V. 



Notes on Ground-ice. 



EVERY student of geological text-books and many observers of 

 nature must be aware of the fact that ice sometimes forms on the 

 bottom of running streams, instead of on the top, as invariably happens 

 with standing v/ater. Sixty years ago the knowledge of this fact 

 seems to have been almost confined to the unlearned, and its reality 

 was denied by the natural philosophers of the day ; it has now become 

 one of the elementary facts of physical and geological science alike, 

 and is only doubted by some few, like Mr. Leslie in his charming 

 " Letters to Marco," who, though keen observers of nature, make no 

 pretence to be regarded as scientific authorities. 



The subject is referred to in every text-book of geology, prin- 

 cipally from the point of view of the action of this form of ice in 

 modifying the shape of the earth's surface, and the question of its 

 mode of origin is treated in a more or less imperfect, where not 

 inaccurate manner. The standard account, to which reference is 

 often made, was published in 1833 by Arago,'' whose paper is princi- 

 pally devoted to a proof of the reality of the phenomenon, which 

 appears at that time to have been generally regarded as incompatible 

 with the fundamental laws of nature. The theory advanced by him 

 was that, owing to the swirls and eddies of the current, the whole 

 body of the stream is reduced to freezing point,- and that in this water 

 the stones and weeds on the bottom serve as so many nuclei for the 

 ice crystals to form upon. 



Within two years of the publication of this theory a totally 

 different one was broached, 3 and supported by some careful and 

 instructive observations. According to this the formation of ground- 

 ice is due to loss of heat by radiation from the stones on the bottom, 

 which are thus cooled below freezing point, and so give rise to the 

 growth of ice crystals in a manner analogous to the formation of hoar 

 frost. 



These two are the theories found, in one form or another, in nearly 

 every text-book, and of them the latter appears to enjoy the greater 



1 Annuaire du bureau des Longitudes, 1833, pp. 244-268. 



2 See Desire Leclerq, Mem. Cony. Acad. Roy. Bruxelles, vol. xviii. (1845), forob serva- 

 tions confirmatory of this. 



•'' J. Farquharson, Phil. Trans, vol. cxxv., 329-3^3 (1835). 



