108 ROYAL SOCIETY OF CANADA 



Now, in the first place, with respect to shade: I was so far from 

 forgetting that it is an obstacle to radiation, that, on the contrary, 

 in my observations in 1835, I had shown by very many instances, that 

 shade had prevented the formation of groimd-gru, just as it prevented 

 dew. Wherever shade intervened to prevent radiation from the bot- 

 toms of the rivers Don and Leochal, there no ground ice was formed; 

 while the unshaded parts of the bottoms were coated with it. ]\Iy 

 explanation thus mainly rested upon the fact that shade prevents radia- 

 tion. In the next place, with i^espect to wind; the writer in the 

 Cyclopaedia himself forgets the difference of the statical conditions of 

 air and water in connexion with temperature. xA-ir becomes heavier 

 by diminution of temperature. Water under 39° Fahr. becomes lighter 

 by diminution of temperature. During wind, on the land, the cold 

 air at the surface of the earth is continually mixed with, or displaced 

 by, the warmer air above; and by this process both the earth and air 

 in contact with it are prevented from being reduced to a very low 

 temperature by radiation. But in a body of moving water, whose 

 temperature is imder 39°, the eddies of the current thro\v down the 

 coldest parts, which in still water would remain at the surface, to 

 come into contaet with the bottom. This last circumstance is the 

 explanation of M. Arago, and it well accounts for the formation of 

 ground-gru taking place first in the most rapid parts of the streams; 

 although neither by itself, nor when taken in conjunction with the 

 other two circumstances to which he refers, namely, aptitude to forma- 

 tion of crystals on asperities at the bottom, and less impediment to 

 the formation of crystals in a slower motion, will it account for the 

 formation of ground gru, as all these circumstances are present when 

 the water forms only surface ice. The formation of ground-gru 

 requires for its explanation an additional element, namely, the radia- 

 tion, into the clear sky, of heat from the bottom of the river; and 

 the formation never occurs but under a clear sky. 



As to the ground-gru, observed by Colonel Jackson in the Neva 

 under three feet of ice and three feet of snow, that can form no valid 

 objection to the explanation I have given, unless it were ascertained 

 that the gru was formed after the surface ice and the fall of the snow, 

 and not before them. All rivers issuing from lakes, like the Neva, 

 have very clear waters to admit of radiation through them, although 

 as deep as it is; and all rivers are very clear during frost, owing to 

 the freezing up of the little land rills that would convey earthy par- 

 ticles into them. Ground-gru formed in the Neva would be much 

 more permanent than in our rivers. The mean temperature of Alford 



