Decadence of the Glaciers. 41 



act. The ends of glaciers are rarely stationary ; they seem always 

 to be advancing or retreating ; the suggestions above show that 

 this is just what is to l^e expected. In other words, glaciers are 

 never exactly in stable equilibrium with the surrounding con- 

 ditions. 



Changes to be expcctecl. 



Main lake and Berg lake are now separated by a very short 

 distance, and it will not be long before they unite. This will 

 result in the draining of Berg lake, which event will probably 

 be marked by a flood. The melting of ice in Main valley 

 must be rapid, for the great extent of its termination there pre- 

 sents a large surface for melting. When this termination has 

 receded two or three miles and the surface of the ice has sunk 

 two or three hundred feet, the ice from the first northern and 

 from the southeastern tributaries will probably be in part de- 

 flected into Main valley. The small lake which occupies a 

 lateral valley opening into Granite canyon will probalily extend 

 as the ice diminishes and perhaps occupy a large part of the 

 canyon itself. 



Professor AVright has kindly sent me some photographs which 

 he took of the glacier in 1886. By comparing these with our 

 own we can readily fix on our map, within 100 j^ards, the position 

 of the ice-front at the time of Professor Wright's visit. This 

 shows that in the four years from 1886 to 1890 the western end 

 of the ice front has receded 1,200 yards and the eastern end 750 

 yards. The center also has receded about 1,200 yards, so that 

 the average recession of the ice-front is a little over 1,000 yards 

 in four years or, say, a mile in seven years. Professor Muir 

 writes me that the notes of his first visit to the glacier in 1879 

 show that the ice then extended about to our station D ; the 

 rate of retreat deduced from this accords fairly well with that 

 given above. The ice-front, therefore, must have extended as 

 far as island C 20 years ago. Below C I think the retreat was 

 more rapid ; for there the glacier presented a much wider front 

 to the water from which a correspondingly larger quantity of ice 

 must have broken off, and this could hardl}^ have been entirely 

 compensated for ])v a greater velocity of flow on account of the 

 many obstructions in the neighborhood of the present position 

 of the ice-front. It does not seem at all incredible that the ice 

 from the various glaciers of Glacier bay may have united to fill 

 a large part of the bay a hundred years ago. Professor Wright's 



