ICE AXD GLACIERS. 123 



where the base of the valley has a steeper slope, as, for instance, 

 the places in the Mer de Glace (Fig. 14), at g, of the Cascade of 

 the Glacier du Geant, and at i and h of the great terminal 

 cascade of the Glacier des Bois. The ice splits there into 

 thousands of banks and clifis, which then recombine towards the 

 bottom of the steeper slope and form a coherent mass. 



This also we may imitate in our ice-mould. Instead of the 

 snow I take irregular pieces of ice, press them together; add new 

 pieces of ice, press them again, and so on, until the mould is full. 

 When the mass is taken out it forms a compact coherent cylinder 

 of tolerably clear ice, which has a perfectly sharp edge, and is an 

 accurate copy of the mould. 



This experiment, which was first made by Tyndall, shows 

 that a block of ice may be pressed into any mould just like a 

 piece of wax. It might, perhaps, be thought that such a block 

 had, by the pressure in the interior, been first reduced to powder 

 so fine that it readily penetrated every crevice of the mould, and 

 then that this powdered ice, like snow, was again combined by 

 freezing. This suggests itself the more readily, since while the 

 press is being worked a continual creaking and cracking is heard 

 in the interior of the mould. Yet the mere aspect of the cylinders 

 pressed from blocks of ice shows us that it has not been formed 

 in this manner; for they are generally clearer than the ice 

 which is produced from snow, and the individual larger pieces 

 of ice which have been used to produce them are recognised, 

 though they are somewhat changed and flattened. This is 

 most beautiful when clear pieces of ice are laid in the form 

 and the rest of the space stuffed full of snow. The cylinder is 

 then seen to consist of alternate layers of clear and opaque 

 ice, the former arising from the pieces of ice, and the latter from 

 the snow ; but here also the pieces of ice seem pressed into flat 

 discs. 



These observations teach, then, that ice need not be com- 

 pletely smashed to fit into the prescribed mould, but that it may 

 give way without losing its coherence. This can be still more 

 completely proved, and we can acquire a still better insight into 

 the cause of the pliability of ice, if, we press the ice between 



