The Worlc of the Ice-mill. 51 



The erosion, of course, cannot be uniform, but must vary much 

 with the nature of the rock and the thickness and rate of motion 

 of the ice. Near the mouth of the glacier all of these conditions 

 cooperate to increase the action, for the rock is slate, the motion 

 more rapid, and the thickness of the ice probabl}^ greater than 

 elsewhere. It does not seem excessive to consider the erosion 

 here five or ten times as great as the average. The sudden fall 

 between G and H proliably marks the line lietween the harder 



tity of ice carried out per day would be a triangle of 7 feet base (see page 

 44) and 920 feet altitude. Near the sides the triangle would have a smaller 

 base and a less altitude. Let us suppose a wedge having 7 feet base and 

 oGO feet altitude (half the greatest depth plus the height of the ice above 

 water) and a iDreadth equal to the whole ice-front, 9,200 feet, to be break- 

 ing off daily ; let us suppose this daily loss constant throughout the year 

 (our ignorance of the law of glacial motion below the surface will not per- 

 mit a closer approximation — double this quantitj^ would certainly be too 

 much, and would still only slightly affect our result). The Signal service 

 sends -nie as the average of six years' observation of rainfall at Sitka 105.62 

 inches, and the average at Juneau 89.30 inches. Oiir own observations at 

 camp Muir for two months gaA'e about the same rainfall as at Juneau for 

 the same period. I have therefore adopted 90 inches for the yearly pre- 

 cipitation (see appendix ii). The rest of the data I liave taken from Pro- 

 fessor Wright's account (Ice Age in North America, p. 64), viz : 708.48 

 grains of sediment in each United States gallon of water of the subglacial 

 streams ; specific gravity of this material, 2.5 ; loss by evaporation, one- 

 eighth of precipitation. We thus find for the total precipitation over the 

 area drained by the glacier, 146,300,000,000 cubit feet ; annual loss in 

 bergs reduced to water, 5,906,000,000 cubic feet; loss by evaporation, 

 18,300,000,000 cubic feet ; leaving, say, 120,000,000,000 cubic feet of water 

 per year carrying off sediment, which gives an average of about f inch 

 eroded from the whole bed of the glacier. It is assumed that all the 

 water coming from the glacier is charged with sediment. This is in ac- 

 cord with observations so far as they go. The clean surface streams near 

 the end of the glacier empty into the subglacial streams, from one of which 

 the determination of the amount of sediment was made. It may be ob- 

 jected that nuich of this sediment comes from the surface moraines, the 

 rocks either there disintegrating into line material or falling through cre- 

 vasses to the glacier bed and being there ground up. The clearness of the 

 surface streams show that the former is not the case ; and the fact that all 

 the moraines on the surface of the glacier would hardly be enough to supply 

 material ecjual to the sediment carried out in a single year is conclusive 

 evidence against them both. We have not taken into consideration the 

 material pushed out from under the glacier before it has been ground fine, 

 and this is probabh^ of large amount (although we have no means of meas- 

 uring it) and Avould increase the alcove estimate of the erosion. 



S— Nat. r-JKOfi. Mao., \m.. IV, 1802. 



