbital; depth of
caudal peduncle about 3 in the depth, 1.5 in its own length.

Compressed, elongate; the dorsal and ventral profiles equally arched; the pre-
ventral area flat, with a nearly regular median series of thirteen scales; predorsal
area rounded or bluntly keeled, with about ten scales; occipital process one-sixth
in the distance from its base to the dorsal, bordered by two or three scales; frontal
fontanel a httle longer than wide, triangular, its length a little more than two in
the parietal fontanel; mouth small, the maxillary scarcely reaching to below the

^^ Caudal spot minute in two.

86 MEMOIRS OF THE CARNEGIE MUSEUM.

anterior margin of the eye. Teeth of the upper jaw narrow, with a large median
point and three small points on the side of the tooth; teeth of lower jaw much wider,
broad-tipped, with five points, three of these of equal size and prominence, the
lateral points minute and much below level of the other three; maxillary usually
with two or three teeth, rarely four; premaxillary teeth six and six or six and seven;
mandible with about six notched, graduated, teeth followed on the side with about
three minute, conical teeth; second suborbital and lower postorbital leaving but
a narrow, naked area between them and the vertical limb of the preopercle.
Gill-rakers 7 + 12.

Origin of dorsal equidistant from tip of snout and some distance behind tip of
adipose, its height little greater than length of head. The first ray in the male pro-
longed into a filament, which sometimes reaches the adipose fin. Anal emarginate,
its origin below the base of the last dorsal ray, base equal to length of head and
about one-fourth of pectoral; ventral usually not reaching anal, its outer ray
sometimes (in some males) filiform, reaching beyond origin of anal; origin of ven-
trals in front of the vertical from the first dorsal ray; pectorals not quite reaching
ventrals in female, a little beyond their base in some males; no prominent inter-
hsemals in males or females ; about seven (beginning with the first fully developed)
raj'S of the anal of the male with hooks.

Scales thin, regularly imbricate, with few if any radial strise; caudal naked;
a small sheath at the base of the anterior anal rays; lateral line but little decurved.

Distinguished from the other members of the genus by the large dorsal spot,
which is wedge-shaped from the upper part of the first fully developed dorsal ray
(the second) to the middle of the sixth or seventh ray; sometimes a minute black
spot near the middle of the first fully developed anal ray; caudal peduncle usually
with a large, conspicuous spot, which extends a little way on the middle caudal
rays and usually upward and downward across the entire caudal peduncle. The
caudal si)ot more rarely (6858a-b) very small, circular. A well-defined silvery
lateral band.

The specimens from the province of Rio Grande do Sul, Nos. 6862 and 6877,
are larger than the rest. The caudal spot is restricted, ovate, but faintly extended
upward and downward. They are slenderer, depth 2.33-2.66; with the six inter-
hsemals and nine interneurals of the caudal peduncle of the male e'qually well-
develoi)ed.

As there is still a little doubt about their identity, I add the original description
of pequira as given by Steindachner, I. c. :

(P. 38) "Seitenlinie vollstiindig; Korperform schr gestreckt. Bauchlinie bis

eigenmann: the cheirodontin^. 87

zur Ventrale bald mehr, bald minder bedeutend gebogen und in der Regel ein
wenig schwacher zur Bauchflosse abfallend, als die nur sehr wenig gebogene Rticken-
linie zur Dorsale ansteigt. Dorsale in der Mitte der Korperlange, hinter der Basis
der Ventralen (in verticaler Richtung) beginncnd. Silbcrfarbige Seitenbinde unter-
halb der Dorsale bis zur Caudale scharf ausgepragt, welter nach vorn an den
Randern verschwommen. Caudalfleck sehr klein, doch deutlich sichtbar. Humer-
alfleck in der Regel fehlend, oder nur ausserst schwach angedeutet. Eine durch
starke Anhaufung dunkler Punkte gebildete schrage Binde in der oberen Halfte
der Dorsale.

"Stirn queriiber gewolbt. Mundspalte sehr klein. Oberkiefer am ganzen
vorderen Rande sehr fein gezdhnt.

"Leibeshohe 33^:4mal, Kopflange 3^mal in der Korperlange; Augendiameter
2^-2%mal, Stirnbreite nahezu 3mal, Schnauzenlange fast 4mal in der Kopflange
enthalten.

(P. 39) "Die Hohe der Dorsale erreicht eine Kopflange; die stark zugespitz-
ten Caudallappen sind merklich liinger als der Kopf. Die Spitze der Ventrale
reicht genau bis zum Beginn der Anale, die der Pectoralen nahezu bis zur Basis
der Ventralen. Rumpf hell goldgelb, gegen die Bauchscite herab hellgelb.

"D. 11. A. 22. L. lat. 35-36 (bis zur Basis d. Caudal). L. tr. 6/1/4.

"Zahlreiche Exemplare bis zu 38 mm. lange, von J. Natterer im Jahre 1824
(Send. VIII, Nr. 59) im Cuyaba gesammelt, und Salmo pequira gennant."

47. Holesthes heterodon sp. nov. (Plate XV, fig. 2.)

Range : Rio San Francisco to Rio Ribeiro.

6875a, C. M., type, female. 48 mm. Jaguara, Rio Grande emptying into
Rio Parana, August 18, 1908. Haseman.

Paratypes all of the following:

6876a-x, C. M., twenty-seven, largest 49 mm. Jaguara, Aug. 18, 1908.
Haseman.

6864a-p, C. M., sixteen, largest 50 mm. Sete Lagoas, May 5, 1908. Haseman.

6865a-x, C. M., twenty -six, largest 50 mm. Bebedouro near Rio Grande and
Rio Parana, Sept. 1, 1908. Haseman.

6866a-h, C. M., eight, largest 38 mm. Pirapora, Dec. 5, 1908. Haseman.

6867a, C. M., one, 36 mm. Rio das Velhas, May 18, 1908. Haseman.

6868a-e, C. M., five, largest 38 mm. Lagoa Pereira, Dec. 23, 1907. Haseman.

6869a-c, C. M., three, largest about 41 mm. Rio Zinga, Nov. 7, 1907. Hase-
man.

88

MEMOIRS OF THE CARNEGIE MUSEUM.

6870a, C. M., one, 28 mm. Barreiras, Lagoa of Rio Grande, San Francisco

basin, Jan. 3 or 4, 1908. Haseman.
6871a-o, C. M., fifteen, largest 46 mm. Sao Joao del Rei, May 17, 1908.

Haseman.
6872a-d, C. M., four, largest about 48 mm. Sao Joao del Rei, May 17, 1908.

Haseman.
6874a, C. M., one, 38 mm. Penedo, March 22, 1908. Haseman.
6873a, C. M., one, 40 mm. Queimadas, Rio Itapicuru, March 2, 1908.

Haseman.
6878a, C. M., one, 48 mm. Iporanga, Dec. 1, 1908. Haseman.
This species is very closely allied to pequira, differing in the characters noted
in the key. It is almost identical in characters with Odontostilbe microcephala.

Fig. 32. Holesthes heterodon Eigenmaim. a, maxillary; b, premaxillary; c, mandible.

Head 4.25-4.75; depth 3-3.5; D. 11, very rarely 10; A. 19-24 most frequently
20 in specimens from Jaguara and Bebedouro; most frequently 22 in specimens
from Pirapora and Sete Lagoas of the San Francisco basin. Scales 5 or 6--/,
¥, ff, -3^4 (35 being more frequent in the specimens from the San Francisco;
36 more frequent in specimens from Jaguara and Bebedouro). Eye 3 in head,
equal to interorbital ; depth of caudal peduncle 2.5 in the depth, 1.5-1.75 in its
length.

Compressed, elongate; the dorsal and ventral profiles equally arched; pre-
ventral area flat or rounded, with about thirteen scales; predorsal rounded, with

eigenmann: the cheirodontin^. 89

ten to twelve scales in a series which is regular, except at about the sixth scale from
the occipital process; occipital process about one-sixth in the distance to the dorsal;
bordered by two or three scales; frontal fontanel a little longer than wide, a little
more than twice the length of the parietal; maxillary reaching to below anterior
margin of eye; teeth as in C. pequira, premaxillary teeth in those examined ^,
---, ^, ^; maxillary teeth ranging from one to four most frequently two and
two in San Francisco specimens, most frequently three and three in Jaguara and
Bebedouro specimens; mandible with about six graduate teeth and a few minute
ones on the side; third suborbital and lower postorbital leaving but a very narrow
naked strip behind them. Gill-rakers 6 or 7 + 11 or 12.

Origin of dorsal about equidistant from tip of snout and the interneurals of
the caudal peduncle, distinctly behind the vertical, from the origin of the ventrals;
origin of anal under base of last dorsal ray. Dorsal rays of male not prolonged;
height of dorsal a little greater than length of head; base of anal a trifle longer
than height of dorsal; the first five or six fully developed anal rays of the male
broad, with hooks; the ventral profile from origin of anal to caudal concave;
ventrals usually just reaching anal in male; pectorals just reaching ventrals, these
fins a little shorter in the female; ventral rays of the male with some hooks below.

Scales thin, regularly imbricate with a few widely divergent radial strise.
Caudal naked; anal with a very rudimentary sheath at the base of the anterior rays.

Lateral line complete, except in one specimen from Sete Lagoas and in nine
specimens from Jaguara, in which the pores were:

Left. Right.

4-3-3-2-1-23 4^3-3-3-lr-3-i|-l-4-8

36 34-2
10-3-1-1-1-6-12-2 14-5-1-3-9

37 4-2-29-2
10-2-6-3-4^6-5-? 13-2-3-1-1-2-3-1-8-2
30-2-1-2-1 4-5-5-6-5-2-9-?
16-2-17 34+(?)
2-21-5-3-3-2 7-5-7-2-15

? (10-2^)-2-13 7 3-5-4-17-1

8-3-2-2-3-1-10-4-4 5-3-6-1-1-1-1-1-13-1-3

The heavy-faced figures indicate scales with pores. It is seen that many
sorts of breaks occur. The pores may not be developed in front, at the end, or at
any point along the lateral line, and at different points on the two sides.

90 MEMOIRS OF THE CARNEGIE MUSEUM.

An ovate spot at the end of the caudal peduncle, continued a little way on
the middle rays; tip of dorsal variously dusky, a distinct narrow, silvery, lateral
band.

Genus XXI. Odontostilbe^" Cope.
Odontostilbe Cope, Proc. Am. Phil. Soc, 1870, p. 566.

Type: Odontostilbe fugitiva Cope.

Teeth notched, in a single series, those of the premaxillary and mandible
similar, with a large median point and smaller lateral points; maxillary with a
few broad-tipped teeth; third suborbital in contact with the pre-opercle below;
adipose fin well-developed; origin of anal under end of dorsal; caudal naked;
lateral line complete; scales with few radial striiE; interhsemals feeble, not pro-
jecting.

Minute fishes found from Panama and Trinidad to the La Plata and the
Peruvian Amazon.

But few of the species of this genus are distinguishable at sight. 0. melandeia
may be distinguished by the absence of a proper caudal spot; paraguayensis ajid
madeirce by their deep form, they are quite similar; drepanon by its filamentous
fin-rays. 0. microcephala and fugitiva are very similar. 0. hastata stands quite
alone, but looks hke microcephala. 0. microcephala differs from Holesthes heterodon
largely in the structure of the teeth of the lower jaw.

It is possible that 0. hastata and 0. melandeia do not belong to this genus.

Key to the Species op Odontostilbe.
a. Caudal in the male with a pouch covered by scales just below the shortest rays at the middle. Eye 3;
A. 18-21 ; raj^s of lower caudal lobe of male with retrorse hooks, an intense well circumscribed spot on the

caudal peduncle 48. hastata Eigemnann.

aa. Caudal naked in the male.

6. A large conspicuous caudal spot; teeth large; 5-7 in the premaxillary, 1-4 in the maxillary,
c. Male with the outer ventral and first developed dorsal rays filamentous. D. 10 to 12.

49. drepanon Fowler.
cc. Male without filamentous rays.
d. Dorsal with 10 or 11 rays.

e. Depth about 3.25; A. 22-34; scales 35 to 37; mouth minute, maxillary half the length

of the eye 50. fugitiva Cope,

ee. Depth 3 to 4; A. 18-22, scales 34r-37; mouth moderate, maxillary more than half the

length of the eye 51. microcephala Eigenmann.

eee. Depth 2.5 to 2.75; A. 22-25; scales 33 or 34; caudal spot imUstiuct.. .,52. pulchra Gill.

eeee. Depth 2.6 to 3; A. 21-24; scales 32-35; mouth minute. Maxillary half the length of

the eye 53. paraguayensis Eigenmaun & Kennedy.

^^ oSoiis, 6 = tooth; o-riX/St), 17 = a lamp, possibly in allusion to the brilliant teeth.

EIGENMANN: the CHEIRODONXINiE.

91

dd. Dorsal with 12 or 13 rays; depth 2i to 3; A. 23-24; scales 34 to 37; mouth minute;

maxillary half the length of the eye 54. madeirae Fowler.

hh. No distinct caudal spot; caudal peduncle narrowly margined with dark; depth 3.6; A. 21; scales 34 or
35; premaxillary with 10 to 14 teeth, maxillary with 4 to 7 55. melandeta Eigenmann.

48. Odontostilbe hastata Eigenmann. (Plate XVI, fig. 1.)
Odontostilbe hastatus Eigenmann, Indiana University Studies, No. 18, 1913, p. 27.
Range: Magdalena and Atrato basins.
5703, C. M., type, cf, 40 mm.; paratypes, twenty-five, largest 37 mm.

5104 a-j, C. M. 12861a-j, I. U. M., Soplaviento. Eigenmann.
5365a, C. M., Certegui. Rio Quito, into Atrato. Wilson.
5387a-f, C. M., 13079, I. U. M., many, Rio Atrato at Quibdo. Wilson.
5366a-x, C. M., 13045, I. U. M., many, Rio Truando, into Atrato. Wilson.
Head 4+; depth 2.8-3.25; D. 11; A. most frequently 19 (18-21); scales

Fig. 33. Odontostilbe hastata Eigenmann. Caudal scales of &.

5.5-32 to 35-3; eye 3 in the head, equal to interorbital. Most nearly like Odon-
tostilbe paraguayensis, but much slenderer.

Compressed; dorsal and ventral profiles equally arched; preventral area
rounded, with about eleven scales; postventral and predorsal areas narrowly
rounded, the latter with a regular median series of about ten scales; occipital
process short and broad, its length one-sixth of the length from its base to the dorsal,
bounded by three scales on each side; frontal fontanel variable, an equilateral
triangle as wide as, and half as long as, the parietal fontanel, or quite minute;
skull convex; snout. blunt, the mouth comparatively large; maxillary-premaxillary
nearly as long as the eye; five teeth in the premaxillary, two or three in the maxil-

92 MEMOIRS OF THE CARNEGIE MUSEUM.

lary; four broad-tipped, seven-pointed teeth in the dentary in front and with as
many as four graduated teeth on the side. Second suborbital covering the entire
cheek. About ten rakers on the lower arch of the gill.

Dorsal pointed, its highest ray longer than head, reaching to within two or
three scales of the adipose, its origin about equidistant from tip of snout and caudal;
middle caudal rays very short; a pouch on the caudal of the male just below the
middle rays covered with scales; in the male the rays of the lower caudal lobe
with retrorse hooks, similar to those of the five anterior anal rays in the male;
anal short, its margin subtruncate (very slightly emarginate), its rays graduate, the
tip of the highest (the third) reaching to last fourth or to base of the last ray; anal
base 3.75-4.6 in the length, its origin behind the vertical from the last dorsal ray;
origin of ventrals in front of the vertical from the anterior dorsal ray, about
reaching the anal; pectorals not quite reaching ventrals or slightly beyond their
base.

Scales thin, regularly imbricate, with as many as ten radial striae; lateral
line complete, nearlj^ straight; anal sheath consisting of a single row of scales along
the bases of the anterior rays; caudal naked, except for a basal sheath on the lower
lobe of the male and the peculiar scales just below the middle rays in the male.

Scales of the back margined with dark; margins of the myotomes above the
anal marked with chromatophores ; no humeral spot ; a silvery band ; a conspicuous
black spot on the end of the caudal peduncle, rounded in front, pointed on the bases
only of the middle caudal rays; peduncle in front of the spot without chromato-
phores. Orange in life above and behind caudal spot.

One specimen from the Calamar Cienega, 32 mm., C. M. No. 5105, and three,
the largest 30 mm., C. M. No. 5106, 1. U. M., No. 12862, may belong to this species.
Chromatophores are limited to the dorsal region and to along the base of the anal;
the caudal spot is smaller, oval.

49. Odontostilbe drepanon Fowler.

Odontostilbe drepanon Fowler, Proc. Acad. Nat. Sci. Phila., 1913, p. 529. (Tribu-
tary of the Madeira river near Porto Velho.)
Range : Madeira basin.
This species is known from the types only. It is quite possible that it will

prove to be synonymous with 0. fugitiva.

Head 3.25; depth 3-1/6; D. 10; A. 24; scales 6-38-5; 9 predorsal scales;

eye 3.25 in the length of the head, maxillary 3.5; interorbital 2.87; depth of caudal

peduncle 2.66.

eigenmann: the cheirodontin^.

93

Elongate, well compressed; mouth small, terminal; lips moderate; maxillary
not reaching eye; teeth with five to seven points, maxillary with two teeth; sub-
orbital completely covering cheek. Origin of dorsal midway between tip of snout
and tip of adipose; third dorsal ray prolonged and filamentous, nearly to origin of
the adipose; origin of anal behind base of dorsal, third ray longest, thence gradually

Fig. 34. Odontostilbe drepanon Fowler. (After Fowler, Proc. Acad. Nat. Sci. Phila., 191.3, p. 529.)

diminishing; pectorals reaching ventrals; origin of ventrals in front of dorsal, first
ray produced in a filament, which extends beyond the origin of anal.

Base of caudal with a dusky blotch a little larger than the eye, not continued
to the end of the middle rays. No humeral spot. Fins all pale.

50. Odontostilbe fugitiva Cope.
Odontostilbe fugitiva Cope, Proc. Am. Philos. Soc, 1870, p. 566, fig. (Pebas);

Eigenmann and Eigenmann, Proc. U. S. Nat. Mus., Vol. XIV, 1891, p. 54;

Eigenmann, Reports Princeton Univ. Exp. Patagonia, Vol. Ill, 1910, p. 429.
Cheirodon fugitiva Ulrey, Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 290. (Lower

Amazon.)

Range: Amazon basin.

6863a-d, C. M., four, largest 41 mm. San Antonio de Rio Madeira. Nov. 3,
1909. Haseman.

Head 4.5; depth 3.25-3; D. 11; A. 22 or 23; scales 37. Eye 2.75-3 in the
head, .6 in snout, equals interorbital; depth of caudal peduncle 3 in the depth,
1.4 in its own length. Gill-rakers 7 + 12.

94

MEMOIRS OF THE CARNEGIE MUSEUM.

51. Odontostilbe microcephala Eigenmann.
Odontostilbe rtiicrocephalus Eigenmann, Proc. U. S. Nat. Mus., Vol. XXXIII, 1907,
p. 10.

Range : Paraguay and Upper Parana basins.
11086, 1. U. M., type, 46 mm. 11086, 1. U. M., paratype, 45 mm. Rio Pilco-

mayo, Bolivia. Exchange with British Museum.
6854a-e, C. M., five, largest about 80 mm. (the caudal is broken), the longest

on record for the genus. Rio Tiete at Salto Avanhandava below the falls.

Sept. 15, 1908. Haseman.
6855a-p, C. M., sixteen, largest 33 mm. Rio Tiete, Salto Avanhandava

above the falls. Sept. 14, 1908. Haseman.
6856a, C. M., one, 38 mm. Asuncion. March 28, 1909. Haseman.
Readily distinguished from the other members of the genus by the teeth, by
the elongate form and prominent caudal spot.

Head 4-4.6; depth 3-4; D. usually 11, rarely 10; A. \^, ¥, ¥, ¥; scales

Fig. 35. Odontostilbe microcephala Eigenmann. a, premaxillary; b, maxillary; 6^ left maxillary of
another individual; c, mandible. (All figures greatly enlarged).

6-^, -0-, ■^, T^-4 or 5; eye 2.5-3 in the head, equal to or a little greater than the
interorbital, .6-.9 (in largest) in the snout; depth of the caudal peduncle 2.33-3
in the depth, .4 of its own length.

Elongate; dorsal and ventral profiles similar; preventral area flat, with a
regular median series of fifteen scales; predorsal area rounded, with a median
series of eleven scales. Occipital process one-fifth to one-eighth in the distance to
the dorsal; frontal fontanel about 2.5 in the parietal; second suborbital and the

eigenmann: the cheirodontin^. 95

postorbitals leaving a very narrow naked area between them and the upper limb
of the preopercle. Snout pointed, maxillary scarcely reaching to origin of eye,
its length about .6 or .7 that of the eye; usually two maxillary teeth, rarely one
or three; premaxillary and mandibular teeth similar, their margins rounded, the
median cusp slightly the larger, five, very rarely six, teeth in the premaxillary;
mandibular teeth eight to ten ; the premaxillary teeth slightly directed backward,
the teeth in the largest specimen (6854a) are much broader, more symmetrical, and
the differences between the points are very slight, the smallest specimen, from the
same locality has fewer points and the median point is distinctly more prominent.
Gill rakers 7 + 10.

Origin of dorsal a little nearer end of adipose than to tip of snout; highest
dorsal ray equals length of head; adipose fin well-developed; origin of anal below
tip or middle of last dorsal ray, the fin emarginate, its base equalling length of head
or a little more; tip of ventrals to the anal or 'to the third scale in front of it; origin
of ventrals under, or slightly in advance of, the origin of the dorsal; the pectorals
to the ventrals, or to the third scale in front of them.

Scales thin, regularly imbricate; lateral line but little decurved; caudal naked;
anal with a few scales forming a sheath in front.

A silvery lateral band ending in a large, oval spot on the caudal peduncle,
which extends a little upon the middle caudal rays and sometimes entirely across
the end of the peduncle.

52. Odontostilbe pulchra (Gill.) (Plate XVII, fig. 1.)
Poecilurichthys pulcher Gill, Ann. Lye. Nat. Hist., Vol. VI, 1858, p. 59 (Trinidad).
Cheirodon (Odontostilbe) pulcher Liitken, Overs. Dan. Selsk., 1874, p. 236.
Odontostilbe pulcher Eigenmann &. Eigenmann, Proc. U. S. Nat. Mus., Vol. XIV,

1891, p. 54; Eigenmann, Reports Princeton Univ. Exped. Patagonia, Vol. Ill,

1910, p. 429.
Cheirodon pulcher Ulrey, Ann. N. Y. Acad. Sci., Vol. VIII, 1895, p. 290.
Chirodon pulcher Regan, Proc. Zool. Soc. London, 1906, p. 385 (Trinidad).

Range: Trinidad.

As far as known this species is confined to the Island of Trinidad. I have seen
no specimens.

The following description is compiled from Regan's account:

Head 4.5-4.66; depth 2.5-2.75; D. 11; A. 23-25; scales 5.5-32 to 34-3.5-4.5,
4.5-5.5 between lateral line and anal. Snout. 5 in the head, eye 2.5, interor-
bital 2.25.

96 MEMOIRS OF THE CARNEGIE MUSEUM.

Maxillary extending to the vertical from the anterior margin of eye. Origin
of dorsal equidistant from snout and base of caudal, its height greater than head.

Pectorals extending to ventrals, origin of ventrals in advance of dorsal, extend-
ing nearly to the anal. Caudal peduncle as long as deep.

Olivaceous, sides silvery, or a silvery longitudinal stripe from operculum to
base of caudal; an indistinct dark humeral spot; a blackish spot at the base of
caudal, ending posteriorly in a point and margined with yellow above and below;
dorsal and anal pink.

53. Odontostilbe paraguayensis Eigenmann & Kennedy. (Plate XVI, fig. 2;

plate XVII, fig. 3.)

Odontostilbe paraguayensis Eigenmann & Kennedy, Proc. Acad. Sci. Phila., 1903,

p. 512 (Asuncion; Arroyo Trementina); Eigenmann, Ann. Carnegie Museum,

Vol. IV, 1907, p. 125 (Corumba); Reports Princeton Univ. Patagonia, Vol.
Ill, 1910, p. 429.

Range: Paraguay basin.

9988, I. U. M., type, 40 mm. Asuncion. J. D. Anisits.

10111, I. U. M., three, 40 mm. Arroyo Trementina. J. D. Anisits.

10178, I. U. M., five, Corumba. J. D. Anisits.

6853a-m, C. M., fourteen, 22-39 mm. Asuncion, March 29, 1909. J. D.
Haseman.

This is the deepest of the species of Odontostilbe, greatly differing from the type,
0. fugitiva, in depth.

Head 3.75-4.5; depth adult female 2.6, adult male 3; D. 11; A. 21-24;
scales 6-32 to 35-4; eye 2.66-3 in head, about equal to interorbital; depth of
caudal peduncle 2.66-3.25 in the greatest depth.

Compressed; dorsal and ventral profiles about equally curved, the depth in-
creasing rapidly to the origin of dorsal and tapering to the slender caudal peduncle;
preventral area flat, with well marked lateral angles, with a perfect median series
of eleven to thirteen scales, or but one scale disarranged; predorsal area keeled
with about nine scales; occipital process pointed, reaching about one-fourth to the
dorsal; skull smooth, convex, frontal fontanel a nearly equilateral triangle about
one-third as long as the parietal fontanel without the groove; third suborbital
leaving naked only a small wedge of the cheek behind its upper angle; postorbitals
very thin, leaving this area practically naked; mouth small, maxillary reaching
to, or not quite reaching, the vertical from the anterior margin of the eye, about
half as long as eye; premaxillary with five or six sub-ovate teeth, each with a

eigenmann: the cheirodontin^.

97

prominent median cusp and four, rarely five, more or less graduate cusps on each
side, the two edges of the teeth not quite symmetric; maxillary with two or three
broad teeth; mandible with eight graduate teeth, the anterior ones very wide at
the tip, contracted at the base, with a prominent median point and three or four
slightly graduate points on each side of it, the tips of the lateral points forming a
slight curve beyond which the median point projects somewhat, the median cusp
at least twice as wide as the two cusps on either side of it; the edges of the teeth
overlapping more and more toward the sides of the jaw. Gill-rakers 6 + 10.

Origin of dorsal about equidistant from tip of snout and tip of adipose, its
height equal to the length of the head; adipose well-developed; caudal lobe a little
longer than head; origin of anal about under vertical from tip of last dorsal ray,
its margin nearly straight from the ninth ray, its base about equal to the head.
Origin of ventrals about equidistant from snout, with the origin of the dorsal just
reaching anal, or falling short of it by one scale ; pectorals not quite reaching ventrals.

Lateral line nearly straight; scales everywhere regularly imbricate; caudal
naked; anal with a few scales at the base of the anterior lobe; about six feeble
interhsemals ; caudal in male without hooks or special scales.

A silvery lateral band; a conspicuous caudal spot, not continued to the middle
of the central caudal rays; anterior margin of dorsal membrane dark.

54. Odontostilbe madeiras Fowler.
Odontostilbe madeiroe Fowler, Proc. Acad. Nat. Sci. Phila., 1913, p. 527 (tributary

of the Rio Madeira near Porto Velho) .

Fig. 36. Odontostilbe madeiroe Fowler. (After Fowler, Proc. Acad. Nat. Sci. Pliila., 1913, p. 527.)

Range : Madeira basin.

This species is known from the types only. It may be synonymous with 0.
Paraguay ensis.

98 MEMOIRS OF THE CARNEGIE MUSEUM.

Head 3.6; depth 2-7/8; D. 12; A. 23; scales 6-37-4; 11 predorsal scales; eye
3-1/8 in the head, maxillary 3-1/8, interorbital 3; height of dorsal 1; height of
anal 1.4; depth of caudal peduncle 2.5; pectoral 1.25; ventral 1.4.

Compressed, elongatel.y ovoid; mouth, short, jaws thin; lips thin; maxillary
reaching the eye; teeth broad, with seven denticles, of which the median largest;
maxillary with two similar teeth; suborbitals entirely covering cheek.

Dorsal inserted nearly midway between tip of snout and end of adipose ; origin
of anal close behind vertical from end of dorsal, lowest edge of anal emarginate;
ventrals inserted slightly behind origin of dorsal, extending to anal.

Margin of scales of back dusted, a patch of dusky dots above base of anal.
Base of caudal with a large rounded spot about as large as eye. A dark line con-
current with vertebral axis from behind shoulder to caudal.

55. Odontostilbe melandeta Eigenmann.
Odontostilbe melandetus Eigenmaim, Mem. Carnegie Mus., Vol. V, 1912, p. 312,

Plate XLIV, fig. 3 (British Guiana, probably Rockstone on the Essequibo

river) .

Range: British Guiana.

1878, C. M., type, 27 mm.; 12160, I. U. M., two, paratypes, 27 and 35 mm.
British Guiana. Eigenmann.

Known only from the types. This species is more cylindrical, its scales firmer,
its teeth smaller. It is quite possible that it represents the type of a distinct genus.

Head 3.75; depth 3.6; D. 10; A. 21; scales 4-34 or 35-3. Eye 2.5-2.7 in
the head, interorbital but little narrower than the eye.

Minute, compressed, head bluntish, mouth terminal; a median series of ten
scales in front of the dorsal. Maxillary very slender, reaching to below the eye;
maxiUary-premaxillary border as long as the eye; ten to fourteen teeth on the pre-
maxiUary, four to seven on the maxillary.

Scales very regularly imbricate, with concentric, but without longitudinal
striae; lateral line complete, scarcely decurved; anal naked; base of caudal with a
few scales.

Gills well-developed; anal deeply emarginate; pectorals not quite reaching
ventrals, ventrals not reaching anal.

No chromatophores on the sides; scales of the back with marginal series of
chromatophores; a series of black specks along the base of anal; caudal peduncle
margined with black.

eigenmann: the cheirodontin^. 99

APPENDIX.

Since the foregoing pages were sent to press the author has discovered among
the material collected by Mr. Henn during his explorations in Ecuador an addi-
tional species, a description of which follows:

56. Megalamphodus ecuadorensis sp. nov.

13628, I. U. M. Type, about 28 mm., 20.5 mm. to base of caudal. Naranjito,
Rio Chan Chan, Ecuador. Henn.

Of this species, the only one of the subfamily found on the Pacific slope of the
northern half of South America, I have but a single, poorly preserved specimen.
Head 3.4; depth about 3; D. ?; A. 23; Scales ?; eye 3 in the head, a little greater than
the interorbital.

Compressed ; dorsal and ventral outlines equally curved. Fontanels very large,
the frontal fontanel entirely separating the frontal bones; third suborbital having
only a narrow naked ridge behind it. Mouth large, the maxillary-premaxillary
border longer than the eye. Seven or eight premaxillary teeth, which are slightly
graduate ; maxillary with two small teeth (on one side at least) ; mandible with
about fifteen teeth, of which the first four or five are nearly equal in size, larger
than the premaxillary teeth, the rest rapidly diminishing in size. Origin of the
dorsal a little behind the middle, 12 mm. from the tip of the snout, 10.5 mm. from
the base of the middle caudal rays; origin of anal on vertical from origin of dorsal;
height of dorsal very nearly equal to length of head; pectorals extending beyond
the base of ventrals, ventrals beyond origin of the anal. The scales are small. A
weU-defined, vertical humeral spot; margin of anal dark.

MEMOIRS OF THE CARNEGIE MUSEUM. Vol VIL

Memoirs Carnegie Museum, Vol. Vll.

Plate II,

, , A A .a^/'AA/^A''
'^ ^ .^ A, A. A A A A / A A '

A A /, /> A /
I V C V-

Fig. 1. Grundulus bogotensis (Humboldt). 5084, C. M., 76 mm. Vicinity of Bogota.
Fig. 2. Aphyocharax paraguayensis Eigenmann. Type. 6096, C. M., 25 mm. Caceres.

Memoirs Carnegie Museum, Vol. VII.

Plate III.

Fig. 1. Spi nthcrobohis papilliferus EiGE-NMANN. Type. 3882, C. M., 41 mm. Alto da Serra.

Fig. 2. Spinthernholus papiUifenis Eigenmann. 3883a, C. M.

Fig. 3. Spintheroholus papilliferus Eigenmann. 3883a, C. M.

Fig. 4. Spintherobolus papilliferus Eigenmann. 3883a, C. M.

Fig. 5. Aplujnchara.r rathhuni Eigenmann. CType of .4. siramineiis Eigenmann.) 10030, I. U. M. 25 mm. to base

of caudaL Arroyo Trementina.

Fig. 6. Aphijocharax anisitsi Eigenmann & Kennedy. Type. 10028, I. U. M., 41 mm. Asuncion.

3

■fl

<
a

^

p "

l-H ^

a ^ci

t^

s -^

CO

^ a i

^5

32 oj j;

^S '' ^

b o

^ ^.

o3 -X

fl

^ d

• Xi

Ps

>>T3

Stein

Boul
mm.
9, C.

t*-.

z -< <1 S .

w ^^

^^

• CI

. O fl

ElGENMANN &

Steindachner

oulenger
5497,
Eigenma

Cope)

CO CO

-§ ^ g^

S <zj e

1 E

steind
filigeru
paragu

t-i «

^ c c5

B ft S

« -2

■gee

S !~ S~

^ ^

0.0-0

c S,

C33 O O

=^ S'

ess

^ s

■goo

~- o

-Tj a, ^ a. Oh

^ IM

CO -* "O

o o

o o o

p^ fe

&H fe p!H

Memoirs Carnegie Museum, Vol, VII.

Plate V.

Fig. 1. Phanagoiu'ates wilsoni EiGENMANN. Type. 5354, C. M., 41 mm. Manigru.

Fig. 2. Parecbasis cydolepis Eigenmann. Type. 5495, C. M., 74 mm. San Antonio de Rio Madeira.

Memoirs Carnegie Museum, Vol. VII.

Plate VI.

Fig. 1. Leptobrycon jatuarance Eigenmann. Type. 20952, M. C. Z., 29 mm. to base of caudal. Jatuarana.
Fig. 2. Macropsobrycon uruguayance Eigenmann. Type. 6895, C. M., 46 mm. Cacequy.

Memoirs Carnegie Museum, Vol. VII.

Plate VIII.

Fig. 1. Megalamphodus micropterus'EiGEfiMANN. Type. 6900, C. M., 30 mm. Lagoa do Porto, Basin Rio S. Franci.sco.
Fig. 2. Microschemohrycon guaporensis Eigenmann. Type. 6910, C. M., about 37 mm. Maciel.

Memoirs Carnegie Museum, Vol. VII.

Plate IX.

Fig. 1. Oligobnjcon ndcrostomus Eigenmann. Type. 6898, C. M., 39 mm. Jacarehy, Rio S. Francisco.
Fig. 2. Aphijocheirodon hemigrammus Eigenmann. Type. 6802, C. M., 4o mm. Jacquara.

Memoirs Carnegie Museum, Vol, Vll

Plate X.

Fig. 1. Coynpsuraheterura Eigenmann. Type rf fiSOS C \i\ -ic ^ • ,

.\ pe. iu.„ J, i. U. M., 35 mm. to base of caudal. Puerto Max, Par

■aguay.

Memoirs Carnegie Museum, Vol. Vll.

Plate XI,

Fig 1. Cheirodon anncv McAtkt,. 4301, I. U. M., 41 mm. S.America. Exact locality unknown.
Fig. 2. Cheirodon parahybw Eigenmann. Type. 6841, C. M., 38 mm. Campos, on R. Parahyba.

Memoirs Carnegie Museum, Vol. VII,

Plate XII.

^.-'-^

FiG.l. CA..rorfo« .Wm-«;.<.s Jen^ 9. 6818, C. M., 44 mm. Cachoeira, R. Jacuhy, R. Grande do SuL
J^IG. 2. Cheirodon notomelns Eigenmann. Type, 9 . 6812, C. M., 35 mm. Miguel Calmone.

Memoirs Carnegie Museum, Vol. VII.

PLATE XIII,

^A.^\aaaaX. #1

l>

Fig. 1. Cheirodon nuularw Eigenmann. Type. 6847, C. M., 32 mm. San Joaquin.
Fig. 2. Cheirodon piaba Lutken, cf. 6822, C. M., 40 mm. Corumba.

Memoirs Carnegie Museum, Vol. VII.

Plate XIV.

Fig. 1. Cheirodon microdon Eigenmann. Type, 9- 6850, C. M., 42 mm. Caceres.
Fig. 2. Cheirodon stenodon Eigenmann. Type. 6848, C. M., 33 mm. Bebedouro.

MEMOIRS Carnegie Museum, Vol. VII.

Plate XV.

» N A A A *> A

^^^s^i^Eg-- ' V y . , . V V V' V V - '
^^^^P?"^, A k/NAAAA)AAX

Fig 1. Holesthes pequira (Steindachner). 6857, C. M., 44 mm. Villa Hays.

Fig. 2. Holesthes heterodon Eigenmann. Paratype. 6876, C. M., about 46 mm. Jaguara.

Memoirs Carnegie Museum, Vol. VII.

Plate XVI.

Fig. 1. Odontostilbe hastata Eigenmann. 5366, C. M., 30 mm. Rio Truando, Colombia.
Fig. 2. Odontostilbe paraguaijensis Eigenmann & Kennedy. 6853, C. M., 36 mm. Asuncion.

MEMOIRS Carnegie Museum, Vol, Vll

Plate XVII.

M-

Fig. 1.
Fig. 2.
Fig. 3.
Fig. 4.
Fig. 5.
Fig. 6.

Odontostilbe pulchra (Gill). (After Regan.)

Cheirodon insignis Steindachner. (After Steindachner.)^

Odontostilbe paragumjensis Eigenmann & Kennedy. 10178, I. U. M.

Cheirodon pisciculus Girard. (After Girard.)

Cheirodon piaba Lutken. (After Liitken.)

Cheirodon piaba LUtken. 10121, 1. U. M. (Greatly enlarged to show the pseudo-

tympanum.)

REPRINTS FROM THE ANNALS OF THE CARNEGIE MUSEUM

1. The Crayfish of Allegheny County, Pa. By B.

B. Williamson. 6 pp. ( Very scarce.)

2. A Preliminary list of the Vascular Flora of Alle-

gheny County, Pa. By J. A. Shatek. Pp.
114. {Out of Print.)

3. Some New and Little Enown Fossil Vertebrates.

By J. B. Hatcheb. Pp. 17, 4 Plates. {Very
scarce.)

4. The BeptUes of Allegheny County, Pa. By D.

A. Atkinson. Pp. 13. {Very scarce.)

5. Osteology of the Herodiones.' By E. W. Shti-

FELDT. Pp. 92, 2 Plates. {Out of print.)

6. Dentition of Titanotherium, By J. B. Hatches.

Pp. 7, 2 Plates. {Out of Print.)

7. Sabal Bigida; a New Species of Palm from the

Laramie. By J. B. Hatchek. Pp. 2. {Out
of print.)

8. Supplement to Dr. John Hamilton's List of the

Coleoptera of Southwestern Pennsylvania. By
Henry G. Klages. Pp. 30. {Out of print.)

9. Osteology of the Flamingoes. By R. W. Smr-

FELDT. Pp. 30, 6 Plates. {Very scarce.)....

10. Description of a New Species of Baena (B.

Hatcheri) from the Laramie Beds of Wyo-
ming. By O. P. Hay. Pp. 2, 1 Plate. {Out
of print.)

11. The Jurassic Dinosaur Deposits near Canyon

City, Colorado. By J. B. Hatchee. Pp. 15.
{Out of pri}it.)

12. A Mounted Skeleton of Titanotherium dlspar

Marsh. By J. B. Hatcher. Pp. 9, 3 Plates.. .

13. Structure of the Fore Limbs and Manns of Bron-

tosaurus. By J. B. Hatcher. Pp. 21, 2 Plates.

14. Genera and Species of the Trachodontidse (Ha-

drosauridae, Claosauridse) Marsh. By J. B.
Hatcher. Pp. 10

15. Some New Pennsylvania Thorns. By W. W.

Ashe. Pp. 12

16. Osteology of the Psittacl. By E. W. Shtipeldt.

Pp. 23, 4 Plates. {Scarce.)

17. An Annotated Catalogue of Shells of the Genus

Partula in the Hartman Collection Belonging
to the Carnegie Museum. By H. H. Smith.
Pp. 64

18. Two New Species of Bahaman Lepidoptera. By

W. J. Holland. Pp. 4

19. Elosaurus Parvus; a New Species of the Sauro-

poda. By O. A. Peterson and C. W. Gilmoeb.
Pp. 10 ,.

20. The Boundary Controversy Between Pennsyl-

vania and Virginia, 1748-1785. By Boyd
Cbumrine. Pp. 20, 3 Maps

21. Minute Book of the Virginia Court Held at Fort

Dunmore (Pittsburgh) for the District of
West Augusta, 1775-1776. Edited by Boyd
Cbumrine. Pp. 44

22. Minute Eook of the Virginia Court Held for

Yohogania County, first at Augusta Town
(now Washington, Pa.), and afterward on the
Andrew Heath Farm near West Elizabeth,
1776-1780. Edited by Boyd Cbumrine. 2 pts.,
pp. 295

23. IVIinute or Order Book of the Virginia Court

Held for Ohio County, Virginia, at Black's
Cabin (Now West Liberty, W. Va.), &c.
Edited by Boyd Cbumrine. Pp. 74

24. The Records of Deeds for the District of West

Augusta, Virginia, for the Court Held at Fort
Dunmore, &c. Edited by Boyd Ceumeine.
Pp. 90

25. Astropecten (?) montanus, &c. By Eael Doug-

lass. Pp. 4

26. Discovery of the Remains of Astrodon (Pleuro-

coelus) in the Atlantosaurus Beds of Wy-
oming, By J. B. Hatchee. Pp. 6. {Out of
print.)

.30

27.
28.

29.

.65

30.

.50

31.

32.

33.

34,

35.

.90

2.25

1.50

1.75
.10

36,

37.

38.

39.

.35

40.

.60

41,

.25

42.

.25

.36

43.

44.

1.25

45.

.10

46.

.15

47.

48.

I .30

49.

50.

.90

51.

52.

53.

54.

65.

56.

57.

Osteology of the Limicolae. By E. W. Shuteldt.
Pp. 56, 1 Plate

New Vertebrates from the Montana Tertiary.
By Earl Douglass. Pp. 64, 1 Plate

Description of a New Genus and Species of Tor-
toise from the Jurassic of Colorado. By O. P.
Hay. Pp. 4, 1 Plate

Osteology of Oxydactylus. By O. A. Peterson.
Pp. 42, 12 Plates

Birds of Erie and Presque Isle. By W. E. C.
Todd. Pp. 115, 3 Plates and Map

In Memoriam. J. B. Hatcher. By W. J. Hol-
land. Pp. 8, 1 Plate

The Tropidoleptus Fauna at Canandaigua Lake,
N. y., with the Ontogeny of Twenty Species.
By Percy E. Eaymond. Pp. 98, 8 Plates.
{Out of print.)

On Two Species of Turtles from the Judith
River Beds of Montana. By O. P. Hay. Pp.
5, 1 Plate. {Out of print.)

A Preliminary List of the Hemiptera of Western
Pennsylvania.. By P. Modestus Wietneb.
Pp. 49. {Scarce.)

The Trilobites of the Chazy Limestone. By
Percy E. Eaymond. Pp. 58, 5 Plates.
{Scarce.)

The Crawfishes of Western Pennsylvania.. By
A. E. Ortmann. Pp. 81. {Scarce.)

Notes on the Geology of Southwestern IJ[ontana.
By Eakl Douglass. Pp. 21, 1 Plate

A New Crocodile from the Jurassic of Wyoming.
By W. J. Holland. Pp. 4, 1 Plate

Procambarus, a New Subgenus of the Genus
Cambarus. By A. E. Ortmann. Pp.8

Presentation of Reproduction of Diplodocus Car-
negei to the Trustees of the British Museum.
By W. J. Holland. Pp. 10, 2 Plates

List of the Birds Collected near Mombasa, E-cSt
Africa, by William Doherty. By W. J. Ho'<-
LAND. Pp. 11 .

The Hyoid Bone in Mastodon Americanus. By
W. J. Holland. Pp. 4

Additions and Corrections to the List of the
Vascular Flora of Allegheny County, Pa. By
Otto E. Jennings. Pp. 7

A New Species of Kneiffla. By Otto B. Jen-
nings. Pp. 2, 1 Plate

Note on the Occurrence of Triglochin palustris
in Pennsylvania. By Otto E. Jennings. P. 1.

A New Species of Ibidium (Gyrostachys). By
Otto E. Jennings. Pp. 4, 1 Plate

The Agate Spring Fossil Quarry. By O. A.
Peterson. Pp. 8

Description of Two New Birds from British
East Africa. By Harry C. Oberholsee. Pp. 3.

The Chazy Formation and Its Fauna. By Percy
E. Eaymond. Pp. 101, 4 Plates

A New American Cybele. By J. B. Naeeaway
and Percy E. Eaymond. Pp. 6

Plastron of the Protostegina. By G. E. Wee-
land. Pp. 7

Description of New Species of Turtles of the
Genus Testudo, collected from the Miocene by
the Carnegie Museum; together with a De-
scription of the Skull of Stylemys Nebrascen'
sis. By Oliver P. Hay. Pp. 6, 8 Plates

The Miocene Beds of Western NebrasV-' _nd
Eastern Wyoming and Their Vertebrate
Faunae. By O. A. Peterson. Pp. 52, 11 Plates.

A New Species of Lonicera from Pennsylvania.
By Otto E. Jennings. Pp. 5, 1 Plate

Merycochoerus and a New Genus of Merycoido-
donts, with Some Notes on Other Agriochoe-
ridae. By Eael Douglass. Pp. 15, 1 Plate.

Some New Merycoidodonts. By Eael Douglass.
Pp. 18, 9 Plates. (Nos. 56 and 57 sold to-
gether.)

1.00
1.25

.10

1.00

.75

.16

.60

$1.00

1.00

.40

.10

.16

.15

.20
.10

.15
.05
.05
.10
.10
.05
1.50
.16
.15

.25

1.00
.05

l.OO

EEPRINTS FROM THE ANNALS OF THE CARNEGIE MUSEUM— Continued

68. On Further Collections of Fishes from Para-
guay. By Carl H. Eigenmann assisted by
Waldo Lee McAtee and David Peekins Wasd.
Pp. 48, 15 Plates 1.25

59. An Undetermined Element in the Osteology of

the Mosasauridae. By W. J. Holland. Pp. 5 $ .20

60. The Gasteropoda of the Chazy Formation. By

Pekcy E. Raymond. Pp. 58, 10 Plates 1.35

61. A Further Occurrence of Wynea Americana in

Pennsylvania. By Otto E. Jennings. Pp. 2,

1 Plate 05

62. A Preliminary Account of the Pleistocene Fauna

Discovered in a Cave Opened at Frankstown,
P'^pisylvania, in April and May, 1907. By
W. , Holland. Pp. 6, 2 Plates 10

63. Desciii- -ion of Vertebrate Fossils from the

Vicinity of Pittshurgh, Pennsylvania. By E.

C. Case. Pp. 8, 1 Plate 15

64. Notes on Ordovician Trilobites: DlaenidEe from

the Black Kiver Limestone near Ottawa,
Canada. By Percy E. Raymond and J. E.
Nareaway. Pp. 14, 3 Plates 20

65. Ehinoceroses from the Oligocene and Miocene

Deposits of North Dakota and Montana. By
Eael Douglass. Pp. 11, 2 Plates 25

66. Fossil Horses from North Dakota. By Eabl

Douglass. Pp. 11, 4 Plates SO

67. Some Oligocene Lizards. By Eael Douglass.

Pp. 8 .20

68. Description of the Type Specimen of Stenomylus

gracilis Peterson. By O. A. Peterson. Pp. 14. .25

69. Brief Description of Some New Species of Birds

from Costa Kica and a Record of Some Species
not Hitherto Eeported from that Country.
By M. A. Cakriker, Jr. Pp. 2 10

70. Notes on Costa Eican Formicariidse. By M. A.

Cabriker, Jr. Pp. 3 05

71. Vertebrate Fossils from the Fort Union Beds.

By Earl Douglass. Pp. 16, 2 Plates 45

72. A Preliminary List of the Lepidoptera of West-

em Pennsylvania Collected in the Vicinity of
Pittsburgh. By Henry Engel. Pp. 110 1.25

73. The Fauna of the Upper Devonian in Montana,

Pt. 1. The Fossils of the Red Shales. By
Percy E. Raymond. Pp. 18, 6 Plates 60

74. Description of a New Species of Procamelus from

the Upper Miocene of Montana, with Notes
upon Procamelus madisonins Douglass. By
Earl Douglass. Pp. 7, 3 Plates 30

75. Some Sections of the Conemaugh Series between

Pittsburgh and Latrobe, Pennsylvania. By
Percy E. Raymond. Pp. 12, 3 Plates 35

76. A Preliminary List of the Unionidae of Western

Pennsylvania, etc. By De. A. E. Oetmann.

Pp. 33 60

77. A Geological Eeconnaissance in North Dakota,

Montana, and Idaho; with Notes on Mesozoic
and Cenozoic Geology. By Eael Douglass.

Pp. 78, 7 Plates 1.00

Cabin (now West Liberty, W. Va.), &c.

78. Botanical Survey of Presque Isle, Erie Co., Pa.

By O. E. Jennings. Pp. 133, 30 Plates 2.75

79. Catalog of Sesciui-Centennial (Pittsburgh) Relics.

By Douglas Stewart. Pp. 30, 6 Plates 46

80. Dromomeryx, a New Genus of American Rumi-

nants. By Earl Douglas. Pp. 23, 5 plates. .30

81. Fossils from the Glacial Drift and from De-

vonian and Mississippian near Meadville,
Pennsylvania. 3y Wm. Millard. Pp.8 10

82. A New Species of Helodus. By Charles E.

Eastman. Pp. 2 05

83. In Memoriam. Charles Chauncey Mellor. By W.

J.Holland. Pp. 12, 1 Plate 15

84. Reports of Expedition to British Guiana of the

Indiana University and the Carnegie Museum,
1908. Report No. 1. By Carl H. Eigenmann.
Pp. 51 60

85. Reports of Expedition to British Guiana of the

Indiana University and the Carnegie Museum,
1908. Eeport No. 2. By Marion L. Duebin.
Pp. 18 25

86. Contributions to a Knowledge of Odonata of the

Neotropical Eegion, Exclusive of Mexico and
Central America. By P. P. Calveet. Pp. 207,
9 Plates 2.26

87. Deinosuchus hatcheri, a New Genus and Species

of Crocodile from the Judith Eiver Beds of
Montana. By W. J. Holland. Pp. 14 20

88. Reports on Expedition to British Guiana of the

Indiana University and the Carnegie Museum.
Report No. 3. By C. B. Blossee. Pp. 6, 3
Plates.

89. Preliminary Description of Some New Titanothe-

res from the Uinta Deposits. By Earl Doug-
las. Pp. 10, 3 Plates 25

90. An Annotated List of the Birds of Costa Rica

Including Cocos Island. By M. A. Caeriker,

Jr. Pp. 601, 1 Plate 3.00

91. The Geology of the Coast of the State of Alagoas,

Braj;il. By J. C. Branner. Pp. 18, 3 Plates ,40

92. Description of a Collection of Fossil Fishes from

the . jriituminous Shales at Riacho Doce, State
of Alagoas, Brazil. By David Stare Jordan.
Pp. 12, 9 Plates 55

93. Notes on Ordovician Trilobites, No. II. Asaph-

idae from the 3tekmantown. By Peecy E.
Raymond. Pp. 10, 1 Plate 35

94. Notes on Ordovician Trilobites, No. III. By

Peecy E. Raymond and J. E. Naeeaway.

Pp. 14, 2 Plates 35

95. Notes on Ordovician Trilobites, No. IV. By

Percy E. Raymond. Pp. 21, 3 Plates 40

96. Notes on a Collection of Fishes Made by James

Francis Abbott at Irkutsk, Sioeria. By David
Starr Jordan and William Francis Thomp-
son. Pp. 8, 4 Plates 30

97. South American Tetrigidse. By Laweence

Brunee. Pp. 55 1.00

98. Preliminary List of the Fauna of the Allegheny

and Conemaugh Series in Western Pennsyl-
vania. By Percy E. Raymond. Pp. 15, 5
plates 30

99. Results of an Ichthyological Survey About the

San Juan Islands, Washington. By Edwin
Chapin Starks. Pp. 52, 3 plates 75

100. Descriptions of a New Species of Pygldium. By

Gael H. Eigenmann. P. 1, 1 plate 10

101. The Brachiopoda and Ostracoda of the Chazy.

By Percy E. Raymond. Pp. 45, 4 plates ... .50

102. A New Camel from the Miocene of Western

Nebraska. By O. A. Peterson. • . Pp. 7, 4
plates 15

103. A Mounted Skeleton of Stenomylus hitchcockl,

the Stenomylus Quarry, and Remarks Upon
the Affinities of the Genus. By O. A. Peter-
son. Pp. 7, 4 plates 15

104. A Mounted Skeleton of Diceratherium cooki,

Peterson. By O. A. Peterson. Pp. 6, 1 plate .15

105. The Carnegie Museum Expedition to Central

South America, 1907-1910. By W. J. Hol-
land, Director. Pp. 4 15

106. A Brief Report Upon the Expedition of the

Carnegie Museum to Central South America.
By John D. Hasbman. Pp. 13 and Localities
at Which John D. Haseman Made Collections.
By Carl H. Eigenmann. Pp. 16 25

107. Descriptions of Some New Species of Fishes and

Miscellaneous Notes on Others Obtained Dur-
ing Expedition of Carnegie Museum to Central
South America. By John D. Haseman. Pp.
13, 7 plates 50

108. An Annotated Catalog of the Cichlid Fishes

Collected by the Expedition of Carnegie Mu-
seum to Central South America, 1907-1910. By
John D. Haseman. Pp. 45, 20 plates 1.25

109. Some New Species of Fishes i'rom the Rio

Iguassu. By John D. Haseman. Pp. 14, 13
plates 65

110. A Contribution to the Ornithology of the Ba-

hama Islands. By W. E. Clyde Todd and W.

W. Woethington. Pp. 77, 1 plate 75

If. 7/6

Publications of the Carnegie Museum, Serial No. 91.

MEMOIRS

OF THE

OAENEG-IE MUSEUM.

VOL. VII. NO. 2.

W. J. HOLLAND, Editor.

THE FOSSIL TUETLES OE THE UINTA EOMATION

By CEARLES W. GILMORE.

.. PITTSBURGH.

PCBUSHED BY THE ACTHOEITY OF THE BOAKD OF TRUSTEES OF THE

CARNEGIE INSTITUTE.
November, 1916.

PRICE LIST OF PUBLICATIONS OF THE CARNEGIE MUSEUM

ANNUAL EEPOETS OF THE DIRECTOR

1898, 30c. (scarce); 1899, '25c.; 1900, 30c. (scarce); 1901-16, 25c. each.

REPORTS OF PROCEEDINGS OF FOUNDER'S DAY

1898-1916, 35c. each.

ANNALS OF THE CARNEGIE MUSEUM

The Anuals are supplied to those who subscribe in advance in parts (paper-bound), as published, @ $3.50 per volume;
^ Vols. I-J?,| 1901-1916, bound in green cloth @ $4.00; bound in * Morocco @ $4.50.

MEMOIRS OF THE CARNEGIE MUSEUM

Tlio Memoirs are supplied to those who subscribe iu advance in parts (paper-bound), as published, at $10 per volume;
bound in green buckram at $10.75, and in half-morocco at $12.00.

VOL. L (1901-3)

No. 1. Diplodocus, Its Osteology, Taxonomy, etc. No. 3. Osteology of the Steganopodes. Shufelbt $2.75

Hatcher '. $3.00 ' ' 4. Clafssification of Chalcid Flies. Ashmead. . 6.50

' ' 2. Oligocene Canidae. Hatcher 1.50

VOL. n. (1904r^6)

No. 1. Osteology of Haplocamthosams, etc. Hatcher $2.00 " 6. Osteology of Diplodocus. Holland 1.50

' ' 2. Osteology of Baptanodon. Gilmorb 1.50 " 7. Osteology of Protostega. Wieland 75

" 3. Fossil Avian Remains from Airmissan. East- " 8. New Suilline Remains from the Jliocene of

man 1.00 Nebraska. Peterson 75

' ' 4. New Rodents and Discussion of Origin of • ' 9. Notes on the Osteology of Baptanodon, etc.

Daemonelix. Peterson 1.75 Gilmoke 1.00

No. 5. Tertiary of Montana. Douglass $1.00 ' ' 10. The Crawfishes of Pennsylvania. Ortmann 4.00

VOL. in.

No. 1. Archaeological Investigations in Costa Rica. No. 2. Osteology of Moropus. Holland and Peter-

Haktman $6.00 SOK $C.O0

VOL. IV.

No. 1. Early Chinese Writing. Chalpant $3.00 No. 5. New Carnivores from the Miocene of West-

' ' 2. Tormosan Fishes. Jordan 25 em Nebraska. Peterson $1.50

' ' 3. Entelodontidse. Peterson 2.50 ' ' 6. MonogTaph of the Najades of Pennsylvania.

' ' 4. Fishes of Formosa. Jordan and Richard- Pts. I and II. Ortmann 2.50

son 1.25 ' ' 7. Catalog Eocene Fishes from Monte Bolea in

Oamegle Museum. Eastman 3.00

VOL. V.

The Fresh Water Fishes of British Guiana. Eigenmann. $10 unbound; $10.75 cloth; $12.00 J morocco.

VOL. VI.

No. 1. Fishes from Korea. Jordan and Metz $1.50 No. 4. Fishes obtained in Japan. 1911. Jordan

' • 2. Lantern-fishes of Japan. Gilbert 1.00 and Thompson $3.50

" 3. Gymnotid Eels of Tropical America. Max Nos. 5-7. Fossil Fishes in the Carnegie Museum.

Mapes Ellis 1.50 Parts II-IV. Eastman 5.00

VII.

No. 1. The Cheirodontiuae. Eigenmann $3.50 No. 3. Ophidia from S. America. Griffin 2.00

' ' 2. Turtles of Uinta Formation. Gilmore 2.00

REPRINTS FROM THE ANNALS OF THE CARNEGIE MUSEUM

Nos. 1-12. See preceding lists. (Out of print.) 16. Osteology of Psittaci. Shufeldt 35

12. Skeleton of Titanotherium dispar. Hatcher. . $ .35 17. Annotated Catalogue of Genus Partula. II. H.

ftMTTTT 1 25

13. Fore Limbs and Manus of Brontosaurus. ' ' ' ' ' '

„ 18. Two New Bahaman Lepidoptera. Holland ... .10

Hatcher 60 ^^ Elosaurus Parvus. Peterson & Gilmore 15

14. The Trachodontidse. Hatcher 25 20. Boundary Controversy Between Pennsylvania

15. New Pennsylvania Thorns. Ashe 25 and Virginia, 1748-1785. Cbumrine 30

)*N 23 1917

MEMOIRS

OF THE

CAENEGIE MUSEUM.

VOL. VII. NO. 2.

THE FOSSIL TURTLES OF THE UINTA FORMATION.

By Charles W. Gilmore.

The finest and most complete assemblage of the remains of fossil turtles as
yet secured from the Upper Eocene of the Uinta formation has been brought to-
gether in the Carnegie Museum through the activities of its various expeditions
to Utah. By the kindness of Dr. William J. Holland, the Director of the Museum,
I have been permitted to study this collection, and the present paper presents the
results of my investigations.

The collection comprises more than fifty individuals, and was made by field-
parties conducted by Messrs. Earl Douglass and 0. A. Peterson and as an in-
cidental part of their search of the Uinta exposures for the remains of extinct
mammals. An important feature of this collection is the determination of the
exact geological horizons in which the specimens were found, thus establishing
a firm foundation for future correlative work.

The chelonian fauna of the Uinta formation is of peculiar interest, since it
marks the last appearance of several forms which had their beginning, so far as
our present records go, in the Upper Cretaceous and Lower Tertiary. Of the six
genera recognized in the present collection from the Uinta formation only three,
Anosteira, Amyda, and Testudo, are known to pass upward into the younger
Tertiaries. Anosteira is known from the Lower Oligocene of England, Amyda
reappears in the Miocene of the Atlantic coast, while Testudo is found in the
overlying Oligocene. It appears that the Uinta thus marks an important stage in
the history of the chelonian life of the Upper Eocene.

The Baenidae make their last appearance. The Dermatemydidse are rep-

101

102 MEMOIRS OF THE CARNEGIE MUSEUM

resented for the first time by tlic single genus and species Anosteira ornata Leidy.
The Emydidfe, suggestive of swampy conditions, in the number of species are the
most abundant turtles in the present collection. Seven species have been recog-
nized and larger collections will doubtless add several more to the list. The soft-
shelled river-turtles, Trionychidae, indicative of flowing water, are represented by
at least three species, one of which is as large as the existing Asiatic species. The
presence of true land-tortoises, Testudinidse, is represented by three species of
the genus Hadrianus, which includes tortoises some of which attain a length of more
than three feet, and the genus Testudo by a single species, the first recorded oc-
currence, in North America, of this genus below the Oligocene. The discovery of
the fossil remains of the lizard-like reptile Ghjptosaurus in the Uinta according to
Osborn' "hints as to the Floridan or south temperate conditions of climate."

There were a considerable number of specimens in the collection which were
too fragmentary for specific determination, and in two instances at least I am in-
clined to the opinion, that, had better material been available, distinctive characters
would have been found to show the presence of additional species new to the fauna.

I take this opportunity to protest most emphatically against the establishment
of new species of turtles based upon inadequate specimens, for it certainly cannot
serve any useful purpose to burden the literature with a lot of useless and meaning-
less names. The difficulties encountered in the present study, in recognizing to
which species certain specimens belonged, when almost perfect individuals were
at hand, shows the futihty of naming scraps with which subsequently discovered
material can never with absolute confidence be identified. There are perhaps
some few exceptions, for occasionally a fragmentary specimen is found which
shows a sculpture, or some character of such striking peculiarity, as to make it
stand out distinctly from all previously described forms. The present study has
demonstrated that a considerable variation within the limits of a species is to be
expected, and, until the range of these variations is determined, it is quite useless
to describe new forms based upon some small part of the carapace or plastron,
which shows some slight difference from described forms, when the very next speci-
men discovered may have these same features and yet have other characters to be
found in an adequate specimen, which show it to belong to a well-established
species.

I wish also to protest against the practice of naming species simply because
the specimen comes from a formation from which the genus to which it belongs
has not previously been recognized. In other words it is assumed that ' it is not

'■ Osborn, H. F., "Age of Mammals," 1910, p. 160.

GILMORE: the fossil turtles of the UINTA FORMATION 103

likely that a species known from a lower horizon continues over into a higher
horizon/ therefore a hunt for characters to separate it from the other species of
the genus is instituted, with the result that minor differences are magnified to
represent specific differences, when, had the specimen come from a formation in
which species of the genus were already known, it would in all probability have
found a resting-place within one of the described species.

In order to facilitate comparisons of the descriptions here given with those of
other described forms I have closely followed the order of arrangement used by Hay
in his monographic study of the fossil turtles of North America.

At this point I wish to acknowledge the assistance rendered me while this
paper was in the course of preparation. First of all I express my gratitude to Dr.
W. J. Holland for his hearty cooperation at all times, for the privilege granted me
of studying this fine collection of fossil turtles, and for his editorial oversight of
the work. I am under obligations to Dr. W. D. Matthew, of the American Museum
of Natural History, New York, for the loan of type-specimens, and to Dr. 0. P.
Hay, to whom, because of his wide knowledge of the turtles, I am especially in-
debted for invaluable advice upon numerous occasions. The text-figures were
made by the well-known artist, Mr. Rudolph Weber, the photographs are by Mr.
Arthur Coggeshall, of the Carnegie Museum.

Geological Occurrence.

All of the specimens considered in the present paper are from the Uinta forma-
tion as exposed in the Uinta Basin at the southern base of the Uinta Mountains,
and from that part of the basin which lies within Uinta County, Utah. The
geological positions of the various specimens as here given were taken from the
original field-labels which accompanied each specimen, so that these determina-
tions are wholly the work of Messrs. Earl Douglass and 0. A. Peterson, whose long
experience in the field insures the accuracy of their observations.

In 1895 the Uinta formation was divided by Peterson' into three levels, or
horizons, designated as follows, A (Lower), B (Middle), and C (Upper) Uinta.

The remains of turtles have now been found in all three horizons, though
judging from the present collection, individuals occur most abundantly in Horizon
B, but the number of species recognized in the collection is about evenly divided
between Horizons B and C. Up to the present time only one species is known from
Horizon A. Douglass-' has pointed out that "the lower portions of these deposits
may be, and probably are, contemporaneous with portions of deposits in the

2 Peterson, 0. A., Bull. Amer. Mus. Nat. History, VII, 1895, p. 74.

3 Douglass, Earl, Bull. Geol. Soc. of America, vol. 25, 1914, p. 418.

104 MEMOIRS OF THE CARNEGIE MUSEUM

Bridger and Washakie Basins and with other deposits elsewhere." If this be the
true condition, it may to some extent account for the presence of many species
common to the two formations.

Below is given a list of the identified species occurring in each of the three
subdivisions of the Uinta formation.

Horizon A (Lower Uinta).
Baena inflata sp. no v.

Horizon B (Middle Uinta).

Baena arenosa Leidy, E. hollandi sp. nov.,

B. emilice Hay, E. uintensis Hay,

B. inflata sp. nov., Hadrianus utahensis sp. nov.,

B. platyplastra sp. nov., Testudo uintensis sp. nov.,

B. gigantea sp. nov., Amyda egregia Hay,

Echmatemys callopyge Hay, A. scutwnantiquum (Cope).

Horizon C (Upper Uinta).

Baena emilice Hay, Hadrianus corsoni (Leidy),

Echmatemys douglassi sp. nov., H. robustus sp. nov.,

E. depressa sp. nov., Anosteira ornata Leidy,

E. obscura sp. nov., Amyda sp.,

E. pusilla? Hay, Glyptosaurus sp. indet.

Six genera and twenty species are recognized in the present collection, whereas
in 1908, at the time Doctor O. P. Hay published his "Fossil Turtles of North
America" only four genera and five species were accredited to the Uinta formation.
These were as follows:

Baena emilice Hay, Hadrianus tumidus Hay,

Echmatemys callopyge Hay, Amyda crassa Hay.

E. uintensis Hay,

The two latter species have not been recognized in the present collection,
although each of the others is represented by from two to six individuals, so that
altogether six genera and twenty-two species of fossil turtles have now been
found in the Uinta formation. The known geological range of these species is
graphically shown in the accompanying table.

GILMORE: the fossil turtles of the UINTA FORMATION

105

Geological Range of Recognized Species.

Wasatch.

Bridger.

Uinta.

Oligocene.

A

JS

c

n

A

£

c

Baenidae:

Baena arenosa Leidy,

X

X

X

X
X
X
X
X

B. emilice Hay,

X

B. inflata sp. nov.,

X

B. platyplastra sp. nov.,

1

B. qiqanlea sp. nov.,

Dermatemydidae :

Anosteira ornala Leidy,

X

X

X

X

Emydidae:

Echmatemys callopijge Hay,

X

E. douglassi sp. nov.,

X?

E. hollandi sp. nov.,

X

E. depressa sp. nov.,

X
X
X?

E. obscura sp. nov.,

E. pusilla? Hay,

X

E. uiniensis Hay,

X

Testudinidee:

Hadrianus corsoni (Leidy),

X

X
X
X

H. utahensis sp. nov.,

H. robustus sp. nov., ; . . . .

Testudo uintensis sp. nov.,

X

X
X
X

Trionychididae:

Amvda eareaia Hav

X

A. crassa Hay,

A . scutumantiquum (Cope) ,

X
X

Lacertilia:

Glyptosaurus sp. indet

X

X

Summary of Material forming the Collection of Turtles from the Uinta Formation in the Carnegie

Museum.

Catalog No. Geol. Horizon.

Baenidse:

Baena arenosa Leidy, 2356 B.

B. emilim Hay, 2159 B.

B. " 3243 B.

B. " 3253 C.

B. " 3257 B, orC.

B. " 3443 Blower.

B. " 3444 Blower.

B. gigantea sp. nov., 3441 B lower.

B. inflala sp. nov., 3406 A. -

B. " 3137 B.

B. " 3442 Blower.

B. platyplastra sp. nov., 3227 B.

B. sp. indet., 2372 B.

B. " " , 3246 ?

B. " " 3255 B.

B. " " 3247 ?

B.f" " 3271 B.

BJ" " (skull) 2956 C.

B. " " 3447 Blower.

106 MEMOIRS OF THE CAENEGIE MUSEUM

Dermatemyclidffi :

Anosteira ornata Leidj'', 2954 C.

Emydidse:

Echmatemys callopyge Hay, 2157 B.

E. " 2371 R.

E. douglassi sp. nov. , 3244 C?

E. depressa sp. nov., 2936 C.

E. hollandi sp. nov., 3249 B.

E. obscura sp. nov., 3252 C.

E. pusilla? Hay, 3282 C.

E. uintensis Hay, 3270 B.

E. " 2158 B, orC.

E. " 2397 B?

E. sp. indet., 2393 B.

E. " " 2361 B.

E. " " (skull and neck) 2387 B.

Testudinidae:

Hadrianus corsoni (Leidy), 3403, 3404 C.

H. robustus sp. nov., 3342 C.

H. utahensis sp. nov., 2343 B, or C.

H. sp. indet., 2376 B.

H. " " 3256 B, orC.

Testudo uintensis sp. nov., 2331 B.

Trionychidae:

Amyda egregia Hay, 3254 B.

A. "? 3258 B.

A. scutumantiquum (Cope), 3272 B.

A. " 3330 ?

A. sp. indet., 3254 B.

A. •' " 3177 C.

A. " " 2981 C.

A. " " 3260 C.

A. " " 3134 B.

A." " ....3050 C.

A." " 3019 B.

A. " " 3245 C.

A." " 3285 C.

Incertx sedis:

Gen. and sp. indet., 2394 B.

" " " " 2374 B.

Fragments of Baena, Amijda, etc., 3250 C.

Gen. and sp. indet., 2982 C.

" " " " 2395 B.

" " " " 3445 B?

Anguidffi:

Glyptosaunis sp. indet., 3405 C.

GILMORE: the fossil turtles op the UINTA FORMATION 107

Family BAENID.E Cope.

In the present collection from the Uinta formation of Utah nineteen specimens
were sufficiently well preserved to be identified as pertaining to the genus Baena.
These were found in all three subdivisions of the Uinta, being distributed as follows:
one specimen, Baena inflata, from Horizon A; thirteen from Horizon B; two from
horizon C; and three for which the data for the horizon were uncertain, or not given.

Five species of this genus are now recognized as occurring in the Uinta forma-
tion, three of which are here described as new. Only one of the recognized species,
Baena arenosa, is found to occur in other geological epochs, and no member of this
family is known to range above the Uinta.

Genus Baena Leidy.

1. Baena arenosa Leidy.

Plate XVIII, fig. 1; text-fig. 1.

Baena arenosa Leidy, Proc. Acad. Nat. Sci. Phila., 1870, p. 123; U. S. Geol. Surv.
Wyoming, etc., 1870 (1871), p. 367; U. S. Geol. Surv. Montana, etc., 1871
(1872), p. 368; Contrib. Ext. Vert. Fauna West. Terrs., 1873, pp. 161, 343,
pi. 13, figs. 1-3; ?pl. 15, figs. 1-5; pi. 16, figs. 8, 9.— Cope, ?Append. LL of
Ann. Report Chief of Engineers, 1875, p. 96; ? Wheeler's Surv. 100th Merid.,
1877, p. 52, pi. 24, fig. 32.— Baur, Proc. Acad. Nat. Sci. Phila., 1891, p. 426.
—Hay, Bibliog. and Cat. Foss. Vert. N. A., 1902, p. 437; Foss. Turtles of
N. A., 1908, pp. 67-71, pi. 12; pi. 13, fig. 1; pi. 14, figs. 1-3, text-figs. 44-51.

Baena affinis Leidy, Ann. Report U. S. Geol. Surv. Wyoming, etc., 1870 (1871),
p. 367.

This species is represented in the collection by a single specimen, Cat. No. 2356.
Collected by Earl Douglass and party, June 18, 1908, from Horizon B, Uinta
formation. Upper Eocene, east of Dragon- Vernal road between White and Green
Rivers, Uinta Basin, Uinta County, Utah.

The specimen consists of a fairly complete carapace, lacking the posterior
borders and the peripherals of both sides, the plastron lacks portions of both
anterior and posterior lobes. It represents an individual of approximately the
same size as the type of the species (Cat. No. 103, U. S. National Museum), with
which it has been carefully compared. This comparison shows several differences,
but such as exist are not considered of sufficient importance to separate the speci-
mens specifically. The sculpture of the carapace is rough and uneven, consisting

108

MEMOIRS OF THE CARNEGIE MUSEUM

of various longitudinal, transverse and oblique ridges, especially within the areas of
the vertebral scutes, this part of the carapace being fairly smooth in the type.

Fig. 1. Carapace of Baena arenosa Leidy. C. M. No. 2356. X H- '«., neural scute; v.s. 1-v.s. 5,
vertebral scutes one and five.

In the measurements of the vertebral scutes it also differs from the type, but
agrees almost exactly with the type of Baena affinis Leidy, which is now regarded
by Hay, following Leidy and Cope, as being a synonym of B. arenosa. The verte-
bral areas of this type are also in accord with the present specimen. In order
to show the close agreement of the present individual with the above-mentioned
specimens a table giving the comparative measurements of the vertebral scutes
of each is herewith appended.

Comparative Measurements of Vertebrals.

Length.

Width.

No. 2356.

Type of B. arenosa.

Type of £. affinis.

No. 2356.

Type of B. arenosa.

Type of B. affinis.

1

44

— .

50

57

72

2

76

77

73

62

73

61

8

76

74

73

64

77

64

4

60

61

60

68

78

63

5

—

68

55

88

96

75

Since the anterior portion of the carapace is missing in the type of the species
and in all subsequently discovered specimens the complete anterior margin in the

GILMORE: the fossil turtles of the UINTA FORMATION

109

present specimen is therefore worthy of brief description. The nuchal scute is
small, having a length of 12 mm., and a transverse diameter of 17 mm. The
median part of the anterior border projects shghtly beyond the general contour of
the shell. The nuchal is flanked on either side by small rectangular marginals.
The arrangement and proportions of the scutes in front of the first vertebral are
very similar to those of Baena antiqua Lambe.

This specimen shows many of the sutures between the costals, but on the
median dorsal surface their complete coalescence renders it impossible to differ-
entiate the neurals. There are no supernumerary costal scutes on either side of the
first vertebral such as found in many species of this genus and occasionally in
individuals pertaining to the present species.

The plastron agrees almost exactly with the type in size and proportions, and
especially in the sculpture of the surface and the course of the various sulci. The
accompanying table gives a comprehensive comparison of the principal measure-
ments of the plastra.

Comparative Measueements of Plastra.

Width of bridge

Length of anterior lobe .
Width of anterior lobe. .
Width of posterior lobe .

No. 2356.

145

82

108

96

Type of B. arenosa.

140

77

107

98

Type ofB. affinis.

125

77
104
114

The discovery of the present specimen in the Uinta formation of Utah in-
creases somewhat the known geographical as well as the geological range of Baena
arenosa. The type of the species is from the Bridger deposits at the junction of
the Big Sandy and Green rivers in southeastern Wyoming. Hay^ has identified
this species from the Washakie, the uppermost division of the Bridger. The type
of B. affinis is from level B, of the Bridger, which represents the middle of that
formation. Both Cope and Hay have identified specimens from the Wasatch of
New Mexico as pertaining to this species, these being the most ancient known at
the present time. With a geological range beginning in the Wasatch and ending
in the middle of the Uinta, Baena arenosa enjoys the distinction of having the widest
geological distribution of any species of the genus.

2. Baena emiliae Hay.

Baena emilice Hay, 0. P., Fossil Turtles of North America, 1908, pp. 80-81, PI.
XX, fig. 1 ; text-figs. 67, 68.
Six specimens in the present collection are identified as pertaining to Baena

* Hay, 0. P., " Fossil Turtles of North America," Pub. Carnegie Institution, Washington, 1908, pp. 67-68.

110 MEMOIRS OF THE CARNEGIE MUSEUM

emilice Hay. The better preserved specimen, C. M. No. 3443, consists of a nearly
complete carapace and plastron, the former lacking small portions of the hinder
borders posterior to the inguinal notches. The missing margin of the right side
was apparently lost in life, as shown by the healed condition of the bone at this
point. The specimen was collected by Earl Douglass during the season of 1915 at
Wagon-hound Bend, on White River, Uinta County, Utah, from the lower part
of Horizon B. It will be seen from the table of comparative measurements given
below that this specimen has about the same dimensions as the type of the species.
It differs, however, in the more angularly rounded contour of the front lobe, a
feature in which it also is different from three of the other specimens here referred
to this species. Whether this difference represents a sexual character, or is only
an individual variation, I am unable to determine.

In many respects the present specimen is very close to the type of Baena clara
from the Bridger formation, but the great length of the third vertebral scute, as
compared with the others of the series, is regarded by Hay as one of the chief dis-
tinguishing characters of the species, and together with the much shorter pos-
terior lobe, as compared with the longer lobe in B. clara, appears to show that its
closest affinities are with the present species.

An anterior portion of a carapace and plastron, C. M. No. 3253, on account
of its close general resemblance to the specimen discussed above, is provisionally
referred to the same species. This specimen is from Horizon C of the Uinta for-
mation and is the only individual in the collection, referred to the present species,
which is positively known to have come from that horizon, all of the others having
been found in strata belonging to Horizon B. It was collected by Earl Douglass
two or three miles west of Well No. 2, Uinta County, Utah.

A third specimen, C. M. No. 2159, consisting of a complete plastron and the
entire central part of the carapace, but lacking portions of both sides, is also
referred to this species. It was collected by Earl Douglass in 1908, in the Devil's
Playground, Uinta Basin, Uinta County, Utah, from Horizon B (near top) or C
(near base).

A fourth individual, C. M. No. 3243, has a nearly complete carapace and
plastron, the latter lacking the posterior lobe. It was collected by Messrs. Earl
Douglass and J. T. Goetschius, October 2, 1908, about one mile northeast of Well
No. 2, "near first gap," Uinta County, Utah, from Horizon B.

The fifth specimen, C. M. No. 3257, consists of a carapace and plastron, the
former lacking some of the posterior border. This turtle also was collected by
Earl Douglass and J. T. Goetschius, July 30, 1908, south of Kennedy's Hole and

GILMORE: the fossil turtles of the UINTA FORMATION 111

west of the Dragon- Vernal road, Uinta Basin, Uinta County, Utah, from Horizon

B, or C.

The sixth specimen, C. M. No. 3244, consists of the greater part of the carapace
and plastron. The carapace has portions of the rim missing on both the anterior
and posterior ends, the plastron lacks the anterior lobe. The specimen was
collected by Earl Douglass in 1915, at Wagon-hound Bend on White River, Uinta
County, Utah, from the lower part of Horizon B.

The type of Baena emilice is in the American Museum of Natural History,
and was collected by Mr. 0. A. Peterson in 1884, from the middle Uinta of Utah.
Geologically therefore all of the known specimens, including the type, two other
specimens referred to the species by Hay, and the six specimens under con-
sideration, came from approximately the same horizon, and from neighboring

localities.

It may be shown hereafter, when larger collections shall have been made, that
more than one species is represented by the six specimens here referred to B. emilice.
When compared with one another there are differences which appear to divide
them into three groups, as follows: Nos. 2159 and 3243, having relatively narrower
vertebral scutes and narrower plastral lobes and bridges than the type, or other
specimens here referred to B. emilice; Nos. 3244 and 3257, having wider vertebral
scutes and a more depressed shell than the type; and No. 3443 with a wider and
more angularly rounded anterior lobe, larger intergulars, and narrower pectorals.
The latter specimen in all of these particulars is different not only from the type,
but from all of the other specimens discussed above, with the exception of the
fragmentary specimen No. 3253, which, in so far as the two can be compared,
appears to be very close to No. 3443. In nearly all other respects these specimens
agree closely with the type of the species. The differences enumerated above are
not considered important enough to warrant the separation of these turtles into
distinct species. When the considerable sexual and individual differences ob-
servable in a series of living turtles of one species and from one locality are con-
sidered, it appears to me that the specimens before me are well within the limits of
a given species. I am inclined to the belief that specimens Nos. 2159 and 3243
may be females of this species, but as to this I cannot be certain. The discovery of
more material may possibly show that more than one species is represented in
these specimens, but at this time, especially in the light of a recent examination
of a large series of living turtles, I do not feel justified in the establishment of new
species on such slender distinguishing characters as have been observed. For the
present, at least, I refer all the six specimens to Baena emilice Hay.

112

MEMOIRS OF THE CARNEGIE MUSEUM

In order to place on record the proportional variations within the species I
have prepared the table of comparative measurements given below:

CoMPAR-^TivE Measurements of Vertebrals.

Length.

Width.

Type.

No.
3443.

No.

3244.

No.
3257.

No.
2159.

No.
3243.

Type.

No.
3413.

No. 3244.

No. 3257.

No. 2159.

No. 3243.

1

55

53

—

51

47

38

68

73

85

65

63

2

80

81

—

82

71

77

63

79

71

87

64

56

3

92

96

97

89

87

88

75

82

80

90

68

63

4

70

69

74

72

69

67

78

77

80

91

69

65

5

70

62

52

—

—

62

98

101

90

—

96

Comparative Measurements of Car^^pace and Plastron.

Greatest length of carapace. . .
mdth ■ "
" length of plastron. . .
" " anterior lobe. .

width " _ "
" length posterior lobe
width
Width of bridge

Type.

No. 3443.

No. 3244.

No. 3257.

No. 2159.

368

364

365e

375e

328

294

292

310

310

320

307

310e

300

80

76

75

77

110

105

106

105

102

83

84

80

98e

80

122

114

119

117

108

160

150

149

146

140

No. 3243.

366

260

73
98

105
146

e, estimated.

3. Baena inflata sp. no v.
Plate XIX; text-figs. 2 and 3.

Type : C. M. No. 3406, consisting of a carapace and plastron, the former lacking
the posterior end back of the middle of the fourth vertebral, the latter a small
portion of the anterior lobe; collected by 0. A. Peterson in 1912.

Locality : McCook Canyon, White River, Uinta County, Utah.

Horizon: Horizon A (near top), Uinta formation. Upper Eocene.

The type of this species is but little crushed and, except for the parts which
are missing, is in a beautiful state of preservation. The surface of the carapace
and plastron are everywhere covered with fine pustular elevations, forming a
shagreened surface. The pustules on the carapace are coarser than those on the
plastron. The surface of the carapace is also somewhat uneven, and laterad to
the second, third, and fourth vertebrals there are some heavy longitudinal wrink-
lings. These are most numerous laterally at the junction of the third and fourth
vertebrals. The pustular ornamentation of the carapace appears to be very
similar to that of Baena sirna Hay, but not so coarse.

The greatest length of the shell is estimated to have been about 400 mm. ; its
greatest width at the center is 310 mm. In outline the front of the shell is evenly,

GILMORE: the fossil turtles of the UINTA FORMATION

113

but broadly, rounded, resembling in its general contour B. sima Hay. The shell
is flat transversely in the region of the vertebral scutes, but from one border to the
other it is broadly convex. One of the distinctive features of this species is the
decided transverse inflation or swelling of the mid-costal region, which gives the
shell the appearance of being puffed out on the sides. This swelling lies largely
within the areas of the second costal scutes, and it is to this feature that the specific
name refers.

Over the posterior legs the margins of the shell begin to flare outward and
slightly upward, and at this point the border is heavy and rounded but becomes
thinner posteriorly. In front of the axillary notches the border has a thickness of
31 mm., but rapidly thins toward the center, where it measures only 8 mm., the
edge being obtusely rounded. The bones of the carapace are so thoroughly coos-
sified that but few of the sutures can now be made out. The sulci, however, can
in most instances be clearly traced.

Fig. 2. Carapace of Baena inflata. C. M. No. 3406, Tyije,
four; V.S. 1, V.S. 4, vertebral scutes one and four. One-fourth natural size

C.S. 1, C.S. 4,

costal scutes one and

The vertebrals as in nearly all Eocene Baenidai are longer than wide. The
sides of the vertebrals, excepting the first, which is hexagonal with a very narrow
anterior end, are bracket-shaped. It wiU be observed, that, as in Baena emilm,
the third vertebral is the longest of the series. Along the center of vertebrals two,

114

MEMOIRS OF THE CARNEGIE MUSEUM

three, and four is a narrow low ridge, on either side of which are parallel grooves
much as in B. emilice. The principal dimensions of the vertebrals as well as those
of a second individual, C. M. No. 3442, referred to this species, are given in the
accompanying table.

Dimensions of Veetebkals.

Length.

Width in front.

Greatest width.

Type.

No. 3442.

Type.

No. 3442.

Type,

No. 3442.

1

61

53

28

75

72

75

2

90

92

62

63

74

77

3

99

98

62

64

78

81

4

—

90

68

70

78

81

5

—

—

—

53

—

—

The nuchal scute is rectangular, being about 17 mm. long and 30 mm. wide on
the free border. It is bordered on either side, as shown in specimen No. 3442,
by small triangular first marginals. The second marginal has its greatest width

Fig. 3. Plastron of Bai-nn inflata C. M. No. 3406, Type, restored after C. M. No. 3442. One-fourth
natural size.

(36 mm.) on the free border. The first vertebral is bordered on either side by
small triangular supernumerary costal scutes, though there is no indication of these
in specimen No. 3442.

It is estimated that the plastron had a total length of about 365 mm. It is

GILMORE: the fossil turtles of the UINTA FORMATION 115

slightly convex transversely throughout its length, and this convexity on the bridge
area continues evenly to the borders of the shell, so that these borders stand 36 mm.
above the level of the plastron at the center. The anterior lobe has its greatest
width (135 mm.) at the base; and at a point half-way to the anterior end it measures
99 mm. The sides of the lobe converge gradually from the base to the anterior
end, which appears to have been rounded. The width of the bridge is 175 mm.

The posterior lobe is tongue-shaped, shallowly, but broadly notched. Its
length on the midline is 97 mm., with a width at the base of 139 mm. The notch
has a depth of 6 mm. at the center.

Excepting those of the anterior lobe all of the sulci and sutures on the plastron
can be clearly made out. The mesoplastrals widen rapidly on either side of the
midline. At the center the right scute measures 51 mm. The width of the right
hypoplastral at the center is 104 mm.; of the left hypoplastral 91 mm. The xiphi-
plastrals are 62 mm. wide on the midline. The pectorals meet on the midline for
a distance of 75 mm.; the abdominals for 49 mm.; the femorals for 80 mm.; the
anals for 47 mm.

On the right side are three inframarginal scutes, the form of which is well
shown in Fig. 3.

A second specimen, C. M. No. 3442, belonging apparently to this species, was
collected by Earl Douglass in 1915 from the lower part of Horizon B of the Uinta
formation, at Wagon-hound Bend on White River, Uinta County, Utah. This
turtle consists of a carapace and plastron, both of which have small portions
missing from their posterior ends. In size, general contour, and the dimensions
of the dermal scutes, the specimen closely resembles the type. The inflation of the
sides of the carapace, which forms such a conspicuous feature in the type, is almost
entirely wanting in this individual. Its absence may be attributed in part, at least,
to crushing, for both sides in this respect have somewhat suffered. There are also
no supernumerary costal scutes at either side of the first vertebral, and in their
absence the first vertebral is tetragonal, whereas in the type it is hexagonal with
the narrow end in front. This specimen shows small triangular first marginals
on either side of the nuchal, and in the drawing of the type (Fig. 2), this region,
which is missing, has been restored after this specimen. It also gives the complete
form of the anterior lobe (See Fig. 3), which in its general contour closely resembles
Baena sima Hay.

The greater part of an anterior lobe, C. M. No. 3137, which was collected by
Earl Douglass in the strata of Horizon B of the Uinta formation, near Well No. 2,
Uinta Basin, Utah, in 1908, is regarded as belonging to Baena inflata. It is from

116

MEMOIRS OP THE CARNEGIE MUSEUM

an individual having tlie same proportions as the type, and shows on the dorsal
surface the triradiate shape of the entoplastron (See Fig. 4, 1). The entoplastron

Fig. 4. Anterior lobes of Baena inflata. 1, and :?, superior and inferior views of C. M. No. 3137, ent.,
entoplastron; hum., humeral scute. S, inferior view of anterior lobe of C. M. No. 3442. AU figures one-fourth
natural size.

has a length of at least 50 mm.; a width of 44 mm. Transversely the lower surface
of the lobe is broadly convex. On the dorsal surface immediately posterior to the
anterior border the bone is scooped out by a shallow transverse depression. The
lateral borders are bevelled off almost perpendicularly, while in front the border
is rounded. The bone along the borders has a thickness of 13 mm., in front of the
center of only 9 mm., at the middle on the posterior broken border of 21 mm.

4. Baena gigantea sp. nov.
Plate XX, figs. 1 and 2; text-figs. 5, 6, and 7.

Type: C. M. No. 3441, consisting of nearly a complete shell. The carapace
lacks portions of the posterior margins on either side of the middle, a small section
of the right anterior border, and the peripherals of the right side above the bridge.
The plastron has the greater part of the posterior lobe missing. Collected by
Earl Douglass, in 1915.

Locality: Wagon-hound Bend, on White River, Uinta County, Utah.

Horizon: Lower part of Horizon B, Uinta formation. Upper Eocene.

The type of the present species is the largest species of the genus as yet dis-
covered. It is estimated that the carapace had an axial length of about 535 mm.
The greatest width, which is near the center, is about 420 mm. The bones of the
carapace are all thoroughly coossified and only the sutures defining the right half
of the mesoplastron can be detected and then only with difficulty. The shell is
oval in outline, in this respect resembling Baena clara Hay, though the oval is
somewhat more elongate than in that species. The carapace has been slightly
crushed on the right side, as may be seen by examining Plate XX, fig. 1.

The front of the carapace is decidedly projecting. The missing posterior

GILMORE: the fossil turtles of the UINTA FORMATION

117

borders render it impossible to determine the character of the scallops on the
hinder end. The surfaces of both the carapace and plastron are roughened with
coarse pustular elevations, though these are more sparsely placed than in either
Baena sima or B inflata. With the exception of this pustular roughening the
surfaces are comparatively smooth, there being no longitudinal ridges or grooves,
such as are commonly found in many species of this genus from the Eocene. The
vertebral areas are a'so free from median ridges and channels.

The nuchal scute resembles in outline that of Baena hatcheri. It has a fore-
and-aft diameter of about 58 mm., and a transverse diameter of 88 mm. The

Fig. 5. Baena gigantea, carapace of C. M. No 34-11, Tyjje. About one-fifth natural size.

unusual length of the nuchal appears to be one of the distinctive features of this
species. All of the sulci are distinctly impressed. At the left side of the nuchal
is a subrectangular first marginal, which has a length on the free border of 43 mm.
The total number of peripherals cannot be determined from this specimen.

There are the usual five vertebrals and these are relatively wide, and differ
from those of all other Eocene Baenidae, except B. emilice, in having the fourth
considerably wider than long. The sides of the vertebrals posterior to the first are

118

MEMOIRS OF THE CARNEGIE MUSEUM

only slightly bracket-shaped. The first is hexagonal, very narrow in front, in this
respect closely resembling the first in Baena inflata. The surfaces within the first,
second, and third vertebral areas are flattened, but the fourth and fifth are trans-
versely broadly convex. The principal dimensions of the vertebrals are given
in the accompanying table.

Dimensions of Vertebrals.

Length.

Width in front.

Greatest width.

1

58

34

. 98

2

122

87

105

3

126

96

113

4

88

94

106

5

102e

88

152c

e, estimated.

As in Baena riparia Hay and B. hatcheri Hay, there are five costal scutes. A
small supernumerary scute is situated on each side of the first vertebral, showing a
difference from the former species by bordering on the nuchal, whereas in B.
riparia these scutes are not in juxtaposition.

On the plastron only the sutures defining the mesoplastron on the right side

Fig. 6. Baena gigantea, pl&stron. Type, C. M. No. 3441. About one-fifth natural s

size.

GILMORE: the fossil turtles of the UINTA FORMATION

119

are traceable, and these show them to be narrow at the midhne (18 mm.), but
expanding toward their outer extremities where the width is 112 mm. The an-
terior lobe is elongated antero-posteriorly, and turns upward with a well-defined
sweep toward the carapace, as shown in Fig. 7. Its greatest length is 137 mm.;
its greatest width 170 mm. at the base; at a point half-way to the tip measuring
117 mm. in width. The sides of this lobe gradually converge from the base to
near the anterior end, which rounds in with a shallow but broad median emargina-
tion on the anterior end. The posterior lobe is largely missing, though enough of
the base remains to show that it had a width of 160 mm. The width of the bridge
is 190 mm.

Fig. 7. Baena giganlca, lateral view of the carapace and plastron, C. M. No. 3-441. Type specimen,
one-fifth natural size.

The sulci defining the intergular scutes cannot be traced. The intergulars
meet on the midline for a distance of 29 mm.; the humerals 112 mm.; the pectorals
97 mm.; the abdominals 55 mm. The number of inframarginals on the bridge
cannot be determined in this specimen.

This species may be distinguished from all others of the genus by its larger size,
the great length of the nuchal scute, and differences in the relative dimensions of
the vertebral scutes. The contour of the anterior lobe of the plastron, its greater
relative length, and especially its decided upward curvature are all features which
serve to distinguish this species. In the presence of five costal scutes the type of
this species agrees with several species of the genus, especially Baena riparia and
B. hatcheri, but it differs from those forms by the decidedly longer nuchal scute and
in the apparent absence of marginal scutella. In the greater length of the third
vertebral this specimen is like B. emilice from the same formation, but the greater
relative widths of all of the vertebrals and especially the shortness of the fourth,
together with other differences to be observed in the plastron, at once distinguish
it from that species.

120 MEMOIRS OF THE CARNEGIE MUSEUM

5. Baena platyplastra sp. nov.
Plate XVIII, fig. 2; text-fig. 8.

Type: C. M. No. 3227, consisting of a plastron lacking the anterior portion
of the anterior lobe, matrix cast of the carapace, at either end of which remain a
few fragmentary parts of the carapace. Collected by Earl Douglass and J. F.
Goetschius, August 5, 1908.

Locality: Northeast of Well No. 2, Uinta Basin, Uinta County, Utah.

Horizon: Horizon B, Uinta formation. Upper Eocene.

The type specimen represents one of the larger species of the genus. It is
distinguished from all other described species of Baena by the extremely flat and
thin plastral bones with sculptured inferior surfaces. Its large size and the absence
of a median emargination on the posterior lobe are features which also assist in
distinguishing this species.

The ornamentation of the plastron consists of low ridges and shallow furrows,
the former being short, sometimes straight, but usually bent or anastomosing.
The effect of the whole may be best expressed as resembling a coarse, shagreened

Fig. 8. Plastron of Baena platyplastra, C. M. No. 3227. Type, one-fourth natural .size.

leather. The sutures in the type have all coalesced, but their courses are indi-
cated by ridges crossing them at right angles. These cross-ridges are especially
pronounced on the median suture between the axillary notches.

GILMORE: the fossil turtles of the UINTA FORMATION 121

It is estimated that the entire shell had a length of about 495 mm., and a height
at the center of about 170 mm. The plastron had a length of about 420 mm. The
anterior lobe at the base is 156 mm. wide. The bridge is 180 mm. wide.

The posterior lobe has a width at the base of 152 mm., and a length of 128 mm.
The lateral borders of the lobe are nearly straight and converge nearly the entire
length of the lobe, there being a slight constriction at the anal-femoral sulcus. The
posterior end of this lobe is broadly but evenly rounded and without median emar-
gination. At the anal-femoral sulcus the transverse measurement is 107 mm.

The mesoplastrals are solidly coossified with the contiguous bones and their
boundaries can only be determined by the ridges which cut them at right angles.
At the midline these bones have a width of 33 mm., at their outer ends they expand
to 55 mm. in width.

As in Ba'ena sima Hay, the median sulcus runs a very tortuous course, as is well
shown in Fig. 8. The pectorals meet on the median line for a distance of 77 mm.;
the abdominals for 70 mm.; the femorals for 68 mm.; the anals for 85 mm. The
anal-femoral sulcus runs in from the border a short distance, then turns abruptly
forward, then again turns at right angles toward the median line to meet the scute
of the opposite side. In the shape of the anal scutes it resembles Ba'ena arenosa
and more especially B. clara.

Owing to the damaged condition of the bridges the number of inframarginal
scutes cannot be determined.

Family DERMATEMYDID^ Gray.

The family Dermatemydidse is represented now for the first time in the Uinta
formation by the single genus and species, Anosteira ornata Leidy. This is the
latest recorded occurrence of this genus for North America, although in England
it is known to range upward into the Lower Oligocene.

Genus Anosteira Leidy.

6. Anosteira ornata Leidy.

Anosteira ornata Leidy, Proc. Acad. Nat. Sci. Phila., 1871, p. 102; Ann. Report
U. S. Geol. Surv. Montana, etc., 1871 (1872), p. 370; Contrib. Ext. Fauna
West. Terrs., 1873, pp. 174, 341, pi. XVI, figs. 1-6.— Hay, Bibliog. and Cat.
Foss. Vert. N. A., 1902, p. 447; Bull. Amer. Mus. Nat. Hist., XXII, 1906,
p. 157, figs. 2, 3; Fossil Turtles of North America, 1908, pp. 279-281, PI. XLIII,
figs. 1, 2; text-figs. 352-354.

122 MEMOIRS OF THE CARNEGIE MUSEUM

Anostira ornata Cope, Ann. Report U. S. Geol. Surv. Wyoming, etc., 1872 (1873),

p. 621; Amer. Naturalist, vol. XVI, 1882, p. 989, fig. 7; Vert. Tert. Form.

West, 1884, p. 128.— DoLLO, Bull. Mus. Roy. Belgique, IV, 1886, p. 93, PI.

XI, figs. 7, 8.

A fragmentary specimen. No. 2954, collected by O. A. Peterson August 24,
1912, from Horizon C, Uinta formation, on White River near Ouray, Uinta County,
Utah, is provisionally identified as pertaining to the above genus and species.
This specimen consists of the articulated nuchal, first and second neurals, with
portions of the abutting costals of both sides, parts of several disarticulated costals,
eleven peripherals, several of which are complete. The plastron is represented
by the right hypoplastron lacking a portion of its outer extremity and many frag-
mentary parts.

The specimen has been carefully compared with the figures and descriptions
given by Leidy and Hay, and especially with one of Leidy's cotypes No. 4062,
now in the U. S. National Museum, and, with the exception of slight differences in
size, it agrees closely in nearly all respects. The present specimen is of about the
same size as one individual in the American Museum of Natural History described
and figured by Hay in his Turtles of North America, but is considerably smaller
than the cotype of Leidy mentioned above.

All of the specimens described by Leidy are supposed to have come from
the lower portion of Horizon B in the Bridger as exposed in the neighborhood of
old Fort Bridger, Wyoming. The specimen described by Hay in the publication
cited is from the third division of Horizon C of the Bridger on Henry's Fork, Wyo-
ming. The discovery of the specimen considered here now extends the geological
range of this species into the uppermost horizon of the Uinta formation.

The nuchal has a length of 15 mm., a width on the free border of 23 mm.
The free border is subacute and is not so deeply excavated in front as in the specimen
figured by Leidy. The thickness of the nuchal at the midline is 5 mm.

The first neural has a length of 13 mm., and a greatest width of 6 mm. The
bone is coffin-shaped with the widest end forward. The second neural is 9 mm.
long, and only 4 mm. wide.

All of the bones of the carapace are delicately sculptured, though those of the
anterior part of the shell appear less distinct than in most of the described specimens.
The few costals present show the usual low undulating ridges crossing them at
right angles to their shorter diameters. This sculpture is most distinct toward
their outer ends. The peripherals have their upper and lower surfaces ornamented
by the usual sharp ridges and pustular elevations.

GILMORE: the fossil turtles of the UINTA FORMATION 123

The few sulci discernible are narrow and delicately impressed. As ill pre-
viously described specimens the intramarginal sulci on the nuchal and anterior
peripherals cannot be traced. The first vertebral has a greatest width of 26 mm.,
whereas in the specimen described by Hay it is only 18 mm. The sulcus forming
the posterior boundary of the first vertebral crosses the first neural as in other
described specimens. The costal sulcus on the second costal is near the center of
that bone, while in the specimen described by Hay it is very close to the posterior
border.

The right hypoplastron is 21 mm. long on the midline, and has a thickness of
5.5 mm. The sculpture on the lower surface of this bone is made up of fine ridges
arranged in a radiating pattern. There is no evidence of epidermal scutes on any
of the plastral bones found with this specimen.

Cope has recognized this species from the Upper Green River beds, so that the
evidence at hand shows that this species ranges from the lowest horizon in the
Bridger deposits to the highest horizon in the Uinta formation, the uppermost
Eocene.

Family EMYDID.E Gray.

Genus Echmatemys Hay.

In 1908 Hay^ recognized nineteen species as pertaining to the genus Ech-
matemys. Since that time he has described one new form,'' so that with the four
new species described in the present paper, twenty-four species have been recog-
nized from the fossiliferous deposits of North America. Seven of these have now
been found in the Uinta formation and increased collections will doubtless show the
presence of several more. The discovery in the present collection of Echmatemys
septaria (Cope) leads to the belief that still other species known in the older
Wasatch and Bridger beds, will sooner or later be found to continue into the upper-
most Eocene.

6. Echmatemys callopyge Hay.

Plate XXI; text-figs. 9 and 10.

Echmatemys callopyge Hay, Fossil Turtles of North America, 1908, 340-342, PL

LII, figs. 1, 2; text-figs. 447, 448.

Two specimens in the Carnegie Museum are identified as belonging to this
species. The better preserved specimen. No. 2371, was collected by Earl Douglass
in 1908, from Horizon B, "above second sandstone with small artiodactyls,"

* " Fossil Turtles of North America," 1908, p. 298.

« Proc. U. S. National Museum, XXXV, 1908, pp. 164-166.

124

MEMOIRS OF THE CARNEGIE MUSEUM

Uinta formation, Upper Eocene, east of Dragon- Vernal road between White and
Green rivers, Uinta Basin, Uinta County, Utah. The second specimen, No. 2157,
was also collected by Douglass from the same geological horizon near Well No. 2,
in the Uinta Basin. Like the type, both of these specimens have the carapace
somewhat crushed over toward the left side. The type of the species is said by
Hay to have come from the middle Uinta, and it appears probable that all of
these specimens were found at about the same geological level.

Hay considered the very narrow first vertebral as the chief distinguishing
character for separating this species from the others of the genus, but both of the

Fig. 9. Echmatemys callopyge Hay. Carapace of C. M. No. 2371. One-fourth natural size. c. S,
costal plate; c.s. 1, first costal scute; n. 1, and n. 8, neurals one and eight; n.p., nuchal plate; s.p., suprapygal.

specimens before me have this scute relatively wider than in the type, although in
nearly all other respects, as is shown by the table of comparative measurements,
the specimens are remarkably similar. So far as the width of the first vertebral is
concerned these specimens are intermediate between the type of the present species
and the figured specimen of Echmatemys septaria (Cope), as illustrated by Hay,
Fossil Turtles of North America, Fig. 415, p. 320. The type of the latter species
is in the U. S. National Museum (No. 4088), and consists of a fairly complete

GILMORE: the fossil turtles of the UINTA FORMATION

125

plastron, a portion of the central part of the carapace, including the third, fourth,
and fifth neurals with portions of the second, third, and fourth costals: and a small
piece of the fifth, sixth, and seventh costals with abutting peripherals. It was
collected in the badlands of South Bitter Creek, Wyoming, from the beds of the
Washakie Basin.

I have carefully compared the specimens before me with the above mentioned
type and except for differences in size, find them, so far as they can be contrasted,
remarkably similar. The broad, hatchet-shaped anterior lobe so characteristic of
Echmatemys septaria is duplicated in these specimens.

In the type of Echmatemys callopyge the front two-thirds of the first vertebral
lies wholly within the lateral borders of the nuchal plate, and, although relatively
wider, this is also true of specimen C. M. No. 2157, but specimen C. M. No. 2371
has the antero-lateral angles of the first vertebral extending across the lateral
sutures of the nuchal. In a specimen identified by Hay as pertaining to Ech-
matemys septaria (See Fossil Turtles of North America, Fig. 415, p. 320) the first
vertebral extends entirely over the lateral boundaries of this plate. From the
intermediate condition observed in the present specimens, the first vertebral
would appear to be subject to considerable variation and therefore its narrowness
cannot be rehed upon as a constant specific difference. Specimen C. M. No. 2371

Fig. 10. Echynatemys callopyge Hay. Plastron of C. M. No. 2371. One-fourth natural size.

126

MEMOIRS OF THE CARNEGIE MUSEUM

has the surface of the carapace smooth, with the exception that, as in the type of
E. septaria, it is relieved by faint striations and growth lines, these being especially
apparent within the areas of the vertebral scutes.

Meastjhements of Neurals.

No.

Length.

Width.

Type..

No. 2157.

No. 2371.

Type.

No. 2157.

No. 2371.

1

52

50

50

33

35

36

2

43

—

40

39

—

41

3

58

—

43

43

—

39

4

41

41

38

37

38

34

5

41

26

35

44

—

41

6

30

—

29

40

—

39

7

29

25

25

45

—

40

8

28

—

23

34

—

35

Measurements of Vertebrals.

No.

Length.

Width.

Type.

No. 2157.

No. 2.371.

Type.

No. 2157.

No. 2371.

1
2
3
4
5

75
89
88
96

74

81

90 ±
96

77
85
87
86

52

85

82

87

100 ±

71

87

95

100

77
86
76
84

Principal Measurements Carapace and Plastron.

Greatest length of carapace

Greatest wdth of carapace

Greatest height of carapace

Nuchal, greatest length

Nuchal, greatest width

First marginal, greatest length. .
First marginal, greatest height. . ,

Plastron, greatest length

Anterior lobe, greatest length. . . .

Anterior lobe, width at base

Posterior lobe, greatest length. . .
Posterior lobe, width at base . . . .

Bridge, width

Lip, width

Entoplastron, width

Gulars, meet on the midline

Humerals, meet on the midline . .
Pectorals, meet on the midline . .
Abdominals, meet on the midline
Femorals, meet on the midline . .
Anals, meet on the midline

Type.

438

270

153

68

90

40

32

410

116

174

135

198

155

38

52

72

35

74

110

38

No. 2157.

420 d

276

160
69
86
43
34

422

125

180

140

205

165
44
66
67
34
76

114
41
62

No. 2371.

415

279

133

66

80

40e

28

370

107

176

195
162
45
75
58
36
73
102
33

Had E. callopyge not been established, I should have unhesitatingly referred
both of the specimens discussed above to Echynatemys septaria (Cope). For the

GILMORE: the fossil turtles of the UINTA FORMATION 127

present, however, it will serve all purposes to assign them to the established Uinta
species, until the discovery of additional Bridger material shall definitely determine
whether two distinct species are represented by this material, or whether E. cal-
lopyge Hay shall become a synonym of the earlier described E. septaria (Cope) .

In order to place on record the variation within the species, some of the prin-
cipal measurements of the two specimens hei'e considered as compared with those
of the type of the species are given in the preceding table.

7. Echmatemys uintensis Hay.

Echmaternys uintensis Hay, Fossil Turtles of North America, 1908, pp. 342, 343.
PL LHI, figs. 1, 2.

The above species is represented in the Carnegie Museum collections by three
specimens. The better preserved specimen, C. M. No. 3270, consists of a carapace
and plastron, the former lacking a portion of the posterior end and a considerable
part of the costals and peripherals of the right side. The plastron is complete.
This specimen was collected by Earl Douglass, May 25, 1908, from Red Bluff Wash,
on the road from Bonanza to Kennedy's Hole, Uinta Basin, Utah, from Horizon B,
"transition beds. First sandstone above red layer," Uinta formation. Upper
Eocene.

The second specimen. No. 2158, consists of a carapace lacking most of the
costals and peripherals of the left side, was collected by Earl Douglass, August 22,
1908, two or three miles below Well No. 2, from Horizon B, Uinta formation, as
exposed in the Uinta Basin, Utah.

The third specimen, No. 2397, consists of considerable portions of the carapace
and plastron of a large individual, both of which are rather fragmentary. This
specimen was collected by Messrs. Earl Douglass and J. F. Goetschius, August 17,
1908, from Horizon B, "grey beds below red and grey beds," Badlands south of
Kennedy's Hole, Uinta County, Utah.

This species is based upon a beautifully preserved specimen. No. 11,198, in
the paleontological collection of Princeton University. It was collected in 1891 from
the middle Uinta, on White River, Utah, and until the discovery of the present
specimens was the only known representative of the species.

The specimens before me add but little to our knowledge of the species, but
I believe it important to give at this time their principal dimensions as compared
with the type in order to show the variations within the species.

128

MEMOIRS OF THE CARNEGIE MUSEUM

Comparative Measurements of Neurals.

No.

Greatest length.

Greatest width.

Type.

No. 3270. No. 21.58.

No. 2397.

Type.

No. 3270.

No. 2158.

No. 2397.

1

60

59 57

.

45

43

47

_

2

43

55 , 45

—

42

48

56

— .

3

50

55 61

69

42

44

54

55

4

50

47 48

59

43

44

45

52

5

34

— 45

53

47

41

48

60

6

30 .

—

. —

38

42

—

45

59

7

20

—

—

28

50

—

—

64

8

25

—

—

37

30

—

—

—

Comparative Measurements of Vertebrals.

No.

Greatest length.

Greatest width.

Type.

No. 3270. No. 2158.

No. 2397.

Type.

No. 3270.

No. 2158.

No. 2.'!97.

1

70

88 i 81

112

82

100

2

115

111 : 99

120

102

77

88

114

3

91

102 100

117

81

80

80

94

4

78

110

94

— .

98

120

5

86

.

132

—

Comparative Measurements of the Pil.4.stron.

Greatest length, pla.stron

Greatest width, plastron

Length, anterior lobe

Width, anterior lobe at base ....

Width of Hp

Length of entoplastron

Greatest width of entoplastron . .

Length of posterior lobe

Width of posterior lobe at base . .
Depth of posterior median notch .

Width of notch

Epiplastrals meet on midline ....
Hyoplastrals meet on midline. . .
Hypoplastrals meet on midline . .
Xiphiplastrals meet on midline . .

Gulars meet on midline

numerals meet on midline

Pectorals meet on midline

Abdominals meet on midline ....

Femorals meet on midline

Anals meet on midhne

Type.

460

295

120

200

65

70

95

150

200

16

55

44

125

120

80

57

40

73

115

70

52

No. 3270.

480

390e

135

240

90

90

96

154

230

19

55

43

108

136

95

68

47

90

110

82

65

No. 2397.

515e

185
240
20e

114

160e

100

e, estimated.

8. Echmatemys douglassi sp. nov.

Plate XXII; text-figs. 11 and 12.

Type: C. M. No. 3244, consisting of a somewhat damaged carapace with a
complete plastron. The carapace lacks portions of the peripheral borders of the
front and sides in addition to several smaU areas out of the costal and neural
regions. Collected by Earl Douglass, May 25, 1908.

GILMOKE: the fossil turtles of the UINTA FORMATION

129

Locality : South Branch of Red Bkiff Wash, above the well on the road between
Bonanza and Kennedj^'s Hole, Uinta Basin, Uinta County, Utah.

Horizon: Lower portion Horizon B, "Transition Beds" (Peterson), ''in
sandstone same as No. 28,"^ Uinta formation. Upper Eocene.

The carapace, although crushed over toward the left side, shows the shell to be
elongated with the median portion high and vaulted. The surface of the shell is
smooth. The peripherals behind the inguinal notches are moderately thin with
acute edges and with a tendency to flare upward. The sulci are narrow, but

Fig. 11. Echmateim/s douglassi. Carapace of the tj-pe, C. M. No. 3244. One-fourth natural size.
c.s. 1, c.s. 8, costals one and eight; n.2,n. 8, neurals two and eight; sp., suprapygal; sp. 2, second suprapygal.

deeply impressed. The total length of the carapace in a straight line is about 470
mm. Its width is 300 mm., its height at the center is about 186 mm.

The nuchal scute is wedge-shaped with the narrow end forward. Some of
the anterior margin of this bone is missing so its length cannot be given. The
posterior end has a greatest width of 22 mm.

' No. 28 is C. M. catalog No. 3270 and is identified as Echmatemys uintensis Hay.

130

MEMOIRS OF THE CARNEGIE MUSEUM

The first neural is represented by the posterior end only, the second to the
sixth inclusive are complete, the seventh and eight are onh^ partially preserved.
In general the neurals are hexagonal with their broadest ends forward. The an-
terior ends of the second to the fifth are concave. There is no indication of a
carina on any of the neurals. Their principal dimensions are given in the table
below:

Length. Width.

40 45

45 42

48 35

47 41

32 44

24 44e

29 —

No.
2

3.
4.
5.
6.

7.

Fig. 12. Echmatemys douglassi. Plastron of type, C. M. No. 3244. One-fourth natural size.

There are as usual eight costals. These vary but little in the width of their
proximal and distal ends. The fifth on the right side has a length of 144 mm.;
the first at the suture with the second a length of 123 mm.

GILMORE: the fossil turtles of the UINTA FORMATION

131

The peripherals are high. On the right side immediately posterior to the in-
guinal notch they extend upward 74 mm. above the margin of the shell; the most
posterior one 45 mm.; the most anterior one 56 mm. above the margin.

The suprapygal is 41 mm. long. The second suprapygal is 52 mm. long, and
93 mm. across the middle, and 42 mm. where it joins the pygal. The pygal has a
length of 32 mm., a width of 55 mm. The posterior boundary of the fifth vertebral
crosses the second suprapygal 20 mm. anterior to its posterior margin. Both the
pygal and second suprapygal are strongly arched above the tail.

The vertebral scutes are long. The sides of the median three are slightly
bracket-shaped, the fourth, as in Echmatemys uintensis Hay, is strongly urn-shaped.
The dimensions of the vertebrals are given in the accompanying table.

Dimensions of Vertebrals.

No.

Length.

Width in Front.

Greatest Width.

1

80e .

97

97

2

93

—

77e

3

98

53

69

4

90

66

92

5

63

—

117

e, estimated.

There appear to be twelve marginals in the complete series of one side, and
some rise to the proximal ends of the peripherals. On the right side the eighth
peripheral is crossed by the marginal sulcus 21 mm. below the costo-peripheral
suture. The most posterior marginal scute is 53 mm. high from the margin of the
shell.

The plastron is perfectly preserved and characters observed in the plastron
show the distinctness of Echmatemys douglassi from all other described forms. The
plastron has a maximum length of 407 mm. At the center it measures 374 mm. in
length. The front of the anterior lobe is broadly and deeply emarginated, the
emargination lying between two toothed projections which extend forward at either
side, as is well shown in Fig. 12. On account of this emargination there is no well-
defined lip. The anterior lobe is broad and at its base measures 190 mm. with a
length at the center of 95 mm. The length therefore is exactly 50 per cent, of the
width. The lateral borders of the lobe in front of the axillary notches are slightly
concave, then expanding a little to the posterior ends of the epiplastra, then turning
inward to the toothed projections at either side of the lip. The latter is wide,
measuring 94 mm., which is nearly one-half of the total width of the lobe.

The posterior lobe has a greatest length at the center of 130 mm. ; a width at
the base of about 199 mm. The hinder lobe is considerably constricted at the anal-

132 MEMOIRS OF THE CARNEGIE MUSEUM

femoral sulcus, and from this point posteriorly the lateral borders converge quite
rapidly to the posterior end, which is deeply notched. The great breadth of the
lobes in E. douglassi leaves but little space for the exit of the limbs, and in front the
exit is still further closed by the upward curve of the anterior lobe as in E. callopyge
Hay. Transversely the plastron is angularly concave, but how much of this de-
pression may be attributed to postmortem causes it is impossible to determine. It
may be largely sexual, and in that case this specimen would represent a male.

The entoplastron is pear-shaped and extends forward to within 11 mm. of
the anterior margin of the lobe, a most unusual position in the Emj^didae. The
length of the entoplastron is 79 mm.; its width 71 mm. The bridge has a width of
167 mm. In this specimen, as shown in Fig. 12, there are intergular scutes, which
overlap the entoplastron. On the left side the gular-humeral sulcus follows the
usual course, reaching the margin of the lobe immediately posterior to the toothed
projection marking the external boundary of the lip. On the right side, however,
there is no trace of this sulcus. Intergulars are not known in any other member of
the Emydidae, and it may be that the scutes here designated as intergulars are the
gulars, and that the extra scute on the left side is supernumerary. Even should
that be the case, the position of the sulcus crossing the border on the mesiad side
of the toothed projection is unusual, and probably constitutes an individual vari-
ation, the true condition of which can only be cleared up by the discovery of ad-
ditional specimens pertaining to this species.

The pectorals do not reach the entop'astron, but at the center pass 6 mm.
behind it. They have their greatest width of 60 mm. at the center. The humerals
meet along the midline for a distance of 65 mm. ; the abdominals for 98 mm. ; the
femorals for 68 mm. ; and the anals for 55 mm.

The hyoplastrals are 85 mm. wide antero-posteriorly at the midline, the left
being slightly more. Each extends laterally about 132 mm. The hypoplastrals
meet on the midline for a distance of 122 mm.; the xiphiplastrals for a distance of
83 mm. The notch between their hinder ends is 20 mm. deep, with a greatest
width of 50 mm.

In the general shape and contour of the shell this species closely resembles
Echmatemys stevensoniana (Leidy), from the Bridger beds of Wyoming. It differs
from that species, however, in the deep emargination of the anterior lobe and the
failure of the pectoral scutes to reach the entoplastron.

Echmatemys douglassi is distinguished from all described species of the genus
by the deep emargination of the anterior median border of the anterior lobe; by the
short and wide anterior lobe, the length of which at the center is only fifty per cent.

GILMORE: the fossil turtles of the UINTA FORMATION

133

of the width at the base, and by the close proximity of the anterior end of the ento-
plastron to the border of the Up. In the unusual proportions of the anterior lobe
it most nearly resembles E. arethusa Hay from the Bridger beds, but is at once dis-
tinguished from it by the concave lip, as contrasted with the projecting lip of the
former species. In having the humero-pectoral sulcus pass behind the entoplastron
this species is distinguished from all other species of the genus with the exception
of E. lativertehralis (Cope), E. megaulax (Cope), and E. rivalis Hay.

