214 F. P. Mennell — Constitution of Igneous RocTcs. 



To obtain a true average we now multiply each silica percentage 

 by the volume of the rock represented, the products being added up 

 and divided by the total volume. The result is a silica percentage 

 of 69-45 for the whole. Whatever composition is assigned to the 

 granites, their percentage will, in fact, be very near the general 

 average, which, it will be noticed, is much higher than Clarke's 

 figure. Granite, indeed, as previously pointed out, is not one extreme 

 of the igneous series, but is substantially the average igneous rock. 



In order not to lay too much stress on observations made in a single 

 area, I have made similar calculations with regard to others of which 

 I happen to have reliable data. England would evidently give very 

 much the same result, but I do not possess a map which discriminates 

 between the various classes of igneous rock with sufficient accuracy. 



There are, however, no large masses of basic rock to set off against 

 the granites of Cornwall, Devonshire, Cumberland, and Leicestershire. 

 The Scottish Highlands would be an excellent area from which to 

 draw conclusions, and it is evident that the granite masses are 

 numerous and large as compared with other classes. My own native 

 colony of Yictoria (Australia) is perhaps not altogether typical, as the 

 exposures of igneous rocks are much smaller than those of the 

 sediments, and the southern portion is largely covered by flows of 

 tertiary basalt. However, taking an area of twenty thousand square 

 miles north of latitude 37°, i.e. between Maldon, Mansfield, and Omeo 

 and the j^Tew South Wales border, the following rough estimate may be 

 made regarding the 2,810 square miles of igneous rocks exposed: — 



Total . 2,715 



In this case the silica percentage works out at 68 '4. Canada is 

 a country where the great development of igneous rocks makes the 

 results as interesting as in the case of Tropical Africa, and I have 

 therefore roughly worked out the data afforded by the West Kootenay 

 (British Columbia) sheet of the Canadian Geological Survey map. 

 Its southern boundary is the United States border, while to the north 

 I have stopped a few miles short of the margin of the sheet at 50° 

 N. lat. in order to avoid inconvenient calculations. The area dealt 

 with is in round figures 5,500 square miles, of which about 4,260 

 show igneous rocks. Among the rocks mapped as metamorphic there 

 are granites, but they are not coloured separately, and are therefore 

 left out of calculation. Of the others there are three types of granite 

 which may be grouped together as averaging 70 per cent, of silica all 

 through. There are also three series of volcanic rocks which appear 

 from the descriptive notes appended to the map to be in each case 

 mainly andesitic in character. There are also exposures of serpentine, 

 and of a ' monzonite ' stated to be related to the gabbros. Tabulating 

 these as before, we have, to a depth of a mile — 



