/

ANNALS

OF THE

CARNEGIE MUSEUM

Volume XXIX

1942-1943

\

Published by the Authority of the Board of Trustees of the Carnegie Institute Pittsburgh, Pa.

October, 1943

ARRANGED FOR PUBLICATION BY

Arthur W. Henn

CARNEGIE

INSTITUTE PRESS

TABLE OF CONTENTS

Title Page i

Title of Contents iii

List of Genera and Species new to Science v

ARTICLE

I. List of the Tinamous in the collection of the Carnegie

Museum. By W. E. Clyde Todd (Feb. 27, 1942) .... 1-29

II. Observations on the life history of a new Chalcidoid Wasp, an internal parasite of Ant-lion larvae. By George E. Wallace, pi. I (March 13, 1942) 31-40

III. Upper Cretaceous fauna of the Asphalt Ridge, Utah. By

I. P. TolmachofT, pis. I ll (May 5, 1942) 41-60

IV. A review of the genus Phoca. By J. Kenneth Doutt,

pis. I-XIV (May 12, 1942) 61-125

V. The Cephalopod fauna of the Conemaugh series in western Pennsylvania. By A. K. Miller and A. G. Unklesbay, pis. I-VIII (June 12, 1942) 127-174

VI. Descriptions of two new Salamanders from peninsular Florida. By M. Graham Netting and Coleman J. Goin, pi. I (June 16, 1942) 175-196

VII. Critical remarks on the races of the Sharp-tailed Sparrow.

By W. E. Clyde Todd (June 16, 1942) 197-199

VIII. The Canadian forms of the Sharp-tailed Sparrow, Am-

mospiza caudacuta. By James L. Peters (Sept. 3, 1942). . 201-210

IX. Description of a new race of Siren intermedia Le Conte.

By Coleman J. Goin (Sept. 3, 1942) 211-217

X. The Chazy Conularida and their congeners. By G. Wins- ton Sinclair, pis. I-III (Oct. 1, 1942) 219-240

XI. Scolecodonts from the Erindale, Upper Ordovician, at Streetsville, Ontario. By E. R. Eller, pis. I-IV (Nov.

6, 1942) 241-270

XII. List of the Hummingbirds in the collection of the Car- negie Museum. By W. E. Clyde Todd (Dec. 31, 1942) . 271-370

XIII. A collection of Lepidoptera (Rhopalocera) from the

Cayman Islands. By G. D. Hale Carpenter and C. B.

Lewis (Jan. 15, 1943). 371-396

XIV. New and rare Ithomiinae (Lepidoptera) in the Carnegie

Museum. By Richard M. Fox, pi. I (March 5, 1943) .397-408

XV. Birds collected during two cruises of the “Vagabondia” to the west coast of South America. By Ruth Trimble

(June 3, 1943) 409-441

XVI. New Neididae (Hemiptera) from South America, with notes on some little-known species. By Halbert M.

Harris (June 30, 1943) 443-450

Index 451-468

IV

LIST OF GENERA AND SPECIES NEW TO SCIENCE

Stomatoceras rubra eriensis, var. nov. Insecta

31

Cardium kayi, sp. nov. Lamellibranchiata

54

Crenella burkei, sp. nov.

. . . . 57

Phoca vitulina mellonae, subsp. nov.

Mammalia

111

Poterioceras subellipticunij sp. nov. Cephalopoda

135

Pennoceras, gen. nov.

u

146

Pennoceras seamani, sp. nov.

u

147

Pseudotriton montanus floridanus, su

bsp. nov. Amphibia. . .

... 175

Pseudobranchus striatus axanthus, subsp. nov. “

183

Siren intermedia nettingi, subsp. nov.

u

211

Conularina, gen. nov. Conularida

220

Conularina undosa, sp. nov.

u

222

Conularina irrasa, sp. nov.

«

.223

Conularina raymondi, sp. nov.

u

.223

Conularina narrawayi, sp. nov.

u

224

Climacoconus, gen. nov.

u

225

Climacoconus rallus, sp. nov.

u

. . .228

Climacoconus humilis, sp. nov.

u

.228

Climacoconus clarki, sp. nov.

u

.229

Climacoconus bromidus, sp. nov.

a

.230

Climacoconus pumilus, sp. nov.

u

.231

Lumbriconereites marlenediesae, sp.

nov. Annelida.

. 243

Lumbriconereites proclivis, sp. nov.

u

244

Lumbriconereites deflexus, sp. nov.

u

... 244

Lumbriconereites copiosus, sp. nov.

u

.245

Nereidavus ineptus, sp. nov.

u

. 246

Nereidavus hamus, sp. nov.

u

.246

Nereidavus procurvus, sp. nov.

a

. ... 247

Ildraites exquisitus, sp. nov.

u

... 248

Ildraites fritzae, sp. nov.

u

. ... 248

Ildraites patulus, sp. nov.

u

249

Palecenonites, gen. nov.

u

250

Paleoenonites accuratus, sp. nov.

a

250

V

Palecenonites latissimus, sp. nov.

u

251

Palecenonites edentulus, sp. nov.

u

251

Paleoenonites accuratus, sp. nov.

u

252

Eunicites denticulatus, sp. nov.

u

252

Eunicites purus, sp. nov.

u

253

CEnonites conterminus, sp. nov.

a

254

(Enonites crepitus, sp. nov.

a

254

CEnonites sinuatus, sp. nov.

a

255

CEnonites caducus, sp. nov.

a

256

Leodicites acclivis, sp. nov.

u

256

Leodicites streetsvillensis, sp. nov.

a

257

Leodicites creditensis, sp. nov.

u

257

Leodicites summus, sp. nov.

u

258

Leodicites barbatus, sp. nov.

u

259

Leodicites densus, sp. nov.

u

259

Staurocephalites cuspis, sp. nov.

u

260

Arabellites perpensus, sp. nov.

a

260

Diopatraites fustis, sp. nov.

u

261

Colibri cyanotus crissalis, subsp. nov.

Aves

292

Chlorostilbon stenurus ignotus, subsp. nov.

a

. ... .305

Lepidopyga luminosa phaeochroa, subsp. nov.

u

. ... .308

Amazilia chionopectus orienticola, subsp.

nov.

u

318

Amazilia fimbriata elegantissima, subsp.

nov.

a

323

Amazilia amabilis costaricensis, subsp. nov.

u

330

Chalybura buffoni interior, subsp. nov.

u

332

Ocreatus underwoodi polystictus, subsp.

nov.

«

347

Brephidium exilis thompsoni, subsp. nov.

Lepidoptera

392

Hypo thy ris meterus deemae, subsp. nov.

«

398

Hyalyris deuscula, sp. nov.

u

399

Hyalyris munda, sp. nov.

u

401

Napeogenes astarte, sp. nov.

u

. . . . .402

Oleria crispinella hemina, subsp. nov.

u

404

Xenoloma, gen. nov. Hemiptera

443

Xenoloma princeps, sp. nov.

444

Phaconotus, gen. nov.

445

Phaconotus ensis, sp. nov.

446

Protacanthus nexus, sp. nov.

446

Parajalysus pallidus, sp. nov.

449

Parajalysus nannus, sp. nov.

vi

449

7- <\ ART. I. LIST OF THE TINAMOUS IN THE COLLECTION OF

THE CARNEGIE MUSEUM

By W. E. Clyde Todd

Introduction

The Tinamous (Family Tinamidae) in the Carnegie Museum total 412 specimens, belonging to 61 species and subspecies. Although only about half of the known forms are represented, so many interesting findings have come to light in the course of their determination that I have thought it well to put the results on record for the benefit of other workers. The Tinamous are an especially difficult group, and their study has been unduly handicapped not only by the paucity of material, but also by its (generally) poor quality. Few collections can boast satisfactory series of more than a few species, and it is usually necessary to assemble ma- terial from several sources for purposes of study and comparison. While the Tinamou material in the Carnegie Museum is above the average in quality, the series of most forms are small. In identifying some of them I have had to draw on the collections of other museums, in particular the American Museum of Natural History and the Academy of Natural Sci- ences of Philadelphia. To the authorities of these institutions my acknowl- edgments are due. I am particularly indebted also to Mr. N. B. Kinnear, who in June, 1938, enabled me to examine the Tinamous in the collection of the British Museum (Natural History). Also, Dr. C. E. Hellmayr and Mr. Ludlow Griscom have been good enough to make comparisons of specimens sent them for that purpose.

It should be explained that the present list includes all the specimens of Tinamous entered in the Museum catalogue, whether now in the collec- tion or alienated therefrom by exchange. Remarks on forms not repre- sented in our collection are also inserted at their proper places. The sequence of the list closely follows that of Mr. J. L. Peters’ “Check-List of Birds of the World,” not because I consider this the final word but merely a standard work of reference and a convenient point of departure.

Since the publication of Mr. Peters’ list, several papers dealing with the systematics of certain species have appeared. Some new forms have been

1

Issued February 27, 1942.

2

Annals of the Carnegie Museum

vol. XXIX

described, and changes in the status of other forms have been proposed. Some authors manifest an increasing tendency to combine related forms under one specific head and to alter their nomenclature accordingly, as the “formenkreis” theory demands. With these views I cannot always agree. While subspecies are of course “representative forms,” the con- verse is not necessarily true. Thus to enlarge and change the original concept of the subspecies is to go far beyond and outside the plan and purpose of the trinomial system of nomenclature, for the adoption of which a bygone generation of zoologists had to contend so long and so vigorously. We realize — all too keenly — the shortcomings of the system to express all the facts as we find them in Nature, but if the present tendency to load too much on the names continues, the system will fall of its own weight. I can only deplore the proposals to reduce a number of well-characterized forms of this particular group to subspecific rank, and in the present paper I have indicated what I consider to be their true status, insofar as my material would justify an opinion.

Two papers dealing with the taxonomy of the Tinamous in general have appeared in the last decade and merit special notice. Dr. Hans von Boetticher has a very important paper in German, published in 1934 (see bibliography). In this paper the author discusses the relationships of the group, the color patterns of the several genera, and the characters ex- hibited by the bill, nostrils, tarsi and toes, and tail. He takes into con- sideration also the ecological distribution of the genera before presenting his scheme of classification and a genealogical tree. He divides the family into three subfamilies: Tinaminae, to include Crypturellus , Tinamus, and Nothocercus; Rhynchotinae, with Rhynchotus , Nothura, Nothoprocta, and Taonisciis; and Eudromiinae, with Tinamotis and Eudromia. This ar- rangement has been carefully worked out, and it has much to commend it. The author makes no attempt to arrange the species within the genera.

Sr. Alipio de Miranda-Ribeiro’s paper in Portuguese appeared in 1938. Although obviously prepared without reference to Dr. von Boetticher’s 1934 paper, it covers the same ground (in some respects more fully), but he bases the systematic discussion solely on the forms found in Brazil. These are treated in some detail and are accompanied by lists of speci- mens and references to the literature. Significantly, he groups the Brazilian genera into subfamilies just as Dr. von Boetticher does, al- though not exactly on the same basis. In addition, he splits Crypturellus and sets up a new genus, Orthocry pturus, for C. variegatus and its allies. For C. cinereus he accepts Crypturornis. (To these groups I would give

1942

Todd: Tinamous in the Carnegie Museum

3

subgeneric status.) This paper embodies certain good suggestions for the proper arrangement of species and races, although the nomenclature is faulty in some respects, and at least two of the new names proposed were anticipated.

In this paper all measurements are in millimeters, and the names of colors are in the main taken from Ridgway’s “Color Standards and Color Nomenclature.”

List of Species Tinamus tao tao Temminck.

One specimen: Apacy, Rio Tapajoz, Brazil.

The upperparts are purer gray than in the other races of this species, and the black bars are distinct. Our single specimen comes from the west bank of the Rio Tapajoz. However, the range of this race is actually more extensive than Peters (1931b, 12) allows, since two specimens from Santatem, on the east bank of the same river, are recofded by Oliveifa Pinto (1938, 2), and one from Cussary, farther down: on the Amazon, is listed by Snethlage (1914, 46). Temminck’s type-specimen came from the “province of Para.”

Tinamus tao septentrionalis Brabourne and Chubb.

Four specimens: La Cumbre de Valencia, Lagunita de Aroa, and Puerto La Cruz, Venezuela.

This appears to be a valid race, easily distinguished from typical tao by its more uniform upperparts, which are vermiculated rather than barred and have a slight olive wash in evidence. A young bird (September 24) resembles the adults except for the very small whitish spots on the wings.

Tinamus tao kleei (von Tschudi).

Four specimens: Cerro del Amboro and Cerro Hosane, Bolivia.

These are fully as heavily barred above as our single specimen of typical tao, but they have a decided olive wash. They have not been directly compared with topotypical Peruvian specimens of kleei, but are referred thereto on the authority of Hellmayr, who (so H. B. Conover writes me) is now inclined to consider Bolivian birds the same. The latter were de- scribed under the name Tinamus weddelli by Bonaparte (1856, 881, 954), which name will thus become a synonym of kleei, as suspected by Peters.

4

Annals of the Carnegie Museum

vol. XXIX

Tinamus major robustus Sclater and Salvin.

Ten specimens: Manatee Lagoon, Quamin Creek, Toledo District, and Cockscomb Mountains, British Honduras.

“Iris hazel: bill leaden brown [brownish plumbeous?]; feet plumbeous” (Peck).

There is considerable individual variation affecting the precise shade of the upperparts, as well as the amount of black barring thereon. A British Honduras bird mentioned by Salvin and Godman (1904, 449) is paler than Guatemala skins, but this is doubtless without significance. A half-grown young bird, readily distinguished by the small buffy spots on the upperparts, is dated July 12. These indications of immaturity persist on the tertiaries of another fully grown bird taken as late as December 7.

On the status and relationships of the Central American races compare Chapman, 1917, 187, and Griscom, 1929, 150. These authors insist that the forms with rufous heads {major, etc.) and those with gray heads (robustus, fuscipennis) are conspecific. Nevertheless, it is interesting to note that both types (according to Griscom) occur in the Canal Zone and still maintain their respective characters. In western Ecuador, according to Chapman (1926, 146), both types are also found indiscriminately.

Tinamus major fuscipennis Salvadori.

Five specimens: Cuabre, Rio Sicsola, Carrillo, and El Hogar, Costa Rica.

No. 23,886 (Cuabre) is practically indistinguishable from British Honduras robustus, to which form both Carriker and I at one time referred it. This specimen has the same olivaceous color above as robustus, and grayish shading below, with little buff. But since the other two speci- mens from this region are clearly fuscipennis, it seems better to consider this odd example an extreme individual variant such as sometimes occurs within the range of a given form. The only other alternative would be to give both robustus and fuscipennis specific rank — a manifest absurdity. The remaining four specimens, although varying considerably among themselves, exhibit the differences pointed out by Griscom ( l.c .) fairly well as a series. One has the back very nearly uniform. The markings on the secondaries vary from fine vermiculations to prominent bars, but the series is much too small to say whether or not this is a character due to age.

Tinamus salvini of Underwood (type-locality Carrillo), as maintained by Salvin and Godman, is unquestionably fuscipennis in immature dress.

1942

Todd: Tinamous in the Carnegie Museum

5

Tinamus major castaneiceps Salvadori.

Five specimens: Pozo Azul de Pirris and El Pozo de Terraba, Costa Rica.

Clearly this form is only a race of the South American T. major , with which it is connected through latifrons and probably through ruficeps also. Our Pozo Azul adult is decidedly brownish above; the others are more olivaceous. A young bird, dated May 16, resembles the adults but has buffy spotting above, as is usual in this group.

Tinamus major saturatus Griscom.

Two specimens: Murindo and Malagita, Colombia.

On geographical grounds these should belong to saturatus , but they fail to show the comparative characters claimed for that race, except for an obvious occipital crest. Their coloration is slightly lighter than that of Costa Rican specimens of castaneiceps , especially on the underparts, which are more uniform, with the barring less in evidence; the crown, however, is not appreciably different. Mr. Griscom has been kind enough to compare these two specimens with the type-series and writes as fol- lows: “Your two specimens agree in coloration with the paler extremes of our excellent series of saturatus; in other words, they are not really typical of this subspecies. You will note, however, that the development of the occipital crest in your two specimens is very much less than in saturatus , nor am I able to see that this difference is due to any defects in your two specimens.”

Pending the receipt of more material I shall provisionally refer these examples to saturatus. It is entirely possible that they may prove to be- long to a race intermediate between saturatus and latifrons.

Tinamus major zuliensis Osgood and Conover.

Seven specimens: Valparaiso, La Tigrera, Las Vegas, and Pueblo Viejo, Colombia; Rio Mocho, Venezuela.

I had at one time (following Chapman) identified these as ruficeps (type from eastern Ecuador), but I now agree with Osgood and Conover that they should be referred to their race zuliensis , a form which ranges from the Caura Valley in Venezuela to the Magdalena Valley in Colombia (in the Tropical Zone). Compared with typical major, it is a pale race; this pallor appears on both the upper- and underparts; the occipital crest is not strongly marked. The Arid Tropical habitat of this race is reflected

6

Annals of the Carnegie Museum

vol. XXIX

in these characters. T. m. zuliensis is very different from castaneiceps , the range of which it touches on the west.

Tinamus major major (Gmelin).

Eight specimens: Rio Yuruan, Venezuela; Tamanoir and Pied Saut, French Guiana; Upper Araucaua and Obidos, Brazil.

Salvadori (1895, 504) calls this form subcristatus Cabanis, mainly on the ground that so-called major had not been found in recent times at Cayenne, and because of the misidentification of the name. But von Berlepsch (1908, 298) insists on the pertinence of the earlier name, after throwing out Marcgrave’s reference.

This form has the following comparative characters: buffy suffusion below; a decided occipital crest; dark-colored head, with the sides and front more or less dusky or ashy, contrasting with the crown; and little or no rufescent barring or squamation on the lower neck in front. The color of the upperparts varies greatly, also the extent and heaviness of the barring. A slightly immature bird from Obidos is heavily barred above and deeply washed with brown, whereas an adult female from the same place is decidedly more olivaceous above and lightly barred.

Tinamus major serratus (Spix).

Two specimens: Tonantins and Manacapuru, Brazil.

Hellmayr (1906, 699 et seq.) argues at length to prove the specific dis- tinctness of serratus and major. He argues that since birds with long occipital crests and ashy foreheads occur together with others having no crests and plain rufous foreheads, there must be two species (this in the Rio Negro region). Evidently he has changed his mind since, if one may judge from Conover’s later remarks (1937, 192). Conover has handled both specimens here listed. They are very different from each other but agree in having a bright rufous and uniform pileum, the sides of the head distinctly rufous, and much buffy and rufous suffusion and squamation on the neck in front; the occipital crest is short.

The specimen! from Tonantins is almost immaculate above as compared with the Manacapuru bird, which is strongly barred.

Tinamus major olivascens Conover.

Nine specimens: Rio Surutu (near Buena Vista) and Buena Vista, Bolivia; Villa Braga (Rio Tapajoz), Hyutanahan (Rio Purus), and Nova Olinda (Rio Purus), Brazil.

1942

Todd: Tinamous in the Carnegie Museum

7

This name has been set up by Conover to cover all the birds of this species from south of the Amazon, from Para to Bolivia. The above specimens agree in generally dark coloration — dark olive gray breast and sides, less buffy, more whitish abdomen, and deep rusty (and uniform) pileum with virtually no occipital crest. The amount of barring above and the exact shade of color are variable characters, as in other forms of this species; accordingly the name olivascens is not entirely appropriate. The sides of the head are strongly rufescent, as in serratus, but this shad- ing does not extend to the neck in front, as in that form.

Tinamus guttatus von Pelzeln.

Sixteen specimens: Benevides, Colonia do Mojuy, Villa Braga, Bella Vista, Hyutanahan, Nova Olinda, Arima, Sao Paulo de Olivenga, and Tonantins, Brazil.

If the female is constantly larger than the male, as is usual with Tina- mous, then several of our skins must be wrongly sexed.

At first glance these seem to fall into two series, according to locality. The lower Amazon birds appear whiter below, while those from the upper Amazon and Rio Purus are more buffy, less vinaceous. Also, the buffy spotting on the neck in the latter is richer, more ochraceous. These dif- ferences are better marked in the males, but since certain specimens from both regions are virtually indistinguishable, I do not think the discrimina- tion of an eastern race is feasible. The upperparts also vary in ground- color from rich umber brown to decidedly olivaceous, and in the extent and width of the black barring, but I am not convinced that these varia- tions are geographical. They parallel those shown in T. major.

Borba, Rio Madeira, has been fixed as the type-locality by Hellmayr (1907, 409), hence if any subdivision of the species were made, the eastern birds would constitute the new race.

Nothocercus nigrocapillus nigrocapillus (Gray).

One specimen: Incachaca, Bolivia.

This example differs from Salvadori’s description and plate (1895, 511, pi. 8) in being more brownish, less rufescent throughout, and in having the sides of the head dusky slate-color like the crown. These discrepancies must be of an individual character, however, since the type, although ascribed by Gray to Chile, could have come only from Bolivia. Carriker (1933, 2) has described a race from Peru.

8

Annals of the Carnegie Museum

vol. XXIX

Nothocercus julius julius (Bonaparte).

One specimen: “Bogota,” Colombia. A young bird.

Nothocercus bonapartei frantzii (Lawrence).

Two specimens: Volcano Turrialba and Ujaras de Terraba, Costa Rica.

The Turrialba specimen is much more rufescent throughout than the other, but a series would be necessary to establish geographical variation. Incidentally, this particular specimen was originally labeled “Guacimo” and later “Cartago.” However, in Carriker’s list (1910, 377) it is said to come from the “Volcan Turrialba, 4,000 feet,” which locality is probably correct, since the species belongs to the Subtropical Zone. It is best re- garded as a race of the South American N. bonapartei , in spite of its separated range.

Nothocercus bonapartei bonapartei (Gray).

Two specimens: La Cumbre de Valencia and Paramo de Rosas, Vene- zuela.

The first is a young bird (September 12); it closely resembles the adult in coloration, but the dusky of the crown is flecked with grayish, and the throat is grayish, although sparsely feathered. In the adult the outer primaries are rufescent, grayish- tinged, and have narrow dusky bars, precisely as in the “Bogota” specimens mentioned by Salvadori (1895, 512). The wing is about nine inches long — longer than in the type. A “Bogota” skin in the American Museum Collection has the wing 210 mm. long.

Crypturellus cinereus cinereus (Gmelin).

Four specimens: Cayenne and Pied Saut, French Guiana; Upper Araucaua, Brazil.

Having already discussed the taxonomy of Crypturellus cinereus at some length (1938, 123-126), I shall need to give here only an abstract of my conclusions.