This species is dedicated to Mr. Earl Douglass, who collected the type speci-
men, as well as the greater number of specimens comprised in this collection of

turtles.

9. Echmatemys hollandi sp. nov.

Plate XXIII, fig. 1; text-fig. 13.

Type: C. M. No. 3249, consisting of a considerable portion of the carapace,
lacking the posterior and the greater part of the peripherals and costals of the
left side and the outer halves of most of the remaining peripheral and marginal

Fig. 13. Echmatemys hollandi. Carapace of the type, C. M. No. 3249. One-third natural size,
c. 6, sixth costal; c.s. 1, supernumerary or first costal scute; n. 1, n. 6, first and sixth neurals.

bones. A considerable part of the plastron is present but the under surface is so
badly shattered that nearh^ all traces of the sutures and sulci have been obliterated.
The impression remaining in the matrix, however, gives some idea of the shape and
dimensions of the anterior and posterior lobes. Collected by Earl Douglass, 1908.

134

MEMOIRS OF THE CARNEGIE MUSEUM

Locality: Skull Butte, southwest of Well No. 2, Uinta Basin, Uinta County,
Utah.

Horizon: Horizon B, Uinta formation, Upper Eocene.

The carapace is elongated oval being more broadly rounded in front than
Eclmiatemys callopyge Hay. The median portion is high and vaulted. The
length of the carapace is estimated to have been about 360 mm. ; its width about
260 mm. The height at the center is 133 mm. The surface of the shell is smooth.
The sulci are narrow, but well impressed.

The nuchal plate has the anterior border missing, so that its greatest length
cannot be determined. It has a greatest width of 82 mm., and where the lateral
sutures cross the costo-marginal sulcus it is 57 mm. wide.

The neurals back of the first are all hexagonal with the anterior ends concave.
Those preserved are all longer than broad. On either side of the anterior end of
the first neural are hollowed out depressions which give this bone the appearance of
being bluntly ridged anteriorly, but otherwise there is no indication of carinse.

Dimensions of Neurals.

No.

Dimensions of Vertebral Scutes.

No.

Length.

Width.

Length.

Width in Front.

Greatest Width.

1
2
3
4
5
6

45
40

41
38
38

30
36
35
35
36
36

1
2
3
4

65
80

87

46
50
40
49

55

77
72
73

The peripherals appear to have been high, but on account of the damaged
condition of the borders it is not possible to give their extent.

Six costals are present in this specimen, the first having a greatest antero-
posterior diameter of 63 mm.; a greatest length of 96 mm. On the upper anterior
half of the first costal there is a low, rounded, obtuse elevation or horn-like pro-
jection, which at once distinguishes this species from all other described forms. The
other costals show nothing unusual. The third has a greatest length of 106 mm.
The costo-peripheral suture between the third and fourth passes about 17 mm.
mesiad of the costo-marginal sulcus.

The first vertebral scute is unusually narrow, in this respect resembling
E. callopyge Hay. The sides of those posterior to the first are strongly urn-shaped.
The anterior end of the third is especially narrow and pointed, and extends well
forward into the second. The dimensions of the vertebrals are given in the table.
On either side of the first vertebral are supernumerary costal scutes, and it is
largely within their boundaries that the horn-hke elevations, described above, arise.

GILMORE: the fossil turtles op the UINTA FORMATION 135

The area of the first costal scute is much reduced, but it is still in contact with the
first vertebral. Supernumerary costals are not unusual in the Baenida?, though
I am not aware of their having been found before in the genus Echmatemys. When
present, they are usually confined to one side, seldom are they symmetrically paired
as in the present specimen.

The plastron is exceedingly thick and heavy. At the center it measures 30
mm. in thickness. Though much of the anterior and posterior lobes are missing
the impressions remaining in the matrix show the plastron to have had a greatest
length of about 315 mm.

The entoplastron though only partially preserved, has a greatest width of 50
mm. Its length cannot be determined.

The bridge has a width of about 145 mm.

There is no suggestion of a notch in the posterior lobe, shown by the im-
pression in the matrix, but it is not possible to state positively that such did not
exist. Judging from the impression left by the anterior lobe the lip was thick and
broad, with an abrupt depression on the dorsal surface some 35 mm. posterior to
the anterior border. The anterior end of this lobe was probably within the forward
end of the carapace.

The pair of horn-like protuberances on the front of the carapace, the presence
of a pair of supernumerary costal scutes on either side of the first vertebral serve to
distinguish this specimen from all other described species of the genus, and I
therefore take great pleasure in naming it Echmatemys hollandi for Dr. William J.
Holland, Director of the Carnegie Museum, in recognition of his activities in the
field of vertebrate paleontology.

10. Echmatemys obscura sp. nov.
Plate XXIV; text-figs. 14 and 15.

Type: C. M. No. 3252, consisting of a carapace, lacking the posterior end
back of the sixth neural; and the plastron, lacking the lip and a small portion of
the extremity of the posterior lobe. Collected by Earl Douglass, August 17, 1908.

Locality : Devil's Play Ground, south of Kennedy's Hole, Uinta Basin, Uinta
County, Utah.

Horizon: Horizon C, "gray beds below red and gray beds," Uinta formation,
Upper Eocene.

Except for the parts missing from the posterior end of the carapace, the type
specimen is well preserved and all of the sutures and sulci are clearly displayed.
The carapace is broadly rounded in front with a wide, but shallow, emargination of

136 MEMOIRS OF THE CARNEGIE MUSEUM

the nuchal border. The median portion is high and vaulted. The length of the
carapace is estimated to have been about 390 mm., its breadth is 286 mm., its height
at the center is 130 mm. The plastron is broad, and, like many other species of the
genus Echmatemys, there is but little space in front of the axillary notches for the
exit of the limbs. The margins of the carapace forward of the axillary notches
are thickened and rounded, but forward it thins rapidly, coming to an obtuse edge
along the median anterior border. The peripherals of this region do not flare up-
ward, though those immediately posterior to the inguinal notches show a tendency
to do so.

The carapace is smooth, except that the median costal areas are crossed antero-
posteriorly by a series of wide, parallel, wavy, flattened ridges. In a line 12 mm.
long three of these ridges may be counted. This ornamentation is inconspicuous
unless the light strikes the surfaces at the proper angle. It is the obscure nature of
this ornamentation which has suggested the specific name.

The nuchal plate is unusually long and narrow. At the point of its greatest
transverse diameter the sides are considerably within the boundaries of the first
vertebral. The length of the nuchal is 65 mm.; its greatest transverse diameter
is 51 mm. ; the free border measures 40 mm. The nuchal surface is without median
elevation. Echmatemys obscura is the only species of the genus which has the
nuchal plate longer than wide.

All of the neurals posterior to the first are broadly hexagonal, and all posterior
to the first are broader than long. The second and third have the anterior end
concave, those posterior being straight. Their principal measurements are given
in the accompanying table.

Dimensions of Neueals.
No. Length. Width.

1 47 35

2 38 40

3 43 43e

4 39 47

5 31 41

6 — 44

e, estimated.

The costals posterior to the first alternate in having the distal ends slightlj^
wider and narrower than the proximal ends.

The peripherals are moderately high, the first extending inward from the
border 50 mm., the seventh 58 mm.; the eighth 55 mm. The border above the
bridge is heavy and rounded, but posterior to the inguinal and anterior to the

GILMORE: the fossil turtles of the UINTA FORMATION

137

axillary notches the borders thin rapidly toward the center. The costoperipheral
sutures pass along the median sides of the shell on an average of about 17 mm.
mesiad of the marginal sulcus. Beyond the bridges the sutures and sulci in some

Fig. 14. Echmatemys ohscura. Carapace of type, C. M. No. 3252. One-fourth natural size. n. 1,
71. 6, neurals one and six; n.p., nuchal plate; I'.s. 4, fourth vertebral scute; c. 6, sixth costal.

places approach one another as close as 4 mm., and in other places are distant as
much as 30 mm.

The nuchal scute is narrow, measuring 8 mm. on the free border. It is 15 mm.
long antero-posteriorly.

The vertebral scutes are wider than long, the second being especially wide.
The sides of those back of the first are bracket-shaped. Their dimensions are
given in the accompanying table.

Dimensions or Vertebrals.

No.

Length.

Width in Front.

Greatest Width.

1

75

87

87

2

83

58

102

3

79

60

81

4

—

58

—

The costo-marginal sulci run below the costo-peripheral sutures. The sulci
on both carapace and plastron are narrow and moderately impressed.

The plastron in life had a length of about 372 mm. The posterior end is

138

MEMOIRS OF THE CARNEGIE MUSEUM

narrow, but deeply notched. The length of the anterior lobe is about 100 mm.;
its width at the base is about 162 mm. The length is therefore only 61 per cent.
of the width. From the axillary notch the free border runs straight forward for a
short distance then curves in regularly to the epiplastral lip, which is missing in

Fig. 15. Echmatemys obscura. Tjr[)e, C. M. No. 3252. Plastron. One-fourth natural size.

this specimen. This free border is relatively thin and acute, thickening somewhat
as it approaches the lip. The lip appears to have been about 50 mm. wide.

The entoplastron has a length of about 60 mm. and a width of 55 mm. It is
crossed by the humero-pectoral sulcus, and is also overlapped by the gular scutes.
It is pear-shaped in outline, in this respect resembling the entoplastron of Ech-
malemys douglassi. The bridge is 150 mm. in width.

The length of the posterior lobe is 120 mm. Its width at the base is 191 mm.
The length is therefore 62 per cent, of the width. The free borders are slightly
contracted at the femoral-abdominal sulcus, and again at the femoral-anal sulcus;
from which point the border turns in rapidly toward the center. The posterior
lobe is terminated posteriorly by rather sharp projecting points on either side of
the narrow, but rather deep, median notch. It is estimated that the notch had a
depth of 30 mm. The posterior borders of the lobe are acute. At the center of
the notch the bone has a thickness of 5 mm.

The anterior lobe curves upward toward the carapace much as in Echmatemys

GILMORE: the fossil turtles of the UINTA FORMATION 139

septaria. The plastron is flat, but in an uncrushed specimen the bridges would
doubtless curve upward to the margin of the shell. In this specimen they are but
little above the level of the plastron.

The hyoplastrals meet on the midline for a distance of 80 mm.; the hypoplas-
trals for 97 mm.; the xiphiplastrals for 66 mm.

The gular scutes along the midline are about 52 mm. long; the humerals 37
mm.; the pectorals 63 mm.; the abdominals 101 mm.; the femorals 48 mm.; and
the anals 53 mm.

Echmatemys ohscura is distinguished from all species of the genus, in which
the nuchal region is known, by the extreme narrowness of the nuchal plate, it
being the only species known, in which the nuchal plate is longer than wide.
From Echmatemys cibollensis, E. megaulax, and E. euthneta (the nuchal region of all
three being unknown) the present species is distinguished: from the former by
having the gulars overlapping the entoplastron ; and from the latter two by having
the gular-humeral sulcus crossing the rear portion of the entoplastron. This
species is further distinguished by the greater relative widths of the neurals and
especially the vertebrals. The obscure, but characteristic, ornamentation of the
costal region of the carapace will also aid in recognizing this species.

11. Echmatemys depressa sp. nov.
Plate XXIII, fig. 2; text-fig. 16.

Type: C. M. No 2936, consisting of the carapace, lacking much of the anterior
margin, the peripherals, and outer halves of the costals of the left side, and most
of the peripherals posterior to the inguinal notch of the right side. The plastron
is represented by a few fragments only, though the impression in the matrix gives
some idea of its proportions. Collected by 0. A. Peterson, August 5, 1912.

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

Horizon: Horizon C, Uinta formation. Upper Eocene.

Although the open sutures of the type specimen give evidence of the im-
maturity of the individual, it appears to represent one of the smaller species of the
genus Echmatemys. I was first inclined to regard it as referable to the genus
Palceotheca on account of its small size and the presence of a dorsal keel, but a
comparison with the types of the two species pertaining to that genus {Palceotheca
terrestris Cope, and P. polycypha Cope) both of which are in the U. S. National
Museum, shows differences which lead me to believe that it can with greater
propriety be referred to the genus Echmatemys. The apparent absence of a second
suprapygal and the extremely wide vertebrals may with other characters to be
observed in a more perfect specimen show its distinctness from that genus.

140 MEMOIRS OF THE CARNEGIE MUSEUM

In a straight line the shell has a greatest estimated length of 135 mm. ; at the
center a greatest width of 115 mm. Though depressed, the upper shell is broadly
convex in all directions, dropping off rather more rapidly toward the back than
toward the front. The upper surface of the carapace is smooth, the sulci lightly
impressed, and nowhere are scutal growth lines to be observed.

Fig. 16. Carapace of Echmatemys depressa. Type. C. M. No. 2936. One-half natural size, c.p.',
c.p. S, costal plates one and eight; nS, n7, neurals two and seven; nu.p., portion of nuchal plate; pij, pygal;
spy, suprapygal; v.s. 2, second vertebral scute.

The nuchal is onlj^ partially preserved, but this portion shows that it had a
greatest width of 26 mm. and was obtusely keeled at the center. The anterior
border, as shown in Fig. 16, is missing.

There are eight neurals, all of which are hexagonal, with the exception of the
first and eighth, the latter being subrectangular in outline. All have their antero-
lateral angles truncated, which serves at once to distinguish this form from Ech-
matemys pusilla Hay, which has the postero-lateral angles of the neurals truncated.
The neurals gradually decrease in size from front to back, and, excepting the first
and third, all are broader than long, as shown in the accompanying table. The
second and third neurals are sharply keeled on their posterior and anterior ends
respectively, as are the fourth and fifth, while the sixth, seventh and eighth are
keeled their entire lengths. The keel on the suprapygal is very low and hardly
discernible.

All of the costals of the right side are present and perfectly preserved. They
are of moderate thickness with pointed distal ends which articulated with the periph-
erals by gomphosis. Portions of the buttresses preserved in the matrix indicate
that they articulate with the costals considerably above the costo-peripheral suture.

GILMORE: the fossil turtles of the UINTA FORMATION 141

Dimensions of Neurals.

No. . Length. Width.
1 17 12

2 13 16

3 14.5 14

4 13.5 14

5 13.5 14

6 11 14

7 10 .....12

8 9 10

The peripherals of the right side above the bridge and somewhat forward of
the axillary notch are perfectly preserved, and, as in Echmatemys pusilla Hay,
have a sharp carina beginning on the third peripheral and continuing backward
across the bridge to the hinder peripherals. The fifth peripheral has a width of
19 mm.; the sixth of 18 mm.; the seventh of 20 mm.; the eighth of 18 mm. The
lateral peripherals from the edge of the carapace to their proximal extremities have
a length of 14 mm., becoming narrower toward the front of the shell.

The sulci on most parts of the carapace are very obscure, being traceable
only here and there, though where they cross the neurals somewhat plainer than
elsewhere. The boundaries of the second vertebral can be partially determined,
and these indicate a very wide scute having at the center angularly pointed outer
borders. The second vertebral has a greatest width at the center of 58 mm.; an
estimated length of about 32 mm. There were four costal scutes. The costal-
marginal sulcus appears to have followed closely the course of the costo-peripheral
suture. The supracaudal scute is divided. The second suprapygal is absent in
this specimen.

The impression in the matrix shows the hypoplastron to have a greatest width
at the midline of 41 mm. The width of the posterior lobe at the base is about
60 mm. Its greatest length was about 47 mm. It cannot be determined whether
this lobe was notched on the midline. At the center the plastron has a greatest
width of 84 mm. The bridge has a width of about 58 mm. The inguinal buttresses
rise well above the costo-peripheral sutures and articulate with both the fifth and
sixth costals.

Echmatemys depressa is distinguished from all other species of the genus by
the greater relative widths of the vertebral scutes, the absence of a second supra-
pygal, and the presence of a dorsal keel. From Echmatemys megaulax (Cope),
which also has a dorsal keel the present species is to be distinguished by having
the sulci less deeply impressed and in having the costo-marginal sulcus follow the

142

MEMOIRS OF THE CARNEGIE MUSEUM

course of the costo-peripheral suture whereas in the former it crosses the peripherals
on their upper third.

12. Echmatemys pusilla? Hay.
Echmatemys pusilla Hay, Fossil Turtles of North America, 1908, pp. 337-339,
text-figs. 445, 446.

A small turtle, C. M. No. 3282, collected by Messrs. Earl Douglass and
Clarence Wilson, November 1, 1911, southeast of Ouray, Uinta County, Utah,
from Horizon C of the Uinta formation, Upper Eocene, is referred with some doubt
to Echmatemys pusilla Hay. The very fragmentary nature of the present specimen
renders its generic and specific affinities difficult of positive determination, but after
a careful comparison of this specimen with the type of E. pusilla, kindly loaned me
by Dr. W. D. Matthew, of the American Museum of Natural History, I am con-
vinced of the very close relationships of the two specimens, even though the dis-
covery of more perfect material may eventually demonstrate their specific distinct-
ness.

Fig. 17. Portions of the carapace and plastron of Echmalemya -pimllal Hay. C. M. No. 3282. (1),
plastron; (2), carapace. c2, c3, costals two and four; n2, nS, neurals two and five; v2, vertebral scute two.
Natural size.

This specimen consists of the anterior lobe of the plastron lacking the lip, a
small portion of the carapace consisting of the second neural complete, and portions
of the third, fourth, and fifth neurals with the upper portions of abutting costals,
as shown in Fig. 17. The second neural is hexagonal in outline and measures 9.5
mm. in length, with a greatest width of 10 mm.; the third is 12 mm. long and 13
mm. wide; the fourth is 13 mm. long. The second vertebral has its greatest width
of about 35 mm. at the center. The scutal areas of the carapace are plainly grooved
by the lines of growth of the scutes. These lines of growth are also present in the
type of the species.

GILMORE: the fossil turtles of the UINTA FORMATION 143

The entoplastron, as in the type, is long, narrow, and pointed in front. Its
greatest length is 19 mm., its greatest width 16 mm. It is overlapped by both the
gulars and pectorals. The specimen before me also agrees with the type in the
great width of the pectorals behind the entoplastron. These scutes reach backward
to the hypoplastral suture, a condition not known in any other species of the genus.
At the point where the pectoro-humeral sulcus crosses the free borders the lobe
has a width of 48 mm. The free border of the lobe is thin and acute, being bevelled
off on the upper surface.

The most important dissimilarity between the two specimens here discussed
appears to be in the wide vertebrals and in the shape of the neurals, those of the
type having the postero-lateral angles truncated, whereas in the specimen from
the Uinta formation the antero-lateral angles of the third, fourth, and fifth are
thus cut off. The second neural is octagonal in the type, hexagonal in No. 3282.
In the shape of the anterior lobe of the plastron, the long pointed entoplastron
overlapped by the gulars and crossed well forward by the pectoro-humeral sulcus
and the extremely wide pectorals reaching backward nearly to the hyo-hypo-
plastral suture these specimens show a remarkably close resemblance.

Family TESTUDINID^ Gray.

The family Testudinidse is represented in the collection of chelonian remains
from the Uinta formation in the Carnegie Museum by the two genera Hadrianus
Cope and Testudo Linnseus. the former genus by three, the latter by but one species.
Four species of Hadrianus are now recognized as occurring in the Uinta formation.
The discovery of Testudo in the Upper Eocene is of interest as being the first time
this genus has been found below the Oligocene in North America. In the Fayum
deposits (Upper Eocene) of northern Africa, however, the genus Testudo has been
recognized by Andrews from well-preserved specimens, which in several respects
closely resemble the species here described.

Genus Hadrianus Cope.
13. Hadrianus corsoni (Leidy).
Plate XXV, fig. 1; text-fig. 18.

Testudo corsoni Leidy, Proc. Acad. Nat. Sci. Phila., 1871, p. 154; Contrib. Extinct
Vert. Fauna West. Terrs., 1873, pp. 132, 339, PI. XI, figs. 1, 2; PI. XV, fig. 7;
PL XXIX, figs. 2-4; PL XXX, figs. 1-4.

Hadrianus octonarius Cope, Palseont. Bull. No. 2, 1872; Vert. Tert. Form. West.,
1884, p. 140, PL XX, figs. 1-4.

144 MEMOIRS OF THE CARNEGIE MUSEUM

Hadrianus corsoni Cope, U. S. Geol. Surv. Terrs., 6th Ann. Rept., 1872 (1873),
p. 631; Hay, Bibliog. and Cat. Foss. Vert. N. A., 1902, p. 450; Fossil Turtles
of North America, 1908, pp. 376-380, PI. LX, LXI; text-figs. 473-479.

A large specimen in the present collection is identified as belonging to the
above genus and species, and represents the first recorded occurrence of Hadrianus
corsoni in the Uinta formation. This specimen consists of a complete plastron,
the peripherals of the right side above the bridge, and a few fragments of the cara-
pace. It was collected by 0. A. Peterson, August 5, 1912, six miles east of Myton,
Uinta County, Utah, from Horizon C of the Uinta formation, Upper Eocene. It
bears the C. M. Catalog No. 3403.

The length of the plastron at the center is 740 mm.; the greatest length over
all 775 mm.; the greatest breadth about 490 mm. The posterior lobe is deeply
and widely notched and the anterior lobe is terminated in front by a wide spade-
like lip which projects prominently from the general contour of the lobe.

Hadrianus corsoni was based upon the anterior portion of a plastron consisting
of the complete lip and lobe back to and including a small portion of the anterior
end of the entoplastron. A comparison of the specimen before me with the type
shows striking similarities in the contour of the lobe and lip, and especially of the
wide anterior end of the entoplastron, which appears peculiar to this species. The
lip has a transverse width at the base of 162 mm., which is greater than that of the
type, or of any subsequently discovered representative of this species. It projects
45 mm. beyond the point where the gular-humeral sulcus crosses the free border.
This measurement is slightly greater than in the type, but less than in the type of
Hadrianus octonarius Cope, now regarded by Hay as being a synonym of the
present species. The anterior border of the lip is subacute, nearly straight, but
slightly notched on the midline as in the type. The upper surface of the lip is
slightly convex along the midline, with shallow longitudinal depressions on either
side. The lower surface is flat, but I am inclined to believe it would have been
broadly convex in life.

The anterior lobe is 345 mm. wide at the base, with a length of 252 mm. Along
the free borders on the upper surface the bone is bevelled off to an acute edge that
becomes obtusely rounded in front of the axillary notches. The lip on the superior
surface extends backward 67 mm.

The entoplastron is unusually broad and angular in front. In the figures of
the type the suture Hmiting the anterior border of the entoplastron runs nearly-
straight across the median line, then turns abruptly backward and outward. In
the specimen here considered the anterior border is inclined more posteriorly, as

GILMORE: the fossil turtles of the UINTA FORMATION

145

in the type of H. odonarius, but it has the same angular turn backward and out-
ward as in the type of the genus and species. The length of the entoplastron on the

Fig. 18. Hadrianns corsoni (Leidy). Plastron of C. M. No. 3403. One-eighth natural size.

midline is 155 mm.; its width 184 mm. It is overlapped slightly by the gular
scutes, but is not crossed by the humero-pectoral sulcus.

The bridge is 290 mm. wide.

The length of the posterior lobe is 234 mm.; its width at the inguinal notches
386 mm. The free edges of the lobe are subacute, except in front of the notches,
where it is rounded. The width of the posterior notch is about 135 mm. ; its depth
about 35 mm. There is no perceptible ridge leading back from the inguinal but-
tresses until a point midway between the buttress and the posterior end of the lobe
is reached, where the surface is concave mesiad, thus causing the ridge to stand
out quite prominently. On the outside of this ridge the surface is bevelled off
steeply near the buttress, but less and less so posteriorly, until on the posterior
outer angle of the xiphiplastral projections the slope is very gentle. The surfaces
of these projections are rugosely roughened as in the type of H. odonarius (No.
2186, U. S. National Museum), which they resemble in shape and size.

The hyoplastron on the midline has a width of 195 mm. The arrangement of

146

MEMOIRS OF THE CARNEGIE MUSEUM

the plastral scutes is not greatly different from that of other Testudinidse. The
gulars have a length on the midline of 90 mm.; pectorals of 60 mm.; abdominals of
255 mm.; femorals of 95 mm.; anals of 86 mm. The principal dimensions and
proportions of specimen C. M. No. 3403, as compared with other specimens iden-
tified as pertaining to Hadrianus corsoni Leidy, are given in the table below.

C. M. No. 3403.

Greatest length of plastron

Greatest length anterior lobe

Greatest width anterior lobe

Width of bridge

Greatest length posterior lobe

Greatest width posterior lobe

Length of entoplastron

Width of entoplastron

Width of lip

Width of bridge to width of anterior lobe . .

Width of anterior lobe to its length

Width of bridge to length of posterior lobe

778
252
.345
290
234
386
155
184
162
81%
72%
80%

HuiManua cot-
soni A. M.N. H.

No. 6027.

550
175
246
215
157
240
115
134
118
87%
71%
73%

Type of H. octo-

narhis U. S.
N.M.No. 2186.

740
247
360
285
240
330
125
160
135
79%
68%
84%

14. Hadrianus robustus sp. no v.
Plate XXV, fig. 2; text-fig. 19.

Type : C. M. No. 3342, consisting of the anterior half of the plastron, collected
by Earl Douglass, July 30, 1908.

Locality: Near Kennedy's Hole, Uinta Basin, Uinta County, Utah.

Horizon: Horizon C, Uinta formation. Upper Eocene.

The specimen upon which the present species is based represents one of the
larger species of the genus. Those parts present are in a good state of preservation,
and all of the sutures and sulci can be clearly traced. It is assigned provisionally
to the genus Hadrianus, until such time as the discovery of a more perfect specimen
makes it possible to determine its true generic affinities.

The anterior lobe is 220 mm. long, and at the base 285 mm. wide. The length
thus being 77 per cent, of the width, whereas in Hadrianus corsoni it is only 71 and
in H. niajusculus 68 per cent. The lip in this species is especially prominent and
exceedingly heavy, having a thickness at the center of 46 mm. At the base it
measures 123 mm. in width, and near the anterior end 117 mm. From the point
where the gular-humeral sulcus crosses the free border the lip extends forward 52
mm. The anterior border of the lip is transversely broadly convex, and bluntly
bevelled dorso-ventrally, the longer bevel being on the lower side. On the superior
surface the lip extends posteriorly 100 mm., at this point the surface descends
perpendicularly, decreasing by one-half the total thickness of the plastron. Slightly

GILMORE: the fossil turtles of the UINTA FORMATION

147

posterior to this drop on the visceral surface tliere is a sharp median ridge on the
midline which measures longitudinally 65 mm. The free borders of the anterior
lobe are bevelled off from the superior surface to a sharp edge, which becomes
obtusely rounded at the base of the lip.

Fig. 19. Hadrianus robustus. Anterior half of the plastron. Tyjje, C. M. No. 3342. One-sixth na-
tural size.

The entoplastron is pointed in front and wide behind. It is 134 mm. long,
and 147 mm. wide. The gular scutes overlap the anterior end.

In front of the axillary notch the plastron has a thickness of 29 mm.; im-
mediately posterior to the entoplastron of 16 mm.; and at the junction of the
hyo-hypoplastron at the center, of 15 mm. This latter measurement in the type
of Hadrianus tumidus Hay is only 9 mm. The gular scutes have their greatest
extent of 114 mm. antero-posteriorly. Within the area of the gular scutes the
surface of the lip is swollen out somewhat below the level of the epiplastral areas
bordering it. The superior surface of the lip anteriorly is broadly convex, but
flattens toward the posterior end.

The humerals meet on the midline for 105 mm.; the pectorals for 40 mm.; the
abdominals for at least 205 mm. The sulci are broad and deeply impressed. The
humero-pectoral sulcus begins just in front of the axillary notch and extends inward
and backward for a short distance, then curves forward to the entoplastron, again
turning backward to the midline, skirting, but not crossing, the entoplastron.

The pectoral scutes have quite a different shape from any of the described
species of the genus. They are narrow (28 mm.) mesiad of the axillary notch

148 MEMOIRS OF THE CARNEGIE MUSEUM

gradually widening to a point a little beyond where the sulcus first reaches the
entoplastron, which measures 63 mm. antero-posteriorly, then again narrowing
to the midline, where it has a fore-and-aft extent of 40 mm. The pectoro-abdominal
sulcus runs straight across the plastron as in Hadrianus corsoni.

The hyo-hypoplastral suture runs a very tortuous course across the plastron.
The hyoplastrals meet on the midline for a distance of 136 mm. The hypoplastrals
at the center have a transverse width of 400 mm.

The pectoral scutes on the midline are less than one-fifth the width of the ab-
dominals, and following Hay's analysis of the various species of the genus this
proportion would be sufficient to show the specific distinctness of the present speci-
men. In the proportions of these scutes the present species is nearest Hadrianus
tumidus Hay, which is also from the Uinta formation, and in which the pectoral
scutes are less than one-third as wide as the abdominals at the center, but it is
distinguished from that species by having a thicker plastron and by the different
form of the pectoral scutes, and the greater width posteriorly of the entoplastron.
From H. majusculus and H. corsoni the difference in the proportions of the length
to the breadth of the anterior lobe will help to separate the present form; that
is to say in Hadrianus robustus the length is 77 per cent, of the width at the base,
while it is 68 and 71 per cent, respectively in the other two species mentioned above.

15. Hadrianus utahensis sp. nov.
Plate XXVI, fig. 1; text-fig. 20.

Type : C. M. No. 2343, consisting of the plastron and portions of the periph-
erals above the bridge on the right side. The anterior portion of the lip, and
parts of the margins of both anterior and posterior lobes are missing. Collected
by Earl Douglass, July 30, 1908.

Locality : South of Kennedy's Hole, Uinta Basin, Uinta County, Utah.

Horizon : Horizon B or C, Uinta formation. Upper Eocene.

The plastron of the type specimen is estimated to have had a greatest length
of about 520 mm., and a greatest width at the center of 320 mm. The anterior
lobe is about 175 mm. long and 300 mm. wide at the base. The width of the lip,
where the gular sulci cross the free border, is 110 mm. The plastron is quite
concave, indicating that the specimen was in all probability a male.

While every thing indicates that the hp extended well forward, it probably
continued the general contour of the lobe. The free borders of the lobe are sub-
acute, being bevelled off from the upper surface, so that the edge is nearly on a
level with the ventral surface of the plastron. At the center on the broken border

GILMORE: the fossil turtles of the UINTA FORMATION

149

the lip is 22 mm. thick, but posteriorly it increases to 27 mm., behind which the
plastron is deeply excavated. On the upper surface of the lip there is a slight
median elevation. Transversely the whole lip is broadly convex rounding down
at the sides to a subacute edge.

Fig. 20. Plastron of Hadrianm utahensis. Type. C. M. No. 2343. One-fourth natural size.

The entoplastron is rhombic in form, being 97 mm. long and 95 mm. wide.
The gular scutes reach, but do not overlap, this bone. In this respect this specimen'
differs from all of the described species of the genus. The bridge has a greatest
width of 210 ram.

The posterior lobe has a greatest length of about 140 mm.; a greatest width
of 230 mm. at the inguinal notches. The posterior lobe is shallowly notched on
the midline. This notch has a greatest depth of 10 mm. Immediately behind the
inguinal notches the borders of the lobe have a thickness of 27 mm. and are rounded
dorso-ventrally. The border at the anal sulcus is 18 mm. thick. The free border
of the entire lobe a short distance posterior to the inguinal notches is bevelled off

150 MEMOIRS OF THE CARNEGIE MUSEUM

from the dorsal surface to a subacute edge. There is no ridge leading back from
the buttresses as in H. tumidus Hay. In front of the notch on the dorsal surface
the plastron is transversely shallowly concave. The plastral buttresses are rela-
tively heavy.

The hyoplastrals meet on the midline for a distance of 108 mm.; hypoplastrala
for 120 mm.; xiphiplastrals for 107 mm.

The sulci on the plastron are relatively narrow but well impressed. The
gular sulci run forward and outward from the center, but, as they approach the
border, suddenly turn outward and then backward continuing in this direction
over the border upon the dorsal surface and fading out at the base of the lip. The
sulcus limiting the humerals behind runs backward and slightly inward from the
border in front of the axillary notch for a short distance, then turns inward and
forward to the posterior boundary of the entoplastron, but does not cross it. The
pectorals occupy 41 mm. of the midline, and have a least diameter antero-poste-
riorly of 29 mm. The abdominals meet on the midline for a distance of 155 mm.;
the femorals for 75 mm.; anals for 58 mm. The plastron at the center has a
thickness of only 7 mm.

The peripherals turn abruptly upwards at the sides of the shell. The sulci
between the plastron and peripheral bones on the bridge, although broadly inter-
digitativc, run quite a straight course antero-posteriorly. The fifth and sixth
peripherals have a width of 64 mm. Their length on account of the missing upper
extremities cannot be determined.

The present specimen is assigned provisionally to the genus Hadrianus,
though later it may be found, when more perfect specimens are available, that it
belongs to Testudo. It is distinguished from the described species of the genus
by the narrow and relatively shallow notch on the posterior lobe; the rounded and
thickened ends of the xiphiplastrals, and by having the bridge longer relative to
the length of the hinder lobe. In Hadrianus majusculus the posterior lobe is 85
per cent, of the length of the bridge; in H. tumidus 77 per cent.; in H. corsoni
73 per cent. ; and in the present specimen only 66 per cent.

Genus Testudo Linnaeus.

16. Testudo uintensis sp. nov.

Plate XXVII; text-figs. 21 and 22.

Type: C. M. No. 2331, consisting of a carapace and plastron, the latter almost
perfectly preserved, the former lacking the peripherals of the right side, and the

GILMORE: the fossil turtles of the UINTA FORMATION

151

anterior portion of the nuclial plate. Collected by Earl Douglass and party,
July 30, 1908.

Locality: South of Kennedy's Hole, and about one hundred rods west of
Dragon- Vernal road, Uinta Basin, Uinta County, Utah.

Horizon: Horizon B or C, Uinta formation. Upper Eocene.

In form the carapace of this tortoise is broad, of moderate height and strongly
arched in all directions. As now preserved the shell is somewhat more flattened
than it would have been in life. The posterior end is not so broad as in Hadrianus
but is more evenly rounded as in many species of Testiido. The areas covered by
the vertebral scutes are decidedly convex. The few anterior peripherals present
suggest that the front was little, if at all, emarginated on the median line. All of
the sutures remain distinct and the sulci can be clearly traced, so there can be no
question raised as to their proper interpretation in the figures. The carapace has
a length of about 360 mm. ; a breadth of 300 mm.

The neurals and costals are highly differentiated. In this species there are
only seven neurals, as in Testudo amnion Andrews, but whether this represents a
constant character in this species or only an individual variation, as in T. amnion,
must await the discovery of additional specimens. The first neural is especially
elongated and oval; the second and si.\th octagonal; the third tetragonal; the fourth,
fifth, and seventh being hexagonal. All of the neurals are longer than wide, whereas
in Hadrianus, Stylemys and most of the species of Testudo the neurals are wider
than long. The dimensions of the neurals and costals are given in the accompany-
ing table.

Dimensions of Neurals.

No.

Dimensions of Costals.

No.

Length.

Width.

Width of Proximal End. ] Width of Distal End.

1
2
3
4
5
6
7

44
39
33
31
34
39
31

23
30
25
29
28
32
24

1

2
3

4
5
6

7
8

44

28
43
30
37
21
20
20

68
47
30
60
14
49
~ 22
35

There are two suprapygals, the anterior being bifurcate and enclosing between
its right and left limbs the lozenge-shaped second suprapygal. The first has a
diameter antero-posteriorly at the center of 32 mm., a breadth of 88 mm.; the
second is 30 mm. in length and 48 mm. in width.

The pygal is wedge-shaped, the narrower truncated end being posterior, the
anterior end is notched for the second suprapygal. The under side is transversely

152

MEMOIRS OF THE CARNEGIE MUSEUM

concave, the upper slightly convex in the same direction. The free edge is acute
with a faint median projection, the whole forming a convex covering for the tail
that apparently projected but little below the level of the carapace.

Fig. 21. Testudo uintensis. Carapace of the type, C. M. No. 2331. One-fourth natural size, n.p.,
nuchal plate; nl and n7, first and seventh neurals ; plus., lip of plastron ; sp., suprapygal; sp;2., second suprapygal.

The costal plates are alternately wide and narrow, reaching as high a degree
of differentiation in this respect as any species of the genus. The diameters of
their proximal and distal ends are given in the accompanying table.

The nuchal plate is largely missing, as well as nearly all of the perii^herals of
the right side. Those of the left side have suffered some damage, so that their
exact dimensions can not always be determined. They are however of moderate
length with thin acute edges on front and back, becoming slightly obtuse along
the sides. There are eleven peripherals in the complete series. The greatest
thickness of the ninth peripheral, the thickest of any of the posterior members, is
16 mm. The second, measured at the suture with the first, is 16 mm. thick.

The plastron has a total length of 332 mm. The width of the anterior lobe
at its base is 150 mm. It has a length of 104 mm. The lip does not extend beypnd
the border of the carapace, and it projects but little beyond the general contour
of the lobe. The lower surface of the lobe is flat, while the upper is bevelled off
from back toward the front, forming rather a sharp anterior border. On the upper

GILMORE: the fossil turtles of the UINTA FORMATION

153

surface of the lip on either side of the low median ridge there are shallow longitudinal
depressions, which run forward to the slight emarginations at either side of the
center. At the gular-humeral sulcus the lateral border of the lobe is slightly
emarginate. The upper surface, 57 mm. back from the front of the lip, is almost
perpendicularly excavated, thus reducing the plastron from 21 mm. to 10 mm. in
thickness. The posterior lobe is 90 mm. long; and is 163 mm. wide at the base.
The posterior notch is 50 mm. wide and 23 mm. deep. At the inguinal notch the

Fig. 22. Testudo uinknsis. Plastron of type, C. M. No. 233L One-fourth natural size.

edge of the lobe forms a wall 22 mm. high. This wall diminishes in height poste-
riorly, so that at the anal-femoral sulcus it measures 15 mm. in height. The outer
face of this wall slopes off rather gradually to the subacute lateral edge, while the
inner edge slopes abruptly to the inner level of the floor of the carapace. The ento-
plastron is rhombic in form, 68 mm. long on the midline, and 69 mm. wide. The
hyoplastra meet along the midline for a distance of 74 mm.; the hypoplastra for
69 mm. ; and the xiphiplastra for 55 mm.

The bridge is 135 mm. wide. The plastron is quite concave and at least
suggests that this individual was a male.

The gular scutes overlap the entoplastron and on the midline have a length
of 57 mm. The humerals at the middle are 53 mm. long; the pectorals 46 mm.;
abdominals 80 mm.; femorals 42 mm.; and anals 28 mm.

154 MEMOIRS OF THE CARNEGIE MUSEUM

The vertebral scutes are longer than wide, their dimensions are given in the

table.

Dimensions of Vertebral Scutes.

No. Length. Width.

1 64 100

2 74 62

3 66 63

4 68 57

5 66 116

The costal scutes are wide, their outer ends joining the marginals at the costo-
peripheral suture. The costo-marginal sulcus appears to follow closely the course
of the costo-peripheral suture. The humero-pectoral sulcus touches the ento-
plastron, but does not cross it.

The supracaudal scute is divided as in Hadrianus.

A comparison of the type of the present species with the upper Eocene tortoise,
Testudo ammon Andrews, from the Fayum deposits of Egypt, shows some striking
resemblances. Both Testudo uintensis and T. amnion are distinguished from all
other species of the genus by the octagonal shape of the second and sixth neurals,
whereas the usual arrangement is for the second and fourth neurals to be octagonal.
Both types agree in having seven neurals, although other specimens referred to
T. anifnon by Andrews have the normal number of eight, and it may be found that
there is a similar variation in the present species.

Hay^ comments upon Testudo ammon as follows: "Dr. A.[C]E.[W] Andrews
(Surv. Dept., Pub. Works Ministry, Geol. Survey, Egypt, 1903; Tert. Vert. Fayum,
Egypt, 1906, p. 278, pi. 24) has described a land-tortoise from the Upper Eocene of
Egypt to which he has given the name Testudo ammon. If a true Testudo, it is the
oldest known. The published figures show that the neurals are variable in form,
but the relationships to typical Testudo are so lose that it may be accepted as
belonging to this genus. In some respects it appears to be intermediate between
Testudo and Hadrianus."

The above remarks would apply equally well to the species here described.
Up to this time the oldest known Testudo found in North America is from the
Lower Oligocene. Two species having been described, Testudo brontops Marsh,
and T. exornata Lambe.

If, as has been inferred, Testudo has derived its ancestry from Hadrianus, the
intermediate characters observed in the present specimen are fully in accord with
its geological position. The axillary and inguinal buttresses rising but little above

* " Fossil Turtles of North America," 1908, p. 368.

GILMORE: the fossil turtles of the UINTA FORMATION 155

the costo-peripheral suture, and the presence of a divided supracaudal scute, in-
dicate its relationship with the genus Hadrianus, as now understood and defined.
On the other hand, the greatly reduced heads of the ribs, and especially the high
degree of differentiation reached by the neural and costal bones show its affinities
to Testudo.

In an attempt to find characters other than those used to separate the three
important genera constituting the family Testudinidse I have determined the width
of the bridge as compared to the total length of the plastron in all available speci-
mens pertaining to the genera Hadrianus, Stylemys, and Testudo, and find the
average to be as follows: In Hadrianus the bridge is thirty-eight per cent, of the
length of the plastron; in Stylemys it is forty-nine per cent., and in Testudo forty-four
per cent. In Testudo uintensis it is forty per cent, thus again demonstrating its
intermediate stage of development between Hadrianus and Testudo.

Cope assigned as the principal character distinguishing Hadrianus from the
other genera of the Testudinidse, "a divided supracaudal scute," and this would
perhaps appear a good reason for assigning the present specimen to that genus if it
were not known that three living species of the genus Testudo as recognized by
Boulenger, also have this scute divided. In the definition of the genus Testudo
the lip is "usually projecting abruptly from the general contour of the lobe."
The present specimen, however, is one of the exceptions, resembling the Oligocene
Testudo amphithorax in this respect.

It will be seen from this brief discussion that the present means of separating
the genera Hadrianus and Testudo is very unsatisfactory, and it would perhaps be
best to combine them until clean cut characters are found, to show that there are
two distinct genera.

Family TRIONYCHID.E Bell.

The soft shelled Trionychidis are represented in the present collection by
thirteen specimens of the genus Amyda, four of which have been identified specif-
ically. These pertain to two species, hitherto known only from the Bridger
formation. Two fragmentary specimens, C. M. Nos. 2396 and 3285, judging
largely from the character of the sculpturing of the parts of the carapace present,
appear to represent undescribed species, but on account of their inadequate nature
I refrain from naming them. The genus Amyda is now represented in the Uinta
formation by the following species, Amyda crassa Hay, A. egregia Hay and A.
scutumantiquum (Cope).

156 MEMOIRS OF THE CARNEGIE MUSEUM

Genus Amyda Oken.
17. Amyda egregia Hay.

Amyda egregia Hay, Fossil Turtles of North America, 1908, p. 531, plate 107,

figs. 1-3, text-fig. 691.

A specimen (C. M. No. 3254), collected by Earl Douglass from Horizon B, of
the Uinta formation, three or four miles northeast of Well 2, Uinta Basin, Utah, in
1908, is identified as belonging to Amyda egregia Hay. This specimen consists of
the articulated nuchal, the greater portion of the first neural, the first, second,
third, fourth, fifth, and sixth costals, all, excepting the first, lacking portions of
their upper extremities. In the matrix on the left side are the free ends of the
second, third and fourth costal ribs. These appear to be in their proper positions
in relation to the other parts of the shell which are preserved, and serve to give
some idea of the width and general contour of the carapace. The shell is broadly
rounded in front and considerably arched transversely. The width is estimated to
have been about 465 mm. The nuchal has a transverse extent of 255 mm. and
measures 52 mm. antero-posteriorly. Hay States that the latter measurement in
the type of the species is but 22 mm., although the figure published by him shows
it to be at least 42 mm. As in the type specimen the outer end of the nuchal over-
laps the free end of the rib of the first costal. The first neural at the anterior end
is 40 mm. wide. The sculpture of the carapace is coarse. On the inner ends of the
costals the pits and ridges form a honeycomb arrangement, but on their outer
fourths the ridges and pits are arranged in rows across the costals, toward their
ends on some of the costals the honeycomb pattern again prevails, on others the
rows persist to the smooth bevelled border.

In the form of the anterior end of the carapace, with slight emargination of
the border at the sutural junction of the nuchal, and the close resemblance of the
sculpture of the carapace, this specimen is in close agreement with the type of the
species. The chief differences observed are the considerably larger size of the
present specimen and the sloping bevel of the smooth ends of the costals, as con-
trasted with the abrupt bevel of the type.

A second specimen, C. M. No. 3255, in this collection consisting of many frag-
ments of the neurals and costals is provisionally referred to this species. This
specimen is from Horizon B, Uinta formation, and from the same locality as the
individual previously discussed. It was collected by Dr. W. J. Holland, and Earl
Douglass in 1908.

The type of Amyda egregia Hay is from the lower Washakie beds south of Hay-

GILMORE: the fossil turtles of the UINTA FORMATION 157

stack Mountain, Wyoming, so that the discovery of the present specimens in the
Uinta formation of Utah, considerably extends the geological as well as the geo-
graphical range of this species.

18. Amyda scutumantiquum (Cope).
Plate XXVI, fig. 2.

Trionyx scutumantiquum Cope, 6th Ann. Report U. S. Geol. Surv. Terr., 1872
(1873), p. 617; Amer. Naturalist, XVI, 1882, p. 988, fig. 6; Vert. Tert. Form.
West., 1884, pp. 118, 121, PI. XVI, figs. 1, la; Hay, Bibhog. and Cat. Foss.
Vert. N. A., 1902, p. 454.

Amyda scutumantiquum Hay, Amer. Geologist, 35, 1905, p. 336; Fossil Turtles of
North America, 1908, pp. 521, 522, plate 100, figs. 2-4, plate 101, fig. 1; text-
figs. 676, 677.

A very large specimen C. M. No. 3272 (see plate XXVI, fig. 2), consisting of a
considerable part of the anterior two-thirds of the carapace, is identified as per-
taining to the above genus and species. This specimen was collected by Earl
Douglass May 25, 1908, from the lower portion of Horizon B, "transition beds
sandstone," Uinta formation. Upper Eocene, as exposed on the south branch of
Red Bluff Wash, above the well on the stage-road between Bonanza and Kennedy's
Hole, Uinta County, Utah.

This specimen lacks all of the carapace posterior to the fifth costals, the
anterior border and left end of the nuchal, and small portions here and there of the
costals forward of the sixth. The sandstone matrix containing the impressions of
the seventh and eighth costals fortunately is preserved and serves to give a fairly
accurate idea of the dimensions of the entire shell.

In form the carapace is broadly oval with the length shghtly exceeding the
breadth. The greatest width, as in the type, appears to have been at the middle.
The extreme width is about 530 mm.; the length, at the very least, was 570 mm.
The shell is broadly arched from the lateral borders to beyond the middle of the
costal plates. Along the middle of the back there is a pronounced longitudinal
depression which is deepest at about the middle of the shell.

The nuchal extends on each side of the midline about 180 mm.; its width on
account of the missing anterior border cannot be given. At the center on the
broken border the nuchal has a thickness of 23 mm. Its outer end is bevelled off
toward the front and outer extremity. The length of the nuchal is .74 the width of
the shell, whereas the type of the species is .76 (not .80 as stated by Hay).

158 MEMOIRS OF THE CARNEGIE MUSEUM

The first neural is exceedingly large, being 93 mm. long and 50 mm. wide toward
the front. The others present diminish in size posteriorly. The second, third,
and fom-th are coffin-shaped, with the widest end posterior, as is usual in the species
of this genus. The principal dimensions of the neurals are given in the table.

No. Length. Width.

1 93 50

2 62 41

3 59 38

4 58 34

All of the costal plates except the first grow wider toward their outer ends.
The expansion of the outer end of the second, as in the type, is especially pronounced.
In the angulation of the free border of the second costal it resembles that of Amyda
salebrosa Hay, more nearly than the type of the present species. At their free ends
all of the costals preserved are bevelled off to a sharp edge, except at the point
where the rib projects. None of the projecting ribs are preserved, so that the
distance they extend beyond the free border cannot be determined. Near their
outer ends at the sutural borders the costals have a thickness of from 5 to 9 mm.
The width of the proximal and distal ends of the costals are given in the table.

No. Width of Proximal End. Width of Distal End.

1 91 62

2 58 103

3 63 79

4 58 79

5 59 84e

e, estimated.

The surface of the carapace is ornamented with the usual pits and ridges, there
being two pits in a line 10 mm. long, and occasionally three. They are large and
distinct, forming a honeycomb arrangement along the middle and on the proximal
halves of the costal plates, becoming smaller and less deeply impressed toward the
front. In small areas here and there at points about the middle of the costals the
pits are arranged in rows across the short diameters of the plates. On the distal
portions of the costals the pits are smaller and less deeply impressed, thus forming
a very distinct pattern, and, as the smooth band is approached, the ridges show a
tendency to break up into tubercles. Nowhere are the ridges wider than the pits.

Except for the considerably greater size of the specimen, and slight differences
in the general distribution of the large and small pits the present individual agrees
very closely with the type. The latter is from the Bridger beds on Cottonwood

GILMORE: the fossil turtles of the UINTA FORMATION 159

Creek, in Wyoming, and the specimen here described is the first record of the oc-
currence of this species in the Uinta formation.

A second specimen, C. M. No. 3330, was collected by Earl Douglass in 1908,
near the region of Well No. 2, Uinta Basin, Uinta County, Utah. It comes from
the Uinta formation (horizon not given) and consists of fragments of costal plates
having sculptured surfaces which are identical with those of the specimen discussed
above, and is therefore regarded as pertaining to the present species.

Order SQUAMATA.

Suborder SAURIA.

Family ANGUID^.

Genus Glyptosaurus Auct.

19. Glyptosaurus sp. indet.

A specimen, C. M. No. 3405, consisting of the greater part of the parietal with
other fragments of the skull and lower jaws, and a few shields, is identified as per-
taining to the lizard-like reptile Glyptosaurus. It was collected by 0. A. Peterson,
August 24, 1912, on White River near Ouray, Uinta Basin, Utah, from Horizon C,
Uinta formation. Upper Eocene.

The few osseous shields present are evidently from the trunk of the body.
These are oblong quadrate, with a smooth, thinned out anterior end, which is over-
lapped by the next plate of the series. The lateral borders are roughened for sutural
union. The external surface, excepting the smooth area mentioned above, is
ornamented with small rounded tubercles closely arranged in more or less con-
centric rows. The cranial shields on the parietal are of irregular sizes with their
surfaces ornamented much in the same manner as the trunk-shields.

This specimen may represent an undescribed species, but at the present time,
on account of the very unsatisfactory type specimens upon which the nine de-
scribed species have been based, it is impossible to make adequate comparisons,
so that the specific determination of the present specimen must await the thorough
revision of the species of the genus. It is of interest, however, as recording for the
first time the occurrence of the genus Glyptosaurus in the Uinta formation, and also
from the fact that it occurs intermediate geologically between the oldest known
specimens from the Bridger described by Marsh, and the youngest specimen dis-
covered and described by Douglass from the Oligocene."

^ Douglass, Earl, "Some Oligocene Lizards," Annals of the Carnegie Museum, IV, 1908, pp. 278-283.

160

MEMOIRS OF THE CARNEGIE MUSEUM

Nine species of the genus have been described from the deposits of North
America. Geologically these are arranged as follows:

Glyptosaurus montanus Douglass.

Lower Wlute River, OUgocene.

Glyptosaurus sp. indet.

Uinta formation, Upper Eocene.

Glyptosaurus sylvestris Marsh.

G. nodosus Marsh.

G. oscellatus Marsh.

G. princeps Marsh.

G. brevidens Marsh.

G. riigosus Marsh.

G. sphenodon Marsh.

G. anceps Marsh.

Bridger, Eocene.

EXPLANATION OF PLATES.

Plate XVIII.

Fig. 1. Carapace of Baena arenosa Leidy. C. M. No. 2356. One-half natural size.
Fig. 2. Plastron of Baiina platyplastra. Type. C. M. No. 3227. One-third natural size.

Plate XIX.
Fig. 1. Carapace of Bae7ia inflata. Type. C. M. No. 3406.
Fig. 2. Plastron of the same. Both figures one-third natural size.

Plate XX.
Fig. 1. Carapace of Baena gigantea. Type. C. M. No. 3441.
Fig. 2. Plastron of the same. Both figures one-fourth natural size.

Plate XXI.
Fig 1. Carapace of Echmafemys caUopyge Hay. C. M. No. 2371.
Fig. 2. Plastron of the same. Both figures one-third natural size.

Plate XXII.
Fig. 1. Carapace of Echniateniys douglassi. Type. C. M. No. 3244.
Fig. 2. Plastron of the same. Both figures one-third natural size.

Plate XXIII.
Fig. 1. Carapace of Echmatemys hollandi. Type. C. M. No. 3249. Viewed from the
right side. One-half natural size.

Fig. 2. Carapace of Echmatemys depressa. Type. C. M. No. 2936. Natural size.

Plate XXIV.
Fig. 1. Carapace of Echmatemys obscura. Type. C. M. No. 3252.
Fig. 2. Plastron of the same. Both figures one-third natural size.

GILMORE: the fossil turtles of the UINTA FORMATION 161

Plate XXV.
Fig. 1. Plastron of Hadrianus corsoni (Leidy). C. M. No. 3403. One-fifth natural size.
Fig. 2. Plastron of Hadrianus robustus. Type. C. M. No. 3342. One-third natural size.

Plate XXVI.

Fig. 1. PlastroQ of Hadrianus utahensis. Type. C. M. No. 2343. One-fourth natural
size.

Fig. 2. Carapace of Amijda scutumantiquum (Cope). C. M. No. 3272. One-fourth
natural size.

Plate XXVII.
Fig. 1. Carapace of Testudo uintensis. Type. C. M. No. 2331.
Fig. 2. Plastron of the same. Both figures one-half natural size.

Memoirs Carnegie Museum, Vol. VII.

Plate XVIII.

1. Carapace of Baena arenosa Leidy. C. M. No. 23.56. X |-

2. Plastbon of Baena platyplastra Gilmore. Type, C. M- No. 3227. X |.

Memoirs Carnegie Museum, Vol. VII.

Plate XIX.

1. Carapace of Baena inflata Gilmore. Type, C. M. No. 3406. X -3-.
2. Plastron of Do. X I-

Memoirs Carnegie Museum, Vol. VII.

Plate XX.

Nj

1. Carapace of Baena gigantea Gilmore. Type, C. M. No. 3441. X i-
2. Plastron of same. X j.

Memoirs Carnegie Museum, Vol. VII.

Plate XXI.

1. Carapace of Echmatemys callopyge Hay. C. M. No. 2.371. X i-
2. Plastron of same. X |.

Memoirs Carnegie Museum, Vol. VII.

Plate XXII.

1. Carapace of Eclunatemijs douglasul Gilmoke. Type, C. M. No. 3214. X |.

2. Plastron of same. X ^.

Memoirs Carnegie Museum, Vol, VII.

Plate XXIII.

1. Carapace of Echmatemijs holknuli Gilmore. Type, C. M. No. 3249. X 5.

2. Carapace of Echmatemijs depressa Gilmore. Type, C. M. No. 2936. X -f-

Memoirs Carnegie Museum, Vol, VII.

PLATE XXIV.

1. Carapace of Echmatemys dbscura Gilmore. Type, C. M, No- 3252. X \-

2, Plastron of same. X 3.

Memoirs Carnegie Museum, Vol, VII,

Plate XXV.

r"'

'-\- -.^r^r-"

1. Plastron of Hadrianus corsoni (Leidy). C. M. No. 3403. X i-

2. Plastron of Hadrianus robustus Gilmore. Type, C. M. No. 3342. X h

Memoirs Carnegie Museum, Vol. VII.

Plate XXVI.

i^*

1. Plastron of Hadrianus utahensis Gilmore. Type, C. M. No. 2343. X 3.

2. Carapace of Aniyda scutummitiquum (Cope). C. M. No. 3272. X J.

Memoirs Carnegie Museum, Vol. vil.

Plate XXVII.

1. Carapace of Testudo uintensis Gilmore. Type, C. M. No. 2331. X i-
2. Plastron of same. X J.

21. Minute Book of the Virginia Court Held at

Fort Dunmore (Pittsburgh) for the District

of West Augusta, 1775-1776. Crummne 90

22. Minute Book of the Virginia Court Held for

Yohogania County, first at Augusta Town
(now Washington, Pa.), and afterward on the
Andrew Heath Farm near West Elizabeth,
1776-1780. Crummne 2.25

23. Minute or Order Book of the Virginia Court

Held for Ohio County, Virginia, at Black's
Cabin (Now West Liberty, W. Va.), &c.
Crumeine 1.50

24. The Records of I>eeds for the District of West

Augusta, Virginia, for the Court Held at Fort

Dunmore, &c. Crdmrine 1.75

25/ Astropecten (?) montanus, &c. Douglass 10

26. Astrodon (Pleurocoelus) in the Atlantosaurus

Beds of Wyoming. Hatcher. (Out of Print.)

27. Osteology of the Limicolae. Shufeldt 1.00

28. New Vertebrates from the Montana Tertiary.

Douglass 1.25

29. New Genus and Species of Tortoise from the

Jurassic of Colorado. O. P. Hay 10

30. Osteology of Oxydactylus. Peterson 1.00

31. Birds of Erie and Presque Isle. Todd 75

32. J. B. Hatcher. By W. J. Holland 15

33. Tropidoleptus Fauna at Canandaigua Lake, etc.

Raymond. {Out of print.)

34. Two Turtles from the Judith River Beds of

Montana. O. P. Hat. {Out of print.)

35. Preliminary List of Hemiptera of Western

Pennsylvania. Wirtner. {Scarce.) 50

36. Trilobites of the Chazy Limestone. Raymond. 1.00

37. Crawfishes of Western Pennsylvania. Ort-

MANN 1.00

38. The Geology of Southwestern Montana. Doug-

lass 40

39. New Crocodile from the Jurassic of Wyoming.

Holland 10

40. Procambarus, a New Subgenus of the Genus

Cambarus. Ortmann 15

41. Presentation of Reproduction of Diplodocus Car-

negei to the British Museum. Holland . .". . .15

42. Birds Collected near Mombasa by William Do-

heity. Holland 20

43. Hyoid Bone in Mastodon Americanus. Hol-

land 10

44. Additions to Flora of Allegheny County, Pa.

Jennings 15

45. New Species of Kneiffia. Jennings 05

46. Occurrence of Triglochin palustris in Pa. Jen-

nings 05

47. New Species of Ibidium. Jennings 10

48. Agate Spring Fossil Quarry. Peterson 10

49. Two New Birds from British E. Africa. Obeb-

HOLSER 05

50. The Chazy Formation and Its Fauna. Ray-

mond 1.50

51. A New American Cybele. Narrawat and U.\y-

KOND ' 15

52. Plastron of the Protosteginae. Wieland 15

53. New Species of Testudo from the Miocene, etc.

O. P. Hay 25

54. Miocene Beds of W. Nebraska and E. Wyoming

and Their Vertebrate Faunae. Peterson .... 1.00

55. New Species of Lonicera from Pa. Jennings. .05

56. MeryocochoBrus, etc. Douglass.

57. New Merycoidodonts. Douglass. (Nos. 56

and 57 sold together.) 1.00

58. Further Collections of Fishes from Paraguay.

Eigenmann, McAtee, and Ward 1.25

59. Undetermined Element in the Osteology of the

Mosasauridae. Holl.us;d 20

60. Gasteropoda of the Chazy. Raymond 1.35

61. Occurrence of Wynea Americana in Pa. Jen-

nings 05

62. Account of Pleistocene Fauna Discovered at

Franl^stown, Pa. Holland 10

63. Vertebrate Fossils from the Vicinity of Pitts-

burgh, Pa. Case 15

64. Dlaenidae from the Black River Limestone near

Ottawa, Canada. Raymond and Narraway. . .20

65. Rhinoceroses from the Oligocene and IVIiocene

of North Dakota and Montana. Douglass . . .25

66. FossU Horses from North Dakota. Douglass. .30

67. Oligocene Lizards. Douglass 20

68. The Type of Stenomylus gracilis. PETiatsoN . . .25

69. New Species of Birds from Costa Rica, etc.

Carriker 10

70. Costa Rican Formicariidae. Carbiker 05

71. Vertebrate Fossils from the Fort Union Beds.

Douglass 45

72. List of the Lepidoptera of Western Pa., etc.

Kngel 1.25

73. Fauna of Upper Devonian in Montana. Fossils

of Red Shales. Raymond 60

74. New Species of Procamelus from Upper Mio-

cene of Montana. Douglass 30

75. Sections of the Conemaugh Series between

Pittsburgh and Latrobe, Pa. Raymond 35

76. List of the Unionidae of Western Pa. Ortmann .50

77. Geological Reconnaissance in North Dakota,

Montana, and Idaho. Douglass 1.00

78. Botanical Survey of Presque Isle. Jennings . . 2.75

79. Sesqui-Centennial (Pittsburgh) Relics. Stewart .45

80. Dromomeryx, New Genus of American Rumi-

nants. Douglass 30

81. FossUs from Glacial Drift and Devonian and

Mississippian near Meadville, Pa. Millard.. '.10

82. New Species of Helodus. Eastman 05

83. Charles Chauncey MeUor. Holland 15

84. Reports of Expedition to British Guiana, etc.

Eigenmann 50

85. Reports of Expedition to British Guiana, etc.

DuKBIN 25

86. Odonata of the Neotropical Region, Exclusive

of Mexico and Central America. Calvert . . . 2.25

87. Deinosuchus hatcheri, a New Genus and Species

of Crocodile. Holland 20

88. Reports on Expedition to British Guiana, etc.

Blosser 15

89. Description of Some New Titanotheres from

the Uinta. Douglass 25

90. Annotated List of the Birds of Costa Rica In-

cluding Cocos Island. Carbiker 3.00

91. Geology of the Coast of the State of Alagoas,

Brazil. Branner 40

92. Fossil Fishes from the Bituminous Shales at Ri-

acho Doce, Alagoas, Brazil. Jordan 55

93. Ordovician Trilobites, No. II. Asaphidae from

the Beekmantown. R.^ymond 35

94. Ordovician Trilobites, No. m. Raymond &

Narraway 35

95. Ordovician Trilobites, No. IV. Baymond 40

96. Fishes from Irkutsk, Siberia. Jordan and

THOJrpsoN 30

97. South American Tetrigidae. Bruner 1.00

98. Fauna of the Allegheny and Conemaugh Series

in Western Pa. Raymond 30

99. Ichthyological Survey About the San Juan

Islands, Washington. Stabks 75

100. New Species of Pygidium. Eigenmann 10

101. The Brachiopoda and Ostracoda of the Chazy.

Raymond 50

102. A New Camel from the Miocene of Western

Nebraska. Peterson 15

103. Skeleton of Stenomylus hitchcocki, etc. Peter-

son 15

104. Skeleton of Diceratherium cooki. Peterson . . .15

105. Carnegie Museum Expedition to Central South

America, 1907-1910. Holland 15

106. Report Upon the Expedition of the Carnegie

Museum to Central South America. Hase-
man & Eigenmann 25

107. New Species of Fishes, etc., from Central South

America. Haseman 50

108. Annotated Catalog of Oichlid Fishes from Cen-

tral South America. Haseman 1.25

109. New Species of Fishes from the Rio Iguassu.

Haseman 65

110. Ornithology of the Bahama Islands. Todd &

WORTHINGTON 75

112. South American Acridoidea. I. Bbuner .... 1.00

113. Species of Hasemania, Hyphessobrycon, and

Hemigrammus. Ellis 25

114. New Characins in the Carnegie Museum. Eigen-

mann 30

115. Jurassic Saurian Remains Ingested within Fish.

Eastman 20

116. Autograph Letter of U. S. Grant to Edwin M.

Stanton. Holland 15

117. Albert J. Barr. By W. J. Holland 10

118. Seventeen New Neotropical Birds. Todd 25

119. Dr. David Alter, the First Discoverer of Spec-

trum Analysis. Holland 10

120. Two Mummy Labels. Allen 10

121. The Families and Genera of Najades. Ort-

mann 1.00

122. Group of Stenomylins in the Carnegie Museum.

Peterson 25

123. Tertiary Fish-remains from Spanish Guinea.

Eastman 25

124. Plated Nematognaths. Ellis 65

125. New Species of Cambarus from the Isle of

Pines. Ortmann 15

126. Sedum Camegiei, a New Species of the Family

Crassulaceae, etc. Hamet 10

127. Two New Species of Fishes from Peru. Eigen-

mann 15

128. South American Locusts (Acridoidea). II.

Bruner 75

129. Revision of the Genus Cheemepelia, Todd 85

130. New Genus and Species of Abyssinian Rodents.

Childs Feick 20

131. New Titanothere from the Uinta. Peterson.. .20

132. Small Titanothere from the Lower Uinta. Peter-

son 10

133. Osteology of Lasiopyga and Callithrix, etc.

Shufeldt 25

134. New Rhynchocephalian from Solenhofen. Grier .15

135. Scales of South American Characinid Fishes.

COCKERELL 25

136. Skeleton of Platigonus Leptorhinus. Peter-

.son 10

137. Lichens Collected in the Thunder Bay District,

Ontario. R. Heber Howe, Jr 10

138. Preliminary List of Fossil Plants in the Roof

of the Pittsburgh Coal. Geieb 10

139. Undescribed Remains of the Uinta Titanothere

Dolichorhinus. Peterson 15

140. Triassic Fishes Belonging to the Catopteridae

and Semionotidae. Eastman 15

141. Osteology of Promerycochoerus. Peterson ... .40

142. Correction of a Generic Name. Peterson 05

143. Skull of Bison Crassicomis. Holland 10

144. Serrasalminae and Mylinse. Eigenmann 50

145. Heads and Tails: Notes on Sauropod Dinosaurs.

Holland 15

146. Dipterus Remains from Upper Devonian of

Colorado. Eastman 10

147. Notes on Tropical American Tettigonoidea (Lo-

custodea). Brunee 1.65

148. New Species of Tortoise from Jurassic of

Utah. GiLMORE 15

149. Faima of Upper Devonian in Montana.

Haynes 45

150. New Sphagebranchus from Bahamas. Eigen-

mann 07

151. Marine Fishes from Colombia and Ecuador.

Wilson 15

152. Apareiodon, a New Genus of Characid Fishes.

Eigenmann 18

153. New and Rare Fishes from S. American Rivers.

Eigenmann 25

154. Three New Species of Characid Fishes. Eigen-

mann & Henn 12

155. The Species of Salminus. Eigenmann 7

156. South American Poeciliid Fishes. Henn 40

157. A New Species of Apatosaurus. Holland 7

158. Birds of the Isle of Pines. Todd 70

159. Reptiles and Amphibians of the Isle of Pines.

Barbour 20

160. Land and Fresh Water Shells of the Isle of

Pines. Henderson 20

161. Pelecypoda of the Chazy. Raymond 45

162. S. American Crickets, Gryllotalpoidea and Ache-

toidea. Bruner 1.30

163. Preliminary Catalog of N. American Sphaeri-

idae. Steeki 75

164. Directions for Collecting Sphaeriidae and

Aquatic Gastropods. Steeki 10

165. Lepidoptera of the Isle of Pines. Holland 50

166. Odonata of the Isle of Pines. Kahl 15

167. A Trip to Islands in Lake Erie. Goodrich 15

168. Land-Shells of Islands at Western End of Lake

Erie, etc. Clapp ^

169. Orthoptera of the Isle of Pines. Holland & ^

Kahl 15

1^. 1IC>

Publications of the Carnegie Museum, Serial No. 92.

MEMOIRS

OF THE

CAENEGIE MUSEUM.

VOL. VII. No. 3.

W. J. HOLLAND, Editor.

A CATALOG OF THE OPHIDIA FEOM SOUTH AMERICA

AT PEESENT (JUNE, 1916) CONTAINED IN THE

CAENEGIE MUSEUM V^ITH DESCEIPTI0N8

OF SOME NEW SPECIES

By LAWRENCE EDMONDS GRIFFIN.

-T" PITTSBURGH.
Published by the AniHORiiy of the Boabd of Trustees of the
CARNEGIE INSTITUTE.
November, 1916.

PRICE LIST OF PUBLICATIONS OF THE CARNEGIE MUSEUM

> ANNUAL REPORTS OF THE DIRECTOR

1898, 30c. (scarce) ; 1899, 25c.; 1900, 30c. (scarce) ; 1901- 16, 25c. each.

REPORTS or PROCEEDINGS OF FOUNDER'S DAY

1898-1916, 35c. each.

ANNALS OF THE CARNEGIE MUSEUM

The Annals are supplied to those who subscribe in advance in parts (paper-bound), as published, @ $3.50 per volume;
Vols. I-X, 1901-1916, bound in green cloth @ $4.00; bound in i Morocco @ $4.50.

MEMOIRS OF THE CARNEGIE MUSEUM

The Memoirs are supplied to those who subscribe in advance in parts (paper-bound), as published, at $10 per volume;
bound in green buckram at $10.75, and in half-morocco at $12.00.

VOL. I. (1901-3)

No. 1. Diplodocus, Its Osteology, Taxonomy, etc. No. 3. Osteology of the Steganopodes. Shupeldt $2.75

Hatchkr $3.00 ' ' 4. Classification of Chalcid Flies. Ashme.u). . 6.50

' ' 2. Oligocene Canidse. Hatcher 1.50

VOL. II. (1904-6)

No. 1. OsteologyofHaplocanthosarus, etc. Hatcher $2.00 " 6. Osteology of Diplodocus. Holland 1.50

' ' 2. Osteology of Baptanodon. Gilmore 1.50 ' ' 7. Osteology of Protostega. Wibland 75

" 3. Fossil Avian Remains from Armissan. East- " 8. New Suilline Remains from the IMiocene of

MAN 1.00 Nebraska. Peterson 75

' ' 4. New Rodents and Discussion of Origin of • ' 9. Notes on the Osteology of Baptanodon, etc.

Daemonelix. Peterson 1.75 Gilmore 1.00

No. 5. Tertiary of Montana. Douglass $1-00 ' ' 10. The Crawfishes of Pennsylvania. Ortmann 4.00

VOL. in.

No. 1. ArchJeological Investigations in Costa Rica. No. 2. Osteology of Moropus. Holland and Petee-

Hastman $6.00 SON $6.00

VOL. IV.

No. 1. Early Chinese Writing. Chalfant $3.00 No. 5. New Carnivores from the Miocene of West-

' ' 2. Fonnosan Fishes. Jordan 25 em Nebraska. Peterson $1.50

' ' 3. Entelodontidse. Peterson 2.50 ' ' 6. Monograph of the Najades of Pennsylvania.

' ' 4. Fishes of Formosa. Jordan and Richard- Pts. I and II. Ortmann 2.50

sojj 1.25 ' ' 7. Catalog Eocene Fishes from Monte Bolea in

Carnegie Museum. Eastman 3.00

VOL. V.

The Fresh Water Fishes of British Guiana. Eigenmann. $10 unbound; $10.75 cloth; $12.00 J morocco.

VOL. VI.

No. 1. Fishes from Korea. Jordan and Metz .... $1.50 No. 4. Fishes obtained in Japan. 1911. Jordan

' ' 2. Lantern-fishes of Japan. Gilbert 1.00 and Thompson $3.50

" 3. Gymaotid Eels of Tropical America. Max Nos. 5-7. FossU Fishes in the Carnegie Museum.

Mapes Ellis 1.50 Parts II-IV. Eastman 5.00

VII.

No. 1. The Cheirodontinae. Eigenmann $3.50 No. 3. Ophidia from S. America. Grippin 2.00

' ' 2. Turtles of Uinta Formation. Gilmore 2.00

REPRINTS FROM THE ANNALS OF THE CARNEGIE MUSEUM

Nos. 1-12. See preceding lists. (Out of print.) 16. Osteology of Psittaci. Shufeldt 35

12. Skeleton of Titanotherium dispar. Hatcher. . $ .35 17. Annotated Catalogue of Genus Partula. H. H.

Smith 1.25

13. Fore Limbs and Manus of Brontosaurus. is. Two New Bahaman Lepidoptera. Holland... .10

Hatcher 60 ^g Elosaurus Parvus. Peterson & Gilmore 15

14. The Trachodontidae. Hatcher 25 20. Boundary Controversy Between Pennsylvania

15. New Pennsylvania Thorns. Ashe 25 and Virginia, 1748-1785. Ckumrine 30

IAN S3 19l7i

MEMOIRS

OF THE

CAENEGIE MUSEUM.

VOL. VII. NO. 3.

A CATALOG OF THE OPHIDIA FROM SOUTH AMERICA AT PRESENT

(JUNE, 1916) CONTAINED IN THE CARNEGIE MUSEUM

WITH DESCRIPTIONS OF SOME NEW SPECIES.

By Lawrence Edmonds Griffin.

The recent reptilia in the Carnegie Museum represent collections made at
various times and places by collectors, who incidentally in connection with other
undertakings preserved such specimens belonging to this class as came to hand.
No systematic collecting of these animals was undertaken by any of the expeditions
which have gone out from the Carnegie Museum to the American tropics, except
by Mr. G. A. Link, Sr., in the Isle of Pines. The collection made by Mr. Link
has already been reported upon in the Annals of this Museum.^ Although the
collections have been acquired for the most part as the result of somewhat desultory
collecting, they nevertheless include a considerable number of interesting species,
some of which appear to be new to science.

Many specimens were preserved by Mr. John D. Haseman during his ex-
tensive travels in South America from 1907 to 1910. For a brief outline of the
journeyings of Mr. Haseman the reader is referred to the articles published in the
seventh volume of the Annals of the Carnegie Museum." Mr. Haseman went to

^ "The Reptiles and Amphibians of the Isle of Pines." By Thomas Barbour, Annals Carnegie Museum,
Vol. X, 1916, pp. 297-308, PL XXVIII.

'^ W. J. Holland, "The Carnegie Museum Expedition to Central South America, 1907-1910," Annals
of the Carnegie Museum, VII, 1910-1911, pp. 283-286.

John D. Haseman, "A Brief Report upon the Expedition of the Carnegie Museum to Central South
America," I. c, pp. 287-299.

C. H. Eigenmann, "A List of LocaUties at which Mr. Haseman Collected," I. c, pp. 299-314.

163

164 MEMOIRS OF THE CARNEGIE MUSEUM

South America primarily to collect fishes. Incidentally he preserved a consider-
able number of such reptilia as he encountered. The localities at which he ob-
tained reptilia are given herewith in alphabetical order:

Arequa, Paraguay.

Asumpcion, Bolivia.

Bom Jesus de Lapa, Bahia, Brazil.

Cacequy (Rio Ibicuhy into Uruguay) Rio Grande do Sul, Brazil.

Cidade de Matto Grosso, Brazil.

Entre Rios, Minas Geraes, Brazil.

Jacarehy, Rio Parahyba, Sao Paulo, Brazil.

Lagoa de Joao Pereira, near Barra, Rio Sao Francisco, Minas Geraes, Brazil.

Mogy das Cruces, Rio Tiete, Sao Paulo, Brazil.

Muniz Freire, Espiritu Santo, Brazil.

Penedo, Alagoas, Mouth of Rio Sao Francisco, Brazil.

Piracicaba, Rio Parana, Sao Paulo, Brazil.

Puerto Suarez, Eastern Bolivia (near Corumba).

Rio Doce, Minas Geraes, Brazil.

Rio de Janeiro, Brazil.

Rio Mamore, Bolivia, below the mouth of the Rio Guapore.

Santarem, Brazil.

Sao Antonio de Guapore, Matto Grosso, Brazil.

Sao Cruz, Campos de Matto Grosso, Brazil.

Sao Joao del Rey, Minas Geraes, Brazil.

Sao Luis de Caceres, Rio Paraguay, Matto Grosso, Brazil.

Sao Matias, Bolivia.

Sete Lagoas (into Rio das Velhas, into Rio Sao Francisco), Minas Geraes, Brazil.

Urucum Mountains, Matto Grosso, Brazil.

Villa Bella, Bolivia.

Xiririca, Rio Ribeira de Iguape, Sao Paulo, Brazil.

Mr. and Mrs. H. H. Smith collected in the Province of Santa Marta, Colombia,
from 1898 to 1901. The snakes contained in their collections were taken at Bonda
(150 ft. alt.); Cacagualito (1,500 ft.); Minca (2,000 ft.); Masinga (2,000 ft.);
Valparaiso (4,500 ft.); Las Nubes (4,500 ft.); El Libano (6,000 ft.).^ The labels
in their collections usually give the month in which the specimen was captured,
but very rarely the year. It is probable, however, that most of the reptiles of
the Smith Collection were taken in the year 1901.

' The localities at which Mr. and Mrs. H. H. Smith collected in Colombia are described in the Annals
of the Carnegie Museum, Vol. VI, 1909, pp. 74-76.

griffin: ophidia from south America in carnegie museum 165

The snakes cataloged by me under the Department Numbers 1841 to 1873
are without original data or field-labels, but they are positively known to have
been received from South America, and are undoubtedly a part of the collections
of H. H. Smith from northern Colombia.

The collections sent in by Senor Don Jose Steinbach were all taken in the
Department of Santa Cruz de la Sierra, Provincia del Sara, Bolivia, except a small
part of his sendings, which are labelled as from Puerto Suarez, Bolivia. Only
two definite localities are given on his labels, Santa Cruz de la Sierra, and Las
Juntas. The altitudes given range from 250 to 450 meters above sea-level.

A small collection of snakes made by Mr. Thomas LeBoutelier was given to
the Museum in 1909 by Mrs. A. Marshall Bell of Pittsburgh. It is unfortunate
that the maker of this collection was not careful in labelling the specimens. Most
of them have no other locality than "South America" assigned to them, while in
a few cases the locality given is manifestly incorrect.

Miss Lola Vance collected a few reptiles at Tarma, Peru. The altitude of
this place is 6,000 feet.

A few serpents were also collected by Mr. J. 0. Kerby at Massasao and Pran-
quina on the Amazon, and donated by him to the Carnegie Museum.

For the greater convenience of those who may refer to this paper I have
arranged the species of each genus in alphabetical order, when more than one species
is listed, and the genera of each family and subfamily are also arranged in like
manner. In the statistical tables I have introduced some modifications of cus-
tomary usage. Under "upper labials" the figure outside of the parenthesis in-
dicates the total number of shields, while the figures within the parenthesis tell
which supralabials are in contact with the eye. The formula for the temporals
is written with a comma, instead of with the plus-sign. When there is a difference
in the number or arrangement of the shields of the left and right sides, both are
given, the formula for the left side being always written above that for the right

side.

Class REPTILIA.

Order SQUAMATA.

Suborder SERPENTES.

Family TYPHLOPID.E.

Genus Helminthophis Peters.

1. Helminthophis bondensis sp. nov.

Rostral half the width of the head, rounded posteriorly, forming a broad
suture with the frontal, extending to a line connecting the posterior margins of

166 MEMOIRS OF THE CARNEGIE MUSEUM

the eyes; frontal broader than the rostral, but little longer than the diameter of the
eye, in contact with the prefrontal and the ocular; prefrontal in contact with
frontal, ocular, and subocular; no preoculars; a subocular; eye under the posterior
and lower part of the ocular; four upper labials, the first the largest, second and
third in contact with the subocular, the second also in contact with the upper
nasal, which reaches the subocular. Twenty-two scale-rows around the body;
diameter of body contained forty-five times in the length. Tail shorter than
broad, ending in a point, but without spine.

Brown, each scale darker in the center; a little lighter beneath. Tip of head
lighter with fine brown markings.

This species resembles H. albirostris Peters more closely than any other. It
is distinguished by having one, instead of two suboculars, by the position of the
eye under the ocular shield, and by the second upper abial being separated from
the prefrontal by the nasal.

Type, d', No. 216, Carnegie Museum Catalog of Ophidia; Bonda, Colombia:
H. H. Smith (coll.). May.

Counts and Measurements.

Scale-rows 22

Upper labials 4

Preoculars 0

Suboculars 1

Diameter of body 4 mm.

Total length 180 mm.

Length of tail 2 mm.

Genus Typhlops Oppel.

2. Typhlops reticulata (Linna;us) .

Anguis reticulata Linn^us, Syst. Nat., Ed. XII, I, 1766, p. 391.
Typhlops reticulatus Boulenger, Cat. Snakes, I, 1893, p. 27.

The collection contains two specimens of this species:
No. 17, c?, Province del Sara, Bolivia, Elev. 350 M., Steinbach coll., Feb., 1911.
No. 334, 9 , Villa Bella, Bolivia, Haseman coll., Oct. 8, 1909.

Counts and Measurements.

(No. 17.) (No. 334.)

Scale-rows 20 20

Upper labials 4 4

Preoculars 1 1

Total length in mm 257 334

Length of tail in mm 8.5 4

Diameter in mm 8 8

griffin: ophidia from south America in carnegie museum

167

Family BOID^ Gray.

Genus Boa* Linnaeus.

3. Boa cooki (Gray).

Corallus cookii Gray, Zool. Misc., 1842, p. 42.
Corallus cookii Boulenger, Cat. Snakes, I, 1893, p. 99.

The five specimens in the collection are listed as follows:

No. 137, (f, Bonda, Colombia, Elev. 150 ft.. Smith coll., June.

No. 139, cf , Bonda, Colombia, Elev. 150 ft., Smith coll., Sept. 6 (in house).

No. 142, c?, Bonda, Colombia, Elev. 150 ft., Smith coll., June.

No. 204, 9 , Bonda, Colombia, Elev. 150 ft.. Smith coll., July.

No. 1860, cf, South America (Colombia, H. H. Smith).

Counts and Measurements.

(No. 137.)

(No. 139.)

(No. 142.)

(No. 204.)

(No. 1860.)

Anal

1
43
270
116
904
170

1

39

269

114

1,596

328

1

43

270

114

1,395

290

1

42
272
115
567
112

1

Scale-rows

Gastrosteges

Urosteges

Total length in mm

Length of tail in mm

41
271
114
593
110

4. Boa hortulana (Linnseus).

Coluber hortulanus Linnaeus, Mus. Ad. Frid., 1754, p. 37.
Boa hortulana Linn^us, Syst. Nat., Ed. XII, 1766, I, p. 374.
Corallus hortulanus Boulenger, Cat. Snakes, I, 1893, p. 101.

The two specimens in our possession are the following:

No. 370, cf , Entre Rios, Brazil, Haseman coll., June 4, 1908.

No. 371, Entre Rios, Brazil, Haseman coll., June 4, 1908 (head only).

Counts and Measurements.

(No. 370 )

Anal 1

Scale-rows 53

Gastrosteges 266

Urosteges 109

Total length in mm 1,616

Length of tail in mm 300

< See Stejneger, Proc. U. S. Nat. Mus., XXIV, 1902, p. 184.

168

MEMOIRS OF THE CARNEGIE MUSEUM

Genus Constrictor Laurcnti.

5. Constrictor constrictor (Linnaeus).

Boa constrictor Linn^us, Syst. Nat., Ed. XII, 1766, I, p. 373.
Boa constrictor Boulenger, Cat. Snakes, I, 1893, p. 117.

The specimens in the collection of the Carnegie Museum are listed as follows:
No. 238, cf , South America, LeBoutelier Collection.
No. 239, South America, LeBoutelier Collection (head onl,y).
No. 240, South America, LeBoutelier Collection (head only).
No. 241, (f, South America, LeBouteher Collection.
No. 242, South America, LeBoutelier Collection (head only).
No. 243, cf , South America, LeBoutelier Collection.
No. 1782, 9 , Bonda, Columbia, H. H. Smith coll., (Skin).
No. 1859, cf , South America, J. H. Smith coll.
No. 2022, cf , Cacaguahto, Colombia, Mrs. H. H. Smith.

Counts and Measurements.

Anal

Scale-rows

Gastrosteges

Urosteges

Total length in mm. .
Length of tail in mm.

(No. 238.) (No. 241.) | (No. 243.) (No. 1782.) (No. 1859.) (No. 2022.)

1
95

237
53

542
55

1
85

233
60

514
56

1

85
243

54
576

60

235

51

1,600

175

1

85

239

62

578
64

246
57

727
90

Genus Epicrates Wagler.
6. Epicrates cenchria (Linnaeus).

Boa cenchria Linn^us, Mus. Ad. Frid., II, 1764, p. 41; Syst. Nat. Ed. XII, 1766,

I, p. 374.
Epicrates cenchris Boulenger, Cat. Snakes, I, 1893, p. 94.
No. 26, cf , Las Juntas, Bolivia, Elev. 50 M., Jose Steinbach coll., Dec. 1913.
No. 115, cf, Santa Cruz de !a Sierra, Jose Steinbach coll.
No. 136, d", Minca, Colombia, Elev. 2,000 ft., H. H. Smith coll., June.
No. 320, cf , Urucum Mts., Matto Grosso, Brazil, Haseman coll., May 2, 1909.

Counts and Measurements.

(No. 320.)

Anal

Scale-rows

Gastrosteges

Urosteges

Total length in mm. .
Length of tail in mm.

griffin: ophidia from south America in carnegie museum

169

6rt. Epicrates cenchria var. fusca (Gray).

Cliftia fusca Gray, Catalogue, 1849, p. 99.

Epicrates cenchris, var. B, Boulenger, Cat. Snakes, I, 1893, p. 96.

No.

Sex.

Locality.

Elevation.

Collector.

■i

a
<

^2

If

O =0

Uro-

steges.

Total
Length
in Mm.

Length

Tail in

Mm.

1861
2025

cT

S. America

Cacagualito, Colombia.. .

1,500 ft.

H. H. Smith
Mrs. H. H. Smith

1
1

50
49

247
247

50
49

1,790

160

Genus Eunectes Wagler.

7. Eunectes murinus (Linnseus).

Boa murina Linn^us, Syst. Nat., Ed. XII, 1766, I, p. 374.
Eunectes murinus Boulenger, Cat. Snakes, I, 1893, p. 115.

No.

Sex.

Locality.

Collector.

"3
c
<

02 >-

O CO

If

■ X

O OJ

(2

Total
Length
in Mm.

Length

Tail in

Mm.

217

9

S. America

LeBoutelier

54

239

69

17

2

725

98

218

9

60

71

17

3

685

98

219

cf

58

248

63

16

2

700

89

220

cf

63

247

68

16

4

68V

95

221

9

62

58

15

3

590

74

222

cf

62

248

66

17

3

5V0

76

223

cf

,58

65

15

3

600

82

224

9

54

241

67

16

2

707

101

225

&

58

240

68

17

3

723

102

226

cf

65

247

69

16

4

617

83

228

cf

iC

58

244

70

16

3

728

101

229

cf

ii

62

242

70

17

J-

3

728 100

230

9

''

56

254

64

16

3

700 : 97

231

9

n

57

246

67

17

2

725

97

232

9

it

66

255

67

16

3

620

77

233

9

li

60

244

67

16

2

716

100

234

cf

((

62

246

68

18

4

698 99

235

cf

ii

60

242

69

15

2

702 ' 101

236

&

a

60

245

69

17

2

720 , 105

237

9

Chile (?)

\ 1

60

240

66

16

1 1 3

706 1 91

8. Eunectes notaeus Cope.

Eunectes nota;us Cope, Proc. Acad. Nat. Sci. Philada., 1862, p. 70.
Eunectes notaeus Boulenger, Cat. Snakes, III, 1896, p. 594.

The only specimen of this species at present in the collection (No. 379) was
taken by Mr. John D. Haseman near Santarem, in a swamp between the Rio
Amazonas and Tapajos, on Dec. 7, 1909. When captured the snake was found
to have a bird in its stomach. It is a female.

170 MEMOIRS OF THE CARNEGIE MUSEUM

Counts and Measurements.

Anal 1

Scale-rows 48

Gastrosteges 227

Urosteges 49+

Upper labials 14

Preociilars 1

Postoculars 3

Total length in mm 1,280

Length of tail in mm 143

Family COLUBRID.E Boulenger.

Series A. Aglypha.

Subfamily Colubrin^e Boulenger.

Genus Aporophis Cope.

9. Aporophis flavifrenatus (Cope).

Lygophis flavifrenatus Cope, Proc. Acad. Nat. Sci. Philada., 1862, p. 80.
Aporophis flavifrenatus Boulenger, Cat. Snakes, II, 1894, p. 158.

This species is represented in the collection by a single individual (No. 363)
taken by J. D. Haseman on February 1, 1909, at Cacequy, Rio Grande do Sul.
It is a female.

Counts and Measurements.

Anal _ 1/1

Scale-rows 17

Gastrosteges 156

Urosteges 77/77

Upper labials 8(4.5)

Preoculars 1

Postoculars 2

Temporals 1,2

Total length in mm 638

Length of tail in mm 162

10. Aporophis lineatus (Linnaeus).

Coluber lineatus Linn^us, Mus. Ad. Frid., 1754, p. 30, PL XII, fig. 1, and PL XX,

fig. 1; Syst. Nat., Ed. XII, 1766, I, p. 582.
Aporophis lineatus Boulenger, Cat. Snakes, II, 1894, p. 158.

There are at hand two specimens of this species, one (No. 5) taken in the
Province del Sara, Bolivia, by Jose Steinbach in January, 1912, the other (No. 346)
taken by Haseman at Bom Jesus de Lapa, Bahia, Brazil. Both specimens are
females.

griffin: ophidia from south America in carnegie museum 171

Counts and Measurements.

(^fo. 5.) (No. 346.)

Anal 1/1 1/1

Scale-rows 19 19

Gastrosteges 165 171

Urosteges 77/77 71/71

Upper labials 8(4.5) 8(4.5)

Preoculars 1 1

Postoculars 2 2

Temporals 1,2 1,2

Total length in mm 675 300

Length of tail in nam 178 68

11. Aporophis melanocephalus sp. nov.

Ma.xillary teeth about twenty, followed after a considerable interspace by two
enlarged, compressed teeth below the posterior border of the eye; mandibular teeth
subequal. Head narrow, not much wider than neck; snout narrow, high; eye large,
pupil round. Body cylindrical, ventrals rounded; scales smooth, without pits, in
fifteen rows. Rostral broader than deep, visible from above; internasals as long as
wide, almost as long as the prefrontals; frontal once and a half as long as broad, sides
straight and nearly parallel, longer than its distance from the end of the snout,
considerably shorter than the parietals; nostril between two nasals; loreal consider-
ably deeper than long; one preocular; two postoculars; temporals 1, 2; eight upper
labials, the fourth and fifth bordering the eye; five lower labials in contact with the
anterior chin-shields, which are as long as the posterior.

Back and sides of body and tail uniform dark brown; each scale has a dark
brown center and light brown edge. The upper surface of the head is nearly black,
this color shading into the brown of the body on the neck. A black vertebral
stripe five scale-rows wide joins the black of the occiput. A black lateral stripe
partly covering the second and third rows of scales extends from the temples along
the neck. These three stripes quickly merge into the brown body color. Close
to the head they are separated by white lines one scale-row wide, which become
darker as they are followed eaudad, and also soon merge into the color of the body.
There is a small white spot back of the eye on the outer part of the parietal, and a
similar spot in front of the ej^e on the lateral part of the prefrontal. The upper
lip and ventral surfaces of the head and body are white, without dark markings.

The type, No. 18, of the Catalog of the Reptilia in the Carnegie Museum, is a
female, taken at Las Juntas, Bolivia, 250 M. above sea-level, by Jose Steinbach
in December, 1913. It is unique.

172 MEMOIRS OF THE CARNEGIE MUSEUM

Counts and Measurements.

Anal 1/1

Scale-rows 15

Gastrosteges 144

Urosteges 72/72

Upper labials 8(4.5)

Preoculars 1

Postoculars 2

Temporals 1,2

Total length in mm 291

Length of tail in mm 47

Genus Atractus Wagler.

12. Atractus badius (Boie).

Brachyorrhos badius Boie, Isis, 1827, p. 540.

Atractus badius Boulenger, Cat. Snakes, II, 1894, p. 308.

The two specimens at hand were taken by H. H. Smith, one, No. 201, a male,
at Minca at an elevation of 2,000 ft. above sea-level, the other, No. 215, a female, at
El Libano at an elevation of 6,000 ft. Both localities are in the Province of Santa
INIarta, Colombia. The female specimen was taken in Maj^, the male in July.
The latter specimen has lost the tip of the tail.

Counts and Measueements.

(No. 201.) (No. 215.)

Anal 1 1

Scale-rows 17 17

Gastrosteges 152 146

Urosteges 30/30-1- 28/28

Upper labials 7(3.4) 7(3.4)

Preoculars 0 0

Postoculars 2 2

Temporals 1,2 .1,2

Total length in mm 270-|- 442

Length of tail in mm 32 49

No. 201 is uniform dark brown above, with about forty pairs of small light
spots making a row on either side of the back. The belly is nearly covered with
dark brown, square, or oblong, spots. The frontal shield is longer than wide.

No. 215 is reddish brown above, with transverse black spots which show a
tendency to produce bars, but are too irregular to actually do so. Each scale of
the ventral surface carries two lateral and one or two median large, black spots,
the whole forming two definite lateral stripes and a less definite median one.

griffin: ophidia from south America in carnegie museum

173

13. Atractus taeniatus^ sp. nov. (Plate XXVIII, figs. 1-3.)

Eight maxillary teeth. Symphysials separated from the chin-shields by the
first lower labials. Snout broad, rounded; rostral considerably broader than deep,
the part visible from above as long as the suture between the internasals; inter-
nasals very small, pentagonal, the combined width of the internasals and posterior
nasals (with which the internasals are in contact) being just equal to the width of
the prefrontals; prefrontals as broad as long, entering the orbit; frontal once and
a quarter as broad as long, longer than the prefrontals, as long as its distance from
the end of the snout, much shorter than the parietals; loreal twice as long as deep,
entering the orbit; two postoculars, the lower very small, both in contact with
the anterior end of the parietal; temporals narrow, 1, 2; six upper labials, the third
and fourth entering the orbit, the third to the last once and one-half to twice as
long as deep; four lower labials in contact with each of the single pair of chin-shields,
which are twice as long as broad.

Dark brown above, with a darker vertebral stripe; a light bar across the
temporals and parietals; edge of upper lip light (white?); white below, chin spotted
with brown, and a large brown spot occupying the center of each gastrostege, form-
ing a midventral brown stripe; urosteges irregularly spotted with brown.

This species resembles A. roulei Despax; the two are compared below.

A. tmniatus.

Rostral wider than deep, portion visible from above
considerable.

Internasals pentagonal.

Prefrontals as long as broad, shorter tlian the fron-
tal.

Frontal once and a quarter as broad as long.

Loreal twice as long as deep.

Two postoculars.

Pupil round.

Four lower labials in contact with chin-shields.

Six upper labials.

Scales in fifteen rows.

Light band across head.

Urosteges lightly spotted with brown.

Scale formula, 1-1&-152-24/24

A. roulei.
Rostral as wide as deep, little visible from above.

Internasals subtriangular.

Prefrontals almost as wide as long, as long as the

frontal.
Frontal once and a half as broad as long.
Loreal twice as long as deep.
One postocular.
Pupil subelliptic.

Thi-ee lower labials in contact wth cliin-shields.
Six upper labials.
Scales in fifteen rows.
No hght band across head.
Urosteges uniformly brown.
Scale formula, 1-1.5-1.54-22/22

The type, which is unique, is No. 117 of the Catalog of Reptiles in the Carnegie

^ Tama = a fillet, or head-band.

174

MEMOIRS OF THE CARNEGIE MUSEUM

Museum. It was taken by Jose Steinbach near Santa Cruz de la Sierra, Bolivia.
The collector failed on his label to indicate the date of capture. It is a male.

Counts and Measurements.

Anal 1

Scale-rows 15

Gastrosteges 152

Urosteges 24/24

Upper labials 6

Preoculars 0

Postoculars 2

Temporals 1, 2

Total length in mm 218

Length of tail in mm 22

Genus Elaphe Fitzinger.
14. Elaphe corals (Boie).

Coluber corals Boie, Isis, 1827, p. 537.

Coluber corais Boulenger, Cat. Snakes, II, 1894, p. 31.

The Carnegie Museum possesses three examples of this species (Nos. 143, 144,
and 149) all of which were captured by H. H. Smith, in the month of June at Bonda,
Province of Santa Marta, Colombia. No. 144 is a female. No. 149 is a male; the
sex of No. 143 I have been unable to determine.

Counts and Measurements.

Anal

Scale-rows

Gastrosteges

Urosteges

Upper labials

Preoculars

Postoculars

Temporals

Total length in mm.
Length of tail in mm

(No. 143.)

(No. 144.)

1

1

17

17

212

214

81/81

82/82

8(4.5)

9(5.6)

1

1

2

2

2,2

2,2

1,762

1,582

375

353

(No. 149.)

1

17
203

88/88

8(4.5)

1

2

2,2

1,832

422

15. Elaphe dichroa (Peters).

Herpetodryas dichroa Peters, Mon. Berl. Akad., 1863, p. 284.
Coluber dichrous Boulenger, Cat. Snakes, II, 1894, p. 30.

The only example of this species at present in the Museum is No. 329, which
was taken by J. D. Haseman at Entre Rios, Brazil, on June 4, 1908. It is a female.

griffin: ophidia from south America in carnegie museum 175

Counts and Measurements.

Anal 1/1

Scale-rows 15

Gastrosteges 181

Urosteges 96/96

Upper labials 8(3.4.5)

Preoculars 1

Postoculars 2

Temporals 2,2

Total length in nun 402

Length of tail in nim 100

Genus Dimades Gray.

16. Dimades plicatilis (Linnaeus).

Coluber plicatilis Linn^us, Mus. Ad. Frid., 1754, p. 23, PI. VI, fig. 1; Syst. Nat.,

Ed. XII, 1766, I, p. 376.
Dimades plicatilis Boulenger, Cat. Snalves, II, 1894, p. 186.

The single specimen we have is No. 378 of the Catalog of Reptiles in the
Carnegie Museum. It was taken by J. D. Haseman at Santarem, Brazil, on
December 7, 1909. It is a female, and was captured under a log in the forested
lowlands between the Amazon and Tapajos rivers in the act of brooding forty-
eight eggs, five of which together with one of the unhatched young were collected
by Mr. Haseman and are preserved as C. M. No. 354. Mr. Haseman states that
"Sucury" and "Sucurujaba" are the local names of this serpent.

The shells of the eggs are membranous, thin, and white. The eggs measure
35 mm. in length and 25 mm. in width. Considering the advanced stage of de-
velopment of the young, there is still a large amount of yolk in the eggs. These
eggs seem to have passed through about half of the period of incubation. The
unhatched j'oung is 160 mm. long; the length of its tail is 35 mm. The central
portion of the back, three scales wide, is light brown; on each side is a dark brown
stripe covering the third scale row and the contiguous half of the scale row on
either side. A row of small dark dots lies on each side of the vertebral stripe. The
intervals between the dorsal and lateral bands, and the ventral surface, are white.
The adult female has well-defined lateral dark stripes on the same scale rows as
the young, but the dorsal surface between these is uniformly colored except for
the two rows of small dark brown spots. The ventral surface of the adult bears
four rows of large brown dots.

I

176 MEMOIRS OF THE CARNEGIE MUSEUM

Counts? and Measurements.

Anal 1/1

Scale-rows 15

Gastrosteges 139

Urosteges 37/37

Upper labials 7(3)

8(3.4)

Preoculars ' 1

Postoculars 2

Temporals 1,1

Total length in mm 1,373

Length of tail in mm 194

Genus Coronella Laurenti.

17. Coronella micropholis (Cope).

Lampropeltis micro'pJiolis Cope, Proc. Acad. Nat. Sci. Philada., 1860, p. 257.
Coronella micropholis Boulenger, Cat. Snakes, II, 1894, p. 203.

The specimen, which is a female (No. 2036), was taken by Mrs. H. H. Smith
at CacaguaUto, Colombia.

Counts and Measurements.

Anal 1

Scale-rows 23

Gastrosteges 228

Urosteges 43/43

Upper labials 7(3.4)

Preoculars 1

Postoculars 2

Temporals 2, 2

Total length in mm 732

Length of tail in mm 128

Genus Drymobius Cope.

18. Drymobius bifossatus (Raddi).

Coluber bifossatus Raddi, Mem. Soc. Ital. Modena, XVIII, (Fis.) 1820, p. 333.
Drymobius bifossatus Boulenger, Cat. Snakes, II, 1894, p. 10.

The collection possesses two specimens:

No. 127, 9 . From Santa Cruz de la Sierra, Bohvia. (Steinbach coll.)
No. 381, cf . From Rio de Janeiro, Brazil. (Haseman coll., July 1, 1908.)
The color of the skin of No. 127 is almost uniform.

griffin: ophidia from south America in carnegie museum 177

Counts and Measurements.

(No. 127.) (No. 381.)

Anal 1/1 1/1

Scale-rows 15 15 •

Gastrosteges 179 167

Urosteges 97/97

Upper labials 8(4.5) 8(4.5)

Preoculars 1 1

Postoculars 2 2

Temporals 2, 2 2, 2

Total length in mm 1.638

Length of tail in mm 455

19. Drymobius boddaerti (Sentzen).

Coluber boddaerti Sentzen, Meyer's Zool. Archiv, II, 1796, p. 59.
Drymobius boddaerti Boulenger, Cat. Snakes, II, 1894, p. 11.

There are twenty specimens in the collection, all of which are presumed to
have come from Colombia, and to have been collected by Mr. and Mrs. H. H. Smith,
except one (No. 29) which was taken by Jose Steinbach at Las Juntas, Bolivia.
Nos. 176-183 inc., and No. 186 are from Bonda, where they were captured by the
Smiths in 1901 at dates ranging from May to September. Nos. 202-203 were
taken by the same parties at Valparaiso, Colombia (elev. 4,500 ft.); Nos. 2006,
2009, and 2020 at Cacagualito, Colombia (elev., 1,500 ft.) by Mrs. H. H. Smith.
Nos. 1866-1868 inclusive, and Nos. 1872-1873 are a part of a collection, which, as
stated in the introductory pages of this paper, had no field data attached to it, but
are known to have been sent in by Mr. H. H. Smith while laboring in Colombia.

Both anterior and posterior temporals are frequently divided into two or three
scales in longitudinal series.

Nos. 202 and 203 are young specimens of the variety rappii. The prevailing
color of the upper surface is dark brown. Numerous white bars cross the back,
alternating with a series of shorter white bars on each side which reach the ventral
surface. The labials are black-margined; the chin and throat are checkered with
black and white. No. 2006 is uniform olive-green above, shading to brown on the
tail; the lower surface is white, with faint cloudy markings on the chin and throat.
No. 2020 is brown above, with faintly defined darker stripes along the middle of
the back and each side. The vertebral band is outlined by a black line in the
posterior part of the body. The lateral stripes become darker and more distinct
on the tail.

178

MEMOIRS OF THE CARNEGIE MUSEUM

Counts and Measxjeements.

Number.

No. 176.
177.
178.
179.
180.

181.

182.

183.

186.
29.

202.

203.
1866.
1867.
1868.
1872.
1873.
2006.
2009.
2020.

Anal.

Scale-
rows.

17
17
17

17
17

17
17
17
17
17
17
17
17
17
17
17
17
17
17
17

Gastro-
steges.

198
172
189
191

188

174
175
188
186
193
179
194
180
187
188
198
196
189
189
185

Urosteges.

120/120
97/97
96/96
86/86
94/94

100/100

95/95

118/118

88/88

53/53

55/55

118/118

82/82

94/94

94/94

121/121

119/119

118/118

'93/93'

Upper Labials.

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

9(4.5.6)

Preoc-
ulars.

Postoc-
ulars.

2
2

2
2
2

2
2
2
2
2
2
2
2
2
2
2
2
2
2
2

Tem-
porals.

2,2
2,2
2,2
2,2
2,2

2,2
2,2
2,2
2,2
2,2
2,2
2,2
2,2
2,2
2,2
2,2
2,2
2,2
2,2
2,2

Total
Length
in Mm.

1,364

916
1,140
1,180
1,165

925

1,023

1,222

975

802

384

505

985

1,005

1,030

1,133

1,270

826

1,020

1,110

Length

Tail
in Mm.

402
268
305
288
307*

265
287
367
242
137*

67*
152
240
266
307
343
372
250

93*
287

Sex.

9
cf
9
9
9

&
d'
&
9
9
cf

9
9
9
9
9
&
cf
9

* Tip of tail lost.

20. Drymobius rhombifer (Giinther).

CorypJwdon rhombifer Guenther, P. Z. S., 1860, p. 236.
Drymobius rhotnbifer Boulenger, Cat. Snakes, II, 1894, p. 14.

The collection contains four specimens as follows:

No. 188, 9 , Minca, Colombia, 2,000 ft. (H. H. Smith coll.), July. Skin.

No. 213, 9 , Bonda, Colombia, 150 ft. (H. H. Smith coll.), June.

No. 1865, 9 , Colombia, S. America (H. H. Smith coll.).

No. 1871, cf , Colombia, S. America (H. H. Smith coll.). Tip of tail lost.

No. 213 has, in addition to the ordinarj^ markings, a brown spot in the center
of each gastrostege, all together making a midventral stripe. Each of these speci-
mens possesses a greater number of urosteges than is given for the species in
Boulenger's Catalog, i. e. 84 to 96.

Counts and Measxjeements.

Number.

Anal.

Scale-
rows.

Gastro-

steges.

Urosteges.

Upper
Labials.

Preocu-
lars.

Postocu-
lars.

Tempo-
rals.

Total
Length
in Mm.

Length

Tail in

Mm.

188
213

1865
1871

1/1

1/1
1/1
1/1

17
17
17
17

151
160
157
152

100/100
102/102
100/100
1 97/97*

9(4..5.6)
9(4.5.6)
9(4.5.6)
9(4.5.6)

1
1

1
1

2 2, 2
2 2,2
2 2,2
2 ,2,2

474

375

1,272

130
103
385

■ Tip of tail lost.

I

griffin: ophidia from south America in carnegie museum 179

Genus Helicops Wagler.

21. Helicops angulata (Linnaeus).

Coluber angulatus Linn^us, Mus. Ad. Frid., 1754, p. 23, PI. XV, fig. 1; Syst. Nat.,

Ed. XII, 1766, I, p. 377.
Helicops angulatus Boulenger, Cat. Snakes, I, 1893, p. 278.

The only specimen in the Museum (No. 279) is a male from the LeBoutelier
collection, which simply bears the label "South America."

COTTNTS AND MEASUREMENTS.

Anal 1/1

Scale-rows 19

Gastrosteges 117

Urosteges 74/74

Upper labials 9(4)

8(4)

Preoculars 1

Postoculars 2

Temporals 2, 4

Total length in mm 300

Length of tail in mm 88

22. Helicops carinicauda (Wied) var. infrataeniata Jan.

Coluber carinicaudus Wied, Beitr. Naturgesch. Bras., I, 1825, p. 300, Plate.
Helicops infratceniatus Jan, Arch. Zool. Anat. Phys., Ill, 1865, p. 245 and Icon.

Gen., 28, PI. Ill, fig. 3 (1868).
Helicops carinicauda, var. B, Boulenger, Cat. Snakes, I, 1893, p. 277.
No. 355, cf , Caceqity, Rio Grande do Sul (Haseman coll.), Jan. 31, 1909.

Counts and Measurements.

Anal 1/1

Scale-rows 19 ~

Gastrosteges 128

Urosteges 61/61

Upper labials 8(3.4)

8(4)

Preoculars 1

Postoculars 2

Temporals 1,2

Total length in unn 543

Length of tail in mm 128

180

MEMOIRS OF THE CARNEGIE MUSEUM

23. Helicops leopardina (Schlegel).

Honialopsis leopardina Schlegel, Phys. Seri3., II, 1837, p. 358.
Helicops leopardinus Boulenger, Cat. Snakes, I, 1893, p. 278.

The Museum possesses eleven examples of this species, as follows:

No. 25, cf . Las Juntas, Bolivia, coll. J. Steinbach, December, 1913.

No. 321, c?. Puerto Suarez, Bolivia, coll. J. D. Haseman, May 7, 1909.

No. 322, 9 . Sao Antonio de Guapore, coll. J. D. Haseman, July 4, 1909.

No. 323, c?. Puerto Suarez, Bolivia, coll. J. Steinbach, Jan., 1909.

No. 324, c?. Aregua, Paraguay, coll. J. D. Haseman, April 7, 1909.

No. 325, cf . Puerto Suarez, Bolivia, coll. J. D. Haseman, May 6, 1909.

No. 326, cf . Puerto Suarez, Bolivia, coll. J. D. Haseman, May 6, 1909.

No. 327, 9 . Sao Luiz de Caceres, coll. J. D. Haseman, May 25, 1909.

No. 338, cf . Santarem, Brazil, coll. J. D. Haseman, Dec. 7, 1909.

No. 357, cf . Santarem, Brazil, coll. J. D. Haseman, December 15, 1909.

No. 358, cf . Santarem, Brazil, coll. J. D. Haseman, December 15, 1909.

The hypapophyses of the posterior trunk vertebrae of H. leopardina are vari-
ably developed. Some specimens show scarcely a keel, some a well-developed
plate, some an intermediate condition. No. 338 has no loreal shields; the loreal
is absent on the left side of No. 357, and the right nasal of this specimen is in con-
tact with the preocular.

Counts and Measurements.

^■°

No. 321 . .

" 322..
" 323 . .
" 324..

" 325..
" 326..
" 327 . .

" 338. .

" 357..

" 358..

" 25...

a

1/1

1/1
1/1
1/1

1/1
1/1
1/1

1/1

1/1

1/1

1/1

4> n

u o

CO '^

19

19
19
19

19
21
19

19

19

19

19

o »

114

115
112
119

116
112
119

120

117

115

122

O I*

74/74

62/62
73/73
64/64

68/68
64/64

72/72

87/87
89/89
72/72
83/83

&.2
P<.o

^^

8(4)

8(3.4)

8(4)

8(4)

8(4)
8(4)
8(4)

8(3.4)

8(4)

8(4)

8(4)

Ph 5

griffin: ophidia from south America in carnegie museum

181

24. Helicops modesta Glinther.

Helicops modesius Guenther, Ann. & Mag. Nat. Hist. (3), VII, 1861, p. 425.
Helicops modestus Boulenger, Cat. Snakes, I, 1893, p. 277.

The specimens at hand are listed as follows :
No. 335, cf , collected by J. D. Haseman at Jacarehy, Sao Paulo, Brazil, July 15,

1908.
No. 336, cf , collected by J. D. Haseman at Jacarehy, Sao Paulo, Brazil, July 15,

1908.
No. 348, 9 , collected by J. D. Haseman at Mogy das Cruzes, Sao Paulo, July 20,

1908.
No. 349, cf , collected by J. D. Haseman at Mogy das Cruzes, Sao Paulo, July 20,

1908.

Nos. 348 and 349 are uniform black above, with no traces of stripes. The
scales of the anterior half of the body scarcely show traces of keels, which are poorly
developed in the posterior half. These two specimens were captured with a
seine in the Rio Tiete. Nos. 335 and 336 are reddish brown above, with three
darker stripes. They were taken in a pond near the Rio Parahyba.

No. 348 has no traces of hypapophyses on the posterior trunk vertebrae; the
hypapophyses of these vertebrae of No. 349 are only low keels.

CotJNTS AND Measurements.

Anal

Scale-rows . . . .
Gastrosteges .
Urosteges . . . .
Upper labials .
Preoculars. . . .

Postoculars.
Temporals. .

Total length in mm. .
Length of tail in nun .

(No. 335.)

(No. 336.)

(No. 348.)

(No. 349.)

1/1

1/1

1/1

1/1

19

19

19

19

121

125

120

125

70/70

54/54

59/59

55/55

8(3.4)

8(4)

8(4)

8(4)

1

1

1

fl

12

2

2

2

2

[2,2
12,3

2,2

2,2

(2,3
12,2

381

285

316

587

111

65

73

121-

25. Helicops polylepis Glinther.

Helicops polylepis Guenther, Ann. Mag. N. H. (3), VII, 1861, p. 426.
Helicops polylepis Boulenger, Cat. Snakes, I, 1893, p. 280.
No. 319, cf , Santarem, Brazil, J. D. Haseman coll., December 12, 1909.
No. 328, cf , Sao Luiz de Caceres, J. D. Haseman coll.. May 25, 1909.
No. 360, cf , Santarem, Brazil, J. D. Haseman coll., December 9, 1909.
No. 2038, cf , Sao Antonio de Guapore, J. D. Haseman coll., July 30, 1909.

182

MEMOIRS OF THE CARNEGIE MUSEUM

Specimen No. 2038 was taken in the Rio Guapore, and according to Haseman's
note is called locally "Cobra de Agua." It presents some variations from the
description given by Boulengcr. The rostral is separated from the internasal by
the nasal. The frontal is once and two-thirds as long as broad, as wide in front
as behind, with perfectly straight and parallel sides. It is shorter than its distance
from the end of the snout, and shorter than the parietals. The anterior chin-
shields are longer and considerably larger than the posterior. There are no keels
on the scales of the outer row except in the posterior fourth of the trunk, and on
the tail. The color of the lower surface is dark brown, with only a few widely
separated small yellow spots. This appears to be the largest specimen of the
species which has been recorded.

Specimens 319, 328, and 360 agree more nearly, but not entirely, with Bou-
lenger's description. The eye of all is longer than one-half the frontal shield; this
character is either more variable than usual or depends to some extent upon the
treatment of the specimen. The lower surfaces of Nos. 319 and 360 are almost
black, with round white (originally yellow?) spots on one or both extremities of
most of the gastrosteges, forming conspicuous rows. The lower surface of No. 328
presents only a few small and indistinct light spots. The dorsal surface of this
specimen is darker than that of any of the others, with more distinct black markings.
The three middle rows of dorsal spots tend in all the specimens to run into each
other and produce a pattern of diagonal crossing lines.

Hypapophyses are not present at all on the posterior trunk vertebrae of two
of these specimens, and are scarcely indicated in the others.

No. 319 was found under a barrel on the bank of the Amazon. No. 360 was
caught in a fish net in the Amazon. A small fish was found in the stomach of
No. 328.

Counts and Measueements.

(No. 319.)

(No. 328.:

(No. 360.)

(No. 2038.)

Anal

Scale-rows

Gastrosteges

Urosteges

Upper labials

Preoculars

Postoculars

Temporals

Total length in mm. . . .
Length of tail in mm. . .

* Tip of tail lost,

1/1

23

131

71/71
8(4)

1

2

1,2

309
83

1/1

25

119

70/70
8(4)

1

2
1,2

344
96

1/1

23

130

79/79

8(3.4)

8(4)

1

2

Jl,2

12,4

407

116

1/1

25

122

28/28*

8(4)

1

2
2,3

940
127*

griffin: ophidia from south America in carnegie museum

183

Genus Herpetodryas Boie.

26. Herpetodryas carina tus (Linnaeus).

Coluber carinaius Linn^us, Mus. Ad. Frid., 1754, p. 31; Syst. Nat., Ed. XII,

1766, I, p. 384.
Herpetodryas carinatus Boulenger, Cat. Snakes, II, 1894, p. 73.
No. 2, cf , Province del Sara, Bolivia, Steinbach coll. (elev. 400 M.), Sept., 1911.
No. 3, 9 , Province del Sara, Bolivia, Steinbach coll. (elev. 350 M.), Jan., 1912.
No. 150, cf , Bonda, Colombia, H. H. Smith coll. (elev. 150 ft.), June.
No. 367, c?. Villa Bella, Bolivia, J. D. Haseman coll., Oct. 7, 1909.

Nos. 2 and 3 belong to the variety flavolineatus Jan. Nos. 150 and 367 are of the

variety bicarinatus (Wied).

Counts and Measurements.

Anal

Scale-rows. . . .
Gastrosteges. .
Urosteges. . . .
Upper labials .

Preoculars. .

Postoculars.
Temporals. .

Total length in mm. .
Length of tail in mm.

(No. 2.)

1/1
12

153
130/130
9(4.5.6)

1

2

1,2

1067
405

(No. 3.:

1/1
12

155
120/120
9(4.5.6)

1

2

1,1

937
366

(No. 150.)

1/1

12

157

126/126

I 8(4.5)

1 9(5.6)

1

2

1,2

1605
521

(No. 367.1

1/1

12

152

136/136
9(4.5.6)

1
2

I 1,2

11,1

1330

455

27. Herpetodryas fuscus (Linnajus).
Coluber fuscus Linn^us, Mus. Ad. Frid., 1754, p. 32, PI. XVII, fig. 1; Syst. Nat.,

Ed. XII, 1766, I, p. 383.
Herpetodryas fuscus Boulenger, Cat. Snakes, II, 1894, p. 75.
No. 365, cf, Villa Bella, Bolivia, J. D. Haseman coll., October 8, 1909.
No. 1428, cf, "Massasao, River Amazon" fide J. 0. Kerby, who presented it to the

Museum.

No. 1428 belongs to the variety saturninus (Linnaeus).

Counts and Measurements.

(No. 365.)

Anal 1

Scale-rows 10

Gastrosteges 150

Urosteges 129/129

Upper labials 9(4.5.6)

Preoculars 1

Postoculars 2

Temporals 1,1

Total length in mm 702

Length of tail in mm 350

(No. 1428.)
1

10

156

120/120

9(4.5.6)

1

2

1,1

385
128

184

MEMOIRS OF THE CARNEGIE MUSEUM

Genus Leptophis Bell.

28. Leptophis ahaetulla (Linnscus).

Coluber ahceiulla Linn^us, Sjst. Nat., Ed. XII, 1766, I, p. 387.
Leptophis liocercus Boulenger, Cat. Snakes, II, 1894, p. 113.

The Carnegie Museum possesses four specimens.
No. 23, 9 , Las Juntas, Bolivia, 250 M., J. Steinbach coll., December, 1913.
No. 268, 9 , South America, LeBoutelier coll.
No. 314, cf, Puerto Suarez, J. Steinbach coll., December; 1908.
No. 366, cf , Villa Bella, Bolivia, J. D. Haseman coll., October 10, 1909.

Nos. 23, 314, and 366 have lost the tips of their tails. A right loreal is present
in the case of No. 268. No. 366 was found to have the intestines stuffed with
winged termites, and the stomach of No. 314 contained four small frogs.

Counts and Measukbments.

(No. 23.)

(No. 268.)

(No. 314.)

(No. 366.)

Anal

Scale-rows. . . .
Gastrosteges. .
Urosteges. . . .
Upper labials .
Preoculars . . . .
Postoculars. . .

