﻿27(5 



DIt. A. HOLMES ON THE TEHTfAHY 



[vol. lxxii, 



These results correspond in a general way with those obtained 

 by other observers for similar rocks from other parts of the world, 1 

 and they support two generalizations already admitted 2 : — 



(a) That rocks belonging to an alkali series tend to be richer in radium 



than those of corresponding silica content belonging to a calc-alkali 

 series. 



(b) That in alkali rocks the radium-content is roughly proportional to the 



amount of alkalies and particularly to the amount of soda, rather 

 than to the silica percentage. 



In the absence of any knowledge of temperature-gradients in 

 Mozambique, it is not possible to apply the results specifically 

 in their relation to the thermal problems of vulcanism. The 

 radium-content of the basalts (including that from the iSanhuti 

 River) is about l/OxlO -12 grms. per grin., and, since this figure 

 agrees closely with the average for other basaltic rocks, it is to 

 be expected that the average thorium-content of the basalts is also 

 normal. I have previously shown 3 that, if granite and gneiss 

 continue downwards from the surface to a depth of about 

 6 kilometres, and if below that depth basaltic rock continues for 

 47 kilometres, then the temperature at the base of the ' radio- 

 active layer ' (about 33 miles below the surface) must be about 

 1000° C. This calculation assumes that all the earth's heat is 

 maintained by the thermal energy liberated during the disinte- 

 gration of uranium and thorium. If, on the other hand, part of 

 the earth's heat (say, one quarter) is a remnant of its original 

 thermal condition, then the temperature at the same depth may be 

 a little above 1000° C., 4 and the temperature will continue slowly 

 to increase at greater depths. With a capping of granite and 

 gneiss thicker than is assumed above, the basal temperature of the 

 radioactive layer is (given the same temperature-gradient) reduced 

 below 1000° C, whereas, if the thickness is less than 6 kilometres, 

 the basal temperature becomes higher. With no granite and 

 gneiss (that is, assuming the whole radioactive layer to be of 

 basaltic material having a radium-content of 1'OxlO -12 grms. per 

 grm.) the temperature would be increased to a maximum of 

 1650° C. at a depth of about 44 miles. 5 



A temperature of 1000° C. at a depth of 33 miles would be 

 barely sufficient to initiate volcanic activity, even if far-reaching 

 relief of pressure by Assuring be admitted. A temperature of 

 1650° C. at a depth of 44 miles is in this case an impossible 

 maximum, since there is a capping of granite and gneiss, though 

 of unknown thickness. We may, therefore, conclude that the 

 lavas came from depths between 33 and 44 miles. Even so, it is 

 not likely that a basaltic magma could be formed by fusion unless 



'■ See, for example, J. Joly, Phil. Mag. vol. xxiv (1912) p. 694 ; and 

 A. Holmes, Sci. Progress, no. 33 (1914) p. 15. 



2 Ibid. p. 17. 



3 Geol. Mag. ser. 6, vol. ii (1915) pp. 6S-69. 



4 Ibid. p. 111. 5 Ibid. p. 67. 



