Anas obscura rubripes. Vol. XIX] 1902 BREWSTER, An Undescribed Form of Black Duck. 187 of October, are, as far as I have observed, invariably obscuræ. Most of the representatives of this race evidently pass further southward to spend the winter, but I have three typical specimens which were shot on our seacoast (at Ipswich and Chatham) during the latter half of February, 1901. Until very recently I had supposed that the Black Ducks which breed about the Gulf of St. Lawrence and to the northward along the eastern coast of Labrador would prove to belong to the red- legged form, but Mr. C. F. Batchelder has shown me seven speci- mens (all but one adult) which were collected for him in New- foundland in June and July and which, although slightly larger than our New-England-breeding birds, are precisely similar to the latter in color and markings. A female in Mr. O. Bang's collection taken in the Straits of Belle Isle on April 25, 1900, must also be referred to obscura. Another, belonging to Mr. J. D. Sornborger, which, with her brood of ducklings, was captured on July 8, 1896, at Okak, on the northeastern coast of Labrador, is intermediate in certain respects between obscura and rubripes, but on the whole perhaps nearer the former. To the red-legged race I can unhesitatingly refer only four of the breeding Black Ducks which I have examined. One of these (a female) was taken by Mr. L. M. Turner on July 1, 1884, at Ungava, Northern Labrador; another (unsexed and without date) by Mr. John McKensie at Moose Factory on James Bay; a third (represented by only the head and wing and bearing no sex mark but evidently a female, for it was "with young") by Mr. C. Drexler, on June 19, 1860, at Cape Hope, Severn River; the fourth (a male) by Mr. E. A. Preble, on July 28, 1900, at Fort Churchill; —the two localities last named being on the western shores of Hudson Bay. Mr. Preble's specimen is in the collection of the Biological Survey while the others belong to the National Museum. All four of these birds are in poor condition for comparison. Two of them were moulting, and the plumage of the other two is worn and faded, while that of the Moose Factory skin is also strongly tinged with rusty chestnut—a mere super- ficial stain, apparently. Nevertheless they show satisfactorily most of the essential characters of rubripes. In respect to size and the character and distribution of the black markings on the