354 SCIENCE PROGRESS 



then by the work of Lacroix, Doelter and Hussak, Salomon, 

 and many others. Hawes, 1 however, had already published a 

 remarkable study of the assimilation of detached blocks by 

 a granite magma ; and since that date the absorption theory, 

 which seems to many workers easily justified by field observa- 

 tion, has found fair but by no means general support. Again 

 and again the u segregation patches," which in some cases are 

 mere black masses of mica, in other cases true diorites, in others 

 composed of garnet, quartz, various felspars, pyroxene, and 

 amphibole, are found to be marginal phenomena, and to lead 

 one, by their increase in abundance in certain directions, towards 

 the contact rocks into which the granite has intruded. Again 

 and again, the streaking out of these patches is responsible for 

 the local conversion of the granite into a banded gneiss. The 

 gneiss of the valley of Boutadiol in the Pyrenees arises thus 

 from the elongation of inclusions. 2 The Transvaal granite 3 

 contains " smaller or bigger masses of the Swaziland beds . . . 

 forced off the main stock of the formation, and arranged in a 

 parallel manner, all following the direction of the main mass." 

 We thus observe in South Africa " long-drawn-out lenses of 

 quartzite, chlorite, actinolite, and other schists, swimming, as it 

 were, in a granitic magma." Observations of this kind are now 

 numerous ; yet the theory of " segregation-patches " by differen- 

 tiation holds its own. Sollas 4 was one of the first to seriously 

 question its general applicability, even in the cases brought 

 forward by J. A. Phillips ; but his suggestive criticism has borne 

 but little fruit. Corstorphine and Hatch, 5 for instance, have quite 

 recently referred the boulders of eclogite found in the diamond- 

 pipes of South Africa to a concretionary origin, because their 

 appearance is quite comparable to that of the basic patches so 

 common in many granites." 



Mention of these eclogite blocks, which are generally re- 

 garded as torn off from some deep-seated mass during the 



1 " The Albany Granite and its Contact Phenomena," Amer. Journ, of Science, 

 vol. xxi. (1881), p. 31. 



2 Lacroix, " Le granite des Pyrenees," 2 me mem. Bull. Carte geol. de la 

 France, No. 71 (1900), pp. 21 and 25. 



3 F. W. Voit, " Gneiss Formation on the Limpopo," Trans. Geol. Soc. S. Ajrica, 

 vol. viii. (1905) p. 142. 



4 Trans. R. Irish Acad., vol. xxx. (1894), p. 502, etc. ; see also Salomon, Zeitsch. 

 deutsch. geol. Gesell. Bd. xlii. (1890). 



5 Geology of South Africa (1905), p. 298. 