Our specimens correspond closely to the description of Tetrao cinereus Gmelin (ex Buffon), which came from Cayenne, French Guiana. The name was later extended to apply to birds from other parts of the range, which was traced southward into Brazil and westward to the Andes. Several writers had noted certain variations in color characters, but it

1942 Todd: Tinamous in the Carnegie Museum 9

remained for Brabourne and Chubb (1914, 320) to discriminate a dark- colored form of Tinamou from British Guiana under the name Crypturus macconnelli. Their description was elaborated by Chubb two years later (1916, 8, pi. 1, fig. 1), in connection with the form supposed by him to be the true C. cinereus, which was figured also on the same plate. But I am convinced that in describing macconnelli Chubb merely renamed cinereus , mistaking for the latter the lighter-colored rufescent form which ranges from the interior of British Guiana to the Amazon Valley. Our topo- typical specimens from French Guiana obviously belong to this dark- colored form, which must be called cinereus , of which macconnelli is thus a pure synonym. Miranda-Ribeiro (1938, 758) has reached precisely the same conclusion.

Since my paper appeared, I myself have examined the series in the col- lection of the British Museum and found my conclusions verified. The differences between the dark and light birds are entirely bridged over in the series of British Guiana specimens. Some marked “ macconnelli ” (by Chubb himself) are decidedly rusty brownish above. Two supposed Cayenne birds (very old skins) are more rusty than ours, but they were probably dark-colored birds originally. The plate of macconnelli is some- what misleading, since the type-specimen is actually not so dark-colored as thereon depicted. The light-colored bird of the plate will stand as

Crypturellus cinereus rufescens Todd.

Seven specimens: Villa Braga, Miritituba, Hyutanahan, and Arima, Brazil.

Miranda-Ribeiro (1938, 760) correctly discriminated this race but wrongly called it assimilis Schlegel — a name which has no standing what- ever, since it is based on a misidentification of Nothura assimilis Gray.

In addition to these specimens I have handled the series in the British Museum, as already said, and certain specimens from the collection of the American Museum of Natural History. “Chubb’s plate is somewhat overdrawn ; the differences between the two forms, while perfectly obvious on comparison, are not so conspicuous as there indicated” (Todd, 1938, 125). Two adults from the Rio Purus vary in the direction of the Bolivian race {cinerascens) , while three Peruvian examples (in the British Mu- seum) vary among themselves, but are probably referable to cinerascens or fumosus. The several forms of this group are certainly very close to each other and are subject to considerable variation.

10

Annals of the Carnegie Museum

vol. XXIX

Crypturellus cinereus subsp.

One specimen from Benevides (near Para), Brazil, as already remarked, may represent an undescribed race, but more specimens are needed to determine its status. Incidentally, I have examined the type-specimen of Crypturus megapodius Bonaparte (1856, 954), now in the British Museum, and I have found it to be a young bird of some form of C. cinereus , but which one is quite indeterminable.

Crypturellus berlepschi Rothschild.

One specimen: Potedo, Colombia.

Examination of additional material of this form in the British Museum confirms my belief that it is a perfectly distinct species, peculiar to the Colombian- Pacific Fauna. “Its darker coloration, pure black pileum, lack of white shaft-stripes on the throat, and differently colored bill seem to be good specific characters when compared with cinereus ” (Todd, 1938, 126).

Crypturellus castaneus (Sclater).

As shown by material in the British Museum, this species resembles C. obsoletus, but the upper- and underparts are chestnut; the head is darker gray.

Crypturellus cerviniventris (Sclater and Salvin).

After examining the type-specimen in the British Museum, I am de- cidedly of the opinion that this form is not conspecific with either C. castaneus or C. obsoletus. I think it should stand as a full species.

Crypturellus obsoletus punensis (Chubb).

Nine specimens: Cerro Hosane, Locotal, San Jose, and Incachaca, Bolivia.

MEASUREMENTS

No. Sex Wing Bill Tarsus

79409 ($ im.) 154 23 42

85274 $ 158 24 44

85284 <? 152 23 45

85584 <f 144 23 43

85641 $ 157 25 46

85676 $ 151 24 43

119582 $ 154 24 45

120420 & 149 23 45

1942

Todd: Tinamous in the Carnegie Museum

11

As represented by specimens in the British Museum, C. obsoletus obsoletus has the head gray all around, and the back and scapulars olive brown, and the underparts dull rufous, posteriorly barred with buff and dusky. C. obsoletus punensis does not differ greatly, but the posterior underparts are somewhat deeper buffy. Three Bolivian birds examined appeared to be the same as those from Peru. The type of punensis is close to the type of castaneus, but is paler, with less chestnut above and below, while the slaty gray of the head extends over the back as a wash. The throat appears grayer, but in the type of castaneus this color may have faded somewhat. In size the two forms agree.

Our specimens have been compared with four in the Rothschild Col- lection from southern Brazil (Parana, Minas Geraes, Sao Paulo) which supposedly represent true obsoletus. The Brazilian birds are appreciably different, in that they are paler, more ochraceous, less cinnamomeous, below, with the abdomen and crissum more buffy. The upperparts, too, are lighter brown (less rufescent), especially the wings and tail, and lack the grayish wash evident in the Bolivian skins. Our series of the latter have not been directly compared with topo typical material from Peru ; they have been identified as punensis on the basis of their agreement in size and coloration with the description, and of the comparisons made at the British Museum. Chubb’s name was based on birds from Puno, Peru, in the same general faunal region from which our birds came. Very recently, Bond and de Schauensee (1941, 1), after examining topotypes of punensis , have described the Bolivian bird as a distinct race, crucis, on the basis of its larger size, darker gray throat, and duller underparts. My measurements do not agree with theirs, but this may well be due to the personal equation. The proposed new race may indeed be a valid one, but I am reluctant to accept it on the present evidence.

Crypturellus griseiventris (Salvadori).

Nine specimens: Santarem and Colonia do Mojuy, Brazil.

The series is fairly uniform; the extent and intensity of the vinous chestnut of the breast is the most variable feature. No. 72,421 (April 30), a young bird, shows some fine white spots above and obsolete bars below.

Admittedly this form is related to C. obsoletus , but in my judgment not so closely as to be reduced to a subspecies thereof, as given by Peters. The ranges of these two forms are not known to approximate each other, since griseiventris is confined to the lower Amazon Valley east of the Rio

12

Annals of the Carnegie Museum

vol. XXIX

Tapajoz. Representatives of this form in the British Museum have been compared with cerviniventris , punensis , and castaneus; in my opinion they are not conspecific with any of these forms.

Crypturellus soui meserythrus (Sclater).

Three specimens: Manatee Lagoon, British Honduras.

In addition to these, I have handled (in the Museum of Comparative Zoology) two other specimens from the same locality.

“Iris dull yellow or yellowish brown; bill leaden brown, paler below; feet dull greenish yellow or yellowish brown” (Peck).

Sclater’s name was originally applied to birds from Playa Vicente, Oaxaca, Mexico, from which region I have seen no specimens. The male of this form is noticeably duller and browner than the female.

Crypturellus soui modestus (Cabanis).

Five specimens: Guapiles, Cuabre, Carrillo, Buenos Aires, and Tucur- riqui, Costa Rica.

Costa Rican specimens, sex for sex, are darker, less rufescent below than either meserythrus or mustelinus. Males are browner, less rufescent than females, but the average sexual difference, so it appears, is less than in the allied races. Peters (1931a, 296) calls attention to the fact that the color phases in this species are not definitely correlated with age or sex.

Crypturellus soui caucae (Chapman).

Twelve specimens: Jaraquiel, La Palmita, and El Tambor, Colombia.

These agree in all essential respects with a pair of birds in the American Museum collection from Las Lomitas and Rio Frio, Colombia. Four specimens from Jaraquiel closely resemble the Santa Marta race ( mus- telinus) in the color of the underparts, but they are much darker above, less rufescent, and have a grayish cast. The specimens from La Palmita and El Tambor show the characters of caucce as remarked by Griscom in his review (1932, 308). Coming as they do from the Magdalena Valley, they may be taken as representing that form correctly. The pileum is darker in these birds than in those from Jaraquiel, which latter are prob- ably to be considered intergrades between caucce and mustelinus.

1942 Todd: Tinamous in the Carnegie Museum 13

Crypturellus soui harterti (Brabourne and Chubb).

Six specimens: Soatata, Potedo, Heights of Caldas, and Yumbo, Colombia.

Griscom (1932, 307) has shown that this is the race of soui which ranges from western Ecuador and western Colombia into eastern Panama. The present series conform with this allocation. The single male specimen is much darker than the same sex of either caucce or mustelinus, as Griscom says, and the females are also dark-colored by comparison and have a decided dusky or grayish wash on the neck.

Crypturellus soui caquetae (Chapman).

Two specimens in the collection of the American Museum have been examined in this connection (Florencia and La Murelia, Caqueta, Colom- bia). The female is similar to the same sex of mustelinus , but it is very dark above (dark chestnut brown), with the pileum dusky slate-color. This race is evidently perfectly distinct.

Crypturellus soui nigriceps (Chapman).

One specimen: Sao Paulo de Olivenga, Brazil.

Reference of this specimen to nigriceps involves an extension of the heretofore known range of the race to western Brazil, south of the Amazon. On the north bank it is replaced by typical soui.

Crypturellus soui mustelinus (Bangs).

Fourteen specimens: Don Amo, Don Diego, Cincinnati, Minca, and Palmar, Colombia; El Hacha, El Trompillo, Sierra de Carabobo, and Guamito, Venezuela.

A well-marked race, whose differential characters I have already dis- cussed at some length (1922, 165). The above listed specimens from Venezuela are clearly referable to this race, the range of which extends eastward to that country, to and beyond the Andes of Merida to the region south of the coast range. Moreover, much to my surprise, I find that our two females from Palmar, east of the Eastern Andes in Colombia, are mustelinus and not caquetce. Hence caquetce does not occupy all of Colom- bia east of the Andes, as supposed by Chapman. It is odd, however, to find mustelinus on both sides of the Venezuelan Andes.

No. 36,271, El Hacha, Venezuela, December 1, is a downy chick. Its

14

Annals of the Carnegie Museum

vol. XXIX

head pattern is characteristic: the front (broad) and sides of the crown are buff, enclosing a median rufous crown patch, produced down the neck, and itself enclosing a longitudinal area of buff; the back is rich rufous, freckled anteriorly with buffy feather- tips; the throat is white; and the rest of the underparts are buffy and the crissum rufescent.

Crypturellus soui soui (Hermann).

Nineteen specimens: El Peru Mine, Venezuela; Cayenne, Mana, and Pied Saut, French Guiana; Obidos, Tonantins, Manacapuru, and Rio Manacapuru, Brazil.

I judge that the brighter-colored birds are females and the duller- colored ones males, but the sex marks on the labels do not all correspond. This race is richly colored by comparison, as indicated by Griscom (1932, 308). The females are chestnut or bay above, the pileum is ashy or sooty blackish, the throat is decidedly buffy, and the breast and sides are rich rusty buffy (between Mars yellow and tawny of Ridgway’s “Color Stand- ards”), without any trace of dusky or ashy color. No. 34,139, an im- mature bird from El Peru Mine in Venezuela, is referred here on geo- graphical grounds. Manacapuru and Obidos specimens are a little brighter above than those from French Guiana, but they are not sufficiently dif- ferent in my opinion to deserve a name. Miranda-Ribeiro (1938, 769) refers a single specimen from Manacapuru to his new race lyardi. All his other specimens of that form come from south of the Amazon. Also, a single bird from Tonantins, on the north bank of the upper Amazon, obviously belongs here, although another race ( nigriceps ) lives across the river. The range of typical soui , therefore, extends from Guiana and Brazil, north of the Amazon, westward to the frontiers of Colombia and Ecuador.

Crypturellus soui andrei (Brabourne and Chubb).

Nine specimens: Santa Lucia, San Rafael, and Mirasol, Venezuela.

These birds, coming from the region north of the coast range in Vene- zuela (in the states of Miranda and Sucre), are not mustelinus, than which they are darker-colored above and below, with a decided dusky band across the neck, and a restricted white throat patch. On the other hand, they are certainly not referable to soui. Upon comparison they prove to be substantially identical with a pair of birds from Trinidad in the collection of the American Museum of Natural History. Incidentally,

1942

Todd: Tinamous in the Carnegie Museum

15

these specimens do not agree precisely with the original description of andrei, but they must nevertheless belong to that very distinct race, which is thus shown to reach the north coast of Venezuela.

Crypturellus soui hoffmannsi (Brabourne and Chubb).

Ten specimens: Benevides, Santarem, Villa Braga, and Miritituba, Brazil.

The name Crypturellus soui decolor Griscom and Greenway (1937, 417) was based on the above series and one other specimen (the type) from the same region.

A few months later birds from this region were described by Miranda- Ribeiro (1938, 767) under the name Crypturornis soui lyardi ( layardi ?). It is a pale, washed-out race, as the name implies. The pileum is brown, rather than ashy, with the sides of the head similar but paler. In the most richly colored female (No. 75,816) the back is argus brown, and the under- parts are deep ochraceous buff. A small young bird (No. 72,079, April 16) is spotted with black above and below, and has some buff on the wings; the pileum still retains some of the rufous feathers of the natal stage.

These have been compared with five specimens (wretched skins!) from the Rio Madeira which Hellmayr calls hoffmannsi , and in my judgment they are the same form. Variation in exact coloration is in evidence, but since all the Rio Madeira birds can be closely matched by others of the lower Amazon series, I can see no point in keeping them apart. Con- sequently decolor falls as a synonym of hoffmannsi , the recognized range of which must be accordingly extended.

Crypturellus soui inconspicuus Carriker.

Nine specimens: Buena Vista (Rio Surutu and Rio Yapacani) and Cerro Hosane, Bolivia; Hyutanahan (Rio Purus), Brazil.

According to Carriker (1935, 315) these represent his new race in- conspicuus, described from the Rio Beni in Bolivia, but since the Rio Purus specimen seems to be the same, comparison with hoffmannsi of the Rio Madeira is indicated. I have recently made this comparison, and I have found that the two series are racially distinct. C. s. inconspicuus is darker, more rufescent, less buffy below, while its upperparts are also darker-colored on an average. I continue to refer the specimen from the Rio Purus here, although I am at a loss to explain the range thereby in- volved. The form is quite distinct racially from nigriceps.

16

Annals of the Carnegie Museum

vol. XXIX

Crypturellus undulatus adspersus (Temminck).

Seven specimens: Santarem, Villa Braga, Goyana Island (Rio Tapajoz), Miritituba, and Apacy, Brazil.

Para (the state) is the type-locality, but Hellmayr (1929, 477) says that Rio Tapajoz birds are typical. The back, neck, and even the pileum are more or less rufescent, with the rump, upper tail-coverts, and wing- coverts grayish olive, in decided contrast. All are vermiculated on these latter areas, and some have decided bars — this character tending thus toward confusus. The breast and underparts generally are plain grayish buffy, becoming decidedly buffy on the tibiae, lower abdomen, and crissum.

Crypturellus undulatus yapura (Spix).

Six specimens: Arima, Manacapuru, and Caviana, Brazil.

These specimens agree with each other, and differ from a series from the Rio Tapajoz (< adspersus ) in being darker, more decidedly grayish, less buffy, below, and also in being darker brown above, with the pileum dusky plumbeous, in decided contrast; the vermiculations on the back are finer.

This form was described from the Rio Yapura, one of the northern affluents of the Amazon coming in not so far above Manacapuru but that specimens from that point would be referred to it on geographical grounds. Its range appears to include also the region south of the Amazon, since the specimen from Arima on the Rio Purus is perfectly typical. Farther up that river, at Hyutanahan, however, it is replaced by confusus , but according to Hellmayr (1929, 477) the Rio Jurua race is yapura.

In its characters this race is farthest of all from undulatus , with which it is connected through adspersus and confusus; otherwise it could stand as a species.

C. balstoni (Bartlett) is a synonym, according to Hellmayr. Examina- tion of the type in the British Museum shows that the colored plate in the “Catalogue” is not a good representation. There is too much cross-barring on the underparts; whereas the breast and abdomen are nearly plain medially.

Crypturellus undulatus confusus (Brabourne and Chubb).

Twelve specimens; Hyutanahan, Brazil.

This form is a connectant between undulatus and adspersus; its char- acters, as shown by the present series, vary now toward the one, and now

1942

Todd: Tinamous in the Carnegie Museum

17

toward the other. Heretofore it has been known only from the type, which was taken at Humaytha, on the left bank of the Rio Madeira. (Peters’ ascription of it to the Rio Purus is based on specimens from the above series examined by him.) Some examples are decidedly grayish- tinged below (like yapura ), others are more buffy (like adspersus) . The barring above varies considerably; some specimens are almost as heavily barred as undulatus , while others are more nearly like adspersus in this respect. The pileum tends to be brownish plumbeous; the upperparts and neck behind are strongly rufescent. Young birds (Nos. 87,043-4) are more or less barred and spotted below with blackish.

The range of this race embraces the region between the rivers Madeira and Purus, south of (about) 5° S. The Amazonian race, yapura , ap- parently occupies both banks of the Amazon between the mouths of these two tributaries, and ranges up the Rio Purus as far as Arima.

Crypturellus undulatus undulatus (Temminck).

Sixteen specimens: Buena Vista (including Rio Surutu), Bolivia.

These fit the description and plate of Crypturus scolopax in the British Museum “Catalogue,” as of course they should do, since they come from the type-locality. But according to Hellmayr (1929, 477) this name is a synonym of the earlier undulatus , based on the bird of Paraguay. It is recognized, however, by Miranda-Ribeiro (1938, 748).

The series shows much individual variation, affecting the color-tone of the underparts, the extent and intensity of the barring on the upper- parts, neck, breast, and flanks, and the coloration of the pileum, which in some specimens is distinctly barred, in others plain slaty gray. A young bird (94,670) shows a little dark spotting on the breast.

Bolivian specimens in the British Museum are much more heavily barred above than those from Chapada, Matto Grosso, Brazil, but ex- amples in the American Museum of Natural History from these respective regions are practically indistinguishable.

No. 80,932, February 28, is a chick, colored as follows: above mummy brown, with slight buffy feather- tipping; broad front and sides of head deep buffy, with a dark stripe through the eye; this stripe behind the ear centered with a median stripe of raw umber; crown raw umber, this color reaching the nape, bordered on either side by a narrow black edge, and divided by a median buffy stripe; throat white; rest of underparts dull buffy.

18

Annals of the Carnegie Museum

vol. XXIX

Crypturellus boucardi boucardi (Sclater) (?)

Two specimens: San Pedro Sula, Honduras; Cockscomb Mountains, British Honduras.

The British Honduras specimen should belong to the recently described race blancaneauxi Griscom (1935, 543) but it does not fit the description, since it is much darker and grayer below than the Honduras skin. I have compared it directly with a specimen from Secanquim, Guatemala (Am. Mus. No. 393,410), than which it is slightly darker above, with pro- nounced black bars on the wings, tail, and lower back. I have examined also the type and one other specimen of blancaneauxi in the British Mu- seum. Both seem to be females. They are more rufescent above and buffier below than Guatemala skins. In the type the pileum is mostly brown; in the other it is slaty. Judging from these two skins alone, I would call blancaneauxi a valid race, but I should like to see more material before finally accepting it.

Crypturellus variegatus variegatus (Gmelin).

Twenty-two specimens: El Llagual, Venezuela; Pied Saut, French Guiana; Cayari Island, Upper Araucaua, Obidos, Tonantins, and Mana- capuru, Brazil.

This species and its component races I have already discussed at some length in a former paper (1937, 175-178). To my remarks there I have only to add that a series from British Guiana in the collection of the British Museum shows much variation in the coloration of the underparts, the barring on the flanks and back, etc. I have examined also the type of Tinamus bimaculatus Gray. It is a young bird in spotted dress, without indication of locality, and absolutely indeterminable as to subspecies, so that the name may be disregarded.

Crypturellus variegatus transamazonicus Todd.

Seventeen1 specimens: Santarem, Colonia do Mojuy, Villa Braga, Hyutanahan, Arima, and Sao Paulo de Olivenga, Brazil.

To the remarks in my paper above cited I have only to add that this is obviously the same form that was described a few months later by Mi- randa- Ribeiro under the name Orthocrypturus variegatus super ciliosus

1 In my former paper I listed nineteen specimens of this form, but I inadver- tently included two specimens of C. bartletti in the total.

1942

Todd: Tinamous in the Carnegie Museum

19

(1938, 741). This author designates no type-specimen, and the example from Manacapuru which he lists is undoubtedly referable to true variegatus , as is our own from the same locality. The range of this pale-colored race would seem to reach the Rio Gy-Parana in western Matto Grosso.

Crypturellus variegatus salvini (Salvadori).

The type-series in the British Museum conforms to the description with respect to the barring on the upperparts. They are in fact by comparison very blackish above and richly colored below (the type especially) ; how- ever, the barring on the flanks is not very distinct. As I have already said, salvini is merely a darker race of variegatus , and it is not a very strongly marked one at that.

Crypturellus brevirostris (von Pelzeln).

Four specimens: Tamanoir, French Guiana; Hyutanahan, Brazil.

Since my previous account of this species appeared ( lx .), a second speci- men from Tamanoir, French Guiana, which had been inadvertently dis- posed of by exchange, has been located and returned for study. H. B. Conover, whose specimen it is now, writes that Peters’ ascription of “Cayenne” in the range of this species was probably based on this speci- men, although it did not come from that exact locality. It agrees per- fectly with the other bird from the same place in having white posterior underparts; this fact increases the probability that two races are repre- sented in the series, as I have already suggested. It is a male, and measures: wing, 129; tail, 39; bill, 19; tarsus, 41. The flanks and tibiae show some barring, although not so much as in the Rio Purus skin.

Oliverio Pinto (1938, 500) records a specimen of this rare species from the Rio Vaupes, on the Colombian frontier of Brazil — a record which considerably extends the known range.

Crypturellus bartletti (Sclater and Salvin).

Eleven specimens: Hyutanahan, Nova Olinda, and Arima, Brazil. Compare my previous remarks on this species (1937, 178). To the list of specimens there given two more must be added. Compare also the re- marks by Miranda-Ribeiro (1938, 750).

20

Annals of the Carnegie Museum

vol. XXIX

Crypturellus cinnamomeus mexicanus (Salvadori).

Two specimens: Rio Sabinas, near Gomez Farias, Tamaulipas; and Valles, San Luis Potosi.

These examples are not alike, but I judge the differences to be sexual and not geographical. While at the British Museum I made a study of the races of this species, and reached precisely the same conclusions as Griscom announced (1935, 541-3).

Crypturellus cinnamomeus vicinior Conover.

One specimen : Monte Redondo, Honduras.

Crypturellus cinnamomeus praepes (Bangs and Peters).

One specimen : Bebedero, Costa Rica.

Crypturellus idoneus (Todd).

Two specimens: Bonda, Colombia.

For a full description and critical remarks on this form compare my former account (1922, 166). Although it was originally described as a full species, Peters (1931b, 21) reduces it to a race of cinnamomeus. Conover (1933, 113) dissents, but Zimmer (1938, 50) insists that this view is cor- rect. My material representing cinnamomeus is scanty I admit, but I should certainly keep idoneus as a species distinct both from the former on the one hand and from erythropus on the other. In my opinion idoneus does not fit in well with either of these and ought to stand alone. It occupies a semi-insular range in the Santa Marta region of Colombia.

Crypturellus noctivagus (Wied).

