F. P. Mennell — Cordierite in Dartmoor Granite. 67 



heat of fusion ; (3) the heut changes incidental to the solution of one 

 constituent of the magma in another. But these three factors do not 

 seem to give an adequate explanation of the total heat of some lavas. 

 Basaltic lavas poured out during fissure eruptions, such as those of 

 the Deccan of India, have often flowed superficially many miles from 

 their parent fissures before solidifying; furthermore, before reaching 

 the surface these lavas have doubtless traversed great thicknesses of 

 relatively cool rock continuously abstracting heat. It is difficult to 

 understand how the initial temperature of the magma, based on the 

 three factors referred to in the previous paragraph, could have been 

 high enough to ensure the great fluidity of the basaltic lavas at the 

 surface, in view of the constant losses of heat suffered en route ; 

 consequently, it seems likely that there is a fourth factor in the heat 

 equation. 



The prol)able exothermic character of the reaction by which the 

 garnetifei'ous phase of a rock passes to its nou-garnetiferous phase (or 

 to its liquid phase) suggests this fourth factor, and enables us to 

 understand how magmas formed by the melting of solid rock due to 

 release of pressure may often possess a surplus quantity of heat due 

 to change of chemical phase above the quantity of heat due to change 

 of physical phase. 



I have discussed these questions on the hypothesis that the infra- 

 plutonic shell is characterized by one particular mineral, garnet, that 

 is formed endothermically under high pressure. Further investigation 

 may show the probability of other condensed minerals playing an 

 important part in the same sense as garnet; and, in conversation, 

 Professor R. D. Daly has suggested to me that it would be better not 

 to base my arguments entirely on garnet, but rather on the general 

 principle of the existence of an infra-plidonic shell composed of dense 

 minerals teyiding to pass exothermically into less dense minerals on release 

 of pressure. 



Finally, it must be stated that the foregoing notes are intended to 

 form merely a rough statement of the bearing of this hypothetical 

 infra-plutonic zone on the interpretation of the problems of isostasy, 

 earthquakes, and vulcanicity. It is recognized that temperature and 

 water factors must modify the reactions, at least to a certain extent ; 

 but it is believed that at the depths under consideration pressure is 

 the compelling factor in chemical changes, the chief function of high 

 temperatures and water (when present) being the ensurance of 

 molecular mobility. 



IV. — CoKDiEKrrE IN Granite from Dartmoor. 

 By F. P. Mennell, F.G.S. 

 ri^HE presence of cordierite does not seem to have been often 

 1 recorded from the granite of Dartmoor, in spite of the abundance 

 of that mineral amongst the contact-altered sediments surrounding 

 the igneous mass. The recently published Survey memoirs on 

 Dartmoor and Ivybridge cover nearly the whole area of the granite, 

 and I can only find references to two specimens in which cordierite, 

 or rather pinite pseudomorphs representing that mineral, have been 



