42 ROYAL SOCIETY OF CANADA 



below zero and when there was no floating- ice, the water instead of being 

 clear as iisnal was lead-eolonred and thick Avith ice, invisible from the 

 .surface and flowing with difficulty. 



The formation of these tine crystals of ice. throughout the mass of a 

 river not protected by a layer of surface ice, is caused by the action t)f a 

 cold atmosphere, abstracting latent heat from the water, which is cooled 

 to the freezing point. 



Mr. Henshaw in bis paper expresses the^ opinion that water never 

 freezes without an independent nucleus ; but gives only a theoretical 

 explanation of what he means by its formation from particles of vapour 

 rising from the water, becoming frozen and falling back. 



It seems to the writer that this idea of nuclei aiding the formation 

 of the crystals in the water is a suggestive one. only it is more probable 

 that these nuclei would be supplied by jmrticles of suspended matter, such 

 as sand or other denudated material. 



In disturbing the water on the underside of the surface ice, I have 

 frec^uently noticed that masses of frazil carrying quantities of earthy 

 matter have risen. On the first appearance of this 1 considered that the 

 -Stem of the thermometer had been lowered too far, and had disturbed 

 the l»ottom. but on closer examination this could not have been possible. 

 The fine earth appeared to be imbedded in among the ciystals, and did 

 not have any tendency to sink when disturbed. 



Mr. Keefer states that frazil fastened to the underside of surface ice 

 is found mixed with earthy material, but considers this to have been torn 

 up from the bottom. He says that this ice has been formed on the bottom 

 in open water during extreme cold weather, and that on becoming 

 detached it is carried under the surface ice, where it becomes fixed by frost. 



There seems to be considerable divergence of opinion in regard to 

 the way in which anchor-ice is formed. Mr. J. B. Prancis, in a paper 

 read before the American Society of Civil Engineers and published in its 

 Transactions for June, 1881, considers it to be due to the shifting of currents 

 in a rapidly flowing river. He states that cold surface currents finding 

 their wa}^ to the bottom, carry down particles of surface formed ice, which 

 become fixed there by regelation. As regelation requires considerable 

 pressure continued at one point, it seems unlikely that this plays any part 

 in the formation of ground ice, especially when the ice is moving in a 

 current. 



Mr. Henshaw has given an explanation in his paper to account for 

 the formation of ground ice and the reasons for its adhering at some 

 points and not at others. Jlis consideration that a sudden cooling of 

 water "up stream" would cause currents colder than the surrounding 

 water, seems tenable, but it is not quite clear what is meant by a " super- 

 cooled" current. He distinctly says that he does not wish it to be con- 

 sidered a current cooled below the freezing point. It is evidently to be 