This form is unrepresented in our collection, but I have examined three authentic specimens from the states of Sao Paulo and Bahia, Brazil, lent by the American Museum of Natural History. The male differs from the same sex of C. erythropus in its darker general coloration and the more decided barring of the lower back, upper tail-coverts, wings, flanks, and posterior underparts. Adult males of erythropus are almost devoid of barring, nearly “solid” brown in fact. The female is still more different from the other species, since it has the superciliaries and cheeks white, the throat refuscent, barred with dusky, the breast dull rufescent ochraceous (with no grayish), and the rest of the underparts rich buff, the sides barred

1942

Todd: Tinamous in the Carnegie Museum

21

with black. The wings and tail are conspicuously barred with buff and the back is black, barred with rufous; the feet (in the skin) are yellowish, instead of dark-colored, as in the male. I consider these differences of specific value.

Crypturellus erythropus erythropus (von Pelzeln).

Fourteen specimens: San Esteban, Venezuela; Obidos, Brazil.

Zimmer (1938, 50) reduces this form to a subspecies of C. noctivagus without apparent hesitation, but I think that he goes too far. According to Hellmayr (quoted by him), C. dissimilis is the same as C. erythropus. Our San Esteban specimen was examined by Hellmayr some years ago and was pronounced dissimilis. On geographical grounds it should be referable to C. spencei Chubb, but I am unable to distinguish it from the series from Obidos, on the Amazon River. Zimmer recognizes spencei , but not on the basis of topotypical material, since all his specimens but one came from points much farther east in Venezuela. Indeed, he says that females of the two forms (< erythropus and spencei) are not certainly distinguishable. In describing spencei , Chubb compares it with cinnamo- meus of Middle America, instead of with the South American erythropus. I have examined the type and one other specimen of spencei in the British Museum. They are in my judgment the same as erythropus , despite the rather prominent buffy bars on the back and wings. Specimens of the latter in the British Museum very considerably among themselves. Some are almost uniform buffy below, while others have a more or less grayish wash on the breast. Again, some are almost uniform olive brown above; others (females or young birds ?) are decidedly barred with buff and black, especially on the lower back and wings, like variegatus. The head is rufescent all around, and the forehead usually slaty.

Our specimens are all sexed as males. Rufescent throats are the rule, but in several individuals this part is white, either partially or wholly. The depth of the rufescent shade below, the amount of grayish wash on the breast, the dark barring of the flanks, the shade of brown on the back, and the black barring on the lower back and upper tail-coverts all vary considerably.

Crypturellus erythropus garleppi (von Berlepsch).

Two specimens: Buena Vista (Rio Surutu), Bolivia.

These fit the original description fairly well, and coming as they do from the type-locality (virtually), they could scarcely be anything else. The

22

Annals of the Carnegie Museum

vol. xxix

form proves to be so close to C. erytliropus as to be only subspecifically separable, in my opinion. It differs from erythropus in the general color of the underparts, which averages darker, deeper rufous, and in the more olive brown (rather than cinnamon brown) shade of the upperparts, with the black vermiculation much more distinct.

The type-specimen of C. garleppi affinis Chubb, examined by me in the British Museum, has been marked by Hellmayr as the female of garleppi. Chubb was evidently misled by von Berlepsch’s statement that garleppi was close to atrocapillus. The latter I have seen (in the British Museum), and I consider it a well-marked and perfectly distinct species.

Crypturellus strigulosus strigulosus (Temminck).

Ten specimens: Santarem, Villa Braga, and Hyutanahan, Brazil.

I cannot accept Zimmer’s arrangement making strigulosus a.n& erythropus conspecific (1938, 50). The differences are far too great, and moreover, as I have just shown, erythropus is represented in Bolivia by an allied race, with strigulosus coming in between. One female of the latter (No. 75,306) is decidedly buffy below, while another (No. 78,199) is the whitest of the series. A young bird (No. 78,200, September 16) is also whitish below. Two males from the Rio Purus are practically indis- tinguishable from the Rio Tapajoz specimens. This fact suggests that the unique type of hellmayri from the Rio Madeira may be only an in- dividual variant. In our series the shade of brown in the upperparts varies considerably.

Crypturellus parvirostris (Wagler).

Thirteen specimens: Buena Vista (also Rio Surutu and Rio Yapacani), Santa Cruz de la Sierra, Guanacos, and Rio Quiser, Bolivia; Santarem, Brazil.

Most recent writers on this species admit no racial variation, but cer- tain characters shown by our series are at least suggestive of such. Two birds (a pair) from Santarem are darker, more grayish below, sex for sex, than the “general run” of Bolivian specimens. They may possibly repre- sent the lately described C. p. fuscus of Miranda-Ribeiro (1938, 775), based on two females from the island of Marajo. Compared with two specimens from Bahia (fixed as the type-locality by Hellmayr [1929, 478]) in the collection of the American Museum of Natural History, Bolivian

1942

Todd: Tinamous in the Carnegie Museum

23

skins are paler above; they have slaty (instead of brownish) pileum and sides of the head, and grayer, less brownish sides and breast. If separable, they may be entitled to the name cervinus of Bonaparte from Chiquitos, Bolivia — unless, indeed, as is possible, this name applies to C. tataupa. (The description is brief and unsatisfactory.) However, Hellmayr (1907, 410) says that in young birds of parvirostris the pileum is mainly pale brown, as it is indeed in our specimen from Santa Cruz de la Sierra, so that the Bahia skins above mentioned may be immature.

Three skins from Chapada, Matto Grosso, and one from Bolivia, in the British Museum, are pale as compared with other skins from Brazil.

No. 79,109 (January 12) is emerging from the natal down, but it still shows traces of the characteristic head pattern of this early stage. No. 43,763 is immature, with the breast obsoletely squamate, and with black subterminal spots, partially concealed, on the feathers of the back and wing-coverts. Females are uniformly darker and more richly colored than males.

Miranda-Ribeiro ( l.c .) has described also a third race of this species {super ciliaris) from the “Planalto Parecis” of Brazil, from which he had a single specimen. In view of the observable differences depending on age, sex, and individual, I feel that no geographical races of this species should be recognized unless on the basis of a very much larger series of specimens than are at present available.

Crypturellus tataupa tataupa (Temminck).

Eight specimens: Puerto Suarez, Palmarito, Rio Grande, and Samai- pata, Bolivia.

These require comparison with specimens from Brazil and Paraguay, although Hellmayr (1929, 478) discounts the chance of any racial varia- tions in this region.

Rhynchotus rufescens rufescens (Temminck).

Nine specimens: Buena Vista, Bolivia.

The buffy barring on the underparts is subject to considerable variation in these specimens. They have not been compared as yet with birds from Sao Paulo, Brazil, which Mrs. Naumburg accepts as the type-locality (1930, 60).

24

Annals of the Carnegie Museum

vol. XXIX

Rhynchotus maculicollis Gray.

Three specimens: Samaipata, Cerro San Benito, and Incachaca, Bolivia.

I see no reason for reducing this form to subspecific rank, even if the neck markings in one specimen are somewhat less prominent. Its range closely approximates that of R. rufescens, and drops down to 1200 meters at Samaipata.

Nothoprocta ornata ornata (Gray).

Eight specimens: La Paz, Colomi, Guaqui, and Incachaca, Bolivia.

In this species the breast is decidedly gray, but this feature is not well shown in the plate (17) in the “Catalogue of the Birds in the British Mu- seum.” The spotting and barring above vary; they are less prominent in the Incachaca specimen, which may possibly belong to a different race.

Nothoprocta cinerascens (Burmeister).

Five specimens: Miraflores and Machomuerto, Argentina.

Nothoprocta curvirostris Sclater and Salvin.

Two specimens: Mount Pichincha, Ecuador.

A chick, collected January 28, is assumed to belong to this species. Above it is mottled and striped with black, brown, and white; the super- ciliaries, malar region, and throat are white; the underparts are dull gray, mottled with brown.

Nothoprocta pentlandi pentlandi (Gray).

Thirteen specimens: Cochabamba, Tiraque, Comarapa, and Pocona, Bolivia.

At first glance these fall into two series, according to size and colora- tion. In the larger birds (wing 130-155 mm.) the forehead, the sides of the head (except a postocular streak), and the throat and breast are decidedly grayish (the latter with some white spots) ; the prevailing tone of the upper- parts is grayish brown; the white streaking is present but not prominent, especially on the upper back; the sides and flanks are shaded and more or less barred with buffy. In the smaller birds (wing 127-139 mm.) the fore- head, the sides of the head, and the throat and breast are buffy with

1942 Todd: Tinamous in the Carnegie Museum 25

brownish streaking (the breast more grayish buffy, with white spots) ; the prevailing tone of the upperparts is dull Brussels brown to rich tawny olive (especially in evidence on the tertiaries) ; the black bars and white streaks are conspicuous, even on the upper back; the sides and flanks are spotted and barred with dusky brown. In one specimen (No. 81,333) the breast is similarly spotted with brownish black. The two lots look like two species, but this could hardly be true, because there is one specimen (No. 120,132) which is clearly intermediate, since it has rich tawny olive tertiaries and brown-spotted sides combined with grayish forehead, sides of the head, and breast. Moreover, the smaller birds (with one exception) were shot at the same time as the larger ones, as I judge from the con- tinuity of the collector’s numbers. These smaller birds fit the description of Nothoprocta moebiusi von Berlepsch, but I agree with Salvadori that this must be the young of pentlandi.

Three birds shot in September and February are obviously more rufes- cent buffy below than those taken in June. This character I believe is one that has been used for discriminating a Peruvian race of this form, which race needs re-investigation.

Females appear to have the throat spotted with grayish, but some doubt attaches to the sexing of the specimens.

Nothoprocta perdicaria perdicaria (Kittlitz).

One specimen: Valparaiso, Chile.

Nothura maculosa agassizi Bangs.

Five specimens: La Paz (market), Guaqui, and Desaguadero, Bolivia.

Mr. Griscom, to whom four of the above specimens were sent for com- parison with the type, writes as follows: “Our specimens of this race of maculosa consist of the type and one other specimen. Your specimen marked, ‘dV from Desaguadero, matches our two specimens of agassizi exactly. Your other three specimens, however, take on the hue of dif- ficulty. They come from a different locality, which may or may not be faunally isolated or distinct from the other.” He goes on to suggest that, being females, they may be different on that account, or else that they are nearer boliviana. Comparison with this latter form suggests that they do actually tend somewhat towards it. Peters reduces both agassizi and boliviana to races of N. maculosa, as also does Laubmann (1934, 280). Pending further study I accept this arrangement provisionally, although

26

Annals of the Carnegie Museum

vol. XXIX

somewhat doubtful of its propriety. Compare also, in this connection, Wetmore’s remarks on the affinities of these forms (1926, 37). At any rate, agassizi and boliviana appear to be very distinct from each other, in spite of variation in both forms. I am not prepared to discuss their re- lationships to the recently described Nothura maculosa oruro Bond and de Schauensee (1941).

Nothura maculosa boliviana Salvadori.

Five specimens: Cochabamba, Chocaya, Tiraque, and Totora, Bolivia.

These specimens were sent to Hellmayr at the British Museum for comparison with the type of boliviana. He reported that they were the same.2 The specimen from Tiraque is a very good match for the type, particularly in the deep coloration below. The Cochabamba birds as well as the original series of boliviana are indeed easily distinguished from agassizi , of Lake Titicaca, which is blacker above and much more coarsely marked with black on the chest and sides. He goes on to say that Bridges’ type-series probably came from the Valley of Cochabamba. The form ranges from Bolivia through the Chaco to Paraguay and Argentina (compare Wetmore, l.c.).

Nothura boraquira (Spix).

Eleven specimens: Santa Cruz de la Sierra and Buena Vista, Bolivia.

These agree with the figure of N. marmorata Gray (in Salvadori, 1895, pi. 18). According to Hellmayr (1906, 705), however, this name is a synonym of Tinamus boraquira Spix — an opinion which he reiterates (1929, 478) after seeing specimens from the same locality from which our own come. Miranda-Ribeiro (1938, 704) renames this form spixi, on the ground that “ boraquira ” is “bad Portuguese” — which reason under present rules is of course no reason at all. Wetmore (1926, 37, note) would refer this species to Nothoprocta on account of the posterior aspect of the tarsal envelope as shown on Spix’s plate. This plate is incorrect in this particular respect, however, although it is otherwise a fair representation.

Two sizes of birds are represented in our series, but not according to sex (as marked). The smaller ones may be younger birds, as I judge from our specimen No. 79,124, which is rather prominently spotted below.

2 In sending these I inadvertently overlooked the fact that Dr. Hellmayr had already compared Bolivian specimens from Mr. H. B. Conover.

1942 Todd: Tinamous in the Carnegie Museum 27

LITERATURE CITED

Berlepsch, Hans Graf von.

1908. On the Birds of Cayenne. Part II. — Novitates Zoologicae, Novem- ber, 1908, 261-324.

Boetticher, Hans von.

1934. Beitrag zu einen phylogenetisch begriindeten, natiirlich System der

Steisshtihner ( Tinami ) auf Grund einiger taxonomisch verwert- barer Charaktere. — Jenaische Zeitschrift fiir Naturwissenschaft, 69, 1934, 169-192.

Bonaparte, Charles Lucien.

1856. Tableaux paralleliques de l’ordre des Gallinaces. — Comptes Rendus A cadernie Sciences Paris , 42, 1856, 874-884. Note sur les Tableaux des Gallinaces. — Ibid., 953-957.

Bond, James, and de Schauensee, Rodolphe Meyer.

1941. Descriptions of New Birds from Bolivia. — Notulae Naturae No. 93, October 14, 1941. pp. 7.

Brabourne, Lord, and Chubb, Charles.

1914. A Key to the Species of the Genus Crypturus, with Descriptions of some new Forms. — Annals and Magazine of Natural History , (8), 14, October, 1914, 319-322.

Carriker, Melbourne A., Jr.

1910. An Annotated List of the Birds of Costa Rica including Cocos Island. — Annals Carnegie Museum, 6, September 7, 1910, 314- 915, 1 map.

1933. Descriptions of new Birds from Peru, with notes on other little- known species. — Proceedings Academy Natural Sciences of Phila- delphia, 85, March 24, 1933, 1-38.

1935. Descriptions of new Birds from Bolivia, with Notes on other little-

known species.- — Proceedings Academy Natural Sciences of Phila- delphia, 87, October 10, 1935, 313-341.

Chapman, Frank M.

1917. The Distribution of Bird-Life in Colombia: A Contribution to a Biological Survey of South America. — Bulletin American Museum of Natural History, 36, 1917, pp. x + 729, 41 pis.

1926. The Distribution of Bird-Life in Ecuador. — Bulletin American Museum of Natural History, 55, 1926, pp. xiv + 784, 30 pis.

Chubb, Charles.

1916. The Birds of British Guiana. Vol. 1, 1916, pp. liii + 528, 10 pis., map.

Conover, H. B.

1933. The Races of the Tinamou Crypturellus cinnamomeus. — Proceed- ings Biological Society of Washington, 46, June 30, 1933, 116-117.

1937. A new race of Tinamus major from Brazil. — Proceedings Biological Society of Washington, 50, October 28, 1937, 191-192.

28

Annals of the Carnegie Museum

vol. XXIX

Griscom, Ludlow.

1929. A Collection of Birds from Cana, Darien. — Bulletin Museum Com- parative Zoology , 69, April, 1929, 149-190.

1932. The Ornithology of the Caribbean Coast of Extreme Eastern Panama. — Bulletin Museum Comparative Zoology, 72, January, 1932,303-372.

1935. Critical Notes on Central American Birds in the British Museum. — Ibis, July, 1935, 541-554.

Griscom, Ludlow, and Greenway, James C.

1937. Critical Notes on New Neotropical Birds. — Bulletin Museum Com-

parative Zoology, 81, May, 1937, 417-437.

Hellmayr, Carl E.

1906. Revision der Spix’schen Typen brasilianischer Vogel. — Abhand-

lungen K. Bayerischen Akademie Wissenschaften, II Kl., 22, 1906, 563-726, 2 pis.

1907. On a Collection of Birds made by Mr. W. Hoffmanns on the Rio

Madeira, Brazil. — Novitates Zoologicae, 14, November, 1907, 343-412, pi. 3.

1929. A Contribution to the Ornithology of Northeastern Brazil. — Field

Museum Zoological Series, 12, March 4, 1929, 235-501, map.

Laubmann, A.

1934. Weitere Beitrage zur Avifauna Argentiniens. — Verhandlungen Ornithologischen Gesellschaft in Bayern, 20, 1934, 249-336, map.

Miranda-Ribeiro, Alipio de

1938. Notas Ornithologicas (xii). Tinamidae. — Revista do Museu Paulista,

23, March 18, 1938, 667-788, pis. 1-18.

Naumburg, Elsie M. B.

1930. The Birds of Matto Grosso, Brazil. — Bulletin American Museum

of Natural History, 60, 1930, pp. vii + 432, 17 pis., map.

Oliveira Pinto, Oliverio M. de

1938. Catalogo das Aves do Brasil e lista dos exemplares que as repre- sentam no Museu Paulista. — Revista do Museu Paulista, 22, “1937,” 1938, pp. xviii -f- 566.

Peters, J. L.

1931a. Additional Notes on the Birds of the Almirante Bay Region of Panama. — Bulletin Museum of Comparative Zoology, 71, February, 1931, 293-345.

1931b. Check-List of Birds of the World. Cambridge, June, 1931, pp. xviii + 345.

Salvadori, Tommaso.

1895. Catalogue of the Birds in the British Museum. Vol. 27, 1895, includes the Chenomorphae, Crypturi, and Ratitae, pp. xv + 636, 19 pis.

1942

Todd: Tinamous in the Carnegie Museum

29

Salvin, Osbert, and Godman, F. DuCane.

1904. Biologia Centrali-Americana. Vol. 3, 1897-1904, pp. iv -f 510. (Tinamidae, pp. 448-459, April, 1904.)

Snethlage, Emilia.

1914. Catalogo das Aves Amazonicas contendo todas as especies descriptas e mencionadas ate 1913. — Boletim do Museu Goeldi, 8, “1911-12,” 1914, 1-530, 6 pis., map.

Todd, W. E. Clyde, and Carriker, Melbourne A.

1922. The Birds of the Santa Marta Region of Colombia: A Study in Altitudinal Distribution. — Annals Carnegie Museum , 14, 1907, 3-611, 9 pis.

Todd, W. E. Clyde.

1937. Critical Remarks on Crypturellus variegatus and its Allies. —

Proceedings Biological Society of Washington, 50, October 28, 1937, 175-178.

1938. Notes on Crypturellus cinereus (Gmelin). — Proceedings Biological

Society of Washington, 51, May 19, 1938, 123-126.

Wetmore, Alexander.

1926. Observations on the Birds of Argentina, Paraguay, Uruguay, and Chile. — Bulletin U. S. National Museum, No. 133, 1926, pp. iv + 448, 20 pis.

Zimmer, John T.

1928. A new form of Crypturellus noctivagus. — Proceedings Biological Society of Washington, 51, March 18, 1938, 47-51.

ART. II. OBSERVATIONS ON THE LIFE HISTORY OF A NEYV^ CHALCIDOID WASP, AN INTERNAL PARASITE OF ANT-LION LARVAE

By George E. Wallace

(One plate)

During the summer of 1940, two ant-lion cocoons, Myrmeleon im- maculatus De Geer, collected at Presque Isle, Erie, Pennsylvania, were found to be parasitized by a chalcidoid wasp belonging to the genus Stomatoceras (family Chalcididae) . One adult wasp emerged from each cocoon. Both individuals were females. Further collections of host material, both larvae and cocoons, were made during July 1941. From this host material forty-five parasites, comprising both males and fe- males, were reared. Several more of the parasites were taken in the field, and the process of oviposition was observed both in the field and in captivity. Dissection revealed pupal cases of the wasp within the remains of the ant-lion larvae.

It was at first thought that this Stomatoceras material from Presque Isle might prove to be a new species. However, in the writer’s opinion, the material should be considered, for the present at least, as a variety of Stomatoceras rubra Ashmead,1 and as such it is here described.

The author is indebted to Mr. C. F. W. Muesebeck and Mr. A. B. Gahan, at the United States National Museum, and Mr. E. T. Cresson,

Jr., of the Academy of Natural Sciences of Philadelphia, for permission to study material in the collections of these two institutions.

Stomatoceras rubra Ashmead eriensis, new variety

Female: length 4.7 mm. Head barely as wide as the thorax, antero- posteriorly thin and somewhat depressed between the lateral ocelli. The distance between each lateral ocellus and the eye margin equal to the diameter of the ocellus; the posterior edge of the median ocellus not quite reaching the vertex. Eyes prominent, with a very few short widely scattered hairs. Face long, triangular in appearance. Clypeus short, lower margin gently concave. Labrum visible. Left mandible bidentate; (the right mandible concealed in the holotype, but in other specimens tridentate). Maxillary palpi 4-jointed; labial palpi 3-jointed. Cheeks

1 Ashmead, W. H., Trans. Amer. Ent. Soc., vol. 21, p. 332, 1894.

31

Issued March 13, 1942.

32

Annals of the Carnegie Museum

vol. XXIX

rimmed posteriorly by a sharp thin carina. Head, excepting antennal furrow, hairy and closely pitted with coarse punctures, each puncture enclosing a hair. Antennal furrow delicately cross-striated. Antennae 11-jointed, situated close together just above the clypeus; a carina present between their scrobes. Scape equal to about one-half the length of the flagellum, and extending to the upper (posterior) margin of the median ocellus. First funicle joint the longest. Following joints of flagellum gradually diminishing in length with the exception of the apical joint which is nearly equal to the first funicle joint. The ring joint a little longer than broad — almost quadrate. Pedicel equal to the third funicle joint. Pubescence of antennae appearing slightly longer and denser from tip of pedicel to and including the first funicle joint.

Thorax slightly longer than wide. Prothorax transverse quadrate,

Left figure, outline of antenna of female, x 15; right figure, outline of antenna of male, x 16.

though somewhat declivous anteriorly; the sides but not the notum bordered anteriorly by a carina. Parapsidal furrows complete. Axillae not meeting at base of scutellum. Scutellum bounded laterally by a carina which posteriorly forms the two small lobes or teeth typical of the genus. Below the carina bounding the scutellum is an area bordered below by a second carina; this area is traversed by rugae connecting the two carinae. Prothorax, mesonotum, and upper portion of the scutellum thickly pitted with coarse punctures. Metepisternum punctate and hairy. Propodeum short, declivous, aerolated by numerous rugae; hairy at the lateral angles. Propodeal spiracles thin, slit-like, directed toward scutellum. All coxae with short silvery hairs. Lower margin of hind femora not markedly convex; denticular ridge straight (slightly notched in some specimens). Hind tibiae two-spurred, the outer spur minute — much the shorter of the two. Forewings overlapping abdomen, fuscous with an embrowned area below the point where the submarginal vein joins the marginal. A hyaline area extends from the embrowned

1942

Wallace: A New Chalcidoid Wasp

33

portion to a short distance beyond the postmarginal. Both embrowned and hyaline areas extend a little less than halfway to the posterior edge of the wing. Pubesence of wings brown ; no silvery white pubescence in the hyaline area.

