* Rivers during Winler. 135 



of the bottom can never descend below -h 4° cent. When this 

 mass is in a state of agitation, the surface, the middle, and the 

 bottom, may be found at zero simultaneously. 



We have only now to examine, why, when this uniformity of 

 temperature exists, and when the entire liquid mass is at zero, 

 that congelation commences at the bottom, and not at the surface. 



But where is the person who does not know, that to produce 

 a speedy formation of crystals in a saline solution, it is merely 

 necessary to introduce a pointed body, or an unequal surface 

 into it ; that it is around the asperities of such a body that crys- 

 tals originate and are promptly increased ? Be it so, every one 

 may be assured that this is the case with crystals of ice ; that if 

 the mud in which the congelation occurs presents a rent or pro- 

 jection, or solution of continuity of any kind, it will become as 

 so many centres, around which the filaments of frozen water 

 will prefer to arrange themselves. 



But is not what we have said exactly the history of the freez- 

 ing of rivers ? This cannot be doubted, if we recollect, that it 

 never takes place in the channel, unless where there are rocks, 

 stones, pebbles, pieces of wood, herbs, &c. 



There is another circumstance which seems to have a certain 

 share in this phenomenon, viz. the motion of the water. At the 

 surface this motion is very rapid and irregular ; it ought of 

 course to put a stop to the symmetrical grouping of needles; to 

 that polar arrangement, without which crystals, whatever be 

 their nature, can neither acquire regularity of form, nor solidity ; 

 it should of course frequently break the crystalline groups, even 

 in their rudimentary state. 



This motion, which is the principal obstacle to crystalliza- 

 tion, if it exists at the bottom as well as the surface of the wa- 

 ter, is at least greatly diminished at the former. It may be 

 supposed, therefore, that its action will merely oppose the for- 

 mation of regular or compact ice, but will not eventually pre- 

 vent a multitude of little filaments becoming irregularly blend- 

 ed, and thus produce that kind of spongy ice through which M. 

 Hugi so easily drove the oars of his boat. 



Having proceeded thus far, the reader may probably ask 

 why I did not present what preceded, as a complete explana- 

 tion of the formation of the grund-eis of Germany, of the 

 glaces defond of our sailors. — This is my answer. 



