[ 167 ] 



XXII. Intelligence and Miscellaneous Articles. 



ON THE DENSITY OF ICE. BY L. DTJFOUR. 



|"N a previous communication* the author determined the density of 

 -*- ice by making a mixture of alcohol and water of exactly the same 

 density as the ice, so that the ice was everywhere in equilibrium in 

 it, and then determining the density of this mixture. This method, 

 however, laboured under the objection that the alcoholic mixture 

 dissolved some of the ice, which vitiated the result. 



In a recent series of determinations he has used a mixture of 

 chloroform and rock-oil for the same purpose. This mixture does not 

 dissolve any ice, it keeps fluid below 0°, and the density of the mixture 

 can be made to vary between 0*82 and 1*50. It is easy to make a mix- 

 ture so that ice remains in it in perfect equilibrium ; if then a little 

 chloroform or a little rock-oil be added, the ice will either rise or sink. 



Preliminary trials showed that the mixture retained its homoge- 

 neity for a longer time than was necessary for an experiment. In 

 these trials he used the method of hollow glass floaters, the density 

 of which could be determined by ordinary methods. Four trials 

 showed a mean difference of 0*0013 between the density of the float- 

 ers obtained directly, and that of the mixture in which they were in 

 equilibrium. This approximation is doubtless too rough, and the 

 method which tolerates it could not be recommended for bodies to 

 which ordinary methods are applicable ; but in the case of the spe- 

 cial difficulties which ice presents, it will be found that this method 

 is not to be despised. By multiplying the determinations, it might 

 be hoped that the mean would not be very far from the truth. 



The choice of the ice is difficult. Fragments must be used which, 

 are free from air-bubbles, and obtained by means of distilled water ; 

 but this absolute elimination of air is very difficult. By causing 

 water, boiled for a long time, to freeze in a kind of barometric 

 chamber, he obtained an opaline ice, quite homogeneous, some frag- 

 ments of which were admirably adapted for the object intended. 

 Their opalescence was not due to the presence of air, but probably 

 to its structure, or to internal cleavages. Although produced in a 

 space where the tension of the remaining air did not exceed half a 

 millimetre, some small bubbles of ice were here and there perceived, 

 and their presence proved by placing the portion of the fragment 

 which contained them at the bottom of a test-tube full of rock-oil. 



The specific gravity of the mixture of chloroform and rock-oil was 

 determined by weighing in it a piece of glass, the weight of which 

 in air and in water was known. The temperature, always below 0° C, 

 varied from — ^° to — 8°. All necessary corrections were made to 

 reduce the result to 0°, admitting for the coefficient of the cubical 

 expansion of ice the number 0*000158, given by Pliicker and Geissler. 



The method gave higher and lower limits, between which the den- 

 sity of ice was certainly comprised ; for these limits different values 

 were obtained, the most extreme being 0*920/ and 0*9133. These 

 limits comprise the numbers given for the density by Berzelius, by 

 Pliicker, and Briinner : the values given by Heinrich, by Kopp, and 

 by Osann are outside these, and, for the specimens of ice operated 



• Phil. Mag. vol. xx. p. 248. 



