834 



SCIENCE 



[N. S. Vol. XXV. No. 647 



ive eroding agents. Among these is W. Kilian, 

 of Grenoble, whose studies have made him 

 intimately familiar with many valleys of the 

 Prench Alps, which most physiographers would 

 regard as overdeepened by intense glacial 

 erosion, but which Kilian explains otherwise 

 (' L'erosion glaciare et la formation des ter- 

 rasses,' La Geogr., XIV., 1906, 261-274). 

 These trough-like or broad U-shaped valleys, 

 sharply incised beneath the gentler upper 

 slopes (' paliers ') which steepen somewhat in 

 ascending to the peaks, he regards as the re- 

 sult of retrogressive torrential erosion during 

 a late interglacial epoch, with slight modifica- 

 tion by ice action during the last glacial 

 epoch; the torrential action thus appealed to 

 being itself explained as the result of an as- 

 sumed change in the attitude of the land-masa 

 with respect to base-level, and the hanging 

 lateral valleys being accounted for as a result 

 of the preponderance of erosion by the main 

 stream. 



It would certainly seem appropriate that 

 those who adopt this hypothesis should, in the 

 interests of thorough investigation, them- 

 selves be the first to give it an impartial test 

 by looking to see if it applies in non-glaciated 

 regions, where independent and acceptable 

 evidence of change of base-level may be found, 

 and where the disputed element of glacial 

 erosion is ruled out of the case. Curiously 

 enough, Kilian does not do this; nor doee 

 Heim, nor Garwood, nor Fairchild, who all, 

 like Kilian, deny the efficiency of glacial ero- 

 sion. It is true that reference is made by 

 Kilian to recent articles by J. Brunhes, of 

 Fribourg (' Sur les contradictions de l'erosion 

 glaciare ' and ' Sur une explication nouvelle de 

 l'erosion glaciare,' C. B. Acad. 8ci., Paris, 

 CXUI., 1906, 1234^1235; 1299-1301), to show 

 that U-shaped valleys occur in non-glaciated 

 districts; but Brunhes's articles are brief and 

 general, and give little aid in solving the 

 problem at issue. 



Kilian himself states that U-shaped valleys 

 may be produced by stream erosion in certain 

 structures, and on this simple point there need 

 be no dissent; but the peculiarity of glaciated 

 valleys is that the U-shape prevails in all sorts 

 of structures; thus indicating that glaeiation 



and not structure is the determining factor. 

 The Allegheny plateau, south of the glaciated 

 area in the eastern United States, contains a 

 good number of open U-shaped valleys, espe- 

 cially where relatively weak underlying strata 

 are capped by stronger overlying strata; but 

 all such open main valleys receive their 

 lateral valleys at accordant grade, and thus 

 differ most significantly from glaciated U- 

 shaped valleys. It is singular that the non- 

 glacialists do not themselves discover and 

 accept this suggestive fact. As to the work- 

 ing hypothesis that retrogressive torrential 

 erosion, excited by favorable change of base- 

 level, may in time produce great trough-like 

 valleys, certainly no one should, on a priori 

 grounds, object to its due consideration; but 

 this hypothesis normally requires the develop- 

 ment of lateral valleys accordant with theii 

 main valley by the time that the main valley 

 has gained an open floor; and as soon as the 

 inquirer sees that the open Alpine troughs are 

 constantly associated with discordant or hang- 

 ing lateral valleys, it would seem to be in- 

 cumbent upon him to set the working hy- 

 pothesis aside as invalidated, and to look for 

 another of greater competence. 



Kilian and Brunhes both emphasize the im- 

 portance of subglacial torrents, as Prech has 

 done even more strongly (' Ueber das Antlita 

 der Tiroler Zentralalpen,' Zft. Deut.-Oesterr. 

 Alpenver., XXXIV., 1903, 1-31; see especially 

 p. 22), in causing whatever erosion may have 

 taken place during the glacial occupation of 

 a valley; and it is to be presumed that no one 

 would wish to minimize whatever aid may 

 thus be given in a difficult problem; but it is 

 inconsistent with all that is known of torren- 

 tial action to think that, even under the con- 

 straint of subglacial flow, torrents can have 

 accomplished the major part of the work that 

 must be attributed to the general process of 

 glacial erosion. Sharp-cleft gorges, such as 

 that of the Aar in the rock-sill above Meir- 

 ingen, and of various other Swiss rivers, may 

 well be ascribed in large part to subglacial 

 torrents ; but the trough-like cross-section of a 

 glaciated valley is not in the least what would 

 be reasonably expected from the work of a 

 high-pressure torrent. Indeed, the fact that 