Abdomen globose. First segment the longest, dorsally extending slightly more than half the length of the abdomen. The second, third, fourth, and fifth segments very short, very finely roughened; the second, fourth, and fifth subequal, each dorsally equal to about one-sixth of the first segment; the third segment small, forming a barely visible trans- verse strip. Posterior margins of the third, fourth, and fifth tergites concave. The sixth tergite forms a vertical plate which is furnished with conspicuous punctures. Ovipositor not visible ; the sheaths visible , but not projecting beyond the sixth tergite.

Fig. 2. Stomatoceras rubra eriensis

Left figure, outline of head of female, front view, x 13; right figure, outline of head of male, x 13.

General body color dark red except for the following structures which are black: antennae, excepting ring and first funicle joint; anterior mar- gin of mesonotum, posterior third of mesonotum, extreme antero-lateral angles of parapsides together with the adjoining portion of prothorax and tegulae, axillae at anterior angles; metathorax except median spot of red; propodeum, mesosternum, metasternum, front and middle tibiae except apices; inner faces and proximal tips of front femora; mid-trochanters and mid-femora, mid- and hind tarsi except apices; hind tibiae except apices; denticulate portion of hind femora, and median portion of first abdominal tergite.

Ring joint and first funicle joint fusco-rufous; in life this portion of antennae appears pale silvery. The middle femora somewhat suffused with fusco-rufous on outer sides. The apices of front and middle tibiae, all of fore tarsi, apices of mid- and hind tarsal joints fusco-rufous. A faint touch of blackish on occiput, vertex, parapsidal furrows, dorsal surface of scutellum, and lower margin of sixth tergite.

34

Annals of the Carnegie Museum

vol. XXIX

Male: length 3.7 mm. Head not as long as in female. Eyes hairy. Ring joint and first funicle joint of antennae apparently fused together, forming a 10-jointed antenna. Antennal scape equal to about one-third the flagellum, barely reaching the median ocellus. Pedicel smaller than than any of the funicle joints.

Thorax similar to that of female. Lower margin of hind femora con- vex; denticulate ridge with an incision so that proximal six to nine den- ticles appear to be on a prominence. Forewing similar to that of female.

Abdomen small, globose, length 1.4 mm. First abdominal tergite the longest, equal to half the abdomen; the second, third, fourth, fifth, and sixth tergites nearly equal in length, the third being smaller than the others. Hind margins of the second and third tergites appear concave when seen from above. All tergites very finely roughened; the fourth, fifth, and sixth with some coarse punctures.

Color black; inner portions of tarsal joints, particularly at apices, tinged with fusco-rufous.

Size and color: the females of the paratype series show considerable variation in both size and color. The largest specimen is 4.9 mm. in length, the smallest 3.6 mm. The average size of all the female type specimens (including the holotype) is 4.5 mm. ; however the majority of the specimens exceed this figure. The lightest colored specimen of the female paratypes has only the tibiae, anterior and posterior margins of the mesonotum, anterior edges of parapsides, mesosternum, metasternum, and basal portion of the antennal scapes, black. The darkest female specimen has the head and thorax almost entirely black, and the abdo- men heavily suffused with black.

The males of the type series range in size from 3.6 mm. to 4.4 mm. The average size of the males is 4.1 mm., but the majority of the speci- mens exceed this figure.

Varietal differences summarized

The chief differences between S. rubra Ashmead2 and the variety eriensis , as shown by female specimens, are summarized as follows:

1. The general color of eriensis is much darker red than rubra. As mentioned previously, the head and thorax are almost uniformly black in the darker specimens of the eriensis paratype material. In both light

2 Of the four types (all females) designated by Ashmead, only the two deposited at the U. S. National Museum have been examined by the writer.

1942

Wallace: A New Chalcidoid Wasp

35

and dark specimens the tibiae are black. The body color of rubra tends more to orange rufous with no such black coloration as mentioned above.

2. In the rubra types, the wing cilia in the hyaline area immediately behind the marginal vein are silvery white and appear slightly flattened. The wing cilia of eriensis are entirely dark with none appearing flattened.

3. The hind femur of eriensis does not have the lower margin as de- cidedly convex nor the denticulate ridge as deeply notched as in rubra.

4. The abdomen of eriensis is globose ; the abdomen of rubra is pointed ovate. Also all of the tergites of eriensis are sculptured, the first to fifth tergites with reticulations and the sixth with coarse punctures. The first tergite of rubra is smooth, the second to fifth not as distinctly reticu- late as in eriensis , and the sixth tergite is not punctate.

5. In average size, the females of eriensis are slightly (0.5 mm.) larger than rubra; however, because of the few rubra specimens available, size differences should not be stressed at present.

Material examined

Holotype: female, Presque Isle, Erie, Pa., July 10, 1941. Taken while ovipositing on an ant-lion larva.

Allotype: male, which emerged from cocoon of ant-lion parasitized by holotype.

Paratypes: twenty-nine females and eighteen males all from Presque Isle, Erie, Pa. Two females taken in the field; July 11, 1933, and July 18, 1941. The other females were all reared from host material; one specimen reared July 1940, the remainder during July and August, 1941. All males, except one, reared during July, August, and September; the specimen excepted was taken on July 18, 1941.

Holotype and allotype in the collection of the Carnegie Museum; paratypes in the Carnegie Museum, and the United States National Museum, Washington, D. C.

In addition to the Presque Isle specimens and the two types of Sto- matoceras rubra studied, other material, here considered as a variety of Stomatoceras rubra , has also been examined. This additional material includes the following specimens: (1) at the U. S. National Museum — a female specimen from Brownsville, Texas; four female specimens from Beach, North Dakota; a female specimen from Raleigh, North Carolina; and a female reared from Myrmeleon immaculatus at Medora, Kansas, by R. C. Smith ; (2) at the Academy of Natural Sciences of Philadelphia —

36

Annals of the Carnegie Museum

yol. XXIX

a female specimen from Montana determined as Stomatoceras rubra Ashmead (the identification label being in Ashmead’s handwriting) ; two females and a male from North Woodbury, New Jersey; and a female specimen bearing only the label “Dac.” All of these specimens are ap- parently intermediate between the rubra types and eriensis, but are closer to the latter. The females possess the lighter color of rubra , and the wing ciliation of eriensis. The abdomen is globose and sculptured as in eriensis. The male specimen at the Academy of Natural Sciences agrees with the males of eriensis except that the prominence on the denticulate ridge is not as pronounced as in eriensis. Possibly these specimens may represent a separate variety of rubra. In order to ascertain this point, however, it seems best to wait until further reared series are available, and also to learn more of the distribution of both rubra proper and of eriensis.

Life History

Parasitism of Myrmeleon immaculatus in North America has been previously recorded by Smith,3 and by Balduf.4 Both of these references pertain to the rearing from ant-lion cocoons by Smith5 of chalcidoid specimens, which were identified as Stomatoceras sp. Other parasites of Myrmelionidae have been mentioned by Wheeler.6

Oviposition of Stomatoceras rubra var. eriensis takes place on the active ant-lion larva in the latter’s pit. Successful oviposition by eriensis in the field was observed on one occasion, and an unsuccessful attempt was observed at another time. In this last instance, the parasite was in- advertently disturbed after it had inserted the ovipositor, but before egg- laying had been completed. In addition to these two cases, oviposition by eriensis in vials was noted once. In a number of other cases, ant-lion larvae in vials were stung, but no egg was deposited.

In searching for the host, the females of S. r. eriensis walk slowly about over the sand — the antennae being bent downward and rapidly tapped or vibrated against the ground. They evince great interest in small depres-

s Smith R. C., Journ. Kan. Ent. Soc., vol. 7, no. 4, p. 137, 1934.

4 Balduf, W. V., The Bionomics of Entomophagous Insects, pt. II, John S. Swift Co., Inc., St. Louis, Mo. (1939), p. 326.

5 One of which is the specimen previously mentioned as having been studied by the author at the U. S. National Museum.

6 Wheeler, W. M., Demons of the Dust. W. W. Norton & Company, Inc., New York (1930), p. 139.

1942

Wallace: A New Chalcidoid Wasp

37

sions, furrows, footprints, etc. When a suitable ant-lion pit is found, the wasp takes up a position facing upward on the side of the pit. In this position the parasite may remain motionless for a considerable period of time providing the ant-lion has shown signs of activity. If the ant-lion senses the presence of the wasp, it begins the characteristic procedure of flipping sand from the lower portion of the pit, causing the sand around the sides of the pit to slide downward. With the commencing of this “flipping” and the downward slide of the sand, the parasite slightly elevates its wings and spreads its hind legs. In this position it slides backward into the jaws of the ant-lion larva and is pulled under the sand. Insertion of the ovipositor and the consequent release of the wasp from the ant-lion’s mandibles evidently take place immediately. In the cases observed, the ant-lion, after pulling the wasp under the sand, could be detected moving hastily about. In one instance the attacked ant-lion came to the surface of the sand in its endeavors to escape. Oviposition behavior was easily observed with a hand lens in those cases where an egg was actually in the process of being deposited. The ovipositor is inserted in the prothorax of the host, the wasp standing either to one side of the mandibles or on the head of the ant-lion. The antennae and fore legs are folded downward, the parasite being supported by the ovipositor, hind, and mid-legs. When the egg has been deposited, the antennae begin to vibrate, and the ovipositor is withdrawn. The exact time required for oviposition was not ascertained; however the length of time from the moment the parasite is seized to the time the ovipositor is withdrawn is judged to be about two minutes.

Successful oviposition on ant-lion larvae in captivity was accomplished only by a single specimen — the holotype (the same specimen that was observed completing oviposition in the field). Captive recently emerged female parasites would use the ovipositor only as a means of defense, and then not always successfully. In all cases, however, the “flipping” ac- tivity of the ant-lion caused an immediate cessation of walking on the part of the wasp, and the assumption of the characteristic attitude of lifted wings, and extended hind legs.

The parasitized ant-lion larvae appeared to be partially paralyzed, able only to feebly move their heads and mandibles. They eventually formed cocoons but did not pupate, the parasites emerging from the en- closed larvae. Parasites were never observed to emerge from larvae not enclosed in cocoons. The ant-lion larvae which were stung in self de- fense by young females of eriensis eventually died without forming co-

38

Annals of the Carnegie Museum

vol. XXIX

coons. Larvae in which eriensis had oviposited, as well as those larvae which merely had been stung, did not construct pits and were indifferent to food, i.e. live insects placed with the larvae in vials.

As stated before, two instances of successful oviposition were observed. In the case where oviposition took place in the field, the egg was de- posited on July 10, 1941, and the adult male offspring (the allotype) emerged September 6. The exact day on which the host larva spun its cocoon is not known, but it was sometime between August 4 and 23. In the case of oviposition in captivity, the egg was deposited July 10, the host larva had spun a cocoon by the morning of July 11, and a female parasite emerged August 22.

The proportion of parasitism of host material collected during 1941 was approximately 41 percent. Of 177 host cocoons, 45 gave rise to adult parasites, 28 were found by dissection to be parasitized, and 104 were not parasitized. No cases of secondary parasitism were observed.

40

Annals of the Carnegie Museum

vol. XXIX

EXPLANATION OF PLATE Stomatoceras rubra variety eriensis Figures 1-3, photographed by transmitted light.

Fig. 1. Hind leg of female (X 13).

Fig. 2. Hind leg of male (X 13).

Fig. 3. Forewing of male (X 13).

Fig. 4. Side view of male (X 8).

Fig. 5. Cocoon of ant-lion host, Myrmeleon immaculatus De Geer, showing emer- gence aperture of adult parasite (X 3).

ANNALS, CARNEGIE MUSEUM, Vol. XXIX

Art. II

Plate I

1*

t y

TMeUlL

/!

a

/

M#** ]^m

(u °0

ART. III. UPPER CRETACEOUS FAUlik OF THE'

G

°yAL KUB&rJ

ASPHALT RIDGE, UTAH By I. P. Tolmachoff

(Plates I-II)

Introduction

In the summer of 1932, the writer together with J. J. Burke and J. LeRoy Kay twice visited Asphalt Ridge, located a few miles southwest of Vernal, Utah. We did not intend to make any detailed geological investigation of the ridge but wished only to collect more fossils in the locality at the northwestern end of the ridge where a small fauna was discovered by J. LeRoy Kay in 1929. In spite of the very short time at our disposal, we succeeded in collecting a small but typical assemblage of the fauna which permits making a fairly exact correlation of the cor- responding strata. A few additions to this fauna were made, in 1933, by the writer and E. R. Eller, and in later years by J. LeRoy Kay. The whole fauna is described in the present paper.

According to H. S. Gale,1 Asphalt Ridge is composed chiefly of the strata of the Wasatch Formation underlaid from the northeast by the Mesaverde Formation of the Cretaceous. The Tertiary and Cretaceous are separated from each other by an unconformity. The Wasatch Formation was identified as such chiefly on the ground of the lithological character of its rocks. E. M. Spieker’s2 interpretation of the stratigraphy of the Asphalt Ridge is different from that of Gale. He recognized unconformity be- tween the Cretaceous and Tertiary, but the latter he considers represented by the Uinta Formation alone, the Wasatch and Green River Formations being absent from this locality. Since he thinks that a considerable thick- ness of Upper Cretaceous is also missing here, the unconformity assumed by him appears to be of considerable magnitude, although nowhere well expressed with the exception perhaps only of his “location 8” where it is possible to suggest “angular discordance between the formations.”3 His correlation of the Tertiary of the Asphalt Ridge is based on tracing the

1 Gale, H. S., Coal Fields of Northwestern Colorado and Northeastern Utah: U. S. Geol. Surv. Bull. 415, pi. XXI (map), Washington, 1910.

2 Spieker, E. M., Bituminous Sandstone near Vernal, Utah: U. S. Geol. Surv. Bull. 822-C, Washington, 1930.

3 Ibidem, p. 82.

41

Issued May 5, 1942.

42

Annals of the Carnegie Museum

vol. XXIX

Uinta Formation from central and western parts of the Uinta Basin and also on the lithological character of the formation. Eight sections described by Spieker along the ridge are placed within the Uinta Formation with a few exceptions. In Section 1, the basal horizon is “doubtless Creta- ceous. The overlying bituminous beds may also be Cretaceous.”4 Basal horizon in Section 8 is referred to the Upper Mesaverde. The whole thickness of bituminous sandstones in Section 2, about 190 feet, may also be Cretaceous. This contradicts the statement that “most and perhaps all of the bituminous sandstone of the Vernal locality is in the Uinta Formation.”6 A certain vagueness of correlation was explained, how- ever, by the lack of paleontological evidence and the necessity of depend- ing upon the general geological structure of the area, lithological char- acter, local unconformities, etc.

Paleontologists of the Carnegie Museum found in sandstones of the Asphalt Ridge, below the lowermost conglomerate, parts of a skull of Mesagriochaerus primus Peterson6 which placed the whole suite of con- glomerates and sandstones above the bituminous sandstones in the Duchesne River Formation.7 There was no attempt made at that time to trace the lower limit of that formation or subdivide formations lying under the conglomerate-sandstone series. As the Cretaceous age of lower horizons in the Asphalt Ridge was established, only after 1932, on the basis of paleontological evidence, all these formations were referred to the Tertiary.

The writer is much obliged to Dr. R. S. Bassler, through whose courtesy he was able to secure important comparative materials from the United States National Museum.

Description of Locality

Sandstones, in which the Cretaceous fauna was collected in the Asphalt Ridge, outcrop at the northwestern end of the ridge under a layer of bituminous sandstone. They are yellowish gray in color, a little shistose in upper horizons, rather loose or massive in places, and in the whole are about a hundred feet thick. Within this sandstone formation were found

4 Spieker, E. M., U. S. Geol. Surv. Bull. 822-C, p. 83.

5 Ibidem, p. 81.

6 Peterson, O. A., List of Species and Description of new Material from the Duchesne River Oligocene, Uinta Basin, Utah: Ann. Carnegie Museum, XXIII, p. 382.

7 Kay, J. LeRoy, The Tertiary Formations of the Uinta Basin, Utah: Ann. Carnegie Museum, XXIII, p. 358.

1942 Tolmachoff: Cretaceous of Asphalt Ridge 43

four layers of hard, calcareous sandstone, each about two feet thick, located as follows: one near the top of the sandstone formation, two about thirty or thirty-five feet below the top two feet distant from each other, and the fourth one about sixty feet below. Because of their hard- ness and brownish color on the surface (grayish inside) these sandstones are very conspicuous in the locality. Massive in general, they break locally into curved crusts because of their concretionary structure. Strongly effervescent with hydrochloric acid the fragments are soon broken into fine quartz.

Cretaceous sandstone underlying the bituminous sandstone corre- sponds apparently to the lowest horizon in Spieker’s sections 1,3, and 8. The southern dip, 13, 20, and 18 degrees, corresponds to one found by the writer — 15° south. The thickness of the same sandstone in the profile 1-50-100 feet also corresponds more or less to that given above.

Correlation

The separation of the Tertiary and Cretaceous has not yet been estab- lished in the Asphalt Ridge. As the well identified Tertiary lies above the bituminous sandstone and the Cretaceous sandstone underlies the latter, it is the stratigraphical position of the bituminous sandstone which has to be determined. The bituminous sandstone is a well expressed horizon. In Spieker’s sections it is subdivided into a number of beds distinguish- able from each other chiefly by the amount of bitumen, and the color and coarseness of the sand grains. In some cases bituminous sandstone is split by the insertion of shales, but such a splitting is apparently only local and the whole horizon remains a unit attaining locally a thickness of several hundred feet, but much thinner in other places. Near the northern end of the ridge a double layer of the bituminous sandstone outcrops about fifty feet below the main layer. However, nowhere on the northeastern face of the ridge could this lower horizon be discovered, although conditions of observation in many places were favorable for such a discovery. Spieker does not mention the lower horizon either, although local people often refer to the upper and lower asphalt. It is quite pos- sible, however, that the lower bituminous sandstone near the northern end of the ridge is the same upper bituminous sandstone but thrown down for about fifty feet along the fault shown here by Spieker on his map. If such an interpretation is wrong and lower bituminous sandstone exists as an independent unit, the Cretaceous sandstone will be located between two bituminous sandstones and this will be a good argument for attribut-

44

Annals of the Carnegie Museum

vol. XXIX

ing Cretaceous age to them. As was mentioned above, Spieker’s correla- tion is rather indefinite. In his sections, however, with the exception of Section 5 in which a horizon with “typical Uinta lithology” underlies a thick bituminous zone, the bituminous part of the sections is always well separated from the overlying Tertiary. This was also the general im- pression of the writer. In the southeastern continuation of the Asphalt Ridge on Green River, the Tertiary, here the Green River Formation, overlies with well expressed discordance on the suite of different, partly bituminous sandstones corresponding to those of the Asphalt Ridge. All these observations bring the writer to the conclusion in favor of the Cretaceous age of the bituminous sandstones of the Asphalt Ridge, at least in their larger part.

With a few exceptions, the fossils collected in the Cretaceous sandstone were found in the form of casts and often could be identified only tenta- tively, even as far as their generic association is concerned. Perhaps a number of the forms described in the present paper should be con- sidered as described from material not sufficiently well preserved for com- plete identification.

The following species were identified from the Cretaceous of the As- phalt Ridge:

Leptosolen conradi Meek Anatina lineata Stanton Cymella hello, Conrad

Enchodus sp.

Apateodus sp. Scapanorhynchus sp. Placenticeras meeki J. Boehm. Turritella sp.

Gy r odes sp.

Margarita sp.

Fissurella sp.

Lucina sp.

Cardium kayi sp. nov.

Tellina sp.

Cymbopora sp.

Yoldia evansi M. & H.

Crenella burkei sp. nov.

This small collection of rather badly preserved specimens shows well the marine origin of the fauna and its correlation with the Mesaverde Formation of the Upper Cretaceous. From five species known before, only Leptosolen conradi Meek is a typical form for the Dakota Sandstone at the base of the Upper Cretaceous. Of other species, Placenticeras meeki J. Boehm, was found in the uppermost part of the Mancos Shale and in the basal section of the Mesaverde Formation. Anatina lineata Stanton was found in deposits corresponding in their age to the Mancos Shale and also in deposits of the Mesaverde Formation. Cymella bella Conr. and Yoldia evansi M. & K. both were described from the Fox Hills

1942 Tolmachoff: Cretaceous of Asphalt Ridge 45

Sandstone stratigraphically covering the Mesaverde Formation and from the Pierre Shale (Fort Pierre Group) corresponding to the Mesaverde Formation.

Descriptions of the Fauna PISCES

A few fish remains found in the Cretaceous of the Asphalt Ridge are so incomplete and isolated that their identification, even only a generic one, cannot be accepted as other than tentative.

TELEOSTEI Family Enchodontidae Enchodus Agassiz Enchodus sp.

(Plate I, figs. 2-3)

A fragment of a tooth about 17.5 mm. long without the base and apex. The restoration of lacking parts should bring the whole length to about 25 mm. or a little more. The general shape is that of a long, narrow, slightly bent cone with an unsymmetricab elliptical cross-section at the upper end with sharpened but not barbed edges. The cross-section at the base is of a rounded, triangular shape. One surface (concave) is covered with delicate ribs, replaced on the convex side with fine striations.

In general appearance this tooth reminds one of Enchodus ( Phascanodus ) dirus J. Leidy from the Upper Cretaceous of Dakota8 and for this reason it is brought, tentatively, into this association.

Carnegie Museum, Catalogue of Vertebrate Fossils, no. 9423.

Family Scopelidae Apateodus Smith Woodward Apateodus sp.

(Plate I, fig. 4)

A small isolated tooth, cylindrical in the lower portion, conical in the middle, and laterally compressed at the upper end thus becoming here

8 Leidy, Joseph, Contributions to the Extinct Vertebrate Fauna of the Western Territories: Rept. U. S. Geol. Survey of the Territories, vol. I, p. 289, pi. XVII, figs. 23-24.

46

Annals of the Carnegie Museum

vol. XXIX

lanceolate with more or less sharp edges. The surface is completely smooth. Enamel yellowish white except for the sharp edges where it is nearly black. The length is 3.2 mm., and the thickness of the lower part in two diameters, 0.7 and 0.8 mm.

In its general shape this tooth reminds one very closely of teeth of different species of the genus Apateodus Sm. Woodward as for example, Apateodus striatus Sm. Woodward9 or Apateodus lanceolatus Sm. Wood- ward, 10 both from the Upper Cretaceous of England. Teeth of English forms are, however, considerably larger than the tooth from the Asphalt Ridge. However, both figures used for comparison represent premaxillary teeth which are much larger than the rest.

Carnegie Museum, Catalogue of Vertebrate Fossils, no. 9424.

Indeterminable scale

(Plate I, fig. 1)

To the Teleostei belongs also a cycloid scale of sub-elliptical outlines, a little distorted, 12.6 mm. long and 11.2 mm. wide. One end has a small triangular incision. Surface shiny, covered with fine concentric wrinkles, distinguishable only under favorable illumination.

Carnegie Museum, Catalogue of Vertebrate Fossils, no. 9422.

SELACHII Family Lamnidae

Scapanorhynchus Smith Woodward.

Scapanorhynchus sp.

