442 PEOF. T. G. BONNET ON DEVITEIFICATION [NOV. I903, 



though a chemical analysis proves that there must be a considerable 

 amount of free silica present. Tor instance, in a spherulite from 

 the Yellowstone Park, the felspar is in it roughly as 169 to 100 ; 

 in one from Corriegills shore (Arran), as 194 to 100 ; and in one 

 from Boulay Bay, as 182 to 100. 1 In the first case Mr. Parkinson 

 suggests a very probable explanation of the invisibility of the silica ; 

 but in the second, third, and other cases of secondary devitrification 

 that I have examined, I have observed that under a high power one 

 of the apparently-felspathic grains or patches, instead of appearing 

 homogeneous, as in ordinary primary devitrification, either seems 

 to be speckled with a material which acts but feebly on polarized 

 light, or sometimes even suggests the presence of an almost ultra- 

 microscopic graphic structure, so that I suspect the apparent felspar 

 to be really a compound of that mineral and quartz. 



Eutectic Structures. 



Mr.Teall pointed out, in his most suggestive Presidential Address, 2 

 that, in accordance with Mr. J. E. Stead's experiments on alloys, 

 eutectic compounds are especially favourable to the formation of 

 spherulitic and graphic structures ; and I have spent some time in 

 calculating from rock-analyses the proportion of felspar to quartz, in 

 order to see whether the ratio which obtained in one case (163 : 100) 

 generally held good. The results, however, as might be inferred 

 from the figures quoted above, are not very satisfactory, and I am 

 not sanguine of success, with our present data. In a rock the 

 problem usually is much less simple than in an alloy, for it consists in 

 finding the eutectic mixture, not of two minerals, but of three or four; 

 since at least two species of felspar must often be present, besides 

 quartz and water, while in regard to the last the amount now 

 present affords no indication of what was originally dissolved in 

 the magma. We might obtain trustworthy results from analyses 

 containing only potash or soda, but these are extremely rare, and 

 even though the substitution of a small quantity of the one alkali 

 for the other may not alter the species of the felspar, yet it may 

 produce some effect on the crystallizing temperature of the mineral, 

 and must, I think, modify that of the eutectic. But that any eutectic 

 is particularly favourable to the formation of spherulitic and peg- 

 matitic structures, may now, I think, be taken for granted; also that 

 the coarseness or fineness of this structure is a question of tempera- 

 ture and time 3 (which is to some extent applicable even to secondary 

 devitrification). Moreover, as I hope that I have made clear, a 

 rectilinear or a curvilinear type of growth is probably dependent on 



1 That is to say, the free silica is always more than one third of the mass. In 

 each case the felspar is of two species, orthoclase and albite— the latter pre- 

 dominating in the second case. 



2 Quart. Journ. Geol. Soc. vol. lvii (1901) p. lxxv. 



3 As this makes it a function of two variables we must remember that (within 

 limits) an increase in the one may compensate for a decrease in the other. 



