Section III., 1899. [ 17 ] Trans. R. S. C. 



IV. — Notes on Frazil and Anchor Ice, with Considerations as to the 

 Freezing Point of Water. 



By Howard T. Baiines, M. Sc. 



Joule Student of the Royal Society (London) ; Demonstrator in Physics, 

 McGill University. 



(Communicated by Prof. John Cox, M.A.) 



In two previous communications ' the author had the honour to 

 submit to the Eoyal Society of Canada the results of certain experiments 

 on the formation of river ice, carried out near Montreal. An attempt 

 Avas made to place the question on a scientific basis by connecting the 

 formation, agglomeration, disintegration and decay of the ice with tem- 

 perature changes in the water, in such a way as to render it possible 

 from climatic conditions to form conclusions as to the state of the ice. 

 In a lecture on this subject, recently given by the author, before the 

 Applied Science Graduates' Society of McCiill University, a number of 

 interesting questions arose, which it is thought might be of interest partly 

 as the subject of the present communication. 



Sec. I. On the Density op Ice. 



Quite recently a paper ^ was published by Prof Nichols of Cornell 

 University, in which is given the results of a very careful series of 

 determinations of the density of ice, by several well selected methods. 

 The work was primarily undertaken to reconcile the discrepancies in the 

 work of previous observers, and with his own more careful work to fix 

 the constant for the ice calorimeter. Incidentally, however, it was 

 shown for the first time that the density varies within narrow limits, 

 depending on the age of the ice and the temperature of formation. The 

 author has stated ^ that it was not likely that ice with a density different 

 from normal could be formed under any condition, and that there were 

 no experimental data to support such a supposition. The recent work 

 of Nichols amply supports this statement, especially so far as the mean- 

 ing of the author had reference to the formation of anchor ice by ice too 

 dense to rise from the bed of a river, an opinion which has been inad- 

 vertently given by previous obseiwers. Within certain limits, however, 

 it appears that natural or old ice is denser than new or artificial ice by 



1 Trans. Royal Society of Canada, vol. II. (N.S.), p. 37, 1896 ; vol. III., p. 17, 1897. 



2 Physical Review, vol. VIII., p. 21, 1899. 



3 First communication, p. 44 ; second communication, p. 29 ; Canadian Engineer, 

 vol. v., p. 8, 1897. 



Sec. Ill, 1899. 2. 