Temporals

Total length in mm. .
Length of tail in mm.

1/1

15

166

129/129*

8(4.5)

(l

1,2
846
430*

1/1

15

160

145/145

9(5.6)

1

2

1,2

1170

437

1/1

15

166

8(4.5)
1
2

1,2
1100
150*

1/1

15

158

145/145*

9(5.6)

1

2

1,2
1338
514*

* End of tail lost.

29. Leptophis bocourti Boulenger.
Leptophis bocourti Boulenger, P. Z. S., 1898, p. 116.

This species is represented by but a single specimen, No. 2011, which was
collected by Mrs. H. H. Smith at Cacagualito, Colombia (elev. 1,500 ft.). It is
a female.

The specimen agrees well with Boulenger's description, except in the greater
number of gastrosteges and the smaller number of urosteges. The terminal shield,
or spine, of the tail is formed normally, yet the tip of the tail is thicker than is
usual in the genus. It is therefore possible that the tip has been amputated and
healed, and that the reduced number of urosteges of this specimen is thus accounted
for. The color is a uniform, very dark green on the upper surface, without the
dots and lines which Boulenger describes; light green below, except the lower jaw,
which is yellow, with a green border covering about half of the lower labials.

griffin: ophidia from south America in carnegie museum 185

Counts and Measurements.

Anal 1/1

Scale-rows 15

Gastrosteges 179

Urosteges 103/103

Upper labials 8(4.5)

Preoculars 1

Postoculars 2

Temporals 1, 2

Total length in mm 1212

Length of tail in mm 355

30. Leptophis nigromarginatus (Giinther).
Ahcetulla nigromarginata Guenther, Ann. & Mag. Nat. Hist. (3), XVIII, 1866, p.

28.
Leptophis nigromarginatus Boulenger, Cat. Snakes, II, 1894, p. 112, PI. Ill, fig. 3.

There are two females of this species in the collection, which were received
from Mr. J. O. Kerby, and have attached to them labels which indicate that they
were taken at Pranquina on the Amazon River. They bear the Catalog Numbers
2007 and 2008. The tip of the tail of No. 2007 has been lost.

CotTNTS AND Measurements.

(No. 2007.) (No. 2008.)

Anal 1/1 1/1

Scale-rows 15 15

Gastrosteges 158 162

Urosteges 111/111 157/157

Upper labials 8(4.5) 9(5.6)

8(4.5)

Preoculars 1 1

Postoculars 2 2

Temporals 1,1 1, 1

1,2 1, 2

Total length in mm 940 1077

Length of tail in nmi 317 392

31. Leptophis occidentalis (Giinther). _

Ahcetulla occidentalis Guenther, P. Z. S., 1859, p. 412.

Leptophis occidentalis Boulenger, Cat. Snakes, II, 1894, p. Ill, PI. Ill, fig. 2.

The Carnegie Museum has five specimens of this species listed as follows :
No. 140, 9 , Bonda, Colombia, H. H. Smith coll., June.
No. 141, 9 , Bonda, Colombia, H. H. Smith coll., August.
No. 1095, 9 , Bonda, Colombia, H. H. Smith coll., August 25.
No. 1869, d", Colombia, H. H. Smith coll.
No. 1870, 9 , Colombia, H. H. Smith coU.

186

MEMOIRS OF THE CARNEGIE MUSEUM

Counts and Mbasuhements.

(No. 140.)

(No. 141.)

1/1

1/1

15

15

184

177

158/158

164/164

8(4.5)

8(4.5)

1

1

o

2

1, 2

1,2

1323

1486

460

575

(No. 1095.)

(No. 1869.)

(No. 1870.)

Anal

Scale-rows

Gastrosteges

Urosteges

Upper labials

Preoculars

Postoculars

Temporals

Total length ia mm. .
Length of tail in mm,

1/1

15

176

146/146

. 8(4.5)

9(5.6)

1

2

1,2

1463

505

1/1

15

168

152/152

9(5.6)

1

2

1,2

1800

621

1/1
15

183

158/1.58

8(4.5)

1

2

1,2

1475

504

32. Leptophis rostralis Lonnberg.
Leptophis rostralis Lonnberg, Ann. & Mag. Nat. Hist. (7), X, 1902, p. 458.

No. 1862 is the only specimen as yet representing this species in the Carnegie
Museum. It is shghtly defective, having lost the end of the tail. It is a female,
and was undoubtedly taken by H. H. Smith in Colombia.

The specimen agrees with Lonnberg's descrijition in all the principal features,
though showing some variation. The principal divergence is that in our specimen
the preocular is in contact with the frontal. The portion of the rostral visible
from above equals one-half the length of the internasals. The internasals are
almost (five-sixths) as long as the prefrontals. The frontal is considerably longer
than its distance from the end of the snout, but shorter than the parietals. The
anterior chin-shields are shorter than the posterior. The color of the upper surface
is a uniform dark brown without any visible markings. The upper lip and the
lower surface of the head are whitish. The scales of the ventral surface appear to
have been dark green; the posterior margin of each is light. The specimen has
not been weU preserved to show color, and the loss of the data regarding it is un-
fortunate. The number of gastrosteges exceeds that of the type.

Counts and Measurements.

Anal 1/1

Scale-rows 15

Gastrosteges 180

Urosteges 86/86*

Upper labials 9(5.6)

8(4.5)

Preoculars 1

Postoculars 2

Temporals 1,2

Total length in mm 373

Length of tail in mm 94*

* Tip of tail lost.

I

griffin: ophidia from south America in carnegie museum 187

Genus Liophis Wagler.

33. Liophis albiventris (Jan).

Liophis regince, vars. albiventris and quadrilineata Jan, Arch. Zool. Anat. Phys., II,

1863, p. 294; Icon. Gen., 16, 1866, PI. VI, figs. 2 and 3.
Liophis albiventris Boulenger, Cat. Snakes, II, 1894, p. 130.

This species is represented by a single male specimen from the LeBoutelier
collection, which has been assigned the Catalog Number 278. The label attached
to it by the collector states that it came from Sipan, South America.

Counts and Measurements.
Anal 1/1

Scale-rows 17

Gastrosteges 153

Urosteges 74/74

Upper labials 8(4.5)

Preoculars 1

Postoculars 1

Temporals 1,2

Total length in mm 495

Length of tail in mm 127

34. Liophis ahnadensis (Wagler).
Natrix almadensis Wagler, in Spix, Serp. Bras., 1824, p. 30, PI. X, fig. 3.
Liophis almadensis Boulenger, Cat. Snakes, II, 1894, p. 134.
No. 112, cf , Santa Cruz de la Sierra, Bolivia, J. Steinbach collector.
No. 352, cf, Sao Joao del Rey, Minas Geraes, Brazil, J. D. Haseman coll., May 17,

1908.

Counts and Measurements.

(No. 112.) (No. 352.)

Anal 1/1 1/1

Scale-rows 19 19

Gastrosteges 158 161

Urosteges 60/60 70/70

Upper labials 8(4.5) 8(4.5)

Preoculars 1 1 ^

Postoculars 2 2

Temporals 1, 2 1,2

Total length in nun 352 485

Length of tail in mm 71 103

35. Liophis elaeoides sp. nov.
Tail 5f to 6^ times in total length. Eye large, its length being equal to the
distance from the front of the eye to the middle of the nostril; head broad and flat;
snout bluntly pointed. Rostral broader than deep, visible from above; internasals

188

MEMOIRS OF THE CARNEGIE MUSEUM

as long as broad or longer, as long as, or longer than, the prefrontals; frontal once
and a quarter to once and a half as long as broad, shorter than its distance from the
rostral (this is true of the mature specimens, the frontal of small specimens is longer
than its distance from the end of the snout) ; frontal as long as, or shorter than,
the parietals; loreal deeper than long; one pre- and two postoculars; temporals 1,2;
eight upper labials, the fourth and fifth bordering the orbit; five lower labials in
contact with the anterior chin-shields which are as long as, or longer than, the
posterior.

Scales in 19 rows: gastrosteges 160-167, obtusely angulate laterally; anal
divided; urosteges in 49-54 pairs.

Uniform dark green above, this color extending to the outer ends of the gas-
trosteges. During life the color is undoubtedly of an olive tone, due to the color
of the horny scales, most of which are rubbed off in our specimens. There are no
spots, nor light or dark edges to scales; neither have the young a dark nuchal band.
The lower surface is a uniform yellowish white. The upper lip is light green,
lighter than the upper part of the head, but not white. (Type, C. M. Cat. Rept.,
No. 32.)

The Museum has fourteen specimens of this species, all of which were collected
by Mr. Jose Steinbach in the Provincia del Sara, Bolivia, and most of them bear
the label Santa Cruz de la Sierra. They bear in the Catalog of Reptiles of the
Museum the numbers 32 (type), 44, 59, and 91-102 inclusive. All after No. 32
are regarded as paratjqjes. The type has a label giving the date of capture as
February, 1913. No. 59 is labelled as having been taken in October, 1911. No
dates are given on any of the other labels.

Counts and Measurements.

M

v

m

"3

a

s s

ii

if

Pi

09

Upper
Labials.

o

Total
Length
in Mm.

Length

Tail
in Mm.

No. 32....

&

19

163

49/49

8(4.5)

2

1,2

612

102

" 44....

9

19

164

50/50

8(4.5)

1^

2

1.2

400

69

' 59. . . .

9

19

160

49/49

8(4.5)

1,2

615

109

' 91....

rf-

19

162

50/50

8(4.5)

1 2

1,2

2.55

41

' 92. . . .

&

19

162

54/54

8(4.5)

1 ' 1

1,2

541

95

' 93. . . .

d'

19

165

54/54

7(4.5)

2

1,2

522

94

' 94. . .

cf

19

167

49/49

8(4.5)

2

1,6

680

106

' 95. . . .

cf

19

163

53/53

8(4.5)

2

1,2

315

50

' 96. ...

cf

19

164

49/49

8(4.5)

2

1,2

475

83

' 97....

cf

1/1

19

163

50/50

8(4.5)

2

1, 2

425

73

' 98. . . .

cf

19

161

49/49

8(4.5)

2

1,2

393

65

' 99. . . .

9

19

162

50/50

8(4.5)

2

1,2

612

97

' 100....

c?

19

162

52/52

8(4.5)

2

1,2

297

48

" 102....

d"

19

161

50/50

■8(4.5)

1

2

1.2

242

40

griffin: ophidia from south America in carnegie museum 189

36. Liophis melanostigma (Wagler) .

Natrix melanostigma Wagler, in Spix, Serp. Bras., 1824, p. 17, PI. IV, fig. 2.
Liophis melanostigma Boulenger, Cat. Snakes, II, 1894, p. 142.

No. 340, 9 , Xiririca, Eio Ribeira, J. D. Haseman coll., December 10, 1908.

This is the only representative of this species in the Museum at the present
time.

A black zigzag band occupies the vertebral region. The lower parts of the
sides are black, this lateral band being broken into large spots on the neck. A
light brown band separates the dorsal and lateral dark stripes. The third and
fourth rows of scales are marked by an indistinct series of small white spots. A
pair of conspicuous white spots lies on the nape just behind the occiput.

Internasals slightly longer than broad; frontal twice as long as broad, consider-
ably longer than its distance from the end of the snout, shorter than the parietals.

Counts and Measurements.

Anal 1/1

Scale-rows 17

Gastrosteges 149

Urosteges : 86/86

Upper labials 9(4.5.6)

8(3.4.5)

Preoculars 1

Postoculars 2

Temporals 1, 2

Total length in mm 770

Length of tail in mm 241

37. Liophis melanotus (Shaw).

Coluber melanotus Shaw, Zoology, III, 1802, p. 534.

Liophis melanotus Boulenger, Cat. Snakes, II, 1894, p. 134.

The Carnegie Museum possesses three examples of this species, as follows:

No. 214, 9 , Bonda, Colombia, H. H. Smith coll., June, 1901.

No. 1863, cf , South America. (No doubt collected by H. H. Smith in Colombia.)

No. 1864, c?, South America. (No doubt collected by H. H. Smith in Colombia.)

Specimen 214 has a row of black spots covering the tips of the second row of
scales in the anterior half of the trunk. The others show traces of the same mark-
ings.

190

MEMOIRS OP THE CARNEGIE MUSEUM

Counts and Measurements.

Anal

Scale-rows

Gastrosteges

Urosteges

Upper labials

Preoculars

Postoculars

Temporals

Total length in mm.. .
Length of tail in mm. .

* End of tail lost.

(No. 214.)

(No. 1863.)

1/1
17

1/1
17

158

155

58/58

8(4.5)

1

36/36*
8(4.5)
1

2

2

1,2

1,2

434

470

85

65*

1/1

17

152

59/59

8(4.5)

1

2

1,2

356

80

38. Liophis poecilogyrus (Wied).

Coluber poecilogyrus Wied, Beitr. Naturgesch. Bras., I, 1825, p. 371, Plate.
Liophis poecilogyrus Boulenger, Cat. Snakes, II, 1894, p. 131.

There are fourteen specimens at hand, which are referable to this species.
Seven of them come from the Provincia del Sara, Bolivia, and were collected by
Jose Steinbach, who neglected except in two cases to indicate the date of capture.
They are listed as follows:

No. 7, c^ , Provincia del Sara, Bolivia, J. Steinbach, Jan., 1912.

No. 53, 9 , Provincia del Sara, Bolivia, J. Steinbach, Jan., 1913.

No. 105, 9 , Santa Cruz de la Sierra, J. Steinbach.

No. 106, 9 , Santa Cruz de la Sierra, J. Steinbach.

No. Ill, cf , Santa Cruz de la Sierra, J. Steinbach.

No. 113, cf , Santa Cruz de la Sierra, J. Steinbach.

No. 274, cf, South America, LeBoutelier Collection.

No. 275, 9 , South America, LeBoutelier Collection.

No. 277, cf , South America, LeBoutelier Collection.

No. 312, cf , Santa Cruz de la Sierra, J. Steinbach.

No. 332, cf , Entre Rios, Brazil, J. D. Haseman, June 4, 1908.

No. 333, cf , Cidade de Matto Grosso, J. D. Haseman, July 1, 1909.

No. 351, cf, Entre Rios, Brazil, J. D. Haseman, June 4, 1908.

No. 359, 9 , Penedo, Alagoas, Brazil, J. D. Haseman, March 22, 1908.

Specimens No. 7, 53, 105, 106, 113, and 312, from the Province del Sara, Bo-
livia, are all very dark on the upper surface, and uniformly colored. The edges
of the dorsal and lateral scales are black, the centers olive. When the surface of a

griffin: ophidia from south America in carnegie museum

191

specimen is dry, the scales appear brownish with an iridescent sheen. The scales
of the upper surface of the head are similarly light brown or olive with dark margins.
The upper labials are white. The lower surface is either entirely white, or some of
the gastrosteges have more or less complete black anterior margins. Nos. 7 and
113, young specimens, are as uniformly marked as the older ones. There is no
nuchal collar, nor are any transverse bars visible. The length of the tail of these
specimens is contained 6| to 7 times in the total length. The number of gastrosteges
and urosteges varies within very narrow limits and about the minimum for L.
poecilogyrus. It is possible that the specimens represent a well defined subspecies.
I feel doubtful about the identification of No. 111. No pits are visible on the
scales, and the length of the tail is contained six and one-half times in the total
length. The dorsal scales are mostly black, but some are white, or white-edged,
so arranged as to form numerous narrow light cross-bars or reticulations. Most
of the gastrosteges are entirely or partly brown.

Counts and Measurements.

03

- a

Is

4S

^3

• CO

Total
Length
in Mm.

Length

Tail
in Mm.

No. 7. . . .

&

1/1

19

149

41/41

8(4.5)

2

1,2

277

42

" 53....

9

1/1

19

148

38/38

1 8(4.5)
1 7(3.4)

2

1,2

510

71

" 105....

9

1/1

19

142

42/42

8(4.5)

2

1,2

535

77

" 106....

9

1/1

19

146

46/46

8(4.5)

2

1,2

512

80

" 111...

c^

1/1

19

148

44/44

8(4.5)

2

1,2

251

39

" 113....

cf

1/1

19

146

41/41

8(4.5)

2

1,1,2

225

33

" 274. . . .

d^

1/1

19

160

61/61

8(4.5)

2

1,2

697

142

" 275. . . .

9

1/1

19

174

57/57

8(4.5)

2

1,2

882

145

" 277....

cf

1/1

19

157

54/54

8(4.5)

2

1,2

480

83

" 312...

c?

1/1

21

141

38/38

8(4.5) ■

2

1,2

494

69

" 332...

cf

1/1

19

162

62/62

8(4.5)

2

1,2

275

52

" 333...

&

1/1

19

153

43/43

8(4.5)

2

1,2

285

40

" 351....

cf

1/1

19

167

56/56

8(4.5)

1,2

513

90

" 359....

9

1/1

19

153

49/49

8(4.5)

2

1,2

800

132

39. Liophis reginae (Linnaeus).

Coluber regince Linn^us, Mus. Ad. Frid., 1754, p. 24, PI. XIII, fig. 3; Syst. Nat.,

Ed. XII, 1766, p. 378.
Liophis regince Boulenger, Cat. Snakes, II, 1894, p. 137.

The Carnegie Museum has three specimens of this species, No. 19, cf , No. 24,
9 , and No. 27, cf , which were taken by Jose Steinbach at Las Juntas, Bolivia, in
December, 1913.

192

MEMOIRS OF THE CARNEGIE MUSEUM

Counts and Measurements.

(No. 19.)

Anal

Scale-rows

Gastrosteges

Urosteges

Upper labials

Preoculars

Postoculars

Temporals

Total length in mm.. .
Length of tail in mm.

(No. 24.)

(No. 27.)

1/1

17

1/1
17

1/1
17

140

149

146

70/70

8(4.5)

1

71/71
8(4.5)

1

76/76

8(4.5)

1

2

2

2

1,2

1,2

1,2

201

729

612

46

182

162

40. Liophis viridis Giinther.
Liophis viridis Gunther, Ann. & Mag. Nat. Hist. (3), IX, 1862, p. 58, PI. IX, fig. 2.
Liophis viridis Boulenger, Cat. Snakes, II, 1894, p. 135.

The specimens in the Carnegie Museum are listed as follows:

No. 281, cf, South America, LeBouteUer collection.

No. 316, 9 , Lagoa de Joao Pereira, Brazil, J. D. Haseman coll., December 23, 1907.
No. 344, 9 , Bom Jesus de Lapa, Bahia, Brazil, J. D. Haseman coU.
No. 344 is defective, having in part lost its tail.

Counts and Measurements.

Anal

Scale-rows

Gastrosteges

Urosteges

Upper labials

Preoculars

Postoculars

Temporals

Total length in mm. . .
Length of tail in mm. .

(No. 281.)

(No. 316.)

1/1

19

1/1
19

1/1
19

182

198

192

72/72
8(4.5)

1

76/76

8(4.5)

1

41/41*
8(4.5)

1

2

2

2

1,2

1,2

1,2

402

597

528

85

125

70*

(No. 344.)

* Tip of tail lost.

Genus Lystrophis Cope.

41. Lystrophis histricus (Jan).

Heterodon histricus Jan, Arch. Zool. Anat. Phys., II, 1863, p. 224.
Lystrophis histricus Boulenger, Cat. Snakes, II, 1894, p. 152.

The only specimen of this species in the possession of the Carnegie Museum
(No. 331) was taken at Sao Matias, Bolivia, by J. D. Haseman, May 7, 1909. It
is a male.

Snout, to level of eyes, brown; a darker brown band across the eyes, followed
by another, which is interrupted in the middle, and ends on each side back of the

griffin: ophidia from south America in carnegie museum 193

mouth. The bands across the head are separated by very narrow light hnes. The
chevron-shaped band of the nape is broad and almost joins the parietal bands.
The remainder of the upper surface of the body is crossed by similar brown, black-
edged, chevron-shaped bands, which almost reach the gastrosteges and are separated
by slightly narrower light bars. A small dark spot lies in the center of each light
bar between the ventral ends of the dark bands. There are a few faint dark spots
on the light ventral surface. The total number of dark bands, including the nuchal
and tail bands, is twenty-six.

Counts and Measurements.

Anal 1/1

Scale-rows 19

Gastrosteges 142

Urosteges 31/31

Upper labials 7(3.4)

Preoculars 1

Postoculars 2

Temporals 1, 2

Total length in mm 176

Length of tail in mm 21

42. Lystrophis semicinctus (Dumeril & Bibron).
Heterodon semicinctus Dumeril & Bibron, Erp. Gen., VII, 1854, p. 774.
Lystrophis semicinctus Boulenger, Cat. Snakes, II, 1894, p. 153.

There are fifteen specimens of this species in the collection, aU taken by Jose
Steinbach in the Province del Sara, Bolivia. It is worthy of note that in the case
of the specimen bearing the catalog number 85 both loreals enter the orbit below
the preocular. The list of species is as follows:

No. 12, 9 , Provincia del Sara, J. Steinbach coll., Feb., 1911.
No. 14, cf , Provincia del Sara, J. Steinbach coll., Feb., 1911.
No. 15, cf , Provincia del Sara, J. Steinbach coll., Feb., 1911.
No. 31, 9 , Provincia del Sara, J. Steinbach coll., Feb., 1913.
No. 33, cf , Provincia del Sara, J. Steinbach coll., Sept., 1913.
No. 36, 9 , Provincia del Sara, J. Steinbach coll., Sept., 1913.
No. 39, cf , Provincia del Sara, J. Steinbach coll., Sept., 1913.
No. 51, cf , Provincia del Sara, J. Steinbach coll., Jan., 1913.
No. 84, cf , Santa Cruz de la Sierra, Jose Steinbach coll.
No. 85, cf , Santa Cruz de la Sierra, Jose Steinbach coll.
No. 86, cf , Santa Cruz de la Sierra, Jose Steinbach coll.
No. 87, 9 , Santa Cruz de la Sierra, Jose Steinbach coll.

194

MEMOIRS OF THE CARNEGIE MUSEUM

No. 88, 9 , Santa Cruz de la Sierra, Jose Steinbach coll.
No. 89, 9 , Santa Cruz de la Sierra, Jose Steinbach coll.
No. 90, cf, Santa Cruz de la Sierra, Jose Steinbach coll.

CotTNTS AND MEASUREMENTS.

Be

No. 12.

14.

15.

' 31.

' 33.

' 36.

' 39.

' 51.

' 84.

' 85.

' 86.

' 87.

' 88.

' 89.

' 90.

9
cf

cf

9
cf

9
cf
cf
cf
&
&

9
9
9

a

1/1
1/1

1/1

1/1
1/1

1/1
1/1
1/1

1)1
1/1
1/1

1/1
1/1
1/1

1/1

« o

21
21

21

21
21

21
21
21
21
21
21

21
21
21

21

151

154

151

155
151

154
155
154
149
153
156

150
140

158

150

27/27
35/35

35/35

25/25

27/27

31/31
30/30
32/32
28/28
37/37
36/36

26/26
25/25
31/31

36/36

P..S

i^3

8(4.5)
8(4.5)

J 8(5)

17(4.5)
8(4.5)

(9(4.5)

18(4..5)
8(4.5)
8(4.5)
8(4.5)
8(4.5)
8(4.5)
8(4.5)

8(4.5)
8(4.5)

(8(5)

18(4.5)
8(4.5)

£.3

1,2

1

1,2

o a^
f-< '^

416
426

488

516

178

488
175
270
237
148
235

178
217
526

500

■2- a

45

58

65

57
17

53
19
32
25
18
27

19
22
59

70

Genus Phrynonax Cope.

43. Phrynonax fascia tus (Peters).

Spilotes fasciatus Peters, Mon. Berl. Akad., 1869, p. 443.
Phrynonax fasciatus Boulenger, Cat. Snakes, II, 1894, p. 21.

This species is represented by a single specimen taken by Mrs. H. H. Smith
at Cacagualito, Colombia, in May, 1901. It is a female and bears the number
2026 in the Catalog of Reptiles in the Carnegie Museum.

Counts and Measurements.

Anal 1

Scale-rows 23

Gastrosteges 209

Urosteges 116/116

Upper labials 8(4.5.6)

7(4.5)

Preocular.s 1

Postoculars 2

Temporals 2, 2

Total length in mm 1240

Length of tail in mm 299

I

griffin: ophidia from south America in carnegie museum 195

Genus Rhadinaea Cope.

44. Rhadinaea merremi (Wied).
Coluber merremii Wied, Reise Bras., II, 1821, p. 121.
Rhadincea merremii Boulenger, Cat. Snakes, II, 1894, p. 168.

The Carnegie Museum has two specimens, bearing the Catalog Nos. 350 and
353, both of which were captured by Mr. J. D. Haseman at Entre Rios, Brazil,
on June 4, 1908. They are males.

Counts and Measurements.

(No. 350.) (No. 353.)

Anal 1/1 1/1

Scale-rows 17 17

Gastrosteges 145 145

Urosteges 47/47 44/44

Upper labials 8(4.5) 8(4.5)

9(4.5.6)

Preoculars 1 1

Postoculars 2 2

Temporals 1,2 1,2

Total length iu mm 842 175

Length of tail in mm 147 31

45. Rhadinaea occipitalis (Jan).
Enicognathus occipitalis Jan, Arch. Zool. Anat. Phys., II, 1863, p. 267.
Rhadincea occipitalis Boulenger, Cat. Snakes, II, 1894, p. 175.

The only specimen we possess is a female (C. M. No. 6) which was caught by
Mr. Jose Steinbach in the Province del Sara, Bolivia, at an elevation of 350 M.

in January, 1912.

Counts and Measurements.

Anal 1/1

Scale-rows 15

Gastrosteges 181

Urosteges 70/70

Upper labials 8(3.4.5)

Preoculars 1

Postoculars 2

Temporals 1, 2

Total length iu mm 507

Length of tail in mm 119

46. Rhadinaea orina sp. nov.
Eighteen small maxillary teeth followed after a short interspace by two much
larger. Eye rather large. Length of parietals equal to their distance from the

196

MEMOIRS OF THE CARNEGIE MUSEUM

internasals (in one case slightly more); rostral a little broader than deep, visible
from above ; internasals broader than long, or as long as broad, much shorter than
the prefrontals; frontal once and two-thirds to twice as long as wide, longer than
its distance from the end of the snout, shorter than the parietals; loreal once and a
half to twice as deep as long; one preocular and two postoculars; temporals 1, 2;
eight upper labials, the fourth and fifth bordering the orbit; five lower labials in
contact with the anterior chin-shields, which are as long as, or (usually) consider-
ably longer than, the posterior.

Scales in seventeen rows; anal divided; gastrosteges 156 to 161; urosteges in
52-60 pairs. Length of tail five and one-third to six times in the total length.

Olive above, the scales tipped or margined with black. The younger specimens
have a chevron-shaped black band crossing the back of the head, the point lying
on the parietal shields, and the ends passing just behind the angle of the mouth.
This is followed by a rather distinct narrow light nuchal collar, behind which are a
number of distinct narrow dark bands separated by much narrower light spaces.
On the posterior part of the trunk and on the tail are four indistinct, longitudinal,
dark stripes, two being dorsal and one along the middle of each side. The upper
surface of the head is dark brown. The upper lip and the lower surface are yellowish,
the anterior margins of the gastrosteges are brown, the brown line being widest at
the outer ends of the scales. The markings are most distinct in the younger speci-
mens. The older ones are nearly uniform olive above, and yellowish beneath,
with only faint traces of the dark bars on the anterior part of the body and of the
lines on the posterior extremity.

This species is very closely related to R. merremi.

Type, C. M. No. 264; paratypes, C. M. Nos. 263, 265, 266, 267.

The specimens belonging to this species form a part of the LeBoutelier Col-
lection presented to the Museum by Mrs. A. Marshall Bell, and are labelled as
from "the Sierras of Bolivia."

Counts and Measurements.

Anal

Scale-rows

Gastrosteges •. . .

Urosteges

Upper labials . ;

Preoculars

Postoculars

Temporals

Total length in mm. .
Length of tail in mm

(No. 263 9.)

(No. 264 J.)

(No. 265 9.)

(No. 266 c?'.)

1/1
17

1/1

17

1/1

17

1/1
17

161

157

156

157

60/60

8(4.5)
1

55/55
8(4.5)

1

55/55

8(4.5)

1

55/55

8(4.5)

1

2

2

2

2

1,2

1,2

1,2

1,2

336

337

274

235

63

60

50

43

(No. 267 (f.)

1/1

17

159

52/52

8(4.5)

1

2

1,2

243

41

griffin: ophidia from south America in carnegie museum

197

Genus Spilotes Wagler.

47. Spilotes puUatus (Linnaeus).

Cohiber piillafus Linn^us, Mus. Ad. Frid., 1754, p. 35, PL XX, fig. 3; Syst. Nat.,

Ed. XII, 1766, p. 388.
Spilotes pullatus Boulenger, Cat. Snakes, II, 1894, p. 23.

The eight specimens we possess were all taken in Colombia by Mr. H. H.
Smith. They are listed as follows :

No. 145, Bonda, Colombia, elev. 150 ft., June.

No. 146, 9 , Bonda, Colombia, elev. 150 ft., June.

No. 147, cf , Bonda, Colombia, elev. 150 ft., August.

No. 148, cf, Bonda, Colombia, elev. 150 ft., June.

No. 187, 9 , Bonda, Colombia, elev. 150 ft., August.

No. 1781, d", Masinga, Colombia, elev. 2,000 ft., August 16.

No. 2039, cf , Bonda, Colombia, elev. 150 ft., June.

No. 2040, 9 , Minca, Colombia, elev. 2,000 ft., July 27, 1899.

Counts and Measurements.

6u
■ 1%

eg

*5

Scale-
rows.

It

Uro-

steges.

L'pper
Labials.

o

Total
Length
in Mm.

Length

Tail
in Mm.

No. 145 ...

16

114/114

8(4.5)
9(5.6)

1

2

2,1

1,1

1,930

527

" 146 . .

9

18

226

122/122

8(4.5)

2

1,1

1,880

450

" 147 ...

d^

16

215

129/129

8(4.5)

2

2,1

2,135

571

" 148 ...

rf"

16

210

122/122

7(3.4)

2

1,1

1,875

528

" 187 . . .

9

16

223

121/121

7(3.4)

2

1,1

760

184

" 1781 . . .

cf

16

213

117/117

S(4..5)

2

1,1

2,010

562

" 2039 . . .

c?

16

231

117/117

7(3.4)

2

1,1

532

125

" 2040. . .

9

16

220

113/113

8(4.5)

2

1,1

520

116

Genus Tropidodipsas Giinther.
48. Tropidodipsas spilogaster sp. nov. (Plate XVIII, figs. 4-6.)

Maxillary short, the thirteen or fourteen maxillary teeth decreasing a little
in front and more behind; the posterior mandibular teeth decrease gradually.
Head very distinct from neck; eye large, its length equal to the distance from the
front of the orbit to the center of the nostril, protuberant, with a small vertically
elliptic pupil. Body cylindrical, slender; scales smooth, without pits.

Rostral much broader than deep, just visible from above; nasal divided;
internasals a trifle more than half as long as the prefrontals; prefrontals entering
the orbit; frontal as broad as long, shorter than its distance from the end of the

198 MEMOIRS OF THE CARNEGIE MUSEUM

snout, much shorter than the parietals, twice as wide as the supra-ocular; supra-
ocular much broader behind than in front; loreal oblong, once and a fifth to once
and a half as long as deep, entering the orbit; no preocular; one crescentic postocular;
temporals 1, 2; seven upper labials, the third and fourth entering the orbit; two
pairs of short chin-shields, the anterior a little the longer and as broad as long, the
posterior broader than long.

Scales smooth, in 15 rows; anal entire; gastrosteges 147-157; urosteges in 44
to 47 pairs.

Upper surface reddish brown, with a dorsal series of black spots narrowly
margined with white; the anterior three or four spots are several times as large as
the others and extend upon the edges of the gastrosteges. Between the first and
second, and the second and third spots, the interspaces are almost white (red in
life?). The sides of the body and tail bear smaller and more numerous black spots
which may reach to the edges of the gastrosteges. The scales of the lighter
areas are finely stippled with brown and black.

A black fleur-de-lys-like marking on the parietals and frontals (on specimen
47 a nearly uniform dark rounded spot) ; lips, chin, and lower surface white (pink
in life ?) with black dots; most of these on specimen 42 are very small, but on 47
they are nearly all large square checks. The urosteges are nearly covered by black
checks.

The type (C. M. No. 42) and the paratype (C. M. No. 47) were both collected
in the Province del Sara, Bolivia, at an elevation of 350 meters, by Mr. Jose Stein-
bach, the former in November, the latter in December, of the year 1912. Both
specimens are males.

Counts and Measurements.

(No. 42 cf.) (No. 47 cf.)

Anal 1 1

Scale-rows 15 15

Gastrosteges 147 157

Urosteges 44/44 47/47

Upper labials 7(3.4) 6(3.4)

7(3.4)

Preoculars 0 0

Postoculars 1 1

Temporals 1,2 1,2

Total length in mm 297 288

Length of tail in mm 56 51

griffin: ophidia from south America in carnegie museum

199

Genus Xenodon Boie.

49. Xenodon colubrinus Giinther.

Xenodon colubrinus Guenther, Catalog of Colubrine Snakes, 1858, p. 55.
Xenodon colubrinus Boulenger, Cat. Snakes, II, 1894, p. 146.

The only representative of this species is a specimen which is derived from the
LeBouteher collection (C. M. No. 280) and which has no other indication of the
locality than the label "South America." It is a male.

Counts and Measurements.

Anal J

Scale-rows jg

Gastrosteges I47

Urosteges 43/43

Upper labials 8(4.5)

Preoculars j

Postoculars 2

Temporals j 2

Total leng-tli in mm 435

Length of tail in mm 53

50. Xenodon merremi (Wagler).

Ophis merremii Wagler, in Spix, Serp. Bras., 1824, p. 47, PI. XVII.
Xenodon merremii Boulenger, Cat. Snakes, II, 1894, p. 150.

There are three specimens at hand, all of which are females. They are cata-
loged as follows :

C. M. No. 52, 9 , Prov. del Sara, Bolivia, J. Steinbach coll., January, 1913.
C. M. No. 56, 9 , Prov. del Sara, Bolivia, J. Steinbach coll., October, 1911.
C. M. No. 315, 9 , Puerto Suarez, Bolivia, J. Steinbach coll., January, 1909.

Counts and Measurements.

Anal

Scale-rows . . .
Gastrosteges. .
Urosteges. . . .
Upper labials .
Preoculars . . .

Postoculars.

Temporals

Total length in mm. .
Length of tail in mm.

(No. 52.)

1/1

19

151

32/32

7(3.4)

li
\l

1,2
740

78

(No. 56.)

1

19

146

35/35

7(3.4)

1

1,2
835
102

(No. 315.)

200 MEMOIRS OF THE CARNEGIE MUSEUM

The anal of No. 56 is single, but shows indications of a fusion of two scales.
This specimen has only four lower labials in contact with the anterior chin-shields.

51. Xenodon neuwiedi Giinther.

Xenodon neuwiedii Guenther, Ann. & Mag. Nat. Hist. (3), XII, 1863, p. 354,

PI. V, fig. C.
Xenodon neuwiedii Boulenger, Cat. Snakes, II, 1894, p. 148.

No. 1234, cf, no data.
No. 1235, d', no data.

Counts and Measurements.

(No. 1234.) (No. 12.35.)

Anal 1/1 1/1

Scale-rows 21 21

Gastrosteges 169 168

Urosteges 59/59 64/64

Upper labials 8(4..5) 8(4.5)

Preoculars 1 1

Postoculars 2 2

Temporals 1,2 1, 2 '

Total length in mm 330 478

Length of tail in mm 54 83

52. Xenodon severus (Linnaeus).

Coluber severus Linnaeus, Mus. Ad. Frid., 1754, p. 25, PL VIII, fig. 1; Syst. Nat.,

Ed. XII, 1766, I, p. 379.
Xenodon severus Boulenger, Cat. Snakes, II, 1894, p. 149.

The only specimen in the collection is a male (C. M. No. 128) taken by Jose
Steinbach at Santa Cruz de la Sierra, Bolivia. Mr. Steinbach furnished no date
upon his label.

(No. 128.)

Anal 1

Scale-rows 21

Gastrosteges 133

Urosteges 37/37

Upper labials 8(4.5)

Preoculars f 2

ll
Postoculars J 2

13

Temporals 1,2

Total length in mm 385

Length of tail in mm 48

griffin: ophidia from south America in carnegie museum

201

Series B. Opisthoglypha.

Subfamily Dipsadomorphin^ Boulenger.

Genus Clelia (Oxyrhopus) Fitzinger.

53. Clelia bitorquata (Giinther) .

Tachymenis bitorquatus Guenther, Ann. & Mag. Nat. Hist. (4), IX, 1872, p. 19.
Oxyrhopus bitorquatus Boulenger, Cat. Snakes, III, 1896, p. 104, PI. VI, fig. 1.

The four specimens of this species in the Museum are listed as follows:

C. M. No. 272, 9 , "South America," LeBoutelier Collection.

C. M. No. 343, cf , Bom Jesus de Lapa, Bahia, Brazil, J. D. Haseman coll., no date.

C. M. No. 369, cf , Piracicaba, J. D. Haseman coll., Sept. 7, 1908.

C. M. No. 376, 9 , Tarma, Peru, Miss Lola Vance coll., no date.

Nos. 272, 343, and 369 have black cross-bands arranged in threes the entire
length of the body and tail. No. 376 has two black bands on the neck, two sets
of threes on the anterior part of the body, traces of a third set, and the remainder
of the upper surface (originally) red with black tipped scales.

Counts and Measuhements.

Anal

Scale-rows

Gastrosteges

Urosteges

Upper labials

Preoculars

Postoculars

Temporals

Total length in inm. .
Length of tail in mm.

(No. 272.)

1

19

208

76/76

8(4.5)

1

2

2,3

808

152

(No. 343.)

1

19

200

63/63

8(4.5)

1

2

2,3

656

111

(No. 369.)

1

19

205

70/70

8(4.5)

1

2

2,3

1150

200

(No. 376.)

1

19

207

75/75

8(4.5)

1

2

2,3

674

110

54. Clelia cloelia (Daudin).

Coluber cloelia Daudin, Rept., VI, 1803, p. 330, PI. LXXVIII.
Oxyrhopus cloelia Boulenger, Cat. Snakes, III, 1896, p. 108.

C. M. No. 330, cf , Entre Rios, Brazil, J. D. Haseman coll., June 4, 1908.
C. M. No. 380, d^, Rio Mamore, Bolivia, J. D. Haseman coll., Sept. 19, 1909.
In specimen No. 330 there is no loreal.

202 MEMOIHS OF THE CARNEGIE MUSEUM

Counts and MKAStrEEMENxs.

(No. 330.) (No. 380.)

Anal 1 1

Scale-rows 19 19

Gastrosteges 239 230

Urosteges 76/76 77/77

Upper labials 7(3.4) 7(3.4)

Preoculars 1 1

Postoculars 2 2

Temporals 2, 3 2, 3

Total length in mm 623 1,857

Length of tail in mm 99 320

55. Clelia doliata (Dumeril & Bibron).

Oxyrhopus doliatus Dumeril & Bibron, Erp. Gen., VII, 1854, p. 1020.
Oxyrhopus doliatus Boulenger, Cat. Snakes, III, 1896, p. 106.

The collection contains one male of this species, which formed a part of the
LeBoutelier Collection, and, as was the case with all of the specimens in that col-
lection, has no definite locality-label, being simply marked as from " South America."
The tip of the tail of the specimen has been lost.

Eye nearly one-half its distance from the end of the snout; the portion of the
rostral visible from above one-half as long as its distance from the frontal; frontal
once and a quarter as long as broad, a little shorter than its distance from the end
of the snout, and a very little shorter than the parietals.

Broad black cross-bands extending to the gastrosteges, posteriorly making
complete annuli. These are much broader than the light interspaces, almost
touching along the middle of the back, narrower on the sides, though still wider
than the light bands. The markings are intermediate between varieties A and B
of Boulenger {loc. cit.).

Counts and Measubements.

(No. 273.)

Anal 1

Scale-rows 19

Gastrosteges 201

Urosteges *

Upper labials 8(4.5)

Preoculars 1

Postoculars 2

Temporals 2, 3

Total length in mm 835

Length of tail in nun 60*

* Tip of tail lost.

griffin: ophidia from south America in carnegie museum 203

56. Clelia euprepa sp. nov. (Plate XXVIII, figs. 7-9.)

Thirteen solid maxillary teeth followed after a short space by two slightly
enlarged grooved fangs. Eye large, its length three-fifths to two-thirds of the
distance from the eye to the tip of the snout. Snout broad, rounded, scarcely
projecting beyond the lower lip. Rostral one-half as deep as broad, the portion
visible from above one-half or less the length of the internasal; internasals two-
thirds as long as the prefrontals; frontal once and a third as long as wide, pentagonal,
longer than its distance from the end of the snout, shorter than the parietals; loreal
once and a half to nearly twice as long as deep; one preocular, just reaching the
top of the head, separated from the frontal; two postoculars; temporals 2, 3 (2, 2
on the left side of No. 109) ; eight upper labials, the third, fourth, and fifth bordering
the orbit. The third upper labial touches the edge of the orbit by its corner only.
Nine lower labials, four in contact with the anterior chin-shields, which are longer
than the posterior.

Nineteen rows of scales; anal undivided; gastrosteges 225-230; urosteges in
94-105 pairs.

The body and tail bear narrow, irregular black rings, separated by narrower
white spaces which are probably bright red in life. The outlines of the black
markings are extremely irregular, many being united by diagonal black bars, while
others are broken along the back. Very few of the ventral shields of the dark rings
are completely black, the color extending half or two-thirds across a shield and
then stopping abruptly. Most of the light dorsal and lateral scales are tipped
with brown. The scales of the upper surface of the head and the lower labials are
black in the center with light (red ?) edges. A large white (red ?) spot crosses the
nape just behind the parietals.

Type, C. M. No. 109; paratype, C. M. No. 108.

The Museum possesses two males of this species, collected by Jose Steinbach
at Santa Cruz de la Sierra, Bolivia, and numbered as above.

Counts anb Measurements.

(No. 108.) (No. 109.) ~

Anal 1 J

Scale-rows 19 2.9

Gastrosteges 230 225

Urosteges 105/105 94/94

Upper labials 8(3.4.5) 8(3.4.5)

Preoculars 1 i

Postoculars 2 2

Temporals 2, 3 f2, 2

12,' 3

Total length in mm 577 797

Length of tail in mm 135 170

204 MEMOIRS OF THE CARNEGIE MUSEUM

57. Clelia peruviana sp. nov.

Eleven subequal solid maxillary teeth followed after a short space by two
slightly enlarged grooved fangs. Eye small, two-fifths as long as its distance from
the end of the snout. Snout rounded, scarcely projecting beyond the lower lip.
Rostral two-thirds as deep as broad, the portion visible from above equal to the
suture between the internasals, not quite one-third the distance between it and
the frontal; internasals one-half as long as the prefrontals; frontal once and one-
fifth as long as broad, as long as its distance from the end of the snout, a httle
shorter than the parietals; loreal once and two- thirds as long as deep; one preocular,
separated, but not widely, from the frontal, and reaching the upper surface of the
head; two postoculars; temporals 2, 3; eight upper labials, the fourth and fifth
entering the orbit; four lower labials in contact with the anterior chin-shields, which
are longer than the posterior.

Scales in 19 rows; anal entire; 179 gastrosteges; 70 pairs of urosteges.

Top of the head orange, with a dark patch on the frontal and parietals. The
back and sides of the anterior third of the body are uniform black. Back of this
the body is marked by broad black rings separated by narrow light (red?) spaces
in which each scale is tipped or margined with black. The black annuli are a little
narrower below than above. In the posterior part of the body and tail the lower
surface is almost entirely black. Passing forward, the black color becomes limited
more and more to the posterior margins of the ventral shields. The gastrosteges
of the anterior quarter have only their outer edges black-margined. The under
surface of the head, the lips, and the throat are uniformly yellowish.

The type (C. M. No. 377) is unique. It is a male, and was collected at Tarma,
Peru, by Miss Lola Vance, who neglected to give the date of capture.

Counts and Measurements.

(No. 377.)

Anal 1

Scale-rows 19

Gastrosteges 179

Urosteges 70/70

Upper labials 8(4.5)

Preoculars 1

Postoculars 2

Temporals 2, 3

Total length in mm 691

Length of tail in mm 148

griffin: ophidia from south America in carnegie museum

205

58. Clelia petolaria (Linnaeus).
Coluber petolarius Linn^us, Mus. Ad. Frid., 1754, p. 35, PI. IX, fig. 2; Syst. Nat.,

Ed. XII, 1766, I, p. 387.
Oxyrhopus petolarius Boulenger, Cat. Snakes, III, 1896, p. 101.

The specimens in the Carnegie Museum are listed as follows:

C. M. No. 22, d", Las Juntas, BoHvia, elev. 250 M., Steinbach coll., Dec, 1913.
C. M. No. 75, 9 , Santa Cruz de la Sierra, Bolivia, Steinbach coll.
C. M. No. 107, cf , Santa Cruz de la Sierra, Bolivia, Steinbach coll.
C. M. No. 110, cf , Santa Cruz de la Sierra, Bolivia, Steinbach coll.

The specimen No. 107 is shghtly defective, having lost the tip of its tail.
Nos. 22, 107, and 110 have broad black bars across the back; the abdomens are
uniformly light.

No. 22 has nineteen bars, three to four times as wide as the light interspaces.
No. 107 has nineteen bars, two to three times as wide as the interspaces. No. 110
has 28 bars, three to four times as wide as the very narrow interspaces.

Counts and Meastjeements.

(No. 22.)

(No. 75.)

(No. 107.)

(No. 110.)

Anal

Scale-rows

Gastrosteges

Urosteges

Upper labials

Preoculars

Postoculars

Temporals

Total length in mm. .
Length of tail in mm.

1

19

197

82/82

8(4.5)

1

2

2,3

313

63

1

19

200

80/80

8(4.5)

1

2

2,3

164

38

1

19

205

54/54*

8(4.5)

1

2

2,3

633

54*

1

19

200

85/85

8(4.5)

1

2

2,3

255

54

* Tip of tail lost.

59. Clelia rhombifer (Dumeril & Bibron).

Oxyrhopus rhombifer Dumeril & Bibron, Erp. Gen., VII, 1854, p. 1018.
Oxyrhopus rhombifer Boulenger, Cat. Snakes, III, 1896, p. 103.

C. M. No. 11, (f, Provincia del Sara, Bolivia, Steinbach coll, March, 1912.
C. M. No. 45, cf , Provincia del Sara, Bolivia, Steinbach coll., December, 1912.
C. M. No. 57, cf , Provincia del Sara, Bolivia, Steinbach coll., December, 1911.
C M. No. 64, cT, Provincia del Sara, Bolivia, Steinbach coll., December, 1911.
C. M. No. 65, cf , Provincia del Sara, Bolivia, Steinbach coll., December, 1911.
C. M. No. 66, cf, Provincia del Sara, BoUvia, Steinbach coll., December, 1911.

206

MEMOIRS OF THE CARNEGIE MUSEUM

Counts and Meastjbements.

Be

GO

<1

<1> n

is

si

o

1.5

■ ID

11

Total
Length
in Mm.

■3^ a

No. 11

cf

1

19

196

75/75

8(4.5)

1

2

2,3

580

125

" 45

cf

1

19

201

87/87

8(4.5)

2

2

2,2

375

80

" 57

cT

19

203

69/69

8(3.4.5)

1

2

2,3

661

118

" 64

ff

19

206

66/66

8(4.5)

1

2

2,3

633

117

" 65

cf

19

195

77/77

8(4.5)

1

2

2,3

336

68

" 66

cf

19

198

77/77

8(3.4.5)

8(4.5)

1

2

2,3

615

130

60. Clelia trigemina (Dumeril & Bibron).
Oxyrhopus trigeminus Dumeril & Bibron, Erp. Gen., VII, 1854, p. 1013.
Oxyrhopus trigeminus Boulenger, Cat. Snakes, III, 1896, p. 104.

The only specimen we have at present is a female from the LeBoutelier Col-
lection, which is simply labelled as from "South America." C. M. No. 271.

Frontal a trifle shorter than its distance from the tip of the snout, much shorter
than the parietals. The triads of dark brown bars are widely separated, but the
individual bars in each are close and not very distinctly marked, as the intervening
red scales are heavily colored with dark brown.

Counts and Measurements.

(No. 271.)
1

Anal

Scale-rows 19

Gastrosteges 199

Urosteges 64/64

Upper labials 8(4.5)

Preoculars 1

Postoculars 2

Temporals 2, 3

Total length in mm 688

Length of tail in mm 120

Genus Pseudoboa Schneider.
61. Pseudoboa coronata Schneider.
Pseudoboa coronata Schneider, Hist. Amph., II, 1801, p. 286.
Oxyrhopus coronatus Boulenger, Cat. Snakes, III, 1896, p. 111.

The species is represented in the collection by a male (C. M. No. 116) taken
at Santa Cruz de la Sierra by Jose Steinbaeh, who failed to give any further in-
formation.

Eye two-fifths as long as the snout; frontal slightly broader than long, shorter
than its distance from the end of the snout; anterior chin-shields a little shorter

geiffin: ophidia from south America in carnegie museum 207

than the posterior; pits of scales extremely small and difficult to see. The color
(in alcohol) is very light, almost white above, except for the dark brown on the
front of the head and the nape.

Counts and Measueements.

(No. 116.)

Anal 1

Scale-rows 17

Gastrosteges 191

Urosteges 76

Upper labials 7(3.4)

Preoculars 1

Postoculars 2

Temporals 2, 2

Total length in mm 297

Length of tail in mm 64

Genus Erythrolamprus Wagler.

62. Erythrolamprus aesculapii (Linnaeus).

Coluber cesculapii Linn^us, Mus. Ad. Frid., 1754, p. 29, PL XI, fig. 2; Syst. Nat.,

Ed. XII, 1766, I, p. 380.
Erythrolamprus cesculapii Boulenger, Cat. Snakes, III, 1896, p. 200.

There are fourteen specimens of this species in the collection. All belong
to the variety having double annuli. The upper surface of the head is largely
black, without distinct bands. The specimens are cataloged as follows:

C. M. No. 129, d", Valparaiso, Colombia, H. H. Smith coll., April.

C. M. No. 130, 9 , Bonda, Colombia, H. H. Smith coll., May.

C. M. No. 131, 9 , Bonda, Colombia, H. H. Smith coll., June.

C. M. No. 132, 9 , Bonda, Colombia, H. H. Smith coll., May.

C. M. No. 133, 9 , Bonda, Colombia, H. H. Smith coll., June.

C. M. No. 134, 9 , Bonda, Colombia, H. H. Smith coll.. May.

C. M. No. 135, c?, Bonda, Colombia, H. H. Smith coll.. May.

C. M. No. 270, 9 , South America, LeBoutelier Collection.

C. M. No. 342, 9 , Entre Rios, Brazil, J. D. Haseman coll., June 4, 1909.

C. M. No. 1096, cf, Bonda, Colombia, H. H. Smith coll., August.

C. M. No. 1841, 9 , South America, LeBoutelier Collection.

C. M. No. 1842, 9 , South America, LeBoutelier Collection.

C. M. No. 1843, 9 , South America, LeBoutelier Collection.

C. M. No. 2035, cf , Cacaguahto, Colombia, Mrs. H. H. Smith coll., no date given.

208

MEMOIRS OF THE CARNEGIE MUSEUM

Counts and Measueements.

3 c
1-

"3

a
<

i on

11
03 <•

It

O to

Upper
Labials.

-1

J.2

o

■ 00

Total
Length
in Mm.

Length

Tail
in Mm.

No. 129...

&

1/1

15

192

60/60

7(3.4)

{?

2

1,2

507

83

" 130...

9

1/1

15

188

47/47

7(3.4)

2

1,2

516

70

" 131 .. .

9

1/1

15

193

52/52

7(3.4)

{\

2

1,2

867

130

" 132...

9

1/1

15

185

47/47

7(3.4)

2

1,2

804

120

" 133...

9

1/1

15

193

37/37

7(3.4)

2

1,2

798

87

" 134...

9

1/1

15

190

48/48

7(3.4)

2

1,2

859

113

" 135...

cf

1/1

15

201

61/61

7(3.4)

2

1,2

780

146

" 270...

9

1/1

15

193

36/36

7(3.4)

2

1,2

558

58

" 342 .. .

9

1/1

15

203

48/48

7(3.4)

2

1,2

756

96

" 1096...

cf^

1/1

15

195

59/59

7(3.41

2

1,2

784

131

" 1841 . . .

9

1/1

16

191

3/3*

7(3.4)

2

1,2

709

5*

" 1842 . . .

9

1/1

15

179

51/51

7(3.4)

2

1,2

842

125

" 1843 . . .

9

1/1

15

192

51/51

7(3.4)

1

2

1,2

810

115

" 2035. . .

&

1/1

15

192

60/60

1 7(3.4)

2

1,2

732

128

* The tail is lost.

Genus Himantodes Dumeril & Bibron.

63. Himantodes cenchoa (Linnaeus).

Coluber cenchoa Linnaeus, Syst. Nat., Ed. XII, 1766, I, p. 389.
Himantodes cenchoa Boulenger, Cat. Snakes, III, 1896, p. 84.

The specimens in the Carnegie Museum are cataloged as follows:

C. M. No. 189, 9 , Bonda, Colombia, H. H. Smith coll., September 1.

C. M. No. 190, c?, Bonda, Colombia, H. H. Smith coll., June.

C. M. No. 362, cf , Muniz Freire, Espiritu Santo, Brazil, Haseman coll., June 18,

1908.
C. M. No. 375, cf , Tarnia, Peru, Miss Lola Vance coll., no date.

CoXraTS AND

Measurements.

(No. 189.)

(No. 190.)

(No. 362.)

(No. 375.)

Anal

243

145/145
8(4.5)
8(3.4.-5)
1

2

2,3
3,3
902
251

1/1
17

236
142/142
8(4.5)
8(3.4.5)

\
2

2,3

820
240

1/1

17

240

167/167

8(4.5)

2

2

2

3

2,2,3

2,3

1285

381

1/1

Scale-rows

17

237

TJrostpffps

157/157

8(3.4.5)

Preociilars

9(4.5.6)
2

Postoculars

Temporals

Total length in mm ....

2

2,3
3,3

1057

Length of tail in mm

334

griffin: ophidia from south America in carnegie museum 209

Genus Tantilla Baird & Girard.

64. Tantilla melanocephala (Linnaeus).

Coluber melanocephalus Linnaeus, Mus. Ad. Frid., 1754, p. 24, PL XV, fig. 2;

Syst. Nat., Ed. XII, 1766, I, p. 378.
Homalocranium melanocephaluni Boulenger, Cat. Snakes, III, 1896, p. 215.

The only example of this serpent which we possess is a male, taken by Mr. J. D.
Haseman at Santa Cruz, in the province of Matto Grosso, Brazil, on June 22, 1909.
It bears the catalog number 337.

Counts and Measurements.

(No. 337.)

Anal 1/1

Scale-rows 15

Gastrosteges 152

Urosteges 63/63

Upper labials 7(3.4)

Preoculars 1

Postoculars 2

Temporals 1,1

Total length in mm 180

Length of tail in mm 43

65. Tantilla semicincta (Dumeril & Bibron).

Homalocranium semicinctum Dumeril & Bibron, Erp. Gen., VII, 1854, p. 862.
Homalocranium semicinctum Boulenger, Cat. Snakes, III, 1896, p. 219.

The specimens in the Carnegie Museum are cataloged as follows:

No. 200, 9 , Bonda, Colombia, H. H. Smith coll., August.

No. 210, d^, Bonda, Colombia, H. H. Smith coll., June.

No. 1094, d", Bonda, Colombia, H. H. Smith coll., August.

No. 1844, 9 , South America, Mr. and Mrs. H. H. Smith coll., no date.

No. 2024, 9 , Cacagualito, Colombia, Mrs. H. H. Smith, no date.

No. 2037, cf, Cacagualito, Colombia, Mrs. H. H. Smith, no date.

Nos. 200, 210, 2024, 2037, and 1844 are all dark brown above with narrow
white bars or half-bars. The lower surfaces are white. The tip of the snout is
yellow touched with brown; there is a yellow spot on the lip behind the eye. The
stomach of no. 1094 contained a centipede 153 mm. long.

210

MEMOIRS OF THE CARNEGIE MUSEUM

Counts and Measueements.

a>:

a
<

is

If

t m

Upper
Labials.

-1

Total
Length
in Mm.

Length

Tail
in Mm.

No. 200...

9

15

178

59/59

7(3.4)

2

482

98

" 210...

cf

15

170

68/68

7(3.4)

2

465

110

" 1094...

d"

15

170

64/64

7(3.4)

2

1, 1 1 425

95

" 1844...

9

15

180

13/13*

7(3.4)

2

1, 1 '•■ 437

22*

" 2024 . . .

cf

15

175

10/10*

7(3.4)

2

1, 1 , 448

17*

" 2037 . . .

9

1/1

15

172

57/57

7(3.4)

2 1, 1 200

41

* Tip of tail lost.

Genus Leptodeira Fitzinger.
66. Leptodeira annulata (Linnaeus).

Coluber annularis Linn^us, Mus. Ad. Frid., 1754, p. 34, PI. VIII, fig. 2.
Leptodira annulata Boulenger, Cat. Snakes, III, 1896, p. 97.

We have twenty-one specimens of this snake, fifteen from Bohvia, and six
from Colombia. The Bohvian specimens were all taken in the Province del Sara
by J. Steinbach, and the Colombian specimens by Mr. and Mrs. H. H. Smith. In
the following table the first fifteen numbers are those of the Bolivian examples,

Counts

AND MeASUEEMENTS.

a^
1-

a

11

02 I-

2»
St

si

^1

^3

a2

=1

, 2

O

^1

1 GO

Is

Total
Length
in Mm.

Length

Tail
in Mm.

No 9

9

1/1

19

188

61/61*

8(3.4.5)

0

{\

11

2

1,2

675

138*

37

9

1/1

19

189

73/73

8(3.4.5)

0

1,2

413

85

70

cf

1/1

19

191

87/87

("9(4.5.6)
18(4.5)

0

2

1,2

502

122

71

9

1/1

19

187

73/73

7(3.4)

0

2

1,2

645

138

72

&

1/1

21

189

80/80

8(4.5)

1

2

1,2

680

150

73

c?

1/1

19

170

74/74

8(4.5)

1

2

1,2

680

148

74

cf

1/1

19

186

92/92

(8(4.5)
18(3.4.5)

0

2

1,2

245

49

" 76

cf

1/1

19

196

55/55*

8(4.5)

{5

1

2

1,2

603

105*

77

cf

1/1

19

196

85/85

8(4.5)

2

1,2

485

111

" 78

d^

1/1

19

190

87/87

7(3.4)

0

2

1,1

610

150

79

9

1/1

19

184

24/24*

8(4.5)

1

-*■

2

1,2

555

52*

80

d'

1/1

19

189

90/90

(head injured)

705

178

" 81

9

1/1

19

191

73/73*

8(4.5)

1

{?

{1

1,1

700

140*

" 82

&

1/1

19

191

85/85

8(4.5)

1

{!

1,2

480

118

83

&

1/1

19

191

84/84

8(3.4.5)

0

2

1,2

616

155

" 191

d"

1/1

21

180

88/88

8(4.5)

1

2

1,2

561

154

" .192

cf

1/1

21

178

8(4.5)

2

1,1

502

71*

" 193

cf

1/1

21

178

93/93

8(4.5)

2

1,2

491

132

" 194

cf

1/1

19

173

8(4.5)

2

1,2

520

75*

" 205

cf

1/1

21

177

88/88

8(4.5)

2

1,2

404

103

" 2023

cf

1/1

21

176

79/79

8(4.5)

2

1,2

805

181

' Tip of tail lost.

griffin: ophidia from south America in carnegie museum

211

and the last six represent the material from Colombia. The Colombian material
was collected from June to September at Bonda, except No. 2023, which was taken
at Cacagualito. No dates accompany the specimens received from Mr. Steinbach,
except in the case of No. 9, which was taken in January, and No. 37, which bears
the label "August, 1913."

Genus Oxybelis Wagler.

67. Oxybelis acuminatus (Wied).

Coluber acuminatus Wied, Abbild. Nat. Bras., 1822, and Beitr. Nat. Bras., I,

1825, p. 322.
Oxybelis acuminatus Boulenger, Cat. Snakes, III, 1896, p. 192.

The specimens in the Carnegie Museum are cataloged as follows:
No. 41, cf , Provincia del Sara, Bolivia, Steinbach coll., November, 1912.
No. 171, 9 , Bonda, Colombia, H. H. Smith coll., 1901.
No. 172, 9 , Bonda, Colombia, H. H. Smith coll., June, 1901.
No. 173, 9 , Bonda, Colombia, H. H. Smith coll., July, 1901.
No. 174, cf , Bonda, Colombia, H. H. Smith coll., July, 1901.
No. 175, 9 , Bonda, Colombia, H. H. Smith coll., August, 1901.
No. 184, 9 , Bonda, Colombia, H. H. Smith coll., August, 1901.
No. 1845, d", Colombia, H. H. Smith.
No. 1846, 9 , Colombia, H. H. Smith.
No. 1847, 9 , Colombia, H. H. Smith.
No. 1848, cf , Colombia, H. H. Smith.
No. 1849, d", Colombia, H. H. Smith.
No. 1850, d", Colombia, H. H. Smith.
No. 2010, 9 , Cacagualito, Colombia, Mrs. H. H. Smith.

Counts and Measurements.

3^
1-

02

"3

<

It

■ QQ

^4

o

^2

^1

Tem-
porals.

Total
Length
in Mm.

Length

Tail
in Mm.

No. 41

&

1/1

17

186

158/158

9(4.5.6)

1

2

1,2

1,083

434

" 171

9

1/1

IV

185

161/161

9(.45.6)

2

1,2

1,165

458

" 172

9

1/1

17

179

159/159

(94.5.6)

2

1,2

1,390

550

" 173

V

1/1

IV

182

1.53/153

9(4.5.6)

2

1,2

1,020

388

" 174

o"

1/1

17

182

170/170

9(4.5.6)

2

1,2

1,172

472

" 175

9

1/1

17

183

153/153

9(4.5.6)

1 2

1,2

1,136

435

" 184

9

1/1

IV

189

159/159

9(4.5.6)

1 i 2

1 2

1,260

481

" 1845

cf

1/1

17

166/166

9(4.5.6)

1 1 2

1,2

1,240

483

" 1846

9

1/1

17

iss

160/160

9(4.5.6)

{1 {?

1 2

1, 2

1,090

318

" 1847

9

1/1

17

179

146/146

9(4.5.6)

1,2

1,180

445

" 1848

&

l/l

IV

180

173/173

9(4.5.6)

1 2

1 2

1,105

454

" 1849

o"

1/1

17

184

159/159

9(4.5.6)

1 ' 2

1 2

1,128

441

" 1850

cr'

1/1

17

183

159/159

9(4.5.6)

1 ; 2

1, 2

455

162

" 2010

9

1/1

IV

176

159/159

8(4.5)

1 ! o

X 1 ^

1.2

1,194

471

212 MEMOIRS OF THE CARNEGIE MUSEUM

68. Oxybelis fulgidus (Daudin).

Coluber fulgidus Daudin, Kept., VI, 1803, p. 352, PL LXXX.
Oxybelis fulgidus Boulenger, Cat. Snakes, III, 1896, p. 191.

The only specimen we have is a female taken at C'acagualito, Colombia, by
Mrs. H. H. Smith. It is C. M. No. 2027.

Counts and Measurements.

(No. 2027.)

Anal 1/1

Scale-rows 17

Gastrosteges 208

Urosteges 158/158

Upper labials 10(5.6.7)

Preoculars 1

Postoculars 2

Temporals 1,2

Total length in mm 1708

Length of tail in mm 592

Genus Philodryas Wagler.

69. Philodryas nattereri Steindachner.

Philodryas nattereri Steindachner, Sitzber. Ak. Wien, LXII, 1870, p. 345, PL

VII, figs. 1-3.
Philodryas nattereri Boulenger, Cat. Snakes, III, 1896, p. 134.

A single male, captured by J. D. Haseman at Bom Jesus de Lapa, Bahia,
Brazil. No other information accompanies the specimen. It bears the number
345 in the Catalog of the Museum.

Brown above, darkest on the head and neck. Faint brown lines run along
the centers of the first and fourth rows of scales on each side of the body.

Counts and Measurements.

(No. 345.)

Anal 1

Scale-rows 21

Gastrosteges 209

Urosteges 132/132

Upper labials 8(4..5)

Preoculars 1

Postoculars • 2

Temporals 2, 2

Total length in mm 540

Length of tail in imn 166

griffin: ophidia from south America in carnegie museum

213

70. Philodryas olfersi (Lichtenstein).
Coluber olfersii Lichtenstein, Verz. DoubL, 1823, p. 104.
Philodryas olfersii Boulenger, Cat. Snakes, III, 1896, p. 129.

The specimens in the Carnegie Museum are listed as follows :
No. 10, cf , Provincia del Sara, Bolivia, Steinbach coll., March, 1912.
No. 13, o^, Provincia del Sara, Bolivia, Steinbach coll., February, 1911.
No. 16, cf , Provincia del Sara, Bolivia, Steinbach coll., Februaiy, 1911.
No. 62, cf , Provincia del Sara, Bolivia, Steinbach coll., October, 1911.
No. 101, cf, Santa Cruz de la Sierra, Bolivia, Steinbach, no date.
No. 103, cf , Santa Cruz de la Sierra, Bolivia, Steinbach, no date.
No. 269, d", Brazil, LeBoutelier Collection.
No. 374, d^, Tarma, Peru, INIiss Lola Vance coll., no date.

Counts and Measurements.

a c

K

^

a

£1
Pi

OS

o

^2

(2"=

1 cc

Total
Length
in Mm.

Length

Tail
in Mm.

No. 10...

(f

1/1

•19

189

108/108

8(4.5)

1

2

1,2

491

138

" 13..

cf

1/1

19

193

103/103

8(4.5)

2

^

1,2

1,275

299

" 16..

ff

1/1

19

188

112/112

8(4.5)

2

1,2

570

166

" 62..

cf

1/1

19

185

80/80*

8(4.5)

a

1,2

496

120*

" 101..

(f

1/1

19

192

96/90

8(4.5)

2

1,2

407

99

" 103..

r?

1/1

19

188

112/112

8(4.5)

2

1,2

960

282

" 269..

cf

1/1

19

176

104/104

8(4.5)

2

Jl, 1
11,2

900

255

" 374....

cf

1/1

19

189

72/72*

8(4.5)

fl

u

\l

1,2

480

108*

* Tip of tail lost.

71. Philodryas schotti (Schlegel).
Xenodon schottii Schlegel, Phys. Serp., II, 1837, p. 91, PL III, figs. 8 and 9.
Philodryas schotti Boulenger, Cat. Snakes, III, 1896, p. 130.

The only example of this species at hand is a female (No. 104) taken by Mr.
Jose Steinbach at Santa Cruz de la Sierra, Bolivia.

Counts and Measurements.

(No. 104.)

Anal 1/1

Scale-row.'i 19

Gastrosteges 193

Urosteges 99/99

Upper labials 7(3.4)

Preoculars 1

Postoculars 2

Temporals 1,2

Total length in mm 1515

Length of tail in mm 379

214

MEMOIRS OF THE CARNEGIE MUSEUM

Genus Rhinostoma Dumeril & Bibron.

72. Rhinostoma guianense (Troschel).

Heterodon guianensis Troschel, in Schomburgk, Reise Brit. Guian., Ill, 1848,

p. 653.
Rhinostoma guianense Boulenger, Cat. Snakes, III, 1896, p. 114.

There are thirteen examples of this species in the Museum of which it may
be confidently asserted that they were all secured in Colombia by Mr. and Mrs.
H. H. Smith, when engaged in collecting in that part of the world. All the speci-
mens which have locality labels attached came from Bonda, and the dates of those
upon which there are records of the time of capture show that they were taken from
April to the first week in September.

Nos. 211, 212, 1857, and 1858 are of the light-colored variety with a dark brown
collar on the nape.

Counts and Measueements.

Be

a;
W

■3

a
<

^2

h a*

• 00

PS

00

. 2

O

o

Total
Length
in Mm.

Length

Tail
in Mm.

No. 163...

9

19

200

61/61

8(4.5)

\l

{i

2,3

916

151

" 164...

(f

19

196

57/57

8(4.5)

{!

1

" 2

2,3

835

134

" 165...

cf

19

201

63/63

8(4.5)

2

2,3

825

140

" 166...

cf

19

199

60/60

8(4.5)

2

2, 3

836

140

" 167...

c?

19

191

69/69

8(4.5)

{[

it

2,3

611

124

" 168...

cf

19

206

62/62

8(4.5)

\ -^

2

2,3

956

157

" 169

cf

19

185

70/70

(9(5.6)
(8(4.5)

3

(3,4
12,3

553

114

" 170...

d'

19

197

57/57

8(4.5)

2

2,3

554

86

" 211...

d"

19

199

55/55+5

8(4.5)

-1

2

2,3

375

58

" 212...

d"

21

192

73/73

(9(4.5.6)
18(4.5)

2

3,3

344

62

" 1856...

c?

19

. 201

52/52+6

8(4.5)

2

2,3

540

83

" 1857...

c^

19

193

70/70

8(4..5)

2

2,3

442

80

" 1858* . .

••

••

••

335*

65*

* Mutilated.

Genus Tachymenis Wiegmann.

73. Tachymenis peruviana Wiegmann.

Tachymenis peruviana Wiegmann, Nov. Acta Ac. Leop.-Carol., XVII, I, 1835,

p. 252, PL XX, fig. 1.
Tachymenis peruviana Boulenger, Cat. Snakes, III, 1896, p. 118.

The specimen (C. M. No. 347) is a female, and was collected by Mr. J. D.
Haseman at Bom Jesus de Lapa, Bahia, Brazil.

griffin: ophidia from south America in carnegie museum

215

CotTNTs AND Measurements.

(No. 347.)

Anal 1/1

Scale-rows 19

Gastrosteges 146

Urosteges 48/48

Upper labials 8(4.5)

Preoculars 1

Postoculars 2

Temporals J2, 3

12,2

Total length in mm 193

Length of tail in mm 33

Genus Thamnodynastes Wagler.

74. Thamnodynastes nattereri (Mikan).

Coluber nattereri Mikan, Delect. Faun. Flor. Bras., 1820, Plate.
Thamnodynastes nattereri Boulenger, Cat. Snakes, III, 1896, p. 116.

C. M. No. 48, d^, Provincia del Sara, Bolivia, Jose Steinbach coll., Dec, 1912.
C. M. No. 339, cf , Santarem, Brazil, J. D. Haseman coll., December 7, 1909.
C. M. No. 361, 9 , Rio Doce, J. D. Haseman coll., May 20, 1908.

No. 339 was captured in the water of a swamp between the Rio Tapajos and
the Amazon. All except the two outer rows of scales are strongly keeled. It
conforms to the characters of T. nattereri except that there are seventeen instead
of nineteen rows of scales. No. 361 was captured in a mill-race.

Counts and Measurements.

Anal

Scale-rows. . .
Gastrosteges .
Urosteges . . . .
Upper labials .

Preoculars

Postoculars

Temporals

Total length in mm. . .
Length of tail in mm.

(No. 48.)

(No. 339.)

1/1

19

1/1
17

1/1
19

156

148

155

74/74
8(4.5)

1

76/76
8(4.5)

2

64/64

8(3.4.5)

_ -8(4.5)

1

2

2

2

2,3

2,3

2,3

568

588

581

142

156

112

(No. 361.)

216 memoirs of the carnegie museum

Series C. Proteroglypha.
Subfamily Elapin^ Boulenger.

Genus Elaps Schneider.
75. Elaps colombianus sp. nov.

Seven upper labials, third and fourth bordering the orbit; snout broad and
rounded, not projecting; sixth and seventh labials the largest of the series; rostral
just visible from above, much broader than deep; internasals much shorter and
narrower than prefrontals; first lower labials in contact behind the symphysial;
posterior nasal in contact with the preocular. Eye two-thirds to three-quarters
as long as its distance from the mouth; frontal once and a quarter to once and two-
fifths as long as broad, much broader than the supraoculars, slightly shorter than
its distance from the end of the snout (longer than this distance in the young speci-
men No. 2031), much shorter than the parietals; parietals longer than their distance
from the internasals; one 'pre- and two postoculars; temporals 1, 1; third upper
labial larger than the fourth; four lower labials in contact with the anterior chin-
shields, which are a little shorter than the posterior.

Scales in 15 rows; anal divided; gastrosteges 187-207; urosteges in 33-44 pairs.

Body with 12 to 14 black rings, three to four scales wide, edged with narrow
yellow rings one scale wide; the latter margined with narrow irregularly outlined
black rings. The last are variably developed, from a row of scales tipped with
black, to two or three rows of all black scales. The red interspaces are more than
twice as long as the length of the black triads, with usually black tipped scales.
Top of the head black from snout to parietals; posterior edges of parietals, the
temporals, fifth, sixth, and part of seventh upper labials crossed by a yellow band,
which passes partly or completely around the lower jaw and is widest at the sides
of the head. Anterior half of lower jaw black. The first black annulus is separated
from the parietals by not more than one scale. A narrow black ring is only faintly
indicated behind this first or nuchal annulus. The last annulus of the body crosses
the anal scales. On the tail are four to six black annuli, separated by narrow red
spaces, in which all the scales may be black-margined, or the central scales are
black, forming a narrow black bar which may extend only to the sides or across
the lower surface of the tail. The tail pattern is, thus, a broad black annulus, a
narrow red ring, a narrow black cross-bar or ring, a narrow red ring, a broad black
annulus, etc.

This species seems not distantly related to Elaps fulvius (Linnseus) .

griffin: ophidia from south America in carnegie museum

217

Type, C. M. No. 197; paratypes, C. M. Nos. 198, 2031, 2033.

There are four specimens representing this species in our collection, which
are cataloged as follows:

No. 197, 9 , Minca, Colombia, H. H. Smith coll., June.

No. 198, 9 , Bonda, Colombia, H. H. Smith coll., June.

No. 2031, cf , Cacagualito, Colombia, Mrs. H. H. Smith, no date.

No. 2033, 9 , Cacagualito, Colombia, Mrs. H. H. Smith, no date.

Counts and Measurements.

Anal

Scale-rows

Gastrosteges

Urosteges

Upper labials

Preoculars

Post oculars

Temporals

Total length in nun. .
Length of tail in mm
Annuli

(No. 197.)

(No. 198.)

(No. 2031.)

(No. 2033.)

1/1

1/1

1/1

1/1

15

15

15

15

207

206

187

203

33/33

36/36

44/44

33/33

7(3.4)

7(3.4)

7(3.4)

6(2.3)
7(3.4)

1

1

1

1

2

2

2

2

1,1

1,1

1,1

1,1

623

508

343

597

56

55

44

64

13+5

12+4

12+6

14+5

76. Elaps corallinus Wied.

Flaps corallinus Wied, N. Acta Ac. Leop.-CaroL, XI, 1820, p. 198, pi. IV.
Elaps corallinus Boulenger, Cat. Snakes, III, 1896, p. 420.

There are four specimens in the collection, as follows:

C. M. No. 199, d^, Cacagualito, Colombia, H. H. Smith coll., Oct. 3, 1901.

C. M. No. 261, 9 , LeBouteher Collection, South America.

C. M. No. 341, 9 , Sao Antonio de Guapore, J. D. Haseman coll., August 5, 1909.

C. M. No. 1236, c?, no data.

Counts and Measurements.

Anal

Scale-rows

Gastrosteges

Urosteges

Upper labials

Preoculars

Postoculars

Temporals

Total length in mm. .
Length of tail in mm
Annuli

218 MEMOIRS OF THE CARNEGIE MUSEUM

77. Elaps frontalis Dumeril & Bibron.

Flaps frontalis Dumeril & Bibron, Erp. Gen., VII, 1854, p. 1223.
Elaps frontalis Boulenger, Cat. Snakes, III, 1896, p. 427.

This species is represented by a female (No. 356) taken at Sete Lagoas by Mr.
J. D. Haseman, May 4, 1908.

Counts and Measurements.

(No. 356.)

Anal 1/1

Scale-rows 15

Gastrosteges 226

Urosteges 21/21

Upper labials 7(3.4)

Preoculars '. 1

Postoculars 2

Temporals 1,1

Total length in nun 726

Length of tail in mm 38

78. Elaps hoUandi* sp. nov. (Plate XXVIII, figs. 10-12.)

Seven upper labials, third and fourth bordering the orbit, the sixth largest,
seventh next in size; rostral just visible from above; internasals about half as long
as the prefrontals; first lower labials in contact behind the symphysial; posterior
nasal in contact with the preocular. Eye slightly longer than its distance from
the mouth; frontal considerably shorter than the parietals, shorter than its distance
from the end of the snout, once and a half as long as broad, hexagonal, considerably
wider than the supraocular; parietals longer than their distance from the inter-
nasals; rostral nearly twice as broad as deep; one pre- and two postoculars; temporals
1, 1, both large and broad; four lower labials in contact with the anterior chin-
shields, which are shorter than the posterior. Tail short, rather bluntly pointed.

Scales in 15 rows; anal divided; gastrosteges 185-204; urosteges in 17-22 pairs.

Body and tail with black annuli arranged in threes, a few of the scales of the
interspaces tipped with black. The central black annulus is little wider than the
other two, covering four to five and a half gastrosteges while the narrower annuli
cover three to four. On No. 206 the three annuli of the tail are similar to those of
the body; on No. 207 there are but two caudal annuli. The upper and lower
surfaces of the head are black as far as the fourth series of scales back of the pa-
rietals; the black of the upper surface is broken by a narrow yellow band passing

* Named in honor of Dr. W. J. Holland, Director of the Carnegie Museum, whose kind interest has
stimulated so many scientific workers.

griffin: ophidia from south America in carnegie museum 219

from lip to lip just behind the eyes. This band covers most of the fifth and sixth
upper labials, the upper postocular, the anterior temporal, the back of the supra-
oculars and frontal, and the anterior third of the parietals. The center of the
chin is light.

There are two specimens in the collection, cataloged as follows:
C. M. No. 206, type, cf , Bonda, Colombia, H. H. Smith coll., June, 1901.
C. M. No. 207, paratype, 9 , Bonda, Colombia, H. H. Smith coll., June, 1901.

Counts and Measurements.

(No. 206.) (No. 207.)

Anal 1/1 1/1

Scale-rows 15 15

Gastrostege.? 185 204

Urosteges 22/22 17/17

Upper labials 7(3.4) 7(3.4)

Preoculars 1 1

Postoculars 2 2

Temporals 1,1 • 1,1

Total length in mm. 366 318

Length of tail in nmi 30 17

Sets of annuli 7+1 8+1

79. Elaps mipartitus Dumeril & Bibron.

Elaps mipartitus Dumeril & Bibron, Erp. Gen., VII, 1854, p. 1220.
Elaps 7nipartitus Boulenger, Cat. Snakes, III, 1896, p. 431.

C. M. No. 208, 9 , Valparaiso, Colombia, H. H. Smith, April, 1901.

C. M. No. 209, 9 , Las Nubes, Colombia, H. H. Smith, December, 1901.

Our specimens differ from Boulenger's diagnosis is not having the anterior
temporal scale noticeably narrowed. In No. 209 the anterior and posterior tem-
poral shields are fused into one. In both specimens the frontal shield lacks a
little of being as long as its distance from the tip of the snout.

The head is black to back of the eyes, the black patch including all or nearly
all of the fourth upper labial, and half of the supraoculars and frontal. The
yellow band extends to the posterior ends of the parietals. The chin is yellow
with scattered, irregularly shaped, dark flecks. The following is copied from the
field label :

"This snake when alive has the head and tail almost precisely alike — bright
red. It coils up with its head underneath its tail and threatens with its tail."

220 MEMOIRS OF THE CARNEGIE MUSEUM

Counts and Measurements.

(No. 208.) (No. 209.)

Anal 1/1 1/1

Scale-row.s 15 15

Gastrosteges 252 249

Urosteges 27/27 24/24

Upper labials 7(3.4) 7(3.4)

Preoculars 1 1

Postoculars 2 1

2

Temporals 1,1 1

Total length in mm 361 477

Length of tail in mm 26 32

SO. Elaps narduccii Jan.
Flaps narduccii Jan, Arch. Zool. Anat. Phys., II, 1863, p. 222.
Elaps narduccii Boulenger, Cat. Snakes, III, 1896, p. 433.

C. M. No. 8, 9 , Provincia del Sera, Bolivia, 350 M., Steinbach coll., Jan., 1912.
C. M. No. 114, d', Santa Cruz de la Sierra, Bolivia, Steinbach coll.

Some variability in the proportions of the head shields of this species may be
deduced from the published descriptions. In our specimen No. 8 the width of the
frontal is equal to that of the supraocular, and the posterior chin -shields are longer
than the anterior. In No. 114 the frontal is much wider than the supraocular,
and the anterior chin-shields are the longer.

Counts and Measurements.

(No. 8.) (No. 114.)

Anal 1/1 1/1

Scale-rows 15 15

Gastrosteges 316 279

Urosteges 18/18 26/26

Upper labials 7(3.4) 7(3.4)

Preoculars 1 1

Postoculars 2 2

Temporals 1,1 2,1

1,1

Total length in mm 541 467

Length of tail in mm 21 30

81. Elaps princeps Boulenger.
Elaps princeps Boulenger, Ann. Mag. Nat. Hist. (7), Vol. XV, 1905, p. 456.
C. M. No. 126, cf , Santa Cruz de la Sierra, Bolivia, Steinbach coll.

This snake was collected at the same place and by the same person as the

griffin: ophidia from south America in carnegie museum 221

original specimens described by Boulenger. Our specimen is considerably larger
than any of those, but otherwise is very hke them. The Carnegie Museum speci-
men differs structurally only in that the eye is slightly larger (three-quarters as
long as its distance from the mouth) ; the frontal is once and a quarter as long as
broad, and the parietals are a little longer than their distance from the internasals.

Counts and Measurements.

(No. 126.)

Anal 1/1

Scale-rows 1^

Gastrosteges 221

Urosteges 22/22

Upper labials 7(3.4)

Preoculars 1

Postoculars 2

Temporals Ij 2

Total length in mm 1602

Length of tail in mm 65

Family AMBLYCEPHALID^ Gunther.

Genus Cochliophagus Dumeril & Bibron.

82. Cochliophagus catesbyi (Sentzen).

Coluber catesbyi Sentzen, Meyer's Zool. Archiv, II, 1796, p. 66.
Leptognathus catesbyi Boulenger, Cat. Snakes, III, 1896, p. 449.
C. M. No. 20, cf , Las Juntas, Bolivia, Steinbach coll., Dec, 1913.
C. M. No. 21, cT, Las Juntas, Bolivia, Steinbach coll., Dec, 1913.
C. M. No. 28, cf , Las Juntas, Bolivia, Steinbach coll., Dec, 1913.
C. M. No. 2034, cf , Brazil, no data.

Counts and Measurements.

(No. 20.)

Anal 1

Seale-rows ' 13

Gastrosteges 176

Urosteges , 94/94

Upper labials I f9(4.5.6)

' t8(3.4..5)

Preoculars

Postoculars

Temporals

Total length in mm. .
Length of tail in mm.

1
1,2
530
141

(No. 21.)

1

13

179

78/78
8(4.5)

2

2

1,2

208

50

(No. 28.)

1

13
175

87/87
18(4.5.6)
18(4.5)

2

2

1,2

186

45

(No. 2034.)

-1

13

193
93/93

8(4.5)

2

2

1,2

747

197*

'Tip of tail lost.

222

MEMOIRS OF THE CARNEGIE MUSEUM

Family VIPERID^ Bonaparte.

Subfamily Crotalin^ Oppel.

Genus Lachesis Daudin.

83. Lachesis lanceolatus (Lacepede).

Coluber lanceolatus Lacepede, Serp., II, 1789, pp. 80, 121, PI. V, fig. 1.
Lachesis lanceolatus Boulenger, Cat. Snakes, III, 1896, p. 535.

Counts and Measurements.

No. 43 9

" 121 cf

159. . .
244. . .
245...
246. . .
247. . .
248. . .
249. . .
250. . .
252. .
253...
254...
2.55. . .
257...
258. . ,
313...
372...
373. . .
2019..

cf
cf
cT
cf
9
<f
&
cf
9
cf
cf

&

cf
cf
9
cf

'S

<0 QD

23
23

25
27
25
23
25
25
27
25
23
25
25
25
25
27
25
27
23
25

o^

181
185

214
196
190
202
191
187
180
188
208
181
184
185
184

182
190
184
218

P5

60/60
66/66

70/70
53/.53
63/63
65/65
60/60
58/58
51/51

54/54
61/61
62/62
59/59
63/63

62/62
62/62
56/56
69/69

7
7
7
7
8
7
7
7
7
8
7
7
7
7

8

7

7
6

8
6
7
7
8
6
6
7
7
10
7
7
6

'7
6

^ S O C c3

■3 S 2 «3

«j a; S 03 =3

-5 q
o eg

550
690

395
1,045
805
820
960
840
885
840
832
923
705
975
857
880
535
595
369
565

-- a
3^.s

72
101

57
116
110
115
137
130
108

'oi

130
105
143
123

118
74
80
47
76

* Tip of tail lost.

The twenty specimens of this species in the collection are cataloged as follows :

C. M. No. 43, 9 , Provincia del Sara, Bolivia, 350 M., Steinbach coll., December,

1912.
C. M. No. 121, cf , Santa Cruz de la Sierra, Bolivia, Steinbach coll.
C. M. No. 159, d", Bonda, Colombia, 150 ft., H. H. Smith coll., August.
C. M. No. 244, cf , South America, LeBoutelier Collection.
C. M. No. 245, c?, South America, LeBoutelier Collection.
C. M. No. 246, cf , South America, Le Boutelier Collection.
C. M. No. 247, 9 , Chili, LeBoutelier Collection.
C. M. No. 248, cf , Colombia, LeBoutelier Collection.
C. M. No. 249, d", Brazil, LeBouteher Collection.
C. M. No. 250, cf , Venezuela, LeBoutelier Collection.

griffin: ophidia from south America in carnegie museum 223

C. M. No. 252, 9 , Peru, LeBoutelier Collection.

C. M. No. 253, d", Paraguay, LeBoutelier Collection.

C. M. No. 254, cf , Paraguay, LeBoutelier Collection.

C. M. No. 255, cf , Paraguay, LeBoutelier Collection.

C. M. No. 257, cf, Montevideo, Uruguay, LeBoutelier Collection.

C. M. No. 258, cf , Argentina (?), LeBoutelier Collection.

C. M. No. 313, 9 , Santa Cruz de la Sierra, Bolivia, Steinbach coll.

C. M. No. 372, d", Tarma, Peru, Miss Lola Vance coll.

C. M. No. 373, Tarma, Peru, Miss Lola Vance coll.

C. M. No. 2019, c?, Cacagualito, Colombia, 1500 ft., Mrs. H. H. Smith coll.

84. Lachesis lansbergi (Schlegel).

Trigonocephalus lansbergii Schlegel, Mag. de ZooL, 1841, Rept. pi. I.
Lachesis lansbergii Boulenger, Cat. Snakes, III, 1896, p. 546.

There are eighteen specimens of this species in the collection all of which came
from the Province of Santa Marta in Colombia, although the last four mentioned
in the following hst have no data accompanying them. These belong to the lot,
the numbers of which run from C. M. No. 1841 to 1873, mentioned in the intro-
ductory notes.

C. M. No. 151, d', Bonda, Colombia, H. H. Smith coll., July.

C. M. No. 152, 9 , Bonda, Colombia, H. H. Smith coll.. May.

C. M. No. 153, &, Bonda, Colombia, H. H. Smith coll., May.

C. M. No. 154, 9 , Bonda, Colombia, H. H. Smith coll., August 31.

C. M. No. 155, d", Bonda, Colombia, H. H. Smith coll., August 29.

C. M. No. 156, 9 , Bonda, Colombia, H. H. Smith coll., June.

C. M. No. 157, 9 , Bonda, Colombia, H. H. Smith coll.. May.

C. M. No. 158, 9 , Bonda, Colombia, H. H. Smith coll., June.

C. M. No. 160, 9 , Bonda, Colombia, H. H. Smith coll.. May.

C. M. No. 185, d, Bonda, Colombia, H. H. Smith coll., August.

C. M. No. 1093, 9 , Bonda, Colombia, H. H. Smith coll., August. -

C. M. No. 1851, d, South America (Colombia) (Mr. & Mrs. H. H. Smith coll.).

C. M. No. 1852, 9 , South America (Colombia) (Mr. & Mrs. H. H. Smith coll.).

C. M. No. 1853, 9 , South America (Colombia) (Mr. & Mrs. H. H. Smith coll.).

C. M. No. 1855, d, South America (Colombia) (Mr. & Mrs. H. H. Smith coll.).

C. M. No. 2016, 9 , Cacagualito, Mrs. H. H. Smith coll., no date.

C. M. No. 2017, d, Cacagualito, Mrs. H. H. Smith coll., no date.

C. M. No. 2018, 9 , Cacagualito, Mrs. H. H. Smith coU.^ no date.

224

MEMOIRS OF THE CARNEGIE MUSEUM

Our Colombian series of L. lansbergi compels a modification of the formula of
the species to include a wider variability, as follows:

Scales in 23-27 rows; anal single; gastrosteges 147-159; urosteges 28-36, as
shown by the accompanj'ing table.

Counts and Measukembnts.

1-

i

Anal.

Scale-
rows.

Gastro-
steges.

Uro-
steges.

Upper
Labials.

Scales
between

Supra-
oculars.

Scales
between
Subocu-
lars and
Labials.

Total
Length
in Mm.

Length

Tail
in Mm.

No. 151

cf

1 25

158

34

9

6 1

376

45

" 152

9

1 25

152

32

9

5 1

412

49

" 153

cf

1 25

155

35

9

5 1

317

38

" 154

9

1 25

153

35

9

5 1

440

56

" 155

cT

1 23

147

33

9

4 1

374

44

" 156

9

1 25

147

28

10

3 1

288

27

" 157

9

1 25

159

31

9

6 1

301

32

" 158

9

1 1 23

153

33

9

6 1

399

47

" 160

9

1 i 23

153

29

9

4 1

254

24

" 185

cf

1 i 25

154

31

9

5 1

452

53

" 1093....

9

1 1 25

152

31

9

5 1

570

62

" 1851...

cf

1

9

6 1

*

" 1852...

9

1 25

i58

28

Ul

5 1

290

26

" 1853....

9

1 1 25

151

29

9

5 1

409

41

" 1855....

cT

1 ' 25

151

31

9

5 1

196

19

" 2016. ...

9

1 23

154

34

/ 9
110

4 1

436

55

" 2017 ..

cf

1 23-24

154

36

f 9
110

5 1

225

23

" 2018...

9

1 25

151

32

9

5 1

420

50

'■ Mutilated.

85. Lachesis mutus (Linnaeus).

Crotalus mutus Linn^us, Syst. Nat., Ed. XII, 1766, I, p. 373.
Lachesis mutus Boulenger, Cat. Snakes, III, 1896, p. 533.

We have but one specimen representing this species in the collection (C. M.
No. 125). It consists only of the head and tail of a very large individual, collected
by Jose Steinbach at Santa Cruz do la Sierra, Bolivia.

86. Lachesis neuwiedi (Wagler).
Bothrops neuwiedi Wagler, in Spix, Serp. Bras., 1824, p. 56, PL XXII, fig. 1.
Lachesis neuwiedi Boulenger, Cat. Snakes, III, 1896, p. 542.

The specimens in the collection are listed as follows :

No. 1, cf , Province del Sara, Bolivia, J. Steinbach coll., Sept., 1911.
No. 4, 9 , Province del Sara, Bolivia, J. Steinbach coll., Jan., 1912.
No. 34, cf , Province del Sara, Bolivia, J. Steinbach coll., Sept., 1913.
No. 35, cf , Province del Sara, Bolivia, J. Steinbach coll., Sept., 1913.

griffin: ophidia from south America in carnegie museum

225

No.

38,

cT,

Province del Sara,

Bolivia,

J.

Steinbach coll..

Sept.

, 1913

No.

40,

9,

Province del Sara,

Bolivia,

J.

Steinbach coll.

Sept.

, 1913

No.

46,

9,

Province del Sara,

Bolivia,

J.

Steinbach coll.

Dec.

1912.

No.

49,

d",

Province del Sara,

Bolivia,

J.

Steinbach coll.

Dec.

1912.

No.

50,

^

Province del Sara,

Bolivia,

J.

Steinbach coll.

Jan.,

1913.

No.

54,

9,

Province del Sara,

Bolivia,

J.

Steinbach coll.

Jan.,

1913.

No.

55,

9

Province del Sara,

Bolivia,

J.

Steinbach coll.

Oct.,

1911.

No.

58,

9,

Province del Sara,

Bolivia,

J.

Steinbach coll.

Oct.,

1911.

Counts and Measurements.

Be

No. 1.

4.
34.
35.
38.
40.

46.

49.
50.

54. , . .

55. . . .

58. . . .

60. . . ,

61...,

67....

68....

69....
119....
120....

122

123..!.

317...
318...
364. . .

9
9

cT

9
9
9
cf
cf
9
cf
cf

9
9

9

cf
cf

9

03 "

23

25
24
25
25
24

25

25
25

23
25
27
25
25
23
23
25
25
25
23
23

23
23
25

1^

Ok

169

170
172
171
178
180

166

177
175

174
179
183
180
181
170
171
179
184
176
167
175

174
184
171

48/48

45/45
49/49
41/41
45/45
46/46

41/41

53/53
50/50

51/51
49/49
42/42
53/53
45/45
49/49
50/50
46/46
49/49
45/45
51/51
45/45

52/52

47/47
43/43

P..2

8

w

8

8

10

8

\l

8

* § rt £

» § Sa.2

2

{\

2
2
2
1
2
1
2
2
2
2
1

i^

2
2

{\

-5 a

442

827
288
467
775
438

438

410
558

380
425
815

i-1 .9

61

93
40
55

83
52

48

53

84

548

70

264

34

502

53

580

83

630

82

365

50

334

45

277

33

455

57

334

29

897

109

314

40

52
49
80

No. 60, cf , Province del Sara, Bolivia, J.
No. 61, cf. Province del Sara, Bolivia, J.
No. 67, 9 , Province del Sara, Bolivia, J.
No. 68, cf , Province del Sara, Bohvia, J.
No. 69, cf , Province del Sara, Bolivia, J.
No. 119, cf. Province del Sara, Bolivia, J
No. 120, 9 , Province del Sara, Bolivia, J

Steinbach coll., Oct., 1911.
Steinbach coll., Oct., 1911.
Steinbach coll., Oct., 1911.
Steinbach coll., Oct., 1911.
Steinbach coll., Oct., 1911.
. Steinbach coll., no date.
. Steinbach coll., no date.

226 MEMOIRS OF THE CARNEGIE MUSEUM

No. 122, 9 , Province del Sara, Bolivia, J. Steinbach coll., no date.

No. 123, 9 , Province del Sara, Bolivia, J. Steinbach coll., no date.

No. 317, cf , Asumpcion, Bolivia, J. D. Haseman coll., June 14, 1909.

No. 318, d', Campos de Matto Grosso, Braz., Haseman coll., June 20, 1909.

No. 364, 9 , Sao Joao del Rey, Minas Geraes, Brazil, Haseman, no date.

87. Lachesis peruvianus Boulenger.
Lachesis peruvianus Boulenger, Ann. & Mag. Nat. Hist. (7), XII, 1903, p. 334.
The Museum has only one example of this species, a female (C. M. No. 373),
which was collected by Miss Lola Vance at Tarma, Peru (elev. 6,000 ft.).

Counts and Measurements.

(No. 373.)

Anal 1

Scale-rows 23

Gastrosteges 184

Urosteges 5 + 51/51

Upper labials 7

Total length in mm • 360

Length of tail in mm 47

Genus Crotalus Linnaeus.

88. Crotalus terrificus (Laurenti).

Caudisona terrifica Laurenti, Syn. Rept., 1758, p. 93.
Crotalus terrificus Boulenger, Cat. Snakes, III, 1896, p. 573.

There are thirteen specimens of this species in the collection, as follows:

C. M. No. 30, 9 , Provincia del Sara, Bolivia, Jose Steinbach, February, 1913.

C. M. No. 63, 9 , Provincia del Sara, Bolivia, Jose Steinbach, October, 1911.

C. M. No. 118, 9 , Santa Cruz de la Sierra, Bolivia, Steinbach, no date.

C. M. No. 124, Santa Cruz de la Sierra, Bolivia, Steinbach, head only, mutilated.

C. M. No. 161, 9 , Bonda, Colombia, H. H. Smith, May.

C. M. No. 162, 9 , Bonda, Colombia, H. H. Smith, May.

C. M. No. 259, cf , Argentina, LeBoutelier Collection.

C. M. No. 276, 9 juv.. South America, LeBouteUer Collection.

C. M. No. 1854, d^. South America, H. H. Smith Coll. (?).

C. M. No. 2012, cf , Cacagualito, Colombia, Mrs. H. H. Smith.

C. M. No. 2013, 9 , Cacagualito, Colombia, Mrs. H. H. Smith.

C. M. No. 2014, 9 , Cacagualito, Colombia, Mrs. H. H. Smith.

C. M. No. 2015, 9 , Cacagualito, Colombia, Mrs. H. H. Smith.

griffin: ophidia from south America in carnegie museum 227

Counts and

Measurements.

Number.

Anal.

Scale-
rows.

Gastro-
steges.

Urosteges.

Upper
Labials.

Scale-rows
Between
Supra-
oculars.

Scale-rows
Between
Eye and
Supra-
labials.

Total
Length
in Mm.

Length of

Tail in

Mm.

Remarks.

30

63
118

124

161
162
259
276

1854
2012
2013
2014

2015

1

1

1

1
1
1
1

1

1
1
1

1

27

29
29

29

27
27
27

29
25

29
25

29

161

176
175

173
176

170
161
174
168

179

27+2/2

19 + 2/2
21 + 1/1

25+3/3
23 + 1/1

20 + 1/1
22+2/2

29+1/1
30

27 + 1/1
30+1/1

30 + 1/1

13
11
14
14
13
14
15
13
13
14
16
17
13
14
13
15
14
14

2

3
5

3

3
3
3
2

3
3
3
3

3

3

3
4

4

3
3
3
3

3
3
3
3

3

620

598
376

1180+
1100
1149
295

364
390
420
410

390

52

46
30

178

100

115

23

36
41
37
43

40

Four rattles.

Three rattles.
One rattle.

Head only.

Ten rattles.

Skin.

Nine rattles.

Young.

Young.
One rattle.
One rattle.
One rattle.

One rattle.

In addition to the specimens cataloged above there are contained in the Car-
negie Museum three specimens which were received in the LeBouteUer Collection,
which are very probably wrongly attributed to the localities given on the original
labels. One is a specimen of Matrix fasciata fasciata (Linnajus), C. M. No. 262,
labeled as having been obtained in the " sierras of Bolivia." This species has been
foimd as far south as Costa Rica, but has never been reported hitherto from any
point south of that country, and in view of the careless labeling of the LeBouteUer
Collection I hesitate to record the species as having been actually obtained at the
locality indicated.

There are also two specimens of Agkistrodon piscivorus (Lacepede), C. M. Nos.
251 and 260, the first labeled "Peru," the second labeled "Brazil." The occur-
rence of this ophidian in the localities cited is more than doubtful.

228 MEMOIRS OF THE CARNEGIE MUSEUM

EXPLANATION OF PLATE XXVIIL

Fig. 1. Atradus taniatus Griffin. Type, C. M. No. 117. X 4.

Fig. 2. Do. Dorsal View of Head. X 4.

Fig. 3. Do. Ventral View of Head. X 4.

Fig. 4. Tropidodipsas spilogaster Griffin. Type, C. M. No. 5085. X 2.

Fig. 5. Do. Dorsal View of Head. X 2.

Fig. 6. Do. Ventral View of Head. X 2.

Fig. 7. Clelia euprepa Griffin. Type, C. M. No. 109. X 2.

Fig. 8. Do. Dorsal View of Head. X 2.

Fig. 9. Do. Ventral View of Head. X 2.

Fig. 10. Flaps hollandi Griffin. Type, C. M. No. 206. X 2.

Fig. 11. Do. Dorsal View of Head. X 2.

Fig. 12. Do. Ventral View of Head. X 2.

Memoirs Carnegie Museum, Vol. VII.

Plate XXVIII.

South American Ophidia.

21. Minute Book of the Virginia Court Held at

Fort Dunmore (Pittsburgh) for the District

of West Augusta, 1775-1776. Crumrine 90

22. Minute Book of the Virginia Court Held for

Yohogania County, first at Augusta Town
(now Washington, Pa.), and afterward on the
Andrew Heath Farm near West Elizabeth,
1776-1780. Cbdmrine 2.26

23. Minute or Order Book of the Virginia Court

Held for Ohio County, Virginia, at Black's
Cabin (Now West Liberty, W. Va.), &c.
Crumrine 1.50

24. The Records of Deeds for the District of West

Augusta, Virginia, for the Court Held at Fort
Dunmore, &c. Crumrine 1.75

25. Astropecten (?) montanus, &c. Douglass 10

26. Astrodon (Pleurocoelus) in the Atlantosaurus

Beds of Wyoming. Hatcher. {Out of Print.)

27. Osteology of the Limicolae. Shufeldt 1.00

28. New Vertebrates from the Montana Tertiary.

Douglass 1.25

29. New Genus and Species of Tortoise from the

Jurassic of Colorado. O. P. Hat 10

30. Osteology of Oxydactylus. Peterson T.OO

31. Birds of Erie and Presque Isle. Todd 75

32. J. B. Hatcher. By W. J. Holland 15

33. Tropidoleptus Fauna at Canandaigrua Iiake, etc.

Raymond. {Out of print.)

34. Two Turtles from the Judith River Beds of

Montana. O. P. Hay. {Out of print.)

35. Preliminary List of Hemiptera of Western

Pennsylvania. Wirtner. {Scarce.) 50

36. Trilobites of the Chazy Limestone. Raymond. 1.00

37. Crawfishes of Western Pennsylvania. Ort-

MANN 1.00

38. The Geology of Southwestern Montana. Doug-

lass 40

39. New Crocodile from the Jurassic of Wyoming.

Holland 10

40. Procambarus, a New Subgenus of the Genus

Cambarus. Ortmann 15

41. Presentation of Reproduction of Diplodocus Car-

negei to the British Museum. Holland 15

42. Birds Collected near Mombasa by William Do-

herty. Holland 20

43. Hyoid Bone in Mastodon Americanus. Hol-

land 10

44. Additions to Flora of Allegheny County, Pa.

Jennings 15

45. New Species of BJieiffia. Jennings 05

46. Occurrence of Triglochin palustris in Pa. Jen-

nings 05

47. New Species of Ibidium. Jennings 10

48. Agate Spring Fossil Quarry. Peterson 10

49. Two New Birds from British E. Africa. Ober-

holser 05

50. The Chazy Formation and Its Fauna. Ray-

mond 1.50

51. A New American Cybele. Narraway and Ray-

mond 15

52. Plastron of the Protosteginae. Wieland 15

53. New Species of Testudo from the Miocene, etc.

0. P. Hat 25

54. Miocene Beds of W. Nebraska and E. Wyoming

and Their Vertebrate Faunae. Peterson 1.00

55. New Species of Lonicera from Pa. Jennings. .05

56. Meryocochcerus, etc. Douglass.

57. New Merycoidodonts. Douglass. (Nos. 56

and 57 sold together.) 1.00

58. Further Collections of Fishes from Paraguay.

Eigenmann, McAtee, and Ward 1.25

59. Undetermined Element in the Osteology of the

Mosasauridae. Holland 20

60. Gasteropoda of the Chazy. Raymond 1.36

61. Occurrence of Wynea Americana in Pa. Jen-

nings 05

62. Account of Pleistocene Fauna Discovered at

Frankstown, Pa. Holland 10

63. Vertebrate Fossils from the Vicinity of Pitts-

burgh, Pa. Case 15

64. lUaenidae from the Black Elver Limestone near

Ottawa, Canada. Raymond and Narraway. . .20

65. Rhinoceroses from the Oligocene and Miocene

of North Dakota and Montana. Douglass . . .25

66. Fossil Horses from North Dakota. Douglass. .30

67. Oligocene Lizards. Douglass 20

68. The Type of Stenomylus gracilis. Peterson . . .25

69. New Species of Birds from Costa Rica, etc.

Carrikeb 10

70. Costa Ricau Formicariidae. Carrikeb 05

71. Vertebrate Fossils from the Fort Union Beds.

Douglass 45

72. List of the Lepidoptera of Western Pa., etc.

Engel 1.25

73. Fauna of Upper Devonian in Montana. Fossils

of Red Shales. Raymond 60

74. New Species of Procamelus from Upper Mio-

cene of Montana. Douglass 30

75. Sections of the Conemaugh Series between

Pittsburgh and Latrobe, Pa. Raymond 35

76. List of the Unionidae of Western Pa. Ortmann .50

77. Geological Reconnaissance in North Dakota,

Montana, and Idaho. Douglass 1.00

78. Botanical Survey of Presque Isle. Jennings . . 2.75

79. Sesqui-Centennial (Pittsburgh) Relics. Stewart .45

80. Dromomeryx, New Genus of American Rumi-

nants. Douglass 30

81. Fossils from Glacial Drift and Devonian and

Mississippian near MeadvlUe, Pa. Millard. . .10

82. New Species of Helodus. Eastman 05

83. Charles Chauncey Mellor. Holland 15

84. Reports of Expedition to British Guiana, etc.

Eigenmann 50

85. Reports of Expedition to British Guiana, etc.

DUKBIN 25

86. Odonata of the Neotropical Region, Exclusive

of Mexico and Central America. Calvert . . . 2.25

87. Deinosuchus hatcheri, a New Genus and Species

of Crocodile. Holland 20

88. Reports on Expedition to British Guiana, etc.

Blosseb 15

89. Description of Some New Titanotheres from

the Uinta. Douglass 26

90. Annotated List of the Birds of Costa Rica In-

cluding Cocos Island. Carriker 3.00

91. Geology of the Coast of the State of Alagoas,

Brazil. Brannee 40

92. Fossil Fishes from the Bituminous Shales at Ri-

acho Doce, Alagoas, Brazil. Jordan 55

93. Ordovician Trilobites, No. II. Asaphidae from

the Beekmantown. Raymond 35

94. Ordovician Trilobites, No. HI. Raymond &

Nabrawat ' 35

95. Ordovician Tiilobites, No. IV. Raymond 40

96. Fishes from Irkutsk, Siberia. Jordan and

Thompson 30

97. South American Tetrigidae. Brtjner 1.00

98. Fauna of the Allegheny and Conemaugh Series

in Western Pa. Raymond 30

99. Ichthyological Survey About the San Juan

Islands, Washington. Staeks 75

100. New Species of Pygidium, Eigenmann 10

101. The Erachiopoda and Ostracoda of the Chazy.

Raymond 50

102. A New Camel from the Miocene of Western

Nebraska. Peterson 15

103. Skeleton of Stenomylus hitchcocki, etc. Peter-

son 15

104. Skeleton of Diceratherium cooki. Peterson . . .15

105. Carnegie Museum Expedition to Central South

America, 1907-1910. Holland 15

106. Report Upon the Expedition of the Carnegie

Museum to Central South America. Hase-
MAN & Eigenmann 25

107. New Species of Fishes, etc., from Central South

America. Haseman 50

108. Annotated Catalog of Cichlid Fishes from Cen-

tral South America. Haseman 1.25

109. New Species of Fishes from the Bio Iguassu.

Haseman 65

110. Ornithology of the Bahama Islands. Todd &

WORTHINGTON 75

112. South American Acridoidea. I. Bbuner .... 1.00

113. Species of Hasemania, Hyphessobrycon, and

Hemigrammus. Ellis 25

114. New Characins in the Carnegie Museum. Eigen-

mann 30

115. Jurassic Saurian Remains Ingested within Fish.

Eastman 20

116. Autograph Letter of U. S. Grant to Edwin M.

Stanton. Holland 15

117. Albert J. Barr. By W. J. Holland 10

118. Seventeen New Neotropical Birds. Todd 25

119. Dr. David Alter, the First Discoverer of Spec-

tnim Analysis. Holland 10

120. Two Mummy Labels. Allen 10

121. The Families and Genera of Najades. Ort-

MANN 1.00

122. Group of Stenomylins in the Carnegie Museum.

Peterson 25

123. Tertiary Fish-remains from Spanish Guinea.

Eastman 25

124. Plated Nematognaths. Ellis 65

125. New Species of Cambarus from the Isle of

Pines. Ortmann 15

126. Sedmu Camegiei, a New Species of the Family

Crassulacese, etc. Hamet 10

127. Two New Species of Fishes from Peru. Eigen-

mann 15

128. South American Locusts (Acridoidea). II.

Brunee 75

129. Revision of the Genus Chsmepelia. Todd 85

130. New Genus and Species of Abyssinian Rodents.

Childs Frick 20

131. New Titanothere from the Uinta. Peterson . . .20

132. Small Titanothere from the Lower Uinta. Peter-

son 10

133. Osteology of Lasiopyga and Callithrix, etc.

Shupeldt _ 25

134. New Rhynchocephalian from Solenhofen. Grier ,15

135. Scales of South American Characinid Fishes.

Cockerell .25

136. Skeleton of Platigonus Leptorhinus. Peter-

son ; 10

137. Lichens Collected in the Thunder Bay District,

Ontario. R. Heber Howe, Jb 10

138. Preliminary List of Fossil Plants in the Roof

of the Pittsburgh Coal. Griee 10

139. Undescribed Remains of the Uinta Titanothere

Dolichorhinus. Peterson 15

140. Triassic Fishes Belonging to the Catopteridae

and Semionotidse. Eastman 15

141. Osteology of Promerycochoerus. Peterson ... .40

142. Correction of a Generic Name. Peterson 05

143. Skull of Bison Crassicomis. Holland 10

144. Serrasalminae and Mylinae. Eigenmann 50

145. Heads and Tails: Notes on Sauropod Dinosaurs.

Holland 15

146. Dipterus Remains from Upper Devonian of

Colorado. Eastman 10

147. Notes on Tropical American Tettigonoidea (Lo-

custodea). Bruner 1.65

148. New Species of Tortoise from Jurassic of

Utah. Gilmoee 15

149. Fauna of Upper Devonian in Montana.

Haynes 45

150. New Sphagebranchus from Bahamas. Eigen-

mann 07

151. Marine Fishes from Colombia and Ecuador.

Wilson 15

152. Apareiodon, a New Genus of Characid Fishes.

Eigenmann 18

1.j3. New and Rare Fishes from S. American Rivers.

Eigenmann , .25

154. Three New Species of Characid Fishes. Eigen-

mann & Henn 12

155. The Species of Salminus. Eigenmann 7

156. South American PoecUiid Fishes. Henn 40

157. A New Species of Apatosaurus. Holland 7

158. Birds of the Isle of Pines. Todd 70

159. Reptiles and Amphibians of the Isle of Pines.

Barbour 20

160. Land and Fresh Water Shells of the Isle of

Pines. Henderson 20

161. Pelecypoda of the Chazy. Raymond 45

162. S. American Crickets, Gryllotalpoidea and Ache-

toidea. Bruner 1.30

163. Preliminary Catalog of N. American Sphaeri-

idae. Sterki 75

164. Directions for Collecting Sphseriidae and

Aquatic Gastropods. Sterki 10

165. Lepidoptera of the Isle cf Pines. Holland 60

166. Odonata of the Isle of Pines. Kahl 15

167. A Trip to Islands in Lake Erie. Goodeich 15

168. Land-Shells of Islands at Western End of Lake

Erie, etc. Clapp 20

169. Orthoptera of the Isle of Pines. Holland &

Kahl 15

MAY 28 19)7

Publications of the Carnegie Museuvx, Serial No. 94

MEMOIRS

OF THE

OAKNEGIE MUSEUM.

VOL. VII. NO. 4.

W. J. HOLLAND, Editor.

PIMELODELLA AND TYPHLOBAGEUS

By CARL H. EIGENMANN.

PITTSBURGH.
Published by the Authoeity of the Boabd of Trustees op the
CARNEGIE INSTITUTE.
April, 1917.

— n

HI

PRICE LIST OF PUBLICATIONS OF THE CARNEGIE MUSEUM

ANNUAL REPORTS OF THE DIRECTOR

1898, 30c. (scarce); 1899, 25c.; 1900, 30c. (scarce); 1901-16, 25c. each.

REPORTS or PROCEEDINGS OF FOUNDER'S DAY

1898-1916, 35c. each.

ANNALS OF THE CARNEGIE MUSEUM

The Annals are supplied to those who subscribe in advance in parts (paper-bound), as published, @ $3.50 per volume,
Vols. I-X, 1901-1916, bound in green cloth @ $4.00; bound in i Morocco @ $4.50.

MEMOIRS OF THE CARNEGIE MUSEUM

The MeAuoirs are supplied to those who subscribe in advance in parts (paper-bound), as published, at $10 per volume;
bound in green buckram at $10.75, and in half-morocco at $12.00.

"VOL. L (1901-3)

No. 1. Diplodocus, Its Osteology, Taxonomy, etc. No. 3. Osteology of the Stegaaopodes. Shufeldt $2.76

Hatcher $3.00 ' ' 4. Classification of Chalcid Flies. Ashmead. . 6.60

' ' 2. Oligocene Canidse. Hatcher 1.50

VOL. n. (1904-6)

No. 1. Osteology of Haplocanthosairus, etc. Hatchek $2.00 " 6. Osteology of Diplodocus. Holland 1.50

' ' 2. Osteology of Baptanodon. Gilmoke 1.50 ' ' 7. Osteology of Protostega. Wieland 75

" 3. Fossil Avian Eemains from Annissan. East- " 8. New SulUine Remains from the Miocene of

MAN 1.00 Nebraska. Peterson 75

' ' 4. New Bodeuts and Discussion of Origin of • ' 9. Notes on the Osteology of Baptanodon, etc.

Daemonelix. Peterson 1.75 Gilmore 1.00

No. 5. Tertiary of Montana. Douglass $1.00 ' ' 10. The Crawfishes of Pennsylvania. Ortmann 4.00

VOL. in.

No. 1. Archaeological Investigations in Costa Rica. No. 2. Osteology of Moropus. Holland and Peter-

Hartman $0.00 SON $6.00

VOL. IV.

No. 1. Early Chinese Writing. Chalpant $3.00 No. 5. New Carnivores from the Miocene of West-

' ' 2. Formosan Fishes. Jordan 25 ern Nebraska. Peterson $1.50

' ' 3. Entelodcntidae. Peterson 2.50 " 6. Monograph of the Najades of Pennsylvania.

' ' 4. Fishes of Formosa. Jordan and Richard- Pts. I and II. Ortmann 2.50

SON 1.26 ' ' 7. Catalog Bocene Fishes from Monte Bolea in

Carnegie Museum. Eastman 3.00

VOL. V.

The Fresh Water Fishes of British Guiana. Eige.n'mann. $10 unbound; $10.75 cloth; $12. GO J morocco.

VOL. VI.

No. 1. Fishes from Korea. Jordan and Metz $1.50 No. 4. Fishes obtained in Japan. 1911. Jordan

' ' 2. Lantern-fishes of Japan. Gilbert 1.00 and Thompson $3.50

" 3. Gymnotid Eels of Tropical America. Max Nos. 5-7. Fossil Fishes in the Carnegie Museum.

Mates Ellis 1.50 Parts II-IV. Eastman 5.00

VII.

No. 1. The Cheirodontinae. Eigenmann $3.50 No. 3. Ophidia from S. America. Qrippin 2.00

' ' 2. Turtles of Uinta Formation. Gilmore 2.00 No. 4. PimelodeUa and Typhlobagrus 2.00

REPRINTS FROM THE ANNALS OF THE CARNEGIE MUSEUM

Nos. 1-12. See preceding lists. (Out of print.) 16. Osteology of Psittaci. Shufeldt 35

12. Skeleton of Titanotherium dispar. Hatcher. . $ .36 17. Annotated Catalogue of Genus Partula. H. H.

13. Fore Iiimbs and Manus of Brontosaurus. „ "^"^, '„'■ ^' '.'■," ]. U ,„

18. Two New Bahaman Lepidoptera. Holland ... .10
Hatcher 60 ^g Elosaurus Parvus. Peterson & Gilmore 15

14. The Trachodontidse. Hatcher 25 20. Boundary Controversy Between Pennsylvania

15. New Pennsylvania Thorns. Ashe 25 and Virginia, 1748-1785. Crumrine 30

MAY 28 1917

MEMOIRS

OF THE

OAENEGIE MUSEUM

Vol. VII. No. 4.

PIMELODELLA AND TYPHLOBAGRUS.^

By Carl H. Eigenmann.

(Plates XXIX-XXXV.)

Pimelodella and Typhlobagrus are two closely related genera of Siluridse of
the fresh waters of South America. Pimelodella is a widely distributed genus
with many species. Typhlobagrus is an offshoot from Pimelodella. Eyes have
disappeared in this genus and it is confined, so far as known, to the caves near
Iporanga in southeastern Brazil.

Pimelodella Eigenmann & Eigenmann.

Pseudorhamdia Steindachner (non Bleeker), Sb. Akad. Wiss. Wien, LXXIV,

1876, Susswasserf. Slidostl. Bras., Ill, p. 46 (lateristriga) .
Pimelodella Eigenmann & Eigenmann, Proc. Cal. Acad. Sci., (2), I, 1888, p. 131;

Occasional Papers Cal. Acad. Sci., I, 1890, pp. 99 and 147.

Type, Pimelodus cristatus MiiUer & Troschel.

Nares remote; teeth viUiform, in bands; gill-membranes free from the isthmus;
dorsal short, with a feeble, pungent spine; anal short, with 11-15 rays; pectoral
with a strong i^ungent spine variously armed with thorn-like teeth on its posterior
(inner) edge; a long, adnate, adipose fin; caudal fin deeply forked, one or the other
lobe frequently wider, or longer; well-developed maxillary barbels reaching to
end of pectoral, or beyond the caudal; two pairs of mental barbels, sometimes in a
nearly straight line; a frontal and a parietal fontanel, the latter reaching to the

^ Contribution from the Zoological Laboratory of Indiana University, No. 154.

229

230

MEMOIRS OF THE CARNEGIE MUSEUM

base of the occipital process, which is narrow and reaches, or nearly reaches, the
plate in front of the dorsal; humeral process spine-like; roof of mouth without teeth;
head covered with thin skin.

