334 White — Specific Heats of Silicates and Platinum. 



Art. XXXIII. — Specific Heats of Silicates and Platinum /* 

 by Walter P. White. 



The thermal properties of the silicates present several points 

 of special interest. First is the value of the data in geological 

 calculations; second, the light likely to be thrown on the 

 nature of matter in general by researches through the long 

 range of temperatures within which silicates are stable ; and 

 third, the opportunity offered of comparing the properties of 

 the same substance in different physical states, for silica and 

 many of its compounds, on account of their great sluggishness 

 of transformation, can be carried over the same wide temper- 

 ature range in the amorphous ( vitreous ) condition and in one 

 or more different crystalline modifications. The present paper 

 describes the beginnings of an investigation in this field upon 

 the subject of specific heats. The present results are prelimi- 

 nary, covering but a small' portion of the field and presenting 

 values possibly differing by a few tenths of a per cent from 

 those likely to be finally reached. Yet as practically no data 

 whatever are now available through much of the region 

 covered, and as the general properties of silicates are shown 

 by a few members of the group, this preliminary publication 

 has seemed worth making at the present time. 



The method adopted as the standard is that in which the 

 heated body is dropped from a furnace into a calorimeter. 

 This, the oldest and most familiar method, seems also the most 

 accurate, since the more delicate measurements are carried out at 

 ordinary temperatures. Indeed, the single temperature deter- 

 mination necessary in the furnace is a source of greater error 

 than all the rest of the process put together. The difficulty of 

 operating a calorimeter near a furnace and of transferring the 

 often white hot body into the water without thermal loss has 

 often been counted very great, and its seriousness has been 

 the subject of dispute among workers in this field. In the 

 present work, it was found relatively easy to make all errors 

 from this source certainly less, and probably very much less, 

 than those arising from the lack of uniformity in the furnace 

 temperature. 



The experimental process naturally divides itself into three 

 parts, the heating of the silicate, the transference to the cal- 

 orimeter, and the measurement of the quantity of heat. 



1. Furnace temperatitre. — The one great difficulty of high 

 temperature measurements, that of obtaining uniformity of 



* Preliminary notices of this work have appeared in the Phys. Kev., xxvi, 

 536, 1908, and'xxviii, 461, 1909. 



