REFUSION OF SEDIMENTS AS FACTOR IN MAGMA GENESIS 445 



ogists believe that igneous rocks are formed by the refusion of sediments, 

 and the similarity of these cores to igneous rock might be regarded as 

 furthering this belief. The original sediments were, however, in this 

 case exceptional in character, and such material can not be regarded as 

 generally available in large quantity for the desired action. Ordinary 

 shales are, however, to be regarded as available in quantity, and Hobbs^ 

 one of the supporters of the refusion hypothesis, would make them the 

 starting point for the formation of lavas.'^ 



By tabulating side by side the chemical composition of average shale 

 and of average igneous rocks (repeated here as columns 6 and 7 of Table 

 lY), Hobbs is able to show a certain broad similarity. At the same time 

 there are distinct differences that are thoroughly characteristic and carry 

 with them definite consequences. The average shale shows a marked 

 dominance of potash over soda and an excess of alumina over that re- 

 quired to make alumino-silicates with the alkalies and lime. Now, as a 

 consequence of these characters, the refusion and recrystallization of 

 such average shale would lead to the formation of such minerals as silli- 

 manite and cordierite, which are not formed from uncontaminated 

 igneous magmas. If igneous rocks are ordinarily the result of fusion of 

 shale, the common absence of such minerals in them is rather surprising. 

 Hobbs would supply the deficiency of alkaline and alkaline-earth bases 

 in his fused shale by supposing that beds of limestone and of salt are 

 absorbed by the melt as it rises toward the surface. Such absorption, 

 though remarkably selective, can not be denied all credence, but one is 

 still puzzled as to how this deficiency is made up in those laccoliths which 

 he imagines to be the result of the fusion of shale in situ. 



The chemical difficulties encountered by this hypothesis are, however^ 

 even less formidable than the thermal difficulties. Hobbs states that 

 "most rocks would fuse at a depth of fifteen miles on the basis of dry 

 conditions only." The temperature at this depth would be 740 degrees 

 centigrade, accepting the geothermic gradient given by Hobbs, and even 

 a salt bed would not melt at that temperature. For the dry fusion of 

 the arkose of the Montezuma core a temperature corresponding to a 

 depth of 23 miles would be required, and this arkose is certainly more 

 fusible than typical shale. It is not, however, to dry fusion that Hobbs 

 appeals for his principal results. Citing the known great effect of water 

 in lowerng the melting point of silicates, he imagines that the temper- 

 ature at a depth of only 6 miles is adequate to melt shale. Again using 



' Earth evolution and its facial expression, New York, 1921, pp. 28-61 ; also Gerland's 

 Beitrage zur Geophysik XII, 2, 1913, pp. 329-361. 