Distribution. — All rivers from Buenos Aires to Guiana and Venezuela and
to the base of the Andes (1,800 feet in places); west of the Andes from northern
Peru to the Chagres river in Panama.

DiSTBIBUTION OF PiMELODELLA AND TyPHLOBAGRUS.

6
J
"S

CO

o

s

^3
3

'3.

M

O

s

•1

2
-§£

^ a

S

OJ 3

i-s a
'" 2
oPh

0)

1

-S

W
'^ .

a

be

3

*w

a

(£

ft

m

3
be

«

O

s

O

MX.

Is

11

s

+
+

'n

s

o

d

pa

a
o

M

p

3-;

.2 a
Ch a

s_^ o
.9 =3

4

W

□
c3

c3

.2
3

a

a

m

i

0

o

a

03
«

a

a.-

■a

d

0)
H

"i .

eS

a

c3
S

P-(

o

■g

u

(2

3
^^
3

O

2

c3

Q
o

2

Is

11

.2S
2 ^

2

1

3
O

o .

S

1

en

-a
o

CTistatfi

+

+
+
+

+

+'

+

steindachneri

+

+

+

+

+

TOCCCB

+
+
+

vfctiiiifpTfi

+

hasetnaui

+

+

+

weonJovfi

+

hvfisilicTis i-^

+

+

+

+

+

TYietCR

+

holivicLTici

+

Inieristriga

viitata

+
+

'+'

+
+

itapicuruensis ....

+

eigenmanni

+
+

hdTiti

tTdnsitovid

+

+
+

meeki . . . .

griffini

t

mucosa ' . .

+

grisea ...

+

+

+
+

modesta . . '

yuncennis

+

chagresi

?

+

+

+

elongata . .

+
+

+

eutcenia

+

+

T. kronei. . . .

+

2

1

1

T

3

5

6

4

5

1

3

3

1

1

1

2

2

4

1

2

1

7

1

The species of this well-marked genus differ from each other in the length
of the adipose fin, the length of the barbels, the length of the pectoral spine, the

eigenmann: pimelodella and typhlobagrus 231

armature of the pectoral spine, the shape of the caudal lobes, and a number of
other minor characters.

The length of the barbels in the same species differs with age. In the young
the barbels are relatively short. They grow disproportionately longer with the
growth of the fish and then lag behind again in their increase in length. The
length of the barbels of the same species not only differs with age, but sometimes
also with locality. The pectoral spines also vary somewhat with growth. With
age the thorns increase in number by the addition of new ones toward the tip
and come to occupy a larger portion of the length of the spine. The spine in-
creases in length by the addition of new sections at the end which are marked off
by notches or hooks on the outer margin of the spine. As the color and the size
of the eye also vary with age and with locality, and, as all of these characters vary
independently, the defining of species of Pimelodella becomes a delicate and diffi-
cult task. The following definitions, especially as far as they concern species of
the Amazon Basin, are therefore more or less tentative.

Most of the species are small. The largest species of the Atlantic slope is
P. cristata with a recorded length of 340 mm. The largest species from the Pacific
slope is P. eutmnia of which I have examined specimens 188 mm. long.

The following notes may help to identify the species. Species in which the
barbel extends to caudal are: gracilis, notomelas, serrata, mucosa, holiviana.

Species in which the pectoral spine is smooth or very feebly serrate are: roc-
coB, filamentosa, eigenmanni, meeki, vittata, buckleyi, holiviana, eutcenia, notomelas,
metce, itapicuruensis, hartti.

Species in which the pectoral thorns are moderate are: cristata, gracilis, avan-
handavce, puruensis, laticeps, australis, vittata, mucosa.

Species in which the pectoral thorns are very strong are: lateristriga, serrata,
hasemani, mucosa, chagresi, macturki, megalops, modesta, grisea, pectinifera, cyano-
stigma, transitoria.

Key to the Cis-Andean Species of Pimelodella.^

a. Head without conspicuous mucous pores.

6. Adipose fin usually less than three in the length, extending beyond the tip of the anal fin.

c. Pectoral spine with strong teeth both in front and behind; maxillary barbel extending

beyond base of anal or caudal 1. serrata Eigenmann.

cc. Pectoral spine with moderate thorns on the posterior margin, but slightly serrate or rough-
ened along the anterior margin.

d. Lower caudal lobe much broader than the upper, upper lobe not produced, the
lower frequently the longer; lateral band if present narrow, linear; 13-14 teeth

^ The species found east of the eastern Andes of Colombia and the Atlantic slope of Ecuador and
southward. See in this connection P. chagresi, p. 253.

232 MEMOIRS OF THE CARNEGIE MUSEUM

on the posterior face of the pectoral in specimens 83 mm. long, about 20-30
much more feeble ones in large specimens, the more distal thorns sometimes
with a broad base and a long free outer edge; maxillary barbels reaching origin,

or beyond tip of adipose 2. cristata (Miiller & Troschel).

dd. Caudal lobes subequal, always longer than the head; 18-29 thorns on the posterior
face of the pectoral spine; maxillary barbels reaching tip of adipose or beyond
tips of middle caudal rays; adipose fin 2.33 in the length; pectoral spine about
equal to the dorsal spine, sometimes shorter; dorsal spine 1-1.25 in the head.

3. steindachneri Eigenmann.

ddd. Upper caudal lobe normally greatly prolonged, but subequal in the young, not

much narrower than the lower lobe; maxillary barbel always extending at

least beyond base of anal; lateral band, if present, broad; pectoral spine a little

longer or a httle shorter than the dorsal spine with a variable number of strong

teeth on over half its posterior margin 4. gracilis (Cuvier & Valenciennes).

dddd. Upper caudal lobe not produced; maxillary barbels reaching to tip of ventrals
or to middle of anal; about 11 teeth on the pectoral spine; adipose fin 2.75-3.25

in the length 5. avanhandavce Eigenmann.

ddddd. Upper caudal lobe not greatly produced; maxillary barbel reaching beyond tip of
ventrals; numerous minute teeth on the pectoral spine; lateral band ill-defined.

6. roccm Eigenmann.

dddddd. Upper caudal lobe not prolonged beyond the lower; maxillary barbels reaching

origin of ventrals; pectoral spine smooth in front, with about nine thorns on

the inner margin, the largest a little more than half the greatest width of the

spine. No lateral band 7. puruensis Fowler.

ccc. Pectoral spine smooth or but shghtly roughened behind; upper caudal lobe prolonged;
maxillary barbels reaching anal; lateral band narrow, well defined; dorsal spine equals

pectoral spine 8. buckleyi (Boulenger).

66. Adipose fin 3 or more than 3 in the length. (See also number 3).
e. Dorsal spine not produced into a filament.

/. Pectoral spine with thorns along all but the very tip of the posterior margin; inter-
orbital three-fourths as wide as eye.

g. Maxillary barbel extending to the origin of the adipose fin, nearly to tip of ventral ;
pectoral spine equals length of head; eye 4 in the head, interorbital 5.

9. peclinifera Eigenmann & Eigenmann.

gg. Maxillary barbel e.xtending beyond base of anal; pectoral spine much shorter

than head; eye 3.33 in the head, interorbital 5.5; inner pectoral thorns large;

a narrow dark band 10. hasemani Eigenmann.

ggg. Maxillary barbel extending to end of the adipose; pectoral spine minutely serrate

within and without 11. cyanostigma Cope.

//. Pectoral spine with IS or fewer teeth, except in the very old.

h. Eye large, 2.5 in the head; interorbital much smaller than the eye, 5.5 in the
head; lower caudal lobe much the longer; pectoral spine but little shorter than
the head; maxillary barbel reaching origin or end of base of anal.

12. megalops Eigenmann.
hh. Eye smaller, more than 3 in the head.

i. D. 1.7; pectoral spine reaching ventrals, which reach to the anal.

13. brasiliensis Steindachner.

eigenmann: pimelodella and typhlobagrus 233

a. D. 1.6 or 1.7; interorbital 3-3.5 in the head; adipose 4-4.3 in the length;
barbels reaching tip of the depressed dorsal or middle of anal; pectoral
spine equals head without opercle, its posterior face smooth on the distal
four-tenths, 10 or 11 small thorns on the proximal six-tenths.
j. Maxillary barbel reaching tip of depressed dorsal.

14. laliceps Eigenmann.
jj. Maxillary barbel extending to end of ventral or middle of anal.

15. ausiralis Eigenmann.
Hi. D. 1.6; interorbital smaller than the eye, except in one specimen of itapicuru-
ensis.

k. Dorsal with a black wedge entering the fin from the middle in front,
or with a hyaline base followed by black and shading to the tip of
the fin.

I. Maxillary barbel reaching middle of caudal; pectoral spine a little
longer than snout and eye, with very few feeble teeth on its pos-
terior face; adipose 3 in the length; interorbital 3.5-4 in the head.

16. notomelas Eigenmann.

II. Maxillary barbel reaching anal; pectoral spine equals snout and

eye, the dorsal spine equals snout and half the eye; adipose 3-3.25

in the length; interorbital 4-5 in the head. . . 17. metm Eigenmann.

III. Dorsal with a black blotch on its middle anterior portion; maxillary

barbel extending to middle of ventrals or to near anal; adipose

3.6-4 in the length; eye 4.5-5 in the head, interorbital about

equal to the eye; upper caudal lobe the longer. See No. 26, meeki

Eigenmann.

kk. Dorsal without a black blotch, its tip usually dark.

m. Maxillary barbel extending to near base or beyond tip of caudal;

interorbital 3.5-4.7 in the head.

?i. Pectoral spine equals head less opercle, with 12-21 strong teeth

on the posterior face; adipose 3.66 in the length; interorbital

3.75 in the length of the head; length of adipose equals its

distance from the dorsal spine or very little more; anal

extends to or beyond tip of adipose. See No. 28, mucosa

Eigenmann.

nn. Pectoral spine a little longer than snout and eye, with 7 scarcely

perceptible teeth on its posterior face. Adipose 3 in the

length; interorbital 4.3-5 in the head; dorsal dark; p basal

light bar through it IS. boliviana Eigenmann.

mm. Maxillary barbel extending to or beyond the origin of the anal.

0. Maxillary barbel extending beyond tip of anal; pectoral
spine equals head without opercle; width of head equals
snout and eye. Eye 3.25 in the head; snout a little
longer than postorbital portion of head; interorbital 5.5
in the length of the head; distance between snout and
dorsal fulcrum 3.2 in the length; width of occipital process
3 in its length; pectoral spine with 14 very strong graduate

234

MEMOIRS OF THE CARNEGIE MUSEUM

teeth on the basal two-thirds of its posterior margin,
about 5 hooks in front. (Based on specimen from Pene-
do 72 mm. long to base of caudal.)

19. lateristriga (MuUer & Troschel).

00. Maxillary barbel extending to the adipose or middle of anal ;

pectoral spine terete, curved, 1-1.25 in the head; width of

head 1.33 in its length; eye 4 in head, 1.5 in the snout,

about equal to the interorbital; pectoral thorns much

smaller than in P. lateristriga 20. vittata (Kroyer).

000. Maxillary barbel extending to origin of anal; pectoral spine
equals snout and eye; width of head equals snout and
three-fourths of the eye; eye 3.25 in the head; snout a
little longer than postorbital portion of the head; inter-
orbital 5 in the length of the head; distance between
snout and dorsal fulcrum 3.4 in the length; dorsal spine
nearer snout than anal; width of occipital process 3.5 in
its length; pectoral spine with very feeble teeth on the
basal half of its posterior margin, 12 hooks in front.
(Based on specimen from Queimadas 72 mm. long to

base of caudal.) 21. itapicuruensis Eigenmann.

0000. Maxillary barbel reaching origin or end of base of anal;
pectoral spine equals snout and eye or a little longer;
eye 3.33-4, in middle of the head; interorbital 4.5; distance
between snout and dorsal fulcrum about 3 in the length;
pectoral spine with twelve long recurved thorns along
the middle of its posterior margin.

22. madurki Eigenmann.
00000. Maxillary barbel extending beyond origin of anal; pectoral
spine equals snout and eye, with about 7 very strong
hooks. See No. 32, chagresi (Steindachner).
000000. Ma.xillary barbel extending beyond origin of anal; pectoral
spine equals head less one-half of opercle; width of head
1.4 in its length; eye 3.5-4 in the head; interorbital 4.4-
5.33 in the head; center of eye in center of head; dorsal
spine equidistant from snout and anal; width of occipital
process 3 in its length; pectoral spine with feeble teeth
not more than one-fifth or one-sixth the width of the
spine. Caudal not split between the middle rays. (Based
on the types 130-180 mm. long, from Rio Parahyba.)

23. eigenmanni Boulenger.
mmm. Maxillary barbel not extending to the anal.

p. Maxillary barbel extending slightly beyond origin of ventrals;
pectoral spine equals length of head less opercle; width of
head a little greater than snout and eye; eye 4 in the head;
snout equals postorbital part of head; interorbital 4.5 in
the head; distance between snout and the process not

eigenmann: pimelodella and typhlobagrus 235

quite reaching the dorsal plate; pectoral spine with about 9
feeble spines on its posterior margin, about 9 hooks in
front, the upper ones sharp, the lower obscure. (Based on
specimen from Rio Doce 72 mm. long to base of caudal.)

24. harlti (Steindachner).
pp. Maxillary barbels extending to end of ventrals;' pectoral spine
equals length of head without opercles; width of head equals
snout and eye; eye 4 in the head; snout longer than post-
orbital portion of head; interorbital 5 in the length of the
head; distance between snout and dorsal fulcrum 3 in the
length; width of occipital process 3 in its length; pectoral
spine with 12 very strong graduate teeth on over half of its
posterior margin, about 5 hooks in front; adipose 3.4 in
the length. (Based on a specimen from Xiririca 91 mm.

long to base of caudal.) 2.5. transitoria Ribeiro.

ppp. Maxillary barbel extends to middle of ventrals (on one side
of the paratype to near anal); pectoral spine equals snout
and eye or shorter; width of head equals length of head
without opercles; eye 4.5-5 in the head; snout longer than
postorbital portion of head; interorbital 5 in the head;
distance between snout and dorsal fulcrum 2.8 in the length;
width of occipital process 3 in its length; pectoral spine
with 12 very feeble teeth on the basal half of the posterior
face; 9-11 recurved hooks near the tip in front; adipose 3.6-4
in the length (largely based on a specimen from Sapina 89

mm. long, and on the types) 26. meeki Eigenmann.

ee. Dorsal spine frequently prolonged into a filament extending consid.^rably beyond the rest of
the fin; interorbital 5 in the head; maxillary barbel reaching tip of ventrals or shorter;
pectoral spine with 9-11 teeth on the basal two-thirds of the posterior face; adipose 3.5-

4.5 in the length 27. griffini Eigenmann.

aa. Lower cheeks and mandible with conspicuous cavities; lower caudal lobe the longer; maxillary
barbel reaching to near origin or middle of caudal; pectoral spine with straight, feeble teeth on its
posterior surface; eye 3.5-4.5 in the head; adipose extending a very little beyond tip of anal.

28. mucosa Eigenmann & Ward.

1. Pimelodella serrata sp. nov. (Plate XXIX, fig. 1).
6967a C. M., type, 67 mm. 6966a C. M., paratype, 110 mm., San Joaquin, BoHvia,

Sept. 5, 6, 1909. Haseman.

Distinguished by its numerous pectoral thorns and the length of the barbels.

Head 3.9-4.33; depth about 7. D. 1.6; A. 13; eye 4-4.2 in head, equal to or
slightly greater than interorbital; snout a little greater than postorbital; maxillary
barbels extending to the end of the base of the anal or beyond the base of the
caudal; postmen tal barbels to the tip of the pectoral. Adipose about 2.5 in the
length.

3 To end of base of anal in No. 6979.

236 MEMOIRS OF THE CARNEGIE MUSEUM

Dorsal spine straight, slender, equal to length of head less opercle, 17 minute
teeth on the ui^per two-thirds of the posterior surface, the anterior margin smooth ;
pectoral spine heavy, little longer than dorsal spine; 18-21 strong retrorse teeth
along nearly the entire posterior margin, longest near the middle of the spine;
25-50 shorter, mostly antrorse teeth along all but a short space near the tip along
the anterior face. Caudal lobes nearly equal, about 3.5 in the length.

Occipital process of nearly the same width throughout, firmly united with
the dorsal plate. No color-markings.

2. Pimelodella cristata (Miillcr & Troschel) (Plate XXIX, fig. 2).

Pimelodus cristatus Muller & Troschel, in Schomb. Reisen in Brit. Guiana,
III, 1848, p. 628 (Takutu and Mahu Rivers); Horse Ichth., Ill, 1849, p. 4
(Essequibo); Gunther, Cat. Fishes Brit. Mus., V, 1864, 117 (Guiana; Esse-
quibo; River Capin; Para); Vaillant, Bull. Soc. Philom. (7), 1880, p. 152
(Calderon); Steindachner, Flussfische Siidam., IV, 1882, p. 4 (Rio Hual-
laga); ? Perugia, Ann. Mus. Geneva (2), X, 1891, p. 631 (Tucuman).

Pimelodella cristata Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), I, 1888,
p. 132 (San Gon^allo; Avary; Villa Bella; Jutahy; Tapajos; Rio Mucuri;''
Tabatinga; Hyavary; Coary); Occasional Papers Cal. Acad. Sci., I, 1890,
p. 150; Proc. U. S. Nat. Mus., XIV, 1891, p. 29; Eigenmann & Bean, Proc.
U. S. Nat. Mus., XXXI, 1907, p. 660 (Amazon); Eigenmann, Reports Prince-
ton Univ. Exped. Patagonia, III, 1910, p. 388; Mem. Carnegie Mus., V,
1912, p. 168 (Tumatumari; creek below Potaro Landing; Rockstone; Kona-
waruk; below Packeoo Falls; Twoca Pan); Fowler, Proc. Acad. Nat. Sci.
Phila., 1915, p. 214 (type of ophthalmicus).

Pimelodus agassizii Steindachner, Sb. Akad. Wiss. Wien, LXXIV, 1876, Flussf.
Siidostl. Bras., Ill, p. 56, footnote (Essequibo).

Pimelodus {Pseudorhamdia) wesselii Steindachner, Sb. Akad. Wiss. Wien, LXXIV,
1876, Siisswasserf. Siidostl. Bras., II, p. 56 (Essequibo).

Rhamdia wesselii Ribeiro, Faun. Bras., Peixes, IV, 1912, p. 268.

Pimelodus ophthalmicus Cope, Proc. Am. Phil. Soc, XVII, 1876, p. 675 (Upper
Amazon) .
Habitat. — Guiana, Amazon Basin to the Huallaga, and the Guapore. Rio

Mucuri.

My reference in former papers of Schomburgk's insignis to this species is

wrong.

* In three of the four specimens from Santa Cruz, Rio Mucuri (No. 7412 Mus. Comp. Zool.) the
caudal is considerably longer than in specimens from the Amazon, 4-4.5 in the length. The upper
caudal lobe is usually narrower and more pointed than the lower.

eigenmann: pimelodella and typhlobagrus 237

6933a C. M., 127 mm., San Joaquin, Bolivia, September 4, 1909. Haseman.

6934a C. M., 131 mm., Villa Bella, Oct. 5, 1909. Haseman.

6936a-/ C. M., 77-220 mm., Santarem, Dec. 8, 9, 1909. Haseman.

6937a-e= C. M., 85-258 mm., Maciel, Rio Guapore, July 23, 1909. Haseman.

???? Field M., 94 mm. to base of caudal, Serra da Lua, near Boa Vista, Amazon.

M. P. Anderson and R. N. Becker.
7586a M. C. Z., 154 mm., Rio Puty, Paranahj^ba basin. Thayer Expedition.
7475a« M. C. Z., 232 mm., Coary. Thayer Expedition.
7492a' M. C. Z., 265 mm., Tapajos. Thayer Expedition.
7541a-6 M. C. Z., 127 and 152 mm., Sao Gon^allo. Thayer Expedition.
7412a-d» M. C. Z., 232-305 mm., Santa Cruz, Rio Mucuri. Thayer Expedition.
7444a9 M. C. Z., 300 mm.. Villa Bella. Thayer Expedition.
7576a-6i'' M. C. Z., 66-103 mm., I^a. Thayer Expedition.
7569a" M. C. Z., 95 mm., Hyavary. Thayer Expedition.

3. Pimelodella steindachneri nom. nov.

Pimelodella wesselii Eigenmann & Eigenmann (non Steindachner), Proc. Cal.

Acad. Sci. (2), Vol. I, 1888, p. 132 (Cudajas; Para; Marajo; Rio Madeira;

Rio Puty; Santarem); Occasional Papers, Cal. Acad. Sci., I, 1890, p. 152;

Eigenmann, Rep. Princeton Univ. Exped. Patagonia, HI, 1910, p. 389.

Habitat. — Amazon Basin.

The species described by Steindachner as P. wesselii from the Essequibo is
identical with P. cristata (Miiller & Troschel) from the same place. It is doubt-
ful whether the specimens enumerated by E. and E. in the papers quoted above
belong to the same species, and they are here set apart under a new name. The
species, if distinct, is very closely related to P. gracilis and P. cristata. Its caudal
is more like that of gracilis. Its maxillary barbel is longer than that of cristata.

^ Maximum number of pectoral thorns 30.

^ Pectoral thorns 24.

' Pectoral thorns 22.

8 a. 232 mm., pectoral thorns 25, upper caudal lobe 39 mm., lower 44, maxillary barbel to end of
base of anal. 6. 262 mm., pectoral thorns 24, upper caudal lobe 52 mm., lower 51, maxillary barbel
to middle of base of anal. c. 300 mm., pectoral thorns 32, caudal broken, maxillary barbel not to origin
of the anal. d. 305 mm., pectoral thorns 29, upper caudal lobe 57 mm., lower 60 mm., maxillary barbel
beyond tip of anal.

' Pectoral thorns 25, upper caudal lobe 45 mm., lower lobe 55 mm., maxillary barbel to middle of
base of anal.

'" Pectoral thorns 14 and 11, maxillary barbel beyond base of anal in the larger specimen, to near
tip of the caudal a in the smaller.

" Pectoral thorns 17, maxillary barbel extending beyond caudal.

238 MEMOIRS OF THE CARNEGIE MUSEUM

The pectoral thorns are 17, 18, 22, and 23 in different specimens, the largest
number being found in 7542 M. C. Z. from Manacapuru. It is quite possible
that most of the specimens from north of the Paraguay basin heretofore referred
to P. gracilis really belong to this species. The specimens at present in the Museum
of Comparative Zoology are :

7566a-b. 95-126 mm. to base of caudal, Maues, Rio Madeira. Thayer Expedi-
tion.
7487a.i- 190 mm.. Para. Thayer Expedition.
7472a. 172 mm., Cudajas. Thayer Expedition.
75G7rt. 137 mm., Santarem. Thayer Expedition.
7588a. 154 mm., Rio Puty. Thayer Expedition.
7542a. 212 mm., Manacapuru. Thayer Expedition.

4. Pimelodella gracilis (Cuvier & Valenciennes).

Pimelodus gracilis Cuvier & Valenciennes, His. Nat. Poiss., XV, 1840, p. 181

(Buenos Ayres; Parana at Corrientes); Valenciennes, Voy. d'Orbigny, Vol.

V, 1847, Poissons, p. 6, plate II, fig. 5; Kner, Sb. Ak. Wiss. Wien, XXVI,

1857, p. 418 (Caigara, Mattogrosso; Rio Guapore; Cujaba); Gunthbr, Cat.

Fishes Brit. Mus., V, 1864. p. 121; Boulenger, Trans. Zool. Soc, 1896, p. 27

(Descalvados) .
Pimelodus {Pseudorhamdia) gracilis? Steindachner, Flussfische Siidam., I, 1879,

p. 9 (Orinoco near Ciudad Bolivar).
Pimelodella gracilis Eigenmann & Eigenmann, Proc. Cal. Acad. Sci., (2), I,

1888, p. 132 (Goyaz); ? Occasional Papers Cal. Acad. Sci., I, 1890, p. 153;

Proc. U. S. Nat. Mus., XIV, 1890, p. 29; Ann. Carnegie Mus., IV, 1907,

p. 114, pi. XXXII, fig. 2 (Corumba; Laguna Ypagarai); Eigenmann, Reports

Princeton Univ. Exped. Patagonia, III, 1910, p. 389.
Rhamdia gracilis Ribeiro, Fauna Bras., Peixes, IV, 1912, p. 269.
Pimelodus lateristriga Boulenger (non Miiller & Troschel), Trans. Zool. Soc. Lond.,

1896, p. 27 (Descalvados).
Pimelodus {Pimelodella) tmniophorus Regan, Ann. and Mag. Nat. Hist. (7), XII,

1903, p. 625 (Descalvados).

Habitat.^La. Plata Basin; Uruguay Basin; ? Amazon; ? Orinoco. More
abundant in the La Plata Basin than elsewhere.

In 6947 the maxillary barbel in specimens between 78 and 112 mm. long extends
beyond the base of the caudal; in the specimens 97 mm. long to the end of the

" In this specimen the dorsal spine is very high, equal to the length of the head. Pectoral thorns 22.

eigenmann: pimelodella and typhlobagrus

239

lower lobe. In the largest specimens it extends a little beyond the base of the
anal. The pectoral spine in the smallest has 9 spines on half of the length, 12 in
the next in size, 15 in the next (125 mm.), 24 on the largest on all but the distal
fourth of the spine. The lateral band extends from snout to caudal. The back
is similar in color to the band, the space between the back and the lateral band
being lighter in color. In the most highly developed the upper caudal lobe is
2.5 in the length; the distance between snout and dorsal fulcrum 3.4 in the length.

Specimens Examined.

Catalogue Number.

Length in
Mm.

Location.

Date.

Collector.

694ia CM

100

Villa Bella, Amazon

Oct. 5, 1909

Haseman

6942a-6 CM...

74-103

San Joaquin, Bolivia

Sept. 4, 1909

Haseman

6943a CM

77

Asuncion, Rio Paraguay

Mar. 29, 1909

Haseman

69440-/ CM...

58-135

Corumba, Paraguay Basin

Apr. 28, May 2,
1909

Haseman

6M5a-g CM...

35-110

Rio Jauru, Paraguay Basin

June 2, 4, 1909

Haseman

6946a-e CM...

75-110

San Luiz de Caceres, Paraguay Basin

May 24, 1909

Haseman

6947a-/i CM...

82-180

Villa Hays, Rio Paraguay

Apr. 13, 1909

Haseman

10133a I. U. M..

About 70

Corumba

Anisits

10792a I. U. M..

About 160

Parana, Entre Rios

von Ihering

IO20O0 I. U. M..

About 170

Laguna Ypagarai, Paraguay

Anisits

6961a-6 C M. . . .

67- 68

Arequa, Paraguay

April S, 1908

Haseman

6935o-i CM

55-140

Uruguayana

Feb. 5 and 7,1909

Haseman

6939a CM

100

Rio Paranahyba (into Parana) at
bridge to Goyaz

Aug. 14, 1908

Haseman

The size of the pectoral teeth varies greatly with the locality, the maxillary
barbel to a less extent. In 6939, 100 mm. long, it extends about to the end of
the base of the anal.

Only one of the twelve specimens recorded by E. & E. from Goyaz (8196
M. C. Z.) is still extant. I am in doubt whether we were correct in referring it to
P. gracilis. The caudal is broken off near the base so that it cannot be determined
whether it possesses one of the most characteristic features of the species, the
prolonged upper caudal lobe. The maxillary barbel extends only to near the
tip of the ventrals and is therefore much shorter than in typical specimens of
P. gracilis. The eye is smaller than in the average given for the twelve specimens
by E. & E., otherwise their description holds. The pectoral spines measured
from above average slightly longer than the snout and eye. The dorsal spine is
slightly less in length than the snout and eye; it is smooth in front and only very
slightly roughened behind. The pectoral spine has nineteen thorns, the longest
one being one third the width of the spine.

A large series of similarly preserved specimens are necessary to properly
define this and the preceding species, if the two are really different.

240 MEMOIRS OF THE CARNEGIE MUSEUM

5. Pimelodella avanhandavae sp. nov. (Plate XXIX, fig. 3).

6969a C. M., type, 85 mm., 6970a-a; C. M., paratypes, 57-96 mm., Rio Tiete at

Salto Avanhandava, above the fall, Sept. 14, 1908. Haseman. 7062a-<7,

56-77 mm., same place, Sept. 15, 1908.

Habitat. — Tiete Basin.

Very similar to P. gracilis and characterized by the long adipose and com-
paratively short barbels.

Head 4.5; depth 5.6; D. 1.6; A. 11-12; eye 4-4.25 in the head; interorbital 4;
adipose 2.75-3.25 in the length. Maxillary barbel extending nearly to tip of
ventrals or the middle of anal; postmental beyond base of pectoral; pectoral spine
equal to or a little longer than snout and eye; dorsal spine a little less than snout
and eye; about 11 thorns, haK as wide as the spine along the basal .7 of the pos-
terior margin of the pectoral spine, 4-6 recurved hooks and numerous tubercles
along the anterior margin; a narrow dark band along the middle of the sides,
faintly continued on the head; a hyaline band through the basal half of the dorsal.

6. Pimelodella roccae Eigenmann MS.

Pimelodella roccce Eigenmann, Mem. Comp. Zool., 1917, (UnpubHshed April, 1917)
(Lower Urubamba Valley).
Habitat. — Urubamba Valley.

7. Pimelodella puruensis Fowler.

Pimelodella puniense Fowler, Proc. Acad. Nat. Sci. Phila., 1915, p. 214, fig. 4

(Peruvian Amazon).

This species is known only from the type, which is 52 mm. long. It is quite
possible that it is the young of one of the other known species.

8. Pimelodella buckleyi (Boulenger).

Pimelodus lateristriga {non Miiller & Troschel) Cope, Proc. Acad. Nat. Sci. Phila.,

1871, p. 270 (Ambyiacu River).
Pimelodus buckleyi Boulenger, Proc. Zool. Soc, 1887, p. 275, pi. XX, fig. 1

(Canelos).
Pimelodella buckleyi Eigenmann, Reports Princeton Univ. Exp. Patagonia, III,

1910, 389.
Pimelodella copei Fowler, Proc. Acad. Nat. Sci. Phila., 1915, p. 216, fig. 5 (based

on Cope's specimens of lateristriga) .

Habitat. — Maranon Basin.

eigenmann: pimelodella and typhlobagrus

241

Boulenger was the first to recognize that Cope's specimens belonged to a
species distinct from P. lateristriga. There is little doubt but that P. copei is a
synonym of P. buckleiji. The P. buckleyi of E. & E. is a distinct species called
P. Eigenmanni by Boulenger, and P. Eigenmanniorum by Ribeiro.

Fig. 1. Pimelodella buekleyi (Boulenger). Reproduced from P. Z. S. London, 1887, PI. XX, fig. 1.

Known from two specimens from Canelos 150 mm. long and now in the British
Museum, and two specimens 160 mm. long collected by Hauxwell in the Am-
byiacu and now in the collections of the Philadelphia Academy of Sciences.

9. Pimelodella pectinifera Eigenmann & Eigenmann.

Pimelodella pedinifer Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), I,

1888, p. 132 (Campos); Occasional Papers Cal. Acad. Sci., I, 1890, p. 154;

Proc. U. S. Nat. Mus., XIV, 1891, p. 29; Eigenmann, Reports Princeton

Univ. Exp. Patagonia, III, 1910, p. 389.
Rhmndia pedinifer Ribeiro, Fauna Bras., Peixes, IV, 1912, p. 270.

This species is known only from the types from the Parahyba Basin.

10. Pimelodella hasemani sp. nov. (Plate XXX, fig. 7).

Pimelodella lateristriga Eigenmann & Eigenmann (part), Occasional Papers Cal.
Acad. Sci., I, 1890, p. 156 (the specimens in the M. C. Z. and I. U. M. enu-
merated below) . *
Habitat. — Madeira and Amazon Basin.

6968a C. M., type, 81 mm., San Antonio de Rio Madeira. Haseman.

4259a-c I. U. M., largest 57 mm., Iga. William James.

7577 M. C. Z., fifty, largest 67 mm., Iga. WiUiam James.

7579-7581 M. C. Z., twelve, largest 72 mm., Obidos. William James.

7502 M. C. Z., three, largest 51 mm., Obidos. Col. Bentos.

7572 M. C. Z., one, 64 mm., Jutahy. James Thayer and Talisman.

242 MEMOIRS OF THE CARNEGIE MUSEUM

Head 4.5; depth 6.6; D. 1.6; A. 12; eye 3.33 in the head, intcrorbital 5.5;
center of eye a little behind the middle of the head; adipose 3.5 in the length.
Maxillary barbel extending to middle of adipose on one side, to near its end on
the other. Distance between snout and dorsal fulcrum 3.2 in the length; dorsal
spine a little more than snout and eye. Pectoral spine very little longer than
dorsal spine, with 16-18 recurved teeth on about the basal three-fourths of the
posterior face of the spine. Anterior face with 4 hooks near the tip and about
40 short teeth along the rest of the spine. Humeral spine extending considerably
beyond the middle of the pectoral spine. Lower caudal lobe longer than the
head. Occipital process of the same width throughout, its width 3 in the length.
A narrow, dark, lateral band; dorsal without a light band.

11. Pimelodella cyanostigma (Cope).
Rhamdia cyanostigma Cope, Proc. Am. Phil. Soc, 1870, p. 569 (Pebas); Eigenmann

& Eigenmann, Occasional Papers Cal. Acad. Sci., I, 1891, p. 164.
Pimelodus cyanostigma Cope, Proc. Am. Phil. Soc, 1878, p. 675.
Pimelodella cyanostigma Fowler, Proc. Acad. Nat. Sci. Phila., 1915, p. 218. (Note

on the relationship.)

Habitat. — Maranon Basin.

Known only from the types. A specimen in the Field Museum, originally
about 65 mm. long, collected by Anderson and Becker at Boa Vista, Serra da Lua,
Feb., 1913, may belong to this species. The pectoral spines have been broken off,
and a definite identification is impossible.

12. Pimelodella megalops Eigenmann.
Pimelodella megalops Eigenmann, Repts. Princeton Univ. Exp. Patagonia, III,
1910, p. 389; Mem. Carnegie Mus., Vol. V, 1912, p. 169, plate XV, fig. 2
(Tumatumari, Potaro River; Crab Falls, Essequibo River).
Habitat. — Near mouth of the Potaro River, British Guiana.

13. Pimelodella brasiliensis (Steindachner).

Pimelodus (Pseudorhamdia) brasiliensis Steindachner, Sb. Ak. Wiss. Wien,

LXXIV, 1876, Siisswasserf. Sudostl. Bras., Ill, p. 50, pi. VII (Rio Parahyba).
Pimelodella brasiliensis Eigenmann & Eigenmann, Proc. Cal. Acad. Sci., (2), I,

1888, p. 133; Occasional Papers Cal. Acad. Sci., I, 1890, p. 162; Proc. U. S.

Nat. Mus., XIV, 1891, p. 29; Eigenmann, Reports Princeton Univ. Exped.

Patagonia, III, 1910, p. 389.
Rhamdia brasiliensis Ribeiro, Fauna Bras., Peixes, IV, 1912, p. 275. Known only

by the types from the Rio Parahyba.

eigenmann: pimelodella and typhlobagrus

243

14. Pimelodella laticeps sp. nov. (Plate XXX, fig. 2).

6957a C. M., type, 62 mm., 6958a-c C. M., paratypes, 75-90 mm., Sapucay,

Paraguay, April 2, 1909. Haseman.

Head 4; depth about 4.66; D. 1.6 in two, 1.7 in the other two; A. 14 or 15;
eye in middle of the head, 4.33-5; interorbital very broad, 3-3.5 in the head;
maxillary barbel reaching nearly to tip of the depressed dorsal or slightly beyond it;
postmental slightly beyond base of last pectoral ray.

Adipose fin 4-4.3 in the length, the space between the dorsal equal to eye
and snout; pectoral spine equal to head without the opercle; 7 recurved notches
in the distal four tenth of the anterior face, the proximal portion with minute

Fig. 2. Pimelodella laticeps Eigenmann. Dorsal view of head, X 2.

tubercles; posterior face smooth opposite the notches of the anterior face and
with 10-11 small thorns, the largest about half the width of the spine on the basal
six-tenths of the spine. Dorsal spine equals snout and eye, with very few faint
recurved notches near the tip in front, otherwise smooth; a large dusky blotch
behind the gill-opening, a dark band along the middle of the sides from below the
dorsal to the caudal; dorsal dusky; a hyaline band through the basal half of the
fin ; a dark stripe along the middle of the back.

15. Pimelodella laticeps australis var. nov. ~

Pimelodus lateristriga Boulenger, Proc. Zool. Soc, 1891, p. 232 (Rio Grande do

Sul).
Pimelodella lateristriga Eigenmann, Ann. N. Y. Acad. Sci., VII, 1894, p. 632.
6950 C. M., type, 75 mm., 6951a-s C. M., 50-87 mm., Uruguayana, Feb. 5, 1909.

Haseman.
6952a-p C. M., 52-76 mm., Cacequy, Rio Ibicuhy, into the Rio Uruguay, Feb. 13,

1909. Haseman.

244 MEMOIRS OF THE CARNEGIE MUSEUM

6953a-& C. M., 55 mm., Cachoeira, Rio Jacuhy, into Lago de Patos, Jan. 26, 1909.
Haseman.

695ia-d C. M., 53-60 mm., Porto Alegre, Rio Grande do Sul, Jan. 19, 1909. Hase-
man.

4876a-c I. U. IVI., 97-123 mm., Rio Grande do Sul. Von Ihering.
Habitat. — Basins of the Uruguay and of the Lago de Patos.
These specimens have longer barbels than the ones from Sapucay and are

much lighter in color, the dark along the back being interrupted at the dorsal plate

and the end of the dorsal, the occipital process is also much lighter than the area

bordering it. Only one of these specimens has the dorsal 1.7, the rest all 1.6.

A. -1-, -T, -1-; the maxillary barbel always longer, extending to end of ventral

or middle anal. Adipose fin 3.75-4.25.

16. Pimelodella notomelas sp. nov. (Plate XXX, fig. 3).

6955 C. M., type, 51 mm.; 6956a-/ C. M., paratypes, 37-53 mm., San Luiz de

Caceres, May 24-27, 1909. Haseman.
6971a-d C. M., 35-60 mm., Rio Jauru, June 2-4, 1909. Haseman.

Habitat. — Upper Paraguay Basin.

Distinguished by the black spot on the dorsal.

Head 4.2; depth 4.5-5; D. 1.6; A. 12 or 13; eye 3-3.5 in the head, a httle
greater than the interorbital ; maxillary barbel extending to tip of middle caudal
rays or a little shorter, postmen tal barbel to near tip of pectoral ; adipose 3 in the
length, its origin in advance of the tip of the dorsal; dorsal spine slender with a
few retrorse hooks near the tip in front, its height not quite equal to snout and
eye; pectoral spine very little greater than snout and eye, with a very few short,
feeble teeth behind; caudal very deeply forked, the lobes equal, about 3 in the
length; a black wedge through middle of dorsal, the black on the first three mem-
branes forming a conspicuous blotch, otherwise plain. No lateral band.

17. Pimelodella metse sp. nov. (Plate XXXI, fig. 1).

13768a I. U. M., 48 mm., Quebrada Cramalote, Villavicencio. Gonzales.

7441a C. M., type, 77 mm., Rio Negro, Villavicencio. Gonzales.

7442a-& C. M., 13769a-c I. U. M., 80-100 mm., Barrigona, Rio Meta. Gonzales.

Habitat. — Upper Meta of the Orinoco Basin.

Head 4.8; depth over 6; eye 4 in the head, intraorbital 4-5; D. 1.6; A. 11 or
12; depth of caudal peduncle equal to postorbital part of head. Maxillary barbel
almost or quite reaching origin of anal; outer mental barbel to middle of pectoral

eigenmann: pimelodella and typhlobagrus 245

or a little farther. Adipose fin 3-3.25 in the length; upper caudal lobe distinctly
the longer, 3 in the length; dorsal spine equals snout and half the eye; pectoral spine
equals snout and eye, with numerous recurved notches in front, very few feeble
teeth or smooth along posterior face. A narrow, well-defined, lateral streak from
eye to caudal; base of dorsal hyaline, then black, shading to the tip of the rays.

This species is closely allied to P. buckleyi, in which the dorsal and pectoral
spines are of equal length, and to P. notomdas with a much longer barbel.

18. Pimelodella boliviana sp. nov. (Plate XXXI, fig. 2).
G964a C. M., type, 90 mm. over all, Santa Cruz de la Sierra, Bolivia. Steinbach.
6965a C. M., 67 mm. to base of caudal, Prov. del Sara, Jan.-Oct., 1911. Stein-
bach.

Habitat. — Central Bohvia; Madeira Basin.

Head 4.5-4.6; depth 6.3-6.5; D. 1.6; A. 13; eye 3.75 in the head; interorbital
4.3-4.7; eye in middle of the head; adipose fin 3 in the length; maxillary barbel
extending to or beyond the base of the caudal; outer mental to or near to tip of
pectoral; pectoral spine a httle longer than snout and eye; dorsal spine less than
snout and eye. Pectoral spine with 10 retrorse hooks along the distal half of the
anterior margin, about 7 scarcely perceptible thorns on the posterior margin;
caudal 3.5 in the length; a narrow band from snout to end of middle of caudal
rays, heaviest just behind gill-openings. Dorsal dusky except along base.

19. Pimelodella lateristriga (MuUer & Troschel).

Pimelodus lateristrigus MtJLLER & Troschel, Horse IchthyoL, HI, 1849, p. 3

(Brazil).
Pimelodus lateristriga GtJNTHER, Cat. Fish. Brit. Mus., V, 1864, p. 118 (Brazil);

? Hensel, Wiegm. Arch., 1870, I, p. 69 (Porto Alegre); Stbindachner, Sb.

Ak. Wiss. Wien, LXXIV, 1876, Siisswasserfische Sudostl. Bras., HI, p. 45

(Rio Parahyba; Rio Doce; Rio Jequintinhonha; Cannavierias ; Muriahe and

Rio Janeiro); Vaillant, Bull. Soc. Philom. (7), IV, 1880, p. 52 (Calderou).
Pseudorhamdia lateristriga Lijtken, Velhas-Flodens Fiske, Dan. Vidensk. Selsk.

Skr., 1875, p. 171, fig. (Rio das Velhas).
Pimelodella lateristriga Eigenmann & Eigenmann (part), Proc. Cal. Acad. (2), I,

1888, p. 133 (Santa Clara; Rio Mucuri; Rio Doce; Cannavierias; San Matheos;

Mendez; Rio Trombetas); Occasional Papers Cal. Acad. Sci., I, 1890, p. 156;

Proc. U. S. Nat. Mus., XIV, 1891, p. 29; Eigenmann, Reports Princeton

Univ. Exped. Patagonia, III, 1910, p. 389.

246 MEMOIRS OF THE CARNEGIE MUSEUM

f Rhamdia lateristriga Ribeiro, Fauna Bras., Peixes, IV, 1912, p. 271 (Calderao,

Amazonas).

Habitat. — Amazon Basin; especially eastern Brazil from the Rio Sao Francisco
to Rio de Janeiro.

A number of references to the name lateristriga have been distributed among
the species to which they actually belonged.

Of this species there appear to be specimens from the Rio Sao Francisco:
6982a-e C. M., 37-85 mm., Penedo, near mouth of the Rio Sao Francisco, March

20, 1908. Haseman.
6983a-c C. M., 45-55 mm., Joazeiro, Nov. 28, 1907. Haseman.
6987a-c C. M., 71-82 mm., Lagoa Pereira, Dec. 23, 1907. Haseman.
and specimens from the Rio Parahyba:

6981a C. M., 60 mm., Sao Joao da Barra, June 23, 1908. Haseman.
6984a-d C. M., 70-95 mm., Campos near the mouth of the river, June 15, 1908.
6985a C. M., 72 mm., Lagoa Feia, a swamp on a sugar plantation south of the

mouth of the Parahyba.

From the Rio Itapemerim we have the following lot:
6986a-/ C, M. 60-71 mm., Muniz Freire, June 18, 1908. Haseman.

Of the above No. 6981 has the barbels extending to the middle of the base of
the anal. The pectoral spine is equal to the head less the opercle and has 11
teeth on the basal three-fourths of the spine; the teeth are graduated from the
second, which is longer than the spine is wide. Width of occipital process 4.5 in
the length; caudal 4 in the length; lateral band faint. With this 6984 and 6985
agree in almost every particular; the lateral band is sometimes conspicuous, con-
tinued on both head and caudal. Those nearest them in the character of pectoral
spines are 6986. The barbels extend to near the tips of the ventrals or a little
beyond them. The pectoral spine is equal to the snout and eye and has 6-8 teeth
on the basal two-thirds of the posterior face; the teeth are graduated from the
last, or the one next to the last, and the longest is greater than the width of the
spine; width of occipital process 3^ in its length; lateral band well defined, con-
tinued on head and caudal.

In 6982 and 6983 the maxillary barbel extends to the middle of the base of
the anal or beyond, on one side of one individual approaching the caudal. The
pectoral spine is equal to the snout and eye, plus or minus, and has 9 or 10 spines
in 6983 and 11-14 in 6982; teeth not as wide as the spine, on the basal half or two-
thirds of the posterior face. In the smallest they are graduated from the tip to

eigenmann: pimelodella and typhlobagrus 247

the base, in the larger from the middle both ways. Width of occipital process 4
in its length. Caudal 4 in the length. Lateral band variable, continued or not
continued to snout and on the tail.

In 6987 the teeth of the pectoral spines, 7 in the smallest, 12 in the two larger,
are distinctly smaller than in 6982, but they are otherwise very similar, and the teeth
are much larger than in the specimens from the Itapicuru = itapicuruensis.

20. Pimelodella vittata (Kroyer).

Pseudorhamdia vittata (Kroyer) LtJTKEN, Velhas-Flodens Fiske; Dan. Vidensk.

Selsk. Skr., 1874, p. 34 (Rio das Velhas); 1. c, 1875, p. 173, with fig.
Pimelodella vittata Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), I, 1888,

p. 133; Occasional Papers Cal. Acad. Sci., I, 1890, p. 159 (Rio Sao Francisco;

Minas Geraes, Sao Matheo; Rio Jequitinhonha) ; Proc. U. S. Nat. Mus., XIV,

1891, p. 29; Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,

1910, p. 389.
Rhamdia vittata Ribeiro, Fauna Bras., Peixes, IV, 1912, p. 272 (Rio das Velhas;

Jaguara) .

Habitat. — Eastern Brazil from the Rio Sao Francisco to Sao Matheo.

Mr. Haseman secured no specimens recognized as belonging to this species.

4258 I. U. M., 69 mm., Rio Sao Francisco, Thayer Expedition.

21. Pimelodella itapicuruensis sp. nov. (Plate XXXI, fig. 3).

6974a C. M., type, 80 mm.; 69746-??z, paratypes, 72-104 mm., Queimadas, Rio

Itapicuru, March 2, 1908. Haseman.
6938a C. M., 76 mm., Rio de Jacobina into Rio Itapicuru, Nov. 9, 1907. Haseman.
6940a-e C. M., 61-74 mm., Bom Fin, Rio Itapicuru, 6 miles north of Fazenda

Amaratu, Nov. 2, 1908. Haseman.
6975a-6 C. M., Agua Branca, into Itapicuru, Nov. 6, 1907. Haseman.

Habitat. — Rio Itapicuru, Eastern Brazil.

Head 4.3-4.5; depth 6.5-7.5; D. 1.6; A. 12; eye 3.26-3.5; width of head 1.5
in its length, depth of head at base of occipital process 1.66 in its length; adipose
2.8-3.33; usually a little over 3 in the length; maxillary barbel extending to caudal
in one, to end of adipose in one, to tip of pectorals in another, usually to some
point above the base of the anal. Dorsal spine nearer the snout than anal; distance
between tip of snout and dorsal fulcrum 3.33-3.16 in the length. Dorsal spine
slender, not equal to snout and eye; about 8 hooks in front, sHghtly roughened
behind; pectoral spine about equal to snout and eye, with 8-12 short teeth along

248 MEMOIRS OF THE CARNEGIE MUSEUM

the basal half or six-tenths of its posterior margin; sharp hooks along the distal
portion of its anterior face; pectoral not nearly reaching ventrals, which are inserted
below the end of the dorsal; ventrals not nearly reaching anal the tip of which does
not extend to the end of the adipose; humeral spine reaching middle of pectoral spine;
upper caudal lobe much longer; a distinct band from snout to caudal; base of dorsal
hyaline.

6974m differs in having the dorsal 1.7; the adipose very short, 6 in the length,
but extending beyond the tip of the anal; caudal lobes nearly equal in length, the
upper 3.75 in the length. Eye larger than in specimens of P. vittata in the Museum
of Comparative Zoology.

22. Pimelodella macturki Eigenmann.

Pimelodella macturki Eigenmann, Repts. Princeton Univ. Exp. Patagonia, III,
1910, p. 389; Mem. Carnegie Mus., Vol. V, 1912, p. 170, plate XVI, fig. 1
(Creek in Mora Passage, trenches at Morowhanna, and Georgetown).
Habitat. — British Guiana near the coast.

23. Pimelodella eigenmann! Boulenger.

Pimelodella buckleyi Eigenmann & Eigenmann {non Boulenger), Occas. Papers

Cal. Acad. Sci., I, 1890, p. 158 (Rio Parahyba; Macacos).
Pimelodella eigenmanni Boulenger, Proc. Zool. Soc. London, 1891, p. 232 (based

on P. buckleyi of Eigenmann & Eigenmann) ; Eigenmann, Reports Princeton

Univ. Exped. Patagonia, III, 1910, p. 389.
Rhamdia eigenmanniorum Ribeiro, Fauna Bras., Peixes, IV, 1912, p. 213. Based

on Eigenmann & Eigenmann I. c.

Habitat. — Eastern Brazil.

Head 4.5-5; depth 5.5-6.5; D. 1.6; A. 12-14 in the head, eye 3.5-4 in the
head, in middle of the head; interorbital 1.33 in the eye, 5.3-6 in the head; maxillary
barbel extending to the origin of the anal; postmentals to middle of pectoral;
dorsal spine equidistant from snout and anal, its height 1.33 in the head; adipose
3.4-4 in the length, its distance from the dorsal equals the base of the latter. Cau-
dal lobes of equal width, the upper longer, 4 in the length. Anal rounded; pectoral
spine a little less than the length of the head, with teeth which are not more than
one-fifth or one-sixth the width of the spine. A dark lateral band.

24. Pimelodella hartti Steindachner.

Pimelodus hartti Steindachner, Sb. Akad. Wiss. Wien, LXXIV, 1876, Siisswassf.
Siidostl. Bras., Ill, p. 53 (Rio Parahyba).

eigenmann: pimelodella and typhlobagrus 249

Pimelodella hartti Eigenmann & Eigenmann, Proc. Cal. Acad. Sci., (2), I, 1888,
p. 133; Occasional Papers Cal. Acad. Sci., I, 1890, p. 158; Proc. U. S. Nat.
Mus., XIV, 1891, p. 29; Eigenmann, Reports Princeton Univ. Exped. Pata-
gonia, III, 1910, p. 389.

Rhamdia hartti Ribeiro, Fauna Bras., Peixes, IV, 1912, p. 270.
Habitat. — Eastern Brazil in Rio Doce and Rio Parahyba.

6973a-x C. M., 42-100 mm., Rio Doce, May 24, 1908. Haseman.

This species is allied to lateristriga. The barbels are shorter, the pectoral

spine with much more feeble teeth behind, and the head is more depressed and

broader. The snout is rounded, the mouth inferior.

25. Pimelodella transitoria (Ribeiro) = lateristriga?

Rhamdia transitoria Ribeiro, Fauna Bras., Peixes, IV, 1912, p. 274 (R. Alambary,

Iporanga) .
6976a-d C. M., 110-about 130 mm., Xiririca, Rio Ribeira do Iguape, Dec. 5-8,

1908. Haseman.
6977a-/ C. M., 35-106 mm., Morretes, on Rio Marunhy, into Rio Nhundiaquara

into ocean at Paranagua, Jan. 2, 1909. Haseman.
6979a C. M., 115 mm., Piracicaba, Rio Tiete, Sept. 9, 1908. Haseman.

Habitat. — Southeastern Brazil.

These specimens representing the P. transitoria of Ribeiro can scarcely be
separated from P. lateristriga. The largest are, however, larger than the largest
of the typical P. lateristriga. The left maxillary extends to the end of the ventrals
in the four specimens of 6976, in two others to the middle of the ventrals, and a
httle beyond the tip of the ventrals; the pectoral spine is about as long as the head
less the opercle and provided with about 13 long recurved teeth along the basal
three-fifths of the spine. The tip of the dorsal is dark.

In 6979 the maxillary barbel extends to the end of the base of the anal. In
the largest of 6977 it extends just about to the middle of the ventrals; eye 4-4.5;
interorbital 4.66-5.

26. Pimelodella meeki Eigenmann. (Plate XXXII, figs. 1-2).

Pimelodella eigenmanni Meek (non Boulenger), Proc. Biol. Soc. Washington,
XVIII, 1905, p. 241 (Sao Paulo, Brazil).

Pimelodella meeki Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,
1910, p. 389 (Sao Paulo).
Habitat. — Rio Tiete, Sao Paulo, Brazil.

6980a-c C. M., 87-117 mm., Sapina, Rio Tiete, Sao Paulo, July 23, 1908. Hase-
man.

250 MEMOIRS OF THE CARNEGIE MUSEUM

6978a-?/ C. M., 37-51 mm., Mogy das Cruzes, Rio Tiete, July 19, 1908. Haseman.

Allied to P. lateristriga and P. transitoria.

Head 4.25; depth 5; D. 1.6; A. 11; eye 4.5 in the head; intcrorbital equals
eye; maxillary barbel reaching a little beyond base of ventrals; adipose 3.8 in the
length; caudal 5; distance from snout to dorsal fulcrum 3 in the length; fins all
small; pectoral spine equals snout and eye; 11 hooks on over half of the distal
anterior face of the spine; 12 feeble teeth not over one-third as wide as the spine
on a little over half of the posterior face of the spine; pectoral fin equals length of
head less opercle; ventral fin equals snout and half the eye; base of anal equals
length of snout; dorsal spine equals snout and half eye; a well-developed band
from snout to caudal.

27. Pimelodella griffini sp. nov. (Plate XXXII, fig. 3).

6962 C. M., type, 87 mm.; 6963a-e C. M., paratypes, 55-88 mm., mountain rills

near Sapucay, Paraguay, April 2, 1909. Haseman.

Distinguished by the filamentous dorsal ray.

Head 4-4.5; depth 5-5.5; depth of caudal peduncle one-half of greatest depth;
D. 1.6; A. 13 counting everything; eye 1.6 in snout, 4 in head; 8 in interorbital;
adipose 3.5-4.5 in the length; maxillary barbel reaching tip of ventrals or shorter;
outer mentals but little if any beyond base of pectoral. Dorsal spine about equal
to the pectoral spine, prolonged into a filament, spine and filament sometimes equal
one-third of the length, the filamentous portion being sometimes longer than the
spine. Pectoral spine with 9-11 spinules on the basal two-thirds of the posterior
edge; the anterior edge with retrorse hooks near the tip and minute serrations
the rest of the way; dorsal spine nearly smooth; upper caudal lobe the longer,
the lower as long as the head; a dark band from anterior nares to end of middle
caudal rays.

One specimen has a light band through the dorsal, the dorsal 1.7; the eye
less than the interorbital.

28. Pimelodella mucosa Eigenmann & Ward. (Plate XXXIII, fig. 1).

Pimelodella lateristriga {non Miiller & Troschel) Eigenmann, part." Ann. Car-
negie Mus., IV, 1907, p. 114 (Villa Rica).

Pimelodella mucosa Eigenmann & Ward, Ann. Carnegie Mus., IV, 1907, p. 114,
plate XXXII, fig. 1; Eigenmann, Reports Princeton Univ. Exp. Patagonia,
III, 1910, p. 389.

'' One of the small specimens referred to Pimelodella lateristriga seems to belong to this species;
the others certainly do not.

eigenmann: pimelodella and typhlobagrus 251

Habitat. — Paraguay Basin.
10125 I. U. M., tjq^e, 126 mm., Bahia Negra, Paraguay, Aiiisits.
10200 I. U. M., part, 50 mm., Villa Rica, Aiiisits.

6959a C. M., 135 mm., Puerto Suarcz, Paraguay, May 7, 1909. Haseman.
6960a-/ C. M., 63-86 mm., Laguna Ypacary, Arequa, Paraguay, April 7, 1909.

Haseman.
6972a C. M., 112 mm., near Berlin, Bolivia, Rio Mamor^, Sept. 14, 1909. Hase-
man.

Very similar to P. gracilis but the adipose distinctly shorter, the tip of the
anal extending to or beyond the tip of the adipose.

Head 4; depth 5; D. 1.6; A. 10-12; eye 3.5-4.5 in the head; interorbital 3.75-4;
maxillary barbel to within an orbital diameter of the caudal or middle of caudal;
adipose 3-3.66 in the length. Pectoral spine equal to length of head less opercle,
with 5-9 recurved hooks near tip in front, these followed by about 30-40 tubercles,
most of which are antrorse; 11-21 recurved teeth along most of the posterior
margin; dorsal spine equal to snout and half of the eye. Caudal deeply forked,
the lower lobe broader and longer, 3.5 in the length; the tip of the anal extends
to or beyond the vertical from the tip of the adipose. A well-developed lateral
band, heaviest behind the gill-opening. The maxillary barbel extends to near the
base of the caudal and may extend beyond its tip. Eye 4 + in the head; inter-
orbital equal to the eye. The adipose is 3-3.66 in the length.

Key to the Trans-Andean Species of Pimelodella."
a. Upper caudal lobe about equal to head in length; pectoral spine with thorns for more than two-
thirds its length.

6. Basal four-fifths of pectoral spine with very strong hooks on its posterior face, the hooks of the
distal half longer than the width of the spine at its base. Dorsal spine with a few teeth on
its posterior face. Adipose fin 4.25-5 in the length, pectoral spine very little longer than
snout and eye, maxillary barbel reaching to somewhere about the ventrals. Head 4.25;

depth 5; D. 1.6; A. 11-12 29. grisea Regan.

66. Basal three-fourths or four-fifths of the pectoral spine with hooks along its posterior face, the
hooks less than the width of the spine, posterior face of dorsal spine rough; adipose fin 34-3i
in the length, pectoral spine about equal to snout and eye. Maxillary barbels reaching some-
where about middle of adipose, shorter in the young. Head 4.8, depth 5.5; D. 1.6; A. 12;

a sharply defined lateral band ,30. modesta (Giinther).

666. Basal two-thirds of the pectoral spine with eight (6-8 fide Steind.) hooks; posterior face of
dorsal spine smooth; adipose fin 2.66 in the length; pectoral spine very little less than snout
and eye; maxillary barbel reaching the last third of the ventrals; head 4.33; depth 5; D. 1.6;
A. 12; no sharply defined lateral band 31. yuncensis Steindachner.

" This account is extracted from the MS. of my forthcoming monograph on the fishes of Colombia.
It was prepared in collaboration with Mr. Homer G. Fisher.

252 MEMOIRS OF THE CARNEGIE MUSEUM

aa. Upper caudal lobe much longer than the lower, much longer than head.

c. Distal one-third or one-fourth of pectoral spine without hooks, the hooks strong.

d. Adipose fin 3-3.5 in the length; pectoral spine equal to length of snout and ej'e; maxillary
barbels reaching beyond the base of anal. Head 4.5-4.8; depth 5-6; D. 1.6; A. 11-12.

32. chagresi (Steindachner).
dd. Adipose fin 2.66-2.75 in the length, pectoral spine a little longer than snout and eye, maxil-
lary barbel reaching beyond origin of anal. Head 4.75; depth 5.5; D. 1.6; A. 11.

33. dongala (Giinther).
cc. Distal half of inner face of pectoral spine without hooks; adipose fin 3.2.5-3.5 in the length;
pectoral spine equal to head less half or whole of operele; maxillary barbels reaching beyond
tip of ventrals in the small or anal in the large; head 4-4.66; depth 5.5; D. I, 6; A. 11-12.

34. eutcenia Regan.

29. Pimelodella grisea Regan. (Plate XXXIII, fig. 2).

Pimelodella griscus Regan, Ann. and Mag. Nat. Hist., Ser. 7, Vol. XII, 1903,
p. 625 (Northwestern Ecuador) ; Eigenmann, Reports Princeton Univ. Exped.
Patagonia, III, 1910, p. 389; Regan, Ann. and Mag. Nat. Hist., Ser. 8, Vol.
XII, 1913, p. 467 (Rio San Juan).

6751a-i C. M., 13588, 1. U. M., 41, largest 165 mm., Rio Dagua at Cordova. Eigen-
mann.

6752a C. M., 13589, I. U. M., 2, 152 and 155 mm., San Juan half-way between
Puerto Negria and Istmina. Eigenmann.

13598 I. U. M., 1, 116 mm., Istmina. Eigenmann.

Habitat. — San Juan River, Colombia, south to Northern Ecuador.

30. Pimelodella modesta (Giinther).

Pimelodus modestus Gunther, Proc. Zool. Soc. Lond., 1860, p. 239, pi. X, fig. C

(Esmeraldas) ; Cat. Fishes Brit. Mus., V, 1864, p. 117 (Esmeraldas) ; Fish.

Cent. Amer., 1866, p. 393 (Rio Chagres, Esmeraldas).
Pimelodella ynodestus Eigenmann & Eigenmann, Proc. Cal. Acad. Sci., (2),

Vol. I, 1888, p. 133 (name); Occasional Papers Cal. Acad. Sci., I, 1890, p. 155

(rivers of Western Ecuador); Eigenmann, Reports Princeton Univ. Exped.

Patagonia, III, 1910, p. 389.

Habitat. — Patia, River of southern Colombia to the Chone River in Ecuador.
6743a-e C. M., 13574 I. U. M., 42, largest 122 mm., Chone, Ecuador. Henn.
6744a-Z C. M., 13575 I. U. M., 19, largest 72 mm., Portovicio. Henn.
6745a-i C. M., 13583 I. U. M., 54, largest 140 mm., San Lorenzo, Rio Telembi,

Jan. 14, 1913. Henn and Wilson.
6746a-i C. M., 13584 I. U. M., 57, .argest 135 mm., Patia, between Magui and

Telembi. Henn.

eigenmann: pimelodella and typhlobagrus 253

6747a-rf C. M., 13585 I. U. M., 7, largest 117 mm., Telembi above Barbacoas,
Jan. 13, 1913. Henn and Wilson.

6748a-c C. M., 3, largest 125 mm., Telembi, above Barbacoas, Jan. 15, 1913.
Henn and Wilson.

6750a-c C. M., 13587 I. U. M., Creeks above Barbacoas, Jan. 17, 1913. Henn
and Wilson.

67Ma-i C. M., 13597 I. U. M., Patia, at mouth of Rio Guaitara. Henn.

6749a-Z C. M., 13586 I. U. M.,21, largest 137 mm., Telembi, 8 miles above Bar-
bacoas, Jan. 16, 1913. Henn and Wilson.

31. Pimelodella yuncencis Steindachner.

Pimelodella yuncencis Steindachner, Denkschr. K. Akad. Wiss. Wien, LXXH,

1912, p. 47 (Pacasmayo, Peru) ; Eigenmann, Reports Princeton Univ. Exped.

Patagonia, HI, 1910, p. 389.
Rhamdia gilli Starks, Proc. U. S. Nat. Mus., XXX, 1906, p. 769, pi. LXV, fig. 1

(Rio Eten, Peru); Eigenmann, 1. c, p. 389.

Field Mus., 74 mm., April 2, 1912. Osgood.

Habitat. — Northwestern Peru.

The single specimen at my disposal has the following characters Head 4.33 ;
depth 5; D. 1.6; A. 12; eye 4.66 in the head, interorbital 3.66; pectoral spine nearly
equal to snout and eye, with 8 thorns on the basal two-thirds of its posterior mar-
gin; maxillary barbel reaching to last third of the ventrals; nearly uniform dusky,
without lateral band; fontanel a very narrow slit; occipital process not quite
reaching the dorsal plate.

This species greatly resembles P. avanhandavce, which, however, has a broader
fontanel and the occipital process meeting the dorsal plate.

32. Pimelodella chagresi (Steindachner). (Plate XXXIII, fig. 3).

Pimelodus (Pseudorhamdia) chagresi Steindachner, Sb. Ak. Wien, LXXIV,
Ichthyol. Beitr., IV, 1876, p. 34 (Rio Chagres and its tributary, near Obispo).

Pimelodella chagresi Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), I, 1888,
p. 134 (Obispo River); Occasional Papers Cal. Acad. Sci., I, 1890, p. 160 (Rio
Chagres and its tributaries); Eigenmann, Reports Princeton Univ. Exped.
Patagonia, III, 1910, p. 389.
Habitat. — Both slopes of Panama, Atrato, and Magdalena Basins. ? Rio

Meta Basin.

6753a C. M., 1, 145 mm., Tambo. Wilson.

254 MEMOIRS OF THE CARNEGIE MUSEUM

13590 I. U. M., 1, 7 mm., Certegui. Eigenmann.

6754:a-d C. M., 13591 I. U. M., 7, largest 96 mm., Bernal Creek. Eigenmann.

6755a-c C. M., 13592 I. U. M., 5, largest 103 mm., Quibdo. Wilson.

6756a C. M., 1, 21 mm., Truando. Wilson.

13599 I. U. M., 1, 63 mm., Puerto Wilches. Eigenmann.

6765a-i C. M., 13600 I. U. M., 18, largest 69 mm., Peilas Blanca. Eigenmann.

6766a C. M., 51 mm., ViUavicencio? (Loc. 7). Gonzales.

6767a-/ C. M., 13543 I. U. M., largest 55 mm., Apulo. Gonzales.

6768a-& C. M., 13544 I. U. M., largest 467 mm., Girardot. Eigenmann.

6769a C. M., 1, 63 mm., Honda. Eigenmann.

If 6766 from ViUavicencio belongs to this species, this species is found both
east and west of the Andes. But it is a small specimen and the positive identifica-
tion across the Andes may be left till larger specimens are available for examination.

Meek & Hildebrand found this species to be abundant on both slopes of the
central range in Panama.

33. Pimelodella elongata (Glinther).

Pimelodus elongatus Gunther, Proc. Zool. Soc. Lond., 1860, p. 238, pi. X, fig. B
(fresh waters of Esmeraldas); Cat. Fish. Brit. Mus., V, 1864, p. 118 (Esmer-
aldas).

Pimelodella elongatus Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), I,
1888, p. 133 (name); Occasional Papers Cal. Acad. Sci., I, 1890, p. 155 (rivers
of Western Ecuador); Eigenmann, Princeton Univ. Exped. Patagonia, III,
1910, p. 389.
Habitat. — Rivers of Western Ecuador.

6741a-6 C. M., 13572 I. U. M., 5, longest 92 mm., Vinces, Ecuador. Henn.

6742a C. M., 13572 I. U. M., 3, longest 81 mm., Colimes, Ecuador. Henn.

34. Pimelodella eutaenia Regan. (Plate XXXIV, fig. 7).

Pimelodella eutcenia Regan, Ann. and Mag. Nat. Hist. (8), Vol. XII, 1913, p. 466

(Rio San Juan).

Habitat. — San Juan, Dagua, and Patia Basins of Western Colombia.
6757a-i C. M., 18593 I. U. M., 66, largest 157 mm., Istmina. Wilson.
6758a C. M., 160 mm., San Juan at mouth of Rio Cucurrupi. Henn.
6759a-b C. M., 2, larger 126 mm., Condoto. Wilson.
6760a-i C. M., 13594 I. U. M., 32, largest 170 mm., Cisnero. Eigenmann.
6761a-j C. M., 13595 I. U. M., 42, largest 188 mm., San Lorenzo, Rio Telembi.

Henn and Wilson.

eigenmann: pimelodella and typhlobagrus 255

6762a C. M., 1, 150 mm., Caldas. Eigenmann.

6763a-d C. M., 13596 I. U. M., 8, largest 150 mm., Barbacoas. Henn and Wilson.

Typhlobagrus Ribeiro.

Typhlobagrus Ribeiro, Kosmos, No. 1, January, 1907.
Type, Typhlobagrus kronei Ribeiro.
In all characters but the eyes like Pimelodella. The eyes lost.

35. Typhlobagrus kronei Ribeiro. " Ceguinho."

Typhlobagrus kronei Ribeiro, Kosmos, No. 1, January, 1907; Eigenmann, Re-
ports Princeton Univ. Exped. Patagonia, III, 1910, p. 387; Ribeiro, Fauna
Bras., Peixes, IV, 1912, p. 250, PI. XLII, figs. 2, 2.4 and 2B (Cavernas das
Areiras, Iporanga, Sao Paulo); Haseman, Ann. Carnegie Mus., VII, 1912,
p. 323 (Cavernas das Areiras).
7U3a-x C. M., 23, 74-202 mm., Caverna das Areiras, Nov. 28, 1908. Haseman.
Head 4-4.2; depth about 5.5; D. usually 1.6, in one case, 1.7; A. 12-15 (usu-
ally 14). Maxillary barbel reaching to near middle of ventrals in youngest and
oldest, sometimes a httle longer or a httle shorter, postmentals a little beyond
base of pectoral. Fontanel narrow, extending to base of occipital crest, which
reaches to dorsal plate, width of occipital crest nearly 4 in its length. Dorsal
spine half or a little more than half the length of the head, a little less than the
length of the pectoral spine; 8-14 moderate-sized thorns on the pectoral spine.
Adipose fin about 3.75 in the length, its tip extending beyond the tip of the anal;
caudal lobes of about equal width, the upper lobe the longer by .25 or .2 of the
length of the head.

A distinct trace of a dark lateral band on some of the specimens; margin of
dorsal sometimes dusky.

The region of the eye varies. No trace of an eye or optic nerve can be noticed
in ordinary dissections of the alcoholic material at my disposal. The skin over the
vanished eye is invaginated in various degrees, a distinct free orbital margin being
quite evident in some specimens, less distinctly so in others. They appear as
though the eye had been removed without greatly altering the region of the skin
over them.

Ribeiro says: "Dentre os exemplares que tivemos em maos, um possue um
olho desenvolvido.

"Este facto prova perfeitamente a reversao, por heranga, a um caracter de
sens antepassados; o orgam mede 4 millimetros no maior diametro e o peixe 150
millimetros.

256 MEMOIRS OF THE CARNEGIE MUSEUM

"Essa predominancia hereditaria, tao frisante aqui, e certanaente um docu-
mento de valor para a theoria genealogica, quando se ve que em 35 outros ex-
emplares de todos os tamanhos, apenas uma estricta fenda indica a posigao que
outr'ora occupou esse importantissimo orgao. No exemplar que nos dissecamos.
nao conseguimos encontrar, siquer, vestigios do nervo optico e o seu logar estava,
ao contrario, occupado pelo grosso ramo nervoso do barbilhao maxillar."

It seems evident from the fact that the skin and the bones near the eye have
undergone little or no change, while the eye has vanished but occasionally de-
velops, that the process of its degeneration was very rapid.

EXPLANATION OF PLATES.

Plate XXIX.

Fig. 1. Pimelodella serrata Kigenmarm. Type, C. M. No. 6967, 67 mm. San Joaquin,

Bolivia.
Fig. 2. Pimelodella cristata (Miiller & Troschel). I. U. M. No. 12064, 176 mm. Below

Potato Landing.
Fig. 3. Pimelodella avanhandavce Eigenmann. Type, C. M. No. 6969, 85 mm. Salto

Avanhandava.

Plate XXX.
Fig. 1. Pimelodella hasemani Eigenmann. Type, C. M. No. 6968, 81 mm. San

Antonio de Madeira.
Fig. 2. Pimelodella laticeps Eigenmann. Type, C. M. No. 6957, 62 mm. Sapucay,

Paraguay.
Fig. 3. Pimelodella notomelas Eigenmann. Type, C. M. No. 6955, 51 mm. Caceres,

Plate XXXI.
Fig. 1. Pimelodella metce Eigenmann. Type, C. M. No. 7141, 77 mm. Rio Negro.

Villavicencio.
Fig. 2. Pimelodella boliviana Eigenmann. Type, C. M. No. 6964, 90 mm. Santa

Cruz de la Sierra.
Fig. 3. Pimelodella itapicuruensis Eigenmann. Type, C. M. No. 6974. Rio Itapicuru.

Plate XXXII.
Fig. 1. Pimelodella meeki Eigenmann. Type, Field M. No. 3400. Sao Paulo.
Fig. 2. Pimelodella meeki Eigenmann. C. M. No. 6980, 117 mm. Sao Paulo.
Fig. 3. Pimelodella griffini Eigenmann. Type, C. M. No. 6962, 87 mm. Sapucay,
Paraguay.

eigenmann: pimelodella and typhlobagrus

257

Fig. 1.
Fig. 2.
Fig. 3.

Plate XXXIII.
Pimelodella mucosa Eigenmann. C. M. No. 6959, 135 mm. Puerto Suarez.
Pimelodella grisea Regan. I. U. M. No. 148 mm. Cordova, Rio Dagua.
Pimelodella chagresi (Steindachner). U. S. N. M., about 145 mm.

Plate XXXIV.

Fig. 1. Pimelodella eutcenia Regan. I. U. M. No. 13593, 147 mm. Istmina.
Fig. 2. Typhlobagrus kronei Ribeiro. C. M. No. 7443, 146 mm. Caverna das x\reiras.

Plate XXXV.
(Figures 1-44, left pectoral spines of various species of Pimelodella as seen from

above. Figures 3, 5, 6, 7, 10, 11, 16, 18, 19, 21, 22, 23, 24, 25, 27, 28, 29, 31, 32, 33,
34, 35, 36, 37, and 38 are camera sketches drawn to the same scale.)

Length Length

of of

Spine Specimen

in mm. in mm.

Fig. 1. Pimelodella serrata 'Ei\ge\\msii\i\. Type, C. M. No. 6966 110

Fig. 2. P. hasemani Eigenmann. Type, C. M. No. 6968 .... 81

Fig. 3. P. hasemani Eigenmann. I. U. M. No. 4259 7 57

Fig. 4. P. mucosa Eigenmann. C. M. No. 6959 135

Fig. 5. P. wmcosa Eigenmann. I. U. M. No. 10200 6 50

Fig. 6. P. mucosa Eigenmann. Type, I. U. M. No. 10125. . .20 126

Fig. 7. P. rocca Eigenmann. Type^jn^l;_Cl_Z- — --- 21 144

' " (to base of

caudal)

Fig. 8. P. avanhandavce Eigenmann. Type, C. M. No. 6969 . . 85

Fig. 9. P. laticeps Eigenmann. C. M. No. 6958 12 89

Fig. 10. P. transitoria (Ribeiro). C. M. No. 6976 16 118

Fig. 11. P. lateristriga (Muller and Troschel). C. M. No. 6982 . 14 93

Fig. 12. P. itapicuruensis Eigenmann. Type, C. M. No. 6974. 80

Fig. 13. P. meeki Eigenmann. Type, Field M. No. 3400 14 12

Fig. 14. P. meeki Eigenmann. C. M. No. 6980 12 105

Fig. 15. P. eigenmanni Boulenger. M. C. Z. No^lO 25 168

Fig. 16. P. vittata (Kroyer). I. U. M. No. 4258 9 60

Fig. 17. P. vittata (Kroyer). M. C. Z. No. 7576. 10 90

Fig. 18. Typhlobagrus kronei Ribeiro. C. M. No. 7443 22 160

Fig. 19. T. kronei Ribeiro. C. M. No. 7443 11 105

Fig. 20. Pimelodella griffini Eigenmann. Type, C. M. No.

6962 87

Fig. 21. P. metoi Eigenmann. C. M. No. 7442 10 95

258 MEMOIRS OF THE CARNEGIE MUSEUM

Fig. 22. P. ynetce Eigenmann. Type, C. M. No. 7441 8 77

Fig. 23. P. yuncensis Steindachner. In Field Mus 8 74

Fig. 24. P. chagresi (Steindachner). In U. S. N. M 8 70

Fig. 25. P. chagresi (Steindachner). In U. S. N. M 14 143

Fig. 26. P. peciinifer Eigenmann. M. C. Z. No. 7508 190

Fig. 27. P. madurki Eigenmann. I. U. M. No. 12068 9 71

Fig. 28. P. megalops Eigenmann. I. U. M. No. 12066 12 95

Fig. 29. P. hartti (Steindachner). C. M. No. 6973 14 101

Fig. 30. P. holiviana Eigenmann. Type, C. M. No. 6964 90

Fig. 31. P. griffini Eigenmann. C. M. No. 6963 10 90

Fig. 32. P. grisea Regan. I. U. M. No. 13588 19 163

Fig. 33. P. grisea Regan. I. U. M. No. 13588 13 115

Fig. 34. P. modesta (Gunther). I. U. M. No. 13583 16 130

Fig. 35. P. eutcenia Regan. I. U. M. No. 13593 8 67

Fig. 36. P. eutmnia Regan. I. U. M. No. 13593 18 144

Fig. 37. P. gracilis (Cuvier and Valenciennes). I. U. M. No.

10205 19 160

Fig. 38. P. cristata (Miiller and Troschel). I. U. M. No.

12064 25 204

Fig. 39. P. cristata (Muller and Troschel). C. M. No. 6936.. . 92

Fig. 40. P. cnstata (Muller and Troschel). C. M. No. 69.36.. . 77

Fig. 41. P. noiomelas Eigenmann. C. M. No. 6956 6 53

Fig. 42. P. steindachneri Eigenmann. M. C. Z. No. 5767 .... 17 136

Fig. 43. P. steindachneri Eigenmann. M. C. Z. No. 7472. . . .20 175

Fig. 44. P. steindachneri Eigenmann. M. C. Z. No. 7542 .... 27 213

Memoirs Carnegie Museum, Vol. Vll

Plate XXIX.

1. Pimelodella serrata Eigenmaxn. Type, C. M. No. 6907, 67 mm. S.\n Joaquin, Bolivia.

2. Pimelodella cristata (Muller & Troschel). I. U. M. No. 12064, 176 mm. Creek below Potaro Landing, Guiana.

3. Pimelodella avanhandavce Eigenmann. Type, C. M. No. 6969, 85 mm. Saltq Avanhandava.

Memoirs Carnegie Museum, Vol. VII.

Plate XXX.

1. Pimelodella hasemani Eigenmann. Type, C. M. No. 6968, 81 mm. San Antonio de Rio Madeira.

2. Pimelodella laiiceps Eigenmann. Type, C. M. No. 6957, 62 mm. Sapucay, Paraguay.

3. Pimelodella notomelas Eigenmann. Type, C. M. No. 695.5, 51 mm. San Luiz de Caceres.

Memoirs Carnegie Museum, Vol. VII.

Plate XXXI.

1. Pimelodelln rnetce Eigenmann. Type, C. M. No. 7141, 77 mm. Rio Negro, Villavicrus s.

2. Pimelodella bolmana Eigenmann. Type, C. M. No. 6964, 90 mm. Santa Cruz de la Sierra, Bolivia.

3. Pimelodella itapicuruensis Eigenmann. Type, C. M. No. 0974, 77 mm. Queimadas, R. Itapicuru.

Memoirs Carnegie Museum, Vol, VII.

Plate XXXII.

1. Pimeludella meeki Eigenmann. Type, Field Mus. No. 3400, 120 mm. .Sao Paulo.

2. Pimelodella meeki Eigenmann. C. M. No. 6980, 117 mm. Sapina, Sao Paulo.

3. Pimelodella griffini Eigenmann. Type, C. M. No. 6962, 87 mm. Sapucay, Paraguay.

Memoirs Carnegie Museum, Vol. VII.

Plate XXXIII.

1. PimelodeUa mucosa Eigenmann. C. M. No. 6959, 135 mm. Puerto Suarez, Paraguay.

2. PimelodeUa grisea Regan. I. U. M. No. 13588, 148 mm. Cordova, Rio Dagua.

3. PimelodeUa chagresi (Steindachner). U. S. N. M. About 145 mm. Monte Liria, C. Z.

>

><

X
X

w

o
>

00

w

3
w

PS

<

00

<;

Q

l-H

<1

1?:

^

«

l-H

w

»J

>

<J

(J

S

s

S

r--

?^

'ft*

T— t

CD

•*

CO

1 — 1

TO

^

"O

CO

CO

^

.— 1

'tl

t^

n

^

6

§

^

o

Iz;

tf

<J

N

o

pq

w

l-H

«

«

e

S^

o

-ii

^

X)

^5

Sj

^

n

•■^.

^i^

o

o

^

-s'

^

a.

-:^

tin Eh

Memoirs Carnegie Museum, Vol. VII.

Plate XXXV.

Pectoral Spines of Species of Pimelodella, viewed from above and greatly magnified.

(See explanation at end of text.)

21. Minute Book of the Virginia Court Held at

Fort Dunmore (Pittsburgh) for the District

of West Augusta, 1775-1776. Crdmeine 90

22. Minute Book of the Virginia Court Held for

Yohogania County, first at Augusta Town
(now Washington, Pa.), and afterward on the
Andrew Heath Farm near West Elizabeth,
1776-1780. Ceumrine 2.25

23. Minute or Order Book of the Virginia Court

Held for Ohio County, Virginia, at Black's
Cabin (Now West Liberty, W. Va.), &c.
Crumrinb 1.50

24. The Records of Deeds for the District of West

Augusta, Virginia, for the Court Held at Fort
Dunmore, &c. Crumeine 1.75

25. Astropecten (?) montanus, &c. Douglass 10

26. Astrodon (Pleuroccelus) in the Atlantosaurus

Beds of Wyoming. Hatchee. {Out of Print.)

27. Osteology of the Limlcolse. Shupeldt 1.00

28. New Vertebrates from the Montana Tertiary.

Douglass 1.25

29. New Genus and Species of Tortoise from the

Jurassic of Colorado. O. P. -Hat 10

30. Osteology of Oxydactylus. Peterson 1.00

31. Birds of Erie and Presque Isle. Todd 75

32. J. B. Hatcher. By W. J. Holland 15

33. Tropidoleptus Fauna at Canandaigua Lake, etc.

Raymond. {Out of print.)

34. Two Turtles from the Judith Eiver Beds of

Montana. O. P. Hay. {Out of print.)

35. Preliminary List of Hemiptera of Western

Pennsylvania. Wirtnee. {Scarce.) 50

36. Trilobites of the Chazy Limestone. Eaymond. 1.00

37. Crawfishes of Western Pennsylvania. Ort-

MANN 1.00

38. The Geology of Southwestern Montana. Doug-

lass 40

39. New Crocodile from the Jurassic of Wyoming.

Holland 10

40. Procambarus, a New Subgenus of the Genus

Cambarus. Ortmann 15

41. Presentation of Reproduction of Diplodocus Car-

negei to the British Museum. Holland 15

42. Birds Collected near Mombasa by William Do-

herty. Holland 20

43. Hyoid Bone in Mastodon Americanus. Hol-

land 10

44. Additions to Flora of Allegheny County, Pa.

Jennings 15

45. New Species of Blneiffia. Jennings 05

46. Occurrence of Triglochin palustris in Pa. .Jen-

nings 05

47. New Species of Ibidium. Jennings 10

48. Agate Spring Fossil Quarry. Peterson 10

49. Two New Birds from British E. Africa. Ober-

HOLSER 05

50. The Chazy Formation and Its Fauna. Ray-

mond 1.50

51. A New American Cybele. Narraway and Ray-

mond 15

52. Plastron of the Protosteginae. Wibland 15

53. New Species of Testudo from the Miocene, etc.

O. P. Hay 25

54. Miocene Beds of W. Nebraska and E. Wyoming

and Their Vertebrate Faunae. Peterson 1.00

55. New Species of Lonicera from Pa. Jennings. .05

56. Meryocochcerus, etc. Douglass.

57. New Merycoidodonts. Douglass. (Nos. 56

and 57 sold together.) 1.00

58. Further Collections of Fishes from Paraguay.

Eigenmann, McAtee, and Ward 1.25

59. Undetermined Element in the Osteology of the

Mosasauridae. Holland 20

60. Gasteropoda of the Chazy. Raymond 1.35

61. Occurrence of Wynea Americana in Pa. Jen-

nings 05

62. Account of Pleistocene Fauna Discovered at

Frankstown, Pa. Holland 10

63. Vertebrate Fossils from the Vicinity of Pitts-

burgh, Pa. Case 15

64. niasnidae from the Black River Limestone near

Ottawa, Canada. Raymond and Naeeaway . . .20

65. Rhinoceroses from the Oligocene and Miocene

of North Dakota and Montana. Douglass. . .26

66. FossU Horses from North Dakota. Douglass. .30

67. Oligocene Lizards. Douglass 20

68. The Type of Stenomylus gracilis. Peterson.. .25

69. New Species of Birds from Costa Rica, etc.

Carriker 10

70. Costa Rican Formicariidas. Careiker 05

71. Vertebrate Fossils from the Fort Union Beds.

Douglass 45

72. List of the Lepidoptera of Western Pa., etc.

Engel 1,25

73. Fauna of Upper Devonian in Montana. Fossils

of Red Shales. Raymond 60

74. New Species of Procamelus from Upper Mio-

cene of Montana. Douglass 30

75. Sections of the Conemaugh Series between

Pittsburgh and Latrobe, Pa. Raymond 35

76. List of the Unionidae of Western Pa. Ortmann .50

77. Geological Reconnaissance in North Dakota,

Montana, and Idaho. Douglass 1.00

78. Botanical Survey of Presque Isle. Jennings . . 2.75

79. Sesqui-Centennial (Pittsburgh) Relics. Stewart .45

80. Dromomeryx, New Genus of American Rumi-

nants. Douglass 30

81. Fossils from Glacial Drift and Devonian and

Mlssissippian near Meadville, Pa. Millaed. . .10

82. New Species of Helodiw. Eastman 05

83. Charles Chauncey Mellor. Holland 15

84. Reports of Expedition to British Guiana, etc.

Eigenmann 50

85. Reports of Expedition to British Guiana, etc.

DuRBlN 25

86. Odonata of the Neotropical Region, Exclusive

of Mexico and Central America. Calveet. . . 2.25

87. Deinosuchus hatcheri, a New Genus and Species

of Crocodile. Holland 20

S8. Reports on Expedition to British Guiana, etc.

Blosser 15

89. Description of Some New Titanotheres from

the Uinta. Douglass 25

90. Annotated List of the Birds of Costa Rica In-

cluding Cocos Island. Carriker 3.00

91. Geology of the Coast of the State of Alagdas,

Brazil. Branner 40

92. Fossil Fishes from the Bituminous Shales at Ri-

acho Doce, Alagoas, Brazil. Jordan 55

93. Ordovician Trilobites, No. n. Asaphidae from

the Beekmantown. R.4.ymond 36

94. Ordovician Trilobites, No. m. Raymond &

Narbaway 35

95.
96.

97.
98.

99.

100.
101.

102.

103.

104.
105.

106.

107.

108.

109.

110.

112.
113.

114.

115.

116.

117.
118.
119.

120.
121.

122.

123.

124.
125.

126.

127.

128.

129.
130.

131.
132.

Ordovlcian Trilobites, No. IV. Raymond 40

Fishes from Irkutsk, Siberia. Joedan and
Thompson 30

South American Tetrigidae. Bruner 1.00

Fauna of the Allegheny and Conemaugh Series
in Western Pa. Eatmond 30

Ichthyological Survey About the San Juan
Islands, Washington. Stasks 75

New Species of Pygidium. Eigenmann 10

The Brachiopoda and Ostracoda of the Chazy.
Raymond 50

A New Camel from the Miocene of Western
Nebraska. Peterson 15

Skeleton of Stenomylus hitchcocki, etc. Peter-
son 1^

Skeleton of Diceratherium cooki. Peterson . . .15

Carnegie Museum Expedition to Central South
America, 1907-1910. Holland 15

Report Upon the Expedition of the Carnegie
Museum to Central South America. Base-
man & Eigenmann 25

New Species of Fishes, etc., from Central South
America. Haseman 50

Annotated Catalog of Cichlid Fishes from Cen-
tral South America. Haseman 1.25

New Species of Fishes from the Rio Iguassu.
Haseman 65

Ornithology of the Bahama Islands. Todd &

WORTHINGTON '^5

South American Acridoidea. I. Beuner 1.00

Species of Hasemania, Hyphessobrycon, and

Hemigrammus. Ellis 25

New Characins in the Carnegie Museum. Eigen-
mann 30

Jurassic Saurian Remains Ingested within Fish.

Eastman 20

Autograph Letter of U. S. Grant to Edwin M.

Stanton. Holland 15

Albert J. Barr. By W. J. Holland 10

Seventeen New Neotropical Birds. Todd 25

Dr. David Alter, the First Discoverer of Spec-
trum Analysis. Holland 10

Two Mummy Labels. Allen 10

The Families and Genera of Najades. Ort-

MANN 1-00

Group of Stenomylins in the Carnegie Museum.

Peterson 25

Tertiary Fish-remains from Spanish Guinea.

Eastman 25

Plated Nematognaths. Ellis 65

New Species of Cambarus from the Isle of

Pines. Ortmann 15

Sedum Camegiei, a New Species of the Family

Crassulaceae, etc. Hamet 10

Two New Species of Fishes from Peru. Eigen-
mann 15

South American Locusts (Acridoidea). II.

Bruner 75

Revision of the Genus Chaemepelia. Todd 85

New Genus and Species of Abyssinian Rodents.

Childs Frick 20

New Titanothere from the Uinta. Peterson.. .20
Small Titanothere from the Lower Uinta. Peter-
son 10

133. Osteology of Lasiopyga and Callithrix, etc.

Shufeldt 25

134. New Rhynchocephalian from Solenhofen. Grier .15

135. Scales of South American Characinid Fishes.

COCKERELL 25

136. Skeleton of Platigonus Leptorhinus. Peter-

son 10

137. Lichens Collected in the Thunder Bay District,

Ontario. E. Heber Howe, Jr 10

138. Preliminary List of Fossil Plants in the Roof

of the Pittsburgh Coal. Grier 10

139. Undescribed Remains of the Uinta Titanothere

Dolichorhinus. Peterson 15

140. Triassic Fishes Belonging to the Catopteridae

and Semionotidae. Eastman 15

141. Osteology of Promerycochoerus. Petersoi: ... .40

142. Correction of a Generic Name. Peterson 05

143. Skull of Bison Crassicomis. Holland 10

144. Serrasalminae and Mylinse. Eigenmann 50

145. Heads and TaUs: Notes on Sauropod Dinosaurs.

Holland 15

146. Dipterus Remains from Upper Devonian of

Colorado. Eastman 10

147. Notes on Tropical American Tettigonoidea (Lo-

custodea). Bruner 1.65

148. New Species of Tortoise from Jurassic of

Utah. Gilmorb 15

149. Fauna of Upper Devonian in Montana.

Haynes 45

150. New Sphagebranchus from Bahamas. Eigen-

mann 07

151. Marine Fishes from Colombia and Ecuador.

Wilson 15

152. Apareiodon, a New Genus of Characid Fishes.

Eigenmann 18

153. New and Rare Fishes from S. American Rivers.

Eigenmann 25

154. Tliree New Species of Characid Fishes. Eigen-

mann & Henn 12

155. The Species of Salminus. Eigenmann 7

156. South American Pceciliid Fishes. Henn 40

157. A New Species of Apatosaurus. Holland 7

158. Birds of the Isle of Pines. Todd 70

159. Reptiles and Amphibians of the Isle of Pines.

Barbour 20

160. Land and Fresh Water Shells of the Isle of

Pines. Henderson 20

161. Pelecypoda of the Chazy. Raymond 45

162. S. Ameridan Crickets, Gryllotalpoidea and Ache-

toidea. Bruner 1.30

163. Preliminary Catalog of N. American Sphaeri-

idae. Sterki 75

164. Directions for Collecting Sphaeriidae and

Aquatic Gastropods. Sterki 10

165. Lepidoptera of the Isle cf Pines. Holland 50

166. Odonata of the Isle of Pines. Kahl 15

167. A Trip to Islands in Lake Erie. Goodrich 15

168. Land-Shells of Islands at Western End of Lake

Erie, etc. Clapp 20

169. Orthoptera of the Isle of Pines. Holland &

Kahl 15

/6;7/^

Publications of the Carnegie Museum, Serial No. lOl-

MEMOIES

OF THE

OAENEGIE MUSEUM.

VOL. VIL No. 5.

W. J. HOLLAND, Editor.

THE PYGIDIIDAE, A FAMILY OF SOUTH
AMERICAN CATFISHES

By CARL H. EIGENMANN.

PITTSBIJKGH.
Published by the Authoeity of the Boakd of Tkustees op the
CARNEGIE INSTITUTE,
September, 1918.

PRICE LIST OF PUBLICATIONS OF THE CARNEGIE MUSEUM

ANNUAL REPORTS OF THE DIRECTOR

1898, 30c. (scarce) ; 1899, 25c.; 1900, 30c. (scarce) ; 1901-18, 25c. each.

REPORTS OF PROCEEDINGS OF FOUNDER'S DAY

1898-1918, 35c. each.

ANNALS OF THE CARNEGIE MUSEUM

The Annals are supplied to those Tvho subscribe in advance in parts (paper-bound), as published, @ $3.50 per volume;
Vols. I-XI, 1901-1917, bound in green cloth @ $4.00; bound in J Morocco @ $4.50.

MEMOIRS OF THE CARNEGIE MUSEUM

The Memoirs are supplied to those who subscribe in advan ce in parts (paper-bound), as published, at $10 per volume;
bound Li gre«r. buckram at $10.75, and in half-morocco at $12.00.

VOL. L (1901-3)

No. 1. Diplodocus, Its Osteology, Taxonomy, etc. No. 3. Osteology of the Steganopodes. Shufeldt $2.75

Hatcher $3.00 ' ' 4. Classification of Chalcid Flies. Ashmead . . 6.50

' ' 2. Oligocene Canidse. Hatcher 1.50

VOL. n. (1904-6)

No. 1. Osteology of Haplocanthosarus, etc. Hatcher $2.00 " 6. Osteology of Diplodocus. Holland 1.50

' ' 2. Osteology of Baptanodon. Gilmobe 1.50 ' ' 7. Osteology of Protostega. Wieland 75

" 3. Fossil Avian Bemains from Aimissan. East- " 8. New Suilline Remains from the Miocene of

MAN 1.00 Nebraska. Peterson 75

" 4. New Eodents and Discussion of Origin of •■' 9. Notes on the Osteology of Baptanodon, etc.

Dsemonelix. Peterson 1.75 Gilmore 1.00

No. 5. Tertiary of Montana. Douglass $1.00 ' ' 10. The Crawfishes of Pennsylvania. Ortmann 4.00

VOL. in.

No. 1. Archaeological Investigations in Costa Eica. No. 2. Osteology of Moropus. Holland and Peter-

Hartman $6.00 son $6.00

VOL. rv.

No. 1. Early Chinese Writing. Chaltant $3.00 No. 5. New Carnivores from the Miocene of West-

' ' 2. Formosan Fishes. Jordan 25 em Nebraska. Peterson $1.50

" 3. Entelodontidae. Peterson 2.50 " 6. Monograph of the Najades of Pennsylvania.

" - 4. Fishes of Formosa. Jordan and Richard- Pts. I and II. Ortmann 2.50

SON 1.25 ' ' 7. Catalog Eocene Fishes from Monte Bolea in

Carnegie Museum. Eastman 3.00

VOL. V.

The Fresh Water Fishes of British Guiana. Eigenmann. $10 unbound; $10.75 cloth; $12.00 J morocco.

VOL. VI.

No. 1. Fishes from Korea. Jordan and Metz $1.50 No. 4. Fishes obtained in Japan. 1911. Jordan

" 2. Lantern-fishes of Japan. Gilbert 1.00 and Thompson $3.50

" 3. Gymnotid Eels of Tropical America. Max Nos. 5-7. Fossil Fishes in the Carnegie Musemn.

Mapes Ellis 1.50 Parts II-IV. Eastman 5.00

vn.

No. 1. The Cheirodontinee. Eigenmann $3.50 No. 3. Ophidia from S. America. Griffin 2.00

' ' 2. Turtles of Uinta Formation. Gilmobe 2.00 No. 4. Pimelodella and Typhlobagrus. Eigenmann. 2.00

No. 5. The Pygidiidae. Eigenmann 2.50

REPRINTS FROM THE ANNALS OF THE CARNEGIE MUSEUM

Nos. 1-12. See preceding lists. (Out of print.) 16. Osteology of Pslttaci. Shufeldt 35

12. Skeleton of Titanotherium dispar. Hatcher. . $ .35 17. Annotated Catalogue of Genus Partula. H. H.

13. Fore Limbs and Manus of Brontosaurus. „ •■" I""lj"l U ,«

18. Two New Bahaman Lepldoptera. Holland ... .10
Hatches, 60 ^^ Elosaurus Parvus. Peterson & Gilmobe 16

14. The Trachodontidae. Hatcher 25 20. Boundary Controversy Between Pennsylvania

15. New Peimsylvania Thorns. Ashe 25 and Virginia, 1748-1785. Crumrine 30

■t-yj 13IO

MEMOIES

OF THE

OAENEGIE MUSEUM.

Vol. VII. No. 5.

THE PYGIDIID^, A FAMILY OF SOUTH AMERICAN CATFISHES.^

By Carl H. Eigenmann.
(Plates XXXVI-LVI.)

Introduction.

The Pygidudce are a family of South American catfishes distinguished exter-
nally by the absence of an adipose fin and by the posterior position of the dorsal.
Most of them are even more readily distinguished by the presence of spines or
thorns on the opercle and interopercle, by twin barbels at the angle of the mouth,
and by the absence of all mental barbels. Other characters of the catfishes may be
present or absent, and by addition, subtraction, or modification of characters,
various subfamilies have been formed. A description of the characteristic struc-
tures is given on pages 276-279. The basal habit of all the members of the family
is that of burrowing. The opercular and interopercular spines are an adaptation
to their habit of insinuation, which is at the root of the commensahsm, parasitism,
and worse, to which some highly specialized members of the family are addicted.

Nematogenys from central Chile, the only representative of the Nematogenyince,
is probably more nearly like the ancestors of the Pygidiidce than the other living
representatives of the family. It recalls the Siluridce by having a pmigent pectoral
spine, serrated on its posterior margin, by having but one barbel at the angle of the
mouth (the remaining subfamihes having two), by having a pair of mental barbels,

' Contribution from the Zoological Laboratory of Indiana University, No. 16-1,

259

260 MEMOIRS OF THE CARNEGIE MUSEUM.

and by the absence of opercular and interopercular spines, which are present in the
other subfamiUes. Its dorsal is farther forward than in tlic other members of the
family, again in this respect approaching the Siluridw. It resembles the rest of the
Pygidiidce in lacking an adij)ose fin and in the case of some specimens by having
a nasal barbel. I am not able to speak with certainty of its air-bladder. -

The principal subfamily is that of the Pygidiime. In addition to the main
characteristics of the family, the members of the subfamily have a barbel on the
anterior nostril, the gill-membranes are free from the isthmus, the teeth are in
bands. The genera of the PygidiincE differ but Httle from each other. Eremo-
philus, found on the plains of Bogota, has lost its ventrals ; Hatcheria, living in the
Andes of Argentina and Chile, has an elongate dorsal fin, Scleronema from the
Uruguay has modified maxillaries and maxillary barbels. The main genus, Pygi-
dium, with sixty-three species, is found everywhere in tlie mountains and sparingly
in the lowlands. It attains the highest altitudes and flourishes in Lake Titicaca,
where it is a food-fish of importance. The only food-fish at Bogota is the closely
related Eremophilus, "El Capitan." Seventy-one of the known species of the
family belong to the subfamily Pygidiince.

■ The genus Pariolius may be related to Nematogcnys, but it is more likely to be related to Phreato-
bius, Heptapterus, Myoglanis, Lepturhamdia (for Leptoglanis which is pre-occupied), etc. The only speci-
men recorded has been lost.

I. PARIOLIUS* Cope.

ParioUus, Cope, Proc. Acad. Nat. Sci. Phihi., 1871, p. 289.

Type. — ParioUus armiliatus Cope.

Similar to Pygidium; no nasal Ijarbel; a single barbel at the angle of the mouth; two pairs of mental
barbels; no armature on the opercles; gill-openings wide; teeth brush-like; origin of the dorsal behind that
of the ventrals; anus under dorsal; anal short.

Little can be said about the relationship of this genus until its skull ami air-bladder are examined.
It appears to be closely related to .some members of the Pimelodiiiw. There arc no specimens available for
examination. It is known only from the type of the species, and that has been lost.

Habitat. — Basin of Peruvian Anuizons.
ParioUus armillalu.'i Cope, I. c. (Ambyiacu); Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II,
1889, p. 50; Occasional Papers Cal. Acad. Sci., I, 1890, p. 324; Proc. U. S. Nat. Mus., XIV, 1891,
p. 36; Eigenmann, Reports Princeton Univ. Exped. Patagonia, III, 1910, p. 398.

"Head flat rounded, eyes small, superior, covered by the skin. Head 4. ,5 times in length to basis
of caudal fin. Depth at D. I. one-half length to basis pectoral fin; width of head two-thirds the same dis-
tance. Interorbital width 3. 60 times in length of head. Maxillary and external mental barbels extending
beyond basis of pectoral; inner mental barbel one-half the same. Radii D. 7; P. 8; V. 6; A. 11; caudal
acuminate. Skin entirely smooth."

* I am not sure of the origin of this generic name. Is it from the proper name Parioli, or from
TTopaioXifu = to trick, hence a trickster or simulator, or from irapa = with, and awXos = speckled'?

EIGENMANN : THE PYGIDIID^, A FAMILY OF SOUTH AMERICAN CATFISHES. 261

Nearest the Pygidiince are the Pareiodontma', which lack a nasal barbel. The
teeth are very peculiar and in a single series (See Fig. 21) and the gill-membranes
are attached. But one species is known. The Nematogenyina', Pygidiinw, and
PareiodonUncE are free-living, and have a terminal or subterminal mouth and pointed
or incisor teeth.

The StegophiUna', V andellivna' , and Tridentince differ widely from the members
of the above-mentioned subfamilies in structure and habit. They are all small or
minute; the mouth is inferior; the head flat below; the lower barbel at the angle of
the mouth is minute. The jaws are weak, the teeth absent or slender. The gill-
opening, in all but Acanthopoma, is greatly restricted, which, put in terms of habit,
means that the mouth is suctorial. Some of them are parasites, or commensals.

The Tridentinai differ in having the anal fin much longer than the others.
Nothing is known of their habits and they are so small (the largest known specimen
is but 27 mm. long) that it is a wonder that any of them have arrived in the bottles
of the naturalist.

In the Vandelliince the teeth are reduced to a minimum, and the rami of the
lower jaw do not meet in the middle. The differences between the genera are
minute, but well marked. The habits of these fishes, as well as those of the next
subfamily, are discussed below.

The Stegophilinw have a very large number of minute teeth in definite series
in both jaws. The rami of the lower jaw meet in the middle. The genus Acantho-
poma stands out in that its gill-membranes, while united, are free from the isthmus.
The genus, Henonemus, is well marked by the small number of opercular spines.
Ochmacanthus has numerous accessory caudal rays above and below, which make
the tail look like that of a tadpole. The remaining genera, Stegophilus, Homo-
dicetus, and Pseudostegophilus , might well be vmitcd. They differ from each other
largely in the position of the ventrals, the shape of the caudal, and in the number of
accessory rays.

Habits.

The habits, as well as the distribution, of various members of the Pygidiidce
have been derived from the general tendency of the catfishes to get under banks,
under logs, out of the way, and out of sight. This general tendency has been
modified into the specialized, insinuating habit of the Pygidiida;, for which the
opercular spines and the eel-like body are adaptations.

On the plains of Bogota the Indians secured the largest specimens of Eremo-
philus by thrusting their arms to the bottom of holes in the banks of streams.
At Honda I found one species buried in the sand in the bottom of the stream. It

262 MEMOIRS OF THE CARNEGIE MUSEUM.

would dart from its hiding place as I raked my fingers through the sand, to dart
into the sand again much like a lancelet or young lamprey, or to dodge under a
rock. Mr. E. B. Williamson wrote me that he noticed another species clinging to
the vertical sides of a waterfall. It looked like a water-weed, but he found by
watching closely that every little while the supposed weed would move up the wall
a short distance, and by using his butterfly-net he secured specimens. It is this
habit assisted by the opercular spines that accounts for the fact that the species of
the genus Pygidium are found in every mountain-stream.

The habit of insinuating themselves into crevices is undoubtedly also the
starting point of the habit of resorting to the gill-cavities and probably other organs
of larger fishes. There is a widely distributed behef among the Indians of the
Amazon Valley, that fishes called "C'andirii" enter the urethrse of bathers. Some
travelers who have had this habit reported to them have simply dismissed the matter
as absurd. Others have made attempts to identify the fish with results that have
not always been fortunate. The native name, Candirii, is applied to some fishes
{Cetopsis of the Cetopsidce) at least a foot long, and at least two inches thick, as well
as to minute slender fishes, species of Vandellia, which might enter the urethra
without violating the law that the greater cannot enter the less. The habit has
been attributed to the large Cetopsis, to Pareiodon, more moderate in size and yet
too large, and to some species of Vandellia and to Acanthopoma. It is, of course,
possible that the young of the larger Candirus have the urinophilous habit. It is
also possible that the Indians consider the small Candirus (species of Vandellia)
as the young of the larger Candirus, members of the genus Cetopsis, which according
to the classification adopted, belong to a different family. The habit is also phys-
ically possible for the species of Tridens, of Miuroglanis, of Paravandellia, Stego-
philus, Branchioica, and for some of the minute species or young of Pygidium.
However, these have not been indicated as being Candirus. As far as I am able
to find, the first notice of the peculiar habit is given by Spix {Selecta Genera et Species
Piscium, 1829, p. viii), who says of Cetopsis:

De alio pisce hoininibus infesto nonnulla afferre debeo, quern Brasilienses Candiru, Hispani
in provincia Maynas degentes Canero niincupaiit. Singulari enira instinctu incitatur in ostia
excretoria corporis humani intrandi, quae quum igitur in iis, qui in flumine lavant, attingit,
summa cum violentia irrepit, ibeque carnein morsu appetens, dolores, irao vitte periculum affert,
Urinse odorc hi pisciculi valde alliciuntur, quam ob causani accolse intraturi flumen amazonum,
cujus sinus hac peste abundant, praeputium ligula constringunt, et a mingendo abstinent. Pertinet
hie piscis ad Cetopseos, quod dcpinximus, genus; at nescio, an descriptaruni specierum (C.
candiru et C. caecutiens) individua juuiora, an tertise cujusdam speciei minoris individua crudeli
hoc instinctu a natura sint donata.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 263

I am indebted to Professor Selatie E. Stout for the following translation:

I should briefly mention another fish which is dangerous to man. The Brazihans call it
Candiru; the Spaniards in Maynas' call it Canero. It is impelled by a curious instinct to enter
the excretory openings of the human body. Whenever it comes in contact with these openings
of persons bathing in the stream, it violently forces its way in, and having entered, it causes
constant pain, and even danger of life, by biting the flesh. These fishes are greatly attracted by
the odor of urine. For this reason, those who dwell along the Amazon, when about to enter the
stream, whose bays abound with this pest, tie a cord tightly around the prepuce and refrain from
urinating. This fish belongs to Cetopsis, a genus which I have already described. But I do not
know whether it is the younger individuals of the two species which I have described (C candiru
and C. coecutiens), or whether a third species of smaller fishes has been given this cruel instinct
by nature.

The habit here described by Spix in reality belongs to fishes of which he did
not secure specimens.

In 1808, Domingo Vandelli, professor of natural history at Lisbon, sent Lace-
pede three small fishes, which he placed with the Loricariidce.. They were described
by Valenciennes (Cuvier & Valenciennes, Hist. Nat. Poiss., XVIII, 1846, p. 386,
pi. 547) as Vandellia cirrhosa, and placed in their Esoces. Nothing was said of the
habits, and even the habitat of the specimens was unknown. The identification
of Vandellia with the urinophilous habit came later.

Castelnau in his Animaux d'Amerique du Sud, Poissons, 1855, says of his
Trichomycterus pusillus = Pareiodon microps Kner:

Cette espece est, de la part des pecheurs de I'Araguay, I'objet d'un prejuge des plus singuliers,
ils pretendent ciu'il est tres dangcreux d'uriner en riviere: car, disent ils, ce petit animal s'elance
hors de I'eau et penetre dans I'urethre en remontant le long de la colonne liquide.

As this species reaches a length of at least si.x inches and a corresponding thick-
ness, Castelnau was probably mistaken in the species acting in this remarkable
manner.

It seems that Paul Marcoy {Voyage a travers VAmerique du Sud, Vol. II, p.
145-147) gives an account with a figure of a Candiru. I have not seen this book,
but Liitken says: "Etude de Candiru signeret med den Rejsendes Initialer, er en
fuldstandig Umuhghed, hvad den saa skal iorst'iWe" {Vidensk. Meddel. Naturh.
Foren. Kjobenhavn, 1891, p. 60).

Lange "In the Amazon Jungle," p. 214, says:

In fact, throughout the Amazon this little worm-like creature, called the kandiroo, is so
omnipresent that a bath-house of a particular construction is necessary. The kandiroo is usually
" Probably Maind, an Igarape tributary to the Amazon, near the Rio Negro; or a province of Peru
with Moyobamba for its capital.

264 MEMOIRS OF THE CARNEGIE MUSEUM.

three Id four inches long and one-sixteenth in thickness. It belongs to the lampreys, and its
particular group is the Myxinos or slime-fish. Its body is coated with a peculiar mucus. It is
dangerous to human beings, because when they are taking a bath in the river it will approach
and with a swift, powerful movement penetrate one of the natural openings of the l)ody, whence
it can l)e removed only by a difficult and dangerous operation.

A small but hard and pointed dorsal fin acts as a barb and prevents the fish from being drawn
back. While I was in Remate de Males the local doctor was called upon to remove a kandiroo
from the urethra of a man. The man subsequently died from the hemorrhage following the
operation.

The Candim does not belong to the lampreys and its particular group is not
Myxinos. The lampreys are not foimd in the Amazon Valley. Its dorsal fin is
neither hard nor pointed, and hence cannot act as a barb to prevent the fish from
being withdrawn. The retrorse spines on the interopercle and opercle are the
obstacles which would prevent it from being withdrawn.

The question naturally arises: Is Lange more trustworthy in his account of
the habit than in his account of the structure and relationship of the Candirii?

The only known specimen of Acanthopoma annectens, another Candirii, seems
to have been collected by Gustav Wallis. In an article, " Mittheilung von C. MiiUer
uber die Reise von Gustav Wallis" in Die-Natur, Zeitung von P. Ule u. K. Mliller,
XIX, No. 23, p. 180, mention is made of the habits of presumably this species,
though it may have been drawn from the general report given the traveler concern-
ing the Candirii. Liitken quotes:

In diesen noch so wenig bekannten Gewassern, namentlich im Huallaga, beobachtete der
Reisende (G. W.) einen Fisch, den ich der Aufmerksamkeit der Wisscnschaft ganz besonders
empfehlen will. Man nennt ihn dort den Candiru und furchtet ihn mit Recht ebensosehr fiir
das Gebiet des Wassers, wic man fur das des Landes die Moskitos und Ameisen fiirchtet. An
sich selbst ist es nur ein kleines, kaum .75 Spannen langes Ding von welsartigem Korperbau,
mit breitem, abgerundetcm Kopfe, auf dem die beiden Augen ziemlich dicht neben cinander liegen,
wahrend die beiden Brustflossen fliigelartig dicht unter ihm sich ausbreiten und der iilirige
Korpertheil keilformig zulauft. Den Riicken ziert eine dunklere Farbung mit undeutlich ver-
laufenden Flecken, so dass das Geschopfchen an sich selbst kaum irgendwie durch eine hervor-
ragende Eigenthiimlichkeit ausgezeichnet ist. Eine umso schrecklichere Plage ist es fur den
Badenden, eine Art Blutegel njimlich, der mit unglaublicher Schwimmfertigkeit jenem zu Leibe
geht, ihm iiberall schropfkopfahnlichc Wunden beibringt und, wcnn es ihm gelungen, sich dadurch
an dem Korper festzusetzen, in der Wunde ein Nadelbiindel ausspreizt, an dem er wie an Wider-
haken sich derart festklammert, dass er nur diuch eine schmerzhafte Operation aus dem Korper
entfernt werden kann. Diese Unart des Fisches ist umso grosser und gefiihrlicher als er am lieb-
sten die geheimsten Korperthcile aussucht; man erziihlt sich Fiille, die bei der Operation mit dem
Tode endeten. Ich werde dafur Sorge tragen, dass dieser seltsame Fisch, den ich in Spiritus vor
mir habe, in die rccliteii wissenschnfl liclicii H-iiidc golangt und sciiicii wissenschaftlichen Namen
empfiingt, den er noch iiicht hat.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATPISHES. 265

Boulenger (Proc. Zool. Soc. London, 1897, p. 901) says of Vandellia cirrhosa,
the iirinophilous Candirii, par excellence:

The "Candyru," as the fish is called, is much dreaded l)y the natives of the Jurua district,
who, in order to protect themselves, rarely enter the river without covering their genitalia by
means of a sheath formed of a small coconut-shell, with a minute perforation to let out urine,
maintained in a sort of bag of palm-fibers suspended from a belt of the same material. The fish
is attracted by the urine, and when once it has made its way into the urethra, cannot be pulled
out again, owing to the spines which arm its opercles. The only means of preventing it from
reaching the bladder, where it causes inflammation and ultimately death, is to instantly amputate
the penis; and at Trcs Unidos, Dr. Bach had actually examined a man and three Ijoys with
amputated penis as a result of this dreadful accident. Dr. Bach was therefore satisfied that the
account given of this extraordinary haliit of the "Candyru" is perfectly trustworthy. Mr.
Boulenger further showed a photograph, taken by Dr. Bach, of two nude Indians wearing the
protective purse.

It is to be noted here that this evidence is only circumstantial. Dr. Bach did
not himself operate or help to operate to remove the Candirii and a much simpler
operation than amputation would be sufficient to remove it.

The literature on the evil repute of members of the Pygidiidce has been re-
viewed by Pellegrin. In Bidletin Societe Philomathique de Paris (10), I, 1909, pp.
101-104 [5-8 of the reprint], he says:

Le Dr. C. Jobcrt qui accomplit au Bresil, en 1877, un voj'age oii il rassembla des materiaux
ichthyologiques considerables, a consacre a la question du Candiru un memoire des plus docu-
mentes, ou il n'admet pas sans reserve Ics declarations du practicien americain cite par G. A.
Boulenger. 'Le Dr. Bach,' ecrit-il, ' n'a pas vu le petit Poisson in situ; la chose est regrettable et,
cctte fois encore, nous ne sortons pas du cercle de la legende.'

Toutefois, le Dr. Jobert rapporte les dires d'un medecin tres estime de Belem (Para), le Dr.
Castro, qui lui affirma avoir extrait de I'urethre d'unc negresse un petit Candiru cpu y avait
penetre pendant la miction, alors qu'elle se baignait en riviere.

Mais ce qui fait le grand interet de I'article du Dr. Joljcrt, ce .sont les renseignements qu'il
a pu lui-meme recueillir sur place au Bresil au sujet des Candirus.

Les Paracuses en distingueraicnt deux especes, I'une petite, qui s'introduirait dans I'urethre
des baigneurs, I'autre de plus grande taille, 'trop grande ]K>ur tenter ces memes operations, mais
redoutable par les blessures ciu'elle fait sur n'importe quelle partie du corps. On donne a cette
derniere le nom de Candiru de Cavallo et les indigenes pretendent qu'elle attaque les chevaux
pendant la baignade.' Au sujet de celle-ci il rapporte en outre les faits suivants:

'Un jour, a un mille environ en aval de Para, je voulus me baigner sans souci des Candirus
qu'on m'assurait etre tres abondants en cet endroit. Je n'etais pas dans I'eau depuis cinq minutes
que je ressentis dans le region lombaire, au ventre, sur le cotes de la poitrine, comme de legers
coups de griffes qui se succedaient rapidement. Voyant I'eau se teinter de rouge autour de moi,
je me hatai de regagner le rivage et je constatai que, dans le region ou j 'avals eprouve la sensation

266 MEMOIRS OF THE CARNEGIE MUSEUM.

de ces coups de griffe, le sang s'6chappait de blessures en scarifications paralleles, qui eussent pu
etre attribuees a un instrument, tant elles etaient regulieres; elles constituaient des groupes de
5 a 6 lignes, longues d'un centimetre au plus et tres rapproch(5es; je n'ai pas cherche a apprecier
a profondeur, mais ces blessures tres ^troites saignaient abondamment.'

Les Poissons qui ont attaqu^ ainsi le Dr. Jobert appartiennent suivant moi, incontestablement
au genre Vandellie, peut-etre meme a I'espece Vandellia Wieneri. Si Ton se reporte a la descrip-
tion donnee plus haut de la bouche et de I'appareil operculaire, on s'expliquera ainsi facilement le
fonctionnement de ces divers organes; on comprendra ais^ment que la dcmi-couronne de dents en
crochet plac6e en avant de la bouche, dents susceptibles d'un certain degre d'eredion et au nomhre
de 5 a 6 principales produit ces scarifications paralleles, reguliere et en groupe de 5 a 6 lignes. Les
epines interoperculaires du dessous de la tete, aussi un peu erectiles, peuvent egalement, dans
une ccrtaine mesure, dechirer les teguments, mais elles doivent sourtout servir a la fixation.
Quant aux epines operculaires du dessus de la tete, elles me semblent plutot, etant donnee la
direction de leur pointe, destinees a faciliter la progression de I'animal et a empecher tout recul
lors-qu'il s'engage dans un conduit 6troit, par example entre les lamelles branchiales des Platy-
stomes.

