CRYSTALLOBLASTIC ORDER AND MINERALS 



513 



Fig. 19 shows a large biotite plate which has replaced a portion 

 of a garnet crystal. The hexagonal outline of the cross-section of 

 the latter is plainly visible in the mica. Between the biotite and 

 the outlying strip of garnet (Fig. 20) is a clear space occupied 

 chiefly by secondary quartz and muscovite. Two or three minute 

 grains of garnet in this space suggest that this mineral once ex- 

 isted as a complete crystal which was replaced partly by biotite 

 and partly by the quartz- 



muscovite aggregate.^ Obvi- « 



ously, biotite developed after ^ * , 



garnet. 



Ottrelite. — Although ottre- 

 lite (Fig. 21) was seen only 

 in rocks with a good flow 

 cleavage, its crystals were 

 never observed to have di- 

 mensional parallelism. Leith 

 states, however, that it may 

 occasionally show a definite 

 orientation.^ Plates of this 

 mineral sometimes wholly or 

 partly include metacrysts of 

 ilmenite and biotite. Ottre- 

 lite, then, originates later than ilmenite, garnet, and biotite. and 

 as a rule subsequent to the development of the schistosity. 



Fig. 21. — Metacnsl ot ottrelite showing 

 hour-glass twin. 25 diameters. 



SUMMARY AND CONCLUSIONS 



The facts presented in the foregoing description of microscopic 

 structures in the Narragansett Basin schists may be summarized as 

 follows : 



I. The commonest minerals in these rocks are sericite, mus- 

 covite, quartz, ilmenite, garnet, biotite, and ottrelite. Quartz and 

 sericite or muscovite form a large percentage of the composition of 



' Cf. B. K. Emerson on "skeleton crystals" of garnet: "Note on Corundum and a 

 Graphite Essonite from Barkhamsted, Connecticut," Am. Jour. Sci. (4), XIV (1902), 

 p- 234. 



= "Rock Cleavage," U.S.G.S., Bull. 2jp (1905), p. 44. 



