4 ROYAL SOCIETY OF CANADA 



from time to time have shown. It has always Been my contention that a 

 knowledge of the physical laws governing the formation of ice would 

 make the task of coping with the ice problem comparatively easy from 

 the practical standpoint. Further study of the question in relation to 

 ice breaking and winter navigation of the St. Lawrence has made this 

 even clearer to me. Observers are too apt to judge of the task before 

 them from the enormous forces at work in the vast accumulations of ice 

 masses which collect in unprotected localities. The well known saying 

 that an " ounce of prevention is worth a pound of cure " holds with un- 

 expected force when dealing with the ice problem. T shall, therefore, 

 merely discuss the properties of ice from the physical standpoint and set 

 before you, in this address, the best determinations of the constants that 

 are now available. 



Density of Ice. 



Passing over the earlier measurements of the density of ice by 

 Boyle, Williams, Berzelius, Heinrich, Dumas and Osann, we find one of 

 the first attempts at an accurate measure was made by Brunner in 1845. 

 His specimens were clear pieces of river-ice, which were weighed first in 

 air and then in petroleum. The value he obtained was .9180. 



Pliicker and Geissler followed shortly after, using a dilatometer 

 method and found the value .9158. Kopp in 1855 found, by the same 

 method, .907. 



Dufour in 1860 determined the density by submerging ice in a 

 liquid, the density of which he could vary and found the value .9175 

 and later, using different liquids, he found .9178. 



In 1870 Bunsen set himself to determine the value in connection 

 with his well-known ice calorimeter. He employed a dilatometer of 

 special design, obviating certain errors connected with previous dilato- 

 meters. He obtained the values .91682 and .91667, giving a mean of 

 .91685. 



Zakrzevski, in 1892, redetermined the value by Bunsen's method, 

 using boiled water. He found the values .916710, .916713, 916708, 

 giving a mean of .916710. 



In 1899 E. L. Nichols made a careful series of measurements by 

 various methods. He found that there were important differences in 

 the density of new and old ice and between natural and artificial ice. 



Differences of the order of 1 part in 1000 were found for such forms 

 of ice. Ice mantles prepared by CO, and ether gave a mean value of 

 .91615, while natural ice, such as icicles, gave .91807. Natural ice from 

 a pond, newly cut, gave .91804, while the same ice, one year old, was 

 only .91644. Still another value found by Nichols, by a different me- 

 thod, for ice newly cut from a reservoir, was .91772. 