Sans vouloir trancher la question de la penetration des Vandellies dans I'urethre, pour laquelle
je ne puis apporter des documents nouveaux, il me parait tout au moins demontre en rapprochant
les details anatomiques que j'ai pu constater sur les Vandellia Wieneri, des observations faites
sur lui-meme au Bresil par le Dr. Jobeii, que les Candirus, veritables Poissons-sangsues, ne sont
pas, ainsi que le pensait Giinther, de simples commensaux des grands Silurides sur lesquels ils
vivent habituellement; leurs dents et leurs epines operculaires ct interoperculaires permettent
non seulement de se fixer sur les branchies de leur hote, mais aussi de faire des blessures amenant
un ccoulement de sang abondant qu'unc disposition speciale leur permet d'ingurgiter. Enfin a
I'etat Ubre, comme la constats le Dr. Jobert, les Vandellies ne craignent pas de s'attaquer a
I'Homme, dont elles percent les teguments, ce qu'elles font aussi certainement sur certains
Mammiferes domestiques. II y a lieu en terminant de noter que les dents volumineuses peu
nombreuses, en forme de crochets aceres de la machoire superieure, sont particulieres au genre
Vandellia, qu'elles sont absentes dans les genres voisins Stegophilus Reinhardt et Acanthopoma
Liitken, ou elles sont remplacees par une bande de tres nombreuses petites dents acerees.^

Les Vandellies representent done, chez les Silurides, le dernier termc de la specialisation en
vue d'un parasitisme des plus caracterises.

That fishes found in the Amazon Valley and called Candirus are a nuisance
is certain. Whether the widely prevalent belief that the Candirii is tropic to
urine, and consequently has a tendency to enter the urethra, or whether the
Candiru's tendency to burrow leads it accidentally to enter the urethra, are all
matters that must for the present remain in debate. A very interesting subsidiary
question is, whether, if Candirus are tropic to urine they do not also enter the

■* While members of the Slegophilini have bands of minute teeth uniform in size in the upper jaw
there are frequently a few elongate, slender teeth in tlie middle of the upper jaw, which are similar and
correspond to those of Vandellia.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 267

urethrse of aquatic mammals and of large fishes. Further study may demonstrate
that some species of Candiriis have become parasitic in the bladders of large fishes
and aquatic mammals. These are all questions that may legitimately be taken
up by future expeditions.

The first of the commensals or parasites of this family to be described is the
Stegophilus insidiosus of Reinhardt. Reinhardt secured all of his specimens from
the gills of the giant catfish of the Rio das Velhas, a tributary of the Rio San Fran-
cisco. Haseman secured one specimen of this fish from the sandy island opposite
Januaria, near the mouth of the Rio San Francisco. The fish therefore may and
does five in the open as well as in the gill-cavities of larger fishes.

The account of Stegophilus insidiosus Reinhardt, given by the author of the
genus and species, which was published in 1858 (Cf. N aturhistorisk Forenings
Videnskabelige Meddelelser, Copenhagen, 1858, reprint, pp. 1-19, PI. II) possesses
great interest. Professor Reinhardt having been repeatedly informed that a large
species of catfish, belonging to the genus Pseudolatystomus and known by the natives
as Sorubim, protects its young by carrying them in its gills, determined, if possible,
to verify the statement. An English translation of a portion of his narrative is here
given :

It deeply interested me to ascertain with exactness the circumstances under which this pecu-
Har method of protection talies place, and also to examine the young at the time when they make
use of it. I therefore offered the fishermen in the vicinity of Lagoa Santa, where I was staying at
the time, a good sum if they woukl bring me a Sorubim with some of its young in the gill-cavities.
Finally on February 27, 1852, a fisherman brought me one, in the gills of which he said there
should be a Httle "young one." On examination I indeed found there a young fish, hardly an
inch long, which was already dead, although the Sorubim still showed faint signs of life. The
little fish looked so unlike the big one that I was astonished, and upon finding out that the old
fish was a male I was strengthened in my doubt as to their relationship. When the same fisher-
man two days later again brought a male Sorubim with a young one, which looked exactly like
the first, but was about three times longer, it became clear to me that these two small fishes
could in no wise be what it was claimed they were. On the other hand they recalled to me the
picture I carried in my mind of a Trichomyderus which I had obtained one year previously from
the Rio das Velhas under the name of Cambeja, or Bagre molle. I naturally concluded that the
fisherman in order to get the reward offered, had brought me the young of this Cambeja and was
passing them off as the young of the Sorubim. I complained to his face about this procedure,
and, though I did not obtain any confession from him, I nevertheless had no doubt that I had been
made the victim of a swindle. During the few weeks I still remained in Lagoa Santa before start-
ing on my homeward journey to Europe, nothing happened to induce me to think otherwise.

Upon my return home, as soon as I could get access to the literature, and could make a direct
comparison between the supposed young of the Sorubim and the Cambeja, I at once saw that I
had made a mistake in assuming that the former were the young of the latter. In- short these

268 MEMOIRS OF THE CARNEGIE MUSEUM.

so-called young of the Sorubim were the little fishes which I have had the honor of exhibiting to
the Society. The whole matter became more involved and enigmatical to me, because it appeared
that the fisherman, if he had been really guilty of an intended fraud, had for this purpose made
use of a fisli which was so rare that I had never found it, although I had collected great quantities
of the various small fishes in the waters around Lagoa Santa; in fact a fish which I was forced to
conclude to be as difficult to obtain as the real young of the Soru])im. In 1854, when I again
visited Brazil, the solution of the riddle was one of my especial aims. Soon after I arrived at
Lagoa Santa in the latter part of November I indeed reached the solution much more quickly
than I had expected, and in the following manner:

A person from the vicinity of Lagoa Santa, liut not the same one, who almost three years
before had brought me the first Stegophilus, came to the village on a Sunday in the middle of
December to attend mass according to the custom of the country. Ho brought with him on this
occasion a Sorul)im, which before he went to church he sold to a Frenchman who had a shop in
the town. When mass was over he returned to get hi.s pay, and watched the shopkeeper cut the
fish into pieces. He remarked that when the fish had been pulled out of the water there had been
five young in its mouth, of which two had remained inside. The shopkeeper looked and actually
found the remaining "young," and was kind enough, as he knew the matter would interest me,
to immediately bring them to me and relate the circumstances.

At the very first glance at the so-called "young" I saw to my surprise that again Stegophili
had been brought me as the young of the Sorubim. That deception shoidd again be at the bottom
of the matter appeared in the highest degree improliable. It could hardly be thought of, except
upon the assumption that the person who had .sold the last Sorubim was in collusion with the
fisherman who during my jirevious stay, three years before, had brought me the first two Steg-
'0])hili. How could it be explained that both had conceived the idea of passing off the very same
fish as the young of the Sorubim, and that a fish, which has no particular resemblance to the latter?
But, even if there had been collusion, would it not have been more likely that the first party con-
cerned would have come directly to me with his "Sorubim young," instead of leaving it to be more
or less of a chance whether or not they should fall into my hands? Even if a trick, prearranged
to allay a possible suspicion, were thinkable, nevertheless it was hard to believe that under the
existing conditions the parties involved would have taken the time and the trouble to deceive me,
unless they had expected to reap advantage from their efi'ort. If a trick had been planned in the
present case it was entirely aimless, as no pay was either asked, or given, for these last "young
Sorubim " ; and neither the last person, nor any one else, came at a later date to offer me " Soru])im
young." There was therefore left for me no other alternative than to conclude that I had been
unjust in my suspicion in the case of the fisherman who on the occasion of my previous stay had
brought me the first Stegophili. In other words, this little fish in reality passes into and abides
in the gill-cavities of the Sorubim. Its presence there has through an easily explained mis-
interpretation on the part of the common people given rise in Minas to the story about the
Sorubim 's care for its young.

The second species, Branchioica berlomi, known to inhabit the gill-cavities of
larger fishes, is recorded in the present volume. It really belongs to the Vanddliina'.
One specimen was sent me several years ago by Sh. A. de W. Bertoni from Puerto

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 269

Bertoni, Paraguay. Later he sent me two more specimens, all three having been
taken from the gills of a large characin, Piaractus hrachyjwmus.

Ribeiro, of the National Museum of Rio de Janeiro, caught another very similar
member of this subfamily, Paravandellia, among the water-weeds of the stream near
San Luis de Caceres, in the Upper Paraguay basin.

With fishes as rare as these and as small as these, the question arises whether
two species are really different, or whether the described differences are due to the
fact that one worker uses a hand lens, and the other a binocular dissecting micro-
scope with an arc spot-light. The results of the two instruments are comparable
to the effects produced by an old-fashioned cannon and a modern forty-two centi-
meter howitzer. Branchioica and Paravandellia may prove to be synonymous.

Distribution (Plates XXXVI-XXXIX.)

In considering the distribution of the fresh-water fishes of South America I
found, among other things {Proc. U. S. Nat. Mus., XIV, 1891, p. 18) "that genera
of many species usually have a wide distribution, and conversely, genera of wide
distribution usually have many species." With one exception the number of
species of any genus of the Pygidiidce varies directly with the greatness of the area
over which it is distributed. Some genera consist of but one species, and that
restricted to but one, or a few neighboring localities. As far as known, Eremophilus
is all but confined to the plateau of Bogota, Scleronema to the center of the Uru-
guay basin, Acanthopoma to a part of the Huallaga basin, Stegophilus to the Upper
San Francisco l:)asin, Paravandellia to the Upper Paraguay basin, Branchioica to
the Lower Paraguay basin. The genera with more than one species invariably have
a wider distribution. Homodicetus, with two species, is limited to the lower and
central La Plata basin, Henonemus, with four species, to the Amazon basin, Hatch-
eria, with six species, to the Andes of central and southern Argentina and Chile,
and Pygidium, with sixty-three species, is found in all the mountain streams from
the Tuyra in southern Panama to central Chile and central Argentina, in the moun-
tain streams from Rio Grande do Sul to the Rio Sao Francisco, and sparingly in the
lowlands of Guiana and Brazil. The only exception to the general rule is Och-
macanthus, with three species, ranging from Guiana to Paraguay.

The Pygidiinm are mountain forms, and while they are found in lowlands near
the mountains, we find the optimum in the plains of Bogota and in Lake Titicaca.
They are sometimes the last species to succumb in the struggle with adverse con-
ditions found in high altitudes, and they range further south (to latitude 47° 30' ),
than any other tropical American fishes.

270 MEMOIRS OF THE CARNEGIE MUSEUM.

The Stegophilimv, Vandelliince and Tridentince are essentially lowland forms,
although some species reach considerable elevations.

Chronology.

The first species of the Pygidiidoe. discovered was taken by Humboldt at Bogota,
and described in 1805 {Recueil d' Observations de Zoologie, etc., pp. 17-19, pi. VI)
as Ereniophilus mutisii.

The habit of one of the species was next described by Spix in 1829, but attri-
buted to a member of another family. See page 262.

The most prominent genus was first described by Meyen (Reise, I, p. 475,
Wiegm. Arch. Naturg., 1835, II, p. 269) as Pygidium. I have at diverse times
defended the name, Pygidium, as against the name TricJwmycterus and its varia-
tions.

The various generic names and their present equivalents are given in the fol-
lowing table :

Name Proposed. Proposed in Present Equivalent.

Eremophilus Humboldt 1805 Eremophibis Humboldt.

Thricho7nycterus Cuvier & Valenciennes 1S0.5 Eremophilus Humboldt.

Trichomycterus Valenciennes 1833 Pygidium Meyen.

Vandellia Cuvier & Valenciennes 1846 Vandellia Cuvier & Valenciennes.

Thrychomycterus Cuvier & Valenciennes 1846 Pygidium Meyen.

Thrichomycterus Girard 18.55 Pygidium Meyen.

Pareiodon Kner 1855 Parciodon Kner.

Centrophorus Kner 1855 Pareiodon Kner.

Stegophilus Reinhardt 1858 Stegophilus Reinhardt.

Astemomycierus Guichenot 1860 Pareiodon Kner.

Pariodon Giinther 1864 Pareiodon Kner.

Trachypoma Giebel 1871 Eremophilus Humboldt.

Tridens Eigenmann & Eigenmaim 1889 Tridens Eigenmann & Eigenmann.

Pseudoslegopkilus Eigenmann & Eigenmann .... 1889 Pseudostcgophilus Eigenmann & Eigenmann

Miuroglanis Eigenmann & Eigenmann 1889 Miuroglanis Eigenmann & Eigenmann.

Acanthopoma Liitken 1891 Acanthopoma Liitken.

Homodiatus Eigenmann & Ward 1907 Homodiwlus Eigenmann & Ward.

Hmoncmus Eigenmann & Ward 1907 Henonemus Eigenmann & Ward.

Hatcheria Eigenmann 1909 Hatcheria Eigenmann.

Ochmacanthus Eigenmann 1912 Ochmacanthus Eigenmann.

Gyrinurus Ribeiro 1912 Ochmacanthus Eigenmann.

Paravandellia Ribeiro 1912 Paravandellia Ribeiro.

Cobitoglanis Fowler 1914 Henonemus Eigenmann.

Urinophilus Eigenmann 1917 Urinophilus Eigenmann.

Branchioica Eigenmann 1917 Branchioica Eigenmann.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 271

Location of the Types and Specimens in the Museums of the World.

The species are, for the most part, but little known. Over forty of the ninety-
five recorded species are known only from the types, which are widely scattered.
Ten or twelve of the types are in Vienna, two are in Berlin, eleven or twelve in
Paris, eleven in London, one in Torino, Italy, two possibly in Munich, one in the
University of Leipzig, two in Copenhagen, three presumably in Santiago, Chile,
three in Buenos Aires, five in Rio de Janeiro, two in Cordoba, Argentina, one in the
Field Museum, two in the Philadelphia Academy of Sciences, eight in the Museum
of Comparative Zoology at Harvard, eight in Indiana University, twenty-four in
the Carnegie Museum, one in Princeton University. The Carnegie Museum pos-
sesses forty-six species, Indiana University is next in line with thirty-two species,
and the Museum of Comparative Zoology comes third with twenty species.

The distribution of the known specimens in the various museums of the world
is given in the following table:

a

3

3
r^
CS

n
c

OJ

>

a

3
d)
w

3

.g

n

a

3
3

a

3
a;
«
3

s
1

•c

P5

d
a

(S
CO

i
<

g

3

.2

'v

a

03

■a
o

1
i

o

'S

a
1

3

C3
02

1

<

a

3

s

3

•3

§

02

w
o

O -

d,

a
6

a

3
3

1
>

a

d
>— 1

a

3

s
s

. ■&
2

6

1. Nematogenys inermis.

2. Scleronema opercula-

tum

type

type
type

+
+

+

type

?

type?
type?
type?

+

type

type

+

type

+

+

+
+

type

+

type

type

3. Hatcheria patagonien-
sis^

+

4. Hatcheria maculata. . .

5. " titcomhi . ..

6. " areolatum..

7. " burmeisteri.

8. " macrcei....

9. Pygidium marmor-

atum

10. Pygidium palleum . . .

11. " tigrinum . .

12. " tenue^

13. " corduvensis^

14. " spegazzinii .

15. " borellii'' . . .

16. " eichornia-
rum

+
+

^ Type in Princeton University.
^ Types in C6rdoba, Argentina?
' Types in Mus. Univ. Torino, Italy?

272

MEMOIRS OF THE CARNEGIE MUSEUM.

S

3

0

t*^

a

J2

s

a

3

S

g

^

fc

;5

N

1

1

S
3

0^

B
:3

3

9

1

73

g
3

^

13

0.

I

3

S

a

3

m

22

3
to

05

0

n
«1

•1

6
Z

C3

0

0
>

'5

a

2

0

IB

0
0

a

i
3

a

■&

a

a

17.

Pygidium riojanum. .

type

18.

" heterodon-
tum

type

19.

Pygidium fuscum ....

type

20.

" eigenmanni

type

21.

" vittatum. . .

type

22.

" dispar ....

type

+

23!

" punctula-
tum

type

+

+

24.

Pygidium taczatiowslcii

type

+

25.

" rivulatum .

type

+

+

+

+

26.

" poeyanum .

type

27.

fassli

type

28.

" barbouri . .

type

+

+

29.

' ' oroyac

+

type

+

+

30.

" quechiwrum

type

31.

" laticeps^ . . .

type?

+

+

32.

" stellatum . .

+

type

33.

" chapmani .

+

type

34.

" tmniurrfi. . .

type?

+

+

35.

" caliense . . .

•

+

type

36.

" latidens . . .

type

37.

" melee

type

38.

" slramineum

+

type

39.

" unicoloT. . .

type

40.

" kneri

type

+

41.

" meridce . . .

type

+

42.

" bogotense . .

+

+

type

43.

Pygidium nlgroma-
cululum

type

+

+

+

44.

Pygidium banneaui. .

+

type

45.

" spilosoma .

type

+

46.

" dorsoslri-
aium

+

type

47.

Pygidium venidosum-

type

48.

" lulldri-
aium

type

49.

Pygidium striatum

type

+

+

+

50.

" regani ....

type

51.

" relropinne.

type

52.

" guianense .

type

53.

" conradi. . .

+

type

54.

" graciliur . .

type

55.

" amazoni-
cum

type

56.

Pygidium hasemani. .

+

type

57.

" nigricans . .

type

58.

" iheringi . . .

type

+

59.

" zonatum . .

type

GO.

" proOps ....

type

+

61.

" paolence . .

type

62.

" reinhardti .

type

63.

" davisi

type

" Types possibly in Munich.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 273

64. Pygidium immacu-

latum

65. Pygidium vermicu-

latum

66. Pygidium alternatum.

67. " gocldii. . . .

68. " brasiliense.

69. " ilatiaym . .

70. " iriguttalum

71. " pundatis-
simum

72. Pygidium minuium . .

73. " sanice-ritcB .

74. Eremophilus muiinii.

75. Pareiodon microps . ..

76. Pseudostcgophilus

iicmurus

77. Homodiwtus anisitsi .

78. " macula-
tus

79. Henonemus macrops .

80. " pundatus

81. " laxis-
.tigma

82. Henonemus interme'

dius

83. Stegophilus insidiu-

SMS'

84. Acanthopoma an-

nedens'°

85. OchmacaiUhus ba-

trachostoma

86. Odtmacanthus rein-

hardli

87. Ochmacanthus flabel-

lifcrus

88. Vandellia cirrhosa. . .

89. " plazai....

90. " wieneri . . .

91. " hasemani..

92. " sanguinea .

93. Paravandellia oxyp-

tera

94. Branchioica bertonii .

95. Tridens melanops.. . .

96. " brevis

97. Miuroglanis plaly-

cephalus

type

type
type

type

m

05

type

type
+

type
type
type

type
?

type

type

type

+

3

+

type

+

type

+
+

type

S3

fa

+

type

a

type

+

type

+
+

type
type

type

+

+

type

+

+

type

+

+

type
type

+
type

type

+

+

+
+

+

+

+

type

+

type
type

+

' Types in Copenhagen.

'" Types in the collection of Professor Leuckhart.

274 memoirs of the carnegie museum.

Sources of the Material Examined.

In 1890 Mrs. Eigenmann and myself published a revision of the Pygidiidce as
part of the general monograph on the Nematognathi of South America (Occasional
Papers California Academj^ of Sciences, Vol. I, 1890, pp. 316-347). Our account
was based on the material in the Museum of Comparative Zoology, which was col-
lected during the Nathaniel Thayer Expedition to Brazil, 1865-1866, during the
U. S. Naval Astronomical Expedition to the Southern Hemisphere, across the Andes
from Lima, 1849-1852, during the Hassler Expedition at Santiago, Chile, and Callao,
Peru, and during Alexander Agassiz's Expedition of 1875 to Lake Titicaca. I have
freely drawn on this monograph, which describes some species, which have not been
duplicated.

From time to time Mr. J. D. Anisits and Sr. A de W. Bertoni have sent col-
lections to Indiana University from Paraguay, containing, among other things,
the types' of Homodicetus and BrancMoica. Similarly collections were sent from
Sao Paulo by Messrs. Hermann and Rudolph von Ihering.

The collections made by the late J. B. Hatcher for Princeton University were
received and reported upon by me in Vol. Ill of the Reports of the Princeton Uni-
versity Expedition to Patagonia.

Miss Lola Vance made a small but valuable collection, containing specimens
of Pygidium oroyce, near Tarma, Peru.

The Yale-National Geographic Society Expedition to Peru collected a few
specimens in the Urubamba Valley, which are being reported upon in the Bulletin
of the Museum of Comparative Zoology.

I collected several species in British Guiana, which were described in the
Memoirs of the Carnegie Museum, V, 1912.

Mr. Thomas Barbour collected Pygidium barbouri in the Beni River in Bolivia.

Several specimens, some of them new, were purchased for the collection of
Indiana University from W. F. H. Rosenberg, London.

By far the greater and most valuable collections were secured in Ecuador and
Colombia, and in Brazil, Uruguay, Paraguay and the Argentine.

The collections from Colombia were made by several field-parties. I collected
between Bogota and Buenaventura and at Istmina. Mr. Arthur W. Henn col-
lected between Buenaventura and Istmina. Mr. Henn also collected in the upper
vaUey of the Patia and southward in the Andes of Ecuador. Mr. Manuel Gon-
zales collected in Colombia along the routes from Bogota west to Honda, north
to Mogotes, and east to Barrigona, securing a wealth of material. Messrs. A. S.
Pearse, M. A. Carriker, Jr., and Alexander Grant Ruthven collected in the Sierra

EIGENMANN: the PYGIDIIDiE, A FAMILY OF SOUTH AMERICAN CATFISHES. 275

Nevada de Santa Marta for the University of Michigan, and Mr. E. B. William-
son secured specimens for me in the Sierra Nevada, and in other places in Colombia.

Mr. J. D. Haseman, who collected for the Carnegie Museum, secured many
species, especially between the Rio Sao Francisco and Buenos Aires, as well as in
the upper Paraguay basin and in the Amazon.

At one time or another I have examined all of the seventy-one species pre-
served in American Museums, fifty-eight species in the Indiana University Museum
and the Carnegie Museum being under my immediate charge. Nine of the eigh-
teen known genera and forty-three of the ninety-five known species were described
by me during the course of my study.

I have attempted to collect what is known of the members of the family. I
hope the result will help the next one who undertakes the study of the group and
stimulate the collection of additional specimens and facts of the commensal or
parasitic members of the family.

The Zoological Position of the Pygidiid^.

Phylum PISCES Artedi.

Class TELEOSTOMI Bonaparte.

Superorder OSTARIOPHYSI Sagemehl.

Order PLECTOSPONDYLI Cope.

Family: Pygidiid^ Eigenmann & Eigenmann.

Subfamilies :
NEMATOGENYiNiE Gi'mther.
Pygidiin^ Eigenmann & Eigenmann.
Pareiodontin^ Eigenmann.
Stegophilin^ Giinther.
Vandelliin^e Eigenmann.
Tridentin^ Eigenmann .

'&"-

Synonymy.
= Siluroidei Trichomycteriformes Bleeker, Nederl. Tijdschr. Dierk., 1, 1863, p. 112.
> Siluridce Opisthoptercv Gunther, Cat. Fish. Brit. Mus., V, 1864, p. 4 and p. 271.

< Siluridce Branchicolw GIjnther, I. c, p. 4 and p. 276.

= Trichomyderidw Gill, Arrangement of Families of Fishes, 1872, p. 19.

< Pygidiidce Eigenmann & Eigenmann, Am. Nat., July, 1888, p. 649; Occasional

Papers California Academy Sciences, I, 1890, p. 316.

< Pygidiidce Gill, Mem. Nat. Acad. Sci., VI, 1893, p. 132.

276 MEMOIRS OF THE CARNEGIE MUSEUM.

< Trichomycteridce Regan, Ann. & Mag. Nat. Hist. (8), VIII, 1911, p. 57.
= Trichomycteridce Ribeiro, Archives do Mllscu Nacional, XVI, 1912, p. 219.

Limits of the family Pygidiid^e. (Plates XL and XLI.)

Giinthcr, in his "Catalogue of the Fishes of the British Museum," V, 1864, pp.
271-277, arranges the then known members of the Pygidiidce under three " Groups,"
belonging to two of his eight Subfamilies of the Siluridw. His seventh Subfamily,
the Siluridce Opisthopterw, consists of his Fifteenth Group, the Nematogenyina
(Hcpiapterus and Nematogenys) and the Sixteenth Group, the TricJwynycterina
{Trichnmyctcrus (= Pygidium) , Erc?tiophilus , Pariodon). His Eighth Subfamily, the
Siluridce Branchicola', consists of his Seventeenth Group, the Stcgophilina {Stego-
philus and Vandellia).

The genus Heptapterus" included in his Fifteenth Group, was shown Ijyus in
the American Naturalist, July, 1888, p. 648, to "have no real affinity with the
PygidiidcE.^'

We do not now feel justified in joining the Cetopsince to the family.

The PygidiidoB, as here understood, are the Pygidiince (exclusive of Pariolius
and the Stegophilince of the family, as described by Eigenmann & Eigenmann, in
the American Ncduralist, Jul.y, 1888, and Occasional Papers of the California
Academy of Sciences, I, 1890. The species known at the time, thirty-six in number,
belonging to eight genera, were reviewed in the last named paper. The Cetopsince,
included in the papers mentioned, constitute a distinct family. Regan (Ann. &
Mag. Nat. Hist. (8), VIII, 1911, p. 574) has united the Pygidiince and Stegopkilince
in his Trichomycterince of his Trichomycteridce = Pygidiidce. The family includes
the South American Nematognaths without an adipose fin, with the dorsal over or
behind the ventrals; posterior air-bladder obsolete; the anterior minute, in two
lateral parts, enclosed in bony capsules with a comj^lete osseous floor, united to the
exoccipital and epiotic bones proximally and to the suprascapula distally; neural
spine of the coalesced vertebrae very low, not as high as that of the vertebrae fol-
lowing them; parapophysis of the vertebra? following the capsule short; skull de-
pressed, entirely closed in front, without an open space between the osseous roof of
the mouth and the ethmoid; vomer and palatines weak, without teeth; clavicles
wide, scoop-shaped, meeting below. The place of the adipose fin sometimes occupied
in part by numerous accessory caudal rays; none of the fin-rays modified into
spines; nares remote from each other, the anterior one frequently provided with
a barbel; the maxillary ending in a short barbel; the lower lip usually ending in

" Related to Ileptaplerus is the genus Phrcatobius, for an account of which see the Appendix to this
paper.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 277

another shorter barbel just beneath the maxillary barbel; this lower labial barbel is
sometimes very minute and has been overlooked in describing some species of
Hcnonomus and Pseudostcgophiliis, and in some species of other genera.

Mental barbels, characteristic of many Nematognaths, are lacking, except in
Nematogenys. Thorn-like spines firmly attached to the opercle and the interopercle
in all but Nematogenys. The opercles and interoperclcs to which the spines are
attached are erectile, and by first erecting those on one side and then those of the
other, the fishes are able to "elbow" their way forward in narrow openings, under
rocks and up waterfalls. In some cases the spines are directed backward, but in
Vandellia the opercular spines point obliquely upward and backward, the interoper-
cular spines downward and backward.

All of the species secrete a copious mass of mucus, and the larger, ones are as

8 PSEUDOSTEGOPHILUS
1 HENONEMUS - '^

.1 HOMOO/^TUS g^^^^p^, M/1C.NTH0POA.A

I OCHMACANTHVS

bPARE10D0N-~^l

5ERE/W0PHILUS-^t

4PYC1DIUM

li Vandellia

/4()f?IN0PHI LUS

15 Paravanoellia

Vn >-n ">^i

IIoBRANCHIOICA

iiTridzns

8/VUUROC.LANlS

3 HATCHERIA
iSCLERONEMA

I Nematogenys ,

Fig. 1. Phylogeiietic tree showing the relationsliip of the Pijgidiidw. The letters correspond to
the letters in the key to the subfamilies and genera. The Ncmalogcnijince are undoubtedly the most
primitive of the family. The Pygidiinx have the family characteristics fully developed. Beyond these
we have the more highly specialized subfamilies, culminating in the parasitic Stcgopliilincc and the uri-
nophilous VandelliincB.

278 MEMOIRS OF THE CARNEGIE MUSEUM.

slippery as the proverbial eel, which they resemble in other respects. The pectoral
gland is very large in the smaller species.

Key to the Subfamilies and Genera of the Pygidiid^.

a. One pair of mental barbels, no opercular or interopercular spines; one barbel at angle of mouth; a
small nasal barbel; pectoral spine pungent; dorsal over ventrals. {Nematogcnyimv.)

I. Nematogenys Girard.
aa. No mental barbels; opercle and iuteropercle with spines; two barbels at angle of mouth; pectoral
spine not pungent.
h. A nasal barbel; mandible with considerable antero-posterior extent, teeth along less than half its
total length; teeth strong; anal short; no mental barbels; opercle and interopercle with
spines; two barbels at angle of mouth; free-living species, some of them of economic
importance. {Pygidiinm.)
c. Opercle with a long dermal flap; maxillary bone longer than the attached barbel; teeth nar-
row incisors; pectoral without a filament; anal short. ..II. Scleronema Eigenmann.
cc. Opercle without a dermal flap; maxillary very small.

d. Dorsal long; caudal peduncle subterete; anal usually entirely under the dorsal; outer

pectoral ray without a filament III. Hatcheria Eigenmann.

dd. Dorsal shorter; caudal peduncle compressed; anal partly or entirely behind the dorsal;
outer ray of the pectoral prolonged or not.

e. Ventrals present IV. Pygidium Meyen.

ee. No ventral fins; otherwise liiie Pygidium V. Eremophilus Humboldt.

66. No nasal barbels.

/. Mouth subtermiual, the teeth strong, in a single series; gill-membranes united with isthmus;

anal short. ( Pareiodontinm.) VI. Pareiodon Kner.

//. Mouth inferior.

g. Anal short, of 7-1 1 rays, its origin usually behind, rarely under that of the dorsal; lower
barbel at angle of mouth minute ; eyes superior. Species small, some of them
commensals or parasites.
h. Mouth wide, teeth very numerous, in several very regular series; rami of the lower
jaw transverse, meeting, with teeth along its entire length; premaxillary large.
{Slegophilinw.)
i. Accessory caudal rays few, not conspicuous; caudal not fan-shaped nor exces-
sively contracted at base; upper lip with fine, hair-like, movable teeth.
j. Gill-membrane confluent with the isthmus; gill-openings reduced to a
narrow slit in front of the pectoral.
k. Opercle with two spines. VII. . .Henonemus Eigenmann & Ward.
kk. Opercle with four to twelve spines.

I. Caudal deeply forked, the upper lobe prolonged; eight or nine
spines on the interopercle; color in bands; origin of ventrals
equidistant from caudal and angle of mouth.

VIII. Pseudostegophilus Eigenmann & Eigenmann.

II. Caudal emarginate or obliquely rounded, origin of ventrals

nearly equidistant from snout and caudal; color, if present, in

spots IX. Homodiaetus Eigenmann & Ward.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 279

III. Caudal rounded; few accessory rays; origin of ventrals one-and-
a-half to twice as far from snout as from caudal.

X. Stegophilus Reinhardt.
jj. Gill-membranes united, free from the isthmus . . XI. Acanthopoma Liitken.
w. Accessory caudal rays very numerous, the tail like that of a tadpole; base of
caudal very narrow; no hair-like teeth on the upper lip.

XII. Ochmacanthus Eigenmann.

hh. Mouth narrower, the rami of the lower teeth feeble, not transverse, not meeting in

the middle; teeth few, slender, pointed. ( Vandelliinw.)

m. A few depressible teeth in a single series in the middle of the upper jaw;

mandibles without teeth, or with a few excessively minute teeth on the

ends of the rami; caudal I'ounded <ir emarginate

XIII. Vandellia Cuvier & Valenciennes.

XIV. Urinophilus Eigenmann.

mm. A band of depressible teeth in the middle of the upper jaw; a series of much
smaller teeth laterad of the median band; no teeth on the mandible;

caudal forked, the upper lobe longer XV. Paravandellia Ribeiro.

mmm. Two series of depressible teeth in tiie middle of the upper jaw; a single
series of much smaller ones laterad of the median series and a claw-like
tooth on the end of the maxillary'^ of the median series; two short series
of teeth on the ends of the mandible; caudal subtruncate.

XVI. Branchioica Eigenmann.
gg. Anal long, with fifteen to twenty-five rays, its origin in front of that of the dorsal; eyes
large, lateral; caudal rounded or emarginate. {Trideidiiuc.)
n. Opercular and aubopercular patches of spines distinct ; head greatly depressed, the
eyes infringing on both the upper and lower surfaces of the head; mouth inferior;
a series of fine labial teeth and strong teeth in the jaws; gill-membranes united,
forming a broad free fold across the isthmus; opercle long, slender, ending in a
few thorns; interopercle with similar but smaller thorns.

XVII. Tridens Eigenmann & Eigenmann.
nn. Opercular and subopercular patches of spines confluent; head less depressed;
mouth subinferior; several series of strong teeth in each jaw; gill-membrane
broadly united with the isthmus, without a free margin.

XVIII. Miuroglanis Eigenmann & Eigenmann.

Genus I. Nematogenys^'' Girard.

Nematogenys Girard, Proc. Acad. Nat. Sci. Phila., 1854, p. 198.

Type. — Trichomyderus inermis Guichenot.

Origin of dorsal over or slightly in front of origin of ventrals, near middle of the
body; nasal barbels small; a single maxillary barbel at angle of mouth; a shorter
mental barbel below it; mouth wide, terminal; teeth in a broad band in each jaw;
gill-membranes narrowly joined to the isthmus; first pectoral ray spinous, with

" Possibly the premaxillary. See under the genus.

'^ vrjiia, TO = thread, yhvs. ij = jaw. In allusion to the maxillary barbel.

280 MEMOIRS OF THE CARNEGIE MUSEUM.

serrae on its posterior margin ; anal short ; fontanel extending to the base of the
occipital process, with a bridge over posterior margin of the eye; opercleand inter-
opercle unarmed.

This genus resembles members of the Pimdodiruc more than the other species
of the family, and is probably nearer the primitive stock of the family than the
following more highly specialized genera, if it is not a member of the Siluridcc.

Hahitat.— Chile.

1. Nematogenys inermis ((kiichenot). (Plate XLII, figs. 1, 2.)
Native name "Bagre."

TrirJwmycterus inermis Guichenot, in Gay, Hist. Ghil. Zool., II, 1848, p. 312;

1854, pi. IX, fig. 2 (Chile).
Nematogenys inermis Girard, Proc. Acad. Nat. 8ci. Phila., 1854, p. 198; U. S. Nav.
& Astron. Exped., 1855, p. 240, pi. XXXII (Rio Maypu near Santiago);
Gunther, Cat. Fish. Brit. Mus., V, 1864, p. 272; Philippi, Mb. Ak. Wiss.,
Berlin, 1866, p. 716; Eigenmann & Eigenmann, Proc. Cal. Acad. Nat. Sci.,
(2), 11, 1889, p. 50 (Curico; Santiago); Occasional Papers Cal. Acad. Sci., I,
1890, p. 323 (Curico; Santiago); Proc. U. S. Nat. Mus., XIV, 1890, p. 36;
Delfin, Catalogo de los Peces de Chile, 1901, p. 29 (Central Provinces of
Chile); Eigenmann, Reports Princeton Univ. Exped. Patagonia, III, 1909,
p. 246, pi. XXXI, fig. 2; 1910, p. 398, pi. XXXII, fig. 2.
Nematogenys nigricans Philippi, Mb. Ak. Wiss., Berlin, 1866, p. 716.
Nematogenys pallidus Philippi, I. c, p. 716.
Habitat. — Fresh-waters of central Chile.

Head 3.8-4.33; depth 6-7; D. 10; A. 11; P. I, 8; Br. 11 12; eye small, superior;
interocular little less than snout; caudal peduncle about as deep as body; origin of
dorsal one-fifth nearer snout than to base of middle caudal rays in specimens 120
mm. long, one-fifth nearer caudal than snout in specimens 260 mm. long; origin of
anal much behind the last dorsal ray ; fins all rounded ; caudal with numerous acces-
sory rays. Light l:)rown, mottled with darker; a series of about five light areas
along the lateral line; fins speckled.

Genus II. Scleronema'* Eigenmann. (Plate XXXVI.)

Scleronema Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. 691.
Type. — Scleronema operculatum sp. nov.
Allied to Pygidium.
Ventrals nearer snout tlian to caudal; outer pectoral rays shortest, without a

" aK\r,pU= hard, pr,y.a, t6 = MiiCMil, ill alliisidu to the hnrd base of the maxillary barbel.

EIGENMANN: the PYGIDIIDiE, A FAMILY OF SOUTH AMERICAN CATFISHES. 281

filament; opercle with a long dermal flap; interopercular spines in a much more
restricted area than in species of Pycjidimn; accessory rays of the caudal inconspi-
cuous; maxillary barbel with a large osseous base (maxillary). Teeth very narrow
incisors; mouth wide, terminal.

In other respects like Pygidium.

Habitat that of the single species.

1. Scleronema operculatum Eigenmann. (Plate XLIV, fig. 1.)

Sclernnema operculalitm Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. 691.
7077, C. M., type, one, 79 mm.; 7539 a-c, C. M. paratypes, 65-80 mm. Cacequy,

Uruguay basin. Feb. 1, 1909. Haseman.

Head 5.2-5.66; D. 12.5; A. 7.5 counting the rudimentary rays; P. 7; eye in
anterior half of the head; interocular five times in the length of the head; width of
mouth nearly half the length of the head.

Nasal barbel short, reaching just beyond posterior nares; maxillary barbel
reaching about half-way to the tips of the opercular spines, the bony base much
longer than the soft filament; a broad, free membrane above from near the anterior
nares to the tip of the osseous base of the barbel, a narrower membrane along the
outer edge of the base of the barbel; six spines in the main row of the interopercle ;
opercular flap reaching to or beyond base of the last pectoral ray; pectoral aliout as
long as the head ; origin of ventrals about equidistant from the snout and from the
base of the middle caudal rays; ventrals reaching beyond the anus, not quite to the
anal, equal to the portion of the head behind the nasal barbels; origin of anal under
the posterior part of the dorsal, the distance from the base of its last ray to the
caudal four times in the length; caudal narrow and long, equal to the length of the
head, its margin slightly obliciue, rounded; origin of dorsal over posterior half of
ventrals, the distance from the first ray to the caudal 1.3 in its distance from the
snout.

Middle of sides with a series of large spots, similar spots along the back.

Genus HI. HatcheriaI''^ Eigenmann. (Plate XXXVI.)

Hatcheria Eigenmann, Reports Princeton Univ. Exped. Patagonia, III, 1909, p. 248.

Type. — Hatcheria patagoniensis Eigenmann.

Origin of dorsal behind that of the ventral, which is near middle of the body;
nasal barbels large; two barbels of nearly equal size at the angle of the mouth; no
mental barbels; mouth terminal; teeth conic or narrow incisors, in a few series;

'* In nioniory of .J. B. Hatcher, lender of t-lie Princeton University Expeditions to Patagonia.

282

MEMOIRS OF THE CARNEGIE MUSEUM.

gill-membrane free or very narrowly joined to the isthmus; first pectoral ray not
continued as a filament; opercles and interopercles with numerous spines; dorsal

AC D

Fig. 2. A-C. Halcheria patagoniensis Eigenmanii. A. Skull from above. B. Opercular appar-
atus, etc., as seen from above when attached to the skull. C. The same spread out flat. D-E. Sclero-
nema operculatum Eigenmann. D. Skull from above. E. Opercular apparatus spread out flat. 1,
Premaxillary; 2, ethmoid; 3, lateral ethmoid; 4, nasal; .5, frontal; 6, sphenotic; 7, pterotic; 8, supra-occipi-
tal; 9, epiotic; 10, supraclavicle; 11, parapophysis of coalesced vertebrae; 12, maxillary; 13, palatine; 14,
metapterygoid; 15, quadrate; 16, pre-opercle; 17, interopercle; 18, opercle; 19, hyomandibular; 20, man-
dible.

long, emarginate; caudal peduncle slender, the fin wider than the peduncle, with
few accessory rays; origin and end of anal under the long dorsal, except in H.
areolata.

Habitat. — In the mountain-streams of northern Argentina, central Chile, and
southward. Replacing the species of the genus Pygidium east of the Andes in the
south, and largely also west of the Andes. Its definite boundaries not known.

Key to the Species of Hatcheria.
a. Dorsal with fourteen to seventeen rays.

b. Origin of dorsal equidistant from tip of caudal and some part of the snout; D. 15; last ray of anal
under the last ray of the dorsal or very little farther forward,
c. Distance between anal and caudal 4.5 in the length; origin of ventrals nearer the caudal than
the snout; origin of anal nearer tip of caudal than head; origin of dorsal equidistant from

tip of caudal and posterior nares 1. patagoniensis Eigenmann.

cc. Distance between anal and caudal 3.75 in the length; origin of ventrals nearer snout than to
caudal; origin of anal nearer the head than tip of caudal; origin of dorsal equidistant from
tips of snout and caudal 2. maculata (Cuvier & Valenciennes).

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 283

66. Origin of dorsal equidistant from tip of caudal and occiput.

d. Last anal ray under the dorsal; D. 17 3. titcombi Eigenmann.

dd. Last anal ray behind the last dorsal ray; D. 14. . . .4. areolata (Cuvier & Valenciennes).
aa. Dorsal with twentj'-one rays; anal entirely under the dorsal.

e. Ventrals nearer tip of caudal than snout 5. burmeisteri (Berg).

.ee. Ventrals nearer snout than tip of caudal 6. macraei (Girard).

1. Hatcheria patagoniensis Eigenmann.
Hatcheria patagoniensis Eigenmann, Reports Princeton Univ. Exped. Patagonia,
III, 1909, p. 250 (Rio Blanco, at base of Andes, latitude 47° 30', longitude 72°
W.; the southernmost record of the family); I. c, 1910, p. 399.
Habitat. — Eastern slope of the Andes between latitudes 47° 30' and 31° 30'.
Mr. Haseman collected the following specimens :

7084, C. M., two, 66 and 82 mm. San Juan, Argentina. Feb. 25, 1909.

7085, C. M., six, 34-77 mm. San Juan, Argentina. Feb. 26, 1909.

11370 and 11371, I. U. M., four, paratypes, 94-120 mm. Rio Blanco; Hatcher.

Fig. 3. Hatcheria -patagoniensis Eigenmann. From a paratype in Indiana University.

Head 5; depth 8; D. (13) 15; A. 6. Elongate, slender; caudal peduncle slender,
its depth nearly three times in the head, about four times in its length; upper
maxillary barbel reaching pectoral, lower maxillary barbel reaching to margin of
interopercle ; a broad lobe of skin joining base of lower maxillary barbel to the lower
lip; snout pointed, mouth narrow, its width 3.5 in head, equal to interorbital ;
nasal barbels reaching beyond eye; width of head but little less than its length;
greatest width of body behind the pectorals, 1.6 in the length of the head. Gill-
opening not extending forward to below eye; origin of dorsal equidistant from tip of
caudal and posterior nares; base of dorsal equal to its distance from the caudal, its
free surface emarginate, the anterior lobe rounded, the posterior pointed; beginning
of last third of dorsal not much more than half as high as anterior lobe. Caudal
moderate, emarginate, its lobes rounded, .8 of the length of the head. Anal
broadly rounded, its last ray about under last ray of dorsal. Ventrals broad, their
middle under origin of dorsal, 1.5 in head, equal to height of anal. Base of pectoral
horizontal, closing edgewise to the body, its lower part folded when appressed, its
first ray sickle-shaped, slightly prolonged. Dark yellowish, more or less regularly
spotted with darker; dorsal, caudal, and pectorals irregularly blotched with black.

284 MEMOIRS OF THE CARNEGIE MUSEUM.

Some of the cotypes are more robust in body; in one the anal is blotched like the
caudal; in some the spots form regular series along the sides, leaving lighter stripes
between them.

2. Hatcheria maculata (Cuvier & Valenciennes). (Plate XLII, figs. 3-5.)

Trichomycterus maculatus Cuvier & Valenciennes, Hist. Nat. Poiss., XVIII,

1846, p. 493 (San lago); Guichenot, in Gay, Hist. Chile, II, 1848, p. 311

(Chile); GDnther, Cat Fishes Brit. Mus., V, 1864, p. 273; Philippi, Mb.

Ak. Wiss. Berhn, 1866, j). 716 (Chile); Delfin, Catalogo, de los Peces de

Chile, 1901, p. 30.
Thrichomycterus maculatus Girard, Proc. Acad. Nat. Sci. Phila., 1854, p. 199;

U. S. Naval & Astron. Exped. 1855, p. 243 (Rio Mapocho).
Pygidium maculatum Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II,

1889, p. 51 (Rio Mapocho) ; Occasional Papers Cal. Acad. Sci., I, 1890, p. 329;

Proc. U. S. Nat. Mus., XIV, 1890, p. 36.
Hatcheria maculata Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,

1909, p. 249, pi. XXXIII, figs. 1, la and 16; 1910, p. 399.

Habitat. — Pacific slope of Chile.

Head 5.33; depth 7.5; D. 15; A. 9. Elongate, somewhat compressed; head as
long as wide; caudal peduncle long and slender. Eye small, midway between tip
of snout and end of opercle. Lips and lower surfaces of the head thickly covered
with warts. Gill-openings not continued forward to below the eye, the membranes
joined to the isthmus for a distance equal to one-third the width of the mouth.
Pectorals rounded, the first ray not produced; origin of dorsal in front of the vent,
but some distance behind the ventrals, equidistant from tip of snout and tip of
caudal, its last ray over the last ray of the anal. Caudal long, truncate. Anal
short and high, its height about equal to the length of the caudal, its distance from
the base of the caudal 3.75 in the length. Origin of the ventrals equidistant from
tip of snout and base of caudal, their tips reaching beyond the vent. Back and
sides marbled with light and dark brown; fins pale, immaculate.

3. Hatcheria titcombi Eigenmann. (Plate XLIV, fig. 2.)

Hatcheria titcombi Eigenmann, Proc. Am. Philos. Soc, LVI, Jan. 1918, p. 692.
Pygidium areolatim Evermann & Kendall {non Cuvier & Valenciennes), Proc.

U. S. Nat. Mus., XXXI, 1906, p. 86. (Rio Comajo, tributary of Lake Traful,

tributary to Rio Limay.)

Habitat. — Eastern slope of the Andes in Argentina, Limay basin.
11110, I. U. M., one, 164 mm. Arroyo Comajo, J. W. Titcomb.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 285

This specimen is one of those mentioned by Evermann and Kendall in the
paper quoted above. It differs from P. nreolalum as described by Cuvier and
Valenciennes, whose specimen came from Chile, west of the Andes. The origin of
the dorsal is farther back, and its last ray is beyond the last ray of the anal.

Head G.33; depth 6.5; D. 17.5 (3 and 14.5); A. 9.5, counting the minute im-
bedded rays in each case; P. 9; front margin of the eye in the middle of the head;
interocular a little over three times in the length of the head, eye three in the inter-
ocular. Teeth very narrow chisels ; nasal barbel reaching to above first interopercular
spines, maxillary barbel to middle of opercular spines. Pectoral rounded, its first
ray not prolonged, nearly two-thirds the length of the head; origin of the ventrals
equidistant from snout and last fifth of the middle caudal rays; first anal ray under
the sixth dorsal ray, the last anal ray under the fourth from the last ray of the
dorsal; distance between anal and caudal 4.75 times in the length; origin of dorsal
equidistant from tip of caudal and middle of pectorals, its distance from the caudal
two times in its distance from the snout.

Sides without distinct markings; faint traces of longitudinal lines.

This specimen differs from a specimen of H . areolata in the Harvard Museum,
in which the last dorsal ray is over the fourth ray of the anal. In a specimen of
nrrolata in the British Museum di'awn by J. Green, the last dorsal ray is over the
l>enultimate anal ray.

4. Hatcheria areolata (Cuvier & Valenciennes).
Native name "Bagre."

Trichomyclerus areolatus Cuvier & Valenciennes, Hist. Nat. Poiss., XVIII, 1846,

p. 492 (coast of Chile); Guichenot, in Gay, Hist. Chile, II, 184S, p. 309;

GiJNTHER, Cat. Fishes Brit. Mus., V, 1864, p. 274 (Chile); Philippi, Mb. Ak.

Wiss. Berlin, 1-866, p. 714; Delfin, Catalogo de los Peces de Chile, 1901, p. 30.
Pygidium areolatum Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889,

p. 51 (Rio Mapocho, Chile); Occasional Papers Cal. Acad. Sci., I, 1890, p. 330;

Proc. U. S. Nat. Mus., XIV, 1891, p. 36; ? Berg, An. Mus. Nac. Buenos Aires,

IV, 1895, p. 143 (Arroyo del Tala, Catamarca, Argentina).
Hatcheria areolata Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,

1909, p. 251, pi. XXXIV, fig. 2; 1910, p. 399.
Thrichomycterus maculatus Girard, yart; U. S. Naval and Astron. Exped., 1855,

p. 243 (Mapocho).

Habitat. — Pacific slope of Central Chile; ? Catamarca, Argentina.

It is doubftul whether the specimens mentioned by Berg, which had come from

286

MEMOIRS OP THE CARNEGIE MUSEUM.

east of the Andes, belong to H. areolata, the definitely known habitat of which is
the western slope of Central Chile.

Fig. 4. Hatcheria areolala (C. & V.) after Eigenmann. From a specimen in the Mus. Comp.
Zool., 103 mm. Mapocho, Chile.

Head 5.75; depth 8.5; D. 14; A. 8. Elongate, subterete. Lips and lower sur-
faces of the head thickly covered with small warts. Gill-openings continued for-
ward to below the eye, the membranes free from the isthmus. Upper maxillary
barbels reaching to the pectorals. Pectorals rounded, the first ray not prolonged ;
origin of dorsal slightly in front of the vent, equidistant from tip of caudal and occi-
put, its last ray over the fourth ray of the anal. Caudal very slightly emarginate.
Distance of anal from the base of the caudal five times in the length. Origin of the
ventrals equidistant from tip of snout and middle of caudal; tips of the ventrals not
reaching the vent. Light brown, with purple longitudinal streaks.

5. Hatcheria burmeisteri (Berg.)

Pygidium burmeisteri Berg, An. Mus. Nac. de Buenos Aires, IV, 1895, p. 128,

Lam. 2, fig. 1 (Rio Mendoza); Eigenmann, Reports Princeton Univ. Exped.

Patagonia, III, 1910, p. 400.
Trichomyderus burmeisteri Boulenger, Ann. & Mag. Nat. Hist. (7), IX, 1902, p.

336 (Palmira, Rio Mendoza, 900 m.).

Habitat. — Province Mendoza, Argentina, elevation 900 meters.

Known from the type and the specimen recorded by Boulenger.

Reaches a length of at least 260 mm.

\

' iC-;*^""

Fig. 5. Hatcheria burmeisteri Berg. After Berg.

Head 7.5 (9 in total); depth 9 (10); D. 21; A. 7; P. 10; eye in the middle of the
head; nasal barbel reaching to the eye, maxillary barbel scarcely to gill-opening;
head much longer than broad, depressed; interopercular spines numerous; pectoral

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 287

ray scarcely produced, shorter than head; anal inserted under the eighth dorsal ray;
caudal emarginate, the upper lobe slightly produced and pointed, the lower obtuse.
Color uniform.

6. Hatcheria macraei (Girard.)

Thrichomycterus macrcei Girard, U. S. Naval and Astron. Exped., 1855, p. 245

(Uspullata, 7,000 feet).
Pygidium macrcei Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889,

p. 51 (Uspullatuo*) ; Occasional Papers Cal. Acad. Sci., I, 1890, p. 328; Proc.

U. S. Nat. Mus., XIV, 1891, p. 36; Delfin, Catalogo de los Peces de Chile,

1901, p. 29.
Hatcheria macrcei Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,

1909, p. 248, plate XXXII, figs. 1, la and 16; 1910, p. 399.

Habitat. — Eastern slope of the High Andes of central Chile.
7458a-j, C. M., 24-113 mm. San Juan, Argentina, Feb. 25, 1909, Haseman.
7549a-/, C. M., 37-70 mm. Rio Colorado, March 5 and 6, 1909, Haseman.

#^^\

*— -ij-- -fc^ A ■■

Fig. 6. Hatcheria macraei (Girard). After Eigenmann. From No. 8298, Mus. Comp. Zool.,
Uspullata, Chile.

Head 6.5; depth 7; D. 21 or 22, rarely 20 or 23; A. 10. Elongate, rather com-
pressed, especially backward. Head nearly or quite as broad as long, snout rounded ;
eye small, midway between tip of snout and end of opercle; none of the barbels
reaching the gill-opening. Gill-opening scarcely continued forward, joined to the
isthmus for a distance equal to half the width of the mouth. Pectorals obliquely
truncate, the first ray not produced in the type, or slightly produced in the speci-

* ? A misprint for Uspullata ? ■ " Uspullatuo " is not found in gazetteers or on maps. Editor.

288 MEMOIRS OF THE CARNEGIE MUSEUM.

mens collected by Haseman; origin of dorsal some distance behind ventrals, equi-
distant from occiput and tip of caudal in the type or from some portion of the snout
and tip of caudal in the specimens collected by Haseman; fourth or fifth dorsal ray
highest, then gradually decreasing in height to the last. Caudal emarginate, the
upper lobe pointed, the lower rounded; anal inserted about under the ninth dorsal
ray and terminating imder about the seventeenth; ventrals inserted nearer tip of
snout than to tips of middle caudal rays, reaching to the vent or slightly beyond.

Sides and back in the 8an Juan specimens profusely spotted, much less so in
the specimens from the Rio Colorado.

Genus IV. Pygidium'*"' Me yen.

Trichomycterus Valenciennes, in Humboldt, Rec. d'Obs. Zool. et Anat., II, 1833,

p. 348 {nigricans) ; not Thrichomycterus Cuvier and Valenciennes, in Humboldt,

of which it is a misspelling. Gunther, Cat. Fishes Brit. Mus., V, 1864, p.

272.

Thrychomyderus Cuvier & Valenciennes, Hist. Nat. Poiss., XVIII, 1846, p. 485

(missi^elled).
Thrichomycterus non Cuvier & Valenciennes, Girard, Proc. Acad. Nat. Sci. Phila.,
VII, 1854, p. 198; Girard, U. 8. Nav. Astron. Exped., II, 1855, p. 242 (mis-
quoted).
Pygidium Meyen,*' Reise, I, 1835, p. 474 {Juscum).
Type. — Pygidium juscum Meyen.

Skin naked; head depressed, nearly or quite as broad as long, its length
five or six times in the length from snout to caudal; body terete, the caudal
peduncle compressed, deej^; a nasal barbel as long as the head or shorter, on the
posterior edge of the anterior nares; two barbels at the angle of the mouth, the
upper, connected with the rudimentary maxillary, may reach to the tip of the

'° Trvyl&wv. TO = a thill rump = the tail mucli compressed.

"In "Archiv fiir Naturgeschichte von Dr. Ar. Fr. Aug. Wiegniaun, Zweiter Band, Berlin, 1835
(Part. II), p. 269," the original description with addenda appears as follows:

"Eine neue Gattung der Siluriden, Pygidium, hat Meyen (Reise, I, p. 475), nach oincm todten
Fische aufgestellt, den er in einem klcincn Bache Peru's antraf.

"Char. gen. Corpus elongatum, caudam versus compressum. Cirri iiuixillares 4, nasales nulli.
Pinnic pectorales ut pinnaj abdominalcs duw cum j)inna anali circa anum posita;. Pinna adiposa parva.
(Die einzige Art P. Juscum ist 5^6" lang). Die Gattung bcdarf einer gonaucren Charakteristik; die
gegebene ist dahin zu berichtigen, dass cirri nasales vorhanden siud, und die Riickenflosse Strahlen hat,
also keine Fettflosse ist. Die Gattung steht demnach uicht Malapterus, sondern Silurus nahe, unter-
scheidet sich von diesem durch Zahnlosigkeit des Vomer, dureh ein operculum acvleato-serratum, und
durch die weit hinten stehende Riickenflosse. Das Exemplar ist im Berliner Museum."

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 289

pectoral, but is usually shorter; no mental barbels; eye small, in the middle of the
head, or just in front of the middle, without a free orbital rim; interopercle with
numerous spines in several series, those of the outer series largest (in the very young
in a bunch as long as the opercular buncli, in the older in a much larger patch);
opercle with a bunch of similar spines; gill-membranes narrowly united with the
isthmus and usually with a narrow, free margin across it; mouth of moderate width,
terminal, the jaws with two or more series of chisel-shaped or conic teeth ; no teeth
on lips or on the vomer; fins without spines, the pectoral short, the outer, simple
ray usually prolonged into a filament extending distinctly beyond the rays; ventrals
small, placed in the middle or considerably behind the middle of the body; anal
short, usually in part below, more rarely behind, the dorsal; caudal short, broadly
rounded, truncate or slightly emarginate, accessory rays variable, sometimes very
conspicuous, sometimes much less so; origin of the dorsal between the vertical from
the origin of the ventrals and anal, always nearer the base of the caudal than to the
snout; the fin is low, rounded, short, with a variable number of rays up to twelve.
Cuvier & Valenciennes state that the first ray of the dorsal of P. nigricans is pro-
longed in a filament. Is this a lapsus diyiti for first ray of the pectoral? The
dorsal and anal have from two to four minute accessory rays entirely hidden in the
thick skin in front of the evident portion of these fins.

The color may be uniform, or there may be one or three longitudinal stripes
or rows of spots, or large spots less regularly arranged, or numerous small spots
which may be discrete, or which may coalesce into vermiculations. There are no
distinct cross-bars. If the markings are longitudinally arranged, a series of spots
may be replaced by a stripe or vice versa in different individuals of the same species.

Some of the species are of very small size, the maximum recorded size is 350
and 390 mm. in P. rivulatum and P. taczanowskii from Peru.

The eggs reach 2.5 mm. in diameter.

The species differ from each other largely in the shape of the teeth, the length
of the barbels, the relative position of the dorsal, anal, and ventrals, and in the
color.

Distribution. — The members of the genus Pygidium belong particularly to the .
mountains, where they live in all waters from small rills to large lakes like Titicaca.
They are frequently found under rocks or buried in the muddy banks of streams.
They extend from Panama southward to Chile and Patagonia, where they are
replaced by the members of the allied genus Hatcheria. In favorable places they
descend to the sea, as at Jequetepec and Callao, and they are among the last or are
the very last to disappear in ascending the mountains, where they are associated

290 MEMOIRS OF THE CARNEGIE MUSEUM.

with a few other mountain forms Uke Grimdulus at Bogota, Astroblepus and Bi^y-
conamericus in the High Andes from Panama to Cuzco, and Oresteas in Lake Titi-
caca. The only fish found by Haseman in the headwaters of the Rio das Velhas
was a member of this genus. Species of Pygidium were found in the most elevated
places visited by Henn in Colombia and Ecuador. In Titicaca they are of con-
siderable economic importance, and on the plains of Bogota, the nearly related
genus Eremophilus is of prime economic importance. They are found in Guiana
and in the Amazon, but only as dwarfs. They also flourish in the mountain-
streams of southeastern Brazil, but the species do not reach the size of those in
Peru. Some of the species are found on both slopes of the Andes, but, unlike low-
land species of other fishes, which if found on both sides of the Andes, usually have
a very wide distribution, the species of the genus Pygidium all have rather limited
ranges. Many of them are restricted to a single small river and no river has many
species. In 1910 I said (Patagonia Report, p. 248), "There is no place on record
harboring more than one species of this genus." This statement requires modifica-
tion. While, so far as known, many basins contain but a single species, a number
of other smaller rivers, the Iguape for instance, contain several. Judging by its
wide distribution, both horizontally and vertically, the genus is probably one of
very long standing.

The species of the genus need a careful revision, but the descriptions usually
omit mention of the character of the teeth, and no collection contains any great
percentage of the total number of species described. Furthermore, judging from
the fact that they are abundant in all the high mountain-rills and even in lowland
rapids, and that from the stretch from Caracas along the eastern slope of the
Andes to Peru we have only the types of the species P. meridoe, kneri, metce, and
dorsostriatum, the revision of the entire genus may be left in abeyance. The species
are grouped according to the areas from which they have been reported.

CHILEAN SPECIES

The species from C'hile where the members of the genus Hatcheria have in part
■ replaced them, are P. marmoratum. (Philippi), P. palleum (Philippi), and P.
tigrinum (Philippi) . ' ^

1. Pygidium marmoratum (Phihppi).

Trichomycterus marmoratus Philippi, Mb. Ak. Wiss. Berlin, 1866, p. 714; Eigen-
MANN & EiGENMANN, Occasional Papers Cal. Acad. Sci., I, 1890, p. 326;
Delfin, Catalogo de los Peces de Chile, 1901, p. 31.

'* In addition to the three species described by Pliilippi, Pygidium nigricans (Cuvier & Valencien-
nes) is recorded from Chile by Gay. This is probably an error.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 291

Pygidium marmoralum Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., XIV,
1891, p. 36; Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,
1910, p. 399.
Habitat. — Chile.

Blackish gray, marbled with many black spots, as in pundatum; fins dark.
Depth 10.82;D. 10; A. 6.

2. Pygidium palleum (PhiHppi.)

Trichomycterus palleus Philippi, Mb. Ak. Wiss. Berlin, 1866, p. 715; Eigenmann &

Eigenmann, Occasional Papers Cal. Acad. Sci., I, 1891, p. 325; Delfin,

Catalogo de los Feces de Chile, 1901, p. 30.
Pygidium palleum Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., XIV, 1891,

p. 36; Eigenmann, Reports Princeton Univ. Exped. Patagonia, III, 1910, p.

399.

Habitat. — Chile.

Light reddish; fins colorless; head 6.5 in total; D. 9-10; A. 6.

3. Pygidium tigrinum (Philippi).

Trichomycterus tigrinum Philippi, Mb. Ak. Wiss. Berlin, 1866, p. 714; Eigenmann

& Eigenmann, Occasional Papers Cal. Acad. Sci., I, 1890, p. 326; Delfin,

Catalogo de los Peces de Chile, 1901, p. 31.
Pygidium tigrinum Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., XIV, 1891,

p. 36; Eigenmann, Reports Princeton Univ. Exped. Patagonia, III, 1910,

p. 399.

Habitat. — Chile.

Light with reddish points; fins immaculate; head 6.5; depth 6.5 in total length;

D. 9 or 10; A. 6.

Members of the genus Pygidium have been recorded from the mountains of
Argentina, north of the latitude of Buenos Aires. South of this latitude species
of Hatcheria take their place.

Key to the species of Pygidium from Argentina and the Paraguay Basin

a. Teeth pointed (not examined in tenue).

b. Plain yellowish, eyes and barbels black; head triangular; opercle and pre-opercle well armed;

body much compressed; D. 6; A. 5 4. tenue (Weyenbergh).

66. Back spotted, sides with a band.

c. Pectoral ray prolonged; head as wide as long, 4.66-5 in the length; eye very small, a little in
advance of middle of head, its diameter three in the interorbital; maxillary barbel reaching
pectoral; depth of caudal peduncle half its length; origin of dorsal behind the ventrals,
its distance from the caudal two and one-half in its distance from the snout; origin of

292 MEMOIRS OF THE CARNEGIE MUSEUM.

ventrals equidistant from snout and tip of caudal, or a little nearer the latter; caudal
truncate, or slightly cniarginate; olive above, more or less distinctly spotted with brown; a
blackish band from opercle to the caudal; D. S or 9; A. 0 . . . . 5. corduvense (Weyenbergh).
cc. Pectoral ray not prolonged; body and head tuberculate; head 6.5-7.5 in the length (with the
caudal); eye a little in advance of the middle of the head; maxillary barbel bnjad and short;
teeth minute, in many series; 6-S spines in the main row of the interopercle; fins small;
posterior part of dorsal over anal; caudal subtruncate or rounded; D. 3 + 8; A. o + 6.

G. spegazzinii Berg.

bhb. Back with spots; no lateral baud; maxillary barbel reaching origin of the pectoral or farther;

distance between origin of dorsal and caudal 2.5-3 in the distance between dorsal and

snout; origin of ventrals equidistant from snout and tip of caudal.

d. Pectoral ray not prolonged; head 5.33-5.5; eye in middle of the head, 3 in interorbital; origin

of anal under end of dorsal; caudal truncate; D. 10; A. 7. Spots of back large, round.

7. borellii (Houlenger).

dd. Pectoral ray mucii i)rolonged; head six times in the length; eye entirely in anterior half of

the head; origin of anal nearly under origin of dorsal; caudal rounded; back, sides, dorsal

and caudal densely siiotted S. eichorniarum Rilx'iro.

aa. Teeth in part, at least, incisors; heatl as long as broad; barbels short; first pectoral ray prolonged.

e. Head S.5 in the length with caudal; eye in anterior half of head, 1.5 in the interorbital; dorsal
obliquely truncate, its posterior tliird over the anal; caudal truncate; D. 2 + 9; A. 1 +6; faint

spots I), riojanum Berg.

ee. Head six times in the length without the caudal; eye in middle of the head; thirteen spines in the
main row of the interopercle; caudal emarginate; D. 4 + C.5; A. 2 + 5.5; nasal barbel extending
to posterior margin of the eye 10. heterodontum Eigenmann.

4. Pygidium tenue'-' (Weyenbergh). (See fig. 7, p. 293.)
TrichoNiijcterus tenuis Weyenbergh, Act. Acad. Nac. Cienc. Exact., Cordoba, III,

1877, p. 12, pi. Ill (Sierra de Cordoba, near C-ruz del Eje); Eigenmann &

Eigenmann, Occasional Papers Cal. Acad. Sci., I, 1890, p. 326.
Pygidium tenue Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., XIV, 1891,

p. 36; Eigenmann, Reports Princeton Univ. Exped. Patagonia, III, 1910,

p. 399.

Habitat. — Rio Primero, Cordoba.

Yellow, eyes and barbels black; head triangular; opercle and pre-opercle well
armed. Body much compressed; D. 6; A. 5.

5. Pygidium corduvense (Weyenbergh).
Trichomyderus corduvensis Weyenbergh, Act. Acad. Nac. Cienc. Exact. Cordoba,
III, 1877, p. 11, pi. Ill (Rio Primero); Eigenmann & Eigenmann, Occasional

" Berg (An. Mus. Nac. Buenos Aires, IV 1895, p. U4) makes this a synonym of Iliiichcrin arcolnta.
Boulengcr (Boll. Mus. Zool. Anat.Oomp. Univ. Torino, XII, 1S97), contends that Weyenbergh is right
in placing it near P. dispar.

EIGENMANN

: THE PYGIDIIDvE, A FAMILY OF SOUTH AMERICAN CATFISHES. 293

Papers Cal. Acad. Sci., I, 1890, p. 32G; Boulenger, Boll. Mus. Zool. Anat.
Comp. Univ. Torino, XII, 1897 (Caiza).

Fig. 7. Pygidiunl tcmie (Weyeiiljergh). Mier Weyenbcrgh.

Pygidium corduvense Eigenmann & Eigenmann, Proc. U. S. Nat. Mus., XIV, 1891,
p. 36; Eigenmann, Reports Princeton Univ. Exped. Patagonia, III, 1910,
p. 399.

Fig. 8. Pygidium curduvense (Weyenljergh). After Weyenbcrgh.

294 MEMOIRS OF THE CARNEGIE MUSEUM.

Habitat. — Sierra dc C6rdol)a, near Cruz del Eje, Argentina; Caiza, Bolivian
Chaco.

The following is from Boulenger's description of specimens up to 62 mm. long:
" Head 4.66-5; D. 8-9; A. 6; eye three times in interorbital, a little nearer snout
than to opercle; maxillary barbel reaching pectoral; caudal peduncle twice as long
as high; distance between origin of dorsal and caudal two and one-half times in the
distance between dorsal and snout; outer pectoral ray prolonged; origin of ventrals
equidistant from snout and tip of caudal, or a little nearer the latter; caudal trun-
cate, or slightly emarginate. A dark lateral band."

6. Pygidium spegazzinii licrg.
Pygidium spegazzinii Berg, An. Mus. Nac. Buenos Aires, V, 1897, p. 267; Eigen-

mann, Reports Princeton Univ. Exped. Patagonia, III, 1910, p. 399.

Habitat. — Rio de Cachi, Province de Salta, northern Argentina, 2,500-2,800 m.

Known only from the types, 29 specimens, the largest of which is 95 mm., in
the National Museum of Buenos Aires.

Head 6.5-7.5 in the length with the caudal; D. 11 (3 + 8); A. 9 (3 + 6); eye
much nearer snout than to edge of opercle; nasal barbel extending beyond the eye,
maxillary barbel short and broad ; gill-membrane without free margin at the middle ;
teeth in many series; interopercular spines few, in three or four rows, the sixth to
eighth in the lower row moderate in size; opercular spines also few and minute; body
verrucose; pectoral obliquely rounded, its first ray not prolonged; anal inserted
under posterior part of dorsal; caudal subtruncate or rounded.

7. Pygidium borellii (Boulenger).
Trichomycterus borellii Boulenger, Boll. Mus. Zool. Anat. Comp. Univ. Torino,

XII, 1897 (Mission d'Aguairenda; Tala; Lesser); Ann. & Mag. Nat. Hist.

(7), IX, 1902, p. 336 (Palmira, Rio Mendoza).
Pygidium borellii Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,

1910, p. 400.
Pygidium schmidti Berg, An. Mus. Nac. Buenos Aires, V, 1897, p. 266 (Rio de

Belen, Prov. Catamarca, Argentina); Eigenmann, I. c, p. 399.

Habitat. — Mission d'Aguairenda, Bolivian Chaco; Tala and Lesser, Province
Salta, northern Argentina; Rio de Belen, Province Catamarca, northern Argentina;
Palmira.

Reaching a recorded length of 110 mm.

Head 5.35-5.5; D. 10; A. 7; eye very small, in middle of the head, three times
in the interorbital; body compressed, caudal peduncle one and one-half times as

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 295

long as high; maxillary barbel reaching pectoral; origin of anal under end of dorsal;
distance of origin of dorsal from caudal two and one-half to three in its dis- .
tance from the snout; pectoral ray not prolonged; origin of ventrals equidistant
from tips of snout and caudal; caudal truncate; sides and back with large dark
spots.

8. Pygidium eichorniarum (Ribeiro). (Plate XLIV, fig. 3).
Trichomijcterus eichorniarum Riljeiro, Comm. Linhas Telegraphicas Estrategicas
Matto-Grosso ao Amazonas, Annexo, 5, 1912, p. 27 (Caceres).
Habitat. — Upper Paraguay.

Evidently allied to P. riojanum, proops, and met(£.
Known from the types, two specimens, the larger 44 mm., and
7556a-c, C. M., 24-30 mm. Caceres, May 27, 1909. Haseman.
7557o & b, C. M., 33-43 mm. Caceres, May 23, 1909. Haseman.
7558a, C. M., 42 mm. San Francisco, Rio Jauru, Paraguay basin, June 10, 1909.

Haseman.
7559a, C. M., 32 mm. Bastos, Rio Alegre, eight miles south of Villa de Matto-
Grosso, June 26, 1909. Haseman.
7560a-c, C. M., 39-41 mm. San Antonio, Rio Guapore, plantation of Maciel,
July 31-Aug. 11, 1909. Haseman.

Head 5-5.75; D. 9-10; A. 8; P. 6; posterior margin of eye slightly in advance
of the middle of the head; eye about 1.5-2 in the snout, 5.5-6.5 in the head, about
equal to the interorbital; maxillary barbel reaching to axil or middle of pectoral;
nasal barbel to the tip of the opercular spines or the axil of the pectoral; teeth
conical, a very narrow band of but two or three irregular series; origin of ventrals
equidistant from tip of snout and tip of caudal; origin of anal under, or but sHghtly
behind, the first dorsal ray; distance from base of last anal ray to base of caudal
about six times in the length; distance from origin of dorsal to base of caudal two
and three-quarters in its distance from the snout; caudal rounded, accessory rays
moderate; first pectoral ray much prolonged, with its filament nearly equal to the
length of the head.

General color of P. brasiliense, back and sides profusely spotted; caudal rays
with numerous spots, dorsal and anal less profusely spotted.

9. Pygidium riojanum Berg.
Pygidium riojanum Berg, Ann. Mus. Nac. Buenos Aires, V, 1897, p. 269; Eigen-
MANN, Reports Princeton Univ. Exped. Patagonia, HI, 1910, p. 399.
Habitat.— Arroyo in the Cordillera de la Rioja, northern Argentina.

296 MEMOIRS OF THE CARNEGIE MUSEUM.

Known from the type, a specimen 85 mm. long, in the National Museum at
Buenos Aires.

Head 8.5 in the length with the caudal; D. 9 (2 + 7); A. 7 (1 + 6); eye 1.5 in
the snout, 1.5 in interorbital, 2.5 in posterior part of the head; nasal barbel scarcely
extending beyond the eye; maxillary barbel scarcely to end of opercle; gill-membrane
with scarcely a free margin ; intero]Dercular spines in two or three series, medium in
size; teeth small, in irregular series, anterior ones larger, their tips broader; fins
small; first pectoral ray prolonged; anal under last third of dorsal; caudal truncate.

10. Pygidium heterodontum Eigenmann. (Plate XLIV, fig. 4.)
Pygidium heterodontum Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p.

692.
13832, 1. U. M., 83 mm., 9, Rio Mcndoza, Palmira, Argentina, 900 m. Purchased

from Rosenberg.

Palmira is probably the southernmost locality on the eastern slope of the
Andes from which species of this genus have been taken.

Head six times in length, as long as broad; D. 10.5 (4 + 6.5); A. 7.5 (2 + 5.5);
P. 9; eye in middle of the head, intcrocular 3.5 in the head; teeth in three series in
each jaw, those of the outer row narrow incisors, of the second row much smaller
incisors and of the third row conic. Head much depressed, interopercular spines
numerous, thirteen in the last row.

Nasal barbel extending to the posterior margin of the eye, maxillary barbel to
the base of the ojiercular spines; first pectoral ray scarcely produced, equal to the
portion of the head behind the posterior nares; origin of ventrals midway between
opercle and caudal, reaching to the vent; origin of anal under posterior part of the
dorsal, the distance between its last ray and the base of the middle caudal ray 4.4
in the length; depth of the caudal ])eduncle 2.5 in its length; caudal narrow, emargi-
nate, a httle more than five in the length; origin of dorsal midway between the tip
of the caudal and the occiput, over the tip of the ventrals, its distance from the
caudal 1.75 in its distance from the snout.

A faint lateral band and obscure spots or marblings.

The members of the genus Pygidium reach their largest size and greatest
economic importance in Peru. The C'arnegie Museum has no specimens from this
region, except P. oroyce Eigenmann & Eigenmann. To the key below should be
added P. fuscum Meyen, the type of the genus.

Key to the Species of Pygidium from Peru and Western Bolivia.
a. Pfictoral ray prolonged.

h. Dorsal entirely in front fif tlic ;inal.
c. Caudal truncate or ronniloil.

EIGENMANN: the PYGIDIIDiE, A FAMILY OF SOUTH AMERICAN CATFISHES. 297

d. Uniform brown, darkest on the back; head flat above; width of head less than its length;
barbels scarcely extending beyond the eyes, which are in the middle of the head; a
broad band of villiform teeth in each jaw; first pectoral ray but slightly prolonged;
origin of dorsal equidistant from tip of caudal and nares, over posterior edge of base of
ventrals; last ray of the dorsal over the origin of the anal; caudal rounded, distance
between anal and caudal 4.5 in the length; ventrals nearer tip of snout than to tip of

caudal; head 5; depth 5.66; D. 10; A. 9 12. eigenmanni (Boulenger).

ilil. Head and body with dark spots; a dark lateral stripe; head as broad as long; barbels
equal to eight-tenths the length of the head; snout slightly shorter than the postor-
bital part of the head; outer pectoral ray as long as the head, longest branched ray
three-quarters as long; origin of dorsal in advance of the vent, its distance from the
base of the caudal one and one-half times in its distance from the snout; origin of anal
slightly behind the last dorsal ray; caudal truncate; distance between anal and
caudal 4.5 in the length; head 6.25; D. with six, A. with four branched rays.

13. vittatum (Regan).
cc. Caudal emarginate.

e. Back and sides profuselj- spotted; head longer than broad; l)arl)els not quite reaching gill-
openings; origin of ventrals equidistant between tip of snout and tip of caudal;
head 5.2-5.66; D, 12; A. 9 or 10.
/. Sides, back, dorsal, and caudal with large spots; spots as large as, or larger than, the-
eye, smallest on the head; origin of dorsal equidistant from tip of caudal and
anterior margin of eye; distance between anal and caudal five or six times in the

length 14. dispar Tschudi.

//. Spots much i-nialler than the eye; origin of dorsal cquiwistant from tip of caudal and
a point between occiput and anterior margin of the eye; distance between anal

and caudal 6-6.5 in the length 15. punctulatum (Cuvier & Valenciennes).

ee. Back and sides unspotted; maxillary barbel reaching past origin of pectoral; origin of
dor.sal varying with age; origin of ventrals a little nearer snout than to tip of cauilal;
distance between anal and caudal five times in the length; head 4.66-5.5; eye minute,
in adult a little behind the middle of the head; teeth conic.™

16. taczanowskii (Steindachner).
hh. Dorsal in part over the anal.

g. Accessory caudal rays conspicuous; caudal rounded; outer row of teeth narrow incisors;
maxillary barbel reaching edge of pre-opercle; origin of dorsal equidistant from tip of
caudal and a point between occiput and posterior nares; distance between caudal and
anal 4.4-4.5 in the length; head 4.-5-5.5; D. 13; A. 11.

17. rivulatum (Cuvier & Valenciennes).

gg. As under g, but "differing in its large, dark blotches." IS. poeyanum (Cope).

ggg. Accessory caudal rays not evident; caudal emarginate; tsct'.i c ji.ic; head 4.66; depth 7;
D. 8; A. 6; eye in middle of the head; head longer than wide; nasal barbels reaching
posterior margin of the eye; maxillary barliel to the gill-opening; distance between dorsal
and caudal about 2 in its distance from the snout; distance between anal and caudal 5.5.

19. barbouri Eigenmann.
gggg. Accessory caudal rays not conspicuous, the caudal truncate; teeth conic; head 4.85; depth

'"' The male of di><par as figured by Tschudi agrees with this.

298 MEMOIRS OF THE CARNEGIE MUSEUM.

6; D. 9; A. 7; eye in micklle of the head; head a little longer than wide; nasal barbels
reaching lateral end of head, maxillary barbel a little beyond origin of pectoral.

20. fassli Steindachner.
aa. Pectoral ray not prolonged; end of dorsal about over the middle of the anal; caudal rounded; head as
long as wide; sides and back with irregular spots.
h. Eye moderate; origin of the dorsal over or in front of the vent, equidistant from eye and tip of

caudal or nearer the latter; head 5.75-6; D. 12; A. 10 21. oroyae Eigenmann.

hh. Eyes very minute; origin of the dorsal in front of the vent, nearer the eye than the tip of the
caudal; head 5; D. 8; A. 6 or 7 not counting the liidden rays; a dark lateral line.

22. quechuorum Steindachner.

11. Pygidium fuscum Meyen.
Pygidium fuscum Meyen, Reise, I, 1835, p. 475; Wicgmann's Arch., 1835, II, p.
269; Eigenmann & Eigenmann, Occasional Papers Cal. Acad. Sci., I, 1890,
p. 325; Proc. U. S. Nat. Mus., XIV, 1891, p. 36; Eigenmann, Reports Prince-
ton Univ. Exped. Patagonia, III, 1910, p. 399.
Habitat. — Peru.

Very little is known about this species. It was imperfectly described by Meyen.
Fortunately the type which was found dead in some stream in Peru, is in the Berlin
Museum {fide Tschudi, Fauna Peruana, Ichthyologie, 1845, p. 21). Tschudi
tells us that fuscum is specifically distinct from his own species dispar. This and
the original description is all we know about the species.

12. Pygidium eigenmanni (Boulenger).
Pygidium knerii Eigenmann & Eigenmann (won Steindachner), Occasional Papers

Cal. Acad. Sci., I, 1890, p. 335 (Cumbaca).
Trichomycterus eigenmanni Boulenger, Boll. Mus. Zool. Anat. Comp. Univ.

Torino, XIII, Dec. 2, 1898, substituted for P. knerii Eigenmann & Eigenmann,

non Steindachner.
Pygidium eigenmanni Eigenmann, Reports Princeton Univ. Exped. Patagonia,

III, 1910, p. 400.

Habitat. — Cumbaca. Location on map not known.

Boulenger based his eigenmanni on the description of P. knerii Eigenmann &
Eigenmann, which, according to Boulenger, was based on a specimen distinct from
knerii. The species is known from the description of a specimen 1 10 mm. long from
Cumbaca, collected by the Thayer Expedition and now at Cambridge, Mass.

Head 5; depth 5.66; D. 10; A. 9. Elongate, compressed; head greatly de-
pressed, flat above, the eyes entirely superior; width of the head less than its length.
Barbels scarcely extending beyond the eyes, which arc equidistant from tip of snout
and end of opercle. A broad band of villiform teeth in each jaw. Pectoral rounded,

EIGENMANN: the PYGIDIIDiE, A FAMILY OF SOUTH AMERICAN CATFISHES. 299

the first ray slightly prolonged. Origin of dorsal above posterior edge of base of
ventrals, equidistant from tip of caudal and nares, the last ray over origin of anal.
Caudal rounded, its distance from the anal 4.5 in the length. Ventrals nearer tip
of snout than tip of caudal. Uniform brown, darkest on the back.

13. Pygidium vittatum (Regan).
Trichomijcterus vittatus Regan, Ann. & Mag. Nat. Hist. (7), XII, 1903, p. 623 (Col-
lected by Ockenden).
Pygidium vittatum. Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,
1910, p. 400.

Habitat. — Marcapata Valley, eastern Peru.
Known from the types, 78 mm. long, in the British Museum.
Head 6.25; D. 6 (branched) ; A. 4 (branched) ; head as broad as long; diameter
of eye 2.33 times in the interocular width, which is 3.5 in the length of the head.
Snout shghtly shorter than the postorbital part of head. Barbels equal to eight-
tenths the length of head. Dorsal originating in advance of the anal opening, the
the distance from its point of origin to the caudal one and one-half times in the
distance from the former to the tip of the snout. Anal originating slightly behind
the vertical from the last dorsal ray, the distance from the base of its last ray to
the caudal four and one-half times in the total length. Longest branched ray of
pectoral three-fourths the length of the simple outer ray, which is as long as the
head. Ventrals extending six-tenths of the distance from their base to the origin
of the anal. Caudal truncate. Head and body with dark spots; a dark longi-
tudinal stripe along the middle of the side.

14. Pygidium dispar Tschudi. (Plate XLV, fig. 5.)
Pygidium dispar Tschudi {partim), Faun. Peruana, Ichthyol., 1845, p. 22, pi. 3,
upper figure. (Eastern slope of the Peruvian Andes at an altitude of 14,000
■ ft.); Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889, p. 52
(Callao); Occasional Papers Cal. Acad. Sci., I, 1890, p. 335 (Callao); Proc.
U. S. Nat. Mus., XIV, 1891, p. 36; Pellegrin, Bull. Soc. Zool., Paris,
XXIX, 1904, p. 91; Starks, Proc. U. S. Nat. Mus., XXX, 1906, p. 770 (Eteri,
Peru) ; Poissons des Lacs des Haut Plateaux de I'Amer. Sud, 1907, p. 17 (Lake
Titicaca); Eigenmann, Reports Princeton Univ. Exped. Patagonia, III, 1910,
p. 400.

Habitat. — High Andes of eastern and western Peru, down to Callao and Eteri.

The P. dispar recorded by Ribeiro from the Rio Iporanga of southeastern

Brazil is a different species. Tschudi says the species is abundant in the highland

300 MEMOIRS OF THE CARNEGIE MUSEUM.

between the two chains of the Cordilleras, but on the eastern slope only. It is
quite possible that Tschudi had two species and that the unspotted male he figured
is the taczanowskii of Steindachner. The specimens recorded by Pellegrin are
probably P. rivulalum.

Head 5.2; D. 12; A. 9. Elongate, compressed, the depth everywhere less than
the length of the head. Head longer than wide by more than a diameter of the eye.
Eye moderate, four times in the interocular, equidistant from tip of snout and end of
opercle. None of the barbels reaching quite to the gill-opening. Gill-openings
continued forward to below the eye. Pectorals obliquely rounded, the first ray
produced in a filament. Origin of dorsal equidistant from tip of caudal and anterior
margin of the eye, the whole fin in front of the anal and behind the ventral fins.
Caudal emarginate. Distance of anal from base of caudal six times in the length.
Origin of ventrals midway between tip of snout and tip of caudal.

Reddish brown; sides, back, dorsal and caudal fins with large dark spots,
those on the head smallest; lower surface plain.

15. Pygidium punctulatum (Cuvier & Valenciennes). (Plate XLV, fig. 4.)
Trichomycterus pnncfulatus Cuvier & Valenciennes, Hist. Nat. Poiss., XVIII,

1846, p. 488 (Lima); Lutken, Velhas Flodens Fiske, 1875, p. 137 (Callao).
Pygidium dispar punctulatum Eigenmann & Eigenmann, Proc. Cal. Acad. Sci.

(2), II, 1889, p. 52 (Rio Remac, near Lima); Occasional Papers Cal. Acad.

Sci., I, 1890, p. 336 (Rio Remac, largest 180 mm.); Proc. U. S. Nat. Mus.,

XIV, 1890, p. 36.
Pygidium punctulatum Starks, Proc. U. S. Nat." Mus., XXX, 1906, p. 771 (Callao);

Eigenmann, Reports Princeton Univ. Exped. Patagonia, III, 1910, p. 400.
TrichoynyctcruH punctatus Cuvier & Valenciennes, I. c, pi. 552.

Habitat. — Rio Remac, Peru.
4234, I. U. M., one, 145 mm., a female with emi)ty ovary. Callao.

From the Harvard collections.

Head 5.33-5.66; D. 12; A. 10.

Teeth conic, in about five series in the middle of the jaws.

Origin of dorsal equidistant from tip of caudal and somewhere between occiput
and anterior margin of eye; distance of anal from base of caudal 6-6.5 in the length.

16. Pygidium taczanowskii (Steindachner). (Plate XLVI, figs. 5-8.)
fPygidium dispar Tschudi (in part), Fauna Peruana Ichthj'ol., p. 22, pi. 3 (lower

figure).
Trichomycterus taczanowskii Steindachner, Flussf. Siidam., IV, 1882, p. 22, pi.

IV, figs. 1-16 (Rio de Huambo; Rio de Tortora).

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 301

Pygidiutn taczanowskii Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2),

II, 1889, p. 52; Occasional Papers Cal. Acad. 8ci., I, 1890, p. 338; Proc. U. S.
Nat. Mus., XIV, p. 37; Eigenmann, Rci)orts Princeton Univ. Exped. Patagonia,

III, 1910, p. 400.

Habitat. — North and central Peru, between the Andes.

Head in specimens 110-113 mm. 5 5.5; in a specimen 390 mm. 4.75; D. 9-10;
A. 7; P. 9; width of head 1.2 in its length, snout 2-2.33; interorbital 3-3.33, nasal
barbels 1-1.25 in the head, in small specimens, 1.4 in larger specimens, maxillary
barbels 1.21-1.25, lower barbels 1.6-2; width of mouth two times in the length of
the head ; anterior margin of eye slightly in front of the middle of the head in smaller
individuals, in the middle in the larger; teeth brush-like; opercular and interoper-
cular spines numerous, in several series, mostly concealed; origin of dorsal variable,
moving back with age, in a specimen 110 mm. its origin much nearer gill-opening
than to caudal; in specimens 390 mm. 1.22 nearer middle caudal rays than tip of
opercle; origin of ventrals in specimens up to 210 mm. almost directlj^ under the
origin of the dorsal, in a specimen 390 mm. half the length of the head farther
forward ; the origin of the anal in front of the end of the dorsal in small specimens,
under it in the specimen 390 mm. long; first pectoral ray prolonged, its length 1-1.37
in the length of the head. Without spots, bands, or stripes.

17. Pygidium rivulatum (Cuvier & Valenciennes). (Plate XLV, figs. 2, 3.)

Trichomyderus rivulatus Cuvier & Valenciennes, Hist. Nat. Poiss., XVIII,
1846, p. 495 (Guasacona); GUnther, Cat. Fishes Brit. Mus., V, 1864; Cope,
Proc. Amer. Philos. Soc, XVII, 1877, p. 46 (Lake Titicaca); Pellegrin,
Bull. Soc. Zool. Paris, XXIX, 1904, p. 91; PoissoNS des Lacs des Hants
Plateaux, de I'Am. 8ud., 1907, p. 17 (Rio de Pazha, lac Poopo).

Pygidium rivulatum Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889,
p. 51 (Cuzco; Moho and Puno, Lake Titicaca); Occasional Papers Cal. Acad.
Sci., I, 1890, p. 330, Proc. U. S. N. Mus., XIV, 1891, p. 36; Starks, Proc. U. S.
N. Mus., XXX, 1907, p. 771 (Lake Titicaca); Eigenmann, Reports Princeton
Univ. Exped. Patagonia, III, 1910, p. 399.

"^Trichomyderus incce Cuvier & Valenciennes, I. c, 496 (Rio Guatanai at Cuzco).

Trichomyderus gracilis Cuvier & Valenciennes, I. c, 497 (Rio Azangaro near
Guasacona; Rio Guatanai near Cuzco; Rio' Pontezualo near Coroico; Lake
Compucila near Cuzco); C'ope, Proc. Amer. Philos. Soc. XVII, 681, 1877
(Tinta).

Trichomyderus barbatula Cuvier & Valenciennes, I. c, 498 (Guasacona; Rio
Pontezualo near Coroico).

302 MEMOIRS OF THE CARNEGIE MUSEUM.

Trichomycterus pentlandi Castelnau, Anim. Noiiv. Am. Sud.,49, pi. XXIV, fig. 1,

1855 (Lake communicating with the Ucayale).
Trichomyctenis pictus Castelnau, Anim. Nouv. Am. 8ud., 59, pi. XXIV, fig. 2,

1855 (Lake Titicaca).
Trichomycterus dispar Gijnther (partim), Cat. Fishes Brit. Mus., V, 273, 1864
(Lake Titicaca; Rio de Pontezualo; Andes de la Paz; Guasacona; Rio de Azan-
garo); Garman, Bull. Mus. Comp. Zool. Ill, 275, 1875 (Lake Titicaca).
?Pellegrin, Bull. 8oc. Zool. Paris, XXIX, 1904, p. 91; Poissons des Lacs des
Hauts Plateaux de I'Am. des Sud., 1907, p. 7 (Lake Titicaca).
Trichomycterus pardus Cope, Proc. Acad. Nat. Sci., Phila., 1874, p. 132.

Habitat. — High Andes of Peru, about Cuzco, Titicaca, Jequetepeque, etc.
13833, I. U. M., one, 93 mm. Tirapata, eastern Peru. 13000 ft. From Rosen-
berg.
13750, I. U. M., one, 145 mm. Ollantaytambo. E. Heller.

Teeth In about five series in the middle of the jaws, those of the outer series
narrow incisors, those of the innermost row conic.

Head 4.5-5.5; depth 3.75-6.5; D. 13; A. 11. Tail compressed, head depressed,
about as wide as long; eye equidistant from tip of snout and end of opercle. Nasal
barbels reaching to the posterior margin of the eye, longer in the young. Upper
maxillary barbel about to edge of pre-opercle. Mouth wide, more than one-third
the length of the head. Pectoral rounded, the first ray prolonged in a short fila-
ment, except in the very young. Origin of dorsal equidistant from tip of caudal
and a point between occiput and posterior nares, its posterior portion always over
the anterior half of the anal. Accessory rays of the caudal very numerous, their
division from the true caudal rays marked. Caudal always rounded, its distance
from the anal 4.5-4.4 in the length. Color of largest specimens dark reddish brown,
sides with fine white or silvery spots and vermiculations. Specimens from 100-
200 mm. greatly variable, grayish or dark brown, with darker markings; sometimes
the ground color predominating, sometimes only forming reticulations between the
dark markings; young with an interrupted dark band along the sides.

18. Pygidium poeyanum (Cope).

Trichomycterus rivulatus Cope {non Cuvier & Valenciennes), Proc. Acad. Nat. Sci.

Phila., 1874, p. 132 (Arequipa).
Trichomycterus poeyanus Cope, Proc. Am. Philos. Soc, 1877, p. 47.
Pygidium poeyanum Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889,

p. 50; Occasional Papers Cal. Acad. Sci., I, 1890, p. 326; Proc. U. S. Nat. Mus.,

EIGENMANN: the PYGIDIIDiE, A FAMILY OF SOUTH AMERICAN CATFISHES. 303

XIV, 1891, p. 36; Eigenmann, Reports Princeton Univ. Exped. Patagonia,
III, 1910, p. 399; Fowler, Proc. Acad. Nat. Sci. Phila., 1915, p. 229 (note on
tyiJe).

Habitat. — Arequipa.

This species named by Cope without any sort of description is said by Fowler
to be "close to rivulatum, differing in its large dark blotches."

19. Pygidium barbouri Eigenmann.
Pygidium barhouri Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,

1910, p. 400; Ann. Carnegie Mus., VII, 1911, p. 214, pi. XXXII.

Habitat. — Beni River, eastern Bolivia.
12566 I. U. M., 37 mm. and 2465a-6, C. M., two, cotypes. Rio Beni, tropical

eastern Bolivia. Obtained by Mr. Thomas Barbour at La Paz, Bolivia, from

the Beni River.

This species has conical teeth.

Fig. 9. Pygidium barbouri Eigenmann. After Eigenmann, Annals Carnegie Museum, VII,
pL XXXII.

Head 4.66; depth 7; D. 8; A. 6; eye 3 in snout, 7 in head, 2.5 in space between
the eyes. Width of head equals its length behind the posterior nares, the body
tapering to the caudal; nasal barbels reaching to posterior margin of the eye, the
longer maxillary barbel scarcely to the gill-opening when laid straight back. Teeth
minute, in bands. First pectoral ray prolonged, not as long as the head; dorsal
subtruncate, none of its rays prolonged; distance of origin of dorsal from caudal 2.6
in the length; origin of anal from base of middle caudal rays 3.75 in the length;
caudal emarginate; accessory rays not evident; ventrals not reaching the short,
scarcely rounded anal. A dark median band from the gill-opening to the tip of the
middle caudal rays, a light stripe above it; the back chocolate.

20. Pygidium fassli Steindachner.
Pxjgidium fassli Steindachner, Anz. K. Akad. Wiss. Wien, 1915, No. XVII, p. 200
(Rio Songo, Province North Yungas, Bolivia).

304 MEMOIRS OF THE CARNEGIE MUSEUM.

Head 8.77; depth 6; D. 9; A. 7; P. 9; teeth pointed, in several irregular scries;
head a little longer than broad; eye in middle of head; nasal barbels reaching about
to lateral end of head, maxillary barbels a little beyond origin of pectoral. First
pectoral ray moderately elongate; caudal truncate; origin of anal behind the vertical
from the middle of the dorsa,!, half a head nearer to base of middle caudal rays than
to the lateral margin of the head; snout rounded. Body velvety with minute
tubercles. Upper part of head, back, and sides light chocolate with darker spots in
tolerably regular rows; a dark lateral band, the spots above it larger than the others,
those of the uppermost rows sometimes confluent. Fins unspotted.

21. Pygidium oroyae Eigenmann & Eigenraann.

Pygidium oroyw Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889,
p. 51; Occasional Papers Cal. Acad. 8ci., I, 1890, p. 334 (Pochachara, Oroyo
River); Proc. U. S. Nat. Mus., XIV, 1891, p. 36; Eigenmann, Reports Prince-
ton Univ. Exped. Patagonia, III, 1910, p. 399; Evermann & Radcliffe,
Bull. 95, U. S. Nat. Mus., 1917, p. 35, pi. IV, fig. 2 (Oroyo).
Habitat. — Rios Oroyo and Perene, central Peru.

5792, C. M., eighteen, 24-90 mm. Spring sui)i)lying water to Tarma, Peru.
Lola Vance.

Fig. 10. Pijyidlaiii oroyiu Kigeiunaun et Eigeunuiiiii. After Evermann & Radcliffe.

Head 5.75-6; depth 5.75-8; D. 12; A. 10-11; P. 10. Head about as long as
wide; none of the barbels reaching the gill-opening; teeth all pointed; gill-membrane
narrowly joined to the isthmus, with a narrow, free margin. Pectoral shorter than
head, fan-shaped, the first ray not prolonged. Dorsal inserted over or slightly
behind the vent, its last ray over or behind the middle of the anal, its origin equi-
distant from anterior margin of eye or occiput and tip of caudal, its distance from
the base of the caudal 1.5 in its distance from the snout. Caudal broadly rounded,
its distance from the anal 4-4.75 in the length. Ventrals extending to or beyond
the anus, their origin about midway between tip of snout and tip of caudal or nearer
the former. Chocolate brown; sides, back and unpaired fins with irregular groups
of dark points; traces of a dar"k lateral line in the young.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 305

22. Pygidium quechuorum Stcindachiicr.
Pygidium quechuorum Steindachner, Anz. Ak. Wiss. Wien, 1900, p. 207 (Arequipa)
Denksch. Ak. Wiss. Wicn, LXXII, 1902, p. 137 (49 of separate), pi. IV, fig.
3-3a (Rio Chile, Arequipa); Eigenmann, Reports Princeton Univ. Exped.
Patagonia, III, 1910, p. 400.
Habitat. — Rio Chile at Arequipa, Peru.

Fig. 11. P!jgidiu7n quechuorum Stchidachimr. After Stciiulucluier.

Known from the types, five specimens, 51-64 mm. long, in the Vienna Museum.

Head about 5; depth 5.66-5.75 in total length; eyes very minute; interorbital 3
in the head, snout 3; width of head equals length of head or only very little less;
first pectoral ray not prolonged, the fin 1.5 in the head; maxillary barbel to some
part of the interorbital spines; origin of ventrals almost exactly in the middle;
origin of dorsal about equidistant from base of caudal and pectoral; origin of anal
behind the dorsal; caudal rounded or subtruncate; upper parts marbled; a narrow,
diffuse lateral band, which is sometimes faintly interrupted.

Key to the Species of Pygidium from Venezuela, Colombia, Panama, and Ecuador.

a. Teetlj incisors, apparently conic in the very young.

6. Last dorsal ray over the last anal ray; distance between last anal ray and base of niidiUe caudal
ray about 2.28 in its distance from the snout; distance between last anal ray and base of middle
caudal rays five and one-half in the length; obscure spots about as large as the eye evenly dis-
tributed over back and sides, no lateral baud 23. laticeps (Kner)

66. Last dorsal ray over the anal.

c. Sides with a lateral band or nearly confluent series of spots.

d. Maxillary barbel extending beyond base of last pectoral ray; origin of ventrals equidis-
tant between caudal and some part of the opercular spines; distance between anal
and caudal four and one-half to four and three-fourths in the length; distance between
origin of dorsal and base of caudal about two in its distance from the snout; a narrow

lateral band, a variable number of small spots 24. stellatum Eigenmann.

dd. Maxillary Ixirbel extending little beyond origin of the pectoral or shorter; a dark band or
series of spots below and another above the median band,
c. Origin of ventrals equidistant from caudal and middle of pectoral; distance between
anal and caudal five in the length; distance between origin of dorsal and caudal
a little more than two in its distance from the snout . . . .25. chapmani Eigenmann.
ee. Origin of ventrals nearer head than caudal; distance between anal and caudal five
to five and one-half in the length; distance between origin of dorsal and caudal
one and four- fifths to two in its distance from the snout. .26. taenium (Kner).

306 MEMOIRS OF THE CARNEGIE MUSEUM.

cc. Sides with numerous spots, those ah.)ng the middle line not forming a Ijand; origin of ven-
trals equidistant from caudal and base or middle of the pectorals; distance between the
anal and caudal four and three-tenths to five in the length; distance between origin of the
dorsal and the caudaj one and three-fourths to one and four-fifths in its distance from

the snout 27. caliense Eigenmann.

ccc. Sides plain; maxillary barbel extending beyond the axil; oiigin of ventrals equidistant from
caudal and pro-opercle; distance between anal and caudal five and one-half in the length;
distance between origin of dorsal and caudal two in its distance from the snout; origin of
anal a little in advance of the middle of the dorsal. See also kneri.

28. latidens Eigenmann.
fflo. Teeth sharp-pointed, conical or recurved conical."'

/. Origin of ventrals nearer to caudal than to tip of pectoral filament, the distance between the ven-
trals and caudal two in the distance from the snout; distance between origin of dorsal and
caudal two and two-fifths in its distance from the snout. Sides densely covered with small

spots with only vermiculating light lines between them 29. metae Eigenmann.

//. Origin of ventrals nearer to tip of pectoral filament than to caudal, usually much further forward.
g. Sides plain.

h. Origin of dorsal nearly over origin of the ventrals, nearer the eye than the tip of the
caudal; distance between origin of dorsal and caudal 1.5 or less in its distance from

the snout, head 4. .5-5. .33 in the length 30. stramineum Eigenmann.

hh. Origin of dorsal above or a little in advance of the vent, its distance from the caudal
1.8 in its distance from the snout; head six times in the length; caudal subtruncate;

barbels as long as head 31. unicolor Regan.

hhh. As under /;, caudal emarginate, barbels reaching considerably beyond origin of pec-
toral 32. kneri (Steindachner).

hhlih. Origin of dorsal distinctly behind the origin of the ventrals, much nearer the tip of the

caudal than the eye striatum, No. 41, which see.

gg. Sides irregularly spotted, more rarely a lateral band, the spots along the middle of the sides
rarely in a distinct series; caudal rounded.
i. Maxillary barbels very slender, reaching to the middle of the pectoral rays; eye entirely
in the anterior half of the head; origin of dorsal equidistant from tip of caudal and

opercle; sides and back with moderate-sized dark spots 33. meridae Regan.

ii. Maxillary barbels not reaching to the middle of the pectoral; eye in middle of the head.

j. Origin of dorsal, on an average, slightly nearer to the caudal than to the eye; dorsal

inclusive of the rudimentary rays most frequently 12.5; head a little longer than

wide; distance between dorsal and caudal 1.5-1.7 in its distance from the snout;

sides and back with numerous large spots, rarely in rows, the spots largest in the

larger specimens 34. bogotense Eigenmann.

jj. Origin of dorsal nearly equidistant from tip of caudal and snout; D. 11.5 or 12.5;
head as wide as long; distance between dorsal and caudal 1.4-1.52 in its distance
from the snout; sides and back in the largest with numerous irregularly arranged
spots about the size of the eye, the spots larger and less numerous in the young.
Sometimes nearly plain, sometimes with a lateral stripe.

35. nigromaculatum (Boulenger).
ggg. Sides with distinct longitudinal bands or rows of spots; caudal emarginate, truncate, or
truncate-rounded.
^' Not examined in unicolor, kneri, and retropinne.

EIGENMANN: the PYGIDIID.E, a family of south AMERICAN CATFISHES. 307

k. Caudal emargiiuite; (irigiu of ilnrsal cciuidistaut from tij} of caudal and opercle or pre-
oporcle, its distance from the caudal 1.5-I.S in its distance from the snout; head five in
the length; sides and hack with numerous dark spots, those along the middle of the
sides forming a distinct row, sometimes confluent along the anterior half of the body.

36. banneaui Eigonmann.
kk. Caudal truncate or rounded.

/. Dorsal, anal, and caudal truncate; origin of dorsal equidistant from tip of caudal and
pre-opercle, its distance from the base of the middle caudal rays l.G in its distance
from the snout; sides with a faint l)road l)and, oversown like the back with spots

about the size of the eye 37. spilosoma Regan.

//. Dorsal and anal rounded.

m. Lateral band above the middle; maxillary l)arl)els exteniliug to the axil.

;(. Caudal triincate-rounded; origin of dorsal equidistant from tip of caudal
and eye or nasal barbel, its distance from the caudal about 1.4 in its dis-
tance from the snout; the lateral band or row of spots above the middle,
from the ui)per part of the gill-o|)ening to above the middle of the caudal.

oS. dorsostriatum Eigenmann.

nil. Harbels very short, about reaching thi' eye; origin of dorsal e(iuidistant

from tip of caudal and opercle, a faint lateral band, sides reticulated,

first pectoral ray not prolonged 39. venulosum Steindachner.

mm. Lateral band, if present, in the middle of the sides.

0. Origin of dorsal equidistant from tip of caudal and nasal barbel, its
distance from the base of the middle caudal rays about 1.4 in its
distance from the snout; a lateral band increasing in width to the
caudal; middorsal area dark, a dark stripe between the lateral stripe
and the dorsal stripe in front of the dorsal.

40. latistriatum Eigenmann.
im. Origin of dorsal equidistant from tip of caudal and a point between the
middle of the pectoral and the pre-opercle, its distance from the
mitldlc caudal rays l.S-2 in its distance from the snout.
/;. Maxillary barbel reaching a little beyond the axil or shorter; color
very variable, plain, or with one to three lateral stripes; origin of
the dorsal typically equidistant from tip of caudal and midiUe of

pectoral rays 41. striatum Meek & Hildebrand.

/*/(. Maxillary barbel reaching to near the end of the lower pectoral ray,
longer than the head; origin of dorsal equidistant from tip of caudal

and opercular spines 42. regani Eigenmann.

000. Distance between origin of dorsal and caudal 2.2-2.4 in its distance from
the snout; head as broad as long; eye in the middle of the head; distance
between base of last anal ray and the caudal 5.4 in the length. An
indistinct darker stripe along the midille of the side and traces of some
darker spots 43. retropinne (Regan).

23. Pygidium laticeps (Kner).
Trichomyderus laticeps Kner, Sb. Acad. Wiss. Miinchen, 1863, p. 228; Kner &
Steindachner, Abhandl. K. Bayer. Akad. Wiss., 11. CL, Vol. X, Part I, 1864,

308 MEMOIRS OF THE CARNEGIE MUSEUM.

p. 54 (western slope of Andes of Ecuador); Gunther, Cat. Fishes Brit. Mus.,
V, 1864, p. 274.
Pygidium laticeps Eigenmann & Eigenmann, Proc. Cal. Acad. Sci., II, 1889, p. 51;
Occasional Papers Cal. Acad. vSci., I, 1890, p. 334; Proc. U. S. Nat. Mus., XIV,
1891, p. 36; Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,
1910, p. 399.
Habitat. — Western slope of the Andes of Ecuador.

13811, I. U. M., one, 79 mm. Below Paramba, Prov. Imbabura, Ecuador. 2600
feet. Henn.

13812, I. U. M., 7094a-d, C. M., seventy-one, largest 89 mm. Mindo. Henn.
Head 5-5.2; D. 10.5; A. 8.5; P. 7.

G. 12. Pygidium laticeps (Kiier & Steindachuer). After Kiier & Steindachner.

The barbels in the specimens average a little longer than in the specimens of
P. tceniuni from Llanos. Distance between last anal ray and base of middle caudal
ray about 5.5 in the length; distance between origin of dorsal and base of middle
caudal rays on an average 2.28 in its distance from the snout; the last dorsal ray over
the last anal ray, the two fins coterminous.

Teeth: Incisors in two series in each jaw.

Dark brown, with obscure spots, about the size of the eye, evenly distributed
over the sides and back. No trace of a lateral band in any of the specimens.

In the very small the last anal ray is a little back of the vertical from the last
dorsal ray, and the color is uniform.

24. Pygidium stellatum Eigenmann. (Plate XLVII, fig. 1.)

7097, C. M., type, 78' mm. Quebrada Sarjento. Gonzales.

7098a-c, C. M.; 13814, I. U. M., paratypes, 05-85 mm. Quebrada Sarjento.
Gonzales.

7096rt-c, C. M.; 13815,1. U. M.,five, 45-86 mm. Quebrada Guamal. Gonzales.

7099a-b, C. M.; 13816, I. U. M., five, 37-55 mm. Quebrada Guadual. Gonzales.

7100a-i, C. M.; 13817, I. U. M., seventeen, largest 50 mm. Rio Guaduas. Gon-
zales.

13807, I. U. M., three, 31-50 nun. Quebrada Cristalina, 28 kilom. above Puerto
Berrio, 1000 ft. E. B. Williamson.
Head 5.5-6; D. 10.5; A. 7.5; P. 7; eye in middle of the head, or very shghtly in

I

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATPISHES. 309

front of the middle; interocular three times in the head; teeth broad incisors, in two
series.

Nasal barbels extending to the tip of the opercular spines or a little further,
maxillary barbels usually beyond base of last pectoral rays; pectoral narrow, the
filament longer than the head; origin of ventrals equidistant from base of middle
caudal rays and some part of the opercular spines, the tips of the ventrals reaching
vent; origin of anal under anterior half of the dorsal, the distance between the base
of the last anal ray and the base of the middle caudal rays 4.5-5 in the length;
caudal rounded, 5 to 6.5 in the length; origin of dorsal about equidistant from tip
of caudal and opercle, its distance from the base of the middle caudal rays about two
times in its distance from the snout.

A narrow, dark lateral stripe, a variable number of dark spots, smaller than
the eye, above the band and below it on the tail.

The specimens from Cristalina have the lateral band very broad, the maxillary
barbels reaching to the middle of the pectoral.

25. Pygidium chapmani Eigenmann. (Plate XLVII, figs. 2, 3.)

Pygidncm chapmani Eigenmann, Indiana University Studies, No. 16, dated Sept.,
issued Dec. 23, 1912, p. 18. (Boquia.)

4817, C. M., type, 106 mm.; paratypes, 4&IMA, C. M.; 12678, 1. U. M., Boquia.
Eigenmann.

7091a-6, C. M.; 13805, I. U. M., 4, 78-118 mm. Rio Dagua at Caldas. Eigen-
mann

Habitat. — Upper Cauca Valley.
Head 5-5.75; D. 10.5; A. 7-8.5; P. 7; interocular 3.5 in the head; eye in middle

of the head; width of head equal to its length in the young, narrower in the adult.

Teeth in 75 mm. specimens long and very narrow chisels, in smaller specimens less

distinctly chisel-shaped, in 65 mm. specimens long and smaller, conical.

Nasal barbels extending to base or near tip of opercular spines; maxillary barbel

just beyond origin of pectorals.

Fig. 13. pygidium chapmani Eigenmann. Type, 106 mm., C. M. No. 4817. Boquia.

310 MEMOIRS OF THE CARNEGIE MUSEUM.

Pectoral narrow, the outer ray about eciual to the head in length; origin of the
ventrals about equidistant from base of caudal and middle of pectorals, the tips
at or slightly beyond anus; origin of anal below middle or posterior part of the
dorsal, the distance between the base of its last ray and the middle caudal rays
five and one-third in the length; caudal distinctly rounded, about six and one-half
in the length; origin of dorsal over tip or middle of the ventrals, its distance from
the base of the middle caudal rays about two in its distance from the snout.

Smallest specimens with a black lateral band, a series of spots above and below

it in the older, the band breaking up into a series of spots in specimens over sixty

millimeters long. The oldest specimens dark with obscure darker spots and mot-

tlings.

26. Pygidium taenium (Kner).

Trichomycterus taenia Kner, 8b. Akad. Wiss. Mlinchen, 1863, p. 228; Kner &
Steindachner, Abhandl. k. Bayer. Akad. Wiss., II. ('!., vol. X, part I, 1864,
p. 52 (western slope of Andes of Ecuador); GtJNTHER, Cat. Fishes Brit. Mus.,
V, 1864, p. 274.

Pygidhcm toenia Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889,
p. 51; Occasional Papers Cal. Acad. Sci., I, 1890, p. 334; Proc. U. S. Nat.
Mus., XIV, 1891, p. 36; Eigenmann, Reports Princeton Univ. Exped. Pata-
gonia, III, 1910, p. 399.
Habitat. — Western slope of the Andes of Ecuador and southern Colombia.

13813 I. U. M., 7095a-d, C. M., forty-two, 31-111 mm. Los Llanos, southern
Colombia. March 8, 1913. Arthur Henn.

Fi(i. 14. Pi/gidiniii tiviiiiiin. (Kuer). After Kner & Steiiiilacliuer.

Head 5.2-5.6; D. 9.5 or 10.5; A. 8.5; P. 7; eye in the middle of the head, inter-
orbital three in the length of the head.

Nasal barbel reaching to the opercular spines, the maxillary barbels to the
pectoral; outer pectoral ray witli its filament a little shorter than the head, the rays
about equal to the head without the snout; ventrals reaching the vent, their origin
equidistant between base of middle caudal rays and opercular spines; origin of
anal under anterior half of the dorsal, the distance between the base of its last ray
and the middle caudal rays about five in the length; caudal rounded, five and five-

I

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 311

tenths to six and five-tenths in the length; origin of dorsal over tips of ventrals, its
distance from the base of the middle caudal rays one and eight-tenths to two in its
distance from the snout.

Teeth in the largest specimen narrow chisels, three rows in the premaxillary and
the middle of the mandible. In the 3'oung the teeth are more nearly conical.

Color variable ; a dark band from the opercle to the middle caudal rays, some-
times in part, or as a whole, replaced by a series of large spots; an irregular series
of irregular spots half way between the lateral band and the ventrals and anal, this
series more rarely replaced by a band; a band or a series of spots between the lateral
band and the mid-dorsal line; a mid-dorsal band; small sjiots sometimes interspersed
among the larger ones.

27. Pygidium caliense Eigenmann."
Pygidium caliense Eigenmann, Indiana University Studies, No. 16, p. 18, dated

Sept., 1912, issued Dec, 1912.
4819, C. M., type, 53 mm. Call. C. H. Eigenmann.

Head 4.88-5.75; D. 10.5 or 11.5; A. 9.5 or 10.5; P. 7; eye about in middle of
the head interocular 3.5-4 in the length of the head.

Fig. 15. Piigidiiiin caliense Eigenmann. T.vpe, 4819, Carn. Mus., .53 mm. Cali.

Nasal barbels extending to base or end of opercular spines, very little shorter
in the type ; maxillary barbel extending to the end of the opercular spines or beyond
the origin of the pectorals; pectoral rays about equal to the length of the eye and
the post-orbital part of the head, the filament extending for more than half the
length of the fin beyond the tip of the divided rays on one side in the largest speci-
men, nearly as long as the head, shorter in other specimens; origin of the ventrals
equidistant from base of middle caudal rays and base or middle of the pectoral
rays, reaching just beyond the vent; origin of the anal about under the middle of
the dorsal, the distance between the base of its last ra.y and the base of the middle
caudal ray four and one-third to five in the length; caudal rounded, five and one-
half to six and one-half in the length; accessory rays large, numerous; origin of dorsal

-- The head in the figure is a Httle too short.

312 MEMOIRS OF THE CARNEGIE MUSEUM.

over middle of the ventrals, its distance from the middle of the caudal one and
three-fourths to one and eight-tenths in its distance from the snout.

Sides and back in the young with black spots, the middle ones of the sides
larger and forming a more or less regular series in the young and half-grown; in the
largest specimen the caudal peduncle and base of the caudal are profusely covered
with spots smaller than the eye, the spots larger, less numerous and less conspicuous
forwards.

Teeth round-tipped incisors in the largest, more pointed but distinct incisors
in the middle-sized specimens; the teeth of the type narrow, pointed incisors.

28. Pygidium latidens Eigenmann. (Plate XLVII, fig. 4.)
Pygidium latidens Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. 693.
13801, I. U. M., one, 53 mm. Small creek near the mouth of Rio Calima, north of

Buenaventura. May 7, 1913. Henn.

Head 5.5; D. 9.5; A. 7.5; P. 7; posterior edge of eye in advance of the middle of
the head; interocular 3.5 in the head.

Nasal barbel extending beyond the tips of the opercular spines; maxillary
barbel extending beyond the axil, longer than the head; pectorals broad, as long
as head without snout; pectoral filament equal to the distance from the snout to the
axil; ventrals not nearly reaching anus, their origin equidistant from the base of
the middle caudal rays and the interopercle ; origin of anal about under middle of
the dorsal, distance between base of the last ray and the middle caudal rays five
and a half in the length; caudal rounded, about six in the length; accessory rays
well developed ; origin of dorsal over anus, its distance from the middle caudal rays
two in its distance from the snout; gill-membrane free to below the anterior spine
of the interopercle, without a free membrane across isthmus; both jaws with two
series of thin, chisel-shaped teeth.

Color plain, without spots or stripes.

29. Pygidium metae Eigenmann. (Plate XLVII, fig. 5.)
Pygidium metm Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. 694.
13770, I. U. M., one, 78 mm. Barrigona. March, 1914. Manuel Gonzales.

Head 6.3 in the length; D. 10.5; A. 9.5, counting the rudimentary rays; P. 6;
width of head nearly equal to its length; eye entirely in the anterior half of the
head, snout 2.75 in the head, interocular 3.5. Teeth conic.

Nasal barbels reaching to tip of opercular spines, maxillary barbel slightly
beyond origin of pectorals; pectorals small, equal to the postorbital portion of the
head, the first ray with its filament equal to the head, origin of ventrals much nearer

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 313

base of middle caudal rays than to tip of pectorals, their tips reaching the anal;
origin of anal under fourth dorsal ray (second fully developed ray), the distance
between the base of its last ray and the base of the middle caudal rays six times in
the length; caudal rounded; origin of dorsal over tip of ventrals, its distance from
the base of the middle caudal rays two and two-fifths times in its distance from the

snout.

Sides and back densely covered with spots about the size of the eye.

30. Pygidium stramineum Eigenmann. (Plate XLIX, fig. 1.)
Pygidium stramineum Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. G94.
7101, C. M., type; 13818, I. U. M., paratype, 46 and 50 mm. Quebrada del

Mango, Santandcr. Gonzales.
7089a, C. M., paratype, 35 mm. Quebrada del Maradat (?) Santander. Gonzales.
7090a-c, C. M., 13804, I. U. M., paratypes, seven, largest 45 mm. Quebrada da

Densino, Santander. Gonzales.
7102o-/i, C. M.; 13819, I. U. M., fifteen, largest 60 mm. Quebrada Deocamante,

Santander. Gonzales.
7103a-6, C. M.; 13826, I. U. M., four, the largest 67 mm. Quebrada de Zuarta,

Santander. Gonzales-.
7104, C. M., one, 41 mm. Quebrada de La Honda, Santander. Gonzales.

Head 4.5-5.33; D. 10.5; A. 8.5-9.5; P. 9; posterior margin of eye in the middle
of the head; interorbital three in the length of the head; teeth bristle-like, in about

three series.

Nasal barbels reaching base of opercular spines or beyond origin of pectorals,
maxillary barbels to tip of opercular spines or axil; pectoral filament about equal to
the length of the head, the rays equal to the length of the head without the snout;
origin of ventrals equidistant from the base of the middle caudal rays and a point
between the axil and a Uttle in front of the opercle (and the tips of the opercular
spines in the type), tips of the ventrals slightly behind the vent; origin of the anal
behind the vertical from the base of the last dorsal ray or under the posterior half
of the dorsal, the distance between the base of the last anal ray and the middle
caudal rays 4.5^5 in the length, accessory caudal rays very large and numerous;
caudal rounded, six and a half times in the length; origin of dorsal over the origin
of the ventrals, or but slightly behind this point, always nearer the eye than the tip
of the caudal, sometimes equidistant from tip of snout and tip of caudal, its dis-
tance from the base of the middle caudal rays one and a half or less in its distance
from the snout.

