34 UNIVERSITY OF COLORADO STUDIES 



mountain or in a gulch, manifestly the movement would be mostly 

 or entirely down the slope from the place of accumulation, and, in 

 case of its occurrence in a gulch, would follow the meandering of the 

 gulch and thus form a "stream" of ice. Such is the typical glacier 

 of a high mountain system. Clearly the original snow-drift was not 

 a glacier. It did not become one when a little was left over and added 

 to by the following year's drift, nor even by a number of such accre- 

 tions. The beginning of a downward or outward movement of itself 

 would not necessarily transform it into a glacier. A body of well- 

 consolidated snow, not fully changed into ice, may slowly creep down- 

 ward or extend a tongue downward upon a steep slope,^ or, if the center 

 of the accumulation be deep enough, it may creep without any slope 

 of the underlying ground. Finally our drift, by small annual accre- 

 tions, may extend down a gulch for many miles, the great body of it 

 composed of true glacial ice, separated from the upper part of the 

 snow of the accumulating area by a well-defined bergschrund. It 

 would be crevassed by slow movement over the variable grade of the 

 bed of the gulch, grinding the rocks beneath it as it moves forward 

 with almost resistless power, sending on down the valley beyond its 

 greatest extension a stream of milky water, laden with the fine sedi- 

 ments scoured from the rocks. No one could then doubt that it is a 

 glacier, yet who can tell at what precise moment in its development 

 it became a glacier? A similar question arises in its retreat when 

 climatic conditions become unfavorable for its continued existence. 

 Just when does it cease to be a glacier and become merely neve ? In 

 some cases a well-defined bergschrund (the great crevasse or system 

 of crevasses extending across a glacier at the point where the ice by 

 accelerated motion breaks away from the partly consolidated snow and 

 ice of the upper neve) may, according to ChamberHn and SaHsbury," 

 be taken as the dividing Hne, and when in its retreat it reaches that 

 line, the term should probably no longer be appHed. The difl&culty 

 of applying this test is great because the transition from snow to ice 

 is gradual and the bergschrund may be not a simple line, but rather a 



' King, Clarence, Report of the Geological Exploration of the Fortieth Parallel, Vol. I, pp. 477-78, 1878. 

 ' Chamberlin, Thojias C, and Salisbury, Rollin D., Geology, Vol. I, p. 258, 1905. 



