112 



SCIENCE. 



[N. S, Vol. IX. No. 212. 



upon various bacteria ; in fatal diseases it some- 

 times loses part of its germicidal j^ower for the 

 colon bacillus shortly before death, but more 

 frequently retains this power for several hours 

 after death ; human blood serum does not lose its 

 germicidal power for typhoid and colon bacilli, 

 even in the late stages of chronic wasting disease. 

 The Philadelphia Medical Journal, which dur- 

 ing its first year has secured a high position 

 among medical journals, will hereafter publish 

 a monthly supplement of 60 pages containing 

 original articles. 



SOCIETIES AND ACADEMIES. 

 NATIONAL GEOGRAPHIC SOCIETY, JANUARY 

 6, 1809. 



Abstract. 



' The Work of Glaciers in High Mountains : ' 

 By Willard D. Johnson. The greater number 

 of the imposing forms in the summit regions of 

 nearly all high mountains are of unknown 

 origin. They are, however, strictly confined to 

 tracts which either have in the recent past been 

 glaciated or are glaciated now. Presumably, 

 therefore, they are of glacial origin. But the 

 difficulty is that, according to the known laws 

 of glacial erosion, they are unintelligible. 



The recognized process in glacial erosion is 

 scour. This process, like aqueous corrasiou, 

 must always tend — in uniformly resistant and 

 unfractured material — to produce graded slopes. 

 But in glaciated summit regions, especially in 

 granite and in tracts of that rock which answer 

 most nearly to ideal condition^ of uniform hard- 

 ness, the topography is essentially that of flat 

 valley floors and of upright cliffs, transverse as 

 well as longitudinal to the direction of flow. 

 In sound rock both glacial scour and aqueous 

 corrasion will be not only incompetent but in- 

 imical to the production of such forms. 



An unrecognized process appears to be that 

 of sapping. The transverse, and therefore 

 buried, cliffs in the glacier's pathway, as well 

 as the amphitheatral cliff at its head, are 

 cliffs of recession. The action of scour is 

 downward and outward with the glacial ad- 

 vance, but the action of sapping is horizontal 

 and backward. It is seldom lateral, and then 

 only for a brief space. The flat valley bottom, 

 as well as the parallel valley walls (where sub- 



sequent scour has not dulled their upright 

 profiles), are by-products of recession of the 

 transverse cliff. 



So long as, along any advancing line, it con- 

 tinues active, sapping will be altogether domi- 

 nant over scour, accomplishing large results in 

 excavation ; but its action, apparently, is by 

 successive attacks, from point to point, and has 

 relatively brief duration. Its forms, thereafter, 

 arrested in development, become obsolescent 

 under the contiuous action of scour, and the 

 rounding-off of angles puts them seemingly into 

 the category of scour forms. 



The following hypothesis is advanced as to 

 the cause of glacial sapping : The glacier pro- 

 tects its bed against the sharp variations of 

 temperature which, by mechanical disintegra- 

 tion, waste exposed slopes. At the same time 

 the covered rock surface is maintained close to 

 zero, Centigrade — a critical temperature. By 

 tearing away at its head from the mountain 

 slope, and by reason of initial irregularities of 

 bed along its line of flow, the glacier is broken 

 across. Jf the depth of ice be not too great 

 these breaks, or crevasses, will penetrate to the 

 bottom. Along the narrow transverse line of 

 bed, or floor, thus exposed — during summer, 

 while the crevasse is open — there will be oscil- 

 lations of temperature, between day and night 

 perhaps, accomplishing an alternation of freez- 

 ing and thawing. This alternation across the 

 freezing point, at the crevasse foot, will be 

 much more frequent than upon the exposed 

 slopes without — a diurnal, rather than a 

 seasonal, change. The crevasse foot will thus 

 be a line of sharply localized and abnormally 

 vigorous weathering, by coarse mechanical dis- 

 integration. The glacier is an agent here, 

 directly, only in the removal of waste products. 

 Frost- fracturing acts vei'tically downward, as 

 well as horizontally backward, into the cliff, 

 which it thus undercuts ; but the products of its 

 downward work are much less readily removed, 

 and failure to remove operates to defeat down- 

 ward action. Thus the cliff recedes, leaving in 

 its trail an approximately flat and horizontal 

 floor. In the slight unevennesses of this floor, 

 after glacial conditions have passed and the 

 caiion has become emptied, rock-basin lakes 

 accumulate. 



