6i4 J- S. L. VOGT 



These figures, however, cannot be transferred so as to be applied at 

 the melting point without being corrected, as the crystalline and the 

 glassy phase will, as a general rule, differ a little as to volume dilation. 



The only precision investigation known to me of the specific 

 volume of a silicate at the melting point, has been carried out by 

 A. L. Day, R. B. Sosman and J. C. Hostetter,^ who determined for 

 a diabase : 



At 20° in the crystalline phase dens. = 2.975 (specific volume = 

 0.3362) and in the glassy phase dens. = 2.763 (specific volume = 

 0.3620), the percentage difference at 20° accordingly, amounting 

 to 7.1 per cent, the glass being taken as the starting point. 



This percentage difference increased at the melting point to 

 9.1 as a minimum and 10.9 as a maximum, and next to the last 

 figure, accordingly to about 10.5 per cent corresponding to a differ- 

 ence =C(2. 0.041 of the specific volume. 



From what we have stated above, we take it for granted that,, 

 at the melting point of the common rock-forming minerals (melting 

 at about 1200-1500°), vuq.— Vsoi. will, as a general rule, vary between 

 the limits of about 0.015 and 0.045. Assuming a melting heat of 

 100 cal. and a melting point of respectively 1200° and 1500°, we 

 shall have, for a pressure of 1000 atmospheres, corresponding to a 

 depth of about 3.7 kilometers, a rise of temperature of respectively 

 about 5-6° and about 15-18°. That is to say, in igneous flows, at 

 depths of up to 0.5 or i kilometer, the melting point of the different 

 minerals only rises between about 1° and about 4 or 5°, while 

 in deep-seated rocks, in which the crystallization takes place at 

 depths of 5, 10, or 15 kilometers, there may be involved a rise 

 varying respectively between about 6° and 30°, between about 12° 

 and '40°, and between about 15 or 20° and 60°. 



Thus, even in deep-seated rocks crystallizing at very great 

 depths, the rise of the melting point is rather inconsiderable. 

 And it should be particularly emphasized that the difference of 

 rise of the melting points of the various minerals crystallizing in 

 one and the same magma, will only amount to some few degrees; 

 even in deep-seated rocks the difference is rather small, in fact 

 rarely amounting to more than 10, 20, or perhaps in some cases 40°. 



I Amer. Jour, of Sci., XXXVII (1914)' 



