ALLEN: AMERICAN COLLARED LEMMINGS. 529 



skulls of both are in fragments. From still farther west, I have seen 

 three skins from Aylmer Lake, western Keew^atin, which in color 

 are intermediate between richardsoni and true rubricatus. Their 

 cheeks are distinctly gray, though not the clear gray of the latter, 

 and though the body is of the general brownish gray typical of rich- 

 ardsoni, there is also a distinct indication of the chestnut shoulder 

 area reaching from the neck nearly to the middle of the body as in 

 rubricatus. Another specimen, from Coronation Gulf, directly 

 north of Aylmer Lake, but on the Arctic Ocean, is also intermediate, 

 but on the whole nearer rubricatus. From this longitude westward, 

 all the specimens examined are more or less typical of the latter 

 form. 



Habits. — A summary of observations on the habits of this Lemming 

 is given by Preble (1902) chiefly from his own field-notes. He found 

 the animals inhabiting burrows mainly in gravelly ridges along the 

 Hudson Bay coast. These burrows were usually not over tw^o or 

 three feet in extent, \\'ith the entrance passage at an angle of 45 degrees 

 or so, then continuing horizontally to the nest of grass and moss. 

 Usually a side gallery branched off near the nest and w^as used as a 

 place of refuge. Preble found no other food in the burrows than 

 remains of the leaves of Bearberry (Arctostaphylos). The breed- 

 ing season seemed to be nearly over at Ft. Churchill by mid-August. 

 Three young w^as the usual number in a litter, and every pregnant 

 female secured contained but three embryos. That two litters are, 

 in at least some cases, brought forth in the short Arctic summer, is 

 shown by the fact that one " captured on the Barren Grounds August 

 12, besides containing the usual three small embryos, w^as suckling 

 three young." As previously shown, the Alaskan Collared Lemming 

 averages more (seven) to the litter and may have as many as eleven 

 young. 



Remarks. — The Collared Lemming of the west side of Hudson Bay 

 is a very strongly marked subspecies and though closely related to 

 the Alaskan rubricatus wath which it intergrades to the northwest- 

 ward, it parallels hudsonius of Labrador in its dull uniform coloration, 

 though it is not so gray as the latter. Specimens w^ere obtained from 

 the northwest part of Hudson Bay during Parry's Second voyage, 

 and described by Richardson as Arvicola grocnJandica. Audubon 

 and Bachman, in referring to the Labrador species, distinctly state 

 that they had seen no specimens of that animal, but Audubon figures 

 a specimen which he saw at the British Museum, that may have been 

 the one recorded by Gray (184S) just before, as having been sent by 



