Manchcsttr Mouoirs, Vol. xliv. (1900), No. 1, 13 



matrix granular, and there are felspar porphyry in which 

 the felspar crystals are large, indicating that in the one 

 case the quartz crystallised first, in the other case the fel- 

 spar ; this, again, implies that in the one case solidification 

 took place nearer the earth's surface and under less pres- 

 sure than in the other, and presents no difficulty if we 

 bear in mind the great diversity amongst the earth's crust 

 temperature gradients. Granite veins, as is well known, 

 are in extreme cases either almost pure quartz, indicating 

 that it was fluid and could force its way into cracks, while 

 the already crystallised felspar remained behind, or these 

 veins are fairly pure felspars — pegmatite veins — showing 

 that the quartz to a great extent had crystallised first, and 

 this again implies that in the one case the pressure was 

 higher than in the other. 



The chemical composition of granite seems to affect its 

 mineralogical composition only partially. The felspars 

 being silicates of alumina and potash, soda or lime ; the 

 micas being silicates of alumina and potash, lithia or 

 magnesia and iron ; the hornblendes being silicates of 

 nearly all the above with or without iron protoxide and 

 alumina, each of these minerals could be formed from one 

 and the same molten rock if it contains all these ingre- 

 dients. We find, however, that it is generally only one or 

 two of these classes of minerals which occur together, and 

 rarely (if ever) are the sub-classes mixed. It is, therefore, 

 but natural to assume that it is not so much the chemical 

 composition as the conditions of solidification, viz., 

 pressure, temperature, and rate of cooling, which deter- 

 mine the mineralogical composition, except as regards 

 quantities. See British Petrology, by J. J. H. Teall, j888. 



Two extreme conditions are conceivable. 



I. The solid rock resting on molten granite is of a low 

 specific gravity and would be a fairly bad conductor of 



