On the Density of Quartz, Corundum, Metals, ^c. 145 



Several fragments of No. 1^ fused and cooled suddenly, pre- 

 sented the following densities : — 



Small rounded globules 2-222 



Fragments drawn out and elongated 2*209 



Same glass in very small fragments 2-221 



Same glass in powder, fine and homogeneous . . 2-228 



Mean .... 2-220 



The density of this quartz glass, referred to that of the primi- 

 tive crystal 2-663, shows a diminution of 0-17. 



Of all the minerals which enter abundantly into igneous rocks, 

 quartz seems to be that which possesses in the highest degree 

 the remarkable property of assimilating to itself, during cooling, 

 a certain quantity of heat, which maintains, even after solidifica- 

 tion, the molecules at an abnormal distance apart. This property 

 is of a nature to justify the hypothesis of a surfusion, which 

 several geologists, and more particularly M. Fournet, have caused 

 to enter into their appreciation of the circumstances which have 

 accompanied the solidification of the rocks which, like granite, 

 exhibit quartz in considerable proportions. 



Sulphur is known to be one of the bodies most easily subjected 

 to the pha9nomena of surfusion. Experiments which I have 

 already communicated to the Academy {Comptes Rendus, vol. xxv. 

 p. 857) gave me, between the density of soft sulphur immediately 

 after its preparation, and that of natural octahedral sulphur, a 

 difierence which amounts only to 0-07 of the latter. But this 

 number is evidently a minimum ; for, as I indicated in the same 

 note, the transformation of the soft or vitreous sulphur is ex- 

 ecuted in the first moments with extreme rapidity. 



The metals and their compounds (excepting the silicates) 

 seem, on the contrary, to have but little tendency to assume this 

 peculiar abnormal condition. The passage to the crystalline 

 state is almost immediate, however quick the cooling may be. 



Crystallized bismuth and the metal suddenly cooled, gave 

 respectively the numbers 9-935 and 9-677. Tin cooled very 

 slowly, and the same metal cooled by being poured into water 

 gave 7-373 and 7-239 ; which indicate for these two metals, in 

 the two circumstances, a difference of density amounting only to 

 about 002 of the maximum. 



With lead the phfenomenon is still less pronounced; for between 

 lead poured into water and small imperfect crystals of the metal, 

 extracted from masses of the same lead cooled with great slow- 

 ness, I found a difference of about one hundredth, but in the 

 inverse sense (11-363 and 11-251)*. 



* Another experiment was made with lead precipitated electro-chemically, 

 and with the same lead melted and run ; I obtained the numbers 11 -542 and 



