animals there to disturb them. Both Mr. Farley and Mr. Prior commented on the large number killed by hunters and dogs, and stated that some of the negroes consider the birds quite a palatable delicacy and gather numbers for the pot.

The Harbor Island Rails face extinction at this very time. A railroad has been built to the island from the mainland and improvements are being made. In addition to this the birds have been almost wiped out by collecting.

Probably the greatest factor the Rails have to contend with in their fight for existence is the flooding of the marshes, both from high tide and from heavy rains. At such times the birds are much exposed and bewildered and many drown. In the seaside marshes they build their nests on the banks of the sloughs or bayous instead of the higher parts of the marsh, and in rainy spells numbers of nests are destroyed.

They are naturally very delicate birds and sensitive to the cold of the more severe winters; many freeze to death where they are unable to secure shelter.

During November and December 1913 Texas was visited by one of the most destructive floods of its history, two of the largest rivers of the state rising and overflowing miles and miles of the lowlands towards the coast. During that time numbers of the Rails left the marshes and took to higher ground until the waters receded. One of these birds was caught in a bewildered condition in Mr. Farley's yard in Port Aransas in October.

XI. WINTER.

Most of the winter habits of the birds have been given in the two preceding divisions of this paper.

In the colder weather they haunt drifts of logs or trash in the marshes, where they take shelter from the cold north wind and from rains. But as the thermometer rarely falls below freezing in this semitropical coast country, the birds are rarely forced to seek shelter, and their actions and habits then are not noticeably different from other times of the year.

In winter in the marshes on Bolivar Peninsula Mr. Pope says

that the birds were fully as common in winter as in summer, if not more so. In travelling through the grass the birds had wellbeaten paths about six inches wide, and from the way these paths were beaten out in the vicinity of the bayous, it would appear that the birds were much more common.

Marshes inhabited by Rails are always intersected by numerous well-worn pathways leading through the tall grass, the grass being pushed to either side and the "marsh hair" being trampled and beaten down into the mud. These pathways are usually about five inches wide at the bottom.

XII. DISCUSSION.

The Louisiana Clapper Rail is fairly common and resident from Louisiana to Corpus Christi Bay. It is apparent that the Caribbean Clapper Rail rarely, if ever, appears in Texas.

The only records of the latter in the United States are the two specimens from Texas in 1889 (Sennett) and two specimens mentioned in Mr. Singley's 1893 paper. Undoubtedly Mr. Singley merely followed the lead of Mr. Sennett, as stated before. So the reader can see that 211.2 Rallus longirostris caribæus is included in the A. O. U. Check-List of North American Birds on the record of Mr. Sennett's two specimens from Texas, the identification of which he himself later questioned and virtually changed. Would it not seem best to drop this bird from the Check-List and extend the range of Rallus crepitans saturatus to include Texas?

a

A NEW SUBSPECIES OF DENDRAGAPUS, (DENDRAGAPUS OBSCURUS FLEMINGI) FROM SOUTHERN

YUKON TERRITORY.1

BY P. A. TAVERNER.

In the examination of a collection of Dendragapus obscurus recently acquired by the Canadian Geological Survey from the Teslin Lake region, on the Yukon-British Columbia boundary, it became evident that a new form of the species was probably represented. Two subspecies have heretofore been recognized as ranging into Canadian territory, D. o. fuliginosus and richardsoni. Fuliginosus is a dark bird with a light tail band and richardsoni a lighter one without a terminal band on the tail. The birds under consideration are without the tail band but are even darker than fuliginosus.

We had a fair series of both recognized forms from southern British Columbia, but all in spring and early summer plumage and hence hardly comparable with the new form which is represented entirely by late summer and autumn birds. The possibility existed that the darkness of the new birds but represented the new and unfaded plumage of the post-nuptial moult; therefore application was made to the U. S. Biological Survey and its chief Dr, Henshaw was courteous enough to place at our disposal what summer and fall specimens there were in the collection of that institution. A comparison of the specimens, including these, confirms our first surmise as to the racial differentiation of the forms and I therefore present the following new subspecies for consideration.

