220 SCIENCE PROGRESS 



illustrated on a large scale in glass-making processes. N. L. 

 Bowen (" The Significance of Glass-making Processes to the 

 Petrologist," Journ. Wash, Acad. Sci. 1918, 7, 88-93) shows, 

 for example, that immiscibility between silicate and silicate 

 does not occur, at least in glass melts. Two layers may separate 

 in the glass pot, but their persistence is entirely due to slowness 

 of diffusion, and the stratification may be eliminated by stir- 

 ring. Glass often shows a density and composition gradient 

 (illustrated by the famous Morozewicz experiment), but this 

 is shown to be due to differential melting, with the sinking of 

 the heavier and more easily fusible constituents at an early 

 stage, and the rising of the difficultly-fusible silica grains at a 

 later stage. 



An important paper by W. J. Miller (Bull. Geol. Soc. Amer. 

 191 8, 29, 399-462) gives the results of an elaborate study of 

 the great anorthosite mass of the Adirondacks, which figured 

 largely in Bowen's recent work on the problem of the anor- 

 thosites. Although, like Bowen, Miller believes that the 

 anorthosite originated by the settling of mafic minerals in a 

 gabbro magma, his view differs from Bowen's in several par- 

 ticulars. The anorthosite did not arise by the subsequent 

 settling of labradorite crystals ; furthermore, the associated 

 syenite-granite masses are intrusive into, and distinctly later 

 than, the anorthosite, and hence are not direct differentiates 

 as advocated by Bowen. Miller shows, moreover, that the 

 anorthosite was at one time effectively molten, and not merely 

 a mass of precipitated plagioclase crystals. The term silexite 

 is proposed by Miller (" Pegmatite, Silexite, and Aplite of 

 Northern New York," Journ. Geol. 19 19, 27, 28-54), for the 

 bodies of pure, or nearly pure, silica of igneous or aqueo- 

 igneous origin, occurring as dykes or segregations within or 

 without their parent rock. The silexite of the area dealt with 

 occurs in a granite, and represents a very siliceous facies of 

 pegmatite, into which there is a complete transition. Both 

 silexite and pegmatite are believed to be segregations which 

 developed while the granite still preserved a considerable 

 degree of fluidity, and continued to form until the granite 

 almost or completely solidified. 



In his recent work on the geology of Fiji (op. cit. supra, pp. 

 97-145), Foye gives brief petrographical descriptions, accom- 

 panied by chemical analyses and quantitative mineral deter- 