(Plate I, fig. 5)

The only specimen of a small, incomplete shark tooth characterized by a long slender principal cusp of flattened, conical shape with smooth, sharp lateral edges. A shallow and narrow groove runs along one edge. The basal part of the tooth is broken off. The length of the fragment is 2.8 mm.; the greatest width LI mm.

Tentatively this form is brought into association with the genus Scapanorhynchus Sm. Woodw., which has similar long narrow teeth.

Carnegie Museum, Catalogue of Vertebrate Fossils, no. 9425.

9 Woodward, Arthur Smith, Catalogue of the Fossil Fishes in the British Mu- seum, pt. IV, p. 260, pi. XIII, fig. 6, London, 1901.

10 Ibidem , p. 264.

1942

Tolmachoff: Cretaceous of Asphalt Ridge

47

MOLLUSCA

Cephalopoda-Ammonoidea Family Cosmoceratidae (Placenticeratinae) Placenticeras Meek Placenticeras meeki J. Boehm.

(Plate II, figs. 1-2)

1927. Placenticeras meeki Reeside, The Cephalopods of the Eagle Sandstone and related Formations in the Western Interior of the United States: U. S. Geol. Surv. Prof. Paper 151, p. 29, pi. 22, figs. 5-7, pis. 23-24, pi. 25, figs. 1-2 (syn- onymy).

A fragment of an inner cast of an ammonite which in spite of its poor and fragmentary preservation can be identified even specifically with a degree of certitude. It belonged to a much compressed discoidal form of about 130 mm. in diameter. Whorls have a high, triangular cross- section and very narrow, slightly concave venter bordered by low ridges. Umbilicus deep, narrow, with abruptly rounded umbilical shoulder and hardly noticeable tubercles. The surface of the whorl was smooth and no remnants of sculpture are preserved on the cast. Some markings on the cast may be perhaps considered to be remains of the suture or at least connected with the suture. The preservation is, however, so poor that nothing more or less definite can be said about the character of the suture. J. B. Reeside described this species from the Telegraph Creek Formation and the Eagle Sandstone of Montana; from the upper part of Cody Shale and Steele Shale of Wyoming; from the lower part of Pierre Shale of the Black Hills and of eastern Colorado11; uppermost part of Mancos Shale and basal part of Mesaverde Formation from New Mexico. The stratigraphic position of this species in the central continental Cretaceous is thus well fixed. At the Wasatch Plateau it would cor- respond to the Emery Sandstone member.12

Carnegie Museum, Catalogue of Invertebrate Fossils, no. 7572.

11 Dane, C. H., Pierce, W. G., and Reeside, J. B., Jr., The Stratigraphy of the Upper Cretaceous Rocks north of the Arkansas River in Eastern Colorado: U. S. Geol. Surv. Prof. Paper 186-K, p. 230.

12 Spieker, E. M. & Reeside, J. B„, Jr., Cretaceous and Tertiary Formations of the Wasatch Plateau, Utah: Bull. Geol. Soc. Am., vol. 36, p. 439.

48

Annals of the Carnegie Museum

vol. XXIX

GASTROPODA Ctenobranchina Family Turritellidae Turritella Lam.

Turritella sp.

(Plate II, figs. 3 and 7)

This species is represented by an incomplete natural mold composed of seven whorls and a fragment of an inner cast of three whorls which may belong to the same form.

Shell small, composed probably, when complete, of about ten whorls tapering under an apical angle of about twenty degrees. Volutions are almost round, separated by deep sutures. The surface of the shell was ornamented with a few spiral ridges, faint traces of which may be per- ceived on the mold here and there. Length of the mold 11.7 mm., width 5.3 mm.; length and width of the fragment of the inner cast are cor- respondingly 6.5 and 4.5 mm.

This species is rather similar to Turritella pumila (?) Gabb. from the Upper Green Marls of New Jersey,13 but may not be more closely compared with the latter because of the poor preservation.

Carnegie Museum, Catalogue of Invertebrate Fossils, nos. 7573, 7574.

Family Naticidae Gyrodes Conrad Gyrodes sp.

(Plate II, figs. 4 and 6)

Shell low with about three volutions increasing rapidly in size. The inner ones scarcely rise above the outer ones. Volutions rather ventricose, ovate in a transverse section. Umbilicus apparently large and open. Suture deep. Faint traces of transverse ridges or lines of growth can be discerned on the casts. The greater diameter of the larger specimen 6.5 mm., the height of the same 3.5 mm.

The species under consideration is closely similar to Gyrodes abyssina

13 Whitfield, R. P., Gastropoda and Cephalopoda of the Raritan Clays and Greensand Marls of New Jersey: U. S. Geol. Surv. Mon. XVIII, p. 187, pi. XXIII, figs. 5-6.

1942 Tolmachoff: Cretaceous of Asphalt Ridge 49

Morton from the Lower Green Marls of New Jersey.14 However, this form is much larger than the Utah specimens and possibly has a smooth shell.

Carnegie Museum, Catalogue of Invertebrate Fossils, nos. 7575, 7576.

Aspidobranchina Family Trochidae Margarita Leach.

Margarita sp.

(Plate II, fig 5)

Shell small with moderately elevated spire having an apical angle of about 65°. Volutions, three to four in number, are round, ventricose, separated by deep sutures. Surface marked by fine, even, spiral ridges over the entire shell, hardly perceivable on the present cast. The height about 4.0 mm., the greatest diameter 5.0 mm.

This species is represented only by a cast which is in such a fragmentary condition that its identification would be considered impossible if it were not very similar to Margarita abyssina Gabb. from the Lower Green Marls of New Jersey.15 So far as both forms may be compared they are similar in their general shape and sculpture, but Margarita abyssina Gabb. is about twice as large as the Utah form.

Carnegie Museum, Catalogue of Invertebrate Fossils, no. 7577.

Family Fissurelidae Fissurella Lam.

Fissurella sp.

(Plate I, fig. 13)

An inner cast of a small patelliform shell of suborbicular outlines with length about 8.1 mm., and width 7.3 mm. The height was hardly more than 3 mm. The summit was located apparently rather close to the middle of the regularly convex shell. On the cast may be distinguished a perforation in the form of a slit reminding one of the genus Emarginula Lam. but not cutting through the border of the shell. The thin shell had the epidermal cover preserved only around the border in the form of a broken narrow fringe. It is smooth, glassy, and without any sculpture.

14 Ibidem, p. 123, pi. XV, figs. 9-12.

l5Whitfield, R. P., U. S. Geol. Surv. Mon., XVIII, p. 133, pi. XVII, figs. 1-5.

50

Annals of the Carnegie Museum

vol. XXIX

What is apparently the impression of a muscular scar may be seen on the side of the cast opposite the slit side.

The shape of the aperture and the lack of any sculpture make its as- signment to the genus Fissurella Lam. very problematic. Perhaps this form may have a more independent systematic position. The material on hand is, however, too poor to permit a more exact description of this form.

Carnegie Museum, Catalogue of Invertebrate Fossils, no. 7578.

LAMELLIBRANCHIATA

Desmodonta Family Solenopsidae Leptosolen Conrad

Leptosolen conradi Meek

(Plate II, fig. 9)

1876. Leptosolen conradi Meek, F. B., A Report on the Invertebrate Cretaceous and Tertiary Fossils of the Upper Missouri Country: U. S. Geol. Survey of the Territories, IX, p. 253, pi. 2, figs. 12a-b.

Shell oblong, about three times as long as high, with the hinge and parallel margins subparallel to each other and almost straight. Anterior end rounded, tapering a little from the pallial line. Posterior end roundly truncated. Both ends gaping, the posterior one more so. Beaks incon- spicuous, located behind the first third of the length of the valve. Thick- ness about half the height. Shell apparently thin and not preserved. The surface was covered with fine lines of growth more or less conspicuous on some inner casts. Found also on inner casts is a deep furrow extending from the beak downwards and a little backwards which becomes also somewhat wider towards its end. It dies out on the lower quarter of the height of the valve.

Dimensions in mm.

Length

35.2

32.4

31.5

18.6

Height

10.3

11.0

10.5

6.1

As far as general form and sculpture are concerned, the Utah form corresponds well with the Kansas form. However, the former attains larger dimensions than those given for the latter. Another difference is

1942 Tolmachoff: Cretaceous of Asphalt Ridge 51

in the character of the transverse inner ridge to which corresponds the furrow on the casts. In the Utah form it has a different direction, pointing a little backwards and is apparently longer and thinner. These differences, however, might be considered as hardly sufficient to distinguish both forms, even as plain varieties. Leptosolen conradi Meek was found in the Dakota Group of Kansas.

Carnegie Museum, Catalogue of Invertebrate Fossils, nos. 7579-7582.

Family Anatinidae Anatina Lam.

Anatina lineata Stanton (Plate II, fig. 10)

1893. Anatina lineata Stanton, T. W.t The Colorado Formation and its Inverte- brate Fauna: U. S. Geol. Surv. Bull. 106, p. 117, pi. XXVI, figs. 3-4.

Shell of medium size, elongate-ovate in outline, inequilateral, repre- sented by one left valve and two fragments. Anterior end produced, broadly and regularly rounded. Posterior end shorter and more narrowly rounded than the anterior one. Hinge border slightly concave behind the small beak and straight or even a little convex in front of it. Shell rather thin with the greatest thickness in the anterior part owing to the slight flattening of the valve in its middle. The surface is covered with abun- dant, regularly distributed rather fine, concentric folds which are better expressed on the anterior part, and fade towards the posterior end where they finally become obsolete. Length of the valve 33.4 mm., height 20.8, thickness of both valves will be about 7-8 mm.

The Utah form corresponds very closely to that of Colorado, par- ticularly with reference to a small variety figured by Stanton in fig. 4. The Colorado form was found in the Pugnellus sandstone at the top of the Fort Benton Shales. The species was also mentioned by Stanton from Utah where, at Coalville, it was found in sandstones of the “First Ridge” which may be correlated with the Pugnellus sandstone of Colorado, and from the “Third Ridge” sandstone which belongs, however, to much higher horizons corresponding to the Montana Formation or Group, per- haps to the basal section of the latter.16

Carnegie Museum, Catalogue of Invertebrate Fossils, nos. 7583-7584.

16 Reeside, J. B., Upper Cretaceous and Tertiary Formations of the Western Part of the San Juan Basin of Colorado and New Mexico: U. S. Geol. Surv. Prof. Paper 134, p. 16.

52

Annals of the Carnegie Museum

vol. XXIX

Cymella Meek Cyme 11 a bella Conrad (Plate II, fig. 8)

1875. Cymella bella Conrad, Descriptions of New Genera and Species of fossil Shells of North Carolina, in the State Cabinet at Raleigh: Rep. Geol. Surv., N. Carolina, vol. I, appendix, p. 10, pi. 2, fig. 9.

1876. Liopista ( Cymella ) undata Meek, A Report on the invertebrate Cretaceous and Tertiary Fossils of the Upper Missouri Country: U. S. Geol. Survey of Territories, vol. IX, p. 236, pi. 39, figs. la-b.

1877. Leiopista ( Cymella ) undata White, Report upon the invertebrate Fossils collected in portions of Nevada, Utah, Colorado, New Mexico, and Arizona, by parties of the Expeditions of 1871, 1872, 1873, and 1874: U. S. Geogr. Survey west of 100th Meridian, IV, p. 187, pi. XVIII, fig. 15a.

1880. Leiopista ( Cymella ) Meeki Whitfield, Paleontology of the Black Hills of Dakota: U. S. Geogr. & Geol. Survey of the Rocky Mountain Region, p. 418, pi. XI, figs. 27-28.

1885. Cymella Meeki Whitfield, Brachiopoda and Lamellibranchiata of the Raritan Clays and Greensand Marls of New Jersey: U. S. Geol. Surv. Mon. IX, p. 142, pi. XX, figs. 6-7.

Shell small, oblongly circular, almost equilateral; rather thick with sub- central prominent beaks located a little nearer to the anterior end. They are not separated from the general convexity of valves. Behind beaks on better preserved specimens it is possible to see a rudimentary escutch- eon-like area. The shell is covered with numerous rather strong, con- centric undulations which are crossed by strong radial ribs very conspicu- ous on the middle part of the valves, but fading on both ends of the latter. Length of the largest specimen 13.0, height 10.0, thickness 7. 0-7. 5 mm.

The general shape of the shell and its sculpture are typical enough to permit bringing this form into association with Conrad’s species, although the latter is twice as large as the Utah form.

Cymella undata M. & H. could hardly be distinguished from Conrad’s species. Meek remarked that both species are alike in almost all respects with the only exception that Cymella bella Conrad “has rather more prominent and ventricose beaks and less closely arranged radiating markings.” Whitfield describing Leiopista ( Cymella ) Meeki Wh. finds that from Cymella undata M. & H. his species only “Differs in the greater size and more robust habit, also in the comparatively stronger undula- tions and in having the radiating costae continuing to the anterior end, if not to the posterior also. The outline of the shell also differs somewhat.17

17 Whitfield, R. P., Paleontology of the Black Hills, p. 419.

1942

Tolmachoff: Cretaceous of Asphalt Ridge

53

Describing Cymella Meeki Wh. from New Jersey, Whitfield compared it with Cymella bella Conrad, but kept both species apart, chiefly because of the difference in the distribution of the radial ribs. In Conrad’s species they are lacking or fading on both ends of the valves, while in Liopista ( Cymella ) undata M. & H. they are supposed to become fainter and thin- ner towards both ends of the valves, being stronger in the middle part of the latter.18 In a few specimens of Liopista ( Cymella ) undata M. & H. available to the writer, radial ribs are obsolete on both ends of the valves thus making the sculpture corresponding to that of Cymella bella Conrad. Whitfield’s name might not be used as it was preoccupied by White for a different form.19 Leiopista ( Cymella ) undata M. & H., described by White, has the same size as the Utah form, but its surfacial markings are apparently thinner than in the latter. Following Grabau and Shimer,20 the finer radii distinguish Liopista ( Cymella ) undata M. & H. from Con- rad’s species. However this character is very variable and rather badly defined and may hardly be accepted for separation of both species. Henderson substituted Liopista ( Cymella ) mountanensis Henderson for Liopista ( Cymella ) undata M. & H., because the specific name undata had been preoccupied a long time before.21 However, including this species in the synonymy ol Cymella bella Conrad makes Henderson’s alteration unnecessary.

Cymella bella Conrad was described from the Ripley Formation of North Carolina. As Leiopista ( Cymella ) undata M. & H., it was described from the top of the Fox Hills Sandstone (Group) at the mouth of Judith River on the Upper Missouri, Idaho, and from the nearer non-subdivided “Strata of the Cretaceous Period” in New Mexico. In the Black Hills, Whitfield described Leiopista ( Cymella ) meeki from the Cretaceous strata which he considered at that time as belonging, probably, to the Fox Hills Sandstone (Group) but later 22 to the Pierre Shale (Fort Pierre Group).

Carnegie Museum, Catalogue of Invertebrate Fossils, nos. 7585-7587.

18 Whitfield, R. P„ U. S. Geol. Surv. Mon. IX, p. 143.

19 White, C. A., U. S. Geol. Survey west of 100th Meridian, IV, p. 186, pi. XVIII, figs. 14a-d. Stanton, T. W., U. S. Geol. Surv. Bull. 106, p. 118, pi. XXVI, figs. 5-7.

20 Grabau, A. W. and Shimer, H. W., North American Index Fossils, I, p. 532.

21 Henderson, Junius, The Nomenclature and systematic positions of some North American fossil and recent Mollusks: Nautilus, XXXII, 1918, p. 60.

22 Whitfield, R. P., Gasteropoda and Cephalopoda of the Raritan Clays and Greensand Marls of New Jersey: U. S. Geol. Surv. Mon. XVIII, p. 28. The species is mentioned here under the name of Cymella meekanum.

54

Annals of the Carnegie Museum

vol. XXIX

Heterodonta Family Lucinidae Lucina Brug Lucina sp.

(Plate I, fig. 6)

Shell suborbicular, lenticular, moderately convex, almost equilateral, with length a little greater than height. Basal margin broadly rounded, posterior and anterior ends more narrowly so. Beaks small, slightly elevated a little nearer to the posterior than to the anterior end. No traces of the sculpture are preserved on the casts. Length 3.7 mm., height 3.1 mm.

This form is represented only by a number of inner casts and hardly can be identified, even only generically, with certitude. If, in spite of this it has been brought into the association of the genus Lucina , this has happened only because of its similarity with Lucina subundata H. & M. from the Cretaceous of the Upper Missouri area.23 Although very small, the latter species is, however, still much larger than the form from the Asphalt Ridge.

Carnegie Museum, Catalogue of Invertebrate Fossils, nos. 7588-7590.

Family Cardiidae Cardium Linn.

Cardium kayi sp. nov.

(Plate I, fig- 7)

Shell small, of angularly elliptical outlines, ventricose and sharply angular along the posterior umbonal ridge. Beaks massive but rather acute, considerably projecting above the short hinge line. Anterior end rather short and regularly rounded. Posterior end long and truncate. Basal line arcuate obliquely, ascending towards the anterior border. The sculpture consists of coarse round radial ribs, thirty in number on the larger specimen. Height of the figured valve 9.8 mm., length 7.4 mm., appreciated thickness 3. 0-3. 5 mm.

The species is represented only by several inner casts of separated valves. In general shape it is pretty closely related to Fragum tenuistriatus Whitfield from the Lower Marls of New Jersey,24 but is distinguished from

23 Meek, F. B., U. S. Geol. Survey of the Territories, vol. IX, 1876, p. 133, pi. XVII, figs. 2a-e.

24 Whitfield, R. P., U. S. Geol. Surv. Mon. IX, p. 139, pi. XX, figs. 15-16.

1942

Tolmachoff: Cretaceous of Asphalt Ridge

55

the latter by its much smaller size and relatively coarser sculpture. It reminds one also of Cardium subcurtum Meek from coal-bearing Cretace- ous of Coalville, Utah.25 Outlines of both forms are, however, different. Meek’s species is less ventricose and lacks the angularity of slopes. Be- sides ribs it has also the marks of growth.

The name is given in honor of J. LeRoy Kay with whom the writer visited and examined the locality.

Carnegie Museum, Catalogue of Invertebrate Fossils, nos. 7591-7594.

Family Tellinidae Tellina Linn.

Tellina sp.

(Plate I, fig. 12)

Shell small, rather flat, transversely elongate, elliptical, almost equi- lateral with protruding umbonal part, but with the beaks not differentiated from the surface of valves. Umbonal angle about 130°. No trace of sculp- ture is preserved on the inner casts which are the only remains present. The character of the hinge and other details of inner structure are not known. The length of a left valve 15.7 mm., height 8.3, depth about 1.5 mm.

Without knowledge of the internal structure, the generic determina- tion of this form is quite tentative and is based chiefly on the close general similarity of this form with Tellina modesta Meek from the Cretaceous of Utah.26

Carnegie Museum, Catalogue of Invertebrate Fossils, nos. 7595-7597.

Family Mactridae

Cymbophora Gabb.

Cymbophora sp.

(Plate I, fig. 11)

Although this form is represented by more than a dozen specimens, even its generic identification is only tentative. All specimens are inner casts of isolated valves without any indication of the hinge structure. Very poor traces of concentric lines of growth may be distinguished in

26 Meek, F. B., U. S. Geol. Expl. 40th Parallel, IV, part I, p. 152, pi. XV, fig. 3a.

26 Meek, F. B., ibidem, p. 157, pi. XV, figs. 4-5; Stanton, T. W., U. S. Geol. Surv. Bull. 106, p. Ill, pi. XXV, fig. 3.

56

Annals of the Carnegie Museum

vol. XXIX

some cases. The form is brought into the association with the genus Cymbophora only on account of the very close similarity of its general form with such forms as Mactra (?) emmonsi Meek,27 or Mactra ( Cym- bophora) ? utahensis Meek,28 both forms from the Upper Cretaceous of Utah. From both species, and especially from the latter, the one under consideration is distinguished by its smaller size. Variations in the general shape remind one of those mentioned for Mactra (?) emmonsi Meek.

Dimensions in mm.

Thickness

Length Height (appreciated)

11.2 10.2 2.5

6.3 6.6 2.0

Carnegie Museum, Catalogue of Invertebrate Fossils, nos. 7598-7600.

TAXODONTA

Yoldia Moeller Yoldia evansi M. & H.

(Plate I, fig. 8)

1876. Yoldia evansi Meek and Hayden, A report on the invertebrate Cretaceous and Tertiary Fossils of the Upper Missouri Country: U. S. Geol. Survey of Territories, vol. IX, p. Ill, pi. 28, figs. lOa-c.

1880. Yoldia evansi Whitfield, Paleontology of the Black Hills of Dakota: U. S. Geogr. & Geol. Survey of the Rocky Mountain Region, p. 409, pi. XI, figs. 1-2.

Shell transversely elongate-sub-elliptical, a little gibbous in the central and umbonal regions. Beaks small, not protruding, located a little in front of the middle of the valves. Both ends narrowly rounded, the posterior one more so. The most prominent part of the rounding is nearer to the hinge border. Hinge border is almost straight, sloping only a little on both sides from the beaks. Numerous teeth can be seen in a number of specimens. Basal border is only slightly convex in the middle and rounds up gradually on both ends. The species is represented only by inner casts on a few of which it is possible to see indications of very fine lines of growth.

Dimensions in mm.

Length

Height

Thickness

15.2

7.0

3.0

12.3

6.3

2.5

15.6

8.5

3.4

27 Meek, F. B.f U. S. Geol. Expl. 40th Parallel, IV, part I, p. 153, pi. XV, fig. 8.

28 Ibidem, p. 155, pi. XV, figs. 9, 9a-b.

1942 Tolmachoff: Cretaceous of Asphalt Ridge 57

From the typical forms this one from the Asphalt Ridge may be dis- tinguished perhaps only by its smaller size. A few details of inner structure mentioned by Meek cannot be found in the specimens under considera- tion.

Yoldia evansi M. & H. was found at the Moreau River in the Fox Hills Sandstone (Group); on the Yellowstone River in the beds containing a mingling of forms found usually in the upper part of the Pierre Shale (Fort Pierre Group) and the Fox Hills Sandstone (Group) ; on the Milk River apparently in the upper part of the Pierre Shale (Fort Pierre Group); and in the Black Hills in the Pierre Shale (Fort Pierre Group). In the Pierre Shale and Fox Hills Sandstone it is mentioned also from Eastern Colorado.29

Carnegie Museum, Catalogue of Invertebrate Fossils, nos. 7602-7605.

ANISOMYARIA Family Mytilidae Crenella Brown Crenella burkei sp. nov.

(Plate I, figs. 9-10)

Shell roundly rhomboidal with the height a little greater than the width, almost equilateral with beaks moved forwards of the median line only a little. Beaks strong, protruding, curved above the hinge line not sepa- rated from the in general strongly convex valves. Hinge border short, sloping both ways from the beaks. Pallial border regularly rounded. The surface of the shell looks smooth but with a magnifier it is possible to discern very fine sculpture consisting of numerous very crowded thin radiating ribs crossed by very thin lines of growth. Length of the figured left valve is 6.2 mm., width 5.0 mm., depth about 1.5 mm. A few other specimens of poorer preservation, which may belong to this species, are a little larger.

