It is not to be expected that all of the volcanoes of the Andes are in exactly the same phase of differentiation, which they undoubtedly are not, but in a general way they have not progressed beyond the limits of olivine-bearing pyroxine-andesite and dacite, and may be considered as having the basalt and rhyolite phases before them. They thus show themselves as comparatively young, or perhaps middle-aged. It will be observed, however, that rhyolite occurs in small amount in Ecuador, as shown by analyses 14 and 15 in Table II.

The chemical similarity of the magmas of the Andes with those of general occurrence in western North America, which are fairly represented by the volcanic rocks of the Yellowstone National Park, is seen upon constructing a diagram of the molecular variation of the essential constituents in a manner employed by the writer and also by Dakyns and Teall in discussing the differentiation of molten magmas.

In Table I are given the chemical analyses of perfectly fresh. rocks from Colombia, published by Dr. Küch. The general molecular character of the magmas is shown by diagram I. In the lowest part of the diagram the molecular variation of all of the essential constituents is represented, that of silica being given by the abscissas, the zero point being some distance to the left. In the middle part of the diagram alumina, soda and potash are separated from the lime, magnesia and iron-oxide, which are given by themselves in the uppermost part of the diagram, in order to avoid the confusion of lines in the lowest part. The iron is represented as ferrous oxide. The character of this diagram is quite similar to that of the diagrams for the rocks from the Yellowstone National Park in the paper on the origin of

1Iddings (J. P.). The mineral composition and geological occurrence of certain igneous rocks in the Yellowstone National Park. Bull. Phil. Soc., Washington, 8vo. Washington, 1890. Vol. 11, pp. 191–220.

100.57 100.16 100.ΟΙ 99.02 98.20 99.47 99.84

100.03

100.24 100.37 100.80 101.47 100.07