Dendragapus obscurus flemingi new subspecies.

FLEMING'S GROUSE.

Named after Mr. J. H. Fleming, C. M. Z. S. in recognition of the value of his ornithological work in Canada.

Subspecific Characters. Like Dendragapus obscurus richardsoni, without terminal tail band, but darker in general coloration even than D. o. fuliginosus.

1 Published by permission of the Deputy Minister, Department of Mines.

Description of Type specimen. Type, adult male, No. 6651 Museum of the Geological Survey, Department of Mines, Canada. Near Teslin Lake, Yukon Territory, October 10, 1912. Clement Lewis, Collector.

Back, rump and tail, dusky neutral gray. Central feathers of tail and some lateral ones double rounded at ends.

Upper tail coverts, fuscous black, finely but sparsely vermiculated with buffish slate.2

2

Scapulars, wings and upper wing coverts fuscous, lightening towards wing tip. All, except flight feathers, finely marked with shades between rusty and grayish. Some scapulars with white shaft streaks, widening at terminal to poorly defined marginal spots. The latter persisting on some of the lesser coverts as vague spots and narrow borders.

Below, between neutral gray and deep neutral gray, and obscured posteriorly by a white tipping of the feather ends that blends into the white of the basal under tail coverts.

Lower breast feathers with white shaft lines broadening towards the flanks and posteriorly where they develop into broad terminal spots, streaks and bars. A fine vermiculation of buffy2 and grayish overlaying the body color of the feather tips of the sides.

1 All color terms except where otherwise stated, from Ridgway's "Color Standards and Nomenclature," 1912 ed.

These terms are not from "Ridgway's Color Standards."

Under tail coverts dusky neutral gray or black with prominent white shaft lines and final borders. White increasing towards base of tail until

body color of feathers is hidden.

Head from deep neutral gray to neutral gray darkest on crown and ear coverts, overwashed with rusty1 strongest on forehead at base of bill and more or less obscuring body colors of darker feathers and at last dying away at back and sides of upper neck as a faint olivaceous1 tinge.

Throat patch, white with irregular deep neutral gray spotting, aggregated on malar region and each spotted feather terminated with small white border.

Lores, deep neutral gray, with white flecking.

Dark feathers of the breast band creamy white basally, the white gradually encroaching on the dark tips towards the sides, finally usurping most of the feather and making a semi-concealed white streak down the sides of the neck to the bend of the folded wing.

The characters of the female are less marked than of the male and without series for comparison may be difficult of recognition. However, they average in the same directions as the male being bluer underneath than richardsoni and darker dorsally, with the rufous1 or rusty markings bolder and more decided in character.

Measurements (in inches). Male, average of four specimens.

Wing, 8.25-8.75, (average 8.51); tail, 6.5-6.75 (average 6.33); tarsus, 1.70-1.80 (average 1.75); exposed culmen, .65–.75 (average .71).

Female average of twelve specimens.

Wing, 7.45-8.10 (average 7.70); tail, 5.10-5.75 (average 5.25); tarsal 1.48-1.68 (average 1.55); exposed culmen, .65-.71 (average .69).

Distribution. This description is based entirely on specimens taken within thirty miles of Teslin Lake, on the boundary between British Columbia and Yukon Territory, Longitude 130°-30', at the west base of the Cassiar Mountains. The range of the form cannot therefore be defined, but a study of the material examined is suggestive.

The male specimens in the series are as follows:- British Columbia: Trail, Midway and Revelstoke: 7 specimens.

Montana and Idaho: Preuss Mts., Fiddle Creek; and Nyak and Beartooth Mts.: 4 specimens.

Mackenzie; Mackenzie River: 1 specimen.

Yukon Territory, Teslin Lake: 4 specimens.

1 These terms are not from Ridgway's "Color Standards."