Very poor material makes it impossible to determine the exact generic characters and the generic association is therefore only tentative. The name is given in honor of J. J. Burke with whom the writer visited the locality.

Carnegie Museum, Catalogue of Invertebrate Fossils, nos. 7606-7609.

29 Dane, C. H., Pierce, W. G., and Reeside, J. B., Jr., The Stratigraphy of the Upper Cretaceous Rocks north of the Arkansas River in Eastern Colorado: U. S. Geol. Surv. Prof. Paper 186-K, p. 230 and 232.

58

Annals of the Carnegie Museum

vol. XXIX

EXPLANATION OF PLATE I

All enlargements are given approximately. Where no enlargement is stated, the figure is of natural size.

Fig. 1. Scale of Teleostei. C.M.V.F., no. 9422.

Figs. 2-3. Enchodus sp. Fig. 3, (sculpture) x 5. C.M.V.F., no. 9423.

Fig. 4. Apateodus sp. x 7. C.M.V.F., no. 9424.

Fig. 5. Scapanorhynchus sp. x 8.5. C.M.V.F., no. 9425.

Fig. 6. Lucina sp. x 5.3. C.M.I.F., no. 7588.

Fig. 7. Cardium kayi sp. nov. x 1.5. C.M.I.F., no. 7591.

Fig. 8. Yoldia evansi M. & H. X 2. C.M.I.F., no. 7602.

Figs. 9-10. Crenella burkei sp. nov. fig. 9x2, fig. 10 x 7. C.M.I.F., no. 7606.

Fig. 11. Cymbophore sp. x 1.8. C.M.I.F., no. 7598.

Fig. 12. Tellina sp. x 1.8. C.M.I.F., no. 7595.

Fig. 13. Fissurella sp. x 2.5. C.M.I.F., no. 7578.

8

11

12

10

60

Annals of the Carnegie Museum

vol. XXIX

EXPLANATION OF PLATE II

All enlargements are given approximately. Where no enlargement is stated, the figure is of natural size.

Figs. 1-2.

Placenticeras meeki J. Boehm. Fig. 1, cross-section of the specimen shown in fig. 2. C.M.I.F., no. 7572.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Fig. 8.

Fig. 9.

Fig. 10.

Turritella sp. Wax cast from a natural mold, x 2. C.M.I.F., no. 7573. Gyrodes sp. x 2. C.M.I.F., no. 7575.

Margarita sp. x 2. C.M.I.F., no. 7577.

Gyrodes sp. x 3. C.M.I.F., no. 7576.

Turritella sp. x 3.5. C.M.I.F., no. 7574.

Cymella bella Conrad, x 1.5. C.M.I.F., no. 7585.

Leptosolen conradi Meek. C.M.I.F., no. 7579.

Anatina lineata Stanton. C.M.I.F., no. 7583.

ANNALS, CARNEGIE MUSEUM, Vol. XXIX

Art. Ill

Plate II

8

9

lO

//iS§%v

(*

S '942

|v4:

' ■ ' - • ■ Ml :

ANNALS, CARNEGIE MUSEUM, Vol. XXIX Art. IV Plate

Type of Phoca vitulina mellonae described on page 111

ART. IV. A REVIEW OF THE GENUS PHOCA

By J. Kenneth Doutt

(Plates I-XIV)

Table of Contents

Introduction 62

Acknowledgments 62

Geography and geology of Seal Lake region 64

Life zones 65

History of the Phoca of Seal Lake 65

Previous exploration of the region 67

Biology of the seals 68

Derivation of the seals in Seal Lake 71

Relationships of the race of Phoca vitulina in Seal Lake 73

Time of isolation 74

Taxonomy 79

Key to the species of the genus Phoca 80

Descriptions of species 82

Phoca vitulina 82

Phoca hispida 85

Phoca groenlandica 87

Phoca fasciata 91

Comparisons 95

Variation in the genus 99

Supernumerary bones in the base of the cranium 99

Age determination 101

Type description Ill

Relationship of the races of Phoca vitulina 114

Summary 118

Specimens examined 119

Literature cited 123

61

Issued May 12, 1942.

62

Annals of the Carnegie Museum

vol. XXIX

Introduction

The seals living in the landlocked lakes of the world present a problem in speciation which is intimately associated with the geological history of the region, and furthermore, they offer excellent examples of the effect of isolation on the development of new species. This paper announces the discovery of a new race of Phoca vitulina from Seal Lake, Ungava, and presents evidence indicating the length of time which has elapsed since the race was separated from the original stock. It also offers descrip- tions and keys for the identification of the northern species of the genus Phoca; discusses methods of age determination ; and presents a list of the recognized races of Phoca vitulina.

Acknowledgments

During the course of this study many persons have assisted in various ways with the progress of the work, and it is a pleasure to express here my gratitude to them. Without the generous contributions of Mr. and Mrs. W. L. Mellon, the expedition could never have left Pittsburgh. For financial assistance I am indebted also to Dr. George H. Clapp, Mr. R. K. Mellon, Mr. John B. Semple, Mr. George D. Lockhart, Mrs. Lawrence C. Woods, Mr. and Mrs. Lawrence C. Woods, Jr., Lieut. Col. Paul C. Hunt, Mr. W. E. Clyde Todd, and Miss Margaret Shaw Campbell. I am in- debted to Mr. Lawrence C. Woods, Jr., for much enthusiastic support in soliciting funds, and also for his pleasant companionship during the latter part of the expedition, when he joined the party in the field. To Dr. Andrey Avinoff, Director of the Carnegie Museum, I am obligated, not only for much encouragement, but also for assistance in raising funds, for the privilege of studying at other institutions, and for many other courtesies.

Mr. W. E. Clyde Todd initiated the plans for the expedition, personally supervised the purchase of all provisions, and gave constant advice and assistance throughout the preparation for the trip.

Without the cooperation of the Hudson’s Bay Company the expedition would have been practically impossible. We are especially indebted to the District Manager, Mr. M. Cowan, and the Post Managers, Mr. Norman Ross and Mr. Robert Cruickshank.

To my companion throughout the trip, Dr. Arthur C. Twomey, at that time Field Collector for the Section of Ornithology, I am grateful for many months of comradery.

1942

Doutt: Review of the Genus Phoca

63

For the privilege of examining specimens under their charge, and for many helpful suggestions, I am indebted to many persons at other in- stitutions, especially to the following: Dr. Glover M. Allen and Mr. Harold J. Coolidge, Jr., Museum of Comparative Zoology, Cambridge, Massachusetts; Dr. R. M. Anderson, National Museum of Canada, Ottawa; Mr. Gerrit S. Miller, Jr., and Dr. Remington Kellogg, U. S. National Museum; Dr. H. E. Anthony, Mr. George G. Goodwin, and Mr. T. Donald Carter, American Museum of Natural History; Mr. Charles M. B. Cadwalader and Mr. Wharton Huber, The Academy of Natural Sciences of Philadelphia; Mr. C. D. Bunker and Mr. Claude W. Hibbard, University of Kansas; Dr. Robert T. Orr, California Academy of Sciences; and Dr. Seth B. Benson, Museum of Vertebrate Zoology, Berkeley, California.

For information concerning the geology of the region around Seal Lake and the probable time of isolation of the seals, I am likewise in- debted to: Dr. Ernst Antevs, Globe, Arizona; Professor Kirk Bryan and Professor R. A. Daly, Geological Museum, Harvard University; Professor Douglas Johnson, Columbia University; Mr. W. A. Johnston, Campbell- ford, Ontario; Professor Richard J. Lougee, Colby College; Mr. D. A. Nichols, Victoria Memorial Museum, Ottawa; Dr. Horace G. Richards, New Jersey State Museum; and Dr. George M. Stanley, Department of Geology, University of Michigan. Where I have made use of specific information received from these men, special mention is made of the fact in the body of the paper.

For information and advice concerning the anatomy of the seals I wish to express my indebtedness to Dr. William K. Gregory and Mr. H. C. Raven, American Museum of Natural History; Dr. Remington Kellogg, U. S. National Museum; Professor A. Brazier Howell, Johns Hopkins Medical School; and Dr. G. C. L. Bertram, Scott Polar Research Institute, Cambridge, England.

I want to express my gratitude to Dr. Glover M. Allen, Dr. Herbert W. Graham, and to my wife, Dr. Margaret T. Doutt, for reading the manuscript and for making many helpful suggestions for its improvement.

Miss Olive L. Bown (now Mrs. Coleman J. Goin) assisted me through- out the study, and I am grateful to her for the unfailing energy and enthusiasm with which she worked, as well as for the many helpful sug- gestions which she made.

Miss Caroline A. Heppenstall has been an invaluable assistant in the preparation of the manuscript, and has assumed the responsibility for all

64

Annals of the Carnegie Museum

vol. XXIX

its literary details. Mr. Sidney Prentice made the drawings; and Mr. Robert M. Burford copied the maps reproduced here.

Geography and Geology of Seal Lake Region

Lower Seal Lake lies approximately ninety to one hundred miles inland from the east coast of Hudson Bay, between fifty-six and fifty-seven de- grees north latitude. According to Low (1898, p. 13 L), it has an elevation of approximately 800 feet above sea level. It is about fifty miles long and varies in width from about a half mile to five miles. Rounded rocky hills rise from 100 to 300 feet above the level of the lake (figs. 1-3, pis. II, III, XIV).

Fig. 1. Map of the Ungava Peninsula, sometimes called the Labrador Peninsula, which is bounded on the west by Hudson Bay, on the north by Hudson Strait, on the east by the Atlantic Ocean, and on the south by the St. Lawrence River and Gulf.

The Indian route from Hudson Bay to Seal Lake passes through a large lake, shown as Clearwater Lake on most maps, but known to the Indians as “We-ya-sha-ga-me.” From Hudson Bay to Clearwater Lake old sea beaches can be seen along the sides of the hills, and numerous rounded

1942

Doutt: Review of the Genus Phoca

65

boulders, perched high on the hilltops, give abundant evidence that the region has been submerged beneath the sea. For fifteen miles, between Clearwater Lake and Seal Lake, the canoe route follows a series of lakes which are connected by small streams; but the actual separation between Seal Lake and the headwaters of Clearwater Lake is a narrow ridge about 500 yards across. In these fifteen miles, a remarkable change takes place. The old sea beaches and the rounded boulders disappear. In their place huge angular blocks of granite and diabase rest on the hillside, where they were dropped by the glacier, and long eskers follow down the valleys.

From this evidence it seems certain that the sea, at one time, extended as far inland as Clearwater Lake, but that it did not reach Seal Lake.

Life Zones

Seal Lake lies at the northern edge of the Hudsonian Zone. In the pro- tected valleys, black spruce and tamarack are common trees, but these are decidedly stunted. The tops of the hills are covered mainly with moss and lichens, typical of the Arctic Zone (Plate III).

The common mammals are the black bear, the marten, the weasel, the otter, the white and the red fox, the lynx, the red squirrel, and the barren ground caribou. Common winter birds are the rock and the willow ptarmigan, the Canada jay, the red poll, the hawk-owl, and the white- winged crossbill (Plate IV).

History of the Phoca of Seal Lake

More than forty years ago, Mr. W. E. Clyde Todd began a study of the birds of the Ungava Peninsula. During the course of his work in the region, he heard a rumor of the occurrence of landlocked seals in Lake Minto. When he financed the 1935 Expedition to Hudson Bay (Doutt, 1935, pp. 195-200), he suggested that I try to reach Lake Minto and col- lect specimens of the seals found there. A study of the Twin Islands in James Bay was the primary objective of that trip, however, so it was August before we reached Great Whale River, the starting point for the trip to Lake Minto. Here I met an Eskimo named Kooke, who had lived on the shores of Lake Minto. He agreed to take me to the lake but, when I told him why I wanted to go, he said there were no seals in that lake. He said he had spent his childhood and youth in the vicinity of Lake Minto; that he knew there were no seals in that lake; and that what he told me was the truth. He said seals were found near the mouth of

66

Annals of the Carnegie Museum

vol. XXIX

the Leaf River and that they ascended the river some distance, but that none ever got into Lake Minto. As I talked with him further about the matter he told me that thirteen years previously he had killed four seals in Lower Seal Lake, and, while he had never been to Upper Seal Lake, the Indians said many seals were to be found there also. I was convinced that what he told me was the truth, so, for the time at least, I gave up the idea of going to Lake Minto. The occurrence of seals in Lake Minto is still an open question. Kooke’s definite statement that there are no seals in Lake Minto is backed up by a similar statement from two of the Indian guides, James and Joseph Sandy, who accompanied us on the trip to Seal Lake in 1938. These men live in the vicinity of Clearwater Lake and frequently travel to Lake Minto while hunting caribou. They should know whether or not seals live in Lake Minto, and I know of no reason why they should not have told me the truth about their presence or ab- sence there. They were very definite about this statement that no seals were to be found in Lake Minto. Opposed to the statements of this Eskimo and the Indians are references by Low and Flaherty. Low makes no defi- nite statement about seals in Lake Minto, but he refers to it as “Kasiaga- luk Lake” (1902, p. 34). “Kasaguea” is the Eskimo name for Phoca vitulina. It seems strange that the Eskimos would call this “Seal Lake” if there were no seals in it. Flaherty (1918, pp. 119-120) makes a more definite statement about the matter. In speaking of Lake Minto he says, “The lake is famous among the Eskimos as the habitat of the fresh-water seal, hunted primarily not as food, but for the pelt, which, much darker, softer, and more lustrous than that of the salt-water variety, is used for their finer garments.” Neither Low nor Flaherty mentions seeing seals in Lake Minto, however. I am unable to reconcile these conflicting state- ments.

The presence of seals in Lower Seal Lake was already known. In 1896, A. P. Low made a traverse of the Labrador Peninsula from Richmond Gulf to Ungava Bay. In the report which he published concerning that traverse (Low, 1898, p. 13 L), he records the fact that three seals were seen in the lake and that the Indians annually kill more than thirty. To the best of my knowledge this is the first published record of seals in Seal Lake, although the Indian name for Lower Seal Lake is “Mushawa Atchiguanipe” which means “Barren Seal Lake,” in contrast to the name for Upper Seal Lake, which they call “Menasqua Atchiguanipe,” meaning “Seal Lake in the Woods.” These native names suggest that for genera- tions the Indians have known that seals lived in these lakes. In the forty-

1942

Doutt: Review of the Genus Phoca

67

odd years that the white man has known about these seals no specimens have ever been collected, and the identity of the species has been open to question (figs. 2, 3, pis. II, XIV).

We returned from Great Whale River to Moosonee by canoe, and dur- ing part of the trip the Revillon Freres and Hudson’s Bay Company interpreters, Ernest Cadot and Roderick McDonald, accompanied us. Ernest gave me a sealskin bag which he said he had obtained from an Indian at Richmond Gulf. He thought the skin was from a seal which had been killed in Seal Lake. The specimen was that of a harp seal, the occurrence of which would be most unusual in a fresh-water lake.

After we returned to Pittsburgh plans were made for an expedition to Seal Lake, the purpose of which was to collect specimens of this fresh- water seal. This second expedition left Pittsburgh on January 2, 1938 (Doutt, 1939, pp. 227-236). After numerous difficulties, we reached Seal Lake on March 12, and ten days later obtained our first specimen of seal. It was a female, carrying a well-developed embryo. On the following day another specimen, a male, was obtained. They were a race of the harbor seal, however, not the harp seal.

Previous Exploration of the Region

Since 1824, five different parties of white men have passed through Lower Seal Lake. The first white man to make the journey, Dr. Mendry, apparently left no account of it except a rough sketch map. What little we know about his journey is recorded by Low (1896, p. 15 L) : “In 1824, a party was fitted out at Moose Factory to proceed overland to Ungava Bay and there establish a post; but it was not until three years later that this was accomplished by Dr. Mendry, who coasted along the east shore to Richmond Gulf, and then passed inland to Clearwater and Seal Lakes, thus reaching the headwaters of the Larch Branch of the Koksoak River, which was descended to near its mouth, and Fort Chimo there established. This trip is the basis of Ballantyne’s ‘Ungava’ a popular story for boys. A map made of the route by Dr. Mendry, is at present at Moose Factory, and a tracing of it is in the Geological Survey office; the part between Clear- water Lake and the forks of the Larch River has been used, in the com- pilation of the map accompanying this report.” In another report (1898, p. 6 L), he comments on Mendry ’s trip as follows: “The route followed between Hudson Bay and Ungava Bay was first passed over in 1824 by Dr. Mendry, when sent by the Hudson’s Bay Company from Moose

68

Annals of the Carnegie Museum

vol. XXIX

Factory to establish a trading post at the mouth of the Koksoak River. The only known record of his trip is a rough map of his journey, from which a copy was taken at Moose Factory in 1887 ; since then the original map has been lost.”

I am likewise indebted to Low for a record of the second white man who passed through Seal Lake. Although an account of his trip was published, I have been unable to locate a copy of it. Low (1898, p. 6 L) recounts his trip in one sentence, “In 1885, the Rev. J. Peck, of the Church Mission Society, crossed by the same route and subsequently wrote a short account of his trip which was printed in a publication of the Society.” Low (1898) gives an excellent account of his own exploration in the region from Richmond Gulf to Ungava Bay.

The next party to pass through Seal Lake was led by Stephen P. Tasker and his wife. Their chief guide was George Elson, now famous for the part he played in the ill-fated expedition of Dillon Wallace and Leonidas Hubbard (Wallace, 1905), and the trip of Mrs. Hubbard (1908). Tasker’s trip was made in 1906 and an account of it was published by Mrs. Florence A. Tasker in “Field and Stream” for February, and March, 1908. Mrs. Tasker lists some of the most essential items in their equipment and gives a very interesting account of their expedition. No especial mention, however, is made of Seal Lake.

Daniel Petacameshcum, our chief Indian guide, told us of a party of eight prospectors who passed through Seal Lake a year or two previous to our visit there, and at one place pointed out to me the stakes where their tents had been pitched. He did not know who they were, and I have been unable to find any account of their trip.

Biology of the Seals

Seal Lake is more than fifty miles long, and has several long narrow arms. The Indians informed me that Upper and Lower Seal Lakes are con- nected, although they are shown on most maps, as separate lakes, each with its own outlet. James Sandy, an Indian who hunts in this region, drew for me a map of Upper and Lower Seal Lakes and explained how the seals were able to get from one part of the lake to the other (figure 2). I talked with him and several other Indians, especially Daniel Petaca- meshcum, Jacob Rupert, Luke Cash, and Thomas George, who hunt in the vicinity of Seal Lake, and also with Jimmie Egomea (Kumiak), an Eskimo who accompanied us on our trip to Seal Lake. They told me many things

1942

Doutt : Review of the Genus Phoca

69

Fig. 2. Sketch map made by Jacob Rupert, showing the connection between Upper and Lower Seal Lakes. Upper Seal Lake is the lowermost elongate body of water, not indicated by name in this sketch, which apparently empties into Hudson Bay through Little Whale River.

about the fresh-water seal. I believe they were telling me the truth as far as they knew it, and I present here the disconnected bits of informa- tion obtained from them, for whatever they may be worth. Before we left Richmond Gulf, the Eskimos said we would not find breathing holes of the seal in this fresh-water lake. When we arrived at Seal Lake the Eskimo, with the help of his dog, hunted diligently, but he did not find anything which he thought was the same as the breathing hole which the jar seal {Phoca hispida ) makes. He did find a small hole in a crevice of the ice which he thought the seals had been using for a breathing hole. He watched at it for several hours on two or three different days, but saw no seals at it.

Jacob Rupert, one of the Indians who was with us on the hunt, was the son of an old Indian known as the “Seal Hunter.” From him Jacob had learned how to make a seal net of a special design for setting in narrow places between the lakes. Jacob set several of these nets for us, but we caught no seals in them. Daniel, another Indian we had with us, said that from one to five seals spend the winter at these places of open water. Usually there are one or two old ones and one or two young ones. Probably all are of one family.

70

Annals of the Carnegie Museum

vol. XXIX

“ Seals come out on the ice only on warm sunny days — on cold windy days they stay in the water.”1

Fig. 3. Sketch map made by Daniel Petacameshcum, showing Clearwater Lake, Lower Seal Lake, and Little Seal Lake. The dates and locations of our camps are indicated.

1 I use quotation marks here to indicate that the enclosed sentences are state- ments of the Indians or the Eskimo, although they are not direct quotations but are the translation and interpretation of what they told me. In many instances these are the summaries of long discussions.

1942

Doutt : Review of the Genus Phoca

71

“The embryo of the fresh-water seal is yellowish in color, with definite evidence of spotting, and with rather smooth hair, as compared with the embryo of Phoca hispida which is white, with no evidence of spotting, and is covered with long, woolly hair.”

“ Phoca hispida probably weighs nearly fifteen pounds when born, and is about 750 mm. long. The young is born in February, or March, in a cavity hollowed out between the ice and the snow which covers it. In Seal Lake the young seals are born about the end of April, or early May, when the birds are here. When born they weigh about thirty pounds. The young are born on land after the snow and ice have melted.”

Jacob said that the male lives about the same hole with the female and that he stays with her and helps when she is having her young. In the summer he said that he saw them together occasionally, but in the winter they were always in pairs.

Open water is the place to get seals when the lakes are frozen. Such places are found in the rivers where steep hills narrow the channel and make swift water. The holes change rapidly with changes in the weather — warm weather opens new holes and makes old ones larger — cold weather closes up the smaller ones entirely and narrows the larger ones to small dimensions.

The stomach of the adult female we obtained was nearly filled with well-digested pieces of fish.

Derivation of the Seals in Seal Lake

Since Seal Lake is about ninety miles from Hudson Bay, the nearest arm of the sea, and lies at an elevation of about 800 to 860 feet, the question which naturally arises is, how did the seals get into Seal Lake? It would be interesting also to know how long they have been there and whence they came. Since this form represents a new race — distinguishable from its closest relatives not only by color, but also by skull characters — we have here a means of determining approximately how long it has taken for a new race to develop. In this case the animal has been marooned in a habitat different from that in which the species normally lives. Although Phoca vitulina customarily ascends the fresh-water rivers in the spring, its usual habitat is the sea. The seals in Seal Lake are confined to a fresh- water lake throughout the year and have no access to the sea. Moreover, the food found in this lake must be markedly different from that found in the sea, or in rivers which have easy access to the sea.

There are several suppositions as to how the seals got into Seal Lake.

72

Annals of the Carnegie Museum

vol. XXIX

The Indians have a legend that a seal was sleeping on the ice and its breath- ing hole froze shut, so it was compelled to travel over the ice and snow to find another breathing hole. This legend disregards so many of the facts, however, that it seems possible to dispose of it with little consideration.

The seals might have gained access to Seal Lake by way of the Koksoak, Larch, and Stillwater rivers. The journey over the height-of-land, from the headwaters of the Stillwater River to the headwaters of Seal Lake, is only a matter of about fifty yards, but to the best of my knowledge seals are not known above the Koksoak and Kenogamistuk rivers. If they entered Seal Lake by this route, shortly after the retreat of the glacier when the land was at a lower elevation, the result would be about the same as if they had entered by way of Hudson Bay drainage at that time. Daniel, our Indian guide, told me that many years previous to our visit someone had seen a seal in Clearwater Lake, but both he and Jacob said they had never heard of one being killed there. At one place, Seal Lake is separated from a tributary to Clearwater Lake by a narrow ridge, scarcely more than 500 yards wide, and certainly less than a hundred feet in ele- vation. In view of the proximity of these two large lakes, it seems strange that the seals have not migrated across this slight barrier and become established in Clearwater Lake too. According to reports of the Indians, Clearwater Lake is much better stocked with fish than is Seal Lake.

