50 II. F. Reld — Studies of Muir Glacier. 



If a glacier reaches water wliicli is so deep tliat it clues not 

 touch the l^ottoni, and the motion of the ice is more rapid at the 

 hottom than the melting, then its end will be forced further and 

 further into and deeper and deeper under the water, following 

 the slope of the bed, until the buoyancy of the water is sufficient 

 to break it off. The place where the fracture will occur, and the 

 size of the iceberg formed, are problems of mechanics. 



Glacial Erosion. 



The general scratching and smoothing of rock l)y glaciers is 

 familiar to all. Another method of erosion, not so generally 

 recognized, was observed here. The spur of Tree mountain, 

 which I have already mentioned and on which we camped one 

 night, is a compact slate ; parts of it were smoothed and scratched ; 

 other parts bore a confusion of mixed rock, the rock of the spur 

 largely predominating ; in still other places the bed-rock showed 

 where angular pieces had been broken out, leaving holes which 

 in some cases contained water. Near thesummi't of nunatak H 

 is a rock-basin lake which must liave been formed in the same 

 way. It is about 40 feet long and 20 feet wicle.^ Its sides are 

 much scratched ; on one side the rock rises vertically eight or 

 ten feet above the water. The rock which formerly filled this 

 hole, separated probably l)y joints from the rock beneath, must 

 have been torn out l)y the ice in its passage over the spot — not 

 necessarily as a whole, but possibly by pieces. The rocks thus 

 torn out are in part pushed liy the glacier to its end, in part 

 rubbed and ground into fine mud and carried off by the sul)- 

 glacial streams. This method of glacial erosion seems to me 

 much more efficient in digging valleys than the simple scratch- 

 ing and smoothing that is so much ftiore noticeable in valleys 

 formerly occupied by glaciers. 



Probably the best method available for determining the rate 

 of erosion is to calculate the amount of sediment carried off by 

 subgiacial streams, as Professor Wright did. Repeating his cal- 

 culation with the more accurate data at our disposal, we find that 

 an average of about three-quarters of inch is eroded annually 

 from the bed of Muir giacier.f 



* These dimensions are given from memory. 



fThe following data were used in this calculation: Area drained by 

 glacier, 700 square miles ; area of glacier bed, o50 square miles. If we 

 assume no motion at bottom, then in tbe middle of the ice-front the qnan- 



