274 ANNUAL, REPORT SMITHSONIAN INSTITUTION, 1920. 



oxide, and with or without traces of other constituents. It is also to 

 be noted that some of the essential minerals enumerated above (the 

 pyroxenes, amphiboles, micas, olivines, and the magnetites), contain 

 small amounts of manganese and titanium. From such a general 

 survey of the rock-forming minerals, then, we obtain the broad lines 

 of the chemical composition of the earth's crust as a whole. 



Another important fact concerning the igneous rock minerals is 

 that, with two exceptions, any one of them may occur in rocks with 

 any one or more of the others. The only exceptions to this are that 

 neither nephelite nor leucite is known to occur along with quartz, 

 and a partial exception is that olivine seldom occurs with quartz, and 

 never in any large amount. Discussion of this and other relations 

 between the various minerals would lead to a consideration of matters 

 outside of our present scope, and would take us too far afield. 



Each rock mineral may be present in very widely varying propor- 

 tions — from practical totality to complete absence. We know igneous 

 rocks that are composed entirely of quartz (arizonite), feldspar 

 (anorthosite), pyroxene (websterite), amphibole (hornblendite), or 

 olivine (dunite), and almost entirely of nephelite (congressite), 

 leucite (italite), or magnetite (some iron ores). Of the essential rock 

 minerals only the micas and apatite do not form the whole, or almost 

 the whole, of any igneous rock. 



From totality of any one mineral we find rocks that are composed 

 of two minerals, more that are composed of three, and still more that 

 are composed of more than three, and with the widest possible varia- 

 tions in the proportions of almost all, with the exceptions noted 

 above as to the non-coexistence of quartz with nephelite and leucite, 

 and its rarity with olivine. 



CHEMICAL CONSTITUENTS OF IGNEOUS ROCKS. 



From what has been said it would appear that the various oxides 

 (in terms of which the chemical composition of rocks is usually for- 

 mulated) may be present in widely different amounts, and, within 

 limits, this is found to be true. All of the constituent oxides have 

 very considerable quantitative ranges, but these differ much with 

 the different oxides. Their possible or recorded maxima are also 

 very different, though in every case the minimum is reached with 

 complete absence. These ranges and maxima will be stated later, 

 after a brief discussion of the oxides that go to make up the igneous 

 rocks. 



Though, as we have seen, most igneous rocks are composed of but 

 few essential minerals, and consequently of but few so-called " major " 

 oxides, yet when we come to study them in detail we find that a very 

 considerable number of different chemical constituents may be present 



