GEOPHYSICAL LABORATORY. 167 



represented graphically in several diagrams. All the determinations which 

 are necessary for the complete description of the crystallization of any mixture 

 have been made and are presented. The facts determined for haplobasalt, 

 haplodiorite, and so forth, are applied to their natural analogues, and it is 

 shown that there can be little reason to doubt that crystallization controls the 

 differentiation of the subalkaline series of igneous rocks. 



(20) Das ternare System Diopsid-Anorthit-Albit. N. L. Bowen. Z. anorg. Chem. (in 



press). 



A German translation of "The crystallization of haplobasaltic, haplodio- 

 ritic, and related magmas" (Am. J. Sci. (4), 40, 161-185, 1915). Reviewed 

 under No. 19 above. 



(21) Obsidian from Hrafntinnuhryggur, Iceland: Its lithophysa; and sm-face markings. 



Fred. E. Wright. Bull. Geol. Soc. Amer., 26, 255-286 (1915). 



The obsidian at Hrafntinnuhryggur, near Myvatn, Iceland, is of special 

 interest to the geologist because of the unusual opportunity it offers for the 

 study of the effects resulting from the physico-chemical conditions of cooling. 

 In this paper the formation especially of spherulitic, lithophysal, and pumi- 

 ceous structures is discussed ; certain remarkable surface markings resembling 

 the pits and grooves on moldavites are also described briefly. They were 

 produced by the etchmg eft"ect of hot volcanic emanations on fragments of 

 obsidian glass. All the evidence indicates that in the formation of lithophysse 

 gases were active. These volatile components, which were released from the 

 magma during the crystallization of the radial spherulites, attacked part of 

 the material of the spherulites ; new crystal compounds, such as tridymite and 

 fayalite, were formed which bespeak conditions of formation different from 

 those under which the original spherulites were crystallized. The pressure 

 of the liberated volatile components aided materially in the original formation 

 and subsequent enlargement of the lithophysal cavities. The general hydro- 

 static tension (external pull) resulting from shrinkage of the central part of 

 the cooling magma probably aided in this development, but it was a less impor- 

 tant factor than the inclosed gas pressing against the walls of the cavity. 



Volatile components set free during the crystallization of a spherulite may 

 either escape along minute cracks and spaces in the spherulite to its margin 

 and there form a bubble in the viscous magma, or the viscosity of the magma 

 may be such that the internal gas pressure forces asunder the spherulite. In 

 the first case the presence of the gas bubble adjacent to the spherulite hinders 

 the further growth of the spherulite at that point, with the result that the 

 spherulites with adjacent bubble cavities are well developed. In the second 

 case it is important to note that the forcing apart of the cavity was a very 

 slow process. The first rupture took place when the spherulite was small and 

 the rigid walls of the cube were forced back and grew accordingly as crystalli- 

 zation proceeded. The edges of the cube were thin and in contact with the 

 magma, which, however, was probably so thick and viscous that less resistance 

 was offered to the slow forcing apart of the walls of the spherulite than to the 

 formation of gas bubbles adjacent to the spherulite. It is not possible to 

 determine, from the scant evidence at hand, the several quantitative factors 

 which are essential to the formation of the type of lithophysal cavities described 

 in this paper. 



Evidence is also presented which shows clearly that the deeply etched sur- 

 faces on irregular fragments of the obsidian are the result of etching by hot 

 circulating solutions from which large amounts of hyalite were deposited. 

 Minute crystals of alunite were also deposited during a later stage of circulat- 



