in Til F BOUNDARY-LINES 



rock uluih in: uc fragments. If in the coune 



of subsequent rumination he discovers that these 

 fragments disappear about the same line all along 

 the hill, he may regard his first surmise as probably 

 correct, and draw a boundary -line accordingly, even 

 although he may never have seen the actual rock itself 

 in sifu. 



Again, in the ascent of streams, similar close observa- 

 tion and sagacious inference will often go far to supply 

 the place of actual sections of rock. The use of the 

 ice in these cases, however, requires still more 

 caution than on the lure hillside, because the tendency 

 of running water is to round the rock fragments exposed 

 to it, and hence in the funnel of a brook 

 not be always possible to distinguish between the pebbles 

 whit h have come as angular fragments from ncightxniring 

 solid rocks, and have been rounded by the attrition of the 

 brook or r . and those which, derived from some 



old gravel, were already rounded and water-worn before 

 they tumbled into the channel A great abundance of 

 fragments of one particular of rock, however, 



would suggest that they had not been washed out of 

 some gravel bed, but had been derived from the waste 

 of a solid rock lying somewhat further up in the drainage 

 basin of the stream. In such a case, moreover, the 

 proportion of these fragments in the channel would 

 probably be found to increase as the stream was traced 

 upwards. Perhaps, at the sanu :liey might be 



observed to become larger in size and less water-worn. 

 If they should suddenly cease, the observer ought at 

 once to note the (act, as possibly indicating that the 





