528 bulletin: museum" of comparative zoology. 



dark fur, in sharp contrast to the tufty remnants of the long winter 

 coat which is shed before the new coat reaches full length. 



It was probably a specimen of this race on which James Clark 

 Ross (1835) made the experiment of delaying the molt to winter 

 pelage by keeping the Lemming in a cage in the warm cabin of the 

 vessel during the course of explorations for the Northwest Passage. 

 It remained in the dark coat until February, though with apparently 

 some indications of white hairs making their appearance. After that 

 time its cage was placed outside on the deck, where the thermometer 

 went down as low as 30 degrees below zero, and at the end of a week 

 the lemming was entirely white. It died of exposure shortly after. 



Geographic distribution. — The type-locality of richardsoni, Fort 

 Churchill, Hudson Bay, is at about the extreme southeastern point 

 of its range. Here it shows its maximum of differentiation from 

 typical rubricatus of Alaska, both in its dull rudd;s'-gra}' coat and in 

 certain cranial characters, as previously pointed out. To the north- 

 westward it intergrades completely with rubricatus, and although the 

 specimens at present available are insufficient for an exact determina- 

 tion of the range-limits, they show in general the area inhabited. 

 Like the other forms of the genus it is confined to the barren-ground 

 areas, and the Arctic coasts. Preble (1902) found it on the barrens 

 below Cape Churchill and farther north at two camps south of Cape 

 Eskimo. His beautiful series of specimens, which I have studied, 

 comprises young, immature, and adult animals, and those of the 

 same age are very uniform in general appearance. A series from the 

 extreme northwestern part of Hudson Bay, collected by Captain G. 

 Comer, at Cape Fullerton, Repulse Bay, Southampton Island, Whale 

 Point, and Frozen Stream, are apparently the same. They are, how- 

 ever, for the most part May or early spring specimens and those that 

 are not white, have not fully developed the summer coat. A few, 

 however, seem to be in nearly complete summer pelage and these 

 average grayer than typical richardsoni, but seem best referred to 

 that race, though the series is not strictly comparable with that 

 from Fort Churchill. In cranial characters, also, there are slight 

 discrepancies with the trend of variation toward rubricatus rather 

 than typical richardsoni. Richardson, in 1825, described in consid- 

 erable detail a specimen taken August 22 at Repulse Bay, and his 

 description accords well with that of richardsoni. To the westward two 

 specimens from Artillery Lake, in extreme eastern Mackenzie, are 

 nearly typical richardsoni in color, and probably mark about the limit 

 of the range of the subspecies in this direction. L^nfortunately the 



