KENNEDY GRAVELS 



329 



obscure stratification, the topographic form, and the position of the 

 mesa, all characterize this occurrence as a remnant of an alluvial cone of 

 Kennedy creek. No earlier record has been detected in the history of 

 that stream. Since that date, however, the valley has been cut down 

 900 feet, a glacial epoch has intervened, and the channel has recently 

 been reexcavated and sunk deeper in the subterrane. 



Certain tabular drift surfaces between Swift Current and South Ken- 

 nedy creeks and on the northern slope of Yellow mountain are probabty 

 not of the Kennedy formation, but are outwash plains beyond moraines. 

 Gravel mesas, that are correlative with the Kennedy and maybe included 

 under the formation name, occur in Canada, one lying 6 to 8 miles north 

 by west from Chief mountain and east of Belly river ; another, a group 

 of three hills, occurring east of lower Waterton lake, a few miles north 

 of the boundary (figure 2). The basis of correlation in these two cases 

 is general form, altitude, and constitution of the masses, which were not, 

 however, examined in detail. 



Figure 2. — Sketch of northern End of Lewis Range. 



Showing flat topped foothills of the Kennedy formation standing above terraces of valley drift. 

 Looking east near the outlet of Waterton lake, Alberta, down Pass creek. 



Gravels are widely spread on the highest tables of the Plains north of 

 Catbank river and between the forks of Milk river. Their position sug- 

 gests a correlation with the Kennedy formation. On the other hand, the 

 gravels of the Plains are composed chiefly of quartzite and presumably 

 have lost the more soluble constituents, which still occur in the Ken- 

 nedy formation. From this distinction, greater antiquity may be argned 

 for the high level gravels of the Plains. Salisbmy, in summarizing the 

 results of Calhoun's observations in 1901 in this region, says : * 



" The high-level quartzite gravels on the plains east of the mountains are be- 

 lieved to be deposits made by streams at the close of the first epoch of baseleveling 

 recorded in the present topography. " 



If this belief be confirmed, the high-level gravels of the Plains and 

 the Kennedy formation are alike in genesis and derivation from the 

 Lewis range. They may, nevertheless, belong to widely different stages 



* Journal of Geology, University of Chicago, January, 1902. 

 XLIX— Bur.r.. Gsor,. Soc. Am., Vor,. 13, 1901 



