44 GLACIERS 



cementing ice, and leave the purer crystals as knob-like 

 glacier grains. I have already mentioned the stratified 

 structure to be seen in the newer ice of a glacier, but there 

 are also the dirt bands which form by the collection of 

 rock debris in transverse fissures of the glacier, and are 

 carried on and spread out as curved bands crossing the 

 glacier from side to side when it has flowed some miles 

 on its course and expanded in a broadened bed. 



A great deal of attention has been given to the question, 

 " Why do glaciers descend ? " Though ice is not a viscous 

 body, it yet has some of the properties of one when it 

 presents itself in huge masses, such as are glaciers. It 

 can bend and spread and alter its shape in response to 

 pressure; it splits and reunites its broken surfaces owing to 

 the property of " regelation " which I described above. In 

 a warm atmosphere a cube of pitch, or of sealing-wax, or 

 wax, as big as a quartern loaf, though solid and apparently 

 keeping its shape, will, if placed on a sloping board, very 

 slowly commence to flow down the slope, the process 

 being so slow that it takes hours, or even days, to give 

 any observable result. In virtue of its " sham " viscosity 

 its power of cracking and healing incessantly by rege- 

 lation a sufficiently large and weighty mass of ice 

 behaves in the same way. But it appears that the size of 

 the mass is a very important condition. You can make 

 a small upright figure, say four inches high, in soft 

 wax, which will hold together and keep its form, but if 

 you make a similarly shaped figure of the same material, 

 ten feet high, it will bend and bulge and droop as a 

 paraffin candle does in hot weather. The same import- 

 ance attaches to actual bulk, height, or depth of the mass 

 in regard to the flow of glacier ice, though it seems that 

 the conditions of its flow or movement are not even yet 

 thoroughly understood. Professor Heim holds that the 



