PROCEEDINGS OF TORONTO MEETIS 



the drift. Bui a has turned again to a study of certain 



phases of the distribution of boulders. This has led to some distinctions and cla&sifi* 



of importance upon the working out of glacial phenomena 

 i nation of the methi il action. Two leading types need dis- 



criminatioi ulder trains, and (2) boulder belts. Boulder 



train- take t h--i r . > r i lt ' " from knobs or prominences of rock which lay in the path 



ffbouldt . ly and abundantly to the over-riding i 



movement, but the boulders are not carried for- 

 ward in They may therefore appropriately be called boulder 

 fun- i kind, or at least of the few kind- repi I by 

 nt knob. They usually grow smaller and more worn as traced away from it. 

 lingle with the underlying drift, and in thi- respect differ from the l>"u'. 



bell tly to b( nsidered A part of the significant !' these trains has bi 



1. hut much additional significance remains t<> !"• developed. Special investiga- 



by Professor Shaler, and at tli«' west by 1' -• - Hindi 



suits of which cannot here be appropriately given. 



Tin- boulder belts differ from tin- boulder train-, in that they lie transverse to the 



direction ol . movement. Thev arc also contrasted with tbem in that the 



boulders, insl f being of one or a few kind-, arc of many kinds, and, instead of 



being derived from somi source, came from distant Bources. The boulder b 



that have 1 n especially studied by myself are found in Illinois, Indiana, and 



Ohio. The boulders of these belts were derived almost exclusively from the crystal- 

 line or Archroai to 500 miles to the northward. There i- an almost com- 

 :' boulders derived from the intermediate I ' ;oic rocks, although such 

 boulders occur in abundance in the moraines with which the boulder licit- arc 

 ind in tin' drift Bheets and gravel hill- with which they arc connected, 

 baracteristic i- the fact that the boulders arc superficial, and do ma 

 mingle deeply with the underlying drift, as i- the case in the boulder fan-. 



listribution, the boulder belts have been found to coincide closely or nearly 



with terminal mora which Btrongly suggests that they were deposited by 



th<- margin of the ice that formed the moraine-. The solution of the problem pre- 



ted by i • der helt- may he found in an analysis of terminal moraines. A 



I - material at il- margin in three way-: il) It pushes matter forward 



mechanically, ridging it at it- edge, forming what may be termed push moraii 



lacier may fail to carry forward to its actual extremity the material which it 

 (i it- base, and this may lodge under the margin, forming a Bub marginal 

 imulation which may !»• called a lodge moraine. \ glacier carries forward 



the material embraced within the ice or borne on it- top until it reaches the extn 



margin, when it i- dropped, forming what may hi- called a dump moraine. 



■older I ■ held to belong to the last class. It i- believed that boulders 



from the high hill- of the Archsean highlands at - distance up in 



and that these were bori ward in tin- une 



until t: bed the margin, where they were necessarily 



[fth : ::■. il follows as an important inference that the 



'i of the ire .11, i . the -in lace, (, , r j n that ra-e the ahuiidalit 



ed with the foreign drift, which i- luously 



I > the doctrine advocated by Bome that, owing to 



'• frontal r< arrying boulders up to heights above 



