ELEMENTS IN IGNEOUS ROCKS—WASHINGTON. 291 
be essentially alike in all their features, as is the case with that of the 
eastern Mediterranean and that which extends from the Andes to 
Alaska. 
IV. THE CORRELATION OF THE ELEMENTS. 
The existence of petrographic provinces is a broad phase of the 
distribution of the elements among igneous rocks, the distribution 
being essentially a spacial one and the evidence resting almost 
entirely 6n the relative proportions of the most abundant elements. 
But apart from this spacial distribution there is evident, also, a cor- 
related variation among the elements; that is, a tendency for certain 
ones to increase or decrease, to be relatively abundant or not, accord- 
ing to the presence or absence of others. The causes of this behavior 
are obscure and apparently complex. In part they may be probably 
referred to certain fundamental relations among the elements, as 
shown by the pericdic classification and chemical affinity; in part to 
the effect. of certain physico-chemical laws leading to the mutual 
segregation of elements affected similarly; and possibly in part to 
the degradation of some of the elements, as indicated by recent ex- 
periments by Ramsay. But any discussion of the causes is outside 
the province of this paper, in which we can only deal briefly with 
some of the facts of distribution. 
The study of these mutual relations among the elements in igneous 
rocks is of recent date, and has been made possible, especially so 
far as the rarer elements go, only by the completeness and accuracy 
of modern chemical analyses of rocks. Such analyses supplement 
the evidence afforded by study of minerals, mineralogical associa- 
tions, and ore deposits, and, deaiing as they do with what must 
be regarded as the ultimate source of the ores, are of the highest 
significance and importance. In the following brief discussion, 
therefore, stress will be laid on the evidence afforded by rock analy- 
ses, with some reference to chemical mineralogy, while ore deposits, 
as being more technical and better known to the mining engineer, 
will be alluded to only occasionally. 
Considering first only the most abundant elements, study of the 
igneous rocks in general shows that silica, alumina, soda, and potash 
tend to increase or decrease together, though not always at the same 
rate; while, on the other hand, the iron oxides, magnesia, and lime 
tend to vary together and in general inversely as the preceding con- 
stituents. The more siliceous rocks almost invariably show rela- 
tively high alumina and alkalies and low iron oxides, magnesia, and 
lime, leading to the common presence in abundance of the alkali 
feldspars and the comparative paucity in calcic feldspars and the 
ferromagnesian minerals, which tend to increase rapidly with dimi- 
nution in the silica content. Highly siliceous rocks which contain 
