Co.MI'OXKXT MIXEHALS. 75 



undergoing' the process of metasoraatic recrystallizution, and which tlius form 

 a link between typical sandstones and the more highly altered rocks in which 

 the clastic origin is not evident on mere inspection. In these rocks minute 

 bacillar augitcs make their appearance in newly formed aggregates limited 

 by the outlines of the original clastic grains. There is clearly a tendency 

 to parallelism and to grouping of these augite cr3stallites, and the evidence 

 points irresistibly to tlie conclusion that under favorable circumstances large 

 solid crj'stals of augite form by the union of these smaller masses. In a 

 considerable number of instances these microlites are actually united in close 

 groups bounded by crystallographic outlines. The usual occurrence of gar- 

 net in metamorphic rocks indicates an entirely similar process of aggrega- 

 tion. It is quite impossible to ascribe any but an authigenetic origin to 

 these characteristic occurrences of augite in newly formed aggregates arising 

 from the alteration of clastic grains, nor is such a formation surprising, since 

 the artificial reproduction of augite by the action of heated water under 

 pressure upon appropriate mineral mixtures is a well known phenomenon. 

 A fine example of a partially formed augite crystal in an altered sandstone is 

 shown in Fig. 2, page 88. 



In the more fully crystallized metamorphic rocks augite is often very 

 abundant. It is of lighter tint under the microscope than the ordinary 

 bamboo-colored augites of eruptive rocks, is monochroitic, and extinguishes 

 at high angles. It readily passes over into uralite, chlorite, epidote, and ser- 

 pentine. The uralite often has a bluish tint approaching that of glaucophane. 

 There is a marked tendency in the larger augite crystals to the development 

 of the orthopinacoidal cleavage, and in a few of the rocks the pyroxene is 

 well developed diallage. 



Hornblende — Tliis mineral occurs in the recrystallized and recrystallizing 

 rocks in two forms. Brown hornblende forms much in the same way as 

 augite and is observed sometimes in the same slides with it. Groups of 

 liornblende microlites also show common crystallographic outlines; a case 

 of this kind is shown in Fig. 3, page 89. 



Either minute chemical differences or certain physical conditions seem 

 to regulate the preponderance of the one mineral over the other, so that the 

 fully recrystallized rocks are divisible with some sharpness into hornblendic 



