FORMED AT THE BOTTOM OF RUNNING WATER. 343 



without part of it being converted into ice, the spiculae and crystals of which find a 

 solid body for their attachment at the very point where the heat is given out*. 



But while in this manner we can explain some of the incidents, may it not be held, 

 as above demonstrated, that the chief cause of the ground-gru is the radiation of heat 

 from the bottoms of the rivers ? Every branch of the phenomenon is of easy expla- 

 nation when we admit the radiation ; and among the rest a circumstance to which I 

 have yet made no reference, and that is, the disappearance at the bottom of the water 

 of the immense quantity of heat, 140° of Fahr., which constitutes the caloric of 

 fluidity disengaged, when water at 32° Fahr. is converted into ice at the same tem- 

 perature. 



The answer to our original question then is. That ice is formed sometimes on the 

 surface of running water, and sometimes at the bottom, because frost sometimes 

 takes place with a clouded sky, which is incompatible with radiation of heat from 

 the bottom of the stream, and sometimes with a clear sky, when that radiation takes 

 place through the water, in the same manner as the experiments of Dr. Wells prove 

 it goes on, under a like sky, through the atmosphere. The bottom is by this cooled 

 down below the freezing point of water, before the water itself; ice is formed on it, 

 and its detachment by transmitted heat from below prevented, as long as the radia- 

 tion continues. 



* We may observe also, that there is a local source of greater cold of the water in the rapids, in its being 

 brought into more active and extensive contact with the air by the sharp ripple and spray. 



MDCCCXXXV. 



2 Y 



