322 Dr. T. Sterry Hunt on 



between 59° and its melting-point ; and the remarkable con- 

 densation effected, even by a gentle heat, when the grey 

 brittle species of tin passes into the ordinary white malleable 

 form (not less than the fact that quartz and many native sili- 

 cates at temperatures below their melting-points acquire that 

 increased solubility which they possess, with augmented value, 

 after fusion), — all tend to show that chemical changes may take 

 place by heat in other solid species where hitherto uususpected. 

 They show moreover the uncertainty which attends any attempt 

 to fix by calculation the augmentation of volume in a solid 

 species by heat beyond the limits within which experiment 

 has determined the rate of expansion. 



8. It is clear that if a species could, without chemical 

 change, be heated to a point at which its augmentation of 

 volume from 4° would be just equal to that of water from 4° 

 to its boiling-point at 760 millim., its specific gravity as 

 compared with this liquid at 100° would be the same as 

 that found for the same species at 4° compared with water 

 at the same temperature. The cubical expansion of water 

 for each degree between 4° and 100° is approximately 

 '00043, while that for iron from 0° to 300° is given as 

 •000044, and that for quartz from 0° to 100° at '000040. . 

 The coefficient of expansion for these bodies being thus about 

 one tenth that of water, it would be necessary, in order 

 to attain an augmentation comparable to that of water at 

 100°, to raise them to 1000° and upward, or to temperatures 

 much above those required to produce chemical changes in 

 quartz and in most native silicates. The coefficient of cubical 

 expansion of these last for each degree between 0° and 100° 

 is moreover much less than that for iron and for quartz, vary- 

 ing in general from -000020 to -000028. The latter rate of 

 expansion, if constant to 500° (even below which many native 

 species are chemically changed), would reduce the specific 

 gravity of the species at that temperature only 0*014 ; and up 

 to 1000° only 0*028. The differences between these numbers 

 and 0*043, which represents the cubical expansion of water 

 from 4° to 100°, thus mark the extent of the errors involved 

 in the determination of the specific gravities of such species 

 with water at 4°. The imperfections and impurities of most 

 natural and artificial crystalline species introduce errors 

 not less considerable in the determinations of their specific 

 gravity. The best figures obtained for such species involve 

 in most cases possible deviations, in the one or the other direc- 

 tion, as great as those due to the different rates of expansion 

 of water and these species ; so that we may take the approxi- 



