318 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1920. 



regions in the United States and their gravity anomalies. Also, the}' 

 appear to be too concordant to be the result of chance, so that we are 

 justified in assuming that the two are causally related and that the 

 theory of isostasy is thus justified. 



SUMMARY. 



After brief consideration of the interior of the earth, the general 

 characters of igneous rocks are discussed, and the presence of water 

 vapor and other gases in the magma, and its analogy with a salt 

 solution, are pointed out. In the discussion of the mineral characters 

 of rocks, stress is laid on the fact that the number of essential rock- 

 forming minerals is very small. These are mostly silicates of Al, Fe, 

 Mg, Ca, Na, and K. Any two or more of these minerals (with two 

 or three exceptions) may occur together and in any proportions. 



The chemical characters of igneous rocks are summarized and 

 the ranges and maxima of the various constituents are given. The 

 average igneous rock is considered, and after some discussion of 

 the sources of error involved in the calculation, a new average in 

 terms of oxides (based on 5,159 analyses) is given. The average 

 rock is shown to be approximately a granodiorite. 



The average composition of the earth's crust in terms of elements 

 is also given. Twelve elements (O, Si, Al, Fe, Ca, Na, K, Mg, Ti. 

 H, P, and Mn) make up 99.61 per cent of the crust. 



The elements are referred to two main groups in the periodic 

 table: (1) The petrogenic elements, characteristic of and most abun- 

 dant in the igneous rocks, of low atomic weight, and occurring nor- 

 mally as oxides, silicates, chlorides, and fluorides; (2) the metallo- 

 genic elements, rare or absent in igneous rocks, but occurring as 

 ores, of high atomic weight and forming in nature " native " metals, 

 sulphides, arsenides, bromides, etc., but not primarily oxides or 

 silicates. The suggestion is made that beneath the silicate crust of 

 petrogenic elements is a zone essentially of nickel-iron and beneath 

 this a central core of the metallogenic elements. This vertical dis- 

 tribution is in accord with Abbot's views as to the distribution of the 

 elements in the sun. 



In igneous rocks and minerals the elements show a correlation, 

 in that certain of them are prone to occur with others, and a similar 

 limited correlation is apparently true of the animal and vegetable 

 kingdoms. 



The idea of "comagmatic regions" — that is, the distribution of 

 igneous rocks in regions of chemically related magmas — is discussed, 

 and some of these are briefly described. 



