204 CHEMICAL DIVERSITY OF GRANITES. 



sometimes in addition to these there are perceptible quantities of 

 oxide of manganese, phosphoric acid, lithia, and fluorine, while not 

 infrequently the protoxide of iron, or even all the iron, may be absent, 

 as may be the magnesia and the water. Even in British granites the 

 percentage of every constituent is very variable ; thus the silica ranges 

 ^rom as low as 55*20 in the granite of Ardara to as high as 80^24 in 

 the granite of Croghan Kinshela ; so that, judged by this test, the 

 Ardara rock might be termed basic, while the Croghan Kinshela rock 

 is typically acidic. 



The alumina varies from 11-14 P er cent, at White Gill, Skiddaw, 

 to 20 per cent, in the granite of Glen in Donegal. The peroxide of 

 iron ranges from '23 at Botallack, to 7*3 in some of the granites of 

 Leinster ; while the protoxide of iron which is so frequently absent, 

 amounts sometimes to upwards of 2 per cent. The lime varies from 

 J per cent, in some of the Cornish granites, to upwards of 5 per cent, 

 in some of those from Donegal. The magnesia, which may be a 

 mere trace, amounts to 3^ per cent, in the granite of Ardara. Soda 

 may be but -J per cent, in Cornish granites, and 5^ per cent, in some 

 of the Leinster rocks. Potash is less than \ per cent, in one of the 

 Leinster granites, and more than 8J per cent, in the granite of Chy- 

 woon Morvah in Cornwall. The manganese never quite amounts to 

 i per cent., and the water is never more than 2 per cent. If we 

 further included the elvans in our survey, we should find in some 

 respects yet greater variations, since the percentage of silica may fall 

 as low as 47, and the water and magnesia rise above 6 per cent. each. 



Turning from the examination of these rocks to discover deposits 

 out of which they might have been formed by metamorphism, we have 

 not to seek far. We may take such clays as are associated with the 

 coal strata, or used in the potteries, and find their composition to 

 include the same chemical elements as granite, though from the 

 analyses available, we may not be able to exactly parallel the granite 

 of any particular district. These clays consist of silica, alumina, 

 peroxide, and occasionally protoxide of iron, lime, magnesia, potash, 

 and sometimes soda, and water ; while occasionally there are traces of 

 phosphoric acid, organic matter, and various elements locally distri- 

 buted. If we further compare the percentages, we shall find the 

 silica to vary between 44 per cent, and 77 per cent., while the 

 alumina varies between 14 per cent, and 34 per cent. The lime may 

 be less than \ per cent., or as much as 3 per cent. ; the magnesia 

 rarely exceeds i per cent. The potash ranges as high as 3 J per cent., 

 and the soda, which is not often detected, is in some cases | per cent., 

 while the water may vary from i per cent, to 25 per cent. These 

 facts may be seen sufficiently set out in the accompanying tables, 

 which give the chemical compositions of certain clays. I'irst we 

 compare the granite of Creetown with a slate from Prague analysed 

 by Zirkel, and detect but little more difference than might occur in 

 different samples of the same rock. The Cornish granite of Redruth 

 may be compared with the American triassic sandstone of Cotton- 

 wood. If this granite is compared with the clay of Hillscheid or 



