ALLEN: AMERICAN COLLARED LEMMINGS. 537 



The Polaris expedition discovered a nest with four young on August 

 11th, at Thank God Harbor (Robeson Channel), but no doubt this 

 is a late date. Feilden found nests in June and July, containing from 

 three to five young, somewhat fewer than the average for rubricatus. 

 In summer they li\'e largely in underground runways, but in winter 

 these are extended beneath the snow, on the ground. 



The varying abundance of lemmings is one of their most interest- 

 ing peculiarities. In a succession of favorable seasons they seem to 

 increase greatly and spread into less thickly populated areas. But 

 in adverse years they become reduced nearly to the point of extinc- 

 tion. From his observations in northeast Greenland, Manniche con- 

 cluded that the lemmings were greatly dependent for safe wintering 

 upon a sufficient blanket of snow (a meter or more) beneath which 

 they might live in comparative security from the extreme cold of the 

 long winter. He found that a winter which commenced with violent 

 winds, preventing an even snow-fall, was very unfavorable. For 

 the grass-grown areas were swept bare of snow, and the lemmings, 

 unable to obtain food with shelter, become exterminated in large 

 numbers. Their natural enemies likewise contribute to their destruc- 

 tion. The Arctic Weasel pursues them in their burrows throughout 

 the year, and at other seasons than winter the Arctic Fox, the White 

 Wolf, Snowy Owl, Glaucous Gull, and Long-tailed Jaeger feed largely 

 on them. 



Manniche asserts that during the long Arctic nights the lemmings 

 do not come to the surface of the snow, but with the approach of 

 spring are often seen above ground, and even make long journeys 

 over the snow, or across the ice to outlying islands. Similar obser- 

 vations have been reported by other explorers. 



Remarks. — Until more exact comparisons with Old W' orld speci- 

 mens can be made, it is unwise to venture an opinion as to the proba- 

 ble relationship of the Greenland Lemming. It seems in some respects 

 like a depauperate form of D. rubricatus of northwestern America. 

 Its geographic distribution in Greenland may indicate that its arrival 

 there is geologically recent, for it seems to have reached only the north 

 and east coasts of that country, and has not yet spread to the south 

 and west coasts. These facts may point to its origin from the west. 

 Specimens from the Arctic Archipelago are much desired to establish 

 whether or not complete intergradation takes place with American 

 continental forms. Present evidence of this, however, is wanting, 

 and the dental and color-characters seem sufficient to accord specific 

 rank to this Lemming. It is certainly not closely related to the 