Low (1898, p. 13 L) sets forth his assumptions concerning the presence of seals in Seal Lake as follows: “The name [Seal Lake] is derived from the seals living in its waters, which are either the common harbour seal ( Phoca vitulind) or a closely allied species. The harbour seal is known to travel overland for considerable distances, but its presence in this lake nearly a hundred miles from salt-water at an elevation of nearly 800 feet above the sea, can hardly be due to its migration up such a rough stream as the Nastapoka. Another way in which it might have reached the lake was during the subsidence of the land at the close of the glacial period, when the lake was nearer sea-level than at present by more than 600 feet, and when the deep bay extended inland up the present valley of the Nastapoka to or near the outlet of the lake, with such conditions it would be easy for seals to reach the lake, and having found it full of fish they probably lost the inclination to return to the sea. Three seals were seen in the lake, and the Indians kill annually more than thirty, showing that the animal breeds freely in the fresh water.”

My own observations along the route into Seal Lake led me to the same conclusion. Although specimens of Phoca vitulina are killed occasionally

1942

Doutt : Review of the Genus Phoca

73

along the coast of Hudson Bay, there is no reason to assume that these animals might migrate in and out of Seal Lake. The Nastapoka River, which is the outlet of Seal Lake, enters Hudson Bay with a fall of 170 feet. Even making ample allowance for the remarkable migratory powers of Phoca vitulina across the land, it still seems unbelievable that a seal would pass around a barrier such as this in order to proceed up such a turbulent stream as the Nastapoka. Furthermore, if we did assume that the seals could migrate from Hudson Bay up the Nastapoka River to Seal Lake, we would have to explain why they did not go up many other rivers of Ungava, and inhabit innumerable lakes in the interior. There are many other rivers which would be easier to ascend than the Nastapoka, and many other lakes which are better stocked with fish than Seal Lake. It seems to me, therefore, that Low’s explanation is the only satisfactory one.

Muskrat Falls on the Hamilton River, Labrador, is an excellent example of the case in point. On our return from the Grand Falls in August, 1939, we saw approximately fifty seals on a sand bar and in the water just below Muskrat Falls. Our guides said that the seals come up the river as far as Muskrat Falls every year, but that they never get beyond that point. We saw no evidence of them above the Falls. Muskrat Falls is not a for- midable barrier. Low (1896, p. 130 L) described it as follows: “For three miles above Muskrat Island, the river narrows to less than a third of a mile, with a narrow island obstructing the channel in the upper mile. Above this narrow, the channel widens out into a nearly circular basin about two miles across, into the west side of which the river pours with a chute of twenty feet called Muskrat Fall. Above this chute is a heavy rapid 400 yards long, with a chute of twenty-five feet at its head, the total fall being seventy feet. At the chutes, where it rushes over ledges of gneiss, the river is only about 100 yards wide. Immediately on the north side of the falls, there is a rounded, rocky hill rising 250 feet above the level of the valley.” Low does not describe the terrain on the south side of the river, but here there is a bench not far above the level of the river, which would be a very insignificant barrier in comparison with that of the Nastapoka River, the outlet of Seal Lake.

Relationships of the Race of Phoca vitulina in Seal Lake

Although specimens from Seal Lake do not have the premaxillaries ex- tended back along the nasals — a character which separates about eighty- five per cent of the specimens of Phoca vitulina richardii from Phoca

74

Annals of the Carnegie Museum

vol. XXIX

vitulina concolor, still in some minor characters they resemble Phoca vitulina richardii more than they do Phoca vitulina concolor. Although this resemblance is slight, it led me to consider the possibility that the stock now living in Seal Lake may have been derived from the Pacific Ocean rather than from the Atlantic.

It is possible that, during the Pleistocene, Hudson Bay was connected with the Pacific Ocean via the Arctic Ocean and Bering Strait, while a barrier of ice or land separated it from the Atlantic, and that the seals which now inhabit Seal Lake came originally from the Pacific into Hudson Bay, rather than from the Atlantic. In a letter to me, dated December 6, 1939, Dr. George M. Stanley of the University of Michigan says, “The existence of an ice barrier across Hudson Straits while a free connection existed between Hudson Bay and the Pacific seems entirely possible for some stage of ice retreat.” In a letter dated at Ottawa, December 15, 1939, Dr. D. A. Nichols says, “Nearly all of Boothia and Melville Penin- sulas were covered by the sea, and a belt of the mainland extending roughly from Queen Maude Gulf to Hudson Bay, taking in the area about Garry Lake, Baker Lake, and the wide belt of Coastal Plain west of Hudson Bay.” Although this presents no positive evidence, it does indicate that the route of migration was open, and that access from the Pacific Ocean into Hudson Bay was much easier then than now.

While seals are able to swim against strong currents, and are not de- pendent on them for distribution, still, it is interesting to note that the currents of Arctic America flow from west to east, rather than from east to west (Nichols, 1940, pp. 18-22). This general trend would seem to aid the migration of forms from the Pacific to Hudson Bay, and would tend to hinder the migration of forms from the Atlantic into Hudson Bay. Un- fortunately I have no specimens from Hudson Bay which shed any im- portant light on the subject. Thus the evidence is not sufficient to justify any conclusions, but it does leave open an interesting problem. To what extent may the fauna and flora of Hudson Bay have been derived from the Pacific Ocean?

Time of Isolation

If the time which has elapsed since the seals became isolated in Seal Lake could be determined, it would help to answer the question — how long does it take to make a species? Obviously there are many factors which influence this transformation, one of the most important of which is whether or not the environment has changed. However, if this question

1942

Doutt : Review of the Genus Phoca

75

is ever to be answered I think it will be answered best by an accumulation of many specific instances. In this case the post-glacial history of the region is especially significant. Of course there are differences of opinion among geologists as to the length of time required for many of the post- glacial events to take place, since these events have progressed at different rates and have reached different magnitudes in different regions where they have been studied. No work of this nature has been done in the in- terior of the Ungava Peninsula. Because of all the unknown factors, I have tried to obtain opinions from several authorities on the subject so that I might present maximum and minimum figures. I was surprised to find the agreement so close and the estimated length of time so short.

Answering my question concerning the possible length of time that the seals had been isolated in Seal Lake, Dr. Nichols, in a letter dated Decem- ber 15, 1939, replied as follows: “Antevs, Memoir 168 ‘Late Glacial Cor- relation and Ice-recession in Manitoba’ (Geological Survey of Canada) states that the Post-Glacial epoch commenced about 9,000 years ago and, at that time, in eastern North America the ice-sheet was confined to the Labrador Peninsula.

“The younger late-glacial epoch may have lasted 2,000 years after the uncovering of the Cochrane area, Ontario. So the Seal Lake area would apparently be free of ice from about 9,000 to 10,0002 years ago, and ac- cessible to oceanic waters.” In answer to the same question, Dr. George M. Stanley (letter dated December 6, 1939) says, “My guess is that some 5,000 to 12,000 years have probably elapsed since the formation of the highest marine beaches about Richmond Gulf, which was about the time when Seal Lake was most accessible from the sea, . . . .”

Dr. Ernst Antevs, in a letter dated at Globe, Arizona, December 17, 1939, says, “My guess is that Lower Seal Lake was separated3 from the Hudson Bay some time between 8,000 and 6,000 years ago.”

Thus it seems that Seal Lake was an embayment of Hudson Bay be- tween 9,000 and 10,000 years ago. Since then, the land has risen about 800 feet. As the land rose, the connection between Seal Lake and Hudson Bay was gradually severed.

It is impossible to say how much the land would have had to rise in order to cut off the seals in Seal Lake from access to Hudson Bay, for it

2 Dr. Ernst Antevs, in a letter to Doutt dated December 16, 1941, suggests that these figures should be 7,000 to 8,000.

3 That is, “formed by being raised above Hudson Bay.” Letter from Dr. Antevs, December 16, 1941.

76

Annals of the Carnegie Museum

vol. XXIX

is a well-known fact that seals can live in fresh water and that they may travel long distances up rivers, even passing strong rapids. It is evident, however, that a rise of something less than 800 feet has been required. It is possible that a rise somewhere between 400 and 600 feet would be sufficient; on the other hand, it may not have required more than 200 feet. These are purely guesses on my part, for nothing is known of the topog- raphy along much of the Nastapoka River, and thus it is impossible to say whether it is the present falls at the mouth of the river or some other falls farther inland which has cut off access to the sea. Within these limits, however, it is possible to make maximum and minimum estimates, which, geologically speaking, are not so very far apart.

Professor R. A. Daly, in a letter dated November 5, 1941, has pointed out that Gutenberg (1941) has estimated that the rate of uplift is nearly two meters per century, but that formerly the rate was two or three times faster. Using a rate of four meters, or about twelve feet per century, and assuming that a maximum uplift of 600 feet was required, it would have taken about 5,000 years after the ice left to complete the isolation of Seal Lake. Thus, if Seal Lake had been free from glacial ice about 9,000 to 10,000 years ago, and it required 5,000 years after this to complete the isolation, the seals have been confined in Seal Lake approximately

4.000 to 5,000 years. If, on the other hand, a rise of only 200 feet was required, this could have been accomplished in about 2,000 years, making

7.000 to 8,000 years since isolation took place. Of course it is well known that the rate of rise was not constant. For some time after the ice melted, uplift was very rapid, but more recently it has been gradually reduced. It seems undesirable, however, to go into such refinements at this time, because they imply a degree of accuracy which is impossible to attain until more is known about the region between Seal Lake and Hudson Bay.

Daly (1934, p. 71) gives a curve showing Nansen’s opinion as to the rate of uplift at the center of the Scandinavian Ice Cap at the time when the last ice was removed from that center. In a letter, dated December 2, 1941, he makes reference to this and says, “The rate then was about ten times the present rate. Two thousand years after, the rate had fallen to three times the present rate. Let us take 800 feet as the elevation of Seal Lake; this is roughly 250 meters. Assuming that the mean rate of uplift up to the time of isolation of Seal Lake was five meters per century, the preceding uplift would have taken about 5,000 years. If the mean rate were seven meters per century, the time taken would be roughly 3,500

1942

Doutt: Review of the Genus Phoca

77

years.4 A good estimate of the total time since uplift began at this point during the ice-free time is 8,000 years. Subtracting 5,000 and 3,500 from 8,000, we have, respectively, 3,000 and 4,500 years as the life of Seal Lake, a freshwater body. Probably the value 3,500 years is the best bet. You see that this estimate agrees with your own rather nicely.”

Dr. W. A. Johnston, in a letter of November 28, 1941, writes, “If the land would have had to rise 400 to 600 feet before tne seals in Seal Lake were cut off from access to the sea, as you suggest, three to four thousand years is a reasonable estimate of the time that has elapsed since the seals were isolated in Seal Lake.”

Dr. Stanley, who has visited Hudson Bay, and has personally examined and measured the region around Richmond Gulf, expresses his opinions as follows (letter dated December 9, 1941), “The very most northerly parts of the Great Lakes region seem to be rising at a rate of approximately one meter per century which is about the same near the center of the glaciated tract in the Baltic. At this rate, as a guide and not to be specifically as- sumed, the coast near the Nastapoka River would have taken almost 4,600 years to rise 150 feet, over 5,100 years to rise 170 feet.

“The rate of rise at the place in question may very likely be two or three times as great as that employed in this calculation, at any rate, greater, almost undoubtedly. If we were to suppose an extreme case — that 150 feet of uplift had occurred here within the last thousand years — the probability of the last century’s share in this having escaped a uni- versal recognition by local people seems remote.

“I think you would do well to avoid specific dating of a matter so questionable. The land-locking of those seals is interesting enough re- gardless of the exact time of isolation.”

There is no evidence to show how long after Seal Lake was isolated that the seals became sufficiently different to be recognizable. It may be that differentiation took place very rapidly at first, and that the progress has already slowed down or ceased, and that little change is now in progress. On the other hand, it is just as logical, and I believe more reasonable, to assume that the process of differentiation is still in progress, and given another five thousand years, the race will be even more distinct. The fact that the seals in Lake Baikal differ more from Phoca hispida than the Phoca vitulina of Seal Lake does from Phoca vitulina vitulina of Europe

4 This is the time required to raise Seal Lake to an elevation of about 800 feet, probably a rise of something less than this was required to isolate the seals.

78

Annals of the Carnegie Museum

vol. XXIX

probably indicates a longer time of isolation, and would suggest that after isolation, differentiation continues indefinitely.

These seals have a comparatively long life span, and breed rather slowly. Although I have made considerable effort to locate a specimen of seal of known age I have so far been unsuccessful. Thus I am unable to state how long a seal may live. I believe, however, that ten to fifteen years may be a reasonable figure to place on the life span of the individuals of this species. S. S. Flower (1931, p. 181) in speaking of Phoca vitulina says that although this species is frequently kept in captivity, it usually lives only a very few years. He continues his statement, however, as fol- lows: u. . . eight individuals that lived for over six years averaged 10 years 9 months, with a maximum of over 14 years.” Of course it is pos- sible that in a wild state they may live considerably longer. Usually only one young is born at a time, and breeding probably takes place once each year, after sexual maturity is reached. The number of generations per hundred years would thus be markedly less than in a small species, such as in the genus Peromyscus, or Mus , where the birth rate is high and the life span short. It would seem that a large species with a long life span and few generations per hundred years, would require a longer period of time to show changes of a subspecific nature than would a species with a short life span and many generations per hundred years.

On the other hand, the stock has been subjected to isolation in a fresh- water lake. Life here must be markedly different from the normal ex- istence with free access to the ocean. The food, too, must be quite dif- ferent. This changed environment would require many adjustments, which would tend to encourage the development of new characters in the species more rapidly than if the environment had remained the same.

To summarize, then, it would seem that we have a large species, with few generations per century; but it has been thrust into a new environ- ment which would cause it to change more rapidly than would other- wise be the case. Thus, these two factors tend to cancel one another. The estimates for the time of isolation range from 8,000 years to 3,000 years, with 5,500 years as a mean. Thus it seems probable that about 5,500 years, plus or minus 2,500 years, has been the time required to produce a new subspecies under the conditions set forth above.5

6 Using the estimates of 7,000 to 8,000 years since the glacial ice left this region, as suggested in Dr. Antevs’ letter of December 16, 1941, these figures be- come 6,000 to 2,000, with an average of 4,000 years.

1942

Doutt : Review of the Genus Phoca

79

Taxonomy

The number of names which have been proposed for the members of the family Phocidae is most perplexing, but in addition to the great num- ber of synonyms, technical names have been switched from one species to another, so that at times it is impossible to know which animal is under discussion. In reviewing the technical history of the Phocidae, Allen (1880, p. 459), remarks upon the number of synonyms as follows: “One hundred and three distinct specific and varietal names have thus been bestowed upon sixteen species, leaving eighty-seven of the names as synonyms, — an average of about six to a species.” The English common names, also, have been misapplied; but, strangely enough, it seems that the Eskimo names, wherever they have been used, have been rather con- sistently applied to the proper species. Wherever the Eskimo name is given in a discussion, therefore, it is usually helpful in allocating the ma- terial to the proper species. Fortunately, too, the Eskimo names, unlike Indian names, are very similar from Greenland to Alaska. In some parts of Alaska, however, the language is quite different, and different names are in use. Throughout the eastern arctic the Eskimo name “Kasaguea” refers to Phoca vitulina, “Netcheck” to Phoca hispida, “Kiolik” or “Kioole” to Phoca groenlandica , and “Ootroo” or “Oogjook” to Erignathus barbatus. Of course there is considerable variation in the spelling of these names. This is due partly to the difference in the local Eskimo dialects, and partly to the way different authors hear and record the same sounds. In general, however, there is sufficient similarity between the various spellings to make the different names recognizable.

The family Phocidae, at the present time, contains eleven genera. Of these, only four, Phoca , Erignathus , Halichoerus, and Cystophora , are known to occur in the North Atlantic, North Pacific or Arctic oceans. In the northern hemisphere, the genus Phoca can be divided into four dis- tinct groups, as follows:

1. Phoca vitulina and related races

2. Phoca hispida and related races

3. Phoca groenlandica

4. Phoca fasciata

Various authors have treated these groups in different ways. Some have considered them to be distinct genera, others have treated them as subgenera, and still others have considered them as species. Such treat- ment, of course, is largely a matter of personal opinion, and depends upon

80

Annals of the Carnegie Museum

vol. XXIX

the author’s approach to the subject. From the present study, however, I can see no reason for elevating these groups above specific rank. Together they form a closely related group, which is very distinct from Halichoerus and Cystophora. Erignathus, however, is more closely related to the genus Phoca than it is to either of these two genera. Its teeth and skull char- acters are not markedly different from other members of the genus Phoca , yet, in general, it is more aberrant than any of the other four species which I have included in the genus. For the present, therefore, I prefer to let it stand as a separate genus. Each of these four species can be recog- nized readily by distinctive skull characters.

Key to the Species of the Genus Phoca

A. Posterior margin of palate notched or incised, resembling a pointed Gothic arch.

a. Greatest length of skull, more than 178 mm.; length of upper second pre-

molar, 6.8 mm. or more; mandibular teeth crowded out of line, and over- lapping; least interorbital width 7 mm. or more P. vitulina

b. Greatest length of skull, 178 mm. or less; upper second premolar, less than

6.8; mandibular teeth not crowded out of line; least interorbital width less than 7 mm P. hispida

B. Posterior margin of palate not notched or incised, resembling a rounded Roman

arch (Plates IX, X).

a. Palatal length,6 more than 86 mm.; posterior palatine foramina in, or an- terior to, maxillo-palatine suture; range, North Atlantic and Arctic

oceans7 P. groenlandica

b. Palatal length, 86 mm. or less; posterior palatine foramina in, or posterior to, the maxillo-palatine suture; range, North Pacific ocean P. fasciata

6 Measured from most anterior part of rostrum to midline at posterior edge of palate.

7 Although Allen (1880, pp. 640-641) states that Phoca groenlandica is circum- polar in its distribution and mentions records from the Pacific, he states that its distribution there is not well known. Furthermore, he had no specimens from there, and his authority for including the Pacific in its distribution was derived from the writings of early explorers, such as Pallas, Steller and Temminck. It is quite possible these authors were referring to some other species. I have seen no specimens from the Pacific. Dr. R. M. Anderson, in a letter dated November 21, 1941, has sent me the following comments concerning its occurrence in waters adjacent to the Pacific: “The only Canadian record that I have from western Arctic waters is a good photograph of a specimen which was caught in a fishnet at Aklavik in 1926. Mr. A. E. Porsild, who spent some time at Aklavik in the west branch of the Mackenzie delta, N.W.T., while engaged in reindeer introduction, sent me a film which was taken at the time by the Rev. Father Trocellier, O.M.I., and I had a print made from it. Mr. Porsild saw part of the skin. The prints plainly show the characteristic pattern of Phoca groenlandica, a broad diagonal

1942

Doutt: Review of the Genus Phoca

81

This key will serve to distinguish all normal specimens, but in any large series certain specimens may be found which are so aberrant that they will not fit into any classification. For example, the rounded arch of the posterior margin of the palate is characteristic of all Phoca groen- landica which I have examined, except one specimen, Carnegie Museum no. 18714. In this specimen the palate is notched, very much as in Phoca hispida or Phoca vitulina. In other respects, however, it is typical of Phoca groenlandica. Another specimen, The Academy of Natural Sciences of Philadelphia no. 2139, is a typical Phoca hispida in most of its char- acters; yet, regardless of the fact that it is quite young, its least inter- orbital width is 8 mm., which is 14.3 per cent greater than the largest specimen I have seen in a series of more than seventy. Its teeth, too, are unusually heavy for a Phoca hispida. Another specimen, Carnegie Mu- seum no. 17852, has the teeth, palate, and interorbital width of Phoca hispida , but in length of skull (193 mm.), width across mastoids (117.5 mm.), and general massiveness, it resembles Phoca groenlandica. Is it possible that these species interbreed on rare occasions?

Except for anomalous specimens such as those mentioned above, the

black area on the side, extending from middle of shoulders on sides and upward again to base of tail. I never met the species during seven years spent along the western Arctic coast in Alaska, Canada, and Northwest Territories, and the Eskimos did not know the animal, so it is evidently only of accidental or casual occurrence.” The above mentioned photograph is reproduced in this paper as figure 5 of Plate XIII.

While this paper was in galley proof I received another letter from Dr. Ander- son, dated Feb. 20, 1942, which follows:

“I am enclosing a copy of another record of Greenland Seal for the Western Arctic district, sent in through the Royal Canadian Mounted Police. The In- spector at Fort Smith thought it might be of interest to the National Museum. I think it is a good record as many of the men in R. C. M. P., and Hudson’s Bay Co., in western Arctic have also served in the eastern Arctic and know the seals pretty well.

“ ‘Re: Seal, Cambridge Bay District.

1. Native Ehakataitok brought the hide of a Greenland seal into the Settle- ment on the 10th instant. I have questioned a few of the natives in this district and they all admit that this is the first time that they have seen this species of seal. This animal was killed off the north end of Melbourne Island [a little east of base of Kent Peninsula].

‘Sgd.’ D. C. Martin A/Cpl.’

“Forwarded through the office of Commissioner, R. C. M. P., Ottawa, 1942, and copy transmitted to National Museum of Canada, Feb. 18, 1942.”

82

Annals of the Carnegie Museum

vol. XXIX

four species of Phoca can be distinguished by skull characters without great difficulty.

Descriptions of Species

Phoca vitulina

(Plates V, VII, IX; text figure 5)

External Characters: In color and markings this species is exceedingly variable, ranging from almost black with a few white spots, to almost white with a few black spots. This wide range of variation has been described so well by Allen (1880, pp. 562-565) that I take the liberty of quoting it here.

“Color variable. Above, usually yellowish-gray, varied with irregular spots of dark brown or black; beneath, yellowish-white, usually with smaller spots of dark-brown. Sometimes uniform brownish-yellow above, and somewhat paler below, entirely without spots; or uniform dark-gray above, and pale yellowish-white below, everywhere unspotted. Not infrequently everywhere dark-brown or blackish, varied with irregular streaks and small spots of yellowish-brown; the head wholly blackish from the nose to beyond the eyes; the lips and around the eyes rusty-yellow. Length of male, 5 to 6 feet; of female, somewhat less. Young at birth uniform soiled-white or yellowish-white, changing to darker with the first moult.