Uniform straw-colored in alcohol.

314 MEMOIRS OF THE CARNEGIE MUSEUM.

31. Pygidium unicolor Regan.
Pygidiurn unicolor Regan, Ann. & Mag. Nat. Hist. (8), XII, Nov., 1913. (Con-

doto.)

Habitat. — San Juan basin.

The following is the original description of Regan :

" Depth of body 7 in length, length of head 6. Head as broad as long. Diam-
eter of eye 12 in length of head or 3 in interocular width; eyes well in advance of
middle of head, close behind nostrils. Barbels as long as head. Dorsal 8-9, with
5 or 6 branched rays, rounded; origin above or a little in advance of vent, 1^ as far
from end of snout as from base of caudal. Anal 7, with 4 branched rays; origin
below last rays of dorsal. Pectoral filament f to as long as head, branched rays f
length of head. Pelvics covering vent. Caudal subtruncate. Coloration uni-
form .

"Two specimens, 80 and 85 mm. in total length, from the Condoto (Spurrell)."

32. Pygidium kneri (Steindachner) . (Plate XLVI, figs. 1, 2.)
Trichoniycterus knerii Steindachner, Ichthyol. Bcitr., XII, 1882, p. 21, pi. V,

figs. 1-1 a.

Habitat. — Canelos, Rio Bobonaza; Rio Zamora, eastern slojie of Ecuador;
Rio Meta, eastern Colombia.
13907, I. U. M., one, 155 mm. Barrigona, Rio Meta. Gonzales.

Head 5.7; depth 6.3; D. 10; A. 10 including the rudimentary rays; eye 9 in the
head, interocular 3, snout 2.3; eye in middle of the head.

Nasal barbel extending a little beyond gill-opening, as long as the head; maxil-
lary barbel reaching to near tip of the shortest pectoral ray; first pectoral ray with
its filament a little longer than the head, the rays about equal to the part of the
head behind the nasal barbels; origin of ventrals equidistant from base of middle
caudal raj^s and the eye ; origin of anal under last dorsal ray ; caudal truncate when
half expanded, slightly rounded or emarginate when fully expanded or compressed,
its middle rays equal to the length of the head; dorsal rays coterminous when de-
pressed; distance of origin of dorsal from base of middle caudal rays 1.6 in its dis-
tance from the snout; depth of caudal peduncle about 1.5 in its length, which is
five in the length..

Shghtly darker above, without spots or streaks.

Steindachner's description and figure give the following variation from the
above.

"Head 5.25-5.66; depth 6.75-7.5; D. 9; A, 7; eye 5-6 in the head; snout 2.5;
interocular 2.66 3.75; width of the head 1.33-1.4 in its length.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 315

"Nasal barbels reaching to gill-opening, maxillary barbel considerably beyond
origin of pectoral; upper pectoral ray prolonged; origin of the dorsal behind that
of the ventrals, its last ray over or a little in advance of the first anal ray; caudal
slightly emarginate in tlie figure, said to be "schwach convex" in the text. Choco-
late brown, thickly peppered with somewhat darker, very small, irregular spots or
points."

33. Pygidium meridae Regan. (Plate XLIX, fig. 2.)
Pygidium meridce Regan, Ann. & Mag. Nat. Hist. (7), XII, Dec, 1903, p. 624

(Merida and from Rio Albireggas above Merida, 3500 meters), Eigenmann,

Reports Princeton Univ. Exped. Patagonia, ITT, 1910, 399.

Habitat. — Cordillera of Merida, Venezuela.
13771, I. U. M., one, 99 mm. Merida. Purchased from Rosenberg.

Head 6 [-7]; D. 10.5 [6-7 branched rays]; A. 9.5 [4 or 5 branched rays]; P. 8;
eye. in front of the middle of the head; interocular contained three and one-third
in the length of the head [3], snout two and one half times.

Nasal barbel reaching a little beyond the eye; maxillary barbel very slender,
reaching about to middle of the pectorals, longer than the head [as long or nearly as
long as the head]; outer pectoral ray equals length of maxillary barbel, the pro-
jecting filament being half as long as the rest of the fin [one and one-third times as
long as head] ; origin of ventrals equidistant from base of middle caudal rays and
posterior portion of head, the ventrals reaching the vent; origin of anal below
penultimate dorsal ray ; distance between last anal ray and the middle caudal rays
four and three-fourths in the length [four and two-thirds to five times]; caudal
rounded [truncate], six and a half times in the length; origin of the dorsal on the
vertical from a jioint just in front of the vent, over tip of ventrals; distance between
origin of dorsal and base of middle caudal rays one and three-fourths times in its
distance from the snout [one and two-thirds to one and four-fifths].

No lateral band ; traces of dark spots.

The characters found by Regan are given in brackets.

34. Pygidium bogotense Eigenmann. (Plate XLIX, figs. 3, 4.)
Pygidium bogotense Eigenmann, Indiana University Studies, No. 16, p. IS, dated
Sept., issued Dec. 23, 1912. (Madrid; Chapinero.)
Habitat. — Plains of Bogota and northward.
4820, C. M., type; 4821, C. M.; and 12679 I. U. M.,paratypes, two hundred thirty-
nine, largest 80 mm. Puente de Supa, beyond Chapinero, north of Bogota.
Eigenmann.

316 MEMOIRS OF THE CARNEGIE MUSEUM.

4834, C. M.; 12680, I. U. M., paratypes, six. Madrid, plains of Bogota. Eigen-

mann.
University of Michigan, 94 mm. Mountains, 4,000 ft., Guira River, Santa Marta.

July 18, 1913. A. S. Pearse.
University of Michigan, 66 and 68 mm. Small stream, San Lorenzo, Santa Marta,

4,500 ft. Julys, 1913. A. S. Pearse.
7087rt-c, C. M., 13802, I. U. M., five, largest 52 mm. Rio Piedras, Santander.

Gonzales.
7457, C. M.; 13845, I. U. M., five, largest 55 mm. Quebrada da Charda. San-
tander.
7088, C. M., 13803, 1. U. M., twenty-six, largest 80 mm. (Label illegible—" Puchada?

de la Pof guira? de Norte Zipa Quira? ") Gonzales.
13806, I. U. M., 85 mm. Mill-stream, Cincinnati (twenty miles from Santa

Marta), Colombia. Dec. 31, 1916, 4,500 feet. E. B. Williamson.

Head 5.25-6; D. 10.5 in two specimens, 11.5 in four, 12.5 in eight, 13.5 in two;
A. 10.5; P. 8; center of the eye very little in front of the middle of the head; inter-
ocular about three in the length of the head; head but little longer than wide; teeth
conical, in three or four irregular series.

Nasal barbel extending to tip or base of opercular spines; maxillary barbel
extending to the base of the opercular spines or beyond the base of the pectoral;
pectoral rays about as long as the head behind the nasal barbel, pectoral filament
about as long as the head ; origin of ventrals equidistant from base of middle caudal
rays and tip or base of the opercular spines, tips of ventrals extending to or very
sHghtly beyond the vent; origin of anal under one of the last three dorsal rays or
just behind the vertical from the last one; distance between the base of the last anal
ray and the middle of the caudal ray four and three-fifths to five in the length;
caudal rounded, six to seven in the length; accessory caudal rays numerous and
large; origin of dorsal over origin or posterior half of the ventrals, equidistant from
tip of caudal and eye,-^ its distance from the base of the middle caudal rays one and
five-tenths to one and seven-tenths in its distance from the snout.

^' In this specimen the origin of the dorsal is over the origin of the ventrals. In the specimens from
the plains about Bogotd examined in this regard, only one had the origin of the dorsal a little further
forward, a number had it equidistant between the tip of the caudal and a point some distance behind the
eye. In the specimens from the Santa Marta Mountains, the origin of the dorsal is a little further for-
ward. These specimens approach nigromaculatum, to which they ought perhaps to be referred. It is
certain that the largest specimens referred to nigromaculatum from Santa Marta l^olong to a species dif-
ferent from those found on the plains of Bogotd. It is possible that P. bogotensc is also found in Santander
and Santa Marta, but it is also possible that the halfgrown of P. bugotense are indistinguishable from the
half-grown of P. nigromaculatum, and that the specimens from Santander and Santa Marta are really the
latter species.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 317

Sides and back with numerous irregular spots, larger in the larger specimens,
sometimes referable to distinct series. The spots are smaller in the specimens from
Santander.

35. Pygidium nigromaculatum (Boulenger). (Plate XLIX, fig. 5.)
Trichomycterus nigromaculatus Boulenger, Ann. & Mag. Nat. Hist. (5), XIX,

1887, p. 349 (Andes of Colombia).
Pygidium nigromaculatum Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2),

II, 1889, p. 52; Occasional Papers Cal. Acad. Sci., I, 1890, p. 336; Proc. U. S.

Nat. Mus., XIV, 1891, p. 37; Eigenmann, Reports Princeton Univ. Exped.

Patagonia, III, 1910, p. 400.

Habitat. — Andes of Colombia, especially Sierra de Santa Marta and Santander.
No. ? , University of Michigan, two, 138 and 165 mm. Small stream, San

Lorenzo, Santa Marta Mountains (4,500 ft.). Sept. 9, 1912. M. A. Carriker.
No. ? , University of Michigan, two, 118 and 150 mm. Same place, Jan.

16, 1913. M. A. Carriker.
No. ? , University of Michigan, one, 73 mm. Same place, July 8, 1913.

A. S. Pearse.
No. ? , University of Michigan, eight,-^ largest 55 mm. Same place, July

9, 1913. A. S. Pearse.

No. ? , one, 115 mm. Locality?

13837, I. U. M.; 7447a-6, C. M., four, largest 93 mm. Quebrada de la Raya,

Santander. Gonzales.
7448a-6, C. M., two, the larger 45 mm. Quebrada Capitanejo, Santander. Gon-
zales.
7449a-c, C. M.; 13838, I. U. M., six, largest 68 mm. Quebrada de Cobarachior,

Santander. Gonzales.

Description of the Species from San Lorenzo.

The characters given by Boulenger are in brackets.

Head 5.24-5.75 [6.5 in total]; D. 11.5 or 12.5; A. 9.5; P. 9; eye in middle of
the head ; interocular 3 in the length of the head ; width of head equal to its length
[longer than the distance between snout and a line connecting the tips of the two
bunches of opercular spines]; teeth conical [pointed, recurved].

Nasal barbels extending to the tips of the opercular spines or to the base of the

" These specimens are so distorted that it is difficult to refer them to their proper place. The
origin of the dorsal seems to be equidistant from the tip of the caudal ant! the front of the eye. There
are traces of longitudinal bands.

318 MEMOIRS OF THE CARNEGIE MUSEUM.

last pectoral ray; maxillary barbels extending to the base of the last pectoral ray
or a little beyond; pectoral broad, rounded, the filament equal or nearly equal to
the length of the head ; origin of the ventrals very little nearer base of middle caudal
rays than e.ye, their tips reaching the vent or very little beyond it; origin of anal
under the penultimate dorsal ivay [anal entirely behind the dorsal], the distance
between its last ray and the middle caudal rhy four and one half to five in the length ;
caudal rounded, about six in the length; accessory caudal rays not conspicuous;
origin of dorsal over the anterior two-thirds of the ventrals; origin of dorsal nearly
equidistant from tip of snout and tip of caudal, its distance from the base of the
middle caudal rays on an average 1.4 in its distance from the tip of the snout.

Sides and back in the largest with numerous irregularly arranged spots about
the size of the eye, the spots larger and less numerous in the young, much as in
P. bogotense.

In the specimens from La Raya, Santander, the dorsal and ventrals are a little
farther back, the ventrals equidistant from base of middle caudal rays and tip of
opercular spines, the distance between the dorsal and the middle caudal rays 1.52
in its distance from the snout. The anal is entirely behind the dorsal. The origin
of the dorsal is equidistant from the eye and the tip of the caudal. Plain or but
faintly spotted.

Specimens from Capitancjo resemble those from La Raya.

Some of the specimens from Cobarachior seem to be typical huyotcnsc, while
others approach the specimens from La Raya and P. latistriakmi from Pinchote.
P. bogotense, typical of the plains of Bogota, grades into P. nigromaculatum of
Santa Marta in Santander and Santa Marta.

36. Pygidium banneaui Eigenmann. (Plate XLVIII, fig. 1.)
Pygidium banneaui Eigenmann, Indiana University Studies, No. 16, dated Sept.,
issued Dec. 23, 1912, p. 19.
Habitat. — Streams near Honda, Colombia.
4815, C. M., type; 4816a-2, C. M., 12677, I. U. M., paratypes, eighty-nine speci-
mens, the largest 44 mm. Bernal Creek, near Honda. Eigenmann.
7456a-b, C. M.; 13844, I. U. M., four, 35-41 mm. Ciuaduas? Gonzales.

These specimens might be considered the young of some of the other species if it
were not for the fact that one specimen, 34 mm. long, contains eggs over a milli-
meter in diameter, which must be nearly mature.

Head 5.33-5.5; depth 5.5-7; D. 10.5 or 11.5; A. 7.5 or 8.5; P. 8; eyes slightly
in advance of the middle of the head; interocular three and five-tenths in the length
of the; head; teeth conical.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 319

Nasal barbels reaching to the tips of the opercular spines, maxillary barbel to
near the middle of the pectoral; upper pectoral ray prolonged in a filament, as
long as the head or a little longer; ventrals reaching thc^ vent, their origin about
equidistant from snout and tii)s of middle caudal rays; origin of the anal under the
middle of the dorsal, the distance between the last ray and the middle caudal rays
4.5-5 in the length; caudal distinctly emarginate, about five in the length; origin
of dorsal a little nearer tip of caudal than eye, the distance between the origin of
the dorsal and the caudal 1.4-1.66 in its distance from the snout.

Sjiecimens 18 mm. long have a black line from the nasal barbel to near the tip
of the middle caudal rays, accented in places; in specimens 20 mm. long the line is
accented more strongly, appearing to be breaking up into spots; there are also
spots on the back; in older specimens the line becomes diffuse; with growth a dis-
tinct series of spots develops along the middle of the back in front of the dorsal,
and another series between these and the lateral series; in the largest the sides and
back are profusely spotted, the spots varying in size and arrangement.

In the specimens from Guaduas the barbels are a little shorter and the dorsal a
trifle farther forward. The color markings are less profuse than in the type.

37. Pygidium spilosoma Regan. (Plate XLVIII, fig. 2.)
Pygidium spilosoma Regan, Ann. & Mag. Nat. Hist. (8), XII, Nov., 1913, p. 468.

(Rio Sipi and Rio Tamana.)

This species from the Pacific drainage of central Colombia is known from three
specimensT 30-250 mm. long, described by Regan, and 7092, C. M., 97 mm. Cor-
dova on the Rio Dagua. Eigcnmann.

In the following description, Regan's data are given in brackets.

Head 6 [to 6.75]; depth 7 [to 8]; D. 11.5 [9, with 6 branched rays]; A. 9.5 [7,
with 4 branched rays] ; P. 8 ; eye very little in front of the middle of the head ; interocu-
lar 3.5 in the length of the head [2.5-3]; head longer than wide, tapering forward, the
space between the nasal barbels 6.5 in the length of the head. Teeth minute, conical.

Nasal barbels extending to the base of the opercular spines, the maxillary barbel
to their tip [to basal part of pectoral] ; pectoral rather narrow, the upper part trun-
cate, the filament eciual to the head; origin of the ventrals equidistant from base of
middle caudal rays and the tip of the opercular spines, a little too far forward in the
figure, origin of the anal under the middle of the dorsal [a little behind end of dorsal],
the distance between the base of its last ray and the base of the middle caudal rays
5 in the length; caudal truncate, six and one-fourth in the length; dorsal obliquely
truncate, origin of dorsal over posterior half of the ventrals, its distance from the
base of the middle caudal rays 1 .6 in its distance from the snout.

320 MEMOIRS OF THE CARNEGIE MUSEUM.

An obscure, dusky band along the middle of the sides; sides and back with
obscure spots about the size of the eye.

38. Pygidium dorsostriatum Eigenmann. (Plate XLVIII, fig. 3.)
Pygidiiim dorsostriatum Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p.

695.
7093a-&, C. M.; 13810, I. U. M., four, 18-76 mm., the largest the type. Villavi-

cencio. Manuel Gonzales.

Distinguished by the eccentric, dark, lateral band.

Head 5; D. 12.5 (of which 4 minute); A. 9.5; P. 9; center of eye very little in
advance of middle of the head,'^ interocular three in the head. Teeth conic.

Nasal barbels extending to, or but shghtly beyond, origin of pectoral; maxillary
barbel to the axil, equal to the length of the head; pectoral filament equal to the
length of the head, the longest ray equal to the length of the head behind the nasal
filament; origin of ventrals equidistant from base of middle caudal rays and base or
tip of the interopercular spines, ventrals nearly reaching the anal; origin of the anal
under the last quarter of the dorsal, the distance between the base of its last ray
and the base of the middle caudal rays about 4.5 in the length; caudal rounded, six
and five-tenths to seven times in the length; the first rudimentary dorsal ray over
the base of the ventrals, its distance from the base of the middle caudal ray equal
to its distance from the tip of the opercular spine, 1.47 in its distance from the snout.

A dark band or row of spots from just above the gill-opening to the base of the
upper caudal lobe; a few spots below the band in the front half of the body in the
larger specimen.

This description is based on the two larger specimens, 68 and 77 mm. long.
The two smaller s^jecimens, 18 and 21 mm. long, are uniform in color.

39. Pygidium venulosum Steindachner.
Pygidium venulosum Steindachner, Anz. K. Akad. Wiss. Wien, 1915, No. XVII,

p. 199 (Paramo de Cruz Verde, Eastern Cordilleras, Colombia, 3,000 M.)

Habitat. — Eastern Andes of Colombia.

I have not seen this species.

I). 10 or 11; A. 10; P. 8. Caudal peduncle greatly compressed. Caudal
rounded; eye very small, a very little in front of middle of head. Barbels short,
about reaching ej^e; origin of anal under middle of dorsal; origin of dorsal equidistant
from tip of caudal and lateral margin of head; teeth pointed. First pectoral ray

" In the plate the eye is placed too far forward; the anterior margin should be where the posterior
nnargin is.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 321

not prolonged. Lateral band above middle of sides; back and sides with dark
reticulations on a lighter background.

40. Pygidium latistriatum Eigenmann. (Plate XLVIII, fig. 4.)
Pygidium latistriatum Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. 696.
7450, C. M., type. 46 mm. Quebrada de Pinchote, Santander. Gonzales.

Head 8 mm., length to base of caudal 39 mm. ; width of head 6 mm., intcrocular
2.5 mm.,eye a little in front of the middle; distance from snout to origin of dorsal
23 mm., to its last ray 27 mm.; distance between origin of dorsal and base of middle
caudal rays 16 mm., distance from snout to origin of ventrals 22 mm., to origin of
anal 28 mm., distance between base of last anal ray and base of middle caudal
rays 9 mm., maxillary barbel 9 mm., nasal barbel 7 mm., length of outer pectoral
ray with its filament 8 and 9 mm., the divided rays 5 mm., D. 8.5; A. 6.5, not count-
ing the imbedded rays in either case; upper caudal rays 8 mm.; lower caudal rays
about 6.5 mm. Accessory caudal rays numerous.

A lateral band from above the opercle to the middle of the caudal, increasing
in width backward; mid-dorsal fine dark; a dark stripe in front of the dorsal between
the lateral stripe and the mid-dorsal stripe.

It is possible that some of the specimens under P. nigromaculatum belong here.

41. Pygidium striatum Meek & Hildebrand.
Pygidium striatum Meek & Hildebkand, Field Mus. Publ., No. 166, Zool., Ser. X,
Dec, 1913, p. 78 (Rio Cana); Publ. 191, Zool., Ser. X, Dec, 1916, p. 266 (Rio
Cana), Tuyra Basin, Panama.
Habitat.— Colomhisi from Santander to the Rio Dagua in west central Colombia

and the Rio Tuyra in Panama.

I have been able to examine the types and the following specimens:

Catalog Numbers.

7113a-i, CM. 13S20I.U.M...
7114a-6, CM., 13821 I.U.M..
7115a-6, CM., 13822 I.U.M..

7105 CM.

7106a-6, CM.,
710So-fc, CM.,
7109a-i, CM.,
7111a-rf, CM.
?7112, CM...
13829 I.U.M..

13823 I.U.M. .
1382-1 I.U.M. .
1382.5 I.U.M..
13828 I.U.M..

Number of
Specimens.

41
5

4

5
20
16
9
1
1

Length in
Mm.

Locality.

79 largest

33-GO

41-61

35 aud 71

58 largest
50 largest
60 largest
43 largest
39
45

Quebrada Sarjento
Quebrada Alban
Quebrada de la Ropera,

Santander
Quebrada de la Hato,

Santander
Guadual
Villeta

San Gil, Santander
Rio Guaduas
Quebrada Chamisal
Caldas, Rio Dagua

Collector.

Gonzales.
Gonzales.

Gonzales.

Gonzales.
Gonzales.
Gonzales.
Gonzales.
Gonzales.
Gonzales.
Eigenmann^

Head 5-6; D. 10.5 or 11.5; A. 8.5 or 9.5; P.8; eye very little in advance of the

322 MEMOIRS OV THK CARNEGIE MUSEUM.

middle of the head, interoculur 3.5 to 4 in tlie lengtli of tlie head; the width of tlie
head equal to its length behind the nasal barbels.

Nasal barbels about reaching the tip of the opei-cular spines, the maxillary
barbels sometimes to the axil; first pectoral ray with its filament about equal to the
length of the head, the rays equal to the length of the eye and postorbital portion
of the head or a little longer ; origin of the ventrals equidistant from the base of the
middle caudal rays and some point in the basal half of the pectorals, their tips
reaching the vent in the young, falling short in the adult; origin of anal under the
middle of the dorsal, the distance between the base of the last ray and the base of
the middle caudal ray five to five and one-half in the length; caudal rounded, six in
the length; origin of the dorsal over some point in the last half of the ventrals,
equidistant between tip of caudal and middle of pectorals or a little farther forward,
its distance from the base of the middle caudal rays 1.8-2.2 in the length.

Sides and back spotted, the spots usually confluent into a narrow median lateral
band and into a narrow band above and below the median band. The very young
with a narrow black lateral band without other markings.

The above description aiiplies particularly to the types and some of the speci-
mens between Honda and Facatativa, Nos. 7113, 7114, 7106, and some specimens
from Santander, No. 7105.

From these typical specimens the following variations were noted. In the
specimens from Villeta, No. 7108, also on the line between Honda and Facatativa,
the caudal is truncate with rounded outer edges, the origin of the dorsal is equi-
distant between the tii) of the caudal and the opcrcle or a little farther forward.
Some of these specimens are more distinctly spotted than the typical striatum,
approaching P. banneaui.

In the six'cimens from Guaduas, also along the line between Honda and
Facatativa, and in those from San Gil, No. 7109, the position of the dorsal agrees
with its position in those from Villeta, /. c, it is in front of the typical position. In
these, the most conspicuous marking is a black lateral band in which the spots are
not recognizable. In some of those from San Gil, the band above the median band
is also prominent, but it can sometimes be seen that both it and the median band
are made up of series of spots. A small specimen from Chamisal, No. 13831, is
nearly like them. The origin of the dorsal is median between the tip of the caudal
and the base of the opercular spines, the barbel extends nearly to the middle of the
pectoral; a narrow black lateral stripe, otherwise plain light.

The specimens from La Ropera are more profusely covered with small spots,
their longitudinal arrangement inconspicuous.

EIGENMANN: the PYGIDIID^, a family op south AMERICAN CATFISHES. 323

In the specimen from Caldas, No. 13829, the origin of the dorsal is equidistant

from the tip of the caudal and the opercular spines. The spots along the middle

of the sides are conspicuous, but not confluent.

To this species probably also belong the following specimens.

7451a, C. M., 13839, I. U. M., 3, largest 60 mm. Quebrada de la Pelada, San-
tandcr.

7452a, C. M., 13840, I. U. M., 2, larger 41 mm. Quebrada de la Callegona, San-
tander.

7453a-6, C. M., 13841, I. U. M., 3, largest (U mm. Rio Mogotes, Santander.

7454a-c, C. M., 13842, I. U. M., 6, largest 03 mm. Quebrada de Horizonte, San-
tander.

7455a-d, C. M., 13843, I. U. M., 8, largest 67 mm. Quebrada de Suescum, San-
tander.
These are like the typical specimens of striatum described above, but lack all

color markings, being uniformly pale.

Very close to striatum if not identical with it, is P. regani.

42. Pygidium regani Eigenmann. (Plate XLVIII, fig. 5.)
Pygidium regani Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. 696.
? Pygidium taenia Regan (non Kner & Steindachner), Ann. & Mag. Nat. Hist.

(8), XII, 1913, p. 469 (Rio Sipi and Rio Tamana).-

Habitat. — San Juan basin.
13772, I. U. M., one, 55 mm. Tado, Rio San Juan. Purchased from Rosenberg.

Head 6; D. 10.5; A. 8.5; P. 8; eye in middle of the head, interorbital four
times in the length of the head.

Nasal barbel as long as the head, reaching beyond axil of pectoral; maxillary
barbel reaching to near the end of the lower pectoral ray, considerably. longer than
the head ; outer pectoral ray as long as the head ; origin of ventrals equidistant from
base of middle ray and tip -of opercle, not quite reaching to the vent; origin of anal
under posterior half of dorsal, the distance from the base of the last ray to the
middle caudal ray contained five and one-half times in the length ; caudal six times
in the length; origin of dorsal equidistant from tip of caudal and opercular si:)ines,
over posterior third of the ventrals, its distance from the middle caudal ray one
and four-fifths in its distance from the snout.

A dark streak from opercular spines to middle of caudal; faint spots above and
below the lateral stripe.

324 MEMOIRS OF THE CARNEGIE MUSEUM.

Table of Measurements.

Length over all '. . 55 mm.

Length to base of caudal ' 47 mm.

Length of head to tip of opercular spines 8 mm.

Distance of origin of dorsal from snout 30 mm.

Distance of origin of dorsal from middle caudal rays 17 mm.

Outer pectoral ray S mm.

Maxillary barbel •. 11 mm.

Nasal barbel 8 mm.

Length of eye 1 mm.

Length of snout 3.5 mm.

Interocular distance 2 mm.

This si^ecies is very similar to driatum, and may be a synonym of it.

43. P*ygidium retropinne (Regan).
Trichomycterus retropinnis Regan, Ann. & Mag. Nat. Hist. (7), XII, Dec, 1903,

p. 624 (headwaters of Magdalena east of Papaganat, St. Augustine, Andes of

Colombia, 5,000 ft.).

Habitat. — Headwaters of the Rio Magdalena.

" Length of head 5| times in the total length. Head as broad as long. Diam-
eter of eye about 4 times in the interocular width, which is 3.33 times in the length
of the head. Snout as long as the postorbital part of head. Barbels equal to
about .8 the length of head. Dorsal with six branched rays, originating above or
slightly behind the anal opening, the distance from its point of origin to the caudal
2.4 times in the distance from the former to the tip of the snout. Anal with 4
branched rays, originating below the anterior third of the dorsal, the distance from
the base of its last ray to the caudal 5.4 times in the total length. Longest branched
ray of the pectoral, .66 the length of the simple outer ray, which is equal to it the
length of head. Ventrals not quite reaching the anal opening. Caudal truncate-
rounded. Brownish, with an indistinct darker stripe along the middle of the side
and traces of some dark spots."

"Total length 80 mm."

"A third specimen, 30 mm. in total length, which I have purposely excluded
from the above diagnosis, has a well-marked broad longitudinal stripe on each side.
In it the longest branched ray of the pectoral is ^ the length of the outer simple ray,
and the distance from the origin of the dorsal to the caudal is 2.2 times in the dis-
tance from the former to the tip of the snout."

Key to the Species of Pycudium from the Amazon to the Esseqdibo.
From the vast lowland area of the Amazon, Orinoco, and Guiana but few species are known.
a. Origin of the dorsal in front of the vertical from the origin of the anal; ma.xillary barbel reaching tip

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 325

of opercle; snout about 2.5 in the head; nasal barbel reaching not quite to the tip of the maxil-
lary barbel; head nearly as wide as long; eye just in front of middle of the head.
6. Head 6 in the length; the first pectoral ray with its filament equals the length of the head; sides
and back with numerous spots, each larger than the eye, in about five series between the dorsal

and anal; origin of dorsal equidistant from tip of caudal and eye 44. guianense Eigenmann.

66. Head 5 in the length; first pectoral ray with its filament equals the length of the head without the
opercle; uniform yellowish-brown above, lighter below, top of head marbted; origin of dorsal

nearer tip of caudal than eye 45. conradi Eigenmann.

aa. Origin of anal under origin of dorsal.

c. Head 6 in the length; maxillary barbel reaching tip of pectoral; upper parts obscurely spotted;
origin of dorsal equidistant from tip of caudal and middle of pectoral; eye in anterior half of

Ijg^jj 46. gracilior Eigenmann.

cc. Head 7 in the length; maxillary barbel reaching middle of pectoral; eye entirely in anterior half

of the head, nearly equal to the interocular and to the snout. . .47 amazonicum (Steindachner).

ccc. Head 5.5 in the length; maxillary barbel reaching tip of last interopercular spine; eye entirely in

anterior half of the head, considerably less than the interorbital or the snout; opercular and

interopercular bunch of spines alike 48. hasemani Eigenmann.

44. Pygidium guianense Eigenmann. (Plate L, fig. 1.)
Pygidium guianense Eigenmann, Ann. Carnegie Mus., VI, 1909, p. 11; Eigenmann,

Reports Princeton Univ. Exped. Patagonia, III, 1910, p. 400; Mem. Carnegie

Mus., V, 1912, p. 212 (Aruataima Cataract above Holmia.)

Habitat. — Upper Potaro River, British Guiana.
1003, C. M., type, 77 mm. Aruataima Cataract. Eigenmann.

Head 6; depth equals head in length; D. 9; A. 7; eye 4 in snout, 9.5 in head;
head nearly as broad as long; maxillary barbel reaching to tip of opercle; teeth in
bands of about four irregular series; origin of anal under middle of dorsal; dorsal
fulcra extending forward to near the dorsal; caudal rounded; first pectoral ray pro-
longed in a filament nearly as long as the rest of the ray; round dark spots every-
where, except on belly and lower surface of head; caudal dusky, the margin light.

45. Pygidium conradi Eigenmann. (Plate L, fig. 2.)
Pygidium conradi Eigenmann, Mem. Carnegie Mus., V, 1912, p. 212 (Amatuk and

Waratuk Cataracts).

Habitat. — Lower Potaro River, British Guiana.
2212, C. M., 41 mm., type. Amatuk Cataract. Eigenmann.
11710, I. U. M., 34 mm., paratype. Waratuk Cataract. Eigenmann.

With the characters given in the key.

The teeth conic.

326 MEMOIRS OF THE CARNEGIE MUSEUM.

46. Pygidium gracilior Eigenmann. (Plate L, fig. 3.)
Pygidium gracilior Eigenmann, Mem. Carnegie Mus., V, 1912, p. 213 (Erukin).

Habitat. — Lower Potaro River, British Guiana.
1730, C. M., 27 mm., type. Erukin. Eigenmann.

Head 6; depth 9; D. 8; A. 6; eye about 2 in the snout; interorbital a little
greater than snout, snout 3 in the head.

Slender, head as broad as long; maxillary barbel reaching tip of pectoral; nasal
barbel to origin of pectoral ; outer pectoral ray prolonged, about equal to the head
in length. Origin of the anal under origin of dorsal; distance from origin of dorsal
to origin of caudal 3.5 in the length; length of caudal 5 in the length.

All upper parts obscurely spotted.

47. Pygidium amazonicum (Steindachner). (Plate XLVI, figs. 3, 4.)
Trichomycterus amazonicus Steindachnt',r, Flussf. Siidam., IV, 1882, p. 29, pi. VI,

figs. 4-4a (Cudajas).
Pygidium amazonicum Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II,

1889, p. 53; Occasional Papers Cal. Acad. Sci., I, 1890, p. 338; Proc. U. S. Nat.

Mus., XIV, 1891, p. 37; Eigenmann, Reports Princeton Univ. Exped. Pata-
gonia, III, 1910, p. 400.

Habitat. — Amazon, at Cudajas.

This species is known from the type, a specimen 60 mm. long.

Head a little over 6, equal to the depth; D. 8; A. 7; P. 6; eye entirely in front
of the middle of the head; interorbital a httle greater than the eye; width of head
nearly equal to its length.

Head greatly depressed, caudal peduncle strongly compressed; nasal barbels
reaching nearly to gill-opening; maxillary barbel to the end of the first third of the
ujiper pectoral ray, lower barbel to the base of the pectoral ; dorsal and anal opposite
each other; ventrals very short, 2 in the head; caudal rounded; upper pectoral ray
nearly equal to the head.

Chocolate brown, with 'faint darker spots on the caudal peduncle; rays of dorsal
and caudal dotted with violet.

48. Pygidium hasemani Eigenmann. (Plate L, fig. 4.)
Pygidium hasemani Eigenmann, Ind. Univ. Studies, No. 20, March, 1914, p. 48.
5238 and 5239, C. M., type and paratypcs, many, largest under 18 mm. Santarem.

Haseman.

Habitat. — Amazon at Santarem to Bolivia.
76023, C. M., about 16 mm. San Joaquin. Sept. 4, 1909. Haseman.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 327

Head 5.5; D. 7 or 8; P. 6. Eye in anterior half of the head, about five times
in the length of the head, less than the snout, about two times in interorbital. Teeth
conical, in a single series; maxillary barbel extending to the tips of the interopercular
spines; axillary glands large; opercular and interopercular bunches of spines similar,
separate from each other; gill-membranes united, free from the isthmus; pectoral
very narrow, about equal to the head in length, the first ray prolonged; origin of
dorsal over origin of anal, its distance from the base of the caudal two in its distance
from the eye or occiput, equidistant from tip of caudal and middle of pectoral;
origin of vcntrals nearer snout than tip of caudal; caudal rounded, merging into
the large accessory rays.

Translucent, with many chromatophores; dusky spots on the back from a short
distance in front of the dorsal to the caudal; a dark bar across base of caudal; dark
spots, similar to those of the back, from between ventrals and anal to the caudal ; a
series of minute spots along the middle of the sides; a small spot on opercle, another
on the interopercle, a dark line forward from the eye; a minute spot on base of
ventrals.

Species from Southeastern Brazil.

The mountain brooks of southeastern Brazil from Rio Grande do Sul to the
Rio San Francisco harbor a number of species. One of these, P. iheringi, has
broad incisors, a number are known to have pointed teeth, while the rest, nigricans,
minutum, goeldii, itatiaycc and punctatissimum, I have not been able to examine.
They have the following distribution :

P. minutum Southern Rio Grande.

P. nigricans Santa Catherina.

P. davisi Rio Tguassu, southern Sao Paulo.

P. proops Ribeira, southern Sao Paulo and Rio Parahyba.

P. iheringi Ribeira, Santos, and Sapina, Sao Paulo.

P. paolence Northern Sao Paulo.

P. goeldii Parahyba basin.

P. vermiculatum Parahyba basin.

P. immaculatum Parahyba basin; Sao Matheos; Goyaz.

P. braziliense ? Rio Grande do Sul, ? Rio Ribeira; Rio das Velhas; Rio Doce.

P. itatiayce Parahyba basin.

P. triguttatum Parahyba basin.

P. reinhardti Burmier, into Rio das Velhas.

P. alternatum Rio Doce. '

P. punctatissimMm Araguay.

P. santoe-rita: Rio Preto.

328 MEMOIRS OF THE CARNEGIE MUSEUM.

Key to the Species of Pygidium found in Southeastern Brazil.
a. First dorsal ray prolonged in a filament; barbels scarcely reaching the edge of eye; back uniform

blackish, lower parts light; D. 11 ; A. 10; P. 9 49. nigricans (Cuvier & Valenciennes).

aa. First dorsal ray not prolonged; teeth incisors; origin of dorsal equidistant from tip of caudal and a
point between the eyes and nasal barbels.
6. First pectoral ray not prolonged, or with only a trace of a projection; first anal ray behind the

dorsal; sides and back with numerous spots 50. iheringi Eigenmann.

bb. First pectoral ray usually iirolonged; first anal ray under the dorsal; markings conspicuous.

51. zonatum Eigenmann.
aaa. First dorsal ray not prolonged; teeth conical.^'^

c. Origin of the dorsal nearer to the tip of the caudal than to the head, or sometimes equidistant
between tip of caudal and occiput in P. paolcnce.
d. Origin of the anal under the second dorsal ray, last dorsal ray over middle of anal; origin of
ventrals nearly equidistant from tip of caudal and snout; eye in anterior half of the head,

near the posterior nares; head 5. .5-6; D. 9; A. 7-9 52. proops (Ribeiro).

(Id. Origin of the anal under the middle of the dorsal, last dorsal ray over third anal ray; origin
of ventnUs nearer to tip of caudal than to tip of the snout; eye just in front of the
middle of the head; head six times in length.
€. D. 8.5; A. 6.5; many faint spots, a dark streak along the middle of the sides, another

above it 53. paolence Eigenmann.

ee. D. 9.5; A. 8.5; a broad lateral band with serrate edges; a series of spots below it and

another above it 54. reinhardti Eigenmann.

ddd. Dorsal opposite the space between ventrals and anal. (See No. 59 gocldii.)
cc. Origin of the dorsal equidistant between the tip of the caudal and some point near the eye."
/. Pectoral ray without a filament; origin of ventral nearer snout than tip of caudal; origin of
anal under jjosterior i)art of dorsal; color variable; D. 7; A. 5.. 55. davisi Haseman.
//. First pectoral ray usually prolonged as a filament.

g. Color plain; anal behind the dorsal; origin of ventral nearer snout than tip of caudal;
caudal very slightly emarginate; eye in the middle of the head; D. 11.5-13; A. 9.5-

10.5 56. immaculatum Eigenmann & Eigenmann.

gg. Variously marked.

/(. Origin of the anal under posterior half of the dorsal.

i. Origin of the dorsal over the origin of the ventrals; origin of the ventrals nearer
tip of caudal than eye; ventrals reaching anal; D. 8.5; A. 8.5; eye in middle
of head; sides and back profusely covered with confluent spots, leaving the
ground-color in irregular vermiculations. See also puctatissimum.

57. vermiculatum Eigenmann.
n. Origin of the dorsal behind the origin of the ventrals.

j. Sides and back with large spots, sometimes alternating across the back;
origin of ventrals equidistant from snout and middle of caudal; dis-
tance between origin of dorsal and caudal about 1.5 in distance of
dorsal from the snout; caudal subtruncate; D. 10. .5-1 1.5; A. 7.5 or 8.5.

58. alternatum Eigenmann.

^^ Not examined in P. minutum, nigricnns, goehlii, and pundatissimum.
" Not examined in P. goehlii.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 329

jj. With ill-defined spots; origin of ventrals equidistant from snout and tip
of caudal; caudal rounded; distance between origin of dorsal and
caudal twice in its distance from the snout; D. 10; A. 7.

59. goeldii (Boulenger).

jjj. With small spots and vermiculations; origin of ventrals equidistant from

snout and tip of caudal; distance between origin of dorsal and caudal

1.66-1.75 in distance of dorsal from the snout; D. 10.5 or 11.5; A. 7.5

or 8.5 60. brasiliense (Reinhardt).

jjjj. A dark lateral band, spots above and below it; origin of ventrals a little
nearer snout than to tip of caudal ; distance between origin of dorsal and
caudal 1.8 in distance of dorsal from the snout; caudal truncate.,

Gl. itatiayae (Ribeiro).
hh. Origin of the anal behind the vertical from the last dorsal ray, the dorsal opposite
the space between the ventral and anal, its origin equidistant from snout and
tip of caudal, or nearer one or the other; origin of ventrals nearer the snout than
to the tip of the caudal; interopercle with about seven spines in a bunch similar
to that of the opercle; a series of dots along the middle of the sides, another along

the back and another between the two 62. triguttatum Eigenmann.

ccc. Origin of dorsal equidistant from tips of snout and caudal; pectorals without filaments; dorsal
entirely in front of the anal.
k. Sides and back with minute dark specks and vermiculations between them; origin of dorsal

over origin of ventrals . . . '. 63. punctatissimum (Castelnau).

kk. Pale brown above with three longitudinal series of squarish brown blotches; origin of dorsal

over middle of ventrals; head 4.5 in the length 64. minutum (Boulenger).

cccc. Origin of dorsal nearer snout than to tip of caudal.

I. Nasal barbels not quite reaching eye, maxillary barbel little beyond posterior margin of the

eye; dorsal entirely in front of anal; sides with large blotches. Head four times in the
length; a long patch of interopercular spines; pectoral ray not prolonged.

65. santae-ritae Eigenmann.

II. Nasal barbels reaching beyond origin of eye; pectoral ray prolonged.

m. Interopercle with a long patch of spines; pectoral ray prolonged. {See P. altcrnatum,
No. 56.)
mm. Interopercle with about seven thorns in a patch similar to that of the opercle; end of
dorsal over anus; pectoral ray much prolonged. (See P. triguttatum, No. 60.)

49. Pygidium nigricans (Cuvier & Valenciennes.)
Trichomyderus nigricans Cuvier & Valenciennes, Hist. Nat Poiss., XVIII,
1846, p. 494; ? Gay, Hist. Chile, 1848, p. 311 (Chile); Gunther, Cat. Fishes
Brit. Mus., V, 1864, p. 274 (copied); Ribeiro, Fauna Braziliense, Peixes,
IV (A), 1912, p. 220.
Pygidium nigricans Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889,
p. 53; Occasional Papers Cal. Acad. Sci., I, 1890, p. 338; Proc. U. S. Nat. Mus.,
XIV, 1891, p. 37; Eigenmann, Reports Princeton Univ. Exped. Patagonia,
III, 1910, p. 400.

330 MEMOIRS OF THE CARNEGIE MUSEUM.

Habitat. — Santa Cathcrina, Brazil.

Valencienncs's description of the only specimen known, 140 mm. long, is very
brief, and I am afraid that it is in part misleading.

"D. 11; A. 10; P. 9; barbels short, scarcely reaching beyond the eyes; caudal
peduncle short and deep; caudal small, truncate, 'le premiere rayon de la dorsal
alonge en fil.'

"Back uniform blackish, lower parts light."

50. Pygidium iheringi Eigenmann. (Plate L, fig. 5.)
Pygidium iheringi Eigenmann, Proc. Am. Philos. Hoc, LVI, Jan., 1918, p. 697.
Trichomyctenis punctulatus {non Cuvier & Valenciennes) Ribeiro, Arkiv. for

Zoologie, IV, No. 19, 1908 (Iporanga).
Trichomycterus dispar {non Tschudy) Ribeiro, Kosmos, V, 1908, and Fauna Brasi-

liense, IV (A), 1912, p. 222 (Rio Iporanga, Sao Paulo).

Habitat. — Sao Paulo in coastal streams and the Parana basin.
7071, C. M., two, 151-160 mm. Sapina, Sao Paulo. July 3, 1908. Haseman.
10785, I. U. M., four, 140-161 mm. Santos. Von Ihering, the largest the type.

Allied to P. punctatissifnvm. from the Aragiiay.

Head 4.5-5 in the length; D. 11.5 or 12.5; A. 7.5 or 8.5 counting the two rudi-
mentary rays in each case ; P. 8 ; width of head equal to its length behind the nasal
barbel; eye in middle of the head, interorbital 3.5-4 in the length of the head.
Teeth incisors with slightly expanded tips, in bands of four or five series.

Nasal barbels reaching about to middle of eye, axillary barbel to above middle
of opcrcle; pectoral rounded, very little longer than snout and eye, the first ray not
prolonged or with only a trace of a projection; distance between origin of ventrals
and eye a little greater or less than that between origins of ventrals and middle
caudal rays; the ventrals as long as the snout, not nearlj' reaching vent, nearly
halfway to the anal; origin of anal on, or behind, the vertical from the base of the
last dorsal ray; distance between bases of last anal ray and middle caudal rays five
or a little over five in the length; caudal slightly rounded, seven to seven and a
half in the length; dorsal low and long, the distance between its origin and the base
of the middle caudal ray about one and a third in its distance from the snout, its
first ray over posterior half of the ventrals.

Sides and back with numerous s]iots, smallest over pectorals, largest over
dorsal, rarely coalescent.

51. Pygidium zonatum sp. nov. (Plate LI, fig. 1.)
7596, C. M., a, the type, 62 mm., /; and c, paratypes, 50 and 55 mm. Agua Quente.
Nov. 27. 1908. Haseman.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 331

7595, C. M., paratype, GO mm. Cubatao, seven miles west of Santos, Sao Paulo,
Brazil. Aug. 1, 1908. Haseman.
In all but the teeth and color very similar to P. davisi.

Head 5; depth 6-6.5; D. 10; A. 7 or 8; P. 7 or 8; eye three times in the snout,
seven to seven and one-half times in the head, two times in the interocular, exactly
in the middle of the head or a Uttle in front of it; posterior nares more than an or-
bital diameter from the eye; three rows of narrow incisors in each jaw. The teeth
of the paratypcs nearly conical.

Maxillary barbel reaching to near tip of last interopercular spine or to tip of
opercular spines, the nasal barbel but little shorter; head pointed, but little longer
than broad; gill-openings extending forward to below the anterior interopercular
spines. First pectoral ray but slightly, if at all, prolonged, its length with the
filament equal to the head without the opercular spines; origin of ventrals about
equidistant from the snout and the middle of the caudal; origin of anal under
penultimate, or fourth from last, dorsal ray; distance between bases of last anal ray
and middle caudal rays four and one-half to five in the length; caudal truncate;
origin of dorsal over last third of the ventrals, its distance from the base of the mid-
dle caudal rays 1.5-1.75 in its distance from the snout.

Color-markings coarse and conspicuous. In No. 7595 C. M., five obscure
bars across the back in front of the dorsal, slightly emphasized at their lower ends
on the middle of the sides; three similar bars behind the dorsal; a dark line from
anterior to posterior nares. In the specimens from Agua Quente a dark lateral
band broken toward the caudal in the smallest. Back and lower part of the sides
with conspicuous spots.

52. Pygidium proops (Ribeiro). (Plate LI, fig. 2.)
Tricomyderus proops Ribeiro, Kosmos, V, 1908, fig. 4; Fauna Brasiliense, Peixes,
IVU), 1912, p. 221, pi. XL, fig. 1.

Habitat. —Riheira de Iguape, southern Sao Paulo, Brazil, and Rio Parahyba.
Known from the types, and from
7593, C. M., one, 60 mm. Agua Quente, Ribeira Basin. Nov 27, 1908. Hase-
man.
7598, C. M., one 32 mm. Sao Joao da Barra, Rio Parahyba. June 22, 1908.

Haseman.

Head 5.5-6; depth 7; D. 9; A. 7-9; P. 7; posterior margin of eye shghtly in
advance of the middle of the head, diameter of eye two times in the snout, six and
one-half in the head, over one and one-half in the interorbital ; posterior nares close
to the eye, their posterior margin on a line with the anterior margin of the eye;

332 MEMOIRS OF THE CARNEGIE MUSEUM.

teeth conical, in very narrow bands; nasal barbels reaching to the base, maxillary
barbels to the tips of the opercular spines; first pectoral ray not prolonged, pectoral
equal to the length of the head without the snout; origin of ventrals equidistant

Fir.. 16. Pygidium -proiips (Ribiero). No. 7593 CM.

from tip of snout and tip of caudal; origin of anal under second or third dorsal ray;
distance between last ray of anal and base of caudal five and one-third times in
the length; caudal rounded; distance between origin of dorsal and caudal two and
one-half times in its distance from the snout.

Back, sides, dorsal and caudal densely spotted.

The above description is based on the specimen from Agua Quente and Ribeiro's
account ; the smaller specimen from the Parahyba has :

Head 5; depth 8; D. 10; A. 9; P. 5; first pectoral ray considerably in-olonged;
origin of ventrals equidistant from tip of caudal and pre-opercle. Sides and back
finely marbled, a faint dusky, lateral hne, and another above it. Its prolonged
first pectoral ray and narrow pectoral, with but five rays indicate a distinct variety,
which may be called parahybce, var. nov.

53. Pygidium paolence Eigenmann. (Plate LI, fig. 3.)-^
Pygidium paolence Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. 698.

Very close to Trichomycterus proops Ribeiro.
7081, C. M., type, 68 mm. Alto da Serra, Rio Tiete, Sao Paulo. July 25, 1909.

Haseman.
7597, C. M., paratype, 61 mm. Rio Paranahyba bridge. Aug. 15, 1908. Hase-
man.

Head 5.33-6; D. 8.5; A. 6.5-8 not counting hidden rudiments, D. 10.5 and
"A. 8.5 with the rudiments; P. 6 or 7; head nearly as wide as long; eye in anterior
half of the head, greater than its distance from the posterior nares; snout 2.33-2.5
in the length of the head, interocular 3-3.5; teeth conic; nasal barbel reaching base
or tip of opercular spines, maxillary barbel reaching tip of opercular spines or a
little farther; outer pectoral ray with its filament equal to head behind the posterior

^' The head is too short in this drawing.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 333

nares, the filament extending very little beyond the other rays; ventrals nearly
reaching anal, their origin nearer caudal than to tip of pectorals; caudal rounded,
six in the length; origin of anal under middle of dorsal, distance between the base
of its last ray and the middle caudal ray 5.2 in the length ; origin of dorsal equidis-
tant from base of middle caudal rays and middle of pectorals, its last ray over the
middle of the anal, the distance between the origin of the dorsal and the base of the
middle caudal rays two in the distance between dorsal and snout in the type, 1.66
in the paratype.

With many faint spots about as large as the eye ; in the type a dark streak along
the middle of the sides, another along the side of the back, and a third along the
edge of the belly.

This species is similar in appearance to P. striatum, from which it differs in the
position of the ventrals, the pectoral filaments, etc.
7n7a-j, C. M., ten, 25-30 mm. Mogy das Cruces, Rio Tietc, July 20, 1908.

These minute specimens came from near the type locality of P. paolence and
are probably the young; the ventrals do not reach the vent; there is a series of
minute spots along the sides, nearly confluent anteriorly; a series of larger spots
above it on the sides of the back and a series along the middle of the back.

54. Pygidium reinhardti Eigenmann. (Plate LI, fig. 4.)
Pygidiuni reinhardti Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. 699.
7078, C. M., one, 65 mm. Burmier on the Rio Itabira, a tributary of the Rio das

Velhas. May 14, 1908. Haseman.

Mr. Haseman notes that this is the only species he secured at the particular
locality where the only specimen of the present species was collected.

Head 6.5; D. 9.5; A. 8.5, counting the minute rudimentary rays in both dorsal
and anal; P. 6; eye in anterior half of head; interocular three times in the head.
Teeth conic.

Nasal barbel nearly as long as the maxillary barbel, which reaches the edge of
the gill-membrane. First pectoral ray with its filament equal to the length of the
head, much longer than the divided rays; ventrals reaching beyond the' vent, their
origin very little nearer tip of pectorals than base of middle caudal rays; origin of.
anal under middle of dorsal ; distance between the base of the last anal ray and the
middle caudal rays five and a half in the length; caudal narrow, a little longer than
the head, the accessory rays inconspicuous; origin of dorsal over middle of ventrals,
its distance from the middle caudal rays nearly two in its distance from the snout
(19 and 36 mm. respectively).

334 MEMOIRS OF THE CARNEGIE MUSEUM.

A broad, dark stripe with notched edges from opercle to middle of caudal,
bordered above and below by light bands; an irregular series of spots below the
lower light band; a series of small spots more or less confluent forming a narrow,
dark stripe above the upper light band; back and fins lightly spotted, a short dark
bar in front of the opercle, a longer one above middle of pre-opercle.

55. Pygidium davisi Haseman. (Plate LI, fig. 5.)
Pygidium davisi Haseman", Aim. Carnegie Mus., VII, 1911, p. 3S0, pi. LXXVII,

fig. 1, and pi. LXXVII.

Habitat. — Coastal streams of southern Sao Paulo.
2862, and2 861, C. M., type and paratypes. Rio Iguassu near Serrinha Parana.

. Dec. 23, 1908. Haseman.
7116a-d, C. M., 52-60 mm. Morretes, Parana, Brazil. Jan. 4, 1909. Haseman.

Head 5.5; D. 7; A. 5 without the imbedded rays; D. 10; A. 9 with the imbedded
rays; P. 8; eye slightly in advance of the middle of the head; interocular four times
in the length of the head; teeth conic.

Nasal barbel as long as the labial barbel, extending to the base of the opercular
spines, maxillary barbel to their tips; pectorals shorter than the head, the first ray
not prolonged ; origin of ventrals equidistant from tip of snout and tip of caudal or
nearer the former, not reaching the vent; origin of anal under the posterior part
of the dorsal, the distance between its last ray and the base of the middle caudal
rays about five in the length; caudal subtruncate, five and one-half in the length;
origin of the dorsal over last half of the ventrals, equidistant between the tip of the
caudal and the nasal barbels or eye ; the distance between it and the caudal one and
six-tenths in its distance from the snout.

Uniform light or dark, or mottled, sometimes with a dark lateral band with
one or more series of blotches above it.

56. Pygidium immaculatum Eigenmann & Eigenmann. (Plate LII, fig. 1.)-"
Pygidium immaculatum Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II,

1889 CJuiz de Fora; Sao Matheos; Goyaz); Occasional Papers Cal. Acad.

Sci., I, 1890, p. 337; Eigenmann, Report Princeton Univ. Exped. Patagonia,

III, 1910, p. 400.
Trichomycterus immaculatus Ribeiro, Fauna Brasiliense, IV (A), 1912, p. 222.

Habitat. — Rios Parahyba and Doce. Goyaz.
7076a-6, C. M., two, 81-93 mm. Rio Doce. May 24 and 25, 1908. Haseman.

" The head is a little too sliort in the figure.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 335

Head 4.75-5; D. 11.5-13; A. 9.5-10.5; P. 8 or 9; eye in the middle of the head
or very nearly so, interocular 4-4.5 in the head. Teeth conic.

Nasal barbels extending to the middle of the interopercle, maxillary barbel
about to opercular spines; pectorals but little longer than snout and eye, the first
ray with its filament equal to head behind the nasal barbel ; ventrals reaching vent,
their origin very little nearer caudal than to eye; origin of anal just behind the
vertical from the base of the last dorsal ray or under the penultimate ray, distance
between the base of its last ray and the base of the middle caudal ray four and a
half to five times in the length; caudal very slightly emarginate, about seven times
in the length; origin of dorsal over posterior part of the ventral, its distance from
the base of the middle caudal rays one and two-thirds tinies in its distance from the
snout.

Sides and back uniform, without trace of spots or vermiculations ; middle caudal
rays dusky.

57. Pygidium vermiculatum Eigenmann. (Plate LII, fig. 2.)
Pygidimn vermiculatum Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918,

p. 699.
Pygidium brasiliense {non Liitken) Ribeiro (partim), Fauna Brasiliense, IV (A),

1912, p. 225 (the specimens from Juiz de Fora).

Habitat. — Rio Parahyba.
7074, C. M., one, 131 mm. Juiz de Fora. June 9, 1908, presented by Dr. Ribeiro.

In general appearance like Liitken 's figure of hrasiliense, differing notably in
the position of the ventrals.

Head 5.4 in the length; D. 8.5; A. 8.5 (counting in each case the two rudi-
mentary rays) ; P. 7 ; width of the head nearly equal to its length ; eye in middle of
the head, interorbital three in the length of the head. Teeth conic, in bands.

Right nasal barbels reaching to above base of the opercular spines, maxillary
barbels of right side nearly as long as head, reaching to the second fourth of the
pectoral, both shorter on left side ; pectoral rather narrow, the outer ray much pro-
longed, as long as the head behind the nasal barbel, the fin without the filament
equal to the part of the head behind a point midway between eye and posterior
nares; origin of ventrals under origin of dorsal, equidistant between base of middle
caudal rays and last third of pectorals, ventrals reaching much beyond vent,
almost to anal, as long as the snout; origin of anal under penultimate ray of the
dorsal, distance between the base of its last ray and the base of the middle caudal
ray a little more than five in the length; caudal rounded, six and one-third in the
length; dorsal short, rounded, the distance between its origin and the base of the

336 MEMOIRS OF THE CARNEGIE MUSEUM.

middle caudal rays one and sixty-seven hundredths in the distance between its
origin and the snout.

Sides and back profusely covered with confluent spots, which leave the light
color as irregular vermiculations.

58. Pygidium alternatum Eigenmann. (Plate LII, fig. 3.)
Pygidium alternatum Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. 700.
Pygidium brasiliensis Eigenmann & Eigenmann (partim), Proc. Cal. Acad. Sci.

(2), II, 1889, p. 51; id. {partim), Occasional Papers Cal. Acad. Sci., I, 1890, p.

332; Ribeiro (partim), Fauna Brasiliensis, IV (A), 1912, p. 223.

Habitat. — Rio Doce.

It is possible that the young specimens of P. brasiliensis mentioned by Eigen-
mann & Eigenmann belong to this species.

7079, C. M., type and paratypes, sixty-seven, largest 81 mm. Rio Doce, May 25,
1908. Haseman.

7080, C. M., eleven, largest 49 mm. Rio Doce, May 25, 1908. Haseman.
7601rt, C. M., 26 mm. Jacarehy, July 15, 1908. Haseman.

Head 5-5.5; D. 10.5-11.5; A. 7.5 or 8.5 counting the rudimentary rays ; P. 7 or
8; eye in middle of the head or slightly farther forward; interocular 3-3.33 in the
length of the head. Teeth conic, in bands.

Nasal barbel very httle shorter than maxillary barbel, which reaches to the
base of the pectoral and is as long as the head; pectoral rays equal to length of head
behind the nasal barbels, the first ray with the filament longer than the head;
ventrals reaching to, or just beyond, vent; origin of ventrals equidistant from base
of middle caudal rays and a point between the posterior nares and the area just
behind the eyes; origin of anal under posterior part of dorsal; distance between base
of last anal ray and middle caudal rays four and a half to five and a third in the
length; caudal subtruncate or rounded, very little longer than head; origin of dorsal
over posterior half of ventrals; distance between origin of dorsal and base of middle
caudal rays 1.54 in its distance from the snout.

Ten to fourteen large spots along the middle of the sides, an irregular series
of much smaller ones below it. Large spots above the median series, frequently
alternating with it, sometimes partly confluent into a longitudinal series, some-
times forming with a mid-dorsal series irregular bars across the back.

As the specimens. No. 7079, ranging up to 81 mm., are essentially alike in color
and entirely different from the specimens of both smaller and larger size (74-120
mm.) from the same place referred to P. brasiliense. No. 7075, they have been

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 337

separated from the latter species. They may be identical with some of the younger
specimens of P. brasiliense mentioned by Eigenmann & Eigenmann in the papers
quoted above.

The specimens, No. 7080, are much shriveled but probably belong to this

species.

59. Pygidium goeldii (Boulengcr).

Trichomycterus goeldii Boulenger, Ann. Mag. Nat. Hist. (6), XVIII, 1896, p. 154.

i/a6ita/..— Colonia Alpina, Therezopolis, nearly 2,600 ft., near Rio de Janeiro.

I give the original description of Boulenger, I. c. :

"Head much depressed, as long as broad, six times in total length; eye small,
midway between end of snout and opercular border, its diameter half interorbital
width; upper maxillary barbel reaching the pectoral; gill-membranes narrowly
joined to the isthmus, extending forward to below the eyes. Body as deep as
broad; caudal peduncle strongly compressed, twice as long as deep. Dorsal with
10 rays, opposite to the space between ventrals and anal, twice as distant from the
end of the snout as from the caudal; anal with 7 rays. Pectorals with the outer
ray produced, filiform. Ventrals equally distant from the end of the snout and
the posterior border of the caudal fin; latter rounded. Yellowish, with ill-defined
brown spots above. Total length 99 millim."

60. Pygidium brasiliense (Reinhardt).

Trichomyderm brasiliensis Reinhardt MS. in Liitken, Overs. Dansk. Vidensk.
Selsk., 1879, No. 3, p. 29 (Rio das Velhas); LUtken, Velhas Flodens Fiske,
p. 15, in Vidensk. Selsk. Skr. (5), Afd., XII, 1875, pp. 135 and I, pi. Ill, fig. 8
(Rio das Velhas); ? Boulenger, Proc. Zool. Soc. London, 1891, p. 235 (Rio
Grande do Sul) ; Ribeiro, Kosmos, 1908 (Ribeiro) ; Archiv. Mus. Nac. Rio de
Janeiro, XIII, 1906 (p. 7 of reprint); Fauna Brasiliense, Peixes, IV (A), 1912,
p. 223, Arch. Mus. Nac. Rio de Janeiro, XVI.

Pygidium brasiliensis Eigenmann & Eigenmann (partim), Proc. Cal. Acad. Sci.
(2), II, 1889, p. 51; Occasional Papers Cal. Acad. Sci., I, 1890, p. 332; Proc.
U. S. Nat. Mus., XIV, 1891, p. 36; Eigenmann, Reports Princeton Univ.
Exped. Patagonia, III, 1910, p. 399.

Trichomycterus brasiliensis tristis LiJTKEN, I. c, p. 138, with figure, and p. I.
■ Habitat.—Rios das Velhas and Doce south to Ribeira do Iguape and ? Rio

Grande do Sul.

7594, C. M., four, 71-87 mm. Rio das Velhas. May 13, 1908. Haseman.

7550, C. M., one, 133 mm. Burmier. May 14, 1908. Haseman.

7075, C. M., fourteen, 74-120 mm. Rio Doce. May 25 and 27, 1908. Haseman.

338

MEMOIRS OF THj5 CARNEGIE MUSEUM.

Head 4.75-5 in the length; D. 10.5 or 11.5; A. 7.5 or 8.5 counting tlie rudi-
mentary rays; P. 7; width of head very nearly equal to its length; eye in the middle
of the head or partly in the anterior half; interocular 3-3.33 in the length of the
head. Teeth conic.

Fig. 17. Pygidium brasiliense (Reinhardt). After Liitken.

Nasal barbels e.xtending to a i)oint above the end of the interopercle, maxillary
barbel to the gill -opening, or very little shorter or longer; pectoral a little longer
than snout and eye, the first ray being very little prolonged in the smallest, equal
to the length of the head less the opercle in the largest ; origin of ventrals equidistant
from base of middle caudal rays and base of pectoral in the specimen from Burmier,
or to a point between the eyes and the opercle in the others, reaching very little be-
yond the vent, about two-thirds to anal, or to the anal in the specimen from Burmier ;
origin of anal below posterior half of dorsal, the distance between the base of its
last ray and the base of the middle caudal rays four and two-thirds to five and a
quarter in the length; caudal rounded, five and a half to six and a half in the length;
origin of dorsal over posterior half or end of ventrals, the distance between its
origin and the base of the middle caudal raj's one and two-thirds to one and three-
fourths times in the distance between its origin and the snout.

Fig. 18. Pygidium brasiliense (Reinhardt). Opercle and P. brasiliunse triste (Liitken).

The color in the specimens from the Rio Doce is the same. Back and sides
with numerous spots and vermiculations, the spots forming an irregular dark line
along the middle of the sides in front. The spots and vermiculations are a little
finer than in Llitken's figure of brasiliensis. In the specimens from Burmier and
the Rio das Velhas there is a distinct median lateral band, the markings below it
being coarse. In the largest the lateral band becomes obscure.

EIGENMANN: the PYGIDIIOiE, A FAMILY OF SOUTH AMERICAN CATPISHES. 339

61. Pygidium itatiayae (Ribeiro).
Trichomyderus brasiliensis itatiayce Ribeiro, Archives Mus. Nac. de Rio de Janeiro,

XIII, 1906, p. 8, pi. I; Fauna Bras., IV (A), 1912, p. 223.

Habitat. — Itatiaya, Serra da Mantiqueira.

Caudal subtruncate; head longer than broad; last ray of the dorsal over the
fourth of the anal; a dark lateral band.

Fig. 19. Pijyidium italiayw (Ribeiro), adult and young. After Ribeiro.

62. Pygidium triguttatum spec. nov. (Plate LII, fig. 4.)
7600, C. M., a the type 36 mm., b to e, paratypes 26-34 mm. Jacarehy. July

14 and 15, 1908. J. D. Haseman.

Readily distinguished by the few spines in the interopercle.

Head 5-5.5; D. 8 or 9; A. 6 or 7.5; P. 6; eye in anterior half of the head, 2 in
the snout, 6 in the head, about 1.5 in the interorbital ; teeth pointed, in very narrow
bands; gill-openings reaching forward to below the eye; nasal barbels reaching to
tip of opercular spines, or but little beyond the eye; maxillary barbels to the base
of the opercular spine or to the axil; pectorals lanceolate, the first ray much pro-
longed, one and a third times as long as the head in the type, longer than the head
in all but one of the paratypes; origin of ventrals equidistant from snout and middle
of caudal; tips of ventrals reaching anus in two of the specimens, falling consider-
ably short of the anus in the rest; origin of anal behind the dorsal; distance between
last anal ray and caudal 5-5.5 in the length; caudal rounded, but few inconspicuous

340 MEMOIRS OF THE CARNEGIE MUSEUM.

accessory rays; distance from origin of dorsal to base of middle caudal rays 1.2-1.4
in the distance between snout and dorsal; distance from origin of dorsal to tip of
caudal sometimes less, sometimes greater, than its distance from the snout.

A row of small spots along the middle of the sides, another along the middle
of the back and a third between the two.

63. Pygidium punctatissimum (Castelnau). (Plate XLV, fig. 1.)
Trichomycterus punctatissimus Castelnau, • Anim. Nouv. Am. Sud., 1855, p. 49,

pi. 24, fig. 3, GtJNTHER, Cat. Fishes Brit. Mus., V, 1864, p. 272; Ribeiro,

Fauna Braziliense, Peixes, IV (A), 1912, p. 221.
Pygidium punctatissimum Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2),

II, 1889, p. 52; Occasional Papers Cal. Acad. Sci., I, 1890, p. 334; Proc. U. S.

Nat. Mus., XIV, 1891, p. 36; Eigenmann, Reports Princeton Univ. Exped.

Patagonia, III, 1910, p. 399.

Habitat. — Rio Araguay.

Known from the types, with the characters given in the Key.

64. Pygidium minutum (Boulenger).
Trichoynycterus minutus Boulenger, Proc. Zool. Soc. Lond., 1891, p. 235, pi.

XXVI, fig. 3.
Pygidium minutu?n Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,

1910, p. 399.

Habitat. — San Lorenzo district,, southern Rio Grande do Sul.

Fig. 20. Pygidium minutum (Boulenger). After Boulenger.

Head 4.5; D. 8; A. 6; eye a httle in advance of the middle of the head, 1.5 in the
interorbital ; gill-opening not continued forward to below the eye ; maxillary barbels
three-fifths the length of the head, not reaching the gill-opening; nasal barbels
extending to the eye; dorsal entirely in front of the anal; origin of dorsal midway
between snout and tip of caudal; caudal rounded. Pale brown above, with three
longitudinal series of large squarish brown blotches. Fins immaculate, the largest
40 mm.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 341

65. Pygidium santae-ritae sp. nov. (Plate LII, fig. 5.)
7599, C. M., type, 24 mm. Santa Rita, Rio Preto. July 10, 1908. Haseman.

Head 4; D.ll; A. about 9; P. 8; eye very nearly in the middle of the head;
about two in the snout, five in the head, a little less than interorbital; teeth pointed,
in a single series or in a very narrow band; gill-openings extending to below the
middle of the eye; nasal barbel extending but little beyond posterior nares; maxil-
lary bai-bels to middle of interopercle ; outer pectoral ray not prolonged, a little less
than head; origin of ventrals equidistant from snout and middle of caudal; anal
entirely behind the dorsal, distance between its last ray and the caudal 4.5 in the
length; caudal rounded, 5 in the length; distance between origin of dorsal and caudal
equal to distance between dorsal and the middle of the eye.

Sides with large spots.

In the length of the barbels and the color this species agrees very closely with
P. minutum from southern Rio Grande do Sul, with which it may be synonymous.
It differs in its longer head, the forward extent of the gill-opening, the more anterior
position of the dorsal. While the number of fin-rays as given differs, not much
weight attaches to this. I have endeavored to count all of the rudiments.

Genus V. Eremophilus^o Humboldt. (Plate XXXVI; Plate XL, figs. A, B.)

Eremophilus sive Thrichomycterus Humboldt, Rec. d'Obs. Zool. et Anat., I,
1805, p. 17, pi. 6, reprinted in 1912, title-page 1911.
Trachypoma Giebel, Zeitschr. Gesellsch. Naturw., Ill, 1871, p. 97 (type marmora-

tum = mutisii).

Type. — Eremophilus mutisii Humboldt.

Like Pygidium, but without ventrals.

1. Eremophilus mutisii Humboldt. (PI. XLI, figs. 1, 2; PL LIV, figs. 1, 2.)
Eremophilus mutisii Humboldt, I. c, I, 1805, p. 17, pi. 6; Valencien-nes, in Hum-
boldt, II, 1835, p. 340; Cuvier & Valenciennes, Hist. Nat. Poiss., XV, 1846,
p. 500, pi. 553 (Bogota); Gijnther, Cat. Fishes Brit. Mus., V, 1864, p. 275;
EiGENMANN & EiGENMANN, Proc. Cal. Acad. Sci. (2), II, 1889, p. 53; Occa-
sional Papers Cal. Acad. Sci., I, 1890, p. 339; Proc. U. S. Nat. Mus., XIV,

'° ipriiJio4>i\v^, o = loving solitude (of the mountain lakes and streams).

" Je I'ai aomme eremophile, a cause de la solitude dans laquelle il vit a de si grandes hauteurs, et
dans des eaux qui ne sont presque habitees par aucun autre etre vivant. Les naturalists qui craignent que
de nouvelles especes de ce meme genre ne viennent a etre d^couvertes dans des situations tres-differentes,
pourroient changer le nom d' eremophile en celui de ihrichomijcterus, tir^ des barbiUons attaches au nez de
ce poisson."

I

342 MEMOIRS OF THE CARNEGIE MUSEUM.

1891, p. 37; EiGENMANN, Reports Princeton Univ. Expcd. Patagonia, III,

1910, p. 400; Indiana University Studies, No. 23, Sept., 1914, p. 230.
Trachijpoma marmoratum Giebel, /. c.

Habitat. — Plain of Bogota and north of it for a short distance.
5046a-p, C. M.; 12840, I. U. M., many, largest 325 mm. Ponta de Suba, north of

Chapinero. Eigcnmann.
5049, C. M., two, largest 300 mm. Laguna near Bogota (bought in the market).

Eigenmann.
5048a-A:, C. M.; 12841, I. U. M., twenty-two, largest 210. Herrera. Eigenmann.
5047a-/t, C. M.; 12842, I. U. M., sixteen, largest 270 mm. Madrid. Eigenmann.
7445a, C. M.; 13834, I. U. M., two, 142 and 165 mm. Rio Chiquiuquiere, Boyaca.

Gonzales.
7446a, C. M., one, 220 mm. Rio Bogota. Gonzales.
13836, I. U. M., three, 130-160 mm. Rio Funjuelo at Usme Sur near Bogota.

Head 5.33-6; depth 5-6.33; D. 11.5; A. 9.5; P. 8; head pointed, a Httle longer
than wide ; center of the eye very little in advance of the center of the head ; inter-
ocular three times, or a little more, in the head; teeth conical, in three to five rows;
gill-openings not extending forward to below the eye; a very narrow free mem-
brane across the isthmus.

Nasal barbel extending to the base of the opercular spines or shorter; maxillary
barbel extending very little if any further than the nasal barbels; pectoral 1.5-2
in the head, its first ray sometimes very shghtly produced, origin of anal about
under the middle of the dorsal, the distance between the last ray and the middle
caudal rays 4.5-5.5 in the length; depth of caudal peduncle 1.25-1.5 in its length,
.66-. 8 in the greater depth; caudal very broad and short, its length seven or more
in the length; origin of dorsal nearly equidistant from tip of caudal and the head,
its distance from the base of the middle caudal rays 1.8-2 in its distance from the
snout.

Blackish, everywhere with well-defined but irregular spots or vermiculations.
The black background most abundant above, the light vermiculations predominant
below. In some specimens the dark predominates everywhere, the light being
reduced to spots, or vermiculations, in others the fight predominates; in the young
there is a narrow dark median stripe, and the dark of the caudal peduncle consists
of a few irregular spots, on the back in front of the dorsal the typical color of the
adult obtains. Very variable.

The specimens from the Rio Funjuelo deserve special mention.

The largest measures a few millimeters over 160; it is not possible to give

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 343

the exact length, owing to the curves. This specimen is without pigment. The
eye is apparent only on account of the lens and its overlying hyaline skin. The
eye measures 2 mm. in diameter. This measurement is taken with the skin re-
moved. The eye is not pigmented.

Another specimen measures 130 mm. It is also without pigment except in the
eye. The eye seems to be fully pigmented and measures about 2 mm.

The third specimen measures 133 mm. The caudal and all but a small patch
on the dorsal surface of the caudal peduncle are without pigment. The region
from the caudal peduncle to the head is pigmented, but much more sparingly than
in normal specimens, and there are irregular pigment-free spots. The sides of the
head behind the eyes are free from pigment, the dorsal surface of the head and
snout are again pigmented. The eye is normally pigmented and measures a little
less than 2 mm.

These specimens seem to be identical with the normal "Capitan" in all respects
except the color.

As stated elsewhere, the "Capitan" is a fish of considerable importance on the
plains of Bogota. The fish are caught with long-handled dip-nets several feet in
diameter. The net is held in a slanting position by one man, one or several others
drive the fishes into the net by beating the water. From time to time the net is
raised and the fish removed. We supplemented this method by having the Indians
drag a small seine, which yielded many smaller specimens. All we caught were
about 80 mm. long or longer, i. e., we caught no very young ones. The Indians
also secured specimens by thrusting their hands and arms into the holes in the banks.
It has heretofore been recorded exclusively from the Plains of Bogota. Mr. Gon-
zales has sent two specimens from the Rio Chiquinquiere north of the plain.

Genus VI. Pareiodon" Kner. (Plate XXXVII.)
Pareiodon Kner, Sb. Ak. Wiss. Wien, XVII, 1855, p. 160.
Centrophorus Kner, Denkschr. Akad. Wiss. Wien, 1859, XVII, p. 167.
Astcmomycterus Guichenot, Rev. et Mag. Nat. Hist., XII, 1860, p. 525, fig. 2

(pusillus) .
Pariodon GtJNTHER, Cat. Fishes Brit. Mus., V, 1864, p. 275.

Type. — Pareiodon microps Kner.

No mental barbels, no nasal barbel, two slender barbels at angle of mouth;
outer pectoral ray not prolonged beyond the other rays; gill-membrane confluent
with the isthmus, without a free fold in the middle, a narrow fold just below the

^' Tvapda = cheek, oSoiis, 6 = tooth.

344

MEMOIRS OF THE CARNEGIE MUSEUM.

gill-opening which is restricted to the space behind the interopercular spines, the
membrane being confluent with the shoulder behind the opercular spines; mouth

subterminal, a single scries of teeth in each jaw, about
sixteen on the pre-opercle and about thirteen on the
mandible. The teeth of the two jaws alike, similar to
the teeth at the end of the premaxillary of Branchioica and
some species of Vandellia. They consist of a narrow
basal section, an enlarged middle section, from the inner
(median) angle of which projects a short spur laterad, the
spur being slightly twisted from the plane of the rest of
the tooth; the teeth not movable; interopercle with a
single series of six backward directed, slightly divergent
spines; opercular spines slightly divergent, four spines
in the main posterior row; five much smaller ones in
the anterior row; caudal forked.

Pareiodon has the appearance of an overgrown Van-
dellia, from which it differs in dentition, in the size of the
eye, and the general shape of the head.

Fig. 21. Pareiodon microps
Kner. a, outline of head from
above; b, the interopercular
spines; c, the opercular spines;
d, a premaxillary and a man-
dibular tooth, very much en-
larged.

1. Pareiodon microps Kner. (Plate LIV, fig. 3.)

Pareiodon microps Kner, I. c. (Borba on the Madeira about four days from its
mouth); Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889,
p. 55; Occasional Papers Cal. Acad. Sci., I, 1890, p. 346; Proc. U. S. Nat. Mus.,
XIV, 1891, p. 37; Ribeiro, Fauna Bras., IV {A), 1912, p. 232; Comm. Linhas
Telegraphicas Estrategicas de Matto-Grosso ao Amazonas. Annexo 5, Sept.,
1912, p. 30 (Manaos) ; Fowler, Proc. Acad. Nat. Sci. Phila., 1915, 229 (Amazon
between mouth of Rio Negro and Peru).

Pariodon microps Gunther, Cat. Fishes Brit. Mus., V, 1864, p. 275; Cope, Proc.
Acad. Nat. Sci. Phila., 1872, 290 (between Rio Negro and Peru.)

Trichomyctenis pusillus Castelnau, Anim. Amer. du Sud, Poissons, 1855, p. 50,
pi. 24, fig. 4 (Araguay; Amazon).

Astemomycterus pusillus Guichenot, I. c. (Araguay; Amazon).

Pareiodon pusillus Ribeiro, I. c, p. 234.
Habitat. — Amazon Basin.
The specimen in the collection of the Philadelphia Academy of Sciences,

mentioned by Cope and Fowler, is about 145 mm. long. I am indebted to Dr.

Fowler for the loan of this specimen.

EIGENMANN: the PYGIDIID^, a family of .south AMERICAN CATFISHES. 345

Head 8; depth 6; D. 10, of which eight are full length; A. 7, of which five are
full length; eye minute, about thirteen times in the head; center of head a full
orbital diameter behind the posterior margin of the eye; interorbital 2.1 in the
length of the head, snout 3 in the head, broad; width of head equal to its length;
maxillary barbel not reaching to the interopercular spines; lower barbel reaching a
little beyond the middle of the maxillary barbel; depth of caudal peduncle two and
one-fourth in its length ; pectoral about equal to the length of the head without the
opercle; distance between origin of dorsal and base of caudal twice its distance from
the snout; last dorsal ray slightly in front of the anal; dorsal and anal truncate or
slightly emarginate; anus under middle of dorsal; ventrals not reaching anus, their
origin equidistant from tip of caudal and interopercle ; caudal deeply forked, the
upper lobe two and one-half times as long as the middle rays; caudal fulcra not
conspicuous.

Genus VLI. Henonemus'' Eigenmann & Ward. (Plate XXXVI).
Henonemus Eigenmann & Ward, Ann. Carnegie Mus., IV, 1907, p. 118 {inter-

niedius) .
Cobitiglanis Fowler, Proc. Acad. Nat. Sci. Phila.-, 1914, p. 268, fig. 16 (taxistigma) .

Type. — Stegophilus intermedius Eigenmann & Eigenmann.

The genus Henonemus was created for Stegophilus intermedius on the obser-
vation that it has but one barbel at the angle of the mouth. The second (lower)
barbel at the angle of the mouth is so minute that it ought not to be considered of
generic value, especially since a minute barbel has been found on closer observation
in a number of cases where but one barbel had been recorded. It is probably present
in the type of Henoyiemus. The types of Humodicetus and Henonemus differ in the
number of opercular spines, four or five in the former, but two in the latter. The
names may, therefore, be at least temporarily retained.

Cobitiglanis was proposed as a subgenus of Ochmacanthus. However, the type
of Cobitiglanis is not related to Ochmacanthus. Cobitiglanis taxistigma is scarcely
distinct from H. punctatus and the name Coblitiganis is a synonym of the subgenus
Henonemus.

This genus, as far as known, consists of free-living species, and is closely re-
lated to the commensal Stegophilus. The mouth is wide, inferior, provided with
numerous teeth in series on the jaws and Ups; those of the middle of the upper lip
are long and in part are homologous with those of Vandellia; the opercle bears two
spines, the pre-opercle five or more. Where observed, the lower barbel is very
minute, followed by an additional barbel or labial lobe; the lower jaw is well formed,

3'^ iv = one, j/ij/ja = thread. A misnomer, since there are two maxillary barbels.

346 MEMOIRS OF THE CARNEGIE MUSEUM.

the rami transverse, united. The genus differs from Stegophilus chiefly in the shape
of the caudal, which is cmarginate instead of rounded, and in the number of oper-
cular spines.

Key to the Species of Henonemus.

a. Origin of dorsal equidistant from tip of caudal and interopercle; origin of vcntrals nearly equidistant
from tip of lower caudal lobe and snout.
6. Sides plain, caudal spotted, last half of its lower lobe black; D. 10; A. 9.

1. macrops (Steindachner).
66. Sides with a median series of spots, and smaller spots above them; tip of lower caudal lobe and

an oblique band across the upper lobe black 2. punctatus (Boulenger).

666. Sides with a regular series of spots, smaller spots above them; lower caudal lobe not black at tip,

several obscure spots on dorsal, caudal and base of pectoral 3. taxistigmus (Fowler).

on. Origin of dorsal equidistant from tip of caudal and occiput; origin of vcntrals equidistant from bases
of caudal and pectoral; caudal with faint dusky spots; upper surface with dark spots; a series of
larger spots along the middle of the sides 4. intermedius (Eigenmann & Eigenmann).

1. Henonemus macrops (Steindachner).
Stegophilus macrops Steindachner, Flussf. Slidam., IV, ]882, p. 28, pi. VI, fig.

2-2a (Lake Manacapuru); Eigenmann & Eigenmann, Proc. C'al. Acad. Sci.

(2), II, 1889, p. 55; Occasional Papers Cal. Acad. Sci., I, 1890, p. 344; Proc.

U. S. Nat. Mus., XIV, 1891, p. 37.
Henonemus macrops Eigenmann, Reports Princeton Univ. Exped. Patagonia, III,

1910, p. 401; RiBEiRO, Fauna Bras., IV (A), 1912, p. 231.

Habitat. — Lake Manacapuru.

Fig. 22. Henonemus macrops Steindachner. (After Steindachner.)

Known from the types only, which are in the Vienna Museum. Head 5;
depth 5; D. 10; A. 9; P. 6; eye 3.4 in the head; width of head 1.25 in its length;
barbel scarcely more than half as long as the eye; pectoral equals the head without
the snout; distance between caudal and origin of dorsal 1.5 in its distance from the
snout; origin of anal under the last dorsal ray; sides of head and body without spots;
tip of lower caudal lobe dark, the other fins plain.

2. Henonemus punctatus (Boulenger). (Plate XL, fig. C.)
Stegophilus punctatus (Boulenger), Proc. Zool. Soc. Lond., 1887, p. 279, pi. XXI,
fig. 4 (Canelos); Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II,

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 347

1889, p. 54; Occasional Papers Cal. Acad. Sci., I, 1890, p. 343; Proc. U. S. Nat.

Mus., XIV, 1891, p. 37.
Henonemus pundatus Eigenmann, Reports Princeton Univ. Exjjed. Patagonia,

III, 1910, p. 401.

Habitat. — Canelos, Ecuador; Santarem, Brazil.

Mr. Haseman collected the following specimens :
7083, C. M., fourteen, largest 88 mm. Santarem, Dec. 9, 1909.
7544, C. M., four, 72-90 mm. San Antonio, Rio Madeira, Nov. 3, 1909.

Fig. 23. Henonemus punctatus (Boulenger). After Bouleiiger.

Head 5.5-6; depth 6.5-7; D. 10 or 11; A. 8; P. 6; eye 4 in the head; width of
the head 1.2 in its length; opercle with two spines, interopercle with four in the
main row, of which the upi:ier is much larger, two in the second row; three rows of
movable teeth in the lip of the upper jaw, the rows close together, the teeth minute.

Fig. 24. Henonemus punctatus (Steindachner). A, opercle with its two spines; B, pre-opercle
with its spines; C, half of the premaxillary with its teeth.

348 MEMOIRS OP THE CARNEGIE MUSEUM.

except for the middle ones of the inner series, which are much larger, similar to those
in the premaxillary of Vmiddlia; four rows on the premaxillary, of which the three
outer rows are similar to, but a little larger than, the labial teeth; the teeth of the
inner series are pressed together, broad, bent toward the middle line near the tip and
then are abruptly bent backward and inward, then outward; about fifty teeth in
the inner row on each side, about seventy-five in the next row; five rows of teeth
in the lower jaw, the outer series labial, those of the next three rows minute, re-
curved hooks, those of the innermost series Uke those of the inner series of the
upper jaw; eye large, without a free orbital rim; posterior nares in a line with the
anterior margin of the eye, much closer together than the anterior; no nasal barbel;
maxillary barbel broad at base entirely concealing the very minute second barbel.
Axillary gland large, the pectorals considerably shorter than the head ; distance
of origin of dorsal from caudal 1.5-1.75 in its distance from the snout; origin of anal
under or behind the last dorsal ray; back with numerous small spots; sides with a
row of much larger spots; lower caudal lobe dark toward tip, base of the fin spotted
like the sides, upper lobe free from chromatophores toward its tip, a few spots along
lower margin of caudal peduncle ; dorsal with spots on its basal half.

3. Henonemus taxistigmus (Fowler).
Ochmacanthus {Cobitoglanis) taxistigma Fowler, Proc. Acad. Nat. Sci. Phila., 1914,

p. 268 {RupununiJi\\eT).

Habitat. — Rupununi River, British Guiana.

Known from the type, 39344 A. N. S. P., 93 mm., which I have had the oppor-
tunity to examine through the courtesy of Dr. H. W. Fowler.

Head 5.5; depth 6.5; D. 10; A. 7.5; P. 7; eye 3-3.33 in the head, width of head
1.1, snout 3.25, width of mouth 1.5, interorbital 3.5, pectoral 1.4, lower caudal lobe
1.17, caudal peduncle 2.5; lower barbel one-fifth as long as the upper; upper jaw
with four scries of teeth, upper lip with three, at least seven series in the mandible;
pre-opercle with five or more spines; opercle with two smaller spines; origin of
dorsal equidistant from tip of caudal and interopercle.

Fig. 25. Henonemus taxistigmus (Fowler). After Fowler.

Predorsal region with about four series of irregular, dusky spots, upper surface

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 349

of head with similar spots, one marking the interopercular, another the opercular
spines; a few median, dusky spots behind the dorsal; sixteen sharply defined, dusky
blotches along the lateral line increasing in size to the caudal peduncle. Fins all
pale or whitish, several obscure spots of dusky on dorsal, caudal, and base of pec-
toral.

4. Henonemus intermedius (Eigenmann & Eigenmann).

Stegophilus intermedius Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II,

1889, p. 54; Occasional Papers Cal. Acad. Sci., I, 1890, p. 343; Proc. U. S. Nat.

Mus., XIV, 1891, p. 37.
Henonemus intermedius Eigenmann, Reports Princeton Univ. Exped. Patagonia,

III, 1910, p. 401 ; Ribeiro, Fauna Bras., IV {A), 1912, p. 230.

Habitat. — Headwaters of the Rio Araguay.

This species, found in a region intermediate between the localities where
punctatus and maculatus are found, combines in a remarkable way the characters
of those species.

Type, No. 9842, M. C. Z., one specimen, 80 mm. Goyaz. Senhor Honorio.

Head 5.5; D. 9; A. 7; eye equals snout, 3.5 in the head; mouth large, upper Up
with two series of teeth, premaxiUary and mandible with four series of depressible
teeth, those of the inner series enlarged at the tip; barbel shorter than eye; head a
little longer than wide; opercle with two spines, interopercle with five or six claw-
like spines

Elongate, compressed behind, depressed forward; head somewhat longer than
wide, snout pointed; eye large, once in the snout, three and one-half times in the
head. Mouth large. Lower lip not dilated. Origin of dorsal about equidistant
from tip of caudal and occiput; caudal emarginate; anal placed entirely behind the
dorsal; origin of ventrals equidistant from bases of caudal and pectoral. Light
brown; entire upper surface with rather large dark brown spots; a series of larger
dark spots along the middle line of the sides, the spots becoming larger towards
the tail; caudal with a few, faint, dark spots.

Genus VIII. Pseudostegophilus^^ Eigenmann & Eigenmann.
Pseudostegophilus Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889,

p. 54.
Type. — Stegophilus nemurus Giinther.

This genus has the characters of Homodimtus but has a deeply forked caudal.

'' tpaiBrii = false; Stegophilus, the name of a related geiuis, see p. 3.53, from (rriyo^, t6 = a roof, and
4>iXos, 6 = a lover.

350 MEMOIRS OF THE CARNEGIE MUSEUM.

1. Pseudostegophilus nemurus (Giinther). (Plate XLIV, fig. 5.)

Stegophilus nemurus Gunther, Proc. Zool. Soc, London, 1869, y>. 429. (Peruvian

Amazon.)
PseAidostegopMlus nenmrus Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2),

II, 1889, p. 54 (Maranon or Ucayale); Occasional Papers Cal. Acad., Sci. I,

1890, p. 341; Proc. U. S. Nat. Mus., XIV, 1891, p. 37; Eigenmann, Reports

Princeton Univ. Exped. Patagonia, III, 1910, p. 400.

Habitat. — Upper Amazon, Rio Mamore.
7547a-/, C. M., 63-78 mm. Rio Mamore. Sept. 19, 1909. J. D. Haseman.

Head 5; depth 6-6.5; D. 9; A. 7; P. 6; eye 4-4.5 in the head, less than snout or
interorbital ; maxillary barbel about as long as the eye, lower barbel very minute.
Five rows of teeth on the upper Hp, four in the upper jaw, six rows in the lower jaw
and lip; gill-membrane not forming a fold across the isthmus; eight or nine spines
on the interoperclc and opercle; pectoral a little shorter than the head; origin of
ventrals about equidistant from tip of snout and tip of middle caudal ray (its base
in a specimen in the Mus. Comp. Zool.) ; origin of anal behind the dorsal, the dis-
tance of the base of its last ray from the base of the middle caudal ray four and one-
half in the length; caudal deeply forked, the lobes pointed, the upper lobe longer
than the lower; distance of origin of dorsal from base of middle caudal rays 1.5-1.6
in its distance from the snout; a dark shade across the head between the eyes,
another between the opercles, four bands across the back and sides about equal to
the interspaces, the margins of the bands darkest; lower caudal lobe and tip of the
upper black.

Genus IX. Homodmtus'*^ Eigenmann & Ward. (Plate XXXVII.)

Homodicetus Eigenmann & Ward, Annals Carnegie Mus., IV, 1907, p. 117, pi.

XXXIV, figs. 2 and 3 {anisitsi).

Type. — Homodiwtus anisitsi Eigenmann & Ward.

Opercle with four or five spines directed upward and backward, interopercle
with more, directed downward and backward; cj^e 3.5-5 in the head. Otherwise
like Henonemus.

Key to the Species of Homodicetus,
a. Caudal slightly emarginate, oblique; accessory rays numerous; origin of dorsal equidistant from tip of
caudal and eye; origin of ventrals equidistant from snout and caudal; D. 8; A. S; back and sides
with chromatophores, but withdut distinct spots; middle caudal rays l)lack.

1. anisitsi Eigenmann & Ward.

'■* ojuoStaiTos = living or eating with others. In allusion to the known parasitic habits of some of its
relatives.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 351

aa. Caudal forked or deeply emarginate, accessory rays few, inconspicuous; origin of dorsal a little nearer
tip of snout than tip of upper caudal lobe, or the reverse; origin of ventrals equidistant from base of
caudal and the eye, or a little nearer the former; D. 10; A. 7; back and sides with conspicuous spots
unsymmetrically distributed, sometimes in part arranged as a median series; a spot at base of
caudal, tips of caudal lobes black 2. maculatus (Steindachner).

1. Homodisetus anisitsi Eigenmann & Ward. (Plate LVI, figs. 3 and 5.)'^
Hoinodicctus anisitsi Eigenmann & Ward, Ann. Carnegie Mus., IV, 1907, p. 117,

pi. XXXIV, figs. 2 and 3 (Villa Rica) ; Eigenmann, Reports Princeton Univ.

Exped. Patagonia, III, 1910, p. 401.

Habitat. — Villa Rica, Paraguay.

Known only from the type in the collection of Indiana University.
10155, I. U. M., 9 , 43 mm., the type. Small creek at Villa Rica, Paraguay.

Anisits.

Head 6.5; depth 5.75; D. 8; A. 8; eye equals snout, 3.5 in the head, about equal
to interorbital; head nearly as wide as long; opercle with about four spines, inter-
opercle with six; the barbel shorter than the eye, the inner barbels much smaller;
upper jaw and Hps each with about four distinct series of teeth; those on the Ups
freely movable; the teeth narrow, more or less spoon-oar-shaped, those of the inner
series slightly larger; lower lip without teeth, three series of teeth on the jaw.

Axillary gland very large ; origin of dorsal equidistant from tip of caudal and
posterior margin of the eye; caudal slightly emarginate, the upper lobe longest;
origin of anal under end of dorsal; ventrals reaching vent, which is equidistant from
tip of mouth and tip of caudal, origin of ventrals equidistant from snout and caudal.
Accessory rays numerous.

In alcohol uniformly pale. The fresh specimen preserved in formalin was straw-
colored, the back with numerous large, conspicuous, stellate, black chromatophores,
and many more smaller, much less consi^icuous, brown ones; sides with a few small,
stellate, black, chromatophores, gradually giving rise to a regular series along the
middle of the tail; a dusky streak along tlie sides between the myotomes of the body
and the thin covering of the abdominal cavity; a small, black spot at the base of
the middle caudal ray; middle caudal rays dark, becoming intensely black toward
tip; oblique bars extending from the end of the second ray below median dark one
downward and forward to the tip of the lower caudal fulcra and then as a black line
forward along the tips of the fulcra; another one like it in all resi)ects from the tip
of the second ray above the median dark one upward and forward to the tip of the
caudal fulcra and then forward along their tips as a black line; remaining fins more
or less dotted.

'^ The caudal should be emarginate in the figure.

352 MEMOIRS OF THE CARNEGIE MUSEUM.

Alimentary canal straight, without convolutions or bends, the thin-walled
stomach lying lengthwise and giving rise to a short, thin intestine, which merges
into the much longer and larger, but thin-walled, large intestine which appears to
be filled with minute grains of sand.

2. Homodiaetus maculatus (Steindachner).
Slegophilus maculatus Steindachner, Denk. Ak. Wiss., Wien, XLI, 1879, p. 25,
IV, fig. 2 (La Plata); Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2),

II, 1889, p. 54; Occasional Papers Cal. Acad. Sci., I, 1890, p. 343; Proc. U. S.
Nat. Mus., XIV, 1891, p. 37.

Henonemus maculatus Eigenmann, Reports Princeton Univ. Exped. Patagonia,

III, 1910, p. 401.

Habitat. — La Plata, in Province Buenos Aires; Uruguay Basin.

Known from the type, a specimen 105 mm. long, and
7545a-d, C. M., 67-69 mm. Uruguayana, Feb. 6, 1907. J. D. Haseman.
7546a-c, C. M., 64-70 mm. Cacequy. Feb. 2, 1907. J. D. Haseman.

Head 5.75-6.75; depth 6-9; D. 9; A. 7; P. 6; eye 3.66-4 in head, equal to snout
or interorbital ; maxillary barbel filamentous, equal to the eye, the inner barbel
much smaller; seven series of teeth in the upper jaw and Hp, five in the lower; about
seven, graduate, opercular spines, about nine larger interopercular spines; gill-
membrane forming a free fold across the isthmus; anal behind the dorsal. In the
type, a series of spots along the back, another row of spots along the middle of the
sides; two or three rows of smaller spots between the two; base of caudal with a
dark cross-bar, several small spots along the upper edge of the caudal; tips of the
caudal dark spotted.

4'

Fig. 26. HomodicBlus maculatus (Steindachner). After Steindachner.

EIGENMANN: the PYGIDIIDiE, A FAMILY OF SOUTH AMERICAN CATFISHES. 353

In the specimens collected by Haseman the color-marking is less regular, the
sides and back with unsymmetrically placed spots, largest on the caudal peduncle,
smallest on top of the head, sometimes arranged in a series along the middle of the
sides; the spot at the base of the caudal largely above the middle.

Genus X. Stegophilus^'' Reinhardt. (Plate XXXVII.)
Stegophilus Reinhardt, Vidensk. Meddel. Naturli. Foren., Kjobenhavn., 1858

(1859), p. 79, pi. II.

Type. — Stegophilus insidiosus Reinhardt.

No nasal or mental barbel, lower barbel at angle of mouth excessively minute,
a minute dermal flap below the lower barbel; mouth very wide, inferior; eye large,
superior; posterior nares between the front parts of the eyes; opercle and inter-
opercle with several spines; gill-opening narrow, about a third as wide as the mouth,
in front of the pectoral, the membrane not forming a free margin ; first pectoral ray
not produced in a filament ; origin of ventral one and a half .to two times as far from
snout as from caudal; caudal rounded, not greatly contracted at base, the accessory
rays not conspicuous; origin of dorsal behind the vertical from the origin of the
ventrals; teeth very numerous, in regular series, those in the middle of the upper
jaw larger than the others.

1. Stegophilus insidiosus Reinhardt.
Stegophilus insidiosus Reinhardt, I. c, p. 79-97; Gunther, Cat. Fishes Brit. Mus.,

V, 1864, p. 276; LDtken, Velhas Flodens Fiske, p. 15; Vidensk. Selsk. Skr.

(5), Afd. XII, 1875, p. 135, and I, text figures 1-3 (Rio das Velhas); Eigenmann

& Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889, p. 55; Occasional Papers

Cal. Acad. Sci., I, 1890, p. 344; Proc. U. S. Nat. Mus., XIV, 1891, p. 37;

Eigenmann, Reports Princeton Univ. Exped. Patagonia, III, 1910, p. 400.

Habitat. — Parasitic in large fishes {Pseudoplatystoma orbignianum = coruscans)
of the San Francisco basin and free on sand bars of the upper San Francisco basin.
7551, C. M., one, 32 mm. Opposite Januaria. Dec. 12, 1907. Haseman.

This example is the first one secured since Reinhardt obtained his specimens
from the gill-chamber of a large catfish, Pseudoplatystoma. Haseman took his
specimen on Dec. 12, 1907, from the sandy shore of an island in the Rio San Fran-
cisco, in front of the town of Januaria. If this specimen is really identical with
those secured by Reinhardt from the same river basin, Stegophilus appears to have
the general habit of members of the family of burrowing in sand as well as the pecu-
har habit of entering the gill-chambers of other fishes. This double, Jekyl and

^^(TTkyos, TO — a, roof; (t>i\o;, o = a lover, i. c, loving a covered home, in allusion to its habit of living
in the gill-cavities of other fishes.

354

MEMOIRS OF THE CARNEGIE MUSEUM.

Hyde, habit of Stegophilus lends probability to the reported habit of Vandellia
of entering the urethrse of bathers. For an account of the habit of the species
see page 267.

Fig. 27. . Stegophilus insidiosus Reinhardt. (After Liitkcn.)

Head about six times in the length; D. 8; A. 7; P. 7; eye about equal to snout
or interorbital, four times in the head; maxillary barbel equal to haK the width of
the mouth, extending to the interopercular spines, lower barbel about four-tenths
as long; a thin, narrow membranous flap below the lower barbel; head flat below,
its width equal to its length less the opercular spines.

This specimen has hardened in the alcohol and it is not possible to describe
the details of the teeth.

Ten or eleven hooks in two scries, on the interopercle, directed downward and
backward, eleven or twelve thorns on the opercle in three or four irregular series,
increasing in size from the minute anterior ones to the strong posterior one.

Pectoral about as long as the head without the snout; origin of ventrals equi-
distant from base of caudal and tip of pectoral; distance between origin of dorsal
and base of middle caudal rays 2.33 in its distance from the snout, distance between
last anal ray and base of caudal 6.5 in the length; caudal rounded, ventral accessory
rays inconspicuous, a few prominent dorsal accessory rays.

No color-markings.

Genus XI. Acanthopoma" Liitken. (Plate XXXVII.)
Acaniho-poma Lutken, Vidensk. Meddel. Naturh. Foren. Kjobenhavn, for 1891,
1892, p. 53, fig. (annectens) .
Like Stegophilus, the gill-membranes forming a free fold across the isthmus.

" aKavBos, b = spiiie; ?rwjua, to = opercle.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 355

1. Acanthopoma annectens Liitken.

Acanthopoma annectens Lutken, I. c. (Huallaga); Eigenmann, Ann. Carnegie Mus.,
IV, 1907, p. 119; Reports Princeton Univ. Exped. Patagonia, III, 1910, p. 401.
Habitat. — Huallaga.

Known from the type 100 mm. long in the collection of Prof. R. Leuckart.
Anterior part, especially the head, depressed; head parabolic; eyes not large; dis-
tance between anterior nares twice the distance between the posterior, neither with
a barbel; a group of four to six large and some small spines on the opercle; another
larger group, ten to twelve, on the interopercle ; mouth inferior; upper jaw with six
to seven very regular rows of very small teeth; in the lower jaw are "naeppe mere
end en enkelt Roekke telstede." Free margin of the gill-membrane begins behind
the interopercle and is continued across the isthmus without uniting with it.

Fig. 28. Acanthopoma annectens Liitken. (After Lutken.)

Origin of ventrals equidistant from bases of caudal and pectoral, or tip of
caudal and the mouth; origin of dorsal nearly twice as far from snout as from base
of caudal; origin of anal under end of dorsal; caudal slightly emarginate; back with
obscure spots.

Called annectens to indicate its supposed position between the Pygidiinse and
Stegophilinse.

Lutken says it is nearest Henonemus microps = macrops?

Genus XII. Ochmacanthus'* Eigenmann. (Plate XXXVII.)

Ochmacanthus Eigenmann, Mem. Carn. Mus., V, June, 1912, p. 213.

Gyrinurus Ribeiro, Comm. Linhas Telegraphicas Estrategicas de Matto-Grosso

ao Amazonas, Annexo No. 5, Sept., 1912, p. 27, pi. with three figures.

Type. — Ochmacanthus flabelliferus Eigenmann.

^* oxi^a, TO = a hold ; dKavBo's, 6 = a spine.

356 MEMOIRS OF THE CARNEGIE MUSEUM.

No nasal or mental barbels; lower barbel at angle of mouth minute or well
developed; mouth very wide, inferior; eye large, superior; posterior nares between
the anterior margins of the eyes; gill-opening very small; first pectoral ray not
spinous, not produced in a filament; accessory caudal rays very numerous, the
caudal greatly contracted at base; origin of anal behind that of the dorsal; teeth
very numerous, in regular series, none on the lip of the upper jaw.

Key to the Species of Ochmacanthus.
a. Maxillary barbel reaching pectoral; origin of dorsal equidistant between tip of caudal and snout or
nearer caudal; accessory caudal rays highest near the middle; distance between last anal ray and

caudal two in its distance from the snout. {Gyrinurus) 1. batrachostoma (Ribeiro).

aa. Maxillary barbel about as long as the eye, reaching tip of interopercular spines; origin of dorsal much

nearer tip of caudal than snout; accessory caudal rays graduate to the caudal, into which they

gradually merge. {Ochmncaiithu^).

b. Distance between last anal ray and caudal 4-4.5 in the length; origin of dorsal in advance of

origin of anal; ventrals not reaching anal; gill-opening extending from the opcrcle to behind the

middle or lower part of the opercular spines 2. reinhardti (Steindachner).

bb. Distance between last anal ray and caudal three in its distance from the snout; origin of dorsal
over origin of anal, ventrals reaching anal; gill-nponings confined to the region between the
interopercular and opercular spines 3. flabelliferus Eigenmann.

1. Ochmacanthus batrachostoma (Ribeiro). (Plate LV, figs. 1-3.)

Gyrinurus batrachostoma Ribeiro, I. c. (S. Luiz de Caceres.)
Habitat. — Upper Paraguay.
Known from the type, a specimen 32 mm. long, and

7553, C. M., about 31 mm. Puerto Suarcz, swampy shore of big bay between
Brazil and BoUvia. May 7, 1909. Haseman.

7554, C. M., 30 mm. Rio Jauru, twenty-eight miles above its mouth at Campos
Alegre, thirty miles southwest of Caceres. June 2, 1909. Haseman.

Head 5.5-6.5; depth 6.5-9; D. 10-12; A. 8-9; P. 4; eyes superio-lateral, three
or four in the head, longer than the snout, about equal to the interorbital ; maxillary
barbel reaching axil of pectoral, the lower one to the opercular spines; head as.
broad as long; six to eight opercular spines arranged in a group; interopercle with
about five to eight spines in one or two series; teeth conical, in three parallel series;
body depressed in front, compressed behind; dorsaP and anal rounded; dorsal
behind the ventrals, its origin equidistant from tip of snout and tip of caudal in the
figure of the type, its distance from the caudal 1.3-1.5 in its distance from the snout
in the specimens enumerated above; anal partly under dorsal, distance of its last
ray from the caudal about 3.5 in its distance from the caudal in the specimens at
hand; caudal minute, rounded, hidden in the accessory rays which are greatly

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 357

developed into fins like those of a larval frog; everywhere except on the belly spotted.
Twenty-one vertebrae behind the anal.
The type was caught among water-weeds {Eichnrnea azurea).

2. Ochmacanthus reinhardti (Steindachner). (Plate LV, fig. 4.)
Stegophilus reinhardti Steindachner, Flussf. Siidam., IV, 1882, p. 28, pi. VI, fig. 1

(Lake Manacapuru; Rio I^a; Montalegre; Te^^e; Tabatinga); Eigenmann &

EiGENMANN, Proc. Cal. Acad. Sci. (2), II, 1889, p. 55; Occasional Papers Cal.

Acad. Sci., 1, 1890, p. 344;Proc. U. S. Nat. Mus., XIV, 1891, p. 37; Eigenmann,

Reports Princeton Univ. Expcd. Patagonia, III, 1910, p. 401 ; Ribeiro, Fauna

Bras., IV (A), 1912, p. 401.

Habitat. — Amazons.
7552, C. M., 38 mm. Amazon, at upper end of island four miles above Santarem.

Dec. 9, 1909. Haseman.
7555a-b, C. M., 35-46 mm. Igarape de Jaura, entering R. Tapajos two miles

above Santarem. Dec. 11, 1909. Haseman.

Head 7; D. 11 13 (9 or 10 developed rays); A. 10-11 (8 or 9); P. 6, partly
adnate; eye equals snout, less than interorbital, entirely in the anterior half of the
head, 4-5 in the head; maxillary barbel reaching interopercular spines, the lower
barbel a third or fourth as long; 8 or 9 interopercular spines, 9 to 12 opercular;
width of head equal to its length. Three series of teeth in the upper jaw, those of
the inner series close set, much more numerous than those of the outer series, the
teeth of the outer two series at the middle of the mouth a little longer and more
slender than the rest; lower jaw with two complete series of teeth, the inner series
similar to inner series of the upper jaw, those of the outer series larger, fewer, and
much more movable than those of the inner, about fifteen in the outer series, about
foity in the inner; four increasingly shorter series from the inner series outward
near the middle of the jaw.

A prominent pectoral pore, pectoral equal to the head or to the head without
the snout. Origin of ventrals equidistant from bases of caudal and pectoral or a
little farther forward, distance between last anal raj^ and caudal 4^.5 in the length;
distance between origin of dorsal and base of caudal 1.75-2 in its distance from the
snout; caudal rounded, with many prominent accessory rays.

Back gray, sides and fins mottled.

3. Ochmac.anthus fiabelliferus Eigenmann. (Plate LV, fig. 5.)
Odunacanthus flabelliferus Eigenmann, Mem. Carnegie Mus., V, 1912, p. 213.

Habitat. — Essequibo basin.

358 MEMOIRS OF THE CAENEGIE MUSEUM.

1729, C. M.; 12111, 1. U. M., type and paratypes, three, 33-35 mm. (Konawaruk.)
Head 5.33; depth 7; D. 8; A. 7; eye 1 in snout, 3.75 in head, 1 in space between
the eyes. Width of head equal to its length; snout semicircular in outline, the head
depressed; mouth very wide, its width equal to the length of the head less half the
snout; upper jaw with three series of teeth; teeth of the two outer series conical,
those of the inner series broad, removed from the others, forming a solid palisade;
no labial teeth; lower jaw with an outer series of long, curved, claw-like teeth in the
lip, and four series in the jaw, of which the first is short, near the middle, the second
extends farther to the sides, the third is longest, extending from the middle to the
side of the jaw, the fourth is shorter again and confined to the sides, not reaching
the median line of the jaw. Interopercle with nine claw-like erectile spines; opercle
somewhat prolonged, carrying a bunch of nine spines similar to those of the pre-
opercle above and behind the gill-opening. Gill-opening small, entirely above the
level of the middle of the pectoral; outer maxillary barbel about as long as the eye,
the inner one minute. Pectorals partly adnate; ventrals small, free, reaching anal;
dorsal about equal to the anal and but slightly farther forward.

''Light, with numerous chromatophores more or less aggregated in places; a
black spot on base of caudal."

The only specimens known were killed with hiari poison in a small ]wol of
back-water from the Essequibo.

Genus XII. Vandellia^^ Cuvier & Valenciennes. (Plate XXXVIII.)

Genus XIII. Urinophilus Eigenmann.

Vandellia Cuvier & Valenciennes, Hist. Nat. Poiss., XVIII, 1846, p. 386, pi.

547.

Type of Vandellia. — Vandellia cirrhosa Cuvier & Valenciennes.

Long, slender fishes with small inferior mouth; a few teeth in a single series in
the middle of the upper jaw; peculiar, claw-like teeth on the end of the maxillary
in some species, probably all of them; teeth on the mandible in some species, none
in other species; the mandibular rami not meeting, separated by a wide membrane;
opercular spines directed obhquely upward and backward, interopercular spines
directed downward and backward; gill-opening small; no nasal or mental barbels,
the lower of the barbels at angle of mouth very minute ; first pectoral ray not pro-
longed in a filament; ventrals very much nearer to caudal than to tip of snout ; origin
of the anal behind that of the dorsal.

'' In honor of Domingo Vandelli, professor of natural history at Lisbon, who sent the types of the
genus to Lac(5pcde.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 359

There are two generic types contained in the genus as here understood. One
of the two genera has teeth on the mandibles and is represented by V. sanguinea
at least; the other genus lacks teeth in the mandible, and plazai and hasemani at
least belong to this genus. I do not know whether the type of VandeUia belongs
to the one or the other of the genera. The new one may be named Urinophilus.
The type is to be selected after the structure of the mandible in VandeUia cirrhosa,
the type of VandeUia, has been examined.

The dentition of the VandeUiirm is very peculiar. There are one or two series
or a patch of pointed teeth in the center of the upper jaw. They are immediately
below the center of the ethmoid at its anterior end, in other words, in the region
occupied by the premaxillary in related forms. The bones are so thin and the fish so
small, that it is difficult to determine all of the outhnes of the bones, or to determine
the identity of all of the bones. The bone on which these teeth are inserted is, in
all probability, the premaxillary. The lateral points of the ethmoid are forked, and
dovetail into the forked ends of maxillaries somewhat after the fashion of the two
hands locked into each other between the thumbs and fingers. On the distal half
of the maxillary of VandeUia there are from two to four comparatively large and
very peculiar "claw-teeth," arranged hke overlapping shingles, the outermost one
being next the bone, the second from the end overlapping this and so on to the
proximal one. The individual teeth consist of a flat, oval disk, from the upper
proximal corner of which the tooth proper points toward the end of the bone. The
bone touches the palatines proximally and the maxillary barbel is joined rather
firmly to the end of it, all of which indicates that this bone is the maxillary. In
Branchioica only one or two teeth of this sort are present, but between them and
the ethmoid there is a series of slender, pointed teeth, similar to those on the pre-
maxillary.

Catfishes with teeth on the maxillaries are very unusual. Outside of the
Vandelliinm, teeth are only found on the maxillary in Diplomysfe of the Diplomy-
stidea of Chile.

Covering the end of the maxillary and joined to the dorsal surface of the base
of the barbel is a thin, comma-shaped bone, which may be the nasal.

Teeth, such as those described on the maxillary, have so far been noticed in
VandeUia hasemani, sanguinea, and plazai. V. cirrhosa and V. wieneri have not
been examined in this respect. Teeth hke these are found in another member of
the family, Pareiodon (compare figures 21 and 34-35).

In ParavandeUia and Branchioica the "claw-teeth" of the proximal part of
the maxillary are replaced by slender, pointed teeth, and there arc more than one

360 MEMOIRS OF THE CARNEGIE MUSEUM.

series of teeth on the premaxillary. In Branchioica a "claw-tooth" is present at
the end of the maxillary.

The rami of the lower mandibles of the Vandelliince seen from below are elon-
gate triangles, converging forward, but not joined, in fact, not meeting in the center.
Teeth may or may not be present in the lower jaw. When present, they are re-
curved, pointed, in two series at the end of the jaws, in apposition to the maxillary
teeth.

The alimentary canal is a simple, straight tul)c, of nearly uniform diameter,
evidently greatly distensible.

Key to the Species of Vandelli/..
a. Mandible without teeth, thorn-like teeth at end of maxillary of V. plnzai and hasemani. (Neither
mandible nor maxillary examined in V. cirrhosa and uncneri.)
b. Caudal truncate or slightly emaria;inate.

c. D. 8-9; A. 9-10; P. 6; depth 9; premaxillaries with five to eight teeth; maxillary barbel two

in the head; caudal slightly cmarginate, the lobes rounded, equal; pectorals longer than

the head 1. cirrhosa Cuvier & Valenciennes.

cc. D. 9; A. 8; P. 7; depth 12; premaxillaries with five to nine teeth; maxillary barbels less than
half the length of the head; caudal cmarginate, the lobes rounded; pectorals as long as the

head 2. plazai Castelnau.

bb. Caudal forked; D. 11; A. 10; P. 6; depth 7-S.

d. Premaxillaries with nine teeth; mouth wide, angle of gape nearly opposite the maxillary

barbel; maxillary barbels about 3 in the head; distance between the origin of the dorsal
and origin of the caudal 2.75 in the distance of the dorsal from the snout; pectorals shorter

than the head 3. wieneri Pellegrin.

dd. Premaxillaries with about six teeth; mouth small, the angle of the gape far in advance of the
base of the barbel; maxillary barbel 2-2.5 in the head; distance between origin of dorsal
and base of middle caudal rays two and a quarter to two and one-half in its distance
from the snout; pectoral about equal to the length of the head.

4. hasemani Eigenmann,

aa. A patch of minute teeth on each mandible; one or two claw-like teeth on the end of the maxillary, just

in front of the barbels; caudal truncate; D. 11; A. 8; P. 6; premaxillaries with five teeth; maxillary

barbel two in the head; distance between origin of the dorsal and the origin of the caudal 2.8 in the

distance of dorsal from the snout; pectorals little shorter than the head . 5. sanguinea I'^igenmann.

1. Vandellia cirrhosa Cuvier & Valenciennes.

Vandellia cirrhosa Cuvier & Valenciennes, Hist. Nat. Poiss., XVIII, 1846,
p. 386, pi. 547; Castelnau, Anim. Amer. du Sud, 1855, p. 51, pi. 28, fig. 2
(Brazil); Gijnther, Cat. Fishes Brit. Mus., V, 1864, p. 277; Eigenmann &
EiGENMANN, Proc. Cal. Acad. Sci. (2), II, 1889, p. 55 (Hyavary); Occasional
Papers Cal. Acad. Sci., I, 1890, p. 345; Proc. U. S. Nat. Mus., XIV, 1891,
p. 37; BouLENGER, Proc. Zool. Soc. Lond., 1897, pp. 901 and 920, Trans. Zool.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 361

Soc. Lond., XIV, 1898, p. 426; Jobert, Archiv. Parasit, I, 1898, p. 494; Pelle-
GRiN, Bull. Soc. Philom. Paris (10), I, 1909, p. ? (Apure, Manaos); Eigenmann,
Reports Princeton Univ. Exped. Patagonia, III, 1910, p. 401; Ribeiro, Coram,
de Linhas Telegraphicas Estrategicas de Matto-Grosso ao Araazonas, Annexo
5, 1912, p. 30 (Manaos).

Habitat. — Amazon and Orinoco basins; Rio Jurua; Hyavary; Manaos; Apure.
This fish, the bad repute of which is widespread (see page 265) in South America,
is represented in Museums by only the few specimens in the following list, in part
copied from Pellegrin:

Jardin des Plantes, Paris, three, locality? Vandelli.

Jardin des Plantes, Paris, one, Apure. Geay.

Jardin des Plantes, Paris, two, Manaos. Anthony.

Mus. Comp. Zool., Cambridge, one, 40 mm., Hj^avary. Bourget.

British Mus., London, ? many, Jurua. Bach.

Mus. Nac. Rio de Janeiro, one, 94 mm., Manaos. Ribeiro.

Fig. 29. Vandellia cirrhosa Cuvier & Valenciennes. (After Cuvier & Valenciennes.)

Head 8-10.5 (9-11.5 including the caudal); depth 9; D. 8-9; A. 9-10; P. 6.
Head slightly longer than wide; eye less than 3 in the head, greater than the snout;
six to ten spines on the opercle, five to ten on the interopercle ; five to eight teeth on
the premaxillary; origin of dorsal twice as far from tip of snout as from margin of
caudal; dorsal partly over'anal.

362 MEMOIRS OF THE CARNEGIE MUSEUM.

2. Vandellia plazai Castelnau. (Plate LIII, fig. 3.)
Vandellia plazaii Castelnau, Anim. Amer. du Sud, Poissons, 1855, p. 51, pi. 28,
fig. 1 (Ucayale) ; Vaillant, Bull. Soc. Philom. (7), IV, 1880, p. 159 (Calderon);
EiGENMANN & EiGENMANN, Proc. Cal. Acad. Sci. (2), II, 1889, p. 55 (Lake
Hyanuary); Occasional Papers Cal. Acad. Sci., I, 1890, p. 345; Proc. U. S.
Nat. Mus., XIV, 1891, p. 37; Pellegrin, Bull. Soc. Philom. Paris (10), I,
1909, p. ? (Calderon); Eigenmann, Reports Princeton Univ. Exped. Pata-
gonia, III, 1910, p. 401; RiBEiRO, Comm. de Linhas Telegraphicas Estra-
tegicas de Matto-Grosso ao Amazonas, Annexo 5, 1912, p. 30 (Manaos).
Vandellia plazce Gunther, Cat. Fishes Brit. Mus., V, 1864, p. 277.

Habitat. — Middle and Upper Amazon basin; Ucayale, Calderon; Lake Hyan-
uary.

This species also is known from but few specimens, as follows:
Jardin des Plantes, Paris, ? Ucayale. Castelnau.
Jardin des Plantes, Paris, one, Calderon. Jobert.

Mus. Comp. ZooL, Cambridge, Mass., one, 125 mm., Lake Hyanuary. Bourget.
Mus. Nac. Rio de Janeiro, one, 67 mm., Manaos? Ribeiro.
7541, C. M., one, 66 mm. Dec. 9, 1909. Santarem. Haseman.
It is principally distinguished by its more elongate form.

Fig. 30. Vandellia plazai Castelnau. Carn. Mus., No. 75-11.

Head 9-11 (10-12 in the total length); depth 12-15.3 in the total length;
D. 9-10; A. 8-9; P. 7; twelve to sixteen opercular spines in 3 or 4 rows, seven or
eight on the interopercle ■ head more rounded than in cirrhosa; barbel less than half
the length of the head; pectoral as long as the head; 8 or 9 teeth in the upper jaw.

3. Vandellia wieneri Pellegrin.
Vandellia wieneri Pellegrin, C. R. Ac. Sc, November 29, Vol. 149, 1909, p. 1016;
Bull. Soc. Philom. Paris (1), X, 1909, p. 199, page of reprint 3, figure in the

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 363

text; Mission Geodesique de TEquateur, XII, 1912, p. 2, p. 10, pi. I, fig. 2,

near the mouth of the Rio Misahually.

Habitat. — Rio Napo, Ecuador.

This species is known from the type. It is distinguished by its short body,
short barbel, and forked tail.
A.9934, Paris Museum, one, 92 mm. Rio Napo near the mouth of the Misahually,

Ecuador. Charles Wiener.

Fig. 31. Vandellia wieneri Pellegrin. (After Pellegrin.)

Head over 7; depth 7; D. 11; A. 10; P. 6; 9 teeth; lower jaw incised in middle,
without teeth; maxillary barbel about three in the head; eye two in the snout;
fifteen opercular spines in four series, directed obliquely upward and backward;
seven or eight interopercular spines in two rows; dorsal about two and three-fourths
times nearer caudal than snout; ventrals a little in advance of the last third of the
body; caudal peduncle 2.5 as long as high. Named for the collector, Mr. Charles
Wiener.

Rio Ma-

Fig. 32. Vandellia wieneri Pellegrin. (After Pellegrin.)

4. Vandellia hasemani Eigenmann. (Plate XVIII, fig. 3.)
7542, C. M., type, 72 mm.; 7543a-6, C. M., paratypes, 68 and 69 mm.

more. Haseman.

Evidently similar to V. wieneri. Head 8-8.5; depth 8; D. 11; A. 10; P. 6.
Five or six teeth in the premaxillaries, two thorn-Uke teeth on the distal part of the
anterior face of the premaxillary, two or three more slender teeth on the distal part
of the lower face of the premaxillary; mandibular rami without teeth, widely
separated from each other, the membrane between the two rami but little emargi-
nate; angle of gape about halfway between the premaxillary and the barbel; maxil-

364

MEMOIRS OF THE CARNEGIE MUSEUM.

lary barbel 2-2.5 in the length of the head; lower barbel minute; about ten inter-
opereular thorns, fifteen or more on the opercle; broad, free, fleshy lobes behind the
opercular and pre-opercular spines; gill-openings about half as wide as the mouth;

Fig. 33. Vandellia hasemani Eigeninann. Type, No. 7543a, Cam. Mus., 72 mm.

eye entirely in the anterior half of the head; the posterior nares nearly as large as
the eyes and between the anterior halves of the latter, anterior nares. with a short
flap; pectorals about equal to the length of the head; origin of ventrals equidistant

Fig. 34. Vnndcllia hasemani Eigcnmann. A, skull from above; B, hyomandibular and opercular
apparatus; C, end of ethmoid and premaxillary teeth from in front,'!, j)remaxinary; 2, ethmoid; .5, frontal ;
6, sphenotic; 7, pterotic; 8, supraoccipital ; 9, epiotic; 10, supraclavicle; 11, parapophysis of coalescep
vertebra;; 12, maxillary; 13, palatine; 14, metapterygoid ; 15, (|uadrate; 16, preopercle; 17, interopercle;
18, opercle; 19, hyomandibular.

from tip of caudal and eye or opercle, reaching a little beyond the anus; origin of
anal under middle of dorsal; distance between anal and base of middle caudal rays
4.5-5.25 in the length; caudal forked for about two-ninths of its length; origin of

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 365

dorsal over the tips of the ventrals, distance of origin of dorsal from caudal 2.25-2.5
in its distance from the snout.

Back and basal part of caudal truncate.

Fig. 35. Vandellia hasemani Eigenmann. A, Left premaxilhiry from below, with two functional
and two relay teeth. B, left premaxillary of another individual with four teeth. C, the same, looking
at the edge of the teeth. D, premaxillary from below.

5. Vandellia sanguinea Eigenmann. (Plate LIII, fig. 2.)
Vandellia sanguinea Eigenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. 701.
7082, C. M., type, 62 mm. San Antonio de Rio Madeira. Nov. 3, 1909. Hase-

man.

This species differs from the others of the genus Vandellia in having concealed
teeth on the ends of the mandibles. It resembles them so much in other points
that it naturally raises the question whether these structures have not been over-
looked in V. cirrhosa and wieneri. They cannot be seen without considerable effort.
The species greatly resembles V. plazai. Mr. Haseman noted that the specimen
was white (translucent?), the alimentary canal straight and gorged with blood.

Fig. 36. Vandellia sanguinea Eigenmann. No. 7082, Carn. Mus. Type, 62 mm.

366 ■ MEMOIRS OF THE CARNEGIE MUSEUM.

Head 11.66; depth 12; D. 4 + 8.5; A. 3 + 7; P. 7; nearly the entire eye in the
anterior half of the head, a httle more than four in the length of the head to the tip
of the opercular spines. (The eye is too small in the drawings.)

Maxillary barbel extending to the tip of the interopercular spines, two in the
head; the lower barbel minute, only about half a millimeter long as compared with
the 2.5 mm. of the maxillary barbel; two flat, recurved teeth on the end of the

Fig. 37. Vandcllia sanguinea Eigcnmann. Left prcmaxillary, showing one of the concealed
teeth.

prcmaxillary concealed just in front of the barbel; five prcmaxillary teeth graduated
from the long middle one to the minute lateral ones; mandibles widely separated
from each other with about five minute teeth; the teeth concealed by the Up; five
spines in the main row of the interopercle, the middle ones very strong, directed
backward, about five spines in supplementary rows; five spines in the main row of
the opercle, about ten in supplementary rows; distance from origin of ventrals to
base of middle caudal rays two times in its distance from the snout; origin of anal
behind the origin of the dorsal, the last dorsal ray over the middle of the anal; dis-
tance between anal and base of middle caudal rays five and one-half times in the
length; distance from origin of dorsal to base of middle caudal rays two and eight-
tenths times in its distance from the snout; caudal truncate, with numerous acces-
sory rays. Translucent, the eyes black.

Genus XV. Paravandellia« Ribeiro. (Plate XXXVIII.)

Paravandellia Ribeiro, Comm. Linhas Telegraphicas Estrategicas de Matto-

Grosso ao Amazonas, Annexo No. 5, Sept., 1912, p. 29.

Type. — Paravandellia oxyptera Ribeiro.

No nasal or mental barbels, one (probably two) barbels at the angle of the
mouth; first pectoral ray not continued as a filament; gill-opening small; mouth
inferior, with a band of teeth in the middle of the upper jaw and a single series
laterally ; no teeth on the mandible ; ventrals much nearer tip of caudal than snout ;
opercular and interopercular spines separate from each other. Caudal forked
("furcada"). Ribeiro says that this genus may be considered between Stegophilus
and Vandellia, having the general appearance of the former.

^° irapa = near, Vandellia = name of a related genus.

EIGENMANN: the PYGIDIIDiE, A FAMILY OF SOUTH AMERICAN CATFISHES. 367

1. Paravandellia oxyptera Riboiro.

Paravandellia oxij'ptera Ribeiro, I. c.

Habitat.— Faragnay River near Cacercs.

Head triangular, eight times in the length including the caudal; D. 12; A. 10;
P. 7; dorsal behind the ventrals, partly over anal, both behind the middle of the
body; origin of dorsal nearer tip of caudal than base of pectorals; eyes without a free
margin, large, two and one-half times in head, equal to snout; maxiUary barbel
reaching at most to tip of interopercular spines; pectorals large, falcate, one-fourth
longer than head, the first ray longest, the next rapidly graduate, the outer rays
longer again; caudal forked, the upper lobe a little the longer; anal similar to the
dorsal, under the last rays of the latter.

White, the eyes black

The single specimen of this genus and species known, 22 mm. long, was taken
in the same locality in which the Ochmacanthus {Gyrinurus) hatrachostoma was
caught, among the " pseudo-rhyzomas de Agua-pe" Eichornea azurea, in the margin
of the Paraguay River near San Luiz de Caceres.

Genus XVI". Branchioica Eigenmann."
Branchioica Eigenmann, Proc. Am. Phil. Soc, LVI, Jan., 1918, p. 702.
Ttjpe. — Branchioica bertonii Eigenmann.

It is quite possible that this genus will, on direct comparison of specimens, prove
a synonym of Paravandellia. It has the same general characters, but comes from
the lower Paraguay, while Paravandellia comes from the upper. The present species
was taken from a fish, while Paravandellia seems to be free swimming. It is quite
possible that teeth will be found in Paravandellia at the end of the maxillary (pre-
maxillary?) and on the mandibles when they are examined minutely. Paravan-
dellia is said to have the caudal forked, while Branchioica has it subtruncate.

No nasal or mental barbels, two barbels at angle of the mouth, of which the
lower is minute; first pectoral ray not spinous, not prolonged in a filament; gill-
openings small, the membrane perfectly confluent with the isthmus; mouth inferior;
two series of teeth in the front of the upper jaw, a single series of much smaller teeth
laterad of these; maxillary with claw-Hke teeth at its end, just in front of the barbel
and entirely concealed; two short series of teeth on the ends of the mandibles,
opposite the lateral series of teeth of the upper jaw, the two rami of the mandibles
not meeting; opercular and interopercular patches of spines separate from each
other; caudal subtruncate.

^' ffpayxM, TOi = the gills of fishes; Sikcu = to inhabit.

368 MEMOIRS OF THE CARNEGIE MUSEUM.

The first specimen of this genus was received and described several years ago.
Both the specimen and description were forgotten, the latter never published. The
two specimens 13950 T. U. M. were received much later and independently
described.

2. Branchioica bertonii Eigcnmann. (Plate XLIII, figs. 3-5.)
Branchioica bertoni Etgenmann, Proc. Am. Philos. Soc, LVI, Jan., 1918, p. 703.
13950, I. U. M., type, 24 mm., paratype about the same length over all, much

curved. Taken from a large Characin, Piaractus brachypomus (Cuvier).

Asuncion, Paraguay. Collected by A. de W. Bertoni.
7545, C. M., paratype, 24 mm. Puerto Bertoni, Alto Parana, from the branchia of

Piaractus brachypomus (Cuvier). Bertoni.

Fig. 38. Branchioica bertonii Eigenmann. a, mandible; b, left maxillary with its five teeth; c,
portion of another ma.xillary, showing the proximal teeth only.

Head about 5.5; depth 5.5; D. 10; A. 7; P. 6; eyes superior, nearly the entire
e3^e in the anterior half of the head, 3.5 in the head, about equal to the length of the
snout, considerably larger than the interorbital ; maxillary barbel extending to very
near the interopercular spines, the lower barbel very minute; caudal peduncle
slender, abdomen well rounded; premaxillary with two irregular series of slender,
pointed teeth, those of the posterior series much the larger, about five in number,
subequal, both series graduated from the larger ones nearer the center outward;
laterad of the median series (on the premaxillary?) are four or five similar but smaller
teeth, graduated from the larger proximal one; the rami of the lower jaw widely
separated from each other, each with about five, recurved, pointed teeth in two
series on its end, in apposition to the lateral series of the upper jaw; gill-opening
minute, circular, gill-membranes perfectly confluent with the isthmus; opercle
with a bunch of about twelve, subequal, upward directed spines; interopercle with
about eleven curved, downward directed spines, arranged in two series; distance
from origin of ventrals to caudal 1 .6 in its distance from the snout, origin of anal
behind the origin of the dorsal; distance between anal and caudal about 5 in the
length; pectoral falcate, the outer ray not prolonged as a filament, about as long
as the head; origin of dorsal between that of the ventrals and anal, twice as far

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 369

from snout as from caudal; caudal narrow, oblitiucly rounded or subtruncate, with
few inconspicuous fulcra.

Translucent; eyes black; chromatophcres on the snout, along the back, along
the base of the anal, on the base of the caudal, along the side of the abdominal
cavity, and a few on the pectoral.

Genus XVII. Tridens Eigenmann & Eigenmann. (Plate XXXIX.)

Tridens Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), I, 1889, p. 53.

Type. — Tridens melanops Eigenmann & Eigenmann.

Anal long, with twenty or more rays, its origin in front of that of the dorsal;
ventrals small, nearer to tip of snout than to base of caudal; head greatly depressed,
the eye lateral, infringing on the upper and lower surfaces; a series of fine labial
teeth, stronger teeth in the jaws; gill-membranes united, forming a broad, free fold
across the isthmus; no nasal or mental barbel, two maxillary barbels; opercle and
interopercle armed, the patches of spines separate.

The two species originally jilaced in this genus differ so greatly that they should
probably be placed in separate genera. The .specimens known are all very small,

27 mm. and less.

Key to the Species of Tridens.
o. Depth 13; head 9; D. 10-12; A. 20-25; opercle with three spines; barbels minute, scarcely evident;
distance between origin of dorsal and tip of caudal three in the length; distance between origin of
anal and tip of caudal two and five-tenths in the length; caudal rounded, without accessory rays.

1. melanops Eigenmann & Eigenmann.

aa. Depth 6; head 6; D. 9; A. 22; opercle with 6 or more spines; maxillary barbel extending to the base of

the pectoral; distance between origin of dorsal and tip of caudal two in the length; distance between

origin of anal and tip of caudal less than two in the length; caudal emarginate; eye large, nearer end

of opercle than tip of snout; first pectoral ray greatly produced.

2. brevis Eigenmann & Eigenmann.

1. Tridens melanops Eigenmann & Eigenmann. (Plate XLIII, figs. 1-2.)

Tridens tnelanops Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889,
p. 53 (Iga) ; Occasional Papers Cal. Acad. Sci., I, 1890, p. 339; Proc. U. S.Nat.
Mus., XIV, 1891, p. 37; Eigenmann, Reports Princeton Univ. Exped. Pata-
gonia, III, 1910, p. 401.

Habitat. — Iga, near boundary between Brazil and Peru.
Known from the types, the largest 27 mm., in the Museum of Comparative

Zoology, one of which was received by Indiana University in 1891 and bears the

number 4245. i

Head 9; depth 13; D. 10-12; A. 20-25.

370 MEMOIRS OF THE CARNEGIE MUSEUM.

Body compressed, extremely slender. Head broad, the snout rounded; mouth
broad, inferior. Opercle long and slender, terminating in three spines, trident-
shaped. Pre-opercle with similar but smaller spines. Barbels minute, scarcely
evident. Distance of origin of dorsal fin from extremity of caudal 3 in the length ;
origin of anal fin from extremity of caudal 2.5 in the length. Anal rays rapidly
decreasing in height backward, the last ray about under the last ray of the dorsal.
Caudal rounded, without accessory rays.

Yellowish; posterior half of the caudal fin dusky; a series of black spots along
the base of the anal.

2. Tridens brevis Eigenmann & Eigenmann.

Tridens brevis Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889, p. 54

(Tabatinga); Occasional Papers Cal. Acad. Sci., I, 1890, p. 340; Proc. U. S.

Nat. Mus., XIV, 1891, p. 37; Eigenmann, Reports Princeton Univ. Exped.

Patagonia, III, 1910, p. 401.

Habitat. — Tabatinga.

Known from the type, 21 mm. long, in the Museum of Comparative Zoology.
A recent search for it has failed to locate it.

Head 6; depth 8; D. 9; A. 22.

Body short and deep. Head as broad as long; mouth broad, inferior. Opercle
with a bunch of six or more spines; pre-opercle with a smaller bunch of spines.
Barbels well developed, the outer one extending to the base of the pectoral, the
inner to the gill-opening. Eye large, nearer end of opercle than tip of snout. Dis-
tance of origin of dorsal from tip of caudal little more than two in the length. Anal
inserted very httle in front of the dorsal and extending some distance beyond it,
its rays decreasing in height toward the caudal. Origin of anal from extremity of
caudal less than 2 in tlie length. First pectoral ray greatly produced. Caudal
emarginate.

Yellowish; blackish dots along the bases of the fins; a series of blackish dots
along the middle line of the sides; similar spots on the back. Head with brown dots.

Genus XVIII. Miuroglanis^- Eigenmann & Eigenmann. (Plate XXXIX.)

Miuroglanis Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2), II, 1889, p. 55.

Type. — Miuroglanis platycephalus Eigenmann & Eigenmann.

Anal long, with fifteen rays, its origin in front of that of the dorsal; no nasal or
mental barbel; two barbels at angle of mouth; head greatly depressed, eye lateral,

^' Mflovpos — curtailed; yXaius, 6 = a catfish.

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 371

behind the angle of the mouth; mouth subinferior; several series of strong teeth in
each jaw; gill-membrane broadly united with the isthmus, without a free margin;
opercular and subopercular patches of spines confluent.

1. Miuroglanis platycephalus Eigenmann & Eigenmann.

Mmroglanis platycephalus Eigenmann & Eigenmann, Proc. Cal. Acad. Sci. (2),

II, 1889, p. 56 (Jutahy); Occasional Papers Cal. Acad. Sci., I, 1890, p. 347;

Proc. U. S. Nat. Mus., XIV, 1891, p. 37; Eigenmann, Reports Princeton

Univ. Exped. Patagonia, III, 1910, p. 401; Ribeiro, Fauna Bras., IV (A),

1912, p. 227.

Habitat. — Jutahy.

Known from the type 17 mm. long, collected during the Thayer expedition by
William James. A recent search for it in the collections of the Mus. Comp. Zool.
has failed to locate it.

Head 5.5; D. 10; A. 15. Body short, compressed, and rather deep. Head
greatly depressed, wider than long. Eye large, lateral, placed behind the angle of
the mouth. Mouth subinferior, the upper jaw projecting slightly. Upper maxil-
lary barbel scarcely extending to the gill-opening; no nasal barbels. The opercular
and pre-opercular patches of spines united. Origin of the dorsal fin little behind
that of the anal, its distance from the tip of the snout somewhat less than twice its
distance from the tip of the caudal.

Appendix to the Monograph on the Pygidiid.e.
PHREATOBIUS''^ Goeldi.

Phreatobius Goeldi, Comptes Rendus Congres Intern. Zool., Berne, 1904, p. 549;

Fuhrmann, Verhandl. Schweitz. Naturf. Gesellsch. Aarau, 1905, p. 50;

Archives d,es Sciences Phys. Nat. Geneve (4), 20, 1906, p. 578.

Type. — Phreatobius cisternarum Goeldi.

Origin of dorsal slightly in front of origin of the ventrals, much nearer snout
than to caudal; no nasal barbel; maxillary barbel similar and about as long as the
two mental barbels, placed nearer the anterior nares than to the angle of the mouth;
the mental barbels of each side close together, but remote from their fellows of the
other side, placed directly below the maxillary barbel; mouth terminal, wide, the
lower jaw projecting; teeth in the upper jaw in about three series, in two series in
the lower jaw in front, in one series on the side; the inner teeth the larger and m
very regular series; gill-membrane extending but little above the base of the pec-
cistern; /3ios, 6 = life,

372

MEMOIRS OF THE CARNEGIE MUSEUM.

toral, narrowly joined to the isthmus at a point about half-way between its pos-
terior angle and the snout ; first pectoral ray not spinous ; anal very long, its origin
under the end of the dorsal, its base more than one-third of the length; caudal small,
accessory rays large and numerous, continuous with the anal fin and extending as a
similar fin on the back for two-fifths of the distance to the snout; opercle and inter-
opercle unarmed; eyes rudimentary, near the posterior nares.

This genus is distinguishable by the concomitant elongation of the caudal
portion of the body, the anal fin and the accessory portion of the caudal, by the
position of the dorsal in relation to the ventrals and by the development of the
barbels and absence of opercular armature.

1. Phreatobius cisternarum tJocldi. (Plate LVI, figs. 1, 2 and 4.)
Phreatobius cisternarum Goeldi, Comptes Rendus Congres intern. ZooL, Berne,
1904, p. 549; Fuhrmann, Verhandl. Schweitz Naturf. Gesellsch. Aarau, 1905,
p. 50; Archives des Sciences Phys. Nat. Geneve (4), 20, 1906; p. 578 (alhed to
Clariida', not to Leptosidce and Trychomycericc: fide Zoological Record).

Fig. 39. Phreatobius cisternarum Goeldi. (By permission of Dr. O. Fulirmanii.) *

The generic as well as specific descriptions of this species are drawn from photo-
graphs lent me by Dr. 0. Fuhrmann, and from a specimen 40 mm. long, also sent

EIGENMANN: the PYGIDIID^, a family of south AMERICAN CATFISHES. 373

me by Dr. Fuhrmann, whose generosity, since he himself proposes to pubhsh an
account of the anatomy of the species, I greatly appreciate.
7603, C. M., 40.5 mm. Marajo. From Dr. O. Fuhrmann.

Head about 7; depth about 12; D. 7 (showing in photograph); A. about 25
(showing in photograph).

Heaviest at back part of head, tapering regularly to the base of the caudal, the
depth of which is about one-third that of the head ; mental barbels in pairs, not reach-
ing pectoral, maxillary barbels sometimes to middle of pectoral; pectoral short and
narrow, but little more than half as long as the head ; distance from snout to origin of
ventrals about one and a half in the distance between caudal and origin of ventrals ;
caudal small, rounded, one and one-half in the length of the head; origin of dorsal
in advance of that of the ventrals, its last ray about over origin of anal; upper acces-
sory caudal rays beginning about over the origin of the second third of the anal, the
highest one but little lower than the dorsal rays; anal continuous with the lower
accessory caudal rays; ventrals a little shorter than the pectorals. Color uniform.

Memoirs Carnegie Museum, Vol, VII.

Plate XXXVI,

KNOWN DISTRIBUTION OF THE PYGIDIIN/E.

^ Known distribution of Pygidium. Probably In all
mountain streams north of the latitude of Buenos
Aires and sporadically in lowlands. The localities in
Southern Chile not indicated. .

Known distribution of Hatcheria.
Known distribution of Sc/er'onema.
Known distribution of Eremopkllus.

Memoirs Carnegie Museum, Vol. VII.

Plate xxxvii.

.1 f k" ■ '''

DISTRIBUTION OF STEGOPHILIN/C.
I Homodiwius anisitsi.

Homodiastus macu/atus.

Henonemus macrops.

Henonemus iniermedius.

Henonemus punctatus.

Henonemus taxistigmus.

Stegophilus insidiosus,

Acanthopoma annectens.
/\ Ochmacanthus balrachosiomus.

Ochmacanthus reinhardti.
^^ Ochmacanthus ffabe/hferus.

"•i^SSi.

Memoirs Carnegie Museum, Vol. VII.

Plate XXXVIII.

DISTRIBUTION OF THE VANDELLIIN/E.

Vandellia cirrho^a.

Vandellia ptazai.
I I Vandellia wieneri.

^r\ Vandellia basemani.

Y//^ Vandellia sangumea.

^B Paramndellia oxyptera.

^L Branchioica bertonii.

Memoirs Carnegie Museum, Vol. vil.

Plate XXXIX.

KNOWN DISTRIBUTION OF THE PAREIODONTIN/€.
^B Pareiodon microps Kner.

KNOWN DISTRIBUTION OF THE TRIDENTIN/E.
I I Miuroglanis p/atycephalus E. & E.
^k Tridens melanops E. & E.
|H Tridens brevis E. & E.

376 MEMOIRS OF THE CAHNEGIE MUSEUM.

EXPLANATION OF PLATE XL.

A and B, Eremophilus tnutisii Humboldt; C, Henonevius pimctatus (Boulenger).

1, Premaxillary ; 2, Ethmoid; 3, Lateral Ethmoid; 4, Nasal; 5, Frontal; 6, Sphenotic;
7, Pterotic; 8, Supraoccipital : 9, Epiotic; 10, Supraclavicle ; 11, Parapophysis of coalesced
vertebrEe; 12, Maxillary; 13, Palatine; 14, Metapterygoid ; 15, Quadrate; 16, Pre-opercle;
17, Interopercle; 18, Opercle; 19, Hyomandibular; a, clavicle from the side; b, posterior
face of clavicle.

I— I
XI
w

0-,

o

>

w

w

o

w

-a;
o

CO

PS

K

H

O

^;

w

1-1

u

o

m

CO

s

c

u

^—^

s

a>

>i.

tjl

OS

CO

a

s

l^i

a

c»

a

^5

o

ai

o

CO

n

.

o

p

h-l

o

M

p

<1)

w

13

•■Si

Ol

CO

•^

•(S-

-^J

Si

Ol

00

O

s

^

"T*

^ar'

«,

o

<i)

1-

kq

378 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE XLL

Figs. 1-2. Eremophilus mutisii Humboldt. The air-bladders .show faintly as two
bags, one on either side of the origin of the vertebral column in fig. 2; and as a small
vesicle just above the column in fig. 1. The outlines are marked with a few dots.

Figs. 3-4. Paracetopsis occidentalis (Steindachner). The air-bladder shows as a
large bag in fig. 3. A few points are placed in its wall to call attention to its outline.

The tails in figures 1 and 3 are from the same negatives as the rest of these figures,
but have been printed heavier, as these portions are thin, and the negatives otherwise
quite faint.

X
w

<

>

o

>

w

CO

o

00

td

H

<
a
g

3

02

"13

a.

CO

Q
O

W

a.
o

I

CO

o

Memoirs Carnegie Museum, Vol, VII.

Plate XLII.

Figs. 1-2. Nematogenys inermis (.Guichenot). After Eigenmann. From No. 9839, M. C.
Z., CuRico, Santiago, Chile.

Figs. 3-5. Hatcheria macidata (Cuvier & Valenciennes). After Eigenmann. From No.
7736, M. C. Z., 92 mm. Mapocho, Chile.

X
w

P-i

r

CO

o

>

w

CO

w

w
«

[/J

P5

N

lO

15

w
2

e

I
CO

aj

O

^1

l-H

'O

-tl

04

^

w"

P

a<

^

H

O

O

o

03

CO

T-H

CM

H

W

CO

W

=y

Z

Z

<!

S

2

W

O

w

CO

^i.

o

380 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE XLIV.

Fig. 1. Scleronema operculatum Eigenmann. Type, No. 7077 C. M., 79 mm.,
Cacequy.

Fig. 2. Hatcheria titcombi Eigenmann. Type, No. 11110 I. U. M., 164 mm.,
Arroyo Comajo.

Fig. 3. Pygidium eichorniarum Ribeiro. No. 7560a, C. M., 40 mm., San Antonio,
Rio Guapore.

Fig. 4. Pygidium heterodontum Eigenmann. Type, No. 138.32 L U. M., 83 mm.,
Rio Mendoza, Argentina.

Fig. 5. Pseudostcgophihis 7iemurus (Giinther). No. 7547 C. M., 78 mm., Rio
Mamore.

Memoirs Carnegie Museum, Vol. VII.

Plate XLIV.

sg..^->«;K^gp»iii^«^^ {i'^.-w

*_?J^«» "£t»

r-*."^

'/^,«

Scleronema, Hatcheria, Pygidium, and Pseudostegophilus.
(For detailed explanation see opposite page.)

382 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE XLV.

Fig. L Pygidmm punctatissimum (Castelnau) after Castelnau.

Fig. 2. Pygidium rivulatum (Cuvier & Valenciennes). After Castelnau's figure of
Trichomycterus pentlandi.

Fig. 3. Pygidium rivulatum (Cuvier & Valenciennes). After Castelnau's figure of
Trichomycterus picius.

Fig. 4. Pygidium punctulatum (Cuvier & Valenciennes). After Cuvier & Valen-
ciennes.

Fig. 5. Pygidium dispar Tschudi. After Tschudi.

Memoirs Carnegie Museum, Vol, VII,

Plate XLV.

Pygidium.
(For detailed explanation see opposite page.

Memoirs Carnegie Museum, Vol. VII.

Plate XLVI.

^ ^

wmnm

"//

Figs. 1, 2. Pygidium kneri (Steindachner). After Steindachner.
Figs. 3, 4. Pygidium amazonicum (Steindachner). After Steindachner.
Fig. 5. Pygidium taczanowskii (Steindachner). After the male of Pygidium dispar
TscHUDi, Faun. Peruana, Ichthyol., pl. Ill, lower figure.

Figs. 6-8. Pygidium taczanowskii (Steindachner). After Steindachner.

384 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE XLVII.

Fig. I. Pygidium stellatum Eigenmann. Type, No. 7097 C. M., 78 mm., Quebrada
Sargento.

Fig. 2. Pygidium chapmani Eigenmann. Type, No. 4817, C. M., 106 mm.,
Boquia.

Fig. 3. Pygiddum chapmani Eigenmann. No. 7091 C. M., 86 mm., Rio Dagua at
Caldas. Eigenmann.

Fig. 4. Ptjgidium latidens Eigenmann. Type, No. 1.3801, I. U. M., .53 mm.,
near mouth of Rio Calima, Colombia. (Head too short ; it is 5.5 in the length.)

Fig. 5. Pygidium metw Eigenmann. Type, No. 13770, I. U. M., Barrigona.

Memoirs Carnegie Museum, Vol. VII.

Plate XLVII.

irf.'c- V'iis&

"'•*'*»*"•.«**.*#.;.■.,

"O-*^ ^»^;^__j;ff=^^s

/e iw

s** ^^_

^

/a,

Pygidium.
(For detailed explanation see opposite page.)

386 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE XLVIIL

Fig. 1. Pygidium banneaui Eigenmann. Type, No. 4815, C. M., 43 mm., Bernal
Creek.

Fig. 2. Pygidium spilosoma Regan. No. 7092, C. M., 97 mm., Cordova, Rio
Dagua, Colombia.

Fig. 3. Pygidium dorsostriatum Eigenmann. Type, No. 7093a, C. M., 76 mm.,
Villavicencio.

Fig. 4. Pygidium latistrialum Eigenmann. Type, No. 7450, C. M., 46 mm.,
Quebrada de Pinchote, Santander.

Fig. 5. Pijgidium regani Eigenmann. No. 13772, I. U. M., 55 mm., Tado.

Memoirs Carnegie Museum, Vol. VII.

Plate XLVIII

V

"T'SSFSi^

;;^:^

/....

JJ"^-^ jip. ii»*l.»«*e' >^ v <»- , « g ,

:f**T^T37Sr

'/i in.

'/■t >n.

Pygidium.
(For detailed explanation see opposite page.)

388 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE XLIX.

Fig. 1. Pygidium stramineum Eigenmann. Type, No. 7101, C. M., 46 mm.,
Quebrada del Mango. (The eye is farther forward than in the specimen.)

Fig. 2. Pygidium 7nerida; Regan. No. 13771, 1. U. M., 99 mm., Merida, Venezuela.

Fig. 3. Pygidium bogotense Eigenmann. Type, No. 4820, C. M., 74 mm., Puente
de Supa, near Chapinero, Plains of Bogota.

Fig. 4. Pygidium bogotense Eigenmann. No. 48216, C. M., 70 mm., Chapinero.

Fig. 5. Pygidium nigromaculatum (Boulenger). Univ. Michigan, 165 mm., San
Lorenzo, Santa Marta Mountains, 4,500 ft.

Memoirs Carnegie Museum, Vol. VII.

Plate XLIX.

^.

~.rti;li^ "^'^^^-.

" J -T''

Pygidium.
(For detailed explanation see opposite page.)

I

390 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE L.

Fig. L Pygidium guianense Eigenmann. Type, No. 1003, C. M., 77 mm., Arua-
taima Cataract.

Fig. 2. Pygidium conradi Eigenmann. Type, No. 2212, C. M., 41 mm., Amatuk
Cataract. (The head is a little too short.)

Fig. 3. Pygidium gracilior Eigenmann. Type, No. 1730, C. M., 27 mm., Erukin,
British Guiana.

Fig. 4. Pygidium^ hasemani Eigenmann. Type, No. 5238, C. M., 15 mm., San-
tar em.

Fig. 5. Pygidium iheringi Eigenmann. Type, No. 10785, L U. M., 161 mm.,
Santos.

Memoirs Carnegie Museum, Vol. VII.

Plate L,

-,«4p:.

Vi vv..

, --"^^'is:^

'/a \-^.

Pygidiwji.
(For detailed explanation see opposite page.)

392 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LL

Fig. 1. Prjgidium zonatum Eigenmann. Type, No. 7596, C. M., 62 mm., Agua
Quente.

Fig. 2. Pygidium proops (Ribeiro). No. 7593, C. M., 60 mm., Agua Quente.

Fig. 3. Pygidium paolence Eigenmann. Type, No. 7081, C. M., 68 mm., Alto da
Serra, Rio Tiete, Sao Paulo, Brazil.

Fig. 4. Pygidium reinhardtii Eigenmann. Type, No. 7078, C. M., 65 mm.,
Burmier.

Fig. 5. Pygidium davisi Haseman. Type, No. 2862, C. M., 52 mm., Serrinha
Parana.

Memoirs Carnegie Museum, Vol. VII.

Plate LI.

- 73

/4 »n.

/«>^

'/sin

(For detailed explanation see opposite page.)

394 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LIL

Fig. 1. Pygidiuin immaculatum Eigenmann & Eigenmann. No. 7076, C. M., 81
mm., Rio Doce.

Fig. 2. Pygidium vermiculatum Eigenmann. Type, No. 7074, C. M., 131 mm.,
Juiz de Fora.

Fig. 3. Pygidium alternatum Eigenmann. Type, No. 7079a, C. M., 79 mm., Rio
Doce.

Fig. 4. Pygidium triguttatum Eigenmann. Type, No. 7600a, C. M., 36 mm.,
Jacarehy.

Fig. 5. Pygidium santce-ritce Eigenmann. Type, No. 7599, C. M., 24 mm., Santa
Rita.

Memoirs Carnegie Museum, Vol. VII.

Plate LI I.

1 %

■'^!?-s^-v

is^'

'<^^'^?^sC!^^S^l^^S":i'^'-'0i'iiem'^'

/V in.

/4 in

Pygidium.
(For detailed explanation of Plate see opposite page.)

w

►J
CL,

>^

o

>

w

CO

o
w

PS

<
o

CO

o

S
w

N

K

3

Q
<!

Q

< '^

CO o

• "(3

H s «

-^ (N g
BO

so.
s .o

sic;

6 .

!z; 2: •
^5

ti s ^

^ ^ s

5 w 2
3 o w

<! S -^

O s s

•2 •§ -2

■^ ^ "TS

e g s

e e e

ti- t^ t-

rt <M CO

6 6 6
Ph P^ P^

<

' 'm

—1

o

>

3
:3

o

w

OS

kt

Ife ^ I.

u

o

«

H

EH

<!

I

K

O

O

m

JJ Fh

,

-<

.^^

O

fH •

n

o o

o

O <!

H

'-/',

H z

^g

S

S ^

s

S H

00 r

S

■ R

Q
O

m
P

Humboldt
NER. 145

■c^

:2W

■^^ S.

3

S o

S b

CO

CO Si

3

S '■

^

^ S

-C

-SS o

a.

s.'e

o

=> .2

S

? S

<!■)

^ b

!-.

^ C

tsj feiq Q,

1—1

(N CO

6

6 a

1—4

^—^ ^-i

Uh

ptH fe

>

0-,

^^~:

?>»w--

o

>

S

W

m

o
w

PS

<

Z

;^

B

O

K

tc

H

s

pq

m"

&

«

«<

^

■<

z

o

»^

hM

s

3

1—1

K 1^

H r

§S

6

H

2 Oi

m

g K

Pi

eg-

(B

tf o

!"

^^

^H

O

'5"

m

a 5

o -^

g.

•^ S

a z

e

Z H

3 w

6

o

•— ' cc

-s;

•^ s

o

-S s~

« IS
-5 'S

S rO

CO

•^ B

1

£ *i.

CO ^

s

s s

s

-«; iS

C3

"3 s

1

5j

o

B 1

-S •«

■o u

CO

1

o o

1

.-H

,

^ lO

CO

C5

6 6

P^ Ph fe

GUIDE TO CONTENTS.

PAGE

Introduction 259

Pariolius (Footnote) 260

Habits 261

Distribution of the Pygidiidse (Plates XXXVI-XXXIX) 269

Chronology 270

Location of types 271

Sources of the material examined 274

Zoological position of the Pygidiidse 275

Synonymy 275

Limits of the family Pygidiidse (Plates XL-XLI) 276

Key to the subfamilies and genera (Text figure 1) 278

I. Nematogenys 279

A', inermis (Plate XLII, figs. 1 and 2) 280

II. Scleronema (Plate XXXVI, text figure 2) 280

>S. operculatum : . . . (Plate XLIV, fig. 1) 281

III. Hutcheria (Plate XXXVI, text figure 2) 281

1. H. patagoniensis (Text figure 3) 283

2. H. maculata (Plate XLII, figs. 3-5) 284

3. H. titcombi (Plate XLIV, fig. 2) 284

4. H. areolata (Text figure 4) 285

5. H. burmeisteri (Text figure 5) 286

6. H. macroei (Text figure 6) 287

IV. Pygidium 288

1. P. inarmoralum 290

2. P. palleum 291

3. P. tigrinum 291

4. P. tenue (Text figure 7) 292

5. P. corduvense (Text figure 8) 293

6. P. spegazzinii 294

7. P. borellii 294

?8. P. eichorniarum (Plate XLIV, fig. 3) 295

9. P. riojanum Berg 295

395

396 MEMOIRS OP THE CARNEGIE MUSEUM.

PAGE

10 P. heferodontum (Plate XLIV, fig. 4) 296

11. F. fuscum 298

12. P. eigenmanni 298

13. P. vittatum 299

299
300
300
301
302
303
303
304
305
307
308
309
310
311
312
312
313
314
314
315
315
317
318
319
320
320
321
321
323
324
325
325
326

14. P. dispar (Plate XLV, fig. 5

15. P. pundulatum (Plate XLV, fig. 4

16. P. taczanoivskii (Plate XLVI, figs. 5-8

17. P. rivulatum (Plate XLV, figs. 2 and 3

18. P. poeyanum

19. P. barbouri (Text figure 9

20. P. fassli -.

21. P. oroyce (Text figure 10

22. P. quecJmorum (Text figure 11

23. P. laticeps (Text figure 12

24. P. stellatum (Plate XLVII, fig. 1

25. P. chapmani (Plate XLVII, figs. 2 and 3; Text figure 13

26. P. tcenium - (Text figure 14

27. P. caliense (Text figure 15

28. P. latidens (Plate XLVII, fig. 4

29. P. metce (Plate XLVII, fig. 5

30. P. stramineum (Plate XLIX, fig. 1

31. P. unicolor

32. P. kneri (Plate XLVI, figs. 1 and 2

33. P. meridce (Plate XLIX, fig. 2

34. P. bogotense (Plate XLIX, figs. 3 and 4

35. P. nigromaculatum (Plate XLIX, fig. 5

35. P. banneaui (Plate XLVIII, fig. 1

37. P. spilosoma (Plate XLVIII, fig. 2

38. P. dorsostriatum (Plate XLVIII, fig. 3

venulosum

latistriatum (Plate XLVIII, fig. 4

striatum

regani (Plate XLVIII, fig. 5

retropinne

guianense (Plate L, fig. 1

45. P. conradi (Plate L, fig. 2

46. P. gracilior (Plate L, fig, 3

39.

P

40.

P

41.

P

42.

P

43.

P

44.

P

EIGENMANN: the PYGIDIID^; a family of south AMERICAN CATFISHES. 397

PAGE

47. p. amazonicum (Plate XLVl, figs. 3 and 4) 326

48. P. hasemani (Plate L, fig. 4) 326

49. P. nigricans 329

50. P. iheringi (Plate L, fig. 5) 330

51. P. zonatum (Plate LI, fig. 1) 330

52. P. proops (Plate LI, fig. 2; Text figure 16) 331

53. P. paolence (Plate LI, fig. 3) 332

54.. P. reinhardti (Plate LI, fig. 4) 333

55. P. davisi (Plate LI, fig. 5) 334

56. P. immaculatum (Plate LII, fig. 1) 334

57. P. vermiculatum (Plate LII, fig. 2) 335

58. P. alternatum (Plate LII, fig. 3) 336

59. P. goeldii 337

60. P. brasiliense (Text figures 17 and 18) 337

61. P. itatiayce (Text figure 19) 339

62. P. triguttatum (Plate LII, fig. 4) 339

63. P. punctatissimwn (Plate XLV, fig. 1) 340

64. P. minutum (Text figure 20) 340

65. P. santce-ritce (Plate LII, fig. 5) 341

V. Eremophilus (Plate XXXVI; Plate XL, figs. A and B) 341

1. E. mutisii . . . (Plate XLI, figs. 1 and 2; Plate LIV, figs. 1 and 2) 341

VI. Pareiodon (Plate XXXIX, text figure 21) 343

1. P. microps (Plate LIV, fig. 3) 344

VII. Henonemus (Plate XXXVII) 345

1. H. macrops (Text figure 22) 346

2. H. punctatus (Plate XL, fig. C; Text figures 23 and 24) 346

3. H. taxistigmus (Text figure 25) 348

4. H. intermedius , 349

VII. Pseudostegophilus 349

1. P. nemurus (Plate XLIV, fig. 5) 350

IX. Homodicetus (Plate XXXVII) 350

1. H. anisitsi (Plate LVI, figs. 3 and 5) 351

2. H. maculatus (Text figure 26) 352

X. Stegophilus (Plate XXXVII) 353

1. S. insidiosus (Text figure 27) 353

XI. Acanthopoma (Plate XXXVII) 354

1. A. annectens (Text figure 28) 355

398 MEMOIRS OF THE CARNEGIE MUSEUM.

PAGE

XII. Ochmacanthus (Plate XXXVII) 355

1. 0. batrachostoma (Plate LV, figs. 1-3) 356

2. 0. reinhardtl (Plate LV, fig. 4) 357

3. 0. flabelliferus (Plate LV, fig. 5) 357

XIII. Vandellia (Plate XXXVIII) 358

XIV. Urinophilus 359

L V. cirrJwsa (Text figure 29) 360

2. V. plazai (Plate LIII, fig. 1 ; text figure 30) 362

3. V. wieneri (Text figures 31 and 32) 363

4. V. hasemani (Plate LIII, fig. 3; text figures 33, 34 and 35) 363

5. V. sanguinea (Plate LIII, fig. 2; text figure 36 and 37) 365

XV. Paravandellia (Plate XXXVIII) 366

1. P. oxyptera 367

XVI. Branchioica (Plate XXXVIII) 367

1. B. bertonii (Plate XLIII, figs. 3-5; text figure 38) 368

XVII. Tridens (Plate XXXIX) 369

L T. melanops (Plate XLIII, figs. 1 and 2) 369

2. T. brevis 370

XVIII. Miuroglanis (Plate XXXIX) 370

1. M. platycephalus 371

Appendix

Phreatobius 371

P. cisternarum (Plate LVI, figs. 1, 2 and 4; text figure 39) 372

21. Minute Book of the Virginia Court Held at

Fort Dunmore (Pittsburgh) for the District

of West Augusta, 1775-1776. Crumrine 90

22. Minute Book of the Virginia Court Held for

Yohogania County, first at Augusta Town
(now Washington, Pa.), and afterward on the
Andrew Heath Farm near West Elizabeth,
1776-1780. Crumrine 2.25

23. Minute or Order Book of the Virginia Court

Held for Ohio County, Virginia, at Black's
Cabin (Now West Liberty, W. Va.), &c.
Crumrine 1.50

24. The Eecords of Deeds for the District of West

Augusta, Virginia, for the Court Held at Fort
Dunmore, &c. Crumrine 1.75

25. Astropecten (?) montanus, &c. Douglass 10

26. Astrodon (Pleurocoelus) in the Atlantosaurus

Beds of Wyoming. Hatcher. {Out of Print.)

27. Osteology of the Limicolae. Shufeldt 1.00

28. New Vertebrates from the Montana Tertiary.

Douglass 1.25

29. New Genus and Species of Tortoise from the

Jurassic of Colorado. O. P. Hay 10

30. Osteology of Oxydactylus. Peterson 1.00

31. Birds of Erie and Presque Isle. Todd 75

32. J. B. Hatcher. By W. J. Holland 15

33. Tropidoleptus Fauna at Canandaigua Lake, etc.

Eaymond. {Out of print.)

34. Two Turtles from the Judith River Beds of

Montana. O. P. Hat. {Out of print.)

35. Preliminary List of Hemiptera of Western

Pennsylvania. Wirtner. {Scarce.) 50

36. Trilobites of the Chazy Limestone. Katmond. 1.00

37. Crawfishes of Western Pennsylvania. Ort-

mann 1.00

38. The Geology of Southwestern Montana. Douo-

lass 40

39. New Crocodile from the Jurassic of Wyoming.

Holland 10

40. Procambarus, a New Subgenus of the Genus

Cambarus. Ortmann 15

41. Presentation of Reproduction of Diplodocus Car-

negei to the British Museum. Holland 15

42. Birds Collected near Mombasa by William Do-

herty. Holland 20

43. Hyoid Bone in Mastodon Americanus. Hol-

land 10

44. Additions to Flora of Allegheny County, Pa.

Jennings 15

45. New Species of Kueiffia. Jennings 05

46. Occurrence of Triglochin palustris in Pa. Jen-

nings .05

47. New Species of Ibidium. Jennings 10

48. Agate Spring FossU Quarry. Peterson 10

49. Two New Birds from British E. Africa. Obek-

holser 05

50. The Chazy Formation and Its Fauna. Ray-

mond 1'50

51. A New American Cybele. Narraway and Ray-

mond 15

52. Plastron of the Protisteginae. Wieland 15

53. New Species of Testudo from the Miocene, etc.

O. P. Hat 25

54. Miocene Beds of W. Nebraska and E. Wyoming

and Their Vertebrate Faunae. Peterson 1.00

55. New Species of Lonicera from Pa. Jennings. .05

56. Meryocochcerus, etc. Douglass.

57. New Merycoidodonts. Douglass. (Nos. 56

and 57 sold together.) 1.00

58. Further Collections of Fishes from Paraguay.

Eigenmann, McAtee, and Ward 1.25

59. Undetermined Element in the Osteology of the

MosasauridsB. Holland 20

60. Gasteropoda of the Chazy. Raymond 1.35

61. Occurrence of Wynea Americana in Pa. Jen-

nings 05

62. Account of Pleistocene Fauna Discovered at

Frankstown, Pa. Holland 10

63. Vertebrate FossUs from the Vicinity of Pitts-

burgh, Pa. Case 15

64. Illsenidae from the Black Elver Limestone near

Ottawa, Canada. Raymond and Narraway . . .20

65. Rhinoceroses from the Oligocene and Miocene

of North Dakota and Montana. Douglass.. .25

66. Fossil Horses from North Dakota. Douglass. .30

67. Oligocene Lizards. Douglass 20

68. The Type of Stenomylus gracilis. Peterson. . .25

69. New Species of Birds from Costa Rica, etc.

Cabriker 10

70. Costa Eican Formicariidae. Carriker 05

71. Vertebrate Fossils from the Fort Union Beds.

Douglass *5

72. List of the Lepidoptera of Western Pa., etc.

Engel 1'25

73. Fauna of Upper Devonian in Montana. Fossils

of Red Shales. Raymond 60

74. New Species of Procamelus from Upper Mio-

cene of Montana. Douglass 30

75. Sections of the Conemaugh Series between

Pittsburgh and Latrobe, Pa. Ratmond 35

76. List of the Unionidaa of Western Pa. Ortmann .50

77. Geological Reconnaissance in North Dakota,

Montana, and Idaho. Douglass 1-00

78.. Botanical Survey of Presque Isle. Jennings. . 2.75

79. Sesqui-Centennial (Pittsburgh) Eelics. Stewart .45

80. Dromomeryx, New Genus of American Rumi-

nants. Douglass 30

81. FossUs from Glacial Drift and Devonian and

Mississippian near Meadville, Pa. Millard.. .10

82. New Species of Helodus. Eastman 05

83. Charles Chauncey Mellor. Holland 15

84. Eeports of Expedition to British Guiana, etc.

Eigenmann ^0

85. Eeports of Expedition to British Guiana, etc.

DURBIN 25

86. Odonata of the Neotropical Region, Exclusive

of Mexico and Central America. Calvert... 2.25

87. Deinosuchus hatcheri, a New Genus and Species

of Crocodile. Holland 20

88. Reports on Expedition to British Guiana, etc.

Blosser ^^

89. Description of Some New Titanotheres from

the Uinta. Douglass 25

90. Annotated List of the Birds of Costa Rica In-

cluding Cocos Island. Caerikeb 3.00

91. Geology of the Coast of the State of Alagoas,

Brazil. Branner ^^

92. Fossil Fishes from the Bituminous Shales at Ri-

acho Doce, Alagoas, Brazil. Jordan 55

93. Ordovician TrUobites, No. n. Asaphids from

the Beekmantown. Ratmond 35

94. Ordovician Trilobites, No. ni. Raymond &

Narraway • ^

95. Ordovician Triloljites, No. IV. Raymond 40

96. Fishes from Iikutsk, Siberia. Jordan and

Thompson 30

97. South American Tetrigidse. Brunee 1.00

98. Fauna of the Allegheny and Conemaugh Series

in Western Pa. Eaymond 30

99. Ichthyological Survey About the San Juan

Islands, Washington. Starks 75

100. New Species of Pygidium. Eigenmann 10

101. The Brachiopoda and Ostracoda of the Chazy.

Raymond 50

102. A New Camel from the Miocene of Western

Nebraska. Peterson 15

103. Skeleton of Stenomylus hitchcocki, etc. Peter-

son 15

104. Skeleton of Diceratherium cooki. Peterson . . .15

105. Carnegie Museum Expedition to Central South

America, 1907-1910. Holland 15

106. Report Upon the Expedition of the Carnegie

Museum to Central South America. Hase-
MAN & Eigenmann 25

107. New Species of Fishes, etc., from Central South

America. Haseman 50

108. Annotated Catalog of Cichlid Fishes from Cen-

tral South America. Haseman 1.25

109. New Species of Fishes from the Rio Iguassu.

Haseman 65

110. Ornithology of the Bahama Islands. Todd &

Worthington 75

112. South American Acridoidea. I. Bruner 1.00

113. Species of Hasemania, Hyphessobrycon, and

Hemigrammus. Ellis 25

114. New Chaxacins in the Carnegie Museum. Eigen-

mann 30

115. Jurassic Saurian Bemains Ingested within Fish.

Eastman 20

lie. Autograph Letter of IT. S. Grant to Edwin M.

Stanton. Holland 15

117. Albert J. Barr. By W. J. Holland 10

118. Seventeen New Neotropical Birds. Todd 25

119. Dr. David Alter, the First Discoverer of Spec-

trum Analysis. Holland 10

120. Two Mummy Labels. Allen 10

121. The Families and G-enera of Najades. Obt-

mann 1.00

122. Group of Stenomylins in the Carnegie Museum.

Peterson 25

123. Tertiary Fish-remains from Spanish Guinea.

Eastman 25

124. Plated Nematognaths. Ellis 65

125. New Species of Cambarus from the Isle of

Pines. Ortmann 15

126. Sedum Camegiei, a New Species of the Family

Crassulacese, etc. Hamet 10

127. Two New Species of Fishes from Peru. Eigen-

mann 15

128. South American Locusts (Acridoidea). II.

Bruner 75

129. Revision of the Genus Chsemepelia. Todd 85

130. New Genus and Species of Abyssinian Rodents.

CmLDS Frick 20

131. New Titanothere from the Uinta. Peterson.. .20

132. Small Titanothere from the Lower Uinta. Peter-

son 10

133. Osteology of Lasiopyga and Callithrix, etc.

Shxjfeldt 25

134. New Rhynchocephalian from Solenhofen. Grier .15

135. Scales of South American Characinid Fishes.

Cockerell 26

136. Skeleton of Platigonus Leptorhinus. Peter-

son 10

137. Lichens Collected in the Thunder Bay District,

Ontario. R. Heber Howe, Jr 10

138. Preliminary List of Fossil Plants in the Roof

of the Pittsburgh Coal. Grier 10

139. Undescribed Remains of the Uinta Titanothere

Dolichorhinus. Peterson 15

140. Triassic Fishes Belonging to the Catopteridae

and Semionotidae. Eastman 15

141. Osteology of Promerycochoerus. Petersoi' ... .40

142. Correction of a Generic Name. Peterson 05

143. Skull of Bison Crassicomis. Holland 10

144. Serrasalminae and Mylinae. Eigenmann 50

145. Heads and Tails: Notes on Sauropod Dinosaurs.

Holland 15

146. Dipterus Remains from Upper Devonian of

Colorado. Eastman 10

147. Notes on Tropical American Tettigonoidea (Lo-

custodea). Bruner 1.65

148. New Species of Tortoise from Jurassic of

Utah. Gilmore 15

149. Fauna of Upper Devonian in Montana.

Haynes 45

ISO^New Sphagebranchus from Bahamas. Eigen-
mann 07

151. Marine Fishes from Colombia and Ecuador.

Wilson 15

152. Apareiodon, a New Genus of Characid Fishes.

Eigenmann 18

153. New and Rare Fishes from S. American Rivers.

Eigenmann 25

154. Three New Species of Characid Fishes. Eigen-

mann & Henn 12

155. The Species of Salminus. Eigenmann 7

156. South American Poeciliid Fishes. Henn 40

157. A New Species of Apatosaurus. Holland 7

158. Birds of the Isle of Pines. Todd 70

159. Reptiles and Amphibians of the Isle of Pines.

Barbour 20

160. Land and Fresh Water Shells of the Isle of

Pines. Henderson 20

161. Pelecypoda of the Chazy. Raymond 45

162. S. American Crickets, Gryllotalpoidea and Ache-

toidea. Bruner 1.30

163. Preliminary Catalog of N. American Sphseri-

idae. Sterki 75

164. Directions for Collecting Sphaeriidae and

Aquatic Gastropods. Sterki 10

165. Lepidoptera of the Isle cf Pines. Holland 50

166. Odonata of the Isle of Pines. Kahl 15

167. A Trip to Islands in Lake Erie. Goodrich 15

168. Land-Shells of Islands at Western End of Lake

Erie, etc. Clapp 20

169. Orthoptera of the Isle of Pines. Holland &

Kahl 15

I^'^lk

Publications of the Carnegie Muteum, Serial No. 102.

MEMOIRS

OF THE

OAENEG-IE MUSEUM.

VOL. VII. NO. 6.

W. J. HOLLAND, Editor.

THE AMERICAN DICERATHERES

By 0. A. Peterson

PITTSBUKGH.

PrBLISHED BY THE AUTHORITY OF THE BoARD OF TRUSTEES OF THE

CARNEGIE INSTITUTE.
July, 1920.

PRICE LIST OF PUBLICATIONS OF THE CARNEGIE MUSEUM

ANNUAL EEPOETS OF THE DIRECTOR

1898,30c (scarce); 1899,25c; 1900,30c (scarce); 1901-19, 25c each.

REPORTS OF PROCEEDINGS OF FOUNDER'S DAY

1898-1919, 35c each.

ANNALS OF THE CARNEGIE MUSEUM

The Annals are supplied to those who subscribe in advanc e in parts (paper-bound), as published, @ $3.50 per volume,
Vols. I-XII, 1901-1919, bound in green cloth @ $4.00; bound in J Morocco @ $4.50.

MEMOIRS OF THE CARNEGIE MUSEUM

The Memoirs are supplied to those who subscribe in advan ce in parts (paper-bound), as published, at $10 per volume;
bound in green buckram at $10.75, and in half-morocco at $12.00.

VOL. L (1901-3)

No. 1. Diplodocus, Its Osteology, Tasonomy, etc. No. 3. Osteology of the Steganopodes. Shufeldt $2.76

Hatcher $3.00 ' ' 4. Classification of Chalcid Flies. Ashmead . . 6.50

' ' 2. OUgoceno Cauldee. Hatches 1.50

VOL. n. (1904-6)

No. 1. Osteology of Haplocanthosaurus, etc. Hatcher $2.00 " 6. Osteology of Dlplodocus. Holland 1.50

' ' 2. Osteology of Baptanodon. Gilmore 1.50 ' ' 7. Osteology of Protostega. Wieland 76

" 3. Fossil Avian Eemains from Armlssan. East- " 8. New Suilline Remains from the Miocene of

man 1.00 Nebraska. Peterson .75

" 4. New Bodents and Discussion of Origin of "9. Notes on the Osteology of Baptanodon, etc.

Daemonelix. Peterson 1.75 Gilmore 1.00

No. 5. Tertiary of Montana. Douglass $1.00 ' ' 10. The Crawfishes of Fennsylvania. Ortmann 4.00

VOL. in.

No. 1. Aichseological Investigations in Costa Rica. No. 2. Osteology of Moropus. Holland and Peteb-

Habtman $6.00 SON $6.00

VOL. IV.

No. 1. Barly Chinese Writing. Chalfant $3.00 No. 5. New Carnivores from the Itliocene of West-

' ' 2. Formosan Fishes. Jordan 25 em Nebraska. Peterson $1.50

" 3. Entelodontidae. Peterson 2.50 " 6. Monograph of the Najades of Pennsylvania.

' ' 4. Fishes of Formosa. Jordan and Eichabd- Pts. I and II. Ortmann 2.50

son 1.25 " 7. Catalog Eocene Fishes from Monte Bolea in

Carnegie Museum. Eastman 3.00

VOL. V.

The Fresh Water Fishes of British Guiana. Eiqenmann. $10 unbound; $10.75 cloth; $12.00 } morocco.

VOL. VI.

No. 1. Fishes from Korea. Jordan and Metz $1.50 No. 4. Fishes obtained in Japan. 1911. Jordan

• • 2. Lantem-flshes of Japan. Gilbert l.CO and Thompson $3.50

" 3. Gymnotid Eels of Tropical America. Max Nos. 5-7. Fossil Fishes in the Carnegie Museum,

Mapes Ellis 1.50 Parts II-IV. Eastman 5.00

vn.

No. 1. The Cheirodontinse. Eiqenmann $3.50 No. 4. Plmelodella and Typhlobagrus. Eiqenmann. 2.00

' ' 2. Turtles of Uinta Formation. Gilmore 2.00 No. 5. The Pygldiidae. Eigfjjmann 2.50

No. 3. Ophidia from S. America. Griffin •. 2.00 No. 6. American Diceratheres. Peterson 2.25

VIII.

No. 1. Monograph of Naiadas of Pennsylvania. No. 2. Fossil Plants from Montana. Jennings. In press.

Part III. Ortmann $5.00

REPRINTS FROM THE ANNALS OF THE CARNEGIE MUSEUM

Nos. 1-40. See preceding lists. (Mostly out of print.) 42. Birds Collected near Mombasa by William Do-

41. Presentation of Reproduction of Dlplodocus Car- herty. Holland 20

negel to the British Museum. Holland 15 43. Hyoid Bone in Mastodon Amerlcanus. Hol-
land 10

MEMOIES

OS THE

CAENEGIE MUSEUM.

Vol. VII. No. 6.

THE AMERICAN DICERATHERES.

By 0. A. Peterson.

(Plates LVII-LXVI.)

Introductory.

At the outset the writer wishes to acknowledge his indebtedness to Dr. W. J.
Holland for much valuable assistance and advice in the preparation of the following
paper, and for permission freely to use the material in the Carnegie Museum.
To the authorities of the Peabody Museum of Natural History at Yale University
thanks are due for their courtesy in allowing me to study and illustrate the types
of Professor 0. C. Marsh. I wish to express my gratitude to Professor F. B.
Loomis, of Amherst, for granting me free access to his collection of types and
drawings. I wish to gratefully acknowledge the kindness of the authorities of
the American Museum of Natural History for granting me the privilege of exam-
ining the material from the John Day beds forming a part of the collection of the
late Professor Cope, and allowing me to describe a new species of the genus Dicera-
therium. Thanks are also due to the Librarian of Congress for literature for-
warded for consultation to the Carnegie Museum, to Mr. James W. Gidley, of
the U. S. National Museum, and to Mr. Harold J. Cook for information. Mr.
Syndey Prentice has carefully executed the drawings here reproduced.

399

400 memoirs of the carnegie museum.

Earlier Investigations.

A number of papers dealing with the subject of this memoir have from time
to time appeared, based upon material obtained ]\v different parties, who in the
past two decades have worked in western Nebraska and contiguous territory.
Some of these papers possess genuine value. Other papers have also appeared,
which indicate that the study given by their authors was hasty and of only a
preliminary nature, often containing mistakes, which cause more or less difficulty
to the student. One marked error has been the attribution of specific value to
certain characters of the dentition and other parts, which after a more exhaustive
study are clearly seen to be misleading. It is hoped that the following pages may
prove to be a stimulus to further study and the exercise of greater care in this
field of investigation.

In earlier contributions relating to the Diceratheres published by the author
it has been stated that a more detailed study of the large collection obtained in the
Agate Spring Fossil Quarries would be forthcoming, after the process of extracting
the fossils from the matrix should be completed. Since that announcement much
work has been done. The writer having at his command material consisting of
the remains of some two hundred or more individuals, was induced to question the
validit.y of some alleged specific characters. It is hoped that the following pages
may supply safeguards against error in the future. \Mth only a few specimens
before him, a student may establish species to his own satisfaction upon characters
selected by him at the time, but which after more abundant material is accessible
to him may jirove to be invalid.

In various publications^ there have been reported to be in the Agate Spring
Fossil Quarries and their immediate neighborhood no less than seven species of
the genus Dicer atherium, besides a new genus, Metacoenopus. At first glance it
might appear that the characters relied upon by the authors in establishing the
different forms are valid, but after a more intensive study it is found that some
species must be abandoned, and others must be regarded as doubtful. The result
of our recent investigations proves that in this case we must either condense the
number of proposed species, or establish an infinite number of additional new forms.
The latter course would be eminently unscientific, though justifiable if we accept
as valid the characters employed and relied upon in discriminating the various

1 Loomis, F. B., Amer. Jour. Sci., Vol. XXVI, 1908, p. 51-64. Cook, Harold J., Amer. Naturalist,
Vol. XLII, 1908, p. 543-545; Nebraska Geol. Surv., Vol. Ill, 1908, p. 245-247. Barbour, E. H., Science,
N. S., Vol. XXIV, Dec. 14, 1906, p. 780-781. Peterson, 0. A., Science, N. S., Vol. XXIV, Aug. 31, 1906.
p. 282-283.

PETERSON: THE AMERICAN DICERATHERES. 401

species which have been proposed. Whether the views expressed in the work
upon the Diceratheres here offered shall prove to be conclusive, can only be ascer-
tained in the light of the future. The improbability of having been able to reach
an absolutely final conclusion is abundantly realized by the writer.

Stratigraphy.

Since the earliest descriptions of the European forms of the Diceratheres by
Pomel (1853), Duvernoy (1854), and the final determination of the genus by
Marsh (1875), many papers treating of the Rhinocerotidse have appeared in America
as well as in the Old World. Through the studies of Professor Osborn and Mr.
Hatcher, based upon some early American forms, we learn that the phylum Dicera-
therince had already acquired incipient nasal horns in the White River Oligocene
of South Dakota. It is now known that the forms, not alone of Diceratherium
from the succeeding John Day beds, but all other mammalian remains available
for comparison from the same horizon of the John Day in which Diceratherium
is found, represent an earlier facies than those from the Nebraska-Dakota Miocene.

In order to give conveniently a clear view of the stratigraphic correlation
the diagram on page 402 is inserted.

The Oligocene in South Dakota, as is well known, is much more extensively
developed than in Nebraska. It comprises, besides the three usually recognized
faunal zones, the Titanotherium beds (= Chadron beds), the Oreodon, and the
Leptauchenia beds (= Brule beds), two other easily recognized divisions, one the
Metamynodon beds included in the Brule beds, and the other, the Protoceras
sandstones, both in the Leptauchenia clays which arose from deposits made by
streams. The Miocene section of South Dakota falls into two (Lower and Upper
Rosebud beds), instead of the four divisions, recognized in Nebraska. The four
divisions of the Nebraska Miocene comprise the Gering, the Monroe Creek, the
Lower and the Upper Harrison. The latter is regarded by the writer as the base
of the Middle Miocene. The lower portion of the John Day beds may be regarded
as of transitional character and should therefore be classed as either uppermost
Oligocene or lowermost Miocene, the only difference being that they are not
separated from the INIiddle John Day beds by any apparent stratigraphic break.
The Mascall beds of the John Day are somewhat later than the Upper Harrison
beds of Nebraska, as indicated by a comparison of the faunae.

We know the earlier progenitors of Diceratherium less clearly, though it is
held that Ccenopus occidentalis (Leidy) from the middle Oligocene and Ccenopus

402

MEMOIRS OF THE CARNEGIE MUSEUM.

South Dakota.

Nebraska.

John Day.

European.

m

CM

Snake River
Beds. Sand
and soft sand-
stones, thinly
bedded.

Rattlesnake
Formation.

Upper Rosebud.
Soft sandstone
and sandy clays.

Upper Harrison
Massive soft
sandstone.

Base of

Middle

Miocene

Mascall.
Tuffs, ashes,
etc.

i

Upper and

Middle

Miocene.

Lower Harrison.
Interstratified
with harder
sandstone led-
ges.

Dicera-
therium
cooki;
D. nio-
brarense.

Columbia.
Lava, basalt,
and tuffs.

o

Loiver Ro.'icbud.
Soft sandstone
and sandy clays.

Dicera-

thenium
gregorii
sp. nov.

Monroe Creek
Beds. Hard
sandstone.

Upper John
Day

Lower
Miocene
(Aquita-
nian).

Gering. Soft
sandstones and
sandj' clays.

Dicera-
ther. sp.

Middle and
Lower John
Day.

D. an-

nectens;
D. arma-
tum.

Dicerathe-
rium pleu-
roceros.

Upper Brule
Leptauchenia
clay and Proto-
ceras sand-
stones.

Coenopus
tridacty-
lus; C. da-
kotensis
sp. nov.

Upper Brule.
Leptauchenia
clay.

Lower John
Day transi-
tional from
Miocene to
Oligocene.

Rhinoce-
roses,
spp. ind.

I

Upper Oli-
gocene

(Stampian)

Lower Oli-
gocene
^Sannois-
an).

2;
w

o

2

O

Lower Brule.
Heavy liedded
clays, Metamy-
nodon sand-
stones.

Cwnopus
occiden-
talis; C.
copei.

Lou'er Brule.
Clays and thin

sandstone

ledges.

Protacer-

atherium
{Dicerathe-
rium) mi-
nutum.

Chadron. Clay
and sandstones,
Titanotherium
Beds.

Chadron. Clays
and sandstones.
Titanotherium
Beds.

copei (Osborn) from the lower Oligocene (Metamydon beds) may be looked upon
as at least in the line of this family. '

While the general trend of the characters of Coenopus tridactylus and Coenopus
dakotensis sp. nov. is obviously in the direction of Dicer atherium, it is in the John
Day that we first recognize the genus as occurring in America. The South
Dakotan and especially the Nebraskan series of Diceratheres are a later group
belonging to the lower Miocene, closely following the species of the John Day, while
the so-called R. oregonensis Marsh is an inadequate type, which furnishes infor-
mation altogether too meager to be assigned to Diceratherium as is done by Pro-

2 Osborn, H. F., Mem. Amer. Mus., Vol. 1, 1898, p. 164.

PETERSON: THE AMERICAN DICERATHERES.

403

fessor Loomis. Among Professor Osborn's third phylum of the later Miocene
Rhinoceroses^ we may find a representative of this phylum.

In Europe Diceratherium pleuroceros (Duvernoy) is the most completely
preserved type from the Aquitanian. Its geological horizon apparently approxi-
mates in age the John Day beds of North America.

From the cast of this European species (Fig. 1) it is seen that the cranium back
of the orbit is very suggestive of D. annectens. The brain-case has similar small

Fig. 1. Diceratherium pleuroceros (Duvernoy). From a plaster replica in the Carnegie Museum. X 1/6.

proportions, the supra-orbital ridges converge gently to form a similarly short
sagittal crest, though less prominent and more rounded in the European form.
The inion is also somewhat higher in the latter. The muzzle is long, though
higher, and perhaps having more the proportions of that part in D. niobrarense
from the Nebraska Miocene. The basi-cranium in the cast of D. pleuroceros
is short and the mastoid process is in touch with the post-glenoid process. Thus
the contour of the skull of the European species apparently has combined char-
acters of D. annectens from the John Day and of D. niobrarense from the Nebraska
Miocene. The dentition of the European form is too much worn to allow accurate
comparison. By regarding such forms as Protaceratherium^ {'^Diceratherium")
minutum (Cuvier) of the Stampian as approximately parallel to Coenopus of the
upper and middle Oligocene of North America, it appears that the family may be
traced back to nearly the same geologic time in Europe and North America,^

'Bull. Amer. Mus., Vol. XX, 190-i, p. 321; Aphelops {IPeraceras) planiceps, p. 322; Aphelops
{1 Diceratherium) brachyodus p. 324.

•■Abel, 0., "Kritische Untersuchungen iiber die paliiogenen Rhinocerotiden Europas," Abh. der K.
K. Geologischen Reichsanstalt, Band XX, Heft 3, 1910, p. 10.

* Osborn, Henry F., "Phylogeny of Rhinoceroses of Europe," Bull. Amer. Mus. Nat. Hist., Vol.
XIII, 1900, p. 229-267; "Age of Mammals," p. 90.

404 MEMOIRS OF THE CARNEGIE MUSEUM.

while according to recent work by European authors Prohyracodon orientale Koch

is regarded as the earUest and most primitive representative of the Rhinocerotidse.*^

Diceratherium armatum of the John Day formation has the dentition as well

as certain other features of the skull in a much less advanced stage of development

2 3

Fig. 2. Diceratherium minutum (Cuvier). M- X i. After Cuvier.
Fig. 3. Diceratherium douvillei. M^ X i. After Osborn.

than D. annectens of the same deposit (Compare PI. LVII with text-figure 11,
also with PL LXIII, fig. 6, and PI. LXVI, fig. 1). In the latter form we naturally
might expect to meet with a greater range of anatomical variations, especially in
connection with the dentition. We may reasonably expect to find grinding teeth,
having crests ranging from those which are quite plain to those which have the
various incipient projections, as crista, crochet, anti-crochet, etc. It is far from
my mind to depreciate some, or all, of these characters; on the contrary, indeed,
it is reasonable to expect that the dentition should be one of the first parts of the
organism to undergo modification with a change in the environment. It is never-
theless questionable whether the absence or presence of a crista, a crochet, and
anti-crochet, more or less developed, or of a cingulum of greater or less prominence,
should constitute a valid specific character in Diceratherium. I very much doubt
whether these characters are of sufficient constancy to be relied upon to establish
specific distinctions in a large collection of individuals from a given locality. Stress
has in times past been laid upon the development of branches or spurs of different
lobes of the cheek-teeth. It is plainly evident that D. annectens, as the result of its
mode of life, was already in the time of the John Day more advanced, having filled
out the grinding surface of its teeth more than its contemporary, D. armatum. In
animals representing a later development in geological time, we should expect
to find similar evidence of progression, and in a large assemblage of individuals
that not all the specimens, say of D. cooki for example, are provided with crista
and crochet united on the premolars and with crista small and crochet larger on the
molars, but that these features, being in a plastic stage of development, would he
found in an endless number of combinations from those less developed to those having
more complex forms, and all within one species.

« Abel, O., I.e., p. 24, 44-4.5, 49.

PETERSON: THE AMERICAN DICERATHERES. 405

Another feature, which often has been misinterpreted in connection with the
study of the Diceratheres, is the diiTerence in the contour of the slcuUs. It is a well-
known fact that in individuals of almost any group of mammals the contour of the
skull changes until well after complete maturity. Furthermore the sexual differ-
ences in the form of the skull in the Diceratherine branch of the Rhinocerotidse are
surprising. In the early development of the phylum the difference between the
sexes was well indicated by the form of the skull. ^ This is undoubtedly due in
great measure to the possession of the prominent nasal horns by the male. In a
young male, for instance of D. cooki, there are found the incipient horn-cores, the
nasals are quite long and pointed in front of the horns, while back of the horns
there is relatively small lateral constriction of the nasals, the temporal ridges are
generally weak and not united to form a sagittal crest, the zygomatic arches are
slender, often without, or with very slight, rugosities on the posterior angles.
This is also quite generally true of the skull of an adult female, with the exception
that in the latter there is a gradation from skulls without any horn-cores in the
young, to those having incipient horn-cores in some of the fully adult and old,
and that there is considerable variation in the prominence of the temporal ridges
and the manner of their convergence before they reach the inion. I have as yet
never seen a well-developed and heavy horn-core, the ends of the nasals short and
blunt, the skull much constricted laterally back of the horn-cores, saddle-shaped on
top, with a sudden lateral expansion and heavy rugosities on the posterior angles
, of the zygomatic arches in D. cooki, except in association with well-worn or very
old dentitions. It is very plain to me that more latitude should be assigned to the
significance of the contour of the skull in the genus Diceratheriurn than has some-
times been done. In study and comparison especial pains should be taken (1)
to ascertain whether the skull is that of a male or a female, or of the young, or
not entirely adult animal ; (2) skulls of fully adult or old males are more uniform
in contour than any others, and therefore more reliable in establishing species;
(3) the significance of the crushing received by the specimen in one direction or the
other should be noted.

The following table of comparisons represents fairly well the large number
of skulls of Diceratheriurn cooki in the collection of the Carnegie Museum. Remains
of very young animals are not included in this table, but will be treated separately.
The object of the descriptions given under H, Nos. 2816, 2463, and 2478 in the
following table are to draw attention to the great ease by which misinterpreta-
tions may result with only a portion of the skull in hand and displaying charac-
ters, some of which may be only pathological.

' Osborn, Henry F., "The Extinct Rhinoceroses," Mem. Am. Mus. Nat. Hist., Vol. I, Part III,
1898, p. 162.

406

MEMOIRS OF THE CARNEGIE MUSEUM.

Contour of Skull compared with the Dentition.

Contour of Skull.

Dentition.

A. No. 1572. Type of D.
cooki. Old male.

B. No. 1841. Paratype of
D. cooki. Young male.

C. No. 1923. Young male.

Skull symmetrical in the region
of the parietals, occiput, zygomatic
arches, and palatal regions, but
very slightly depressed by crushing
in the frontal and nasal regions.

Horn-cores prominent, nasals
blunt and constricted back of
horns, frontal region broad and flat,
temporal ridges moderately promi-
nent, but not united to form a sagit-
tal crest, brain-case broad, arches
suddenly expanded posteriorly,
heavy and rugose on posterior angle.
Skull comparatively broad and low.

Skull somewhat distorted by
crushing and otherwise unreliable
on account of immature age.
Horn-cores incipient, nasals pointed,
not projecting over the premaxil-
laries, due partly to crushing, a
considerable constriction of nasals
back of horn-cores; frontals com-
paratively narrow and slightly con-
vex from side to side, due partly to
crushing and partly to immaturity.
Temporal ridges less prominent and
not united to form a sagittal crest,
but quite broadly united with the
occiput, brain case broad, zygomatic
arches expanded posteriorly and
plainly indicating the usual rugo-
sities formed in mature males.
Skull comparatively narrow, partly
due to immaturity and partly to
crushing.

Top of skull more symmetrical
than in No. 1841, but considerable
lateral crushing has taken place,
especially noticeable in the region
of the palatines. Base of skull
open at suture and basioccipital lost.

Horn-cores incipient, nasals pro-
jecting over the premaxillaries,
more than in some specimens,
slightly less pointed anteriorly and
more constricted back of the horns
than in No. 1841; frontals convex
hum side to side and proportionally
narrow, due to crushing and imma-
turity; temporal ridges quite weak
especially on right side, not united
to form a sagittal crest; brain-case
wide, zygomatic arches heavy, indi-
dicating that on further develop-
ment of the skull the rugosities
on posterior angle would be heavy
as in old males generally. Skuil
comparatively narrow and high.

Dentition considerably worn.
Crista of premi)lars worn off, me-
dian valley on P- and P^ isolated by
wear, crochet of P- and P'' united
with ectoloph, cingulum on P^ quite
strong, crochet on P^ not entirely
united with ectoloph; crochet united
with ectoloph, median valley open
and post-fossette isolated on M';
crista slight, crochet heavy, and
post-fossette open posteriorly on M-;
crista weak and crochet strong on

Permanent incisors just appearing
in the alveoli of the premaxillaries.
P' considerably worn, causing the
post-fossette to have already become
isolated. P- very slightly worn,
crista and crochet very slightly de-
veloped and not united on tooth
of the right side, while that on the
left side has crista and crochet better
developed and would on much wear
form an isolated medifossette; anti-
crochet slightly indicated. P' has
crista and crochet c^uite well de-
veloped and united. P* is well
worn, has a small tubercle in the
median valley between proto-and
mctalophs, crochet well developed,
nearly meeting the crista, which is
only slightly indicated on the tooth
of the right side. M' with well de-
veloped crista and crochet nearly
meeting to isolate the medifossette.
M- just erupted, and shows even less
development of crista, but with strong
crochet. M^ buried in the maxillary.

P' in same stage of wear as in No.
1841, P- quite simple, no crista, a
weak crochet, which on further
wear of the tooth would practically
disappear; post-fossette very large.
P^ not as yet worn, both crista and
crochet weak, but showing a ten-
dency to unite so as to isolate the
medifossette on further wear of the
tooth; post-fossette broadly open
posteriori}'. D.P'' much worn, me-
dian valley nearly isolated by wear
of proto- and metalophs; a tubercle
of small size in the median val-
ley; post-fossette isolated. M' with
double, though small, crista; crochet
heavy nearly reaching tlie protoloph;
post-fossette broadly open poste-
riorly. IVP just starting to receive
wear, crista extremely weak, crochet
stronger than usual. ]VP buried in
the maxillary.

PETERSON: THE AMERICAN DICERATHERES.

407

Contour of Skull compared with the Dentition.— Confinwed.

Contour of Skull.

Dentition.

D. No. 2467.
female.

Fullv adult

E. No. 1855. Paratype of
D. cooki. Female.

F. No. 2809. Male.

Skull somewhat crushed back-
ward and to one side, though of
quite symmetrical appearance.

Only a suggestion of horn-cores,
nasals not projecting over the pre-
maxillaries, quite long and pointed
in front of the horn-cores and very
little constricted back of them;
frontals quite broad, though slightly
injured by lateral crushing; tem-
poral ridges prominent, and remain-
ing well apart throughout to the oc-
cipital crest; brain-ca.se large, zygo-
matic arches light. Skull propor-
tionallv high and narrow.

Skull somewhat crushed to one
side.

Nasals broken well back, but the
sides do not indicate as great a con-
striction as in skulls of males.
Frontals broad; temporal ridges
quite prominent and placed quite
close together before reaching the
inion, but not forming a sagittal
crest. Brain-case broad, zygo-
matic arches light. Skull rather
broad and low, to a great extent
brought about by crushing.

Skull slightly depressed by crush-
ing.

Nasal horn-cores very robust,
nasals extend in front of premaxil-
laries, bluntly pointed anteriorly
and gently constricted back of horn-
cores; frontals broad; temporal
ridges prominent and remaining far
apart all the way back to the occi-
put. Brain-case large; zygomatic
arches very robust on the posterior
angle. Skull rather low and broad.

Small functionless and persistent
canines indicated on both sides.
P' much worn. P- with no crista,
but a weak crochet, the latter being
double on tooth of left side; a small
antecrochet present on this tooth of
the right and left jaws and the
post-fossette nearly enclosed. P'
without crista, crochet also poorly
developed, crenulated, and post-fos-
settes isolated by wear. P* with
weak crista and crochet. All the
premolars with strong cingulum in-
ternally. M' with no crista, but
heavy crochet, a minute tubercle in
the median valley; a large post-
fossette and no cingulum on internal
face of the tooth. M- with weak
crista but strong crochet, a very
minute tubercle in the median
valley, post-fossette broadly open
and cingulum only faintly indicated
on the internal faces of the tooth.
M' has received little or no wear.
Crista and crochet very poorly de-
veloped.

P' much worn. P'^ with medi-
and prefossettes quite distinct;
crochet with crenulated border,
post-fossette large. P' quite worn,
but medifossette indicated. Crochet
nearly united with ectoloph. Post-
fossettes quite large. P^ with medi-
fossette (especially on the tooth of
right side) isolated, prefossette and
median valley united. Post-fossette
large. Premolars with heavy cin-
gulum. ]VP much worn, no crista,
crochet nearly united with ectoloph;
post-fossette large, cingulum weak.
M- with strong crista and crochet
(the two nearly meeting on tooth of
right side). Post-fossette large. M'
with crista and crochet much better
developed on tooth of left side,
cingulum weak on molars.

Dentition represented only by P'
and P^ and M' and M^ P' with no
crista; crochet strong, crenulated
internally and nearl)' united with
ectoloph. Post-fossette large. P*
with no crista, crochet strong an
crenulated, as on tooth in advance
of it. Post-fossette large. M' with
no crista, but strong crochet, which
on a little further wear would unite
with the ectoloph. Post-fossette
nearly isolated by wear. M- with
prominent crista and crochet. A
tendency to develop a small tubercle
in the median valley. Post-fossette
large. The premolars have cingulum
better developed than on the molars.

408

MEMOIRS OF THE CARNEGIE MUSEUM.

Contour of Skull compared with the Dentition — Continued.

Contour of Skull.

G. No. 2408. Paratype of
D. cooki. Rather young
female.

H. No. 2463. Old female
with pathfilogical de-
formity.

No. 2816, Male.
No. 2478, Male?

Skull somewhat depres.sed by
cru.shiug, especially over the poster-
ior part of the nasals and the fron-
tals.

Nasals not projected in front of
the premaxillaries, ixiinted, no horn-
cores indicated, but nasals thickened
in this region ; little or no const riction
back of the thickened region; fron-
tals broad and flat; temporal ridges
fairly prominent and early united to
form a sagittal crest before the inion
is reached (the latter is broken off) ;
brain-case broad; zygomatic arches
slender. Skull projiortionally broad
and low, due partly to crushing.

Dentition.

Skull crushed so as to produce an
unusually, high occiput. Frontal
and nasal regions ciuite symmetrical.
Anterior portion of nasals and pre-
maxilhr broken off.

Anterior portion of skull not un-
like that of No. 1572 (type of D.
cooki) in fact the general contour is
similar. However, the sagittal re-
gion is narrower, there being a de-
cided sagittal crest in the present
specimen.

Zygoma lighter, without the ru-
gose area on the posterior angle.
The comparatively light structure
of the skull clearly indicates a fe-
male specimen.

The greater portion of skull No.
2816 is preserved, while No. 2478 is
only represented by a portion of the
top and back.

These two specimens are no less
unique than No. 2463 just de-
scribed.

The chief feature is the inflated
condition of thefrontals, which is not
unlike that in Rhinoceros hicornis,
except that in the fossil specimens
the swollen area is confined more to
the posterior portion of the frontals.
In No. 2478 the inflated area is
more pronounced than in 2S16 and
also differs in the median line from
the latter, 'having this inflated region
continued backward as a promi-
nent rounded ridge between the tem-
poral ridges.

Judging from the heavy and ru-
gose zj'gomatic arches on No. 2816
the skull is undoubtedly that of a
male.

Dentition comparatively little
worn. P' well worn. P^ with medi-
fos.sette isolated, especially on tooth
of right side, tiny crenulation on the
face of the crochet wall, post-fossette
widely open behind. P-* with medi-
fossette isolated, crenulation on
crochet as on preceding tooth, post-
fossette large. P'' with medifossette
isolated, especially on tooth of left
side, post-fossette large; prominent
cingulum on internal faces of pre-
molars. M' with moderate rounded
crista and heavy crochet, but not
united to form a closed medifossette,
post-fossette broadly open poster-
iorly. M^ with more prominent
crista, which very nearly unites with
the crochet, post-fossette large. M^
well erupted, but not yet in contact
with the lower teeth, crista and
crochet rather delicately developed.
Cingulum little developed on the
internal faces of the molars.

P^ of left side represented only by
an extremely thin band of dentine.
P^ and M' closely succeeding one
another, nearly closing up the space
for P''. This was accomplished dur-
ing the life of the animal. M' of
both sides have curious metastyles
located on the postero-internal angle,
a deep fissure separating them from
the main body of the teeth; median
valley open, but crista and crochet
well shown. Dentition nuich worn
(See Fig. 4.)

The dentition of No. 2816 is essen-
tially that of D. cooki, while in No.
2478 there are no teeth represented.

PETERSON: THE AMERICAN DICERATHERES.

409

In comparing (D) No. 2467 in the above table with the description by Dr.
Loomis of his pi'oposed species Diceratherium schiffi it will at once be observed
that while the general contour of the skull agrees fairly well, the dentition totally
disagrees in the presence of the minute canines^ and the difference of the config-

FiG. 4.

Upper dentition of Diceratherium cooki, C. M., No. 2463, showing the reduced condition of
?•* of right side and accessory tubercles on IVP. X i-

uration of the grinding surfaces of the teeth. We are not permitted, therefore,
according to the usually accepted view to refer this specimen to the above pro-
posed species. The same comparison with (G) No. 2408 shows that while the
dentition agrees, the contour of the skull is less in accordance with the above-
mentioned description, and corresponds better with the original type of D. cooki,
sexual characters excepted. With another female skull (E) No. 1855, one of the
original .specimens used as a paratype in my earliest paper, D. schiffi seems to agree
best, except that the temporal ridges come closer together before reaching the
inion. It is thus seen that in comparing female skulls it is frequently found that
dentition and contour of skull do not both agree; on the contrary the material
affords numerous different combinations. There are of course female skulls which
occasionally answer to the description by Loomis somewhat more closely than in
the cases stated above. However, it is quite obvious that we would create a
difficult task for the systematist and student, not to say a non-scientific record
of the subject, were we to establish additional species founded upon our abundant
material. The different patterns of the premolar and molar teeth which were
formerly regarded as satisfactory for the establishment of species are obviously
not to be relied upon, at least not in connection with the study of the material
from the Agate Spring Fossil Quarries. The differences to which allusion is here

* The canines are probably deciduous teeth, which sometimes abnormally persist and their presence
may be regarded as only an individual peculiarity. Professor Owen speaks of a canine in the foetal
skull of Rhinoceros indicus ("Odontography," p. 592).

410 MEMOIRS OF THE CARNEGIE MUSEUM.

made should rather be regarded as in the main due to the varying extent to which
speciahzation has operated in the individual. The teeth, especially of the smaller
American species of the Diceratheres of the Nebraskan Miocene, may be said to be
in a stage of rather rapid and progressive change. It is hardly probable that
we shall be able to perfect any satisfactory adaptive radiation of forms, such as
has recently been suggested,^ from the study of this material. In paleontology
we are debarred from the finer subdivisions used in recent zoology. We have to
content ourselves with characters which stand out more prominently and which may
be used not only to clearly determine species, but to give aid in the question of cor-
relations of faunae and demarcations in geology. From the study of the collection
above tabulated, we are forced to regard the variations shown as being individual,
sexual, juvenile, and pathological.

1. Rhinoceros (?Diceratherium) pacificus Leidy," incertce sedis.

Type. — Upper molar, left side. United States National Mu-
seum.

Horizon. — ? Miocene.

Locality. — "Alkali Flat" John Day region, Oregon.
Fig. 5. Di- Pciratype. — A mutilated fragment of the upper jaw of the right

ceratherium pa. gide, with portions of the fangs of the true molars and an inferior
cificum Leidy, ^^^^^^. ^^^^^

.c^ T j^ Horizon. — ? Miocene.

After Leidy.

Locality. — Bridge Creek, John Day region, Oregon.

As indicated in Leidy 's original description this material from "Alkali Flat"
in the John Day region, Oregon, was provisionally referred to Coenopus (R.) occi-
dentalis. Receiving more material from the same general region Leidj^ again
restudied the "Alkali Flat" specimens and finally placed them, together with the
material from Bridge Creek, under his species R. pacificus.

This type like that of the John Day material referred to as R. hesperius we
now find to be inadequate, or of very doubtful generic value. Leidy was appar-
ently not clear as to the true association of these different fragments and teeth.
On page 222 (I.e.) he states that the second molar described, "may be a true molar
of the preceding species" [R. hesperius] described in his report.

I am unable to agTee with Dr. Loomis in accepting this species as valid and
am obliged, as the result of the study I have made, to regard this type as incertce

' Loomis, F. B., I.e., p. 53.

'» Proc. Acad. Nat. Sci., Philadelphia, 1870, p. 112; 1871, p. 248; U.S.G.S. Terr., Vol. I, 1873, p. 221.
Plates II, VII, Figs. 6-7, 24-25; Ainer. .lour. Sci., Vol. XXVI, 1908, p. 55-56, Fig. 6.

1

PETERSON: THE AMERICAN DICERATHERES. 411

sedis. It pertains to an animal no larger than, for instance, D. annedens (Marsh),
but that is about all I feel justified in positively stating.

2. Rhinoceros (?Diceratherium) hesperius Leidy,'^ incertce sedis.

Type. — A third upper molar. Location of the type uncertain.

Horizon. — Miocene?

Locality. — John Day region, Oregon.

The material from the John Day of Oregon, which Leidy finally referred with
a question to the Californian species "Rhinoceros hesperius,"^- is, as Leidy himself
states, inadequate. The more important features of the remains
of the skull appear to be the position and size of the infra-orbital
foramen and the position of the base of the zygomatic process of the
jugal. Of material referred to R. hesperius Leidy says [I.e., U. S. G.
S., Vol. 1, p. 220) : "The anterior extremity of the space included by Fig. 6. Di-
the zygoma extends to a line with the interval of the second and he.speri um
third molars; in Rhinoceros [Coenopus] occidentalis it extends onlj' to a (Leidy) MK
line with the back part of the last molar. The infra-orbital foramen ^ ■'/
is large and occupies a position above the second premolar; in R. [C]
occidentalis it is over the third premolar." This description agrees with D.
cooki so far as the zygomatic arch is concerned, but the infra-orbital foramen
of the latter species is usually opposite the interval between P- and P^ both
in D. annedens and D. cooki. In D. annedens the space included by the zygoma
referred to by Leidy is slightly more posterior. In comparing the measurements
of well-known species of Diceratherium with the figures of specimens referred
to R. hesperius and R. pacificus it is seen that M'* of hesperius might well
go with the molar of pacificus. So far as the difference in size and even the
configurations of the crowns in these teeth go, there is now no valid reason for
separating the two on the evidence produced. The tubercle of the median valley
of M^ of R. hesperius may well be questioned as a specific character, and is in all
probability, as Leidy suggests, "merely an individual peculiarity." In my opinion
these remains are generically and specifically unidentifiable, but hold the historic
position of being the first material of the Rhinocerotidse obtained in the John
Day region of Oregon.

" Proc. Acad. Nat. Sci. Phila., 1865, p. 176-177; 1870, p. 112; 1871, p. 248. U. S. Geol. Surv.
Terr., Vol. I, 1873, p. 220, PI. II, Figs. 8-9. Amer. Jour. Sci., Vol. XXVI, 1908, p. 55, Fig. 5.

'- Professor Osborn has placed this Californian specimen with Ccenopus platycephalus, "The Extinct
Rhinoceroses," Mem. Amer. Mus., Vol. I, 1898, p. 144.

412

MEMOIRS OF THE CARNEGIE MUSEUM.

No.

<='-^-^

3. Rhinoceros (?Diceratherium) oregonensis Marsh," incertce sedis.

Type. — Penultimate upper molar. Peabody Museum of Natural History

10,002.

Horizon. — ("Pliocene deposits of Oregon") Mascal formation.

Locality. — John Day region, Oregon.

In reference to this fragment Marsh says : ' ... At the union
of the transverse posterior ridge with the outer cusp, there is a
deep cavity, nearly circular, and enclosed by a vertical cylinder
of enamel. The anterior crest, also, is divided, a strong branch
being sent inward and backward from the posterior side into the
main transverse valley."

Whether or not this specimen pertains to a Dicerathere may
never be settled. I have recently examined this tooth and may
state that it may equally well belong to a middle Miocene Rhino-
ceros (Teleoceras) . I cannot now see any reason for regarding this
type as anything except of indeterminate value.

4. Diceratherium (?) truquianum (Cope), inccrtoc sedis.

Type. — A symphysis of the lower jaws with all the incisors and the posterior
portion of the ramus with M2 and M3. American Museum Natural History (Cope
Collection) No. 7333.

Horizon. — Lower John Day, Miocene(?).

Fig. 7. (?Di-
ceraiherium) Rhi-
noceros oregonen-
sis Marsh.

M^ X i.
After Loomis.

Fig. 8. Diceratherium truquianum (Cope). No. 733.3, Coll. Am. Mus. Symphysis and portion of left

ramus. X i.

In describing an incomplete mandible from the ("Truckee beds"") (Lower
John Day) Professor Cope says that the specimen "supports molar, canine
[= lateral incisor], and incisor teeth. . . . The crowns of the canines [= lateral
incisors] are considerably wider than those of the incisors [= median incisors],
but do not project very far beyond them. They are sub-triangular in outline,

" American .Journal Science, Vol. V, 1873, p. 410. Ibid., Vol. XXVI, 1908, p. 00, Fig. 13.
"American Naturalist, Vol. XIII, 1879, p. 333.

PETERSON: THE AMERICAN DICERATHERES.

413

having a prominent shoulder at the base on their inner side. . . . Diastema long;
ascending ramus vertical, flat in front. Depth of ramus at last molar .065; length
of crown of canine [= lateral incisor] .027; width of do. at the base .024."

This type specimen, now in the Cope collection of the American Museum of
Natural History, has recently been studied by the writer. After a comparison
with fragments of the lower jaw associated with a skull (No. 10,005) of Diceratherium
armatum in the Yale Museum I think it possible that this specimen may pertain
to that species. The thick and rather shallow ramus of Cope's type is character-
istic of D. armatum. The symphysis is similarly long and heavy, the mental
foramen is below Pi, as in the latter species, and the comparative measurements
of the two specimens agree fairly well. The question of the relationship of these
two species cannot, however, be entirely satisfactorily settled until more complete
material of the John Day forms is obtained.

Additional Measurements of Type of D. truquianum Cope.

Antero-posterior diameter of crown of median incisor 6 mm.

Transverse " " " " " " 8 "

Height " " " " " 6J "

" " " lateral " 26| "

(( (( a !• (( 90 it

'' " " M^ 44 "

" " M^ 46 "

Transverse

Antero-posterior

Transverse

Anterior-posterior

Transverse

Diceratherium petersoni Loomis^'' incertoe sedis.
Type. — First and second molars of left side. Amherst Museum, No. 1583.
Horizon. — Miocene.
Locality. — Agate Spring Fossil Quarries (quarry A) Sioux County, Nebraska.

Fig. 9. Diceratherium petersoni lioomii^. M- and AP left side. No. 1.58.3, Coll. Amherst Museum. X |.

After Loomis.

In the extensive collection from the Agate Spring Quarries and neighborhood
now in the Carnegie Museum, there are not found any teeth or other remains

'^ Amer. ,Jour. Sci., Vol. XXVI, July, 1908, p. .57, Fig. 7. Cook, Harold J., Neb. Geol. Surv.,
Vol. VII, Aug., 1912, p. 40.

414 MEMOIRS OF THE CARNEGIE MUSEUM.

comparable in size with this species. On the whole the configuration of the crowns
of these teeth, as represented, does not greatly disagree with that of the type of
D. niobrarense; one of the teeth being of a considerably younger individual than
the latter. The size of the species of Loomis is, however, decidedly larger than
D. niobrarense. If this does not prove to be a very large individual of the latter
form, it may be a distinct species; possibly in a more direct line from the large
form D. cjregorii sp. nov., of the lower Rosebud beds of South Dakota (See page 421).

6. Diceratherium armatum Marsh."' (Plate LVII and text-figure 10.)

Type. — Complete skull somewhat crushed dorso-ventrally. Bones of fore
foot associated. Peabody Museum of Natural History No. 10,003.

Horizon. — Lower John Day Formation (?Lowermost Miocene).

Locality. — Near John Day River in eastern Oregon.

As is well known, the genus Diceratherium established by Professor 0. C.
Marsh in 1875 rests on this famous specimen in the Peabody Museum of Natural
History. The type was only briefly described by Marsh. Since that time no

Fig. 10. Diceratherium armalum Marsh., No. 10003, Coll. Peabody Museum of Natural History.

Top of cranium. X 1.

complete illustrations or detailed description of its osteological structure have ap-
peared. For the purpose of more detailed records the writer was accorded the
privilege of studying the type material in the Peabody Museum. The descriptions
and illustrations of the type material follow the modified generic determination
of Diceratherium and the specific characters of D. armatum.

Generic Characters Established by Professor Marsh (Modified).

Males with osseous protuberances on the anterior portion of the nasals. Females
ranging from those with light or incipient nasal protuberances to those with nasals
more or less smooth. Incisors -§•, Canines 1,--^ [in rare cases a very minute canine per-
sists]. Premolars f, Molars f. Fore and hind feet functionally tridactyl.

'« Amer. Jour. Sci., Vol. IX, 1876, p. 242; Ibid., Vol. XXVI, 1908, p. .54, Fig. 2.

PETERSON: THE AMERICAN DICERATHERES. 415

Specific Characters. — Diceratherium armatum may he characterized as follows:
Frontals relatively broad over the orbits. Sagittal crest short. Inion light. Broad and
heavy nasals. Anterior nares not extended as far back of the horn-cores as in D.
cooki. Muzzle and ijremaxillaries long. Postglenoid and paroccipital processes
well separated. Cheek-teeth comparatively simple in the configuration of their crowns.
Animals of larger size than tapirs. [?Median lower incisors proportionally large.]*

General Description. — As stated above, the type specimen of this species is
somewhat depressed by crushing. In the Peabody Museum collection is an
additional specimen, a skull with fragments of the lower jaws, re-identified by
Professor F. B. Loomis. This second specimen has the contour of the top much
better preserved. From these two specimens it is at once observed that the fron-
tals are quite broad over the eyes, which causes a rather short, sharp, emargination
of the muzzle back of the horn-cores. The latter are, as in D. niobrarense, located
near the lateral border of the nasals; they are well developed though somswhat
less in proportion to those in D. annectens of the same horizon. The nasals as a
whole are, however, heavier than in the later forms from Nebraska, D. niobrarense
and D. cooki, while one might look for evidence to the contrary. Even when the
crushed condition of the John Day type is duly considered, it seems clear that the
nasals were not elevated over the premaxillaries as high as in the later Diceratheres
from Nebraska. The temporal ridges are rather weak and the sagittal crest is
not greatly developed and occupies a smaller antero-posterior area than in D.
niobrarense. The inion is not nearly as heavy as in the latter form, resulting in
a comparatively less saddle-shaped top. The nasals are unfortunately broken off
in front of the horn-cores, but enough is present to indicate that they were of con-
siderable length and possibly terminated in a rather sharp point. The premaxil-
laries are also broken off, but what remain of them indicates that they were of
greater length than in the Nebraskan species. There is present a heavy lachrymal
process in D. armatum. The palate is broad as are also the posterior nares. The
anterior margin of the posterior nares extends even with the anterior face of M-.
The post-glenoid and paroccipital processes are well separated, indicating the
condition found in earlier Rhinoceroses. That the alveolar border of the maxil-
lary terminates more abruptly back of M' and the paroccipital process is less
robust than in some of the Nebraskan species are perhaps characters of less im-
portance.

The dentition of the type is perhaps better known than any other part of the

* The only lower jaw of a large species with median incisors present is D. tmquianum Cope. If
the latter species should prove to be identical with D. armatum Marsh, then the above specific character
is valid.

416 MEMOIRS OF THE CARNEGIE MUSEUM.

type specimen. This is due to the studies of Professor Marsh and later students.
Therefore it is only necessary to here state that the ectoloph is perhaps thinner,
the excavations or valleys of the crowns larger, and the cross-crests simpler, than
in D. annedens or later forms from the Nebraska Miocene. P^ is also observed
to be proportionallj^ large, and the cingulum seems to be well developed, especially
on the internal faces of the teeth.

A case of reversion, or at least a non-uniformity of tooth-structure worthy of
note, is seen in the second specimen of D. armatum in the Yale Museum collection
(No. 10,055). The metacone on P- of the right side of this specimen displays a
curious primitive roundness, though connected with the ectoloph by the usual
thin cross-crest, while the corresponding tooth of the opposite side has this postero-
internal tubercle of the usual type seen in the Diceratheres. There is otherwise
little or no differences in the dentition from that in the type, except that No. 10,055
represents a younger animal. The crista, crochet, etc. are very little or not at
all indicated, while the cingula are prominent, especially internally.

With the skull No. 10,055 of the Yale Museum collection, referred to D. arma-
tum, there are associated fragments of lower jaws, which undoubtedly belong
with the skull, inasmuch as the third molars, both upper and lower, arc not yet
entirely developed. These fragments of the lower jaw indicate that the hori-
zontal ramus was very heavy, but rather shallow, the symphysis strong, the mental
foramen of large size, and located directly below Pi. The roots of the lateral
incisors indicate that the crown was large and most likely of the usual type met
with in the family. The symphysis is broken off too far back to show any indica-
tion of the median incisors. Pi has a rather small antero-posterior diameter, but
the crown is quite high ; the tooth is broken externally and the grinding and internal
faces are buried in the matrix. This is also true of P2. The external face of the
latter tooth shows a very heavy cingulum, which extends around the entire pos-
terior face, but has a less upward oblique trend than is seen in the later forms
from Nebraska. The crowns of the cheek-teeth are little worn, indicating the
juvenile stage of the specimen. Mo has also a cingulum on the external face which
is, however, less developed than on P2; this is especially true of the posterior lobe
of M2.

The fourth metacarpal associated with the type of D. armatum is rather long
and broad, having a comparatively small antero-posterior diameter. The bone is
somewhat crushed, but the proximal end is not distorted and indicates that the
bone was not very thick fore-and-aft. The distal trochlea extends well up upon
the anterior face of the metacarpal. Judging from the unciform, which is present,
the carpus was fairly high.

PETERSON: THE AMERICAN DICERATHERES. 417

Measurements of the Type of Diceratherium armatum.
Length of skull from condyles to end of nasals as preserved (Points of nasals broken off) 503 mm.

Length from occipital condyles to M' 20S "

M^ to end of maxillary (Point of maxillary broken off) approximatelj' 166 "

Greatest width across the zygomatic arches 145 "

Greatest transverse diameter of occipital condyles 112 "

Transverse diameter of occipital plate 146

Inferior surface of condyles to end of iuion 159

Length of dentition (molar-premolar series) 248

Antero-posterior diameter of P' 29

Transverse " " P' (greatest diameter) 24 "

Antero-posterior " " p2 " " 31

Transverse " " P- " " 40 "

Antero-posterior " " ps " " 35

Transverse " " P' " " 45 "

Antero-posterior " " P^ " " 38

Transverse " "P^ " " 49 "

Antero-posterior " " M' " " 44 "

Transverse " " M' " " 53 "

Antero-posterior " " M^ " " 53 "

Transverse " " M" " " 53 "

Antero-posterior " " M^ " " 45 "

Transverse " " IVP " " 50 "

Length of McIV 187 "

Greatest transverse diameter of head of Mc IV 51

Transverse diameter midway of shaft of Mc IV 40

Antero-posterior diameter of shaft of Mc IV (approximately) 15

Height of unciform 54

7. Diceratherium annectens (Marsh)." (Plates LXIII, Fig. 6; LXVI, Fig. 1,

and text-figs. 11 and 11a.) (See PL LVIII, Figs. 1, 2, 3.)

Synonym. — Diceratherium nanum (Marsh) .^*

Type.— A set of superior premolars of the left side ; one superior incisor asso-
ciated. Peabody Museum of Natural History No. 10,001.

Hy polypes. Skull nearly complete. Cope Collection, American Museum
Natural History, No. 7324, a male. Front of skull and lower jaws of Professor
Marsh's type D. nanum in the Marsh Collection, Peabody Museum of Natural
History, No. 10,004, a male.

Horizon. — Lower to Middle John Day Formation.

Locality. — Near John Day River in eastern Oregon.

" Marsh, 0. C, Amer. Jour. Sci., Vol. V, 1873, p. 4. Loomis, F. B., Amer. Jour. Sci., Vol. XXVI,
1908, p. 54, Fig. 3.

1* Marsh, 0. C., Amer. Jour. Sci., Vol. IX, 1875, p. 243.

418

MEMOIRS OF THE CARNEGIE MUSEUM.

Specific Characters. — Premaxillaries long and slender. Nasals and nasal horn-
cores of males broad and heavy. Muzzle long. Anterior nares excavated back of the horn-
cores in the same proportion as in D. armatum. A well-defined and quite heavy sagittal
crest. Occiput overhanging, and the cranium well extended back of the posterior angle
of the zygomatic arches. Liberal separation between the postglenoid and paroccipital
processes. First premolar relatively large. Cheek-teeth with swollen cross-crests
and crowns otherwise complicated; crista and crochet present, especially on the pos-
terior premolars and the molar series. Median and lower incisors proportionally
large. Animal about the size of, or larger than, a tapir.

General Description of the Type Material.
From recent studies of the type material of Professor Marsh's collection in
the Peabody Museum and the splendidly preserved skull in the Cope Collection of
the American Museum there is now no valid reason for regarding the types of

Fig. 11. Diceratherium anncctcns (Marsh)- No. 10001, Coll. I'imK.mIv Museum of Natural History.
Premolar teeth of left side and superior incisor. X 5.

D. annectens and D. nanum as belonging to separate species. D. annectens, having
been described before D. nanum, and also being now found to possess sufficient
characters for identification and comparison, must be regarded as the type.

In his description of D. annectens Professor Marsh was apparently not entirely
clear as to the composition of the specimen. Professor Loomis correctly associates
the type, but mistook some of the premolars for molars.

There is no doubt in the mind of the present writer that this series of pre-
molars belongs to one individual. In placing the teeth together one finds that
they fit against one another perfectly and the grinding surfaces form a natural
gradation generally obtained in specimens of D. cooki which have reached an equal
stage of wear. Whether or not the associated incisor tooth belongs to the type
is less satisfactorily determined, as it has received comparatively little wear and
appears small in proportion. The small amount of wear of the upper incisors
is, however, often found in skulls of D. cooki, when the cheek-teeth have been well
ground down.

PETERSON: THE AMERICAN DICERATHERES. 419

P* is very much worn, so that its configuration is practically obliterated.
P- is also much worn, but plainly indicates that the cross-crests are more swollen
than in the larger species, D. armatum, so that on extreme wear of the tooth the
two crests become almost united internally and more nearly approximate the
condition in the Nebraskan form D. cooki; there is a slight indication of a crochet
in P'. P^ has the internal portion of the cross-crests even more closely united,
so that on extreme wear the tooth has a remarkably close similarity to that in
D. cooki. There is, however, no crochet shown in this worn tooth; the crista might
be said to be represented by a heavy fold on the inner face of the ectoloph. P* in
its general characters is practically a repetition of P'*, except that the crista and
crochet are more plainly shown. The crochet of P* of this species represents,
undoubtedly, the most external process of the comb-like plate on the posterior
border of the medifossette in D. cooki; that is to say, the true crochet, which unites
with the ectoloph on extreme wear of the tooth. In the forms of the John Day it
appears that this crochet does not entirely unite with the ectoloph.

No. 10,004 of the Yale Museum Collection (Marsh's type of D. nanum)
is laterally compressed by crushing. As a consequence the nasals appear less
broad than otherwise would be the case, and they are also possibly somewhat
lengthened by crushing. The horn-cores are well-developed and the points of
the nasals are quite heavy, and extend well in front; their tips are broken off.
The nasals as a whole are heavy and are elevated above the premaxillaries much as
in later forms, thus presenting large anterior nares. The infra-orbital foramen is
large, well up upon the maxillarj' and its posterior margin is opposite the middle
of P^ The premaxillary is long; it is also slender, though somewhat heavier than
in the later forms from Nebraska. There is a large upper incisor of the usual
cutting pattern. The premolars are very much worn.

The lower jaws of the same specimen are also slightly crushed laterally.
The most noteworthy feature of these jaws are the proportionally large median
incisors and the long diastemata from the cheek-teeth to the incisors. The lateral
incisor is robust, well sharpened by wear and procumbent in position. The cheek-
teeth are much worn, indicating the senility of the individual. There is a fairly
well developed cingulum on the lower premolars (PI. LVIII, Figs. 1-3.)

As in the lower jaw of D. armatum, the ramus is quite heavy, but somewhat
deeper in proportion. The internal face is also less convex supero-inferiorly.
This latter character may in part be due to crushing.

As stated above, the skull (No. 7324 Cope Collection) in the American Museum
is by far the best of the three specimens here described. (See Pis. LXIII Fig. 6;

420

MEMOIRS OF THE CARNEGIE MUSEUM.

LXVI, Fig. 1 and text fig. 11a.) From this material combined we are now in pos-
session of practically all the anatomy of the skull of D. annectens. With the
exception of the ends of the nasals, the anterior portion of the left horn, and the
points of the premaxillaries this specimen is quite complete. The skull is somewhat
depressed, so that the region about the horns and anterior portion of the maxil-
laries appears broader than in the New Haven specimen described above; however,
the present specimen is in reality more robust. In proportion the horn-cores
of D. annectens from the John Day are considerably heavier than in the later Ne-
braskan species and the tips of the nasals were evidently quite long. The constric-

FiG. llfl. Diceratherium annectens (Marsh). No. 7.324, Cope Collection, American Museum of Natural

History. Hypotype. X {.

tion between the orbit and the nasal horn is, as in D. armatum, much shorter and
sharper than in D. niobrarense or D. cooki, and the occiput extends further back of
the posterior angle of the zygomatic arch and overhangs the occipital condyles
to a greater degree. There is also a well-defined and quite heavy sagittal crest.
The supratemporal ridges are distinct, but more gently oblique or more gradually
converging towards the sagittal crest than in the Nebraskan species, which is
due to the smaller brain-case in the form from the John Day. The occiput is
somewhat more elevated above the occipital condyles and the transverse diameter
of the occipital plate is actually less, though the skull is larger than that of the
average skulls of D. cooki found in the Nebraskan quarries. As stated, the pre-
maxillaries are broken off anteriorly, but it is very evident that the diastema from
the cheek-teeth to the upper incisor in this specimen was as long as in the type
at New Haven. The pre-orbital foramen is located above the anterior, or rather
the middle, region of P'' as in D. cooki, but it is further back of the narial border,

PETERSON: THE AMERICAN DICERATHERES. 421

due to the longer muzzle of the form from the John Day. The zygomatic arch of
the latter is somewhat lighter, especially at the posterior angle, which forms a
less direct right angle with the side of the skull and is not nearly so rugose. The
region back of the pterygoid processes is decidedly longer in the John Day species
than in the eastern species. Thus there is a wide separation between the post-
glenoid and paroccipital processes in the earlier species, while in the later form
{D. cooki) these processes are closely united. Even the occipital condyles of the
John Day specimen are less sessile. The anterior margin of the posterior nares
is opposite the interval between M' and M^ as in the smaller Nebraskan form,
D. cooki, and the size of the nares is of the same proportion.

The premolars of the specimen in the American Museum are even more worn
than in the type, which is brought out in the illustration, PL LXIII, Fig. 6. They
are, however, not too far gone for identification and comparison, and they are
seen to agree with the type in New Haven. M^ has a decided antecrochet-like
swelling of the anterior lobe, which is as great as, or perhaps greater than, in any
of the Nebraskan specimens which I have seen. The crochet is of quite large
size and totally separated from the ectoloph, so that at no stage of wear will this
process apparently ever become united with the ectoloph as in D. cooki. M- has
the crochet generally less developed than in D. cooki. M' has on the right side a
curious basal cusp on the posterior margin of the exit of the median valley which is
very similar to the same tooth in a specimen of D. cooki at the Carnegie Museum,
though less deeply separated from the main body of the tooth. (See Fig. 4, p. 409.)
On M^ of the left side there is also a minute tubercle situated in a position similar to
the one described above. With the exception of the relative size of the median
incisor and the first premolars D. annectens from the John Day and D. cooki from
the Nebraska Miocene differ less in the detailed structure of the dentition than
was anticipated.

On page 422 are tabulated measurements of the specimens above described.

8. Diceratherium gregorii" sp. nov. (Plate LIX and text-figure 12.)

Type. — Skull,? female. American Museum, No. 12,933.

Horizon. — Miocene, Lower Rosebud beds.

Locality. — Near Rosebud Indian Agency, South Dakota.

Specific Characters: Occiput low, but overhanging, as in the John Day form
{D. annectens) Sagittal crest low, but well defined. Postorbital ridges converging
very gradually, as in the John Day form, but the brain-case proportionally larger in
size. Greater robustness of the inion, shorter basicranium and premaxillaries, when

" In honor of Dr. W. K. Gregory, of the American Museum of Natural History, who found the type-

422

MEMOIRS OF THE CARNEGIE MUSEUM.

Measurements of D. annectens.

Type, Y.M.

No. lOtlOl.

Neotype,

Y. M.
No. 1UU04.

Neotvpe,

A. M.
No. 7324.

Length of skull from inion to broken points of nasals

Occipital condyles to and including P'

Occipital condyle.s to M^

Greatest transverse diameter of skull at posterior angle of zygomatic

arches

Transverse diameter of occipital plate. (Measurement taken

superiorly)

Antero-posterior diameter of P', - and ^

Length of upper dentition

Length of premolars

Length of molars .'

Antero-posterior diameter of P'

Transverse

Antero-posterior

Transverse

Antero-posterior

Transverse

Antero-posterior

Transverse

Antero-posterior

Transverse

Antero-posterior

Transverse

Antero-posterior

Transverse

Length from Pi to and including the lateral incisor . .

Length of P2 and P3

Length of Pj

Length of P3

Antero-posterior diameter of crown of median incisor.
Transverse diameter of same

P'..

P2..
P=..
P'..
P^.
P^.
P^.

Ml.
M'.

M\
M-.
M'.
M^

95 mm.

21 mm.
17 mm.
23 mm.
27 mm.
27 mm.
34 mm.
29 mm.
38 mm.

*19 mm.
*1S mm.
*23 mm.
*28 mm.

110 mm.
48 mm.
23 mm.
26 mm.

8 mm.
6 mm.

410 mm.

368 mm.
185 mm.

235 mm.

83 mm.

63 mm.
185 mm.

93 mm.

98 mm.

19 mm.

17 mm.

24 mm.
*28 mm.

28 mm.

35 mm.

30 mm.

39 mm.
35 mm.
41 mm.

40 mm.

41 mm.
33 mm.
38 mm.

* Approximate measurements.

compared with the John Day species, D. annectens. Paroccipital and postglenoid

processes in close proximity to one another as in D. niobrarense. Border of anterior

nares extending further back than in the latter species. Animal considerably larger

than the tapir.

General Description.

The type specimen was discovered by Dr. W. K. Gregory of the American Mu-
seum party of 1906. The skull is somewhat depressed by crushing, which fact
has been taken into due consideration. That the cranium may i)robably be that
of a female should also be noted. The animal was of advanced age, as the den-
tition is greatly worn down and of no practical service for specific determination.

There is no true contact between the broken end of the prema.xillary and the
maxillary bone in the type at present, but Dr. W. D. Matthew assured me that it
was complete when discovered, and that the length of the premaxilla is not far
from correct as restored. (See PI. TJX). Whether or not there was a lateral
incisor, as in Coenopus tridactylus from the Protoceras beds, cannot positively be

PETERSON: THE AMERICAN DICERATHERES. 423

determined from the type. It is, however, most probable that this tooth is wanting,
especially when we consider the proportionally small development of the premaxil-
lary in the type, which is apparently much lighter and is no doubt shorter than,
for instance, in Coenopus tridactylus of the upper Oligocene. The later geological
formation in which this new species was found is also to be considered. The

Fig. 12. Diceratheriurn gregorii. No. 12933, Coll. Ainer. Museum. Top of cranium. X i

nasals were apparently of considerable length in front of the very slight swelling
on the anterior portion of the nasals. The crushing of the anterior region of the
skull gives to the anterior nares an unusually low position, low even when proper
allowance is made for the distortion which has occurred. This may, or may not,
be a valid character. The postorbital ridges of the frontals converge very grad-
ually, somewhat as in the John Day form {D. annectens), but the brain-case is
somewhat larger in proportion. The sagittal crest is low, but well-defined, and
the inion is intermediate between the John Day form and D. niobrarense of the
Nebraska Miocene, that is to say, the rise from the sagittal crest proper to the
top of the inion is very much more gradual than in D. niobrarense, even when the
difference of sex and the crushing sustained by the specimen is taken into account,
thus more like what is seen in D. annectens, but the slight emargination on the
posterior face of the inion is more as it is in D. niobrarense. The inion itself is
less rugose and the lambdoidal crests are thinner than in D. annectens, which may
be a sexual character. The top of the skull when in perfect condition was on the
whole less saddle-shaped; the zygomatic arches lighter, less prominent posteriorly,
and united with the sides of the squamosals more obliquely than in D. cooki. The
postglenoid and paroccipital processes are in touch with one another, but are
not so completely fused as in the latter species. The external ear is large and in
shape more nearly as in 7). niobrarense.

The incisor present in the premaxilla is of unusually small size in comparison
with the cheek-teeth and the size of the skull. P^ has about the same relative

424 MEMOIRS OF THE CARNEGIE MUSEUM.

size as in D. niobrarense. As stated above, the cheek-teeth are much worn. The
last molar does not indicate any crochet or crista as in the last named species; the
cingulum is, however, equally prominent. M'* on the right side has a small cone in
the median valley.

It is very likely that the above specific characters may be modified, or added
to, when more material representing both sexes of this species is obtained.

Measurements.

Greatest length of skull from inion to the end of the premaxillaries approximately 'AO mm.

