16 PEOF. T. G. BONIfEY ON SOME RESULTS OF 



mica is chieflj^ brown, and rather more commonly outlines the 

 quartz grains, which are a little larger, but which contain abundant 

 enclosures of the mica. Needles, possibly of sillimanite, are 

 abundant, especially piercing into the quartz grains; there are 

 a few grains of tourmaline, one or two of epidote, and perhaps a tinj^ 

 garnet or two. Iron-oxides as usual. Here, then, as has been said, 

 the mica is both disseminated and to some extent aggregated in 

 bands, but these have at most been compressed, not puckered. 



No. 2 is a quartzite containing numerous microliths of mica. 

 Its grains are often about '04" in diameter, but smaller occur ; they 

 are irregular in outline. Flakes of white (rare), brown and 

 greenish mica, perhaps sometimes of a chlorite, granules of ferrite 

 and opacite, and one or two possibly of epidote, are scattered over 

 the field. The rock obviously is a quartzite, but there is this pecu- 

 liarity: — In an ordinary quartzite mineral inclusions are compara- 

 tively rare and obviously were present in the materials of the original 

 sand. Adventitious or accidental minerals usually occur between 

 the constituent quartz grains, and not seldom the distinction between 

 the original grains of the sandstone and the secondary quartz which 

 has made it a quartzite can be observed. But in this rock from 

 Brittany there is no sure indication of an original nucleus, for the 

 little mica flakes are constantly included in the grains. jSTow and 

 then a whole granule or an inner space in a grain seems compara- 

 tively free, and one or two also of such grains are cracked, as I have 

 seen happen with quartz grains which have been exposed to heat ; 

 but generally the mica is so uniformly disseminated that either the 

 rock must originally have been a very fine-grained one, or its con- 

 stituents must have been reduced to such a plastic condition that a 

 large amount of molecular movement became possible. 



Nos. 4 and No. 5 may be more quickly dismissed, as they are 

 nearer tj^pes which have already been noticed. The former rock 

 consists largely of mica (chiefly the rich brown variety so cha- 

 racteristic of contact-metamorphism) and small granular quartz. 

 The rock is foliated, but not banded, the mica flakes being parallel 

 with the original stratiflcation, and the bed itself less disturbed 

 (where it is exposed) than the other, so that here there is little 

 " waving " and no strain-slip cleavage. There are the usual vaguely 

 defined spots of granular andalusite, associated with quartz and a 

 little mica, and here and there a more perfect crystal. No. 5 

 is a similar rock, the chief diff'erence being that the andalusite is 

 better developed, and fairly complete crystals are common. They 

 are practically colourless, and frequently occur in cruciform twins. 

 Evidