6i2 /. E. L. VOGT 



G. Tammann, as is well known, has pointed out that for many- 

 salts, viiq.— Vsoi. decreases with increasing pressure so that, when the 

 pressure is extraordinarily high, for instance 5000-10,000 atmos- 

 pheres, it approaches zero; at still higher pressure the difference 

 may be negative. However, this is not the case of substances in 

 general. Thus P. W. Bridgman^ has experimentally ascertained 

 that, in a great many substances (metals and some salts, etc.), 

 on pressures rising to 12,000 atmospheres (corresponding, for a 

 magma, to a depth of about 40 kilometers), vijq.— Vgoi. remains posi- 

 tive — indeed with gradually (somewhat) decreasing magnitude 

 for increasing pressure. Some substances, even at the enormous 

 pressures just mentioned, show but an inconsiderable decrease of 

 the difference of volume, and J. Johnston and L. H. Adams^ showed, 

 for a number of metals, under a pressure of up to 2000 atmospheres 

 (corresponding, for a magma, to a depth of about 7.5 kilometers), 

 a rectilinear course of the difference of volume. 



Under a pressure of one atmosphere all the silicate minerals 

 hitherto examined show, at room-temperature, a lower specific 

 gravity for the glass, that is the extreme viscous fluid phase, than 

 for the crystalline, that is the solid phase. 



According to what we have stated above, we must take it for 

 granted that for rock-forming minerals VHq.— Vsoi. is in all cases 

 positive, and that the difference of volimie existing for a pressure 

 of one atmosphere will remain nearly constant, at any rate down to 

 depths of a few kilometers, and that even to depths of 5 or 10 

 kilometers it will show practically no or only a little decrease. 



As far as the common rock-forming silicate minerals are con- 

 cerned, the latent melting heat is very high throughout. In fact 

 for minerals melting at about 1200° to 1400° or 1500°, it amounts 

 to about 90 or 100 gr. cal. pr. gr. — for some a little more and for 

 some a little less.^ 



^Proc. Amer. Acad., XL VII (1911-12) and Physic. Rev., Ill (1914)- 



^Zeits.f. anorg. Chemie, 72 (1911), and Amer. Jour, of Sci., XXXI (1911). 



3 As for anorthite (melting point 1550°), diopside (1391°), akermanite (about 



1310°) and fayalite (about 1075°), I have in earlier publications (Silikatschmelzlos. 



II [1904], with correcting calculation in my publication on slags in Doelters, Handb. d. 



Mineralchemie, I [1912], p. 942) determined the latent melting heat at respectively 



about 105, 94, 90, 80 cal., with an error limit of =t 15 or 20 cal. For anorthite Bowen 