From occipital condyles to anterior end of maxillary, approximately 490 "

Occipital condyles to P' 362 "

Occipital condyles to M' 230 "

From incisor to orbit 197 "

From orbit to occipital condyles 310 "

Incisor to and including M' 266 "

Greatest transverse diameter of skull at posterior portion of zygomatic arches 258 "

Greatest transverse diameter of frontals 191 "

Greatest transverse diameter at constriction back of the enlarged portions of the nasals 97 "

Length of molar-premolar series 217 "

Length of premolar series 102 "

Length of molar series 119 "

Antero-posterior diameter of P' (greatest diameter) 21 "

Transverse " " P' " " 20 "

Antero-posterior " " P^ " " 26 "

Transverse " " P^ " " 32 "

Antero-posterior " " P^ " " 31 "

Transverse " " P^ " " 44 "

Antero-posterior " " P< " " 34 "

Transverse " " P-" " " 48 "

Antero-posterior " " M^ " " 38 "

Transverse " " M' " " 4.5 "

Antero-posterior " " M'^ " " 45 "

Transverse " " M^ " " 47 "

Antero-posterior " " M' " " 38 "

Transverse " " M^ " " 44 "

Antero-posterior " " crown of incisor 19 "

Transverse " " " " " 9 "

9. Diceratherium niobrarense Peterson.'-'^ (Plate LX, Fig. 2; PI. LXI, Fig. 2;
PI. LXII, Fig. 2, and text-figure 13.)
Synonym. — M. (Aceratherium) egrerius}^

-1 Science (N.S.), Vol. XXIV, 1906, p. 28; Ann. Car. Mus., Vol. IV, 1906, p. 46, Pis. XIII-XIV;
Loomis, F. B., Amer. Jour. Scl., Vol. XXVI, 1908, p. 56.

^' Loomis, F. B., ibid., p. 61 [Aceratherium, egrerius]; Cook, Harolil ,1., Amer. Naturalist, Vol. XLII,
1908, p. 543, 2 figs [Aceratherium egregius]; [Metacocnopus cgregius]; Neb. Geol. Surv., Vol. Ill, 1908,
p. 245, PI. I.; Vol. VII, 1912, p. 41.

PETERSON: THE AMERICAN DICERATHERES. 425

Type.SkuW of young male. C. M., No. 1,271.

Horizon. — Miocene .

Locality. — Agate Spring Fossil Quarries (Quarry A.) Sioux County, Nebraska.

Paratype. — Posterior portion of skull from same quarry as the type. Ver-
tebrae and limb-bones referred to same species.

Specific Characters. — Premaxillary somewhat reduced in length. Grinding
surface of cheek-teeth comparatively simple. Nasals long in front of the horn cores
especially in females. Muzzle long. Border of anterior nares comparatively little
extended backward. Skull quite saddle-shaped, especially in males, due to the develop-
ment of the horn-cores and the high inion. Postorhital ridges less oblique than in the
John Day forms due to the enlargement of the brain-case. A sagittal crest present;
this is proportionally long, but not especially strong. Zygomatic arches somewhat
more expanded posteriorly and the basi-cranium shorter than in earlier John Day
forms. Paroccipital and postglenoid processes sometimes touching one another at
their bases so as to enclose the external auditory meatus. Lower jaws heavy and the
angle little or not at all everted. Animal smaller than D. armatum of the John Day

formation.

General Description.

Since the earlier descriptions of this species the type has been restudied.
Illustrations, which in some respects are more accurate than those which appeared
earlier, are also herewith presented. I, furthermore, add data recently obtained,
and have corrected certain errors, which occurred in earlier publications.

In Science (I.e.) it was stated that the nasals were found in the talus below the
point where the skull was taken out. The nasals were separated from the skull at
the fronto-nasal suture, but agree with the skull found in situ, with the correspond-
ing parts missing. I at once associated the difTerent parts as those of one indi-
vidual, and have not since found any reason for changing my mind. Confirming
my view, a good skull of this species in Dr. Loomis ' collection at Amherst has the
fronto-nasal suture quite open, as in the type. Dr. Loomis assures me that the
nasals belong with the skull of his specimen, which is of approximately the same
age as the type. (See Fig. 13.)

In the original description it was said that the brain-case is large, while Loomis
states that the brain-case is comparatively small, a statement which only holds
good so far as the present species and D. cooki are concerned. From the earlier
John Day forms D. niobrarense may be distinguished by its having the brain-
cavity of larger size. I stated that there is a well-formed sagittal crest, but I did
not especially emphasize the fact that the crest is strong. From the Amherst

426 MEMOIRS OF THE CARNEGIE MUSEUM.

material it appears that there is some variation in this respect, judging from the
statements of Dr. Loomis. The statement made by the latter author that the
nasals project considerably beyond the horn-cores is characteristic of this species,
while in D. cooki the points of the nasals are much abbreviated in fully adult or
old males.

The muzzle of the skull in D. niobrarense is apparently not shorter than in the
John Day forms, while the constrictions back of the horn-cores and in front of
the orbits are longer and gentler, due to the relative narrowness of the nasals
across the horn-cores and the narrower frontals. The location of the infra-orbital
foramen is similar to that in D. armatum, located well back from the border of

Fig. 13. Dicer alherium niobrarense Peterson, No. 1022, Coll. Amherst Museum. X i- After an

outline drawing by Dr. F. B. Loomis.

the anterior nares due to the slight backward excavation of the latter. The
occiput is, however, not overhanging, as in the John Day forms, while the external
ear is sometimes enclosed below.

As has been shown by Loomis and Peterson the molar-premolar dentition of
D. niobrarense is more primitive than in D. cooki, and more nearly like that of
D. armatum. In the latter species, which is clearly an older and rather primitive
type, we find that P^ is somewhat larger, the ectoloph of the grinders is thinner,
the different valleys wider, and the cingulum perhaps somewhat more prominent.
To judge from the scanty remains of D. armatum which we possess, it certainly
is indicated that the crista is practically wanting, while the crochet is in a very
much more rudimentary stage of development on the teeth of the latter species
than in D. niobrarense.

In comparing the descriptions and figures of Aceratherium egrerius, later
called Metacoenopus (I.e.) by Mr. Harold J. Cook, it is very evident that the remains
of an adult female of Diceratherium niobrarense has been used as the type. Cook
admits that there is a "thickening of the nasals at the point where a horn visually
occurs in the Rhinocerotidse, which may indicate a rudimentary horn." Indeed
one should expect to find this thickened condition, and we usually do find it in

PETERSON : THE AMERICAN DICERATHERES. 427

the .young males and in female skulls of D. cooki of more adult stages. It appears
that in males of D. niobrarense the nasal horn-cores are located more laterally and
point more outward,-' while in D. cooki they are nearer the median line and point
more directly upward. (See PL LXII.)

With the exception of the longer nasals in front of the thickening portion or
the incipient horn-cores (undoubtedly a sexual character), Mr. Cook's description
agrees quite well with the type of D. niobrarense.

The premaxillaries are complete in this splendidly preserved specimen in
Mr. Cook's collection, and show some reduction in length from those in the older
John Day forms.

There is a considerable portion of the left mandible in Mr. Cook's specimen,
which was found in an articulated position on the skull. Cook states that this
mandible is "heavier and lacks the outward turn or flange commonly found in
the Diceratheres." A splendid pair of lower jaws in the Loomis collection at
Amherst (see Fig. 14) referred to D. niobrarense also agrees with the characteri-

FiG. 14. Diceratherium niobrarense Peterson, No. 1022, Coll. Amherst Museum. Internal view of
ramus. X i- After an outline drawing by Dr. F. B. Loomis.

zation in Cook's paper, and plainly indicates that the masseteric muscle was
much less developed in this species. Unfortunately the outline drawing kindly
furnished by Professor Loomis from the Amherst specimen does not indicate the
external view of the mandible.

The atlas and axis of Mr. Cook's specimen were found attached to the occipital
condyle and are so illustrated in his paper. No description in detail of these
vertebrae is, however, furnished.

A series of cervicals (atlas, 3d, 4th, and 6th), an anterior dorsal (?5) and three
lumbar vertebrae were found isolated in the same quarry (quarry A) in which
the type of D. niobrarense and the Amherst material was found. These bones,
No. 1910, are here provisionally referred to D. niobrarense, inasmuch as the size

''^ Figures on Plate I of Cook's illustrations show admirably well these lateral eminences although of
an incipient stage most likely due to sex. Cook states that there is no double-horn tendency in his type.

428

MEMOIRS OF THE CARNEGIE MUSEUM.

corresponds very well to the type. All these vertebrae, except the atlas, are more
or less mutilated, but enough is preserved to show that they are very similar to
those bones in D. cooki, size excepted. In its ]iroportions the atlas referred to is
not unlike that of the smaller form [D. cooki), save that the transverse process
is less extended forward and is somewhat heavier, especially along the terminal
border. A second marked difference of this bone in the two species is the presence

Fig. 1.5. Diccralheriwm niohrarense Peterson, No. 1910, Coll. Carnegie Museum. Anterior and ventral

views of atlas. X i.

in D. niohrarense of a round venal foramen, (remnant of the inferior exit of the
vertebraterial canal) on the ventral face at the base of the transverse process, and
a strong antcro-posteriorly directed ridge immediately internal to the foramen.
While there is in D. cooki a groove and occasionally a minute foramen, located
in the same position as the foramen described on the atlas of D. niohrarense, there
is found no evidence of the ridge in D. cooki. If the heavy terminal border of the
transverse process and the venal canal are constant in D. niohrarense, this may be
regarded as an additional specific character. (See Fig. 15.) The third cervical

PETERSON: THE AMERICAN DICERATHERES. 429

has the ventral keel of the same proportions as in the smaller species, and the same
faintly indicated neural spine, and the strong transverse process similarly expanded
lateraUy. Cervical four has the neural spine as prominent, but the back part of
its transverse process is perhaps somewhat heavier than in D. cooki. Cervical six
again has the same downwardly directed inferior lamella of the transverse process,
which is, however, proportionally less developed fore-and-aft than in D. cooki.
The dorsal vertebra referred to presents the same characters as the corresponding
bone in the latter species, including the mammillary process on the upper anterior
surface of the transverse process. With the exception of a somewhat more de-
cided ventral keel and possibly a lighter spine, the lumbar series associated are of
approximately the same relative size and detailed structure as in the smaller form,
D. cooki.

The remains of the limb-bones (No. 1910) which were found in this same
quarry, and are provisionally associated with the type of D. niobrarense, show that
the scapula is proportionally shorter than in the smaller species, and possibly also
somewhat broader; the coraco-scapular notch shorter and shallower; the bicipital
groove of the humerus smaller; and the shaft of the ulna straighter. The propor-
tionate length of the femur cannot positively be ascertained from the material at
hand, but the tibia is decidedly shorter. The remains of the foot-bones asso-
ciated bear no marked differences from those of D. cooki, except their larger size.

The type (No. 1040) of " Metacoenopus" (Aceratherium) stigeri in the Amherst
College Museum is described by Dr. Loomis, and the right upper molar-premolar
series is illustrated. Dr. Loomis says: "The small skull is elongated, light in
build and rather narrow. The orbit is large and the zygomatic arch light. The
premolar teeth are crowded, there being neither an anterior nor posterior cingulum,
though one is developed along the inner face around the protocone, running out
on the hypocone. Crista and crochet are wanting on these teeth of a rather old
individual, except that on the fourth premolar there is a faint trace of a crista,
and on the third premolar a small antecrochet is developed. On the molars the
cingulum is reduced as in the premolars, and both crochet and crista are wanting.
The protocone, however, is swollen, making a considerable fold as in European
Diceratheres. A. stigeri is closely related to A. egrerius but is smaller, and has
the cingulum on the premolars and the crochet on the molars less developed."

In the description of this complete skull the greatest stress is laid on the
configuration of the grinding surface of the teeth, principally for the purpose of
comparison. The illustration indicates an animal of old age, as Loomis states.
The crista, if there was one, has consequently disappeared by wear. The crochet

430 MEMOIRS OF THE CARNEGIE MUSEUM.

most likely has likewise disappeared, or rather united with the ectoloph, while
the cingulum may be lighter or even wanting. The cingulum has united (by the
wear of the tooth) with the internal border of the grinding face, especially on the
posterior portion. It is very evident, from the enclosed post-fossette that the
dentition illustrated by Loomis is very much more worn than that of the type of
D. niobrarense. Finally through the courtesy of Professor Loomis. I had the
opportunity of restudying his type and find that the Amherst specimen as well
as the description agree quite well with the average old female skulls of D. cooki
in the Carnegie Museum.

Measurements of the Type of D. niobrarense.

Greatest length of skull, approximately 450 mm.

Length of skull from occipital condyle to and including P^ 370 "

Length of skull from occipital condyle to M' 190 "

Greatest transverse diameter of skull 235 "

Greatest transverse diameter of brain-case 130 "

Greatest transverse diameter of frontals 150 "

Greatest transverse diameter of occipital condyles 103 "

Greatest transverse diameter of palate 68 "

Vertical diameter of the orbit 60 "

Length of 2d, 3d and 4th premolars and the molar series 185 "

Antero-posterior diameter of P^ 26 "

Transverse diameter of P- 29 "

Antero-posterior diameter of P* 32 "

Transverse diameter of P'' 36 "

Antero-posterior diameter of M' 39 "

Transverse diameter of M' 37 "

Antero-posterior diameter of M' 35 "

Transverse diameter of M' 39 ''

Scapula, approximate height 280 "

Humerus, length 340 "

Tibia, length (No. 1910) 335 "

Tibia, length (No. IQlOo) 310 "

Tarsus, height tuber of calcaneum not included, approximately 85 "

Tarsus, length of tuberosity of calcaneum 58 "

Metatarsal II, length 130 "

Metatarsal III, length 145 "

The occurrence of the material of the above described species in the Agate
Spring Fossil Quarries is of considerable interest. Diceratherium niobrarense
has only been found in Quarry A. The remains of this species may be said to be
practically absent in the representative fauna of the quarries in the Carnegie or
the University Hills (See Mem. Car. Mus. Vol. IV, Fig. 1, p. 205), while similar

PETERSON: THE AMERICAN DICERATHERES. 431

remains representing D. cooki of these latter quarries are quite as abundantly
mingled with the remains of D. niobrarense in Quarry A. The latter quarry is
only a very short distance (300 yards) to the north of the main quarries and may
possibly represent a somewhat earlier time; or, more probably, the sediments
accumulated at this spot represent a different stream, which had its origin in, and
flowed through a locality more favorable to this species.

10. Diceratherium cooki Peterson.-^ (Plates LX, LXI, LXII, fig. 1, LXVI,
figs. 2, 4).

Synonyms. — Diceratherium arrikarense Barbour; D. schiffi Loomis; Acera-
therium stigeri Loomis; Diceratherium aberrans Loomis; D. loomisi H. J. Cook.^^

Ti/pe.— Skull of old male. CM. No. 1572.

Paratypes. — Eight skulls, a number of lower jaws, and other skeletal material
C. M. Nos. 1573, 1575, 1581, 1841, 1848, 1853, 1855, 1888, 2408, 2443, 2799.

Specific Characters. — Skull, especially of males, short and broad in its proportions.
Median lower incisor small. Crowns of the upper cheek-teeth coynpUcated. Crochet
often uniting with ectoloph in much worn teeth. Muzzle short. Horn-cores of males
prominent, but nasals not broad across the horns, and ends of nasals abbreviated.
Females varying from those with no horns to those with incipient horns. Postorbital
ridges seldom entirely meeting to form a sharp sagittal crest. Brain-case large. Occi-
put quite broad, of moderate height, and not overhanging. Premaxillaries short.
Margin of anterior nares much excavated, extending well back of the horn-cores and the
infra-orbital foramen situated close to the border. Posterior point of zygomatic arch
greatly expanded laterally and covered with heavy rugosities in fully adult or old males.
Basicranium short. Post-glenoid and paroccipital processes united to enclose the
external auditory meatus. Animal considerably smaller than D. niobrarense.

General Description.

Skull.

As stated in the original description, this species rests on a number of skulls,

lower jaws, and other skeletal material from which a male skull No. 1572 was

originally selected as the type. In the first description it was stated that "the

occiput is rather low, . . . the temporal ridges quite prominent, not uniting

-^ Science (N.S.), Vol. XXIV, Aug. 31, 1906, p. 282-283; Annals Car. Mus., Vol. IV, 1906, p. 47,
PI. XV, text-figs. 12-13; Vol. VII, 1910, p. 274-279, PI. LXV; Loomis, F. B., Araer. Jour. Science, Vol.
XXVI, July, 1908, p. 58.

-'D. arrikarense Barbour, Science (N.S.), Vol. XXIV, Dec. 14, 1906, pp. 780-781, figs. 1, 2; Acera-
therium stigeri Loomis, Amer. Journal Sci., XXVI, July, 1908, p. 60; Dieeralherium schiffi Loomis, I.e.,
p. 57; D. aberrans Loomis, I.e., p. 59; D. loomisi Cook, Harold J., Neb. Geol. Surv., VII, p. 48-32, figs. 2-3.

432 MEMOIRS OF THE CARNEGIE MUSEUM.

to form a sagittal crest, but continuing separate to the inion where they join the
lambdoidal crest." Loomis on the other hand states that the skull is relatively
short and high, with high occipital crest and a moderate sagittal crest, formed by
the confluence of the two ridges from over the orbits. This mistake is undoubtedly
due in part to the illustration (Ann. Cam. Mus., Vol. IV, 1906, p. 48, Fig. 12)
which does not accurately represent the top of the skull. This is remedied in the
illustration given with this paper, PL LXII, Fig. 1.

Loomis further states that "on the premolars, the cingulum is greatly reduced,
while the strong crochet is united with the feeble crista, thus isolating the median
fossette. In like manner on the molars the cingulum is reduced to traces on the
front, inner side, and rear of the teeth." While this description was undoubtedly
based on the material in the Amherst Museum and answers some individuals of this
species in the Carnegie Museum, the type specimen does not agree with his de-
scription. In the first place the internal faces of the premolars, except P- are
incomplete. The said premolar has a prominent cingulum on the anterior and
internal faces of the protocone, which is confluent by wear with the grinding
face of the metacone. On the antero-internal angle of P'' a prominent cingulum
is indicated, the inner face of the tooth is otherwise, as stated, broken off as is
also P* on the anterior inner angle (See PI. LX, Fig. 1.) M^ and M^ are also dam-
aged in this same region, but enough is preserved to indicate that the cingulum
is as prominent as is the case in other skuUs, which are more complete in this
respect. P- and P'* have the crochet united with the ectoloph through wear;
no crista is shown, while the post-fossette is entirely isolated on P\ P"* has only a
trace left of the crochet, but no crista; post-fossette nearly enclosed. M' has
crochet united with the ectoloph by wear, and post-fossette enclosed. M- has a
strong crochet and the crown is injured in the region of the crista. M^ has a
strong crochet and a fairly prominent crista. The entire dentition is much worn,
plainly indicating an old individual. The distinguishing characters of Acera-
therium stigeri and Diceratherium schiffi given by Dr. Loomis now appear to rest
entirely on sexual characters and individual variation, the type of his proposed
species being female skulls of D. cooki, while his species D. aberrans is established
on D. P.^ of the left side as has already been pointed out.-'^

From Professor Barbour's description and figures of his proposed species
D. arrikarense in Science, N.S., Vol. XXIV, 1906, p. 780, it is clear that he has
described a male skull of D. cooki minus the dentition, while Mr. Cook's proposed

"Peterson, 0. A., "Recently PiMpdsed Specie.? of the Genus Diceratherium," Science (N.S.), Vol.
XXXVI, 1912, p. 801.

PETERSON: THE AMERICAN DICERATHERES.

433

species, D. loomisi, is also established on a maxilla of D. cooki with deciduous
teeth. (Neb. Geol. Surv., VII, 1912, p. 29.)

In comparing the crania of the abundant material of D. cooki with the earlier
John Day forms, or even with D. niobrarense fovmd in the same beds in which
D. cooki is found in Nebraska, it is at once clear that D. cooki is a comparatively
more specialized and modified type of the Diceratherina'. We find in the male
skull (1) a pair of prominent horn-cores set closely together on the nasals; the
nasals themselves not nearly as heavy as in the earlier John Day forms; and the
ends of the nasals much abbreviated as in more specialized or modified forms of
the TitanotheridcB] (2) muzzle, premaxillaries, and the front of the lower jaws
shortened and the lateral margin of the anterior nares extended further back of
the horn-cores; (3) brain-case enlarged; occiput broadening and not overhanging;
basicranium short, analogous to Teleoceras from the middle Miocene and the recent
Rhinoceroses {R. bicornis) ; (4) zygomatic arches much expanded with heavy rugo-
sities on the posterior angle, and the angle of the lower jaws heavy and very much

Fig. 16. Dicerathcrium cooki Peterson. No. 1853, Coll. Carnegie Mu.seum. Skull of an old female. X i-

everted to support heavy masseteric muscles ; (5) the grinding surface of the cheek-
teeth further complicated with a tendency on the part of the crochet to become
united with the ectoloph, especially in teeth having received some wear.

As has already been stated in the introductory paragraphs of this paper, a
greater range of variability must be allowed in dealing with female crania of this
species. They range from those without horns in very young and immature
females to those with incipient horns in fully adult and old individuals. The
top of the skull is consequently comparatively little concave antero-posteriorly
and the supratemporal ridges vary so much in their course to the inion that this
region of the cranium may be said to range from a broad surface to a completely
formed sagittal crest (this variation of the supratemporal and sagittal ridges
holds good even in males, though to a somewhat less extent). Another feature
of the female skull, no less noticeable, is seen in the longer pointed nasals, the

434 MEMOIRS OF THE CARNEGIE MUSEUM.

lighter zygomatic arches without the heavy rugosities on the posterior angle, and
the less everted angle of the lower jaws. In connection with these characters
the skeletal frame is lighter and the pelvic cavity proportionally somewhat larger,
judging from the material at hand.

Mandible.
(PL LXVI, Figs. 2 and 4.)

As stated elsewhere-" the lower jaws are heavy, especially those of the males.
The depth of the ramus, however, is not great, and the diastema between the lateral
incisor and P2 is rather short. The symphysis, though short, is very heavy, and
the median suture is entirely obliterated in old individuals. The angle is very
greatly everted; in males the border of this everted area is very rugose, while in
females and young the angle is less everted and is also less rugose, but still furnishes
an unusual heavy surface for the masseteric muscles. The glenoid condyle is quite
broad transversely and the coronoid process is strongly directed forward.

Dentition. — In proportion the median incisor is extremely small; it has a
rounded enamel-covered crown, and is implanted in the symphysis by a thick
short root; in many specimens this rudimentary tooth has dropped out. The
lateral, or cutting incisor, is, as usual, comparatively large, but varying in different
individuals. The canine is generally absent, even in the young. Occasionally
this tooth is present in young individuals, and sometimes there is found on the
alveolar border of the diastema a shallow groove, or scar, in which a minute canine
is found in a very procumbent position. Pi is absent. In young individuals a
small milk-tooth is often found immediately in front of Po which persists in the
alveolar border until all the cheek-teeth are erupted. On PI. LXVI, Figs. 2 and 4
the characters of the cheek-teeth are well shown, and require no detailed de-
scription.

Only a very few hyoid bones are found mixed with the general mass of ma-
terial in the quarries, there evidently having been during deposition too much
disturbance for the preservation of these delicate parts.

In this connection it is interesting to turn to the work of Professor Henry
F. Osborn, "The Extinct Rhinoceroses" Mem. Amer. Museum Natural History,
Vol. I, 1898, pp. 136-140. In this publication, Osborn characterizes Ccenopus
{Aceratherium) mitis from the lower Oligocene as follows (p. 139): "Ci", P3, M3,
Diastema short. Canine" alveoli semi-procumbent. Premolar-molar series

" Peterson, 0. A., Ann. Car. Mus., Vol. VII, 1910, p. 275.
" [= Lateral incisor.]

PETERSON: THE AMERICAN DICERATHERES. 435

142 mm. Mandibular symphysis very short. Locahty Colorado. Amer. Mus.,
Cope Collection, No. 6325."

This type specimen is accompanied by an upper maxillary with teeth and
other skeletal material, but, as there seems to be some doubt as to their association,
they will not here be considered.

A second specimen in the American Museum Collection from the upper
Oligocene (Protoceras Beds) illustrated on page 139 of Osborn's work is especially
interesting. This lower jaw (No. 1110) is doubtfully referred to Coenopus {Acera-
therium) mitis Cope, and exhibits just such characters as one might expect to find
in an ancestral form of D. cooki: the short symphysis; the short diastema between
incisor and cheek-teeth; the curving of the lower border of the ramus; and the angle
everted as in D. cooki. The ramus itself is, however, deeper, the vertical ramus
having a gTeater antero-posterior diameter, and the coronoid process a more nearly
vertical position than in D. cooki, as indicated by the illustration. The measure-
ments (p. 140) do not appear to agree completely with the illustrations.

This lower jaw undoubtedly represents a distinct species, judging from the
great vertical depth of the sediments lying between the Titanotherium beds in
which Coenopus mitis (Cope) was found, and the Protoceras sand-stones, together
with the general change in the fauna of these two geological horizons. This
lower jaw from the Protoceras beds may here be provisionally regarded as the type
of Coenopus dakotensis sp. nov. and also provisionally placed in the line more or
-or less directly leading to D. cooki found in the lower Harrison beds of Nebraska
as indicated in the table found in the introduction to this paper.

Vertebral Column.

The vertebral formula is provisionally given as follows: Cervicals, 7; Dorsals,
19 (?); Lumbars, 5; Sac7'als, 4-5; Caudals, 26.

Cervical Vertebrce (Figs. 17-21). In a fully adult animal the width of the
atlas is almost double that of the length. The anterior projection of the trans-
verse process extends well forward. The neural canal is of moderately large size,
while the arterial canal on the ventral face of the transverse process is generally lack-
ing. The median area of the neural arch varies in different individuals in robustness
and rugosity. The transverse process and the median tubercle of the lower pos-
terior face of the body also vary in robustness and size.

The odontoid process and body of the axis are heavy; the neural spine is
generally heavy and overhanging, while the transverse process projects rather
strongly backward. The latter process is subject to much variation in size, as is

436

MEMOIRS OF THE CARNEGIE MUSEUM.

also the ventral keel. The arterial canal is indicated by a deep groove on the
anterior border of the pedicel, which is often found completely bridged over by a

Fig. 17. Diceratherium cooki Peterson. No. 18.53, Coll. Carnegie Museum. Ventral and anterior views

of atlas. X h

thin splint of bone. Additional features of this bone are well shown in the illus-
trations. (See Fig. 18.)

The third, fourth, and fifth cervicals are, as usual, very uniform in their details

Fig. is. Diceratherium cooki Peterson. No. 1853, Coll. Carnegie Museum. Lateral anil posterior

views of axis. X i-

of structure. There is, however, no neural spine on the third; the fourth has a spine
more or less clearly indicated ; the fifth a spine of considerable size varying in length
in different individuals.

The sixth cervical is characterized by the broad, thin, hatchet-shaped, inferior
lamella of the transverse process, which sometimes terminates in a rounded process

PETERSON: THE AMERICAN DICERATHERES.

437

behind. The superior, or transverse process proper, is located well up on the
centrum; it is trihedral in cross-section, rather short, and projects strongly back-
wards. The neural spine of this vertebra is quite high, attenuated, of great antero-
posterior diameter, and terminates rather abruptly.

Fig. 19. Diccratherium cooki Peter.son. No. 2499, Coll. Carnegie Museum. Lateral and posterior views

of fourth cervical. X i.

The neural spine of the seventh cervical is quite high and generally terminates
in a sharp point. The transverse process is abruptly reduced and there is no
vertebrarterial canal.

Fig. 20. Diccrntherium cooki Peterson. No. 2499, Coll. Carnegie Museum. Lateral and posterior views

of sixth cervical. X i.

Dorsal Vertebrce (Figs. 22-24). We are not yet in a position to positively
state the number of vertebrae in the dorsal series. In the skeletons articulated
and those assembled for articulation and sent to other institutions by the Carnegie
Museum there have been inserted nineteen. This number is thought to be ap-

438

MEMOIRS OF THE -CARNEGIE MUSEUM.

proximately correct, as it corresponds with those in Coenopus tridactylus Osborn
which was found, articulated, in the upper Oligocene of South Dakota.-*

The first dorsal vertebra is characterized by the proportionally large and high

Fig. 21. Fig. 22.

Fig. 21. Diceratherium cooki Peterson. No. 2499, Coll. Carnegie Museum. Lateral and posterior views

of seventh cervical. X h-
Fig. 22. Diceratherium cooki Peterson. No. 2499?, Coll. Carnegie Museum. Lateral and posterior

views of first dorsal. X 5.

neural spine and the depressed centrum; the prezygapophyses are also much ex-
panded laterally to receive the postzygapophyses of the last cervical. The articu-
lations for the ribs are located very low on the side of the centrum and the pedicel
is broad and heavy.

With the exception of the skull, there is probably no other part of the skeleton
in this species which is subject to greater variation than the neural spines of the
dorsal vertebrae. In specimens undoubtedly referable to adult and old males

28 Osborn, H. F., Bull. Amer. Mus. Nat. Hi.st., Vol. V, 1893, p. 85.

PETERSON: THE AMERICAN DICERATHERES.

439

the first dorsal spine is high, broad, and rugose, as shown in Fig. 22, while in
many specimens, fully adult and old, this spine is 50 mm. shorter, and sometimes
even more. In the anterior dorsals the curvature of the neural spine also varies
from a comparatively straight spine to one with a gentle sigmoid curve. The
latter are generally those with the longer and heavier spines. The neural spine of
the second dorsal is suddenly reduced in size, but back of the second the reduction
is more gradual. The anterior dorsals have short, broad, and depressed centra,
while further back they are higher, narrower, and terminate ventrally in better
defined keels. The intervertebral notch is deep and in the posterior upper side
of the centrum it continues downward in a broad and well-defined groove, princi-
pally due to the greatly elevated border of the capitular facet on the centrum.

Fig. 23. Fig. 24.

Fig. 23. Diceratherium cooki Peterson. No. 2499, Coll. Carnegie Museum. Lateral and posterior

views of tenth dorsal. X i-
Fig. 24. Diceratherium cooki Peterson. No. 2470a, Coll. Carnegie Museum. Lateral and posterior

views of eighteenth dorsal. X i.

In the neighborhood of the eighth, ninth, and tenth dorsals there is usually a fora-
men formed at this notch, which is characteristic of all posterior dorsals except
the last. (See Figs. 23-24.) The fourteenth and fifteenth dorsals have the neural
spines broader and more lumbar-like; the mammillary processes, so characteristic
of the transverse processes of the dorsals are also longer and project forward in
these vertebrae to a greater extent. The inferior aspect of the centra vary from a
gently rounded to a more decided ventral keel, possibly due to sex.

Lumbar Vertebrce. Figs. 25-27. There are five lumbar vertebrae. This
causes this part of the spinal column to be rather short. In general outline the

440

MEMOIRS OF THE CARNEGIE MUSEUM.

centrum of the first lumbar is not unlike that of the posterior dorsals, while further
back the lumbar vertebrae are more depressed and gradually broadened. The
last is broader than long. The last lumbar vertebra is otherwise conspicuous on

Fig. 25. Dicerolhcrium cooki Peterson. No. 2470a, Coll. Carnegie Museum. Posterior view of second

lumbar. X l-

account of its suddenly reduced neural spine. This reduction is principally in the
fore-and-aft direction so that there are broad vacuities between the spines in front
and behind.

The more noticeable variations in the lumbar series result from the presence
or the absence of an articulating buttress between the fourth and fifth lumbars.

Fig. 26. Diceratherium cooki Peterson. No. 2470«, Coll. Carnegie Museum. Lateral and posterior

views of lumbar four. X \.

This articulation is located on the transverse process (posterior face) of the fourth,
and meets a corresponding surface on the anterior face of the process on the fifth
lumbar. (See figs. 26 and 27.) The first lumbar is sometimes found to possess
an unusually long transverse process, which tapers rapidly and is rib-like.

PETERSON: THE AMERICAN DICERATHERES.

441

Sacrum. Fig. 28. There are four and often five coossified centra of the
sacrum. The two first vertebra? support the ilium, while those in the rear have
sharp lateral edges and gradually taper toward the caudals. The neural spine of

Fig. 27. Diccratherium cooki Peterson. No. 2470a, Coll. Carnegie Museum. 1, lateral view of fifth
lumbar; 2, anterior view of same; 3, posterior view of same. X i-

the first sacral is generally quite slender, but further back the spines are more
robust. The sacral foramina are of large size and the coosification between the

Fig. 28. Diceratherium cooki Peterson. No. 2797, Coll. Carnegie Museum. Lateral and ventral views

of sacrum. X i-

442

MEMOIRS OF THE CARNEGIE MUSEUM.

centra is complete, especially in fully adult or old individuals, there being little
or no trace of a suture.

Caudal Vertebrae. Twenty-six vertebrae have been attributed to the tail.
This is thought to be approximately correct, inasmuch as seventeen were found in
consecutive order from the first to and including the seventeenth. The seventh
caudal is the last with a complete neural arch, and the eighth is the last with
traces of a transverse process. The tail as a whole is moderately long and tapers
to a fine point as indicated by many very small vertebrae found in the collection.

Ribs. — The ribs, even the posterior, are rather long. In the anterior region
they are flat, though not broad, while further back their cross-sections have a
tendency to be more trihedral. There are well-defined tubercular facets throughout
the entire series. Altogether the thorax forms a rather solid cylinder.

Fig. 29. Diceralherium cooki Peterson. No. 2817, Coll. Carnegie Museum. 1, Dorsal view of sternum;
2, ventral view of same; 3, lateral view of same. X h

Sternum. Fig. 29. The sternum was described in part as follows:'-^ 'The
manubrium is an elongated, laterally compressed, and vertically deep plate of
bone. Anterior to the contact for the first pair of ribs there is a long heavy process,
extending forward, and constituting the greater half of the entire length of the
presternum. Posteriorly the bone is slightly expanded transversely and has a
rough surface for the attachment of the mesosternum. The first two segments
of the latter are somewhat deeper than wide. The j^osterior end of the fourth
sternebra is nearly square in outline, while the fifth and sixth are broader than deep.'

2S Peterson, 0. A., Ann. Car. Mus., Vol. VII, 1911, p. 277.

PETERSON: THE AMERICAN DICERATHERES.

443

Fore Limb.
Scapula. The scapula is long, narrow, and recurved. It is perhaps some-
what narrower than in earlier types (C. tridadylus, Osb.) and nearer the propor-
tions found in more recent forms (R. pachygnathus or R. bicornis Wagn.) In
general the outlines are very similar to those of these species. The coracoid is
prominent, the supra-scapular notch quite deep. The spine, which nearly equally
divides the supra- and infra-spinous fosss, terminates in a very heavy and retro-
verted process. There is a third fossa at the coracoid border immediately above
the suprascapular notch, which is separated from the supraspinous fossa by a

Fig. 30. Diccraiherium cooki Peterson. Xo. 2473, Coll. Carnegie Museum. External views of scapula.

X i.

somewhat prominent vertical ridge. The fossa itself, however, is rather shallow
and of relatively small size. (See Fig. 30). There is very iittle variation in the
details of the scapula in fully adult animals, robustness and size excepted.

Humerus. The humerus is short and heavy. The tuberosities of the proximal
end, though not as heavy in proportion as in some of the recent Rhinoceroses
{R. bicornis), are nevertheless, very prominent and the bicipital groove has a
tendency to become double, i.e., separated by a broad, but very low ridge, approach-
ing the condition in the recent Rhinoceros where the bicipital tubercle is more
prominent. Distally the bone has furthermore a great transverse diameter due to

444

MEMOIRS OF THE CARNEGIE MUSEUM.

The anconeal fossa is

the large entepicondyle and prominent supinator ridge,
very deep and of considerable height (See Fig. 31.)

The proportionate length of the radius and ulna is approximately like that of
Rhinoceros bicornis. In the fossil form the shafts of both radius and ulna are, how-

FiG. 31. Diceratherium. cooki Peterson. No. 2473, Coll. Carnegie Museum. Posterior and anterior

views of humerus. X J.

ever, flatter than in the African species. In fully adult and old individuals this is
chiefly due to the rugose and prominent ridges, which actually come in contact with
each other throughout the whole length of the bones, while in the recent form the
median region of the shafts is rounded and the two bones are separated by a consider-
able space. The shaft of the radius is quite straight, while the ulna as a whole is much
bent backward, especially the upper half, a characteristic seen in the recent form.
The carpal articulations differ from those in the African form to a marked degree.
Thus the lunar articulates exclusively with the radius in Diceratherium, while
in R. bicornis it encroaches to a considerable extent on the distal face of the ulna.
On the other hand it is observed in a number of cases in Diceratherium that the
cuneiform extends over upon the radius, forming a minute facet on the extreme
ulnar border as well as on the palmar face. There is a considerable variation,
especially in length and robustness, of the fore-arm of D. cooki, which is undoubtedly
due to sexual and individual variation as well as age.

Manus. Pis. LXIII; LXIV. The height and breadth of the carpus are

PETERSON: THE AMERICAN DICERATHERES.

445

practically equal, the height being sometimes very slightly greater than the breadth,
while in the recent rhinoceros the breadth is a little greater than the height.
The proximal facets of the scaphoid and cuneiform are not unlike those in recent
forms, while the lunar lacks the facet for the ulna, so plainly shown in the African
species. The large facets and the heavy palmar hook of the lunar uniting the
lateral bones on the proximal row of the carpals in R. hicornis are conspicuously
absent in Dicer atherium. The second row of the carpals in the latter are higher than

Fig. 32. Fig. 33.

Fig. 32. Diceratherium cooki Peterson. No. 2473, Coll. Carnegie Museum. Radial and dorsal views of

ulna. X h
Fig. 33. Diceratherium cooki Peterson. No. 2499, Coll. Carnegie Museum. Oblique ulnar view of

radius and ulna. X i.

those in R. hicornis; otherwise there are only minor details of difference between
the two forms. With regard to size, the trapezium and metacarpal V have ap-
proximately the same proportions as in the living species. The three functional
metacarpals in D. cooki are decidedly longer, slenderer, and the shafts of II and IV
are more curved. This curvature of the shafts of Mc II and IV is to conform to

446

MEMOIRS OF THE CARNEGIE MUSEUM.

the sides of Mc III to which bone they he rather close with comparatively little
divergence distally. All the phalanges are broad, short, and depressed. (See
PI. LXIII).

In Coenopus tridactylus the metacarpals also are close to one another, but
the lateral metacarpals are heavier in proportion than in D. cooki. This is also
true of the bones in the hind foot. Another noticeable feature of the hind foot in
C. tridactylus is seen in the proportionally larger size of the entocuneiform, which is
not remarkable, when we consider the differences in size of the lateral metapodials
in the two genera here compared.

Hind Limb.

The pelvis is short and broad. The area for the gluteal muscle is broadly
expanded, but the supra-iliac border is emarginated as in Coenopus of the Oli-

Fio. .34. Diceratherium cooki Peterson. No. 2797, Coll. Carnegie Museum. Left half of pelvis, dorsal

view. X h

gocene. The ischium and pubis are relatively short when compared with Coenopus
tridactylus, indicating quite an advance in the direction of the recent Rhinoceroses.

PETERSON: THE AMERICAN DICERATHERES.

447

The acetabulum is well rounded and deep; the pit for the round ligament is quite
deep and the cotyloid notch broad. The obturator foramen is very large and
ovate in outline. The sciatic notch of the ischium is well-defined by the sudden
termination of the spine and the heavy and suddenly upward, or outward, turned
tuber ischii. The prominence of the latter tuberosity is subject to some variation

Fig. 35. Fig. 36.

Fig. 35. Diceralherium cooki Peterson. No. 2460, Coll. Carnegie Museum. Posterior and anterior

views of femur. X i.
Fig. 36. Diceratherium cooki Peterson. No. 1840, Coll. Carnegie Museum, 1, anterior view of tibia

and fibula; 2, posterior view of same. X 3.

in different individuals. In fact there is in this species a considerable degree of
variation in the robustness of the pelvis in the large collection before us. This,
in my judgment, is due to age and sex.

The femur is quite long. The shaft, when seen from the front, is quite straight,
but on a rear view appears curved, due to the prominence of the different trochanters
(See Fig. 35.) The third trochanter, though large, is not the long, forward, and

448 MEMOIRS OF THE CARNEGIE MUSEUM.

outward extending process seen in R. bicornis. Nor is the proximal end so much
expanded laterally as in the African species.

The tibial border of the rotular trochlea is greatly developed; it has very nearly
reached the extreme modernized stage of development seen in the recent Rhinoceros
and the horse. The antero-posterior diameter of the distal end is therefore greater
than the transverse in approximately the same proportion as in R. bicornis.

The patella is triangular in general outline, due to the large development of
the internal process in order to cover the greatly developed internal border of the
rotular trochlea described above. The trochlear grooves of the patella are quite
uneven in size and the bone as a whole unlike that of the horse.

In fully adult and old individuals both ends of the tibia and fibula have a
strong tendency to become coossificd. This is a direct indication of the progressive
development which has reached its culmination in the completely united tibia and
fibula of R. bicornis. Like the femur, the tibia and fibula are rather long and slender,
when compared with these bones in the recent species, and it also appears that

Fig. 37. Diccrathcrium cooki Peterson. No. 1840, Coll. Carnegie Musonm. Posterior and anterior

views of patella. X i.

these bones in Coenopus tridactylus (Osborn) are proportionally shorter and pos-
sibly also somewhat heavier than in D. cooki.

Pes. Pis. LXIII, LXIV. As is already known, the pes is strictly tridactyl.
It is on the whole narrow and quite high, especially when compared with R. bicornis.
It is also somewhat higher and slenderer than the pes in Coenopus tridactylus.

The tuber of the calcaneum has about the same general proportions as the
European species D. (?) croizeti Pomel*, i.e., it is quite heavy and of medium length,
while the sustentacular facets are similar in detail. The broad and rather low
astragalus also agrees in detail with that of this European form. The cuboid is
quite high and has an extremely heavy process posteriorly. The metatarsals are
quite elongated, the lateral metatarsals with curved shafts somewhat similar to
those in the manus. The ungual jihalanges are shorter than in the manus.

The remains of Diceratherium cooki constitute by far the greatest percentage of
all the material found in the Agate Spring Fossil Quarries. Another significant

* Pomel referred the species to Aceratherium. Professor Max Schlosser identifies it as Diceratherium
in the case of material sent to us from the Royal Museum in Munich.

PETERSON: THE AMERICAN DICERATHERES. 449

Measurements of the Type of D. cooki.

Greatest length of skull 350 mm.

Length from occipital condyle to and including P- 307

Length from occipital condyle to M^ 150

Greatest transverse diameter of skull 215

Greatest transverse diameter of brain-case 107

Greatest transverse diameter of frontals 140

Transverse diameter of nasals back of horn-cores 65

Transverse diameter of nasals at the horn-cores 70

Transverse diameter of palate at M' 55

Vertical diameter of the orbit 30

Antero-posterior diameter of premolars two, three, and four 68

Antero-posterior diameter of the molar series 90

Antero-posterior diameter of P- 22

Transverse diameter of P- 23

Antero-posterior diameter of P* 28

Transverse diameter of P* 29

Antero-posterior diameter of M' 34

Transverse diameter of M' • 32

Antero-posterior diameter of M^ 26

Transverse diameter of M' 32

Measurements of Limb Bones of Skeleton. '

Scapula, height 273 mm

Scapula, width at superior border 138

Humerus, length 250

Ulna, length 315

Radius, length 250

Carjius, heiglit 59

Carpus, transverse diameter 60

Mc II, greatest length 70

Me III, greatest length 138

Mc IV, greatest length 115

Phalanges, median digit 70

Pelvis, total length 335

Pelvis, diameter across ilia 355

Pelvis, Diameter across acetabulum 220

Femur, length 323

Tibia, length _, 275

Tarsus, height of tuber of calcaneum not included 70

Tarsus, length of tuber of calcaneum 48

Tarsus, greatest transverse diameter 53

Mt. II, length 110

Mt. Ill, length 125

Mt. IV, length 110

Phalanges, third digit 30

450 MEMOIRS OF THE CARNEGIE MUSEUM.

fact is the great number of bones representing young animals and females in pro-
portion to those of males. This would appear to indicate (1) that the animals
were polygamous to a great degree and that the males were either struggling for
the possession of the herds after the manner of recent ungulates (Equus) , and were
few, or, that they were strong enough to extricate themselves when overtaken by
the calamities which destroyed the herds.

The articulated skeleton of Diceratherium cooki has been fully discussed in
the Annals of the Carnegie Museum, Volume VII, pp. 274-279.

Modes of De\t:lopment of Certain Dental and Bony Structures of the

Cranium in Diceratherium.
(Plates LXV and LXVI.)

Important facts, in connection with the evolution of the dental formula and
other features of the cranium of the Rhinocerotidse, are obtained from the large
collection under study in the Carnegie Museum. Some studies bearing on the
evolution of the incisors and canines of Diceratherium were already presented before
the Paleontological Society at Pittsburgh in 1910. The following pages are given
to a further discussion of the appearance and shedding of the different deciduous
teeth, the appearance of the permanent series, and other changes of contour of
the head from the young to the fully adult form of Diceratherium cooki.

1. A skull of a young Diceratherium, No. 1848 (See PI. LXV, Figs. 1, 2, 4)
which belongs to the original series from which the type of D. cooki was selected,
is especially complete and furnishes an excellent opportunity for study. In viewing
this skull from above, the most noticeable characters are the following: brain-case
proportionally broad; occiput short; frontals broad; horn-cores little developed,
and nasals gradually pointed, more like that of adult females. Back of the horn-
cores on the lateral margin of the nasals there is also less constriction in skulls of
young individuals and adult females than is the case in males. The supra-orbital
ridges are so varied that one cannot attach great importance to them, though it
would appear that in female skulls they are generally less prominently developed
and in their backward progression to the occiput they possibly have a tendency to
be further separated from the median line. On either a direct side view or a palatal
view of the young skull the most noticeable feature is the great backward extent of
the alveolar border of the maxillary. The alveole for M' is seen to be nearly oppo-
site the pterygoid, while in fully adult forms this tooth is well in advance of this
region. In very young individuals, the base and the supra-occipital of the skull
are often slipped off at the sutures, not an unusual feature of the mammalia.
In the skull here described, the base is lost, but the supra-occipital is in position.

PETERSON: THE AMERICAN DICERATHERES. 451

Buried deep in the small round alveolus, the point of the upper permanent
incisor is found. Judging from the size of the alveole, the deciduous tooth was
rather small and had a root of more rounded outline than the permanent one,
and the crown was perhaps also of an entirely different shape. There is no canine
present and, if there were a deciduous canine in this individual, it dropped out
early and the alveole was closed, there being in this region a small groove which
extends for a short distance back of the maxillary-premaxillary suture. If there
was a deciduous first premolar in the Diceratheres, it was possibly shed very early
in life.^" P^ is somewhat worn, but not enough to lose the characters of the grinding
face. (See PI. LXV, Fig. 2.) The ectoloph is, as usual, well developed, the
protoloph is less prominent than the metaloph, which gives to the tooth the char-
acteristic triangular outline. The post-fossette is sometimes constricted in such a
manner as to form an isolated fossette on the metaloph on further wear, while the
main post-fossette continues to the posterior edge of the tooth. This fossette is
not always present. D.P- is considerably worn, but the detailed structure is yet
easily made out. The tooth is longer than the permanent tooth, the ectoloph is
heavy, the protoloph is well developed internally as is also the metaloph. The
crista is enormously developed, extending on an even internal line with the proto-
and metalophs. In a young or unworn tooth this ridge is often constricted so
as to form an internal tubercle, which on further wear unites with the true crista.
In the young of the John Day forms both the crista and this internal tubercle are
less developed and apparently entirely separated, judging from the material in
the American Museum. This is admirably illustrated on PL LXV Fig. 3. The
crochet of D.P- in D. cooki is quite distinct though much less developed than the
crista, and the cingulum is well developed on the internal face of the tooth. In
excavating the maxillary above D.P- it is seen that P- is quite well advanced. See
PL LXV, Fig. 1. D.P^ is well worn. The median valley is open, but the crochet is
evidently united with the ectoloph, while the post-fosette is isolated by wear of
the tooth. There is a small tubercle on the internal cingulum in the median
valley. D.P^ has the well developed crochet still separated from the ectoloph,
but the crista is rather poorly developed or wanting. The median valley is open
and, as in the preceding tooth, there is a small tubercle on the cingulum at the
exit of the valley. The post-fossette is broad and open. M' is fully erupted and
has already received some wear. The ectoloph is yet quite thin, but in excavating

™ If P' in Diceratherium did not succeed a milk-tooth in extremely early stages of the animal, this
tooth may be regarded as a persistent milk-tooth which would agree with the studies of Huxley ("Anatomy
of Vertebrate Animals," p. 362); Lydekker ("Notes on the Dentition of Rhinoceroses," Jour. Asiatic
Society of Bengal, Vol. 49, 1880, pp. 1.35-6).

452 MEMOIRS OF THE CARNEGIE MUSEUM.

the median and prefossettes, the walls of the internal face of the ectoloph and the
external face of the crochet are rapidly slanting toward one another, so that on
extreme wear the tooth would have the usual api:)carance seen in old individuals
of this species. The post-fossette is deep and broad. The cingulum is less de-
veloped internally than on the milk premolars. M" is just appearing in its alveolus
and M' is entirely buried in the maxillary.

2. In a somewhat younger individual (No. 2476) it is observed that the
roots of D.P- are longer and heavier and in excavating the maxillary, P-
is found in an extremely early stage of formation (often no evidence of it
is found). M' in this individual is just cutting through the alveolar border,
while that in the specimen described above has received slight wear. There
can be only a comparatively small difference in age of these two individuals, and
it thus appears that the permanent teeth developed extremely rai^idly after they
began to show the form of tooth in the maxillary bone. This rapid formation
and development of the permanent dentition in Diceratherium should not be
regarded as out of the ordinary when comparison is made with the shedding of the
deciduous and the appearance of the permanent teeth in man and other mammals.

3. In the collection of the Carnegie Museum are two left rami (Nos. 1820
and 1821) representing very young animals, most probably foetal. The total
length of the rami of each of these young specimens is approximately 180 mm.,
while the depth in the middle antero-posterior region is 28 mm. The most char-
acteristic features are as follows:

The lunate-shaped outline of the ramus due to the greatly downward curved
under border of the jaw in the fore-and-aft direction, the close proximity of the
cheek-teeth to the canine and the incisors, due to the absence of a diastema on
the alveolar border of the jaw, the very slight constriction in front of the cheek-
teeth and back of the incisors which is so very pronounced in adult and even in
quite young specimens, the small transverse diameter of the symphysis, and the
deep groove on the external face of the jaw extending from the symphysis about
20 mm. back in a parallel line with the long axis of the ramus. The glenoid condyle
is not present in either one of the rami; the coronoid process on the other hand is
present in No. 1820. The latter is rather low, and terminates in an attenuated
process.

The second deciduous incisor is in place, while the alveolus for the first is
empty. The lateral incisor is not present and a small opening external to D.I2
of this individual may or may not have contained this tooth. The small alveolus
for the canine is immediately in front of that for D.Pi; the latter is a round opening

PETERSON: THE AMERICAN DICERATHERES. 453

of considerable size. The two succeeding round openings are for the roots of
D.P2. Back of this point the two succeeding cheek-teeth are partly erupted.
(See PI. LXVI, Figs. 3, 8 and 9.) The general pattern of these teeth is quite
similar to that of the permanent set, indeed it is not easy to distinguish one set
from the other. Portions of a fifth cheek-tooth (Mi) lie buried deep in the jaw
behind the two just described.

4. Two pairs of lower jaws (Nos. 2476, 2477) have been selected from the
collection to represent the next stage of evolution in the development of the ramus.
No. 2476 being represented on PL LXVI, Fig. 7. The total length of jaws Nos.
2476, 2477 is 250 and 270 mm. and their depth is 36 and 40 mm. respectively. At
this stage the jaw is easily recognizable, as all the characteristic generic features
are present. The jaw is less lunate-shaped, the characteristic diastema in front
of the cheek-teeth is quite well developed, including the constriction of the alveolar
border, which in the younger stage is represented only by the deep groove on the
external face of the ramus referred to above. In this stage of development the chin
is broader, due to the lodgment of the already well-advanced lateral incisor. The
temporal fossa is well developed, as are the condyle and the coronoid process.

The median incisors are just through the alveolar border and present the
same small and conical-shaped crowns met with in older forms. These teeth
are succeeded by a short diastema before the alveoli of D.I 2 is reached. The latter
is situated somewhat posterior to Ii and I3 in the alveolar border and is thus placed
in an irregular position. In the .specimens of the Carnegie Museum under obser-
vation there is sometimes found a delicate septum separating the second and third
incisors. This bony bridge is often broken. D.I3 is frequently found in i:)osition
as is the case in the specimen here described, see PI. LXVI, Fig. 7. This tooth has
a long root, quite robust, on which sits a small enamel-covered crown very little
larger in circumference than the root itself. The two lateral incisors are succeeded
by a diastema; the alveolar border here forms a heavy rounded edge with a notice-
able swelling on the external face. This swelling of the incisive alveolar border
is due entirely to the rapid development of the cutting incisor of the second set of
teeth which is yet buried underneath the deciduous dentition. The deciduous
canine is found in many individuals. This is a small tooth with a conical enamel-
covered crown and a rather short root. The canine is quite generally found in a
procumbent position, isolated by diastemata in front and behind, and drops out
very early. The diastema back of the canine constitute a long and sharp border
which has first a slight inward curvature and then suddenly takes an outward
direction to meet the cheek dentition. The first deciduous cheek-tooth is rather

454 MKMOIRS OF THE CARNEGIE MUSEUM.

small, simple-crowned, receives comparatively little wear, and is closely crowded
to the anterior faceof the succeeding tooth. D.Pa is proportionally longer and
narrower than P.. The configuration of the crown is otherwise c^uite similar in
the two. D.P3 has by far the greatest wear and is, if one may judge by the degree
of wear, the first cheek-tooth to appear in the young. D.P4 is less worn and cuts
the alveolar border simultaneously with or perhaps a httle sooner than D.P2. Back
of D.P4 are seen the points of the crown of Mi and back of the last-mentioned tooth
the alveolar border is deeply marked to indicate the position of M2 which is still
entirely buried in the jaw.

5. The next stage of development in the succession of teeth from the deciduous
to the permanent series is interesting. Three individuals have been selected
which fairly well cover the main points in individual variation and irregularities of
development. Of these three individuals No. 1923a might be considered as the
most normal and will be first discussed. The small median incisor occupies the
usual position, while the permanent lateral incisor has broken through the alveolar
border (PI. LXVI, Fig. 5), uniting the alveoli for D.l2and D.I3 into a large trans-
versely oblongate fissure for receiving the cutting and procumbent incisor. ^^ In
this individual the alveolus for the canineof the right ramus is present, though very
small, while in the left there is no trace of an alveolus for the canine. D.Pi is in
place while D.P2 has been shed and the crown of Po appears through the alveolar
border. D.P.s is still in place butPs is well advanced and the deciduous tooth was
almost ready to drop off before the death of this individual. D.P4 is apparently
quite solid in the jaw and still served well for masticating purposes. Mi has
already received considerable wear, while M2 is almost entirely erupted. M3 is
quite undeveloped and is buried deep in the jaw.

6. The next specimen to be considered is No. 1841, a pair of lower jaws. This
specimen presents some irregularities worthy of note. From the illustration
(PL LXVI, Fig. 6) it is seen that the permanent lateral incisors of this specimen are
retarded, i.e., they have not yet appeared above the surface of the alveolar border;
the alveoli for the canines are quite large. D.Pi is still in place, but contrary
to the specimen just described both D.P2 and D.P3 have disappeared and P2and P3
have already received some wear, D.P4 is solidly inserted in the jaw, Mi has been in
use for some time and is considerably worn as in No. 1923a, while M2 has received
slight wear on the anterior portion. M, on the other hand is apparently no further
developed than in the specimen previously described.

'' In the judgment of the writer this incisor is probably I2 while I3 and the canine of the Diceratheres
are atrophied.

1

PETERSON: THE AMERICAN DICERATHERES. 455

7. In No. 1854 it is seen that the lateral incisor is no further advanced than in
No. 1841 just described. The alveole or deep groove^^ for the deciduous canine is
still quite prominent while D.Pi is shed and all traces of its alveole entirely oblit-
erated. P2 and P3 have received slightly more wear than those teeth in the previous
specimen described, but D.P4 is still well rooted in the alveolar border. Mi is quite
well worn and the anterior part of M2 is also more worn than that in No. 1841.
The deep fissure back of Mo indicates the position of M3. The latter is very little
further developed than in the two preceding specimens and is yet buried in the
angle of the jaw. The three lower jaws just described are of approximately the"
same age as the skull No. 1848, referred to in the opening paragraph of this dis-
cussion.

In connection with the probable manner in which the upper and lower incisors
of Diceratheriuni^^ developed in size, and modified into the shape in which we find
them, it is interesting to return to the foetal specimens Nos. 1820 and 1821 just
described (page 452). We have already found that the deciduous dentition of
these specimens forms practically a close series, without the constricted and thin
areas of the alveolar border between the cheek-teeth and the incisors of older ani-
mals, the alveolus for the canine is deep though small; in excavating the chin,
the continuation of the dental canal is found, but the germ for the permanent
incisor had not yet started, hence the small transverse diameter of the chin.

In the next stage represented by Nos. 2476 and 2477, the specimens are of
quite young animals. We observe here a sudden change. It is likely that the
characters so prominently developed in this young animal had already been
well advanced during the latter part of the intra-uterine stage. At all events
the jaw was still in a very plastic condition in order to have transformed so quickly
between the two stages represented in the illustrations (see PL LXVI, Figs. 7 and
9). In the specimens of this second stage we find a broad and heavy chin in order
to support the heavy and long-rooted permanent incisor. The alveolus for the
canine, which we originally found quite deep and placed close to the cheek-teeth,
is now shifted well forward and is transformed into a shallow groove, which in

'- In more matured individual!?, the fissure in the alveohir border which lodged the canine cannot be
regarded as a true alveolus.

'' Not only Diceraiherium but the Rhinocerata in general (such forms as the Amynodonts excepted)
undoubtedly developed the cutting incisors along the same general line.

The result of the present study is contrary to the statement by Professor Marsh (Amer. Jour. Sci.,
Vol. XIV, 1877, p. 251). It may be said here that the presence of the canines in the Amynodonts does
not prove "that the large lower teeth, usually regarded as incisors in Acerathcrium and many other
members of the Rhinoceros family, are really canines."

456 MEMOIRS OF THE CARNEGIE MUSEUM.

the majoritj' of cases is empty, the small canine lodged therein having dropped
out, while back of the canine we find a long diastema which is very much constricted
forming externally a deep and rather broad groove for the lodgment of the inferior
labial muscles.

Let us suppose that the foetal specimens referred to have the jaws more or
less like those of the early Tertiary forms. We must in any event expect that the
early progenitors of the family had, first a complete dental series, i.e., f • {■ i- f
(abundantly proven by the genus Trigonias of the lower Oligocene); and secondly
quite likely the absence of a diastema back of the incisors.''^ It follows that
advancing influences effected gradual changes of the bony structure simultaneously
with that of the teeth. If we have, for instance, a set of lower incisors of subequal
size and a normal canine in its natural position (we actually do find evidence of a
canine in young Diceratheres), we should expect the upper incisors to meet the
lower. When the atrophied and hypertrophied changes took place, which trans-
formed the original sub-equal teeth to those which obtained in later forms, it was
not the lower canine, hut l-i, which received the constant impact from the upper median
tooth. The diastemata between the incisors, canine, and cheek-teeth was most
likely an early development of the group. The modification of the cutting incisor
was cotemporaneous with the reduction of Ii the atrophy of I3, the broadening of
the chin, and the constriction of the ramus in the region of the canine which, in
turn probably, caused the reduction and final disappearance of the latter tooth.
After the present study of the Diceratherinw, I cannot accept the designation
given to this tooth, as "canine," by some authors. Professors Marsh, Cope, and
Gaudry having been the first to promulgate this view.

Since the preceding jmragraphs were submitted for publication, Professor
W. B. Scott of Princeton University has published his splendid work on the "His-
tory of Land Mammals in the Western Hemisphere." On consulting the history
of the Rhinocerotidse in Scott's volume, pp. 326-340, it is very evident that he
does not regard the large cutting incisor of the lower jaw in the early Rhinoceroses
as a canine. In fact since the genus Trigonias from the Lower Oligocene of America
was established by Mr. Lucas '^ and more completely described by Mr. Hatcher'^
the morphology of the incisors and canines of the Rhinocerotidse rests on a firmer
foundation.

'* Even in Colonoccran agrestis Marsh, a genus which might be regarded as possibly near the ancestral
line of the Diceratheres, there is a well-established diastema back of the superior canines.
"Proc. National Museum, Vol. XXIII, 1900, pp. 221-223.
'I' Ann. Carnegie Museum, Vol. I, 1901, pp. 135-144.

458 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LVII.

Fig. 1. Diceratherium armatum, type. Side view of skull. Peabody Museum,
No. 1000.3.

Fig. 2. Diceratherium armatum, type. Palatal view of same specimen as Fig. 1.

All figures about | natural size.

460 MEMOIRS OP THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LVIIL

Fig. 1. Diceratherium nanuni, (Marsh) type. Peabody Museum of Natural
History, No. 10004. Front of skull from the side.

Fig. 2. Diceratherium 7ianum, type. Front of jaws from the side.
Fig. 3. Diceratherium nanum, type. Alveolar border and dentition.
Fig. 4. Diceratherium cooki. Carnegie Museum, No. 1555.

All figures | natural size.

>

<

>

o

>

C3

00

w

3
w

Pi

00

462 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LIX.

Fig. 1. Dicerathcrium gregorii, type. Side view of skull. American Museum of
Natural History, No. 12933.

Fig. 2. Diceratherium gregorii, type. Palatal view of the same skull.

All figures \ natural size.

><

w

o
>

W
3

w

PS

o
to

K

H
B

464 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LX.

Fig. L Diceratherium cooki type. Carnegie Museum, No. 1572.

Fig. 2. Diceratherium niobrarense type. Carnegie Museum, No. 127L

Fig. 1, I natural size; Fig. 2, | natural size.

Memoirs Carnegie Museum, Vol. VII.

Plate LX.

D. cooki Peterson and D. niobrareiise Peterson.

466 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LXL

Fig. 1. Diceratherium cooki, type. Carnegie Museum, No. 1572.

Fig. 2. Diceratherium niobrarense, type. Carnegie Museum, No. 127L

Fig. 1, 5 natural size; Fig. 2, I natural size.

Memoirs Carnegie Museum, Vol. VII.

Plate LXI.

D. cooki Peterson and D. niohrarcnse Peterson.

468

MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LXIL

Fig. L Diceratherium cooki, type. Carnegie Museum, No. 1572.

Fig. 2. Diceratherium viohrarense, type. Carnegie Museum, No. 127L

Fig. 1, I natural size: Fig. 2, % natural size.

Memoirs Carnegie Museum, Vol. VII.

Plate LXII.

D. Cdoki Peterson ^nd D. niobrarenae Peterson.

470 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LXIII.

Fig. L Diceratherium cooki, ulnar view of manus, Carnegie Museum, No. 2473.

Fig. 2. Diceratherium cooki, palmar view of manus, Carnegie Museum, No. 2473.

Fig. 3. Diceratherium cooki, radial view of manus, Carnegie Museum, No. 2473.

Fig. 4. Diceratherium cooki, dorsal view of manus, Carnegie Museum, No. 2473.

Fig. 5. Diceratherium cooki, dorsal view of pes, Carnegie Museum, No. 1888.

Fig. 6. Diceratherium annectens hypotype, American Museum Natural History,
No. 7324, Cope Coll.

Fig. 7. Diceratherium cooki, plantar view of pes, Carnegie Museum, No. 1888.

All figures | natural size except Fig. 6, which is \ natural size.

Memoirs Carnegie Museum, Vol. VII

Plate LXIII.

D. cooki Peterson and D. annedens (Maksh).

472

MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LXIV.

Diceraiherium cooki.

Carnegie Museum, No. 1888.

Fig. 1. Calcaneum, dorsal view.

Fig. 2. Calcaneum, distal view.

Fig. 3. Calcaneum, plantar view.

Fig. 4. Calcaneum, tibial view.

Fig. 5. Calcaneum, fibular view.

Fig. 6. Astragalus, tibial view.

Fig. 7. Ectocuneiform, proximal view.

Fig. 8. Ectocuneiform, distal view.

Fig. 9. Ectocuneiform, tibial view.

Fig. 10. Astragalus, fibular view.

Fig. 11. Astragalus, plantar view.

Fig. 12. Astragalus, dorsal view.

Fig. 13. Cuboid, proximal view.

Fig. 14. Cuboid, tibial view.

Fig. 15. Entocuneiform, fibular view.

Fig. 16. Entocuneiform, distal view.

Fig. 17. Navicular, distal view.

Fig. 18. Navicular, posterior view.

Fig. 19. Navicular, fibular view.

Fig. 20. Navicular, proximal view.

Fig. 21. Cuboid, distal view.

Fig. 22. Mesocuneiform, distal view.

Fig. 23. Mesocuneiform, fibular view.

Fig. 24. Mesocuneiform, tibial view.

Fig. 25. Mesocuneiform, proximal view.

Fig. 26. Metatarsal III, dorsal view.

Fig. 27. Metatarsal III, fibular view.

Fig. 28. Metatarsal II, fibular view.

Fig. 29. Metatarsal II, tibial view.

Fig. 30. Metatarsal IV, tibial view.

Fig. 31. Metatarsal III, tibial view.

Fig. 32. Metatarsal III, jjlantar view.
Carnegie Museum, No. 2453.

Fig. 33. Scaphoid, proximal view.

Fig. 34. Scaphoid, distal view.

Fig. 35. Scaphoid, ulnar view.
Carnegie Museum, No. 2453.

Fig. 36. Lunar, proximal view.

Fig. 37. Lunar, radial view.

Fig. 38. Lunar, distal view.

Fig. 39. Lunar, ulnar view.

Fig. 40. Cuneiform, radial view.

Fig. 41. Cuneiform, distal view.

Fig. 42. Cuneiform, ulnar view.
Carnegie Museum, No. 1853,

Fig. 43. Pisiform, radial view.

Fig. 44. Trapezium, ulnar view.

Fig. 45. Trapezoid, radial view.
Carnegie Museum, No. 2453.

Fig. 46. Trapezoid, ulnar view.

Fig. 47. Trapezoid, distal view.

Fig. 48. Trapezoid, proximal view.

Fig. 49. Magnum, distal view.

Fig. 50. Magnum, ulnar view.

Fig. 51. Magnum, radial view.

Fig. 52. Magnum, proximal view.

Fig. 53. Unciform, radial view

Fig. 54. Unciform, proximal view.

Fig. 55. Unciform, ulnar view.

Fig. 56. Metacarpal II, radial view.

Fig. 57. Metacarpal II, ulnar view.

Fig. 58. Metacarpal III, radial view.

Fig. 59. Metacari^al III, ulnar view.

Fig. 60. Metacarpal IV, radial view.

Fig. 61. Metacarpal IV, ulnar view.

Fig. 62. Metacarpal V, radial view.

Fig. 63. Metacarpal V, palmar view.

All figures are | natural size.

><
w

Cu

>

o

00

w
5

-<

O

s

Z

o

a:

K

r-"
K

474 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LXV.

Fig. L Diceralherium cooki, young male, side view of skull. Carnegie Museum,
No. 1848.

Fig. 2. Diceralherium cooki, Palatal view same as Fig. 1 .

Fig. 3. Diceratherium? annectens, deciduous upper teeth, American Museum
collection.

Fig. 4. Diceratheriurn cooki, top view of skull, same as Figs. 1 and 2.

All figures | natural size.

Memoirs Carnegie Museum, Vol. Vll

Plate LXV.

D. cooki Peterson and D. anncctens (Marsh).

476 MEMOIRS OF THE CARNEGIE MUSEUM.

EXPLANATION OF PLATE LXVI.

Fig. 1. Diceratherium anneclens, hypotype, Side view of skull, American Museum,
No. 7324, Cope Coll.

Fig. 2. Diceratherium cooki, upper contour of lower jaws and crown view of den-
tition. Carnegie Museum, No. 2499.

Fig. 3. Diceratherium cooki, outer view of mandible in very young stage of develop-
ment. Carnegie Museum, No. 1820.

Fig. 4. Diceratherium cooki, outer view of mandible of fully adult male. Carnegie
Museum, No. 2499.

Fig. 5. Diceratherium cooki, alveolar border of lower jaw and crown view of den-
tition. Carnegie Museum, No. 1923a.

Fig. 6. Diceratherium cooki, alveolar border of lower jaw and crown view of den-
tition. Carnegie Museum, No. 1841.

Fig. 7. Diceratherium cooki, alveolar border and crown view of dentition. Carnegie
Museum, No. 2476.

Fig. 8. Diceratherium cooki, inner view of mandible, very young stage of develop-
ment, Carnegie Museum, No. 1820.

Fig. 9. Diceratherium cooki, upper contour of lower jaw and crown view of den-
tition, same as Nos. 3 and 8.

Fig. 1 is I natural size; Figs. 2 and 4 are | natural size; Figs. 3, 5, 6, 7, 8, and 9 are

1

2

- natural size.

>

X

w

<

0-,

o

>

w

CO

w

o
w
2:

OS

<
o

en

o

s
w

s

a
z
■<

z

o

«

H

H
Oh

INDEX

aberrans, Diceratherium, 432
Acanthopoma, 261, 269

annectens, 354, 355
acuminatus, Oxybelis, 211
sesculapii, Erythrolamprus, 207
affinis, Baena, 109
agassizi, Aphyocharax, 12, 23, 36

Cheirodon, 3
Agkistrodon piscivorus, 227
Aglypha, 170
ahaetulla, Leptophis, 184
albirostris, Helminthophis, 166
albiventris, Liophis, 187
alburnus, Aphyocharax, 3, 12, 13, 23, 24,
25, 27, 29

Paragoniates, 3, 12, 40, 41
almadensis, Liophis, 187
alternatum, Pygidium, 327, 328, 336
amazonicum, Pygidium, 327, 328, 336
amazoni, Prionobrama, 39
Amblycephahdse, 221
amnion, Testudo, 151, 154
amphithorax, Testudo, 155
Amy da, 101

crassa, 155

egregia, 155, 156

scutumantiquum, 155, 157
anceps, Glyptosaurus, 160
Anguidse, 159
angulata, Helicops, 179
Anisits, J. D., 2, 74

anisitsi, Aphyocharax, 12, 13, 23, 29, 30,
33

Homodisetus, 351
anna;, Cheirodon, 7, 12, 64, 67, 68, 69, 76
annectens, Acanthopoma 355

Diceratherium 403, 411, 416, 417,
420, 422
annulata, Leptodeira, 210

477

Anosteira, 101

ornata, 102, 121, 122
antiqua, Baena, 109
Aphyocharax, 9, 15

agassizi, 12, 23, 36

alburnus, 3, 12, 13, 23, 24, 25, 27, 29

anisitsi, 12, 13, 23, 29, 30, 33

avary, 12, 23, 24, 30, 31

dentatus, 3, 12, 13, 23, 24, 25, 26, 28

erythrurus, 12, 23, 24, 28

fiUgerus, 37

maxillaris, 12, 23, 25, 37

melanotus, 23

nattereri, 12, 23, 24

paraguayensis, 12, 24, 25, 33

pulcher, 3

pusilhis, 12, 13, 22, 23, 24, 28, 29

rathl3uni, 12, 24, 31, 32

stramineus, 32

Key to the species of, 24
Aphyocheirodon, 5, 8, 16

hemigrammus, 12, 58, 59
Aphyodite, 5, 16, 48

grammica, 12, 47
Aporophis flavifrenatus, 170

lineatus, 170

melanocephalus, 171
areolata, Hatcheria, 285
arenosa, Baena, 107, 108, 109, 121
Argopleura, 11
armatum, Diceratherium, 404, 413-416,

419, 426
armillatus, Pariolus, 260
Atractus badius, 172

roulei, 173

tseniatus, 173
austraUs, Pimelodella, 230, 231, 233
avanhandavffi, Pimelodella, 232, 240, 253
avary, Aphyocharax, 12, 23, 24, 30, 31

478

MEMOIRS OF THE CARNEGIE MUSEUM.

azurea, Eichornea, 357, 367

badius, Atractus, 172
Baena affiiiis, 109

antiqua, 109

arenosa, 107, 108, 109, 121

Clara, 110, 116, 121

emili«, 109, 111, 114, 119

gigantea, 116, 117, 118, 119

hatcheri, 117, 118, 119

inflata, 107, 108, 112, 113, 115, 117,
118

platyplastra, 120

riparia, 118

sima, 112, 115, 117, 121
Baenidae, 101, 107
"Bagre," 280

molle, 267
banneaui, Pygidium, 307, 318
Barbour, Thomas, 274
barbouri, Pygidium, 274, 297, 303
batrachostoma, Ochmacanthus, 356, 367
Bertoni, Sr. A. de, 274
bertonii, Branchioica, 268, 269, 367, 368
bicornis, Rhinoceros, 444
bifossatus, Drymobius, 176
bitorquata, Cleha, 201
Boa cooki, 167

hortulana, 167
bocourti, Leptophis, 184
boddaerti, Drymobius, 177
bogotense, Pygidium, 306, 315, 316, 318
bogotensis, PoeciUa, 17
Boidse, 167

boliviana, Pimelodella, 230, 231, 233, 245
bondensis, Helminthophis, 165
borellii, Pygidium, 294
Bramocharax, 10
Branchioica, 262, 359

bertonii, 268, 269, 274, 367, 368
brasiliense, Pygidium, 295, 327, 329, 337
brasiliensis, Pimelodella, 230, 232, 242

Pygidium, 336, 338
brevidens, Glyptosaurus, 160
brevis, Tridens, 370
brontops, Testudo, 154

Bryan, W. L., President Indiana Uni-
versity, 1

buckleyi, Pimelodella, 230, 231, 232, 240,
241, 245

burmeisteri, Hatcheria, 283, 286

caecutiens, Cetopsis, 263
Caenopus dakotensis, 402

mitis, 435

tridactylus, 402, 446, 448
caliense, Pygidium, 306, 311
calliurus, Cheirodon, 76, 77
callopyge, Echmatemys, 123, 124, 125,

126, 127, 132, 134
Cambeja, 267
"Candiru," 262-263

Cetopsis, 263
"Capitan," 343
carinatus, Herpetodryas, 183
carinicauda, Helicops, 179
Carriker, Jr., M. A., 274
catesbyi, Cochliophagus, 221
Cenchoa, Himantodes, 208
cenchria, Epicrates, 168
Cetopsinse, 276 ■
Cetopsis caecutiens, 262, 263

candiru, 263
chagresi, Pimelodella, 230, 231, 252, 253
chapmani, Pygidium, 305, 309
Cheirodon, 16

agassizi, 3

annai, 6, 12, 64, 67, 68, 69, 76

calliurus, 76, 77

eques, 3, 52 .

gorgonse, 70

ibicuhyensis, 12, 65, 74

insignis, 12, 65, 70

interruptus, 3, 6, 12, 57, 65, 71, 72,
74, 76, 77

madeirae, 12, 65, 76

microdon, 12, 64, 66, 80, 81

micropterus, 77

monodon, 12, 65, 71, 72, 73, 74

nattereri, 35

notomelas, 7, 12, 65, 74, 75

parahybae, 12, 65, 70

INDEX.

479

Cheirodon pequira, 83

piaba, 7, 9, 11, 12, 65, 71, 72, 76, 77,
79, 80

pisciculus, 2, 6, 12, 13, 64, 65, 67, 68

pulcher, 3, 35

ribeiroi, 63

steindachneri, 35

stenodon, 66, 80, 82

Key to the species of, 65
Cheirodontinae (The), a Subfamily of
Minute Characid Fishes of South
America, By Carl H. Eigenmann,
1-99

Anal fin of, 6

Distribution of, 11, 12

Dorsal fin of, 6

Generalized type of, 3, 4, 5

Highest altitude of, 14

In various museums, 2

Key to the genera of, 14

Mouth of, 6

Museums containing species of, 2

Phylogenic arrangement of the genera
of, 13

Relationship of the, 10

Secondary sexual differences, 11

Teeth of, 6, 9

See also Errata and Corrigenda, p. xxiii
Chelonian fauna of the Uinta, last appear-
ance of the, 101
cibollensis, Echmatemys, 139
cirrhosa, Vandellia, 263, 358, 360
cisternarum, Phreatobius, 371, 372
clara, Baena, 110, 116, 121
Clelia bitorquata, 201

cloelia, 201

doliata, 202

euprepa, 203

petolaria, 205

rhombifer, 205

trigemina, 206
Cobitiglanis taxistigma, 345
" Cobra de Agua," 182
Cochliophagus catesbyi, 221
Coggeshall, Arthur S., 103
Colubrida;, 170

Colubrina?, 170
colubrinus, Xenodon, 199
columbianus, Elaps, 216
Compsura, 10, 13, 16

heterura, 6, 12, 60, 61
conradi, Pygidium, 325
Constrictor constrictor, 168
cooki. Boa, 167

Diceratherium, 403-409, 411, 420,
421, 425, 428, 431, 433, 448
Corais, Elaphe, 174
corallinus, Elaps, 217
corduvense, Pygidium, 292, 293
coronata, Pseudoboa, 206
Coronella micropholis, 176
corsoni, Hadrianus, 143, 144, 145, 146, 148
crassa. Amy da, 155

cristata, Pimelodella, 230, 231, 236, 237
cristatus, Pimelodus, 229
croizeti, Diceratherium, 448
Crotalina?, 222
Crotalus terrifieus, 226
cyanostigma, Pimelodella, 230, 231, 232,

242
cyclolepis, Parecbasis, 12, 44, 45, 46

dakotensis, Caenopus, 402
davisi, Pygidium, 327, 331, 334
dentatus, Aphyocharax, 3, 12, 13, 23, 24,

25, 26, 28
depressa, Echmatemys, 139, 140, 141
Dermatemj^didse, 121
Diceratheres, The American. By 0. A.
Peterson, 399-476

Early investigations, 400-401

Stratigraphy, 401
Diceratherium aberrans. 432

annectens, 403, 411, 416, 417, 420,
421, 422

armatum, 404, 413, 416, 419, 426

arrikarense, 432

cooki, 403-409, 411, 420, 421, 425,
428, 431, 433, 448

croizeti, 448

douvillei, 404

gregorii, 421

480

MEMOIRS OF THE CAENEGIE MUSEUM.

Diceratherium hesperius, 411

loomisi, 433

minutum, 404

niobrarense, 403, 414, 420, 422, 424,
425

pacificus, 410

petersoni, 413

schiffi, 432

truquianum, 412
dichroa, Elaphe, 174
Dimades plicatilis, 175
Dipsadomorphinse, 201
dispar, Pj^gidium, 297, 299
doliata, Clelia, 202

dorsostriatum, Pygidium, 290, 307, 320
Douglass, Earl, 101, 103
douglassi, Echmatemys, 128, 130, 131, 132
douvillei, Diceratherium, 404
di-epanon, Odontostilbe, 3, 12, 90, 92
Drymobius bifossatus, 176

boddaerti, 177

boddaerti var. rappii, 177

rhombifer, 178

Echmatemys callopyge, 123, 124, 125,
120," 127, 132,' 134

cibollensis, 139

depressa, 139, 140, 141

douglassi, 128, 130, 131, 132

euthneta, 139

hollandi, 132, 135

lativertebralis, 133

mpgaulax, 132, 139, 141

obscura, 135, 136, 137, 138, 139

pusilla, 140, 141, 142

septaria, 124, 125, 126, 127

stevensoniana, 132

uintensis, 127, 131
ecuadorensis, Megalamphodus, 2, 12, 99
egregia, Amyda, 155, 156
egrerius, Aceratherium, 426
Eichornea azurea, 357, 367
eichorniarum, Pygidium, 295
Eigenmanu, Carl H., The Cheirodoutina",
a Subfamily of Minute Characid
Fishes of South America, 1-99

Eigenmann, Carl H., Pimelodella and
Typhlobagrus, 229-258

Pygidiidse (The), a Family of South
American Catfishes, 259-398
eigenmanni, Pimelodella, 230, 231, 234,
241, 248

Pygidium, 297, 298
eigenmanniorum, Pimelodella, 241
elseoides, Liophis, 187
Elaphe corais, 174

dichroa, 174
Elapina?, 216
Elaps colombianus, 216

corallinus, 217

frontalis, 218

hollandi, 218

mipartitus, 219

narduccii, 220

princeps, 220
elongata. Pimelodella, 230, 231, 252, 254
emiliae, Baena, 109, 111, 114, 119
Epicrates cenchria, 168
var. fusca, 169
eques, Cheirodon, 3, 52

Megalamphodus, 12, 50, 52
Eremophilus, 260

mutisii, 270, 341

a food fish, 260
Erythrolamprus sesculapii, 207
erythrurus, Aphyocharax, 12, 23, 24, 28
Eunectes murinus, 169

notaeus, 169
euprepa, Clelia, 203

euta;nia, Pimelodella, 230, 231, 252, 254
euthneta, Echmatemys, 139
exornata, Testudo, 154

fasciata, Natrix, 227

fasciatus, Phrynonax, 194

fassli, Pygidium, 298, 303

filigera, Pi'ionobrama, 38, 39, 40

filigerus, A])hyocharax, 37, 3, 12, 13

Fisher, Carl G., 1

flabelliferus, Ochmacanthus, 356, 357

flavifrenatus, Aporophis, 170 *

frontaUs, Elaps, 217

INDEX.

481

fugitiva, Odontostilbe, 90, 92, 93
fulgidus, Oxybelis, 212
fuscum, Pygidium, 288, 296, 298
fuscus, Herpetodryas, 183

Gephyrocharax, 37
gigantea, Baena, 116, 117, 118, 119
Gilmore, Charles W., The Fossil Turtles
of the Uinta Formation, 101-161
Glyptosaurus, 102, 159
anceps, 160
brevidens, 160
montanus, 160
nodosus, 160
ocellatus, 160
princeps, 160
rugosus, 160
sphenodon, 160

Geological arrangement from de-
posits of North America, 160
goeldii, Pygidium, 327, 328, 337, 339
Gonzales, Manuel, 274
gorgonse, Cheirodon, 70
gracilior, Pygidium, 325, 326
gracilis, Pimelodella, 230, 231, 232, 237,

238, 239, 251
grammica, Aphyodite, 12, 47
gregorii, Diceratherium, 421
Griffin, Lawrence Edmonds, A Catalog
of the Ophidia from South America
at present (June, 1916) in the Car-
negie Museum with Descriptions of
Some New Species, 163-227
griffini, Pimelodella, 235, 250
grisea, Pimelodella, 230, 231, 251, 252
Grundulus, 3, 6, 8, 10, 14, 20

bogotensis, 12, 17, 18
guaporensis, Microschemobrycon, 12, 56
"Guapuche," 17
guianense, Pygidium, 325

Rhinostoma, 214
Gyrinurus, 367

fiadrianus, 102

corsoni, 143, 144, 145, 146, 148
majusculus, 146, 148, 150

Hadrianus octonarius, 145
robustus, 146, 147, 148
tumidus, 147, 148, 150
utahensis, 148, 149
hartti, Pimelodella, 230, 231, 235, 248
Haseman, John D.: Travels in South

America, 163, 164, 275
hasemani, Pimelodella, 230, 231, 232, 241
Pygidium, 325, 326
Vandelha, 359, 360
hastata, Odontostilbe, 10, 12, 90, 91
Hatcher, J. B., 274
hatcheri, Baena, 117, 118, 119
Hatcheria, 260
areolata, 285
burmeisteri, 283, 286
macrsei, 283, 287
maculata, 282, 284
patagoniensis, 281, 282
titcombi, 283, 284
Key to the species of, 282
Hay, Dr. 0. P., 103, 104
Helicops angulata, 179

carinicauda (Wied) var. infratseniata

Jan.
leopardina, 180
modesta, 181
polylepis, 181
Helminthophis albirostris, 166

bondensis, 165
Hemigrammus, 5

hemigrammus, Aphyocheirodon, 12,
58, 59
Henn, Arthur W., 99, 274
Henochilus, 2
Henonemus, 261, 269, 277
intermedins, 349
macrops, 346
microps, 353
punctatus, 345, 346
punctatus, 345, 346
taxistigma, 346, 348
Heptapterus, 260, 276
Herpetodryas carinatus, 183
var. flavolineatus, 183
fuscus 183

482

MEMOIRS OF THE CARNEGIE MUSEUM.

Herpetodryas fuscus var. saturninus, 183
hesperius, Diceratherium, 411

Rhinoceros, 410
heterodon, Holesthes, 3, 12, 84, 87, 88
heterodontum, Pygidium, 296
Heterognathi, 14
heterostomus, Probolodus, 12
heterura, Compsura, 6, 12, 60, 61
Himantodes cenchoa, 208
histricus, Lystrophis, 192
Holesthes, 17, S3

heterodon, 3, 12, 84, 87, 88

pequira, 3, 12, 84, 86

Key to the species of, 84
Holland, W. J., 1, 101, 103
lioUandi, Echmatemys, 132, 135

Elaps, 218
Holoprion maxillaris, 23

nattereri, 23

paraguayensis, 23
Homodiajtus, 261, 269, 274, 345

anisitsi, 351

maculatus, 351, 352
hortulana. Boa, 167
Hyphessobrycon, 5, 62

ibicuhyensis, Cheirodon, 12, 65, 74
Ihering, H. von, 2, 274
iheringi, Pygidium, 327, 328, 330
immaculatum, Pygidium, 327, 328, 334
Indiana University Expedition to South

America, 2
inermis, Nematogenys, 280

Trichomycterus, 279
inflata, Baena, 107, 108, 112, 113, 115,

117, 118
insidiosus, Stegophihis, 267, 353, 354
insigjiis, Cheirodon, 12, 65, 69, 70
intermedius, Henonemus, 349
interruptus, Cheirodon 3, 6, 12, 57, 65,

71, 72, 74, 76, 77
itapicuruensis, Pimelodella, 230, 231, 234,

247
itatiayffi, Pygidium, 327, 329

jatuaranse, Leptobrycon, 7, 12, 40

Kennedy, Clarence, drawings by, 1
Kerby, J. O. Serpents collected by, 165
kneri, Pygidium, 290, 306, 314
kronei, Typhlobagrus, 230, 255

Lachesis lanceolatus, 222

lansbergi, 223, 224

mutus, 224

neuwiedi, 224

peruvianus, 226
lanceolatus, Lachesis, 222
Landon-Fisher Expedition to South Amer-
ica, 2
lansbergi, Lachesis, 223, 224
lateristriga, Pimelodella, 230, 231, 245,

249, 250
laticeps, Pimelodella, 230, 231, 233, 243

Pygidium, 305, 307
latidens, Pygidium, 306, 312
latistriatum, Pygidium, 307, 318, 321
lativertebralis, Echmatemys, 133
leopardina, Helicops, 180
Leptagoniates, 6, 10, 15

steindachneri, 3, 12, 42
Leptobrycon, 15, 48

jatuaran£e, 7, 12, 40
Leptodeira annulata, 210
Leptophis ahsetulla, 184

bocourti, 184

nigromarginatus, 185

occidentalis, 185

rostralis, 186
Leptorhamdia, 260
lineatus, Aporophis, 170
Link, G. A., Collections of snakes made

by, 163
Liophis albiventris, 187

ainiadensis, 187

elseoides, 187

melanostigma, 189

melanotus, 189

poecilogyrus, 190

reginjE, 191

viridis, 192
Localities in South America at which
John D. Haseman collected reptiles, 164

INDEX.

48

Uf<'

loomisi, Diceratherium, 433
Loricariidje, 263
Lystrophis histricus, 192
semicinctus, 193

macrsei, Hatcheria, 282, 287
macrops, Henonemus, 346
Macropsobrycon, 16, 47

uruguayaniB, 7, 12, 48, 49
macturki, Pimelodella, 230, 231, 234, 248
maculata, Hatcheria, 282, 284
maculatus, Homodiajtus, 351, 352
madeira;, Cheirodon, 12, 65, 76
Odontostilbe, 3, 12, 90, 91, 97
Prionobrama, 37
majusculus, Hadrianus, 146, 148, 150
marmoratum, Pygidium, 290
Matthew, W. D., 103
maxillaris, Aphyocharax, 12, 23, 25, 37

Holoprion, 23
meeki, Pimelodella, 230, 231, 235, 249
Megalamphodus, 16, 47, 48
ecuadorensis, 2, 12, 99
eques, 12, 50, 52
heteresthes, 12, 50, 53, 54
megalopterus, 4, 7, 12, 49, 50, 51, 54
melanotus, 12, 50, 53
micropterus, 7, 10, 12, 50, 54, 55
megalops, Pimelodella, 230, 231, 232, 242
megalopterus, Megalamphodus, 4, 7, 12,

49, 50, 51, 54
megaulax, Echmatemys, 132, 139, 141
melandeta, Odontostilbe, 9, 12, 90, 91, 98
melanocephala, Tantilla, 209
melanocephalus, Aporophis, 171
melanops, Tridens, 369
melanostigma, Liophis, 189
melanotus Aphyocharax, 12, 23, 25, 37
Liophis, 189

Megalamphodus, 12, 50, 53
merida?, Pygidium, 290, 306, 315
merremi, Rhadinea, 195, 196

Xenodon, 199
" Metacaenopus " (Aceratherium) stigeri,

429
metJB, Pimelodella, 230, 231, 233, 242

met£e, Pygidium, 290, 295, 306, 312
microcephala, Odontostilbe, 12, 85, 94
microdon, Cheirodon, 12, 64, 66, 80, 81
micropholis, Coronella, 176
microps, Henonemus, 355

Pareiodon, 343, 344
micropterus, Cheirodon, 77

Megalamphodus, 7, 10, 12, 50, 54, 55
Microschemobrycon, 5, 7, 16

guaporensis, 12, 56
microstomus, Oligobrycon, 6, 12, 56, 57
minutum, Diceratherium, 404

Pygidium, 327, 329, 340
mipartitus, Elaps, 219
Miuroglanis, 262

platycephalus, 370, 371
Mixobrycon, 5, 10, 16
ribeiroi, 12, 62, 63
modesta, Helicops, 181

Pimelodella, 230, 231, 251, 252
monodon, Cheirodon, 12, 65, 71, 72, 73, 74
montanus, Glyptosaurus, 160
mucosa, Pimelodella, 230, 231, 235, 250
murinus, Eunectes, 169
mutisii, Eremophilus, 270, 341
mutus, Lachesis, 224

Myoglanis, 260 {en) Cf. Errata and Cor-
rigenda

narduccii, Elaps, 220
Natrix fasciata fasciata, 227
nattereri, Aphyocharax, 12, 23, 24

Cheirodon, 35

Holoprion, 23

Philodryas, 212

Thamnodynastes, 215
Nematogenyinse, 259, 275
Nematogenys, 259, 279

inermis, 280
nemurus, Pseudostegophilus, 349, 350
neuwiedi, Lachesis, 224

Xenodon, 200
nigricans, Pygidium, 289, 327, 328, 329
nigromaculatum, Pygidium, 306, 316, 317,

318
nigromarginatus, Leptophis, 185

484

MEMOIRS OF THE CARNEGIE MUSEUM.

niobrarense, Diceratherium, 403, 414, 420,

422, 424, 425
nodosus,'Glypto.saurus, 1 GO
notseus, Eunectes, 169
notomelas, Cheirodon, 7, 12, 65, 75, 74
Pimolodella, 230, 231, 233, 244

obscura, Echmatemys, 135, 13(), 137, 138,

139
occidentalis, Leptophis, 185

Rhinoceros, (Caenopus) 411
occipitalis, Rhadinea, 195
ocellatus, Glyptosaurus, 160
Ochmacanthus, 261, 345

batrachostoma, 356

flabelliferus, 356, 357

(Gyrinurus) batrachostoma, 367

reinhardti, 356, 357

Key to the species of, 356
octonarius, Hadrianus, 145
Odontostilbe, 17

drepanon, 3, 12, 90, 92

fugitiva, 90, 92, 93

hastata, 10, 12, 90, 91

melandeta, 9, 12, 90, 91, 98

microcephala, 12, 85, 94

paraguayensis, 12, 90, 96, 97

pulchra, 2, 12, 95

Key to the species of, 90
olfersi, Philodryas, 213
Oligobrycon, 7, 16

microstomus, 6, 12, 56, 57
operculatum, Scleronema, 280
Ophidia: A Catalog of the Ophidia from
South America at Present (June, 1916)
Contained in the Carnegie Museum
with Descriptions of some New Species.
By Lawrence Edmonds Griffin, 163-227
Opisthoglypha, 201
orbignianum, Pseudoplatystoma, 353
oregonensis, Rhinoceros, 412
orient ale, Prohyracodon, 404
orina, Rhadinea, 195
ornata, Anosteira, 121, 122
Oroya;, Pygidium, 274, 296, 298, 304
Ostariophysi, 14, 275

Oxybelis acuminatus, 211

fulgidus, 212
oxyptera, Paravandellia, 366, 367

pacifieus. Rhinoceros, 410

Diceratherium, 410
Pala^otheca polycypha, 139

terrestris, 139
palleum, Pygidium, 291
paolence, Pygidium, 327, 328
papilliferus, Spintherobulus, 12, 19, 20
Paragoniates, 10, 15

alburnus, 3, 12, 40, 41
paraguayensis, Aphyocharax, 12, 24, 25,
33

Holoprion, 23

Odontostilbe, 12, 90, 96, 97

Prionobrana, 3, 12, 13, 38, 40
parahybffi, Cheirodon, 12, 65, 70
Paravandellia, 262, 269, 359

oxyptera, 366, 367
Parecbasis, 3, 15, 48

cyclolepis, 12, 44, 45, 46
Pareiodon microps, 343, 344
Pareiodontinse, 261, 275
Pariolus armillatus, 260, 276
patagoniensis, Hatcheria, 281, 282, 283
Pearse, A. S., 274

pectinifera, Pimelodella, 230, 231, 232, 241
pequira, Cheirodon, 83

Holesthes, 3, 12, 84, 86
peruviana, Tachymenis, 214
peruvianus, Lachesis, 226
Peterson, O. A., 101, 103; The American

Diceratheres, 399
petersoni, Diceratherium, 413
petolaria, Clelia, 205
Phanagoniates, 6, 10, 15

wilsoni, 12, 43
Philodryas nattereri, 212

olfersi, 213

schotti, 213
Phreatobius, 260, 371, 372
Phrynonax fasciatus, 194
piaba, Cheirodon, 7, 9, 11, 12, 65, 71, 72,
76, 77, 79, 80

INDEX.

485

Pimelodella australis, 230, 231, 233
avanhandava^, 232, 240, 253
boliviana, 230, 231, 232, 242
brasiliensis, 230, 232, 242
buckleyi, 230, 231, 232, 240, 241, 245
chagresi, 230, 231, 252, 253
cristata, 230, 231, 236, 237
cyanostigma, 230, 231, 232, 242
eigenmanni, 230, 231, 234, 241, 248
eigenmanniforum, 241
elongata, 230, 231, 252, 254
euta^nia, 230, 231, 252, 254
gracilis, 230, 231, 232, 237, 238, 239,

251
griffini, 235, 250
grisea, 230, 231, 251, 252
hartti, 230, 231, 235, 248
hasemani, 230, 231, 232, 241
itapicuruensis, 230, 231, 234, 247
lateristriga, 230, 231, 245, 249, 250
laticeps, 230, 231, 234, 248
laticeps australis var. nov., 243
macturki, 230, 231, 234, 248
meeki, 230, 231, 235, 249
megalops, 230, 231, 232, 242
meta?, 230, 231, 233, 244
modesta, 230, 231, 251, 252
mucosa, 230, 231, 235, 250
notomelas, 230, 231, 233, 244
pectinifera, 230, 231, 232, 241
puruensis, 230, 231, 232, 240
rocese, 230, 231, 232, 240
steindachneri, 230, 232, 237
transitoria, 230, 231, 235, 249
vittata, 230, 231, 235, 249
vittata, 230, 231, 234, 247
serrata, 235
yuncencis, 251, 253
Distribution of, 230
Key to the Cis- Andean species of, 231
Key to tlie Trans-Andean species of,
251
Pimelodella and Typhlobagrus. By Carl

H. Eigenmann, 229-258
Pimelodina;, 280
Pimelodus cristatus, 229

Pisces, 275

pisciculus, Cheirodon, 2, 6, 12, 13, 64, 65,

67, 68
piscivorus, Agkistrodon, 227
platycephalus, Miurogianis, 370, 371
platyplastra, Baena, 120
plazai, Vandellia, 359, 360, 362
Plectospondyli, 14, 275
plicatilis, Dimades, 175
Poecilia bogotensis, 17
poecilogyrus, Liophis, 190
poeyanum, Pygidium, 297, 302
polycj^pha, Palcotheca, 139
polylepis, Helicops, 181
princeps, Elaps, 220

Glyptosaurus, 160
Prionobrama, 6, 11, 15
amazoni, 39
filigera, 38, 39, 40
filigerus, 3, 12, 13
madeirse, 37

paraguayensis, 3, 12, 13, 38, 40
Probolodus, 7, 15
heterostomus
Prohyracodon orientale, 404
proops, Pygidium, 295, 327, 331
Protaceratherium (' ' Diceratherium ' ' )

minutum, 403
Proteroglypha, 216
Psalidodon, 2
Pseudoboa coronata, 206
Pseudoplatystoma orbignianum corus-

cans, 353
Pseudoplatystomus, 267
Pseudostegophilus, 261, 277

nemurus, 349, 350
pulcher, Aphyocharax, 3

Cheirodon, 3, 35
pulchra, Odontostilbe, 2, 12, 95
jniUatus, 8i")ilotes, 197
punctatissimum, Pygidium, 327, 329, 340
punctatus, Henonemus, 345, 346
punctulatum, Pygidium, 297, 300
puruensis, Pimelodella, 230, 231, 232, 240
pusilla, Echmatemys, 140, 141, 142
pusillus, Aphyocharax, 12, 13, 22, 24, 28, 29

486

MEMOIRS OF THE CARNEGIE MUSEUM.

pusillus, Trichomycterus, 263
Pygidiidse, 275

Appendix to, 371
Chronolog}^ of, 270
Distribution of, 269
A Family of South American Cat-
fishes. By C. H. Eigenmann, 259-
398
First species discovered, 270
Generic Names and present equiv-
alents, 270
Guide to contents of monograph, 395
Habits of, 261
How distinguished, 259
Limits of the family, 276
Key to the subfamilies and genera,

278
Location of the types and specimens
in the museums of the world 271
Phylogenetic Tree showing relation-
ship, 277
Sources of material examined, 274,

275
Synonymy, 275
Zoological position, 275
Pygidiina;, 260, 275

Pygidium : Key to the species from Argen-
tina and Paraguay Basin, 291 ; from the
Amazon to the Essequibo, 324; from
Peru and Western Bolivia, 296; from
Southeastern Brazil, 328; from Vene-
zuela, Colombia, Panama, Ecuador, 305;
from Chile, 290; from Southeastern
Brazil, 327
Pygidium alternatum, 327, 328, 336
amazonicum, 325, 326
banneaui, 307, 318
barbouri, 274, 297, 303
bogotense, 306, 315, 316, 318
borellii, 294

brasiliense, 295, 327, 329, 336, 337
cahense, 306, 311
chapmani, 305, 309
conradi, 325
corduvense, 292
davisi, 327, 331, 334

Pygidium dispar, 297, 299

dorsostriatum, 290, 307, 320

eichorniarum, 295

eigenmanni, 297, 298

fassli, 298, 303

fuscum, 288, 296, 298

goeldii, 327, 328, 337, 339

gracilior, 325, 326

guianense, 325

hasemani, 325, 326

heterodontum, 296

iheringi, 327, 328, 330

immaculatum, 327, 328, 334

itatiays, 327, 329

kneri, 290, 306, 314

laticeps, 305, 307

latidens, 306, 312

latistriatum, 307, 318, 321

marmoratum, 290

merida^, 290, 306, 315

meta?, 290, 295, 306, 312

minutum, 327, 329, 340

nigricans, 289, 327, 328, 329

nigromaculatum, 306, 316, 317, 318

croya;, 284, 296, 298, 304

palleum, 291

paolence, 327, 328

poeyanum, 297, 302

proops, 295, 327, 331

parahybse, 332

punctatissimum, 327, 329, 340

punctulatum, 297, 300

quechuorum, 298, 305

regani, 307, 323

reinhardti, 327, 328, 333

retropinne, 306, 307, 324

riojanum, 295

rivulatum, 289, 297, 300, 301

santse-rit£e, 327, 329, 341

spegazzinii, 294

spilosoma, 307, 319

stellatum, 305, 308

stramineum, 306, 313

striatum, 306, 307, 321, 324

taczanowskii, 289, 297, 300

tsenium, 305, 310

INDEX.

487

Pygidium tenue, 291, 292
tigrinum, 291
triguttatum, 327, 329, 339
unicolor, 306, 314
venulosum, 307, 320
vermiculatum, 327, 328, 335
vittatum, 297, 299
zonatum, 328, 330

quechuorum, Pygidium, 298, 305

rathbuni, Aphyocharax, 12, 24, 31, 32

regani, Pygidium, 307, 323

reginse, Liophis, 191

reinhardti, Ochmacanthus, 356, 357

reinhardti, Pygidium, 327, 328, 333
Reptilia in the Carnegie Museum, 163-165

Localities in South America where
collected by J. D. Haseman, 164
reticulata, Typhlops, 166
retropinne, Pygidium, 306, 307, 324
Rhadinea merremi, 195, 196

occipitalis, 195

orina, 195
Rhinoceros bicornis, 444

hesperius, 410

oregonensis, 412

pacificus, 410
Rhinostoma guianense, 214
rhombifer, Clelia, 205

Drymobius, 178
Ribeiro, Dr. Alipio de Miranda, 64
ribeiroi, Cheirodon, 63

Mixobrycon, 12, 62, 63
riojanum, Pygidium, 295
riparia, Baena, 118

rivulatum, Pygidium, 289, 297, 300, 301
robustus, Hadrianus, 146, 147, 148
rocc«, Pimelodella, 230, 231, 232, 240
Rosenberg, W. F. H., 274
rostralis, Leptophis, 186
roulei, Atractus, 173
rugosus, Glyptosaurus, 160
Ruthven, Alexander G., 274

sanguinea, Vandellia, 359, 360, 365
santa^-ritse, Pygidium, 327, 329, 341

saturninus, Herpetodryas fuscus var., 183

Sauria, 159

schiffi, Diceratherium, 432

schotti, Philodryas, 213

Scleronema, 260

operculatum, 280
scutumantiquum, Amyda, 155, 157
semicincta, Tantilla, 209
semicinctus, Lj^strophis, 192
septaria, Echmatemys, 124, 125, 126, 127
Serpentes, 165
serrata, Pimelodella, 235
severus, Xenodon, 200
Siluridffi, 259, 260

Branchicolfe, 275, 276
Opisthopterse, 275, 276
sima, Baena, 112, 115, 117, 121
Smith, Mr. & Mrs. H. H., Collections of

snakes made by, 164
Sorubim, 267

spegazzinii, Pygidium, 294
sphenodon, Glyptosaurus, 160
spilogaster, Tropidodipsas, 197
spilosoma, Pygidium, 307, 319
Spilotes pullatus, 197
Spintherobulus, 6, 7, 14, 19
papilliferus, 12, 19, 20
Squamata, 159, 165
Stegophilinffi, 261, 275, 276
Stegophilus, 261

insidiosus, 267, 353, 354
intermedins, 345
Steinbach, Jose, Collections of snakes

made by, 165
steindachneri, Cheirodon, 35
Leptagoniates, 3, 12, 42
Pimelodella, 230, 232, 237
stellatum, Pygidium, 305, 308
stenodon, Cheirodon, 66, 80, 82
stevensoniana, Echmatemys, 132
stigeri, Metaca>nopus ( Accra therium), 429
Stout, Prof. Selatie E., 263
stramineum, Pygidium, 304, 313
stramineus, Aphyocharax, 32
striatum, Pygidium, 306, 307, 321, 324
Stylemys, 151
" Sucurujaba," 175

488

MEMOIRS OF THE CARNEGIE MUSEUM.

"Sucury," 175

Tachymenis peruviana, 214
taczanowskii, Pygidium, 289, 297, 300
tsniatus, Atractus, 173
ta:'nium, Pygidium, 305, 310
Tantilla melanocephala, 209

semicincta, 209
taxistigma, Cobitiglanis, 345
taxistigmus, Henonemus, 346, 348
Teleostomi, 14, 275
tenue, Pygidium, 291, 292
terrestris, Palseotheca, 139
terrificus, Crotalus, 226
Testudinidse, 143
Testudo ammon, 151, 154

ampliithorax, 155

brontops, 154

exornata, 154

uintensis, 150, 152, 153
Thamnodynastes nattereri, 215
Thayer Expedition to South America, 2
tigrinum, Pygidium, 291
titcombi, Hatcheria, 283, 284
transitoria, Pimelodella, 230, 231, 235, 249
Trichomj^cteridse, 276
Trichomycterus, 267

inermis, 279

pusillus, 263
tridactylus, Csenopus, 446, 448, 402
Tridens, 262

brevis, 370

melanops, 369

Key to the species, 369
Tridentina?, 261, 275
trigemina, Cleha, 206
Trigonias, 456

triguttatum, Pygidium, 327, 329, 339
Trionychida;, 102, 155
Tropidodipsas spilogaster, 197
trufiuianum, Diceratherium, 412
tumidus, Hadrianus, 147, 148, 150
Turtles (Fossil) of the Uinta formation.
By Charles W. Gilmore, 101-161; Ac-
credited to the Uinta formation, 104;
in the Carnegie Museum, 105.

Typhlobagrus kronei, 230, 255
Typhlopida>, 165
, Typhlops reticulata, 166

I'

unicolor, Pygidium, 306, 314
uintensis, Echmatemys, 127, 131

Testudo, 150, 152, 153
Urinophilous habit of members of the

Pygidiida', 262
Urinophilus, 358
uruguayanae, Macropsobrycon, 7, 12, 48,

49
utahensis, Hadrianus, 148, 149

Vance, Miss Lola, Fishes collected by, 274

reptiles collected by, 165
Vandellia cirrhosa, 262, 263, 358, 360

baseman i, 359, 360

plazai, 359, 360, 362

sanguinea, 359, 360, 365

wieneri, 360, 362, 363

Key to the species of, 360
Vandelliina;, 261, 275
venulosum, Pygidium, 307, 320
vermiculatum, Pygidium, 327, 328, 335
Viperidse, 222
viridis, Liophis, 192
vittata, Pimelodella, 230, 231, 234, 247
vittatum, Pygidium, 287, 299

Weber, Rudolph, 103
wesselii, Pimelodella, 237
Williamson, E. B., 262, 275
wieneri, Vandellia, 360, 362, 362
Wilson, Charles, 2
wilsoni, Phanagoniates, 12, 43

Xenodon colubrinus, 199
merremi, 199
neuwiedi, 200
severus, 200

yuccencis, Pimelodella, 251, 253

gonatum, Pygidium, 328, 330

44. Additions to Flora of Alleglveny County, Pa.

Jennings 16

45. New Species of Kneiffia. Jennings 05

46. Occurrence of Triglochln palustris in Pa. Jen-

nings 05

47. New Species of Ibidium. Jennings 10

48. Agate Spring Fossil Quarry. Peterson 10

49. Two New Birds from British E. Africa. Ober-

. HOLSER 05

50. The Chazy Formation and Its Fauna. Ray-

mond 1.50

51. A New American CJybele. Nabraway and Ray-

mond 15

52. Plastron of the Prot-^steginae. Wieland 15

53. New Species of Testudo from the Miocene, etc.

O. P. Hay 25

54. Miocene Beds of W. Nebraska and E. Wyoming

and Their Vertebrate Faunae. Peterson .... 1.00

55. New Species of Lonicera from Pa. Jennings. .05

56. Meryocochoerus, etc. Douglass.

57. New Merycoidodonts. Douglass. (Nos. 56

and 57 so!d together.) 1.00

58. Further Collections of Fishes from Paraguay.

EiOENMANN, McAtee, and Ward 1.25

59. Undetermined Element in the Osteology of the

Mosasauridae. Holland 20

60. Gasteropoda of the Chazy. Raymond 1.36

61. Occurrence of Wynea Americana In Pa. Jen-

nings 05

62. Account of Pleistocene Fauna Discovered at

Frankstown, Pa. Holland 10

63. Vertebrate Fossils from the Vicinity of Pitts-

burgh, Pa. Case 15

64. Elaenidae from the Black Biver Limestone near

Ottawa, Canada. Raymond and Naeraway.. .20

65. Rhinoceroses from the Oligocene and Miocene

of North Dakota and Montana. Douglass . . .25

66. Fossil Horses from North Dakota. Douglass. .30

67. Oligocene Lizards. Douglass .20

68. The Type of Stenomylus gracilis. Peterson . . .25

69. New Species of Birds from Costa Rica, etc.

Cakeiker 10

70. Costa Rican Formicariidae. Carriker 05

71. Vertebrate Fossils from the Fort Union Beds.

Douglass 45

72. List of the Lepidoptera of Western Pa., etc.

Engel 1.25

73. Fauna of Upper Devonian in Montana. Fossils

of Bed Shales. Raymond 60

74. New Species of Procamelus from Upper Mio-

cene of Montana. Douglass 30

75. Sections of the Conemaugh Series between

Pittsburgh and Latrobe, Pa. Raymond 35

76. List of the Unionidae of Western Pa. Obtmann .50

77. Geological Reconnaissance in North Dakota,

Montana, and Idaho. Douglass 1.00

78. Botanical Survey of Presque Isle. Jennings . . 2.75

79. Sesqui-Centennial (Pittsburgh) BeUcs. Stewart .45

80. Dromomeryx, New Genus of American Rumi-

nants. Douglass 30

81. FossUs from Glacial Drift and Devonian and

Mississippian near Meadville, Pa. Millard . . .10

82. New Species of Helodus. Eastman 05

83. Charles Chauncey Mellor. Holland 15

84. Reports of Expedition to British Guiana, etc.

Eigenmann 50

85. Reports of Expedition to British Guiana, etc.

Durbin 25

86. Odonata of the Neotropical Region, Exclusive

of Mexico and Central America. Calvert. . . 2.26

87. Deinosuchus hatcherl, a New Genus and Species

of Crocodile. Holland 20

88. Reports on Expedition to British Guiana, etc.

Blosser 15

89. Description of Some New Titanotheres from

the Uinta. Douglass 26

90. Annotated List of the Birds of Costa Rica In-

cluding Cocos Island. Carriker 3.00

91. Geology of the Coast of the State of Alagoas,

Brazil. Branner 40

92. Fossil Fishes from the Bituminous Shales at Ri-

acho Doce, Alagoas, Brazil. Jordan 55

93. Ordovician Trilobites, No. n. Asaphidae from

the Beekmantown. Raymond 36

94. Ordovician Trilobites, No. m. Raymond &

Nareaway .' 36

95. Ordovician Trilobites, No. IV. Raymond 40

96. Fishes from Irkutsk, Siberia. Jordan and

Thompson 30

97. South American Tetrigidae. Bkuner 1.00

98. Fauna of the Allegheny and Conemaugh Series

in Western Pa. Raymond 30

99. Ichthyological Survey About the San Juan

Islands, Washington. Starks 75

100. New Species of Pygidium. Eigenmann 10

101. The Brachiopoda and Ostracoda of the Chazy.

Raymond 50

102. A New Camel from the Miocene of Western

Nebraska. Peterson 15

103. Skeleton of Stenomylus hitchcocki, etc. Peter-

son 15

104. Skeleton of Diceratherium cooki. Peterson . . .16

105. Carnegie Museum Expedition to Central South

America, 1907-1910. Holland 15

106. Report Upon the Expedition of the Carnegie

Museum to Central South America. Hase^
MAN & Eigenmann 26

107. New Species of Fishes, etc., from Central South

America. Haseman 50

108. Annotated Catalog of Cichlid Fishes from Cen-

tral South America. Haseman 1.26

109. New Species of Fishes from the Rio Iguassu.

Haseman 65

110. Ornithology of the Bahama Islands. Todd &

WORTHINGTON 75

112. South American Acridoidea. I. Bbuner 1,00

113. Species of Hasemacla,- Hyphessobrycon, and

Hemigrammus. Ellis . .25

114. New Characins in the Carnegie Museum. Eigen-

mann 30

115. Jurassic Saurian Remains Ingested within Fish.

Eastman 20

116. Autograph Letter of U. S. Grant to Edwin M.

Stanton. Holland 15

117. Albert J. Barr. By W. J. Holland 10

118. Seventeen New Neotropical Birds. Todd 25

119. Dr. David Alter, the First Discoverer of Spec-

trum Analysis. Holland 10

120. Two Mummy Labels. Allen 10

121. The Families and Genera of Najades. Ort-

mann 1.00

122. Group of Stenomylins in the Carnegie Museum.

Peterson 25

123. Tertiary Fish-remains from Spanish Guinea.

Eastman 25

124. Plated Nematognaths. Ellis 65

125. New Species of Cambarus from the Isle of

Pines. Ortmann 15

126. Sedum Oamegiei, a New Species of the Family

Crassulaces, etc. Hamet 10

127. Two New Species of Fishes from Pern. Eioen-

MANN 15

128. South American Locusts (Acrldoidea). II.

Bbuner 75

129. Revision of the Genus Cheemepelia. Todd 85

130. New Genus and Species of Abyssinian Bodents.

CmLDS Frick 20

131. New Titanothere from the Uinta. Peterson.. .20

132. Small Titanothere from the Lower Uinta. Peter-

son 10

133. Osteology of Laslopyga and Callithrix, etc.

Shufeldt 25

134. New EhynchocephaUan from Solenhofen. Gbier .16

135. Scales of South American Characinid Fishes.

COCKERELL 26

136. Skeleton of Platigonus Leptorhlnus. Peter-

son 10

137. Lichens Collected in the Thunder Bay District,

Ontario. E. Heber Howe, Jr 10

138. Preliminary List of Fossil Plants in the Boof

of the Pittsburgh Coal. Grier 10

139. Undescribed Bemalns of the Uinta Titanothere

Dolichorhinus. Peterson 15

140. Triassic Fishes Belonging to the Catopteridse

and Semlonotldse. Eastman . . -. 15

141. Osteology of Promerycochoerus. Petersoi' ... .40

142. Correction of a Generic Name. Peterson 05

143. Skull of Bison Crassicomls. Holland 10

144. Serrasalminae and Myllnse. EigenmaKn 50

145. Heads and Tails: Notes on Sauropod Dinosaurs.

Holland 15

146. Dipterus Bemalns from Upper Devonian of

Colorado. Eastman 10

147. Notes on Tropical American Tettigonoldea (Lo-

custodea). Bruner 1.65

148. New Species of Tortoise from Jurassic of

Utah. GiLMORE 15

149. Fauna of Upper Devonian In Montana.

Haynes 45

150. New Sphagebranchus from Bahamas. Eioen-

MANN 07

151. Marine Fishes from Colombia and Ecuador.

Wilson 15

152. Apareiodon, a New Genus of Characld Fishes.

ElGENMANN 18

153. New and Bare Fishes from S. American Bivers.

ElGENMANN 25

154. Three New Species of Characld Fishes. Eigen-

MANN & HeNN 12

155. The Species of Salmlnus. Eigenmann 7

156. South American Pceciliid Fishes. Henn .40

157. A New Species of Apatosaurus. Holland 7

158. Birds of the Isle of Pines. Todd 70

159. BeptUes and Amphibians of the Isle of Pines.

Barbodr 20

160. Land and Fresh Water Shells of the Isle of

Pines. Henderson 20

161. Pelecypoda of the Chazy. Eatmond 45

162. S. American Crickets, Gryllotalpoidea and Ache-

toidea. Bruner 1.30

163. Preliminary Catalog of N. American Sphasrl-

idae. Sterki 75

164. Directions for Collecting Sphseriidae and

Aquatic Gastropods. Sterki 10

165. Lepidoptera of the Isle cf Pines. Holland 50

166. Odonata of the Isle of Pines. Kahl 15

167. A Trip to Islands In Lake Erie. Goodrich 15

168. LaQd-Shells of Islands at Western End of Lake

Erie, etc. Clapp 20

169. Orthoptera of the Isle of Pines. Holland &

Kahl 16

170. Obituary Notes. G. A, Link, T. A. MiUs, Boyd

Crumrine, E. M. Bigelow. Holland.

171. Two New W. African Bhopalocera. Holland. $0.15

172. Botany of the Isle of Pines. Jennings 2.00

173. List of Hymenoptera of Isle of Pines, Holland .10

174. Some Species of Farlowella. Eigenmann 30

175. List of S. American Lizards. Griffin 35

176. Leptodeira albofnsca (Lacdp6de). Griffin 10

177. List of Coleoptera collected on Isle of Fines.

Holland & Schwarz ' 15

178. Bhynchota of the Isle of Pines. Heidemann

& Osborn 10

179. Mammals of the Isle of Pines. Holland 05

180. Beport upon Fossil Material Collected In Cave

in the Isle of Pines. Peterson 10

181. New Species of Fern (Polystichum jenningsi).

Hopkins 10

182. Notes upon the Genus Leucophenga Idik with

Descriptions of New Species. Kahl 30

183. Some Species of Bhamdia, a Genus of S.

American SUuridae. Eigenmann & Fisher. .15

184. New and Bare Species of South American

SUurldse. Eigenmann 30

185. List of the Hypopthalmidse, Dlplomystidae, and

some Unrecorded Species of SUurldse. Fisher. .30

186. A Synopsis of the Saurian Genus Prionodac-

tylus. Griffin 05

187. Notes on a Collection of Fishes from Ceylon

with Descriptions of New Species. Jordan

& ST.4KKS 35

188. Collection of Fishes from Port Said, Egypt.

Jordan & Hubbs 20

189. A Fossil-bearing Alluvial Deposit In Saltville

Valley, Virginia. Peterson 10

190. Catalog of Collection of Watches belonging

to H. J. Heinz. Stewart, Holland, Cogges-
HALL Paper, 30 eta.; hoards .50

191. Contributions to the Natural History of the

Isle of Pines: Eeprinted from Annals, Vols.
VIII-XI, 560 pp., 34 plates, index. Bound
in cloth 5.00

192. Material Discovered in the Upper Eocene of

the Uinta Basin by Earl Douglass and O. A.
Peterson. Peterson 2.25

193. The Fresh-Water Fishes of the Island of For-

mosa. Masamitsu Oshima 2.75

194. On Elepheuor, a New Genus of Fishes from ?

Japan. DAvro Starr Jordan 50

195. A Description of Cypripedium passerlnmn.

Jennings 10

196. Obituary of Charles Bochester Eastman 10

197. Obituary of Andrew Arnold Lambing 10

198. Obituary of Herbert Huntington Smith 10

199. Obituary of Henry John Heinz 10

3 2044 072 225 618

Date Due

s*^^^:e=^9

i4

., 1 ''"■■■

'»!,(.!