“The variations in color are almost endless, ranging from uniform yellowish-brown to almost uniform dark-brown, and even nearly black, with, between these extremes, almost every possible variation, from dark spotting on a light ground to light spotting on a dark ground. The mark- ings vary in size from very small spots to large, irregular patches and streaks. The more common color is brownish-yellow, varied with spots and patches of darker, but not unfrequently the general color is blackish, more or less varied with spots, patches and streaks of lighter. The lower surface is generally thickly marked with small oval or roundish spots, smaller and less confluent than those of the upper surface. Specimens from Denmark and the Atlantic coast of North America are indistinguish- able from those of Lower California, Washington Territory, and Alaska. Specimens from the Pacific coast present the same wide range of color- variations, and precisely the same phases as those from the shores of the Atlantic. . . .

“Unlike the Phoca foetida, P. groenlandica , and most other Phocids

1942

Doutt : Review of the Genus Phoca

83

of the northern waters, the first coat is shed before or soon after birth, but as to the exact time at which it is cast authorities disagree. Mr. Bartlett, in describing a young Seal of this species (wrongly identified at the time as Phoca foetida), born in the Garden of the London Zoological Society, June 8, 1868, says: ‘It was born near the edge of the water, and in a few minutes after its birth, by rolling and turning about, was com- pletely divested of the outer covering oifur and hair , which formed a com- plete mat, upon which the young animal lay for the hour or two after its birth.’

“It is sometimes stated that the foetal coat is retained for four or five days after birth, but other writers affirm that it is shed at the time of birth.”

Allen’s descriptions apparently were taken from preserved specimens, and he has been misled by the oxidation which takes place in these skins when they are allowed to dry with grease on them. Under these conditions the normally white or silvery white colors change to yellowish or brown- ish, so that in many places where he has described the color as yellowish or brownish, the natural color was probably whitish or silvery white.

On the Belcher Islands I examined a young seal of this species which was apparently only a few hours old. Part of the navel cord was still attached, but it was not yet dry. There was no evidence of the white foetal coat.

Table I

Skull Measurements of Phoca vitulina (Text figures 9-11)

Maximum

Minimum

Average

Total length

221.5 mm.8

157.3

184.1

Width across mastoids

128.0

100.0

111.6

Interorbital width

16.0

9.5

11.6

Length of nasals

67.3

33.4

45.7

Width of nasals at tip

21.0

9.4

14.2

The nasals are broad and taper gradually from the tip to the maxillo- frontal suture, but from here back they narrow rapidly. For the anterior half of their length they lie between the maxillaries and pre-maxillaries. The posterior half of their length lies between the frontals. Posteriorly the palate ends acutely like a pointed Gothic arch. The posterior palatine

8 The measurements used throughout this paper are expressed in millimeters unless otherwise indicated.

84

Annals of the Carnegie Museum

vol. XXIX

foramina, with but a few exceptions, lie anterior to the maxillo-palatine suture. The maxillary teeth are large, being both wide and long. The third upper molariform tooth is usually the largest. It has a large central cusp, and frequently two accessory posterior cusps, although in some instances only one posterior accessory cusp is present. Usually there is one small anterior cusp as well. The anterior end of each ramus of the mandible is heavy and blunt and curves gradually to meet the lower border of the ramus. The coronoid process is short and broad, and usually does not extend backwards beyond the glenoid process. The mandibular teeth, like the maxillary teeth, are large and heavy. The third molariform tooth is the largest. It has one large central cusp, two small posterior cusps and one small anterior cusp. One or two small tubercles are fre- quently found near the base of this small anterior cusp. These teeth are frequently set diagonally in' the jaw, so that the posterior portion of the one tooth overlaps the anterior portion of the one behind it. This is particularly true of young specimens in which the jaw has not attained its full growth. As age advances the jaw grows; thus the teeth receive more space, and gradually come into line with the jaw.

Range : Along the Atlantic coast of North America the species is most abundant from Maine to Labrador. It has been reported from as far south as North Carolina (Allen, 1880, p. 585), and as far north as Ellesmere Island (Anderson, 1934, p. 75), but is rare, or uncommon, at the extremes of its range. On the Pacific coast of North America it is common from California to Alaska, ranging from Lower California (Allen, 1902, p. 495) to Pt. Barrow, Alaska (Allen, 1902, p. 484). Along the Pacific coast of Asia it has been reported from Bering Strait (Allen, 1902, p. 485) south to the mouth of the Yangtze River (G. M. Allen, 1938, p. 493). A specimen has recently been described from Chefoo on the Shantung coast of China by Leroy (1940).

On the European side of the Atlantic it has been reported as occurring occasionally in the Mediterranean. From France northward to Scandi- navia it is the commonest species of the family. It ranges from here north- ward and eastward along the Arctic coast, but apparently does not reach Spitsbergen and Jan Mayen Islands (Allen, 1880, pp. 586-587). Accord- ing to Smirnov (1908, pp. 69-70), it reaches Upernivik in northern Green- land and eastward, Novaya Zemlya, but is not known east of there. It frequently travels inland up fresh-water rivers and lakes and has been reported from Lake Ontario and Lake Champlain, and from the Columbia River near the Dalles, above the Cascades and approximately 200 miles

1942

Doutt : Review of the Genus Phoca

85

from the sea (Allen, 1880, pp. 587-588). In Hudson Bay it is not common along the east coast, but is occasionally seen as far south as Great Whale River and the Belcher Islands. Here it is known to the Eskimos as “Kasaguea.”

Phoca hispida

(Plates V, VII, IX; text figure 4)

External characters: The color markings of Phoca hispida are so similar to those of Phoca vitulina that I could not be certain of distinguishing every specimen by color alone. However, there is a difference in the tex- ture of the hair, which, combined with general differences in color, makes it possible to distinguish practically every specimen. The wide range of variation makes description very difficult, however. In general, the skins are white, silvery white, or yellowish white on the belly. In preserved specimens, where the grease has not been removed completely, the belly is usually yellowish or brownish. The hair around the lips and sometimes the sides of the head, the axilla of the front flippers, the underside of the tail and the inside of the hind flippers is usually a salmon or light brownish color (“Tawny” or “Russet” of Ridgway, 1912). This is particularly noticeable in fresh specimens, but becomes more or less obliterated if the specimens have become yellow from age and grease. The back is spotted, streaked or marbled with black. White spots with dark centers are char- acteristic, but not always present. The dark markings may be confined to the mid-dorsal region, or may extend down on the sides to the belly. These dark markings are usually black or some shade of dark brown, and usually begin on the top of the head between the eyes and extend back- wards to, and include the tail. The face, from the nose to the eyes, is usually light. The hair is coarse and stiff and usually points directly backward.

The coat of the newly born young is quite different, being soft and woolly and of a white or yellowish white color. According to Allen (1880, p. 600), “At the age of about four weeks this gradually gives place to the coarser, more rigid pelage of the adult, and the color changes to dusky, marked sparsely with small blackish spots. Yearlings are often yellow- ish-white; dusky along the middle of the back, with here and there small spots of blackish.

“There is a wide range of individual variation in color, in the newly- born young as well as in the adults. . .”

86

Annals of the Carnegie Museum

vol. XXIX

In addition to this wide range of color, Kumlien (1879, p. 60) mentions hairless and albino specimens. Degerb^l and Freuchen (1935, p. 196) mention variation in total length of from one to two meters. This dif- ference is attributed to the amount of food available.

Table II

Skull Measurements of Phoca hispida

Maximum

Minimum

Average

Total length

177.0

155.5

165.8

Width across mastoids

105.8

90.4

100.05

Interorbital width

6.2

4.0

5.04

Length of nasals

42.5

30.0

37.33

Width of nasals at tip

12.1

9.5

11.26

The nasals are long and narrow and taper gradually from the tip to the maxillo-frontal suture; at this point they narrow rapidly. The portion of the nasals lying posterior to the maxillo-frontal suture is usually con- siderably less than one-half the length of the nasals. The posterior margin of the palate ends acutely in a pointed Gothic arch. The posterior pala- tine foramina lie in or posterior to the maxillo-palatine suture. The mo- lariform teeth are small, the third one being the largest. It has a large central cusp and may have two small posterior cusps, though usually only one is present. It also has one small anterior cusp. There is seldom any evidence of the small tubercles which are frequently seen in Phoca vitulina. Although possessing nearly the same number of cusps as Phoca vitulina, the teeth of Phoca hispida are readily distinguished by their much smaller size. The anterior end of each ramus of the mandible is narrow and pointed and slopes backward in a nearly straight line, to the lower margin of the ramus. The coronoid is long, slender and pointed, and ex- tends much further backward than it does in Phoca vitulina. As in the maxillary teeth, the third mandibular tooth is the largest. It has one large central cusp, two smaller posterior cusps, and one small anterior cusp. These teeth are seldom crowded out of line as they are in Phoca vitulina , but occasionally the second molar tooth may be set diagonally in the jaw.

Range: The species is reported to be circumpolar in its distribution. It has been found as far north as 82°40' and ranges southward on the Atlantic coast of North America to Labrador (Allen, 1880, p. 615). On the Pacific coast of North America it has been reported as far south as St. Michaels, Alaska (Allen, 1902, p. 477), but Osgood (1904) did not mention it in his “Reconnaissance of the Base of the Alaska Peninsula.” On the

1942

Doutt : Review of the Genus Phoca

87

Atlantic coast of Europe, it has been reported as far south as the British Channel (Allen, 1880, p. 615) and along the Pacific coast of Asia, it has been reported from the Okhotsk Sea (Allen, 1902, p. 480). Smirnov (1908, p. 57) states that its southern limit here is the Amur River. He says that it is found in the Bering Sea, White Sea, and Baltic Sea. Degerb0l and Freuchen (1935, p. 46) record specimens from King William Land. It shares part of its range with Phoca vitulina, but is a more northern species, being found commonly much farther north, and never ranging as far south as that species.

In Hudson Bay it is the most common seal along the east coast, and at the Belcher Islands. The Eskimos here call it “Netcheck.” I have speci- mens from as far south as the Twin Islands, and on September 24, 1935, I saw a seal which I took to be this species at the mouth of the Moose River.

Phoca groenlandica

(Plates VI, VIII, X; text figure 6)

External characters : The small number of skins which have been avail- able to me have not been sufficient to permit a description of this species, therefore I have borrowed the following description from Robert Brown (1868, pp. 416-420).

“It seems to be almost unknown to most writers on this group that the male and female of the Saddlebacks are of different colours; this, how- ever, has long been known to the Seal-hunters. Male. — The length of the male Saddleback rarely reaches 6 feet, and the most common length is 5 feet; while the female in general rarely attains that length. The colour of the male is of a tawny grey, of a lighter or darker shade in different individuals, on a slightly straw-coloured or tawny-yellowish ground, hav- ing sometimes a tendency to a reddish-brown tint, which latter colour is often seen in both males and females, but especially in the latter, in oval spots on the dorsal aspect. The pectoral and abdominal regions have a dingy or tarnished silvery hue, and are not white as generally described. But the chief characteristic, at least that which has attracted the most notice, so much as to have been the reason for giving it several names, from the peculiar appearance it was thought to present ( e.g ., ‘harp’ Seal, ‘saddleback,’ etc.), is the dark marking or band on its dorsal and lateral aspects. This ‘saddle-shaped’ band commences at the root of the neck posteriorly, and curves downwards and backwards at each side superior

88

Annals of the Carnegie Museum

vol. XXIX

to the anterior flippers,* reaches downwards to the abdominal region, whence it curves backwards anteriorly to the posterior flippers, where it gradually disappears, reaching further in some individuals than in others. In some this band is broader than in others and more clearly impressed, while in many the markings only present an approximation, in the form of an aggregation of spots more or less isolated. The grey colour verges into a dark hue, almost a black tint, on the muzzle and flippers; but I have never seen it white on the forehead as mentioned by Fabricius. The muzzle is more prominent than in any other northern Seal.

“ Female. — The female is very different in appearance from the male: she is not nearly so large, rarely reaching 5 feet in length; and when fully mature her colour is a dull white or yellowish straw-colour, of a tawny hue on the back, but similar to the male on the pectoral and abdominal regions, only perhaps somewhat lighter. In some females I have seen the colour totally different; it presented a bluish or dark grey appearance on the back, with peculiar oval markings of a dark colour ap- parently impressed on a yellowish or reddish-brown ground. These spots are more or less numerous in different individuals. Some Seal-hunters are inclined to think this is a different species of Seal from the Saddleback, because the appearance of the skin is often so very different and so ex- tremely beautiful when taken out of the water; yet as the females are always among the immense flocks of the Saddleback, and as hardly two of the latter females are' alike, but varying in all stages to the mature female, and on account of there being no males to mate with them, I am inclined to believe with Dr. Wallace that these are only younger female Saddlebacks. The muzzle and flippers of the female present the same dark- chestnut appearance as in the male. . .

“(a) The colour after birth is a pure woolly white, which gradually assumes a beautiful yellowish tint when contrasted with the stainless purity of the Arctic snow; they are then called by the sealers ‘white-coats’ or ‘whitey-coats’ f ; and they retain this colour until they are able to take

*1 use this very convenient sealers’ vernacular term to express the ‘paws,’ ‘hands,’ etc. of systematic authors.

fThese are rarely seen in Danish Greenland, and then are called ‘Isblink’ by the Danes from their colour; at least, so Fabricius says. He, moreover, informs us that the third year they are called Aglektok (as mentioned above), the fourth Millaktok, and after a winter Kinaglit, when they are beginning to assume the harp-shaped markings of the male (Nat. Selsk. Skrift., i, p. 92). I never heard these names in North Greenland.

1942

Doutt : Review of the Genus Phoca

89

the water (when about fourteen or twenty days old). . . The white-coat changes very quickly. In 1862 the late Capt. George Deuchars, to whom science is indebted for so many specimens, brought me two alive from near Jan Mayen; they were white when brought on board, but they changed this coat to a dark one completely on the passage, of a week or ten days.

“I consider that about three years are sufficient to complete these changes. This is also the opinion held in Newfoundland, though the Greenland people consider that five years are necessary. I wish, however, to say that these changes do not proceed so regularly as is usually de- scribed, some of them not lasting a year, others longer, while, again, several of the changes are gone through in one year; in fact the coats are always gradually changing, though some of the more prominent ones may be re- tained a longer, and others a shorter time. It would require a very care- ful and extended study of this animal to decide on this point, which, owing to their migrations, it is impossible to give. After all, these changes and their rapidity vary according to the season and the individual, and really will not admit of other than a general description.”

Table III

Skull Measurements of Phoca groenlandica

Maximum

Minimum

Average

Total length

221.0

190.0

204.7

Width across mastoids

123.6

107.5

114.6

Interorbital width

20.0

8.3

11.7

Length of nasals

52.6

35.8

42.1

Width of nasals at tip

19.2

14.9

16.6

The nasals are long and narrow and in some respects resemble those of Phoca hispida more than those of Phoca vitulina. Except in very young specimens, the nasal sutures are frequently so ossified that the margins are not distinctly visible. The nasals taper rather gradually from the anterior to the posterior end. Thus the abrupt change observed in Phoca vitulina and Phoca hispida at the junction of the maxillo-frontal sutures is not so evident. The posterior margin of the palate is round like a Roman arch, and is seldom notched or incised like that of Phoca vitulina and of Phoca hispida. The posterior palatine foramina usually lie in or anterior to the maxillo-palatine suture. The teeth are comparatively small. The third

90

Annals of the Carnegie Museum

vol. XXIX

molariform tooth is usually the largest. It has one large central cusp and one small posterior cusp, and occasionally there is evidence of a still smaller second posterior cusp. Occasionally, also, there is a very small anterior cusp. The anterior end of each ramus of the mandible is narrow and pointed and slopes backward in a nearly straight line to the lower margin of the ramus. In this respect it is very similar to Phoca hispida, but the angle made by this sloping portion with the base of the ramus is much more acute than in Phoca hispida (Pis. VII, 1; VIII, 1). The coronoid is long, slender and pointed. It is prolonged backward much as in Phoca hispida . The mandibular teeth are small, the third being the largest of the set.. It has one large central cusp, two small posterior cusps and one small anterior cusp. In number of cusps they resemble Phoca hispida , but in groenlandica the Central cusp is proportionately larger. The teeth are not crowded in the jaw, but are spaced even farther apart than in Phoca hispida.

Range: The home of this species is the North Atlantic. It is found in great numbers off the coast of Newfoundland and Labrador, especially in the early spring. It has been reported as far south on the American coast as New Jersey (Allen, 1880, p. 640), although its occurrence there is rare. It has been reported as far north as Annanactook at about lat. 67°N., long. 68°50'W., Kumlien (1879, p. 61). Sverdrup (1904, vol. II, p. 40) reports them from Jones Sound, about lat. 76°N., long. 85°W. Along the Atlantic coast of Europe they have been reported from as far south as “Morecombe Bay, England.”9 Northward they are common about “Ice- land,” “Jan Mayen,” “Spitzbergen,” and also occur about “Nova Zembla,” “Franz Josef Land” and the “Kora Sea” (Allen, 1880, p. 641). Plehanoff (1933) found them on Marjovez Island in the White Sea. They enter Hudson Bay and are found at least as far south as Great Whale River and the Belcher Islands. The Eskimos here know them by the name of “Kioole.”

Allen (1880, p. 640) says the species is circumpolar, but later (p. 641) he says that the distribution of the species in the North Pacific is not well known. The only authorities he gives for its occurrence there are Pallas and Temminck. He says he saw no specimens from there. I have seen no specimens from the Pacific either, and am inclined to believe that the species does not occur there.

* I have used quotes here to show that I am copying these place names di- rectly from Allen (1880). Different names and spellings are now employed for some of these places.

1942

Doutt : Review of the Genus Phoca

91

Phoca fasciata

(Plates VI, VIII, X; text figure 7)

External Characters: The males of this species can be recognized at once by the unique color pattern. It may be considered as a black seal with a white band around the neck, around each foreleg, and around the body just anterior to the junction of the hind flippers. Since I have seen only two skins of this species I am unable to describe the range of variation or the color of the female or young, so I have borrowed the following de- scription from Allen (1880, pp. 676-678), who, for the most part, was quoting von Schrenck.

“Adult male. General color, dark brown. A narrow yellowish-white band surrounds the neck extending forward to the middle of the head above; another broader yellowish-white band encircles the hinder portion of the body, from which a branch runs forward on each side to the shoulder, the two branches becoming confluent on the median line of the body below, but widely separated above. In other words, the (1) front part of the head, the (2) hind limbs, and the posterior fourth of the body, the (3) top of the neck and the whole anterior half of the back, as well as (4) the fore- limbs and a considerable area at their point of insertion, are dark brown; these four regions being separated by bands of yellowish-white, of variable breadth over different regions of the body. The brown of the anterior part of the dorsal region also extends laterally in the form of a narrow band around the lower part of the neck, where it expands to form a small shield- like spot on the breast. There are also very small spots of brown on the posterior part of the abdominal region.

“Adult female. — Uniform pale grayish-yellow or grayish-brown, with the exception of an obscure narrow transverse whitish band across the lower portion of the back. The extremities and the back are darker, with a faint indication of the dark ‘saddle’-mark seen in the male.

“ Young. — The young of both sexes are said to resemble the adult female.

“Von Schrenck’s detailed description, on which the foregoing is mainly based, is substantially as follows: The dark-brown of the head, in the male, is followed by a broad dusky yellowish-gray neck-band, which on the mid- dle line, both above and below, passes forward, but on the sides has the convexity pointing backward. Behind this light neck-band is a broad, long saddle-shaped patch upon the back, which, on the middle line, runs forward in a point, but which extends itself laterally in two narrow bands

92

Annals of the Carnegie Museum

vol. XXIX

meeting and expanding on the breast into a pointed spot; posteriorly the dark dorsal patch is also prolonged backward and laterally, but without meeting below. Along the sides of this dorsal area runs a broad, curved, light, soiled yellowish-gray band, with the convexity upward; these lateral light bands become deflected downward, both anteriorly and posteriorly, and form, by their union, a light band along the belly. Within these light bands anteriorly, on each side, is a large oval dark-brown spot, in which are inserted the anterior extremities. The light ventral area encloses posteriorly two small oval dark-brown spots, and in front of these a third narrower and larger. Behind the dark area on the back is a very broad dorsal cross-band of light yellowish-gray, joining the light bands on the side of the body. Behind this light cross-band the whole posterior part of the body, as well as on the tail and hind limbs, is blackish-brown. As a rule the above-described dark and light color areas are very sharply defined. Sometimes, however, there extends from the dark areas a smaller spot more or less isolated. According to the same writer the color varies considerably in different individuals, one of those he describes having the dark color of a dark grayish-black, and the light markings whitish or straw-yellow. He also states that in the figures given by Siemaschko the light neck-band is deflected backward from the back of the neck to the fore-limbs, leaving the whole breast of the same dark-brown color as the head. Besides this the dark-brown color of the back extends, both posteriorly and anteriorly, to the lower sides of the body, occupying the whole of the ventral surface, with the exception of two light bands which run crosswise around the base of the anterior extremities, and a separate light band that crosses the hinder part of the body. In consequence of the wide departure of the pattern of coloration in Siemaschko’s figure from his own examples, von Schrenck is left in doubt as to whether the figure is really a true copy from nature.

“The single specimen I have examined . (Nat. Mus. No. 9311, Cape Romanzoff, W. H. Dali), a flat skin, lacking the flippers and the facial portion, agrees with von Schrenck’s figure in respect to the form and size of the neck-band, but there is a far greater preponderance of light color, which occupies rather more than half the entire surface. Only the pos- terior sixth of the body is black, and the dark area of the back is very much more restricted, and differs somewhat in outline. In this specimen the breadth of the dark dorsal portion occupies scarcely more than one- third of the whole width of the skin, the light portion on either side nearly equalling it in breadth. It widens over the neck and sends down a lateral

1942

Doutt : Review of the Genus Phoca

93

branch on each side, the two meeting on the breast. It is contracted over the shoulders, behind which it again expands, and at its posterior border sends down a very narrow branch from the right side to the middle of the belly ; its fellow on the opposite side is nearly obsolete, forming merely a broken chain of small dusky spots. There is hence in this example a wide departure from the specimens described by von Schrenck, while the want of symmetry in the two posterior branches of the dorsal spot, and the relatively nearly equal amount of light and dark color, lead one to ap- prehend a much wider range of individual variation in coloration than von Schrenck apparently suspected, and that after all Siemaschko’s figure merely represents a variation in the opposite direction from that here indicated, or an unusual extension of the dark color at the expense of the lighter markings.

“Size. — Von Schrenck states that this animal is reported to sometimes attain the length of 634 feet. He gives the length of a full-grown male as 5 feet, 634 inches (1683mm.), and that of a full-grown female as 5 feet, 3 inches (1600mm.), based on Wosnessenski’s specimens obtained in Kamtschatka, which his hunters informed him were not of the largest size. In other words, it appears to be a Seal of the medium size, or about as large as Phoca groenlandica .”

Table IV

Skull Measurements of Phoca fasciata

Maximum

Minimum

Average

Total length

201.3

193.0

197.0

Width across mastoids

133.4

122.0

127.7

Interorbital width

14.6

9.8

12.2

Length of nasals

43.3

42.2

42.75

Width of nasals at tip

10.0

9.8

9.9

Only two adult specimens were available for the above measurements. The nasals are long and narrow and taper gradually from the anterior to the posterior end. Like Phoca groenlandica there is little change in con- tour at the maxillo-frontal suture. The palate is broad and the posterior