Page:The Journal of geology (IA journalofgeology11893univers).pdf/316
contribution to the study of gneissic structure, since it shows the possibility of a part, at least, of the banding in these rocks being due to a primary banding of igneous masses through some process of segregation or through differentiation of the magma into layers. A parallel banding of igneous rocks in the neighborhood of a plane of contact has been known, but its magnitude is generally inconsiderable. The structure in the gabbro on the Isle of Skye, however, which was described by Geikie, is on a large scale, and without apparent relation to a plane of contact. No attempt was made to suggest a cause for such a mode of segregation, since the study of the locality where it is best developed is not yet completed.
Prof. Brögger's paper, "On the Genetic Relations of the Basic Eruptive Rocks of Gran, Christiana Region," presented an array of facts with regard to the differentiation of rock magmas. By means of chemical analyses and field observations he showed that basic magmas of like composition in neighboring localities had separated into pairs of magmas, which were quite unlike one another chemically; producing dissimilar pairs of rocks. This proves that a given magma may differentiate in more than one manner, according to circumstances. The entire paper is to appear in the Quarterly Journal of the Geological Society.
Mr. Harker discussed the question of magmatic concentration, or differentiation, with reference to its probable cause, and pointed out what seemed to him obstacles to the application of Soret's principle. He suggested that a more probable explanation would be found in Berthelot's principle, or that of maximum dissipativity. The applicability of Soret's principle to the differentiation of magmas is also assailed by Prof. Bäckström in an article to appear in the next number of this Journal, and the principle of liquation advocated. While it is quite probable that all of the phenomena of segregation and differentiation may not be accounted for by one law of diffusion dependent on osmotic pressure, and while this law finds its most perfect realization in the most dilute solutions, and while certain separations of rock magma may take place near the point of saturation, still it can