252 ON CONGELATIOjr. 



itself determines the crystallization, the same thing hap- 

 " pens with regard to thespicular crystals thus beginning to 



be formed. This is the reason why rivers always bt'gin 



to freeze near their marshy sides, and at places where 



the current is the least rapid. 

 — And stand- But to return to the stagnant waters. However abun- 

 mg water, ^^^^^ ^^ y^^ ^j^^ deposition of hoar frost, the water bc- 

 which explains , / . , .,, . - . %¥,, ~ « 



the phenome- neath abridge will receive no part of it. Ihe surface of 



■a* the water has therefore this cause of refrigeration less than 



\ at the other parts. Its caloric is not taken away but by 



the mere contact of the air. This condition would bo 

 sufficient to render its congelation much more slow ; but 

 its fluidity is not less preserved by its surface being de- 

 j fended from the predominating action of the hoar frost, 



I which would follow if that obstacle were not interposed. 



Other facts of These facts enabled me to explain certain experiments 

 the same kind, urged by a philosophical gentleman at Paris a few years 

 i.^ ago, to support his opinion on the existence of a mate- 



rial principle under the name offrigoric. The author of 

 these experiments assured me, that in a frosty night the 

 frigoric fell perpendicularly from the atmosphere upon 

 the surface of the earth ; and he offered the following 

 proof. If plates filled with water be exposed to the 

 open air at night, and it be cold enough, the water will 

 freeze; but if one of these vessels be covered with a pane 

 of glass, or any other body, that water will not freeze, 

 even though the covering body do not rest immediately 

 upon the plate. It is sufficient, continued the author, 

 that the fall of the frigoric be interrupted, no matter 

 whether from an higher or lower distance; and to com- 

 plete his demonstration, he added the following experi» 

 ment, which at first aspect seems very cogent, and is cer- 

 tainly very interesting. It is as follows : place in the 

 evening, at a certain distance above a plate filled with 

 water, a funnel, of which the diameter shall be less than 

 that of the plate, you will find the next day a ring of 

 ice formed round the circumference ; but all the water 

 situated perpendicularly beneath the funnel will remain 

 fluid. 

 Explanation, I h^ve not repeated this last experiment ; but every 

 thing leads me to conclude that it would succeed in favor* 



able 



