THE TILL. 535 



elongate in the direction of its motion, like an inverted canoe. These 

 one may compare with the bars of a river, and thus complete the almost 

 perfect parallel between the two. 



The study of these deposits is very difficult (the hardest problems in 

 the g-eoloo-ical book are at tlie beginning and the end) because they often 

 blend intricately with succeeding deposits and are largely concealed by 

 them, and because they can be successfully studied only in fresh exca- 

 vations. In a few days the exposure caves and sinks into a, sl(>j)c wliich 

 often loses all its characteristic peculiarities of the deposit. For this reason 

 the following descriptions have i-eference almost always to fresh exposures, 

 and especially where the color or consistency of the bed is discussed the 

 reference is to a surface newly opened up and still moist. 



The deposits of the basin which we may refer to the "moraine pro- 

 fonde" of the inland ice and which we may believe to have rested beneath 

 tlie ice wholly completed in the form in which we now find them, at a time 

 when the ice was so fixr thinned by melting that it had ceased to advance 

 and only awaited its final dissokition, may be divided into three gi-oups of 

 only local value, whose differences in structure depend in large degree upon 

 their position in the valley or their altitude above it. These are the upland 

 di'ift, the fine valley drift, and the coarse valley drift. 



THE UPLAND DRIFT. 



Using the old word drift (althougli it has somewhat gone out of 

 fashion in late years, and although it contains always some reminiscences of 

 earlier theories now wholly abandoned) for the explanation of the phenomena 

 with which we are now occupied, we will take the section exposed by the 

 ditch for the water main from the west village of Pelham eastward to where 

 the Shutesbury road branches off. In this exposure a face 1,300 feet long 

 and 5 feet deep, 320 feet above sea, in many places showing the underlying 

 rock, Avas open for study. 



This mass is wholly free from clay or fine sand, and consists in the 

 main of fragments of rock of various sizes up to 4 feet on a side, with a 

 considerable preponderance of those about 1 foot in diameter. These 

 bowlders are almost wholly local — that is, they consist of the ordinary 

 Pelham gneiss u})on which they rest — with very rarely a fragment of the 

 compact trap which occurs a few rods north. 



