432 Chronicles of Science. ^July, 



contracting instead of expanding on the increase of temperature. 

 But this negative dilatation is not quite the same in the amorphous 

 state, in the state of compressed precipitate, as in the crystalline 

 state. Eaised to a temperature of 40° C. the co-efficient is, 

 - 0-00000137 in the first case, - 0-00000139 in the second. This 

 co-efficient is, besides, variable with the temperature and the direc- 

 tion in the interior of the crystal. In the direction in which the 

 contraction is greatest, parallel to the axis, it is the r+m part for 

 100°, about a tenth of the dilatation of mercury. To see if this is an 

 exceptional case, he has experimented upon other series of analo- 

 gous salts such as the chlorides, bromides, and iodides. But he 

 finds that whilst the co-efficient of dilatation of all ih<3 chlorides, 

 bromides, and iodides is positive, as it is in the case of other sub- 

 stances experimented upon, as far as he has hitherto found out, 

 the co-efficient of dilatation of the iodide of silver is alone negative. 



V. Eegnault has determined the specific heat of several 

 specimens of natural and artificial graphite. Graphite from Canada, 

 No. 1, gave a mean number -19866; from Canada, No. 2, mean 

 •20198; from Canada, No. 3, mean *19113; from Siberia, mean 

 •19879. Graphite from gas retorts, after being heated white hot, 

 gave mean '1^68 ; the same kind of graphite, after being heated in 

 a current of chlorine, gave "2-00, and that from Canada, No. 3, 

 after like treatment, gave '1977. Hard burnt clay gave *1;;40; 

 therefore the ash in graphite (consisting principally of clay) has no 

 appreciable effect on the results obtained. H. Kopp remarks on 

 the above numbers, that they are obtained by immersing the 

 graphite heated up to J 00° in the water of the calorimeter ; that 

 graphite is porous ; therefore the heat given off by the water as it 

 enters the pores of the substance reappears in the final result of the 

 experiment. Kegnault's numbers are higher than those obtained 

 by Kopp. 



The behaviour of some minerals at a high temperature has been 

 studied by Dr. Eisner. He has heated a great variety of minerals 

 and other rock constituents in a porcelain furnace to an estimated 

 temperature of 2,500° to 3,000°, and finds in general that silicates 

 containing the alkaline metals, or iron, are much more fusible than 

 those which contain much alumina, or no iron ; with the exception 

 of obsidian, all the rocks experimented upon, even those of volcanic 

 origin, presented after heating totally different characteristics from 

 those which accompany them in nature ; the crystalline varieties 

 became compact and semi-fused ; therefore he says they must have 

 been produced in nature under different conditions from those of his 

 experiments The conversion of pumice into an obsidian-like body 

 is especially interesting. After heating, the rocks experimented on 

 showed lower specific gravities. 



