248 THE GLACIAL LAKE AGASSIZ. 



of the measured portions of the glacier is much less than 7 feet per clay, 

 and certain immeasured lateral lobes are nearly stagnant. It is further 

 to be observed that the Muir glacier is a trunk stream, the joint product 

 of the ice streams draining a large area. Moreover, these descend from 

 mountain heights. They are therefore radically different from ice-sheets 

 which spread out in all directions on plains or plateaus. 



The evidence cited from Greenland is selective and exceptional ; indeed, 

 selective and exceptional evidence is about the onlj- class that can be cited 

 from Greenland. No average measurements, nor anything approaching 

 to average measurements, have been made. The high rates of move- 

 ment of the Jacobshaven glacier, as given by Helland, and of the Great 

 Karaiak glacier, as given by Drygalski, and other similar measurements, 

 are not at all questioned, but these are quite exceptional, and almost as far 

 as 230ssible from being representative. They exhibit extraordinary move- 

 ments tlu'ough deep, constricted straits, where the ice is forced by the vast 

 accumulations of great areas in the rear, and where the warm season 

 appears to exert its earliest and greatest effects. The amounts of ice dis- 

 charged in the form of bergs from these two glaciers is very much greater 

 than from any other known points on the ice front of Greenland. It is 

 perfectly obvious that the average border of the Greenland ice-sheet does 

 not move at a rate even distantly approximating that of these two straits. 

 If it did so, the whole coast of Greenland must be overwhelmed almost 

 immediately, because the competency of the summer heat of that region to 

 hold back the edge of the ice by melting- is very slight. Drj-galski esti- 

 mated the annual surface melting at 7 feet. Even this is much greater 

 than the annual surface melting of the Inglefield Gulf region, judged by 

 that of 1894. While estimates are few, and even these may need much 

 qualification, it is nevertheless certain that the average movement of that 

 portion of the border of the Greenland ice-cap which lies upon the land is 

 extremely small. Of that portion which ends in the sea only a small frac- 

 tion has a high rate of motion, as is shown by the lack of activity in the 

 discharge of icebergs. When it is considered that the land border is verj- 

 much greater than the sea border, and that of the sea border a portion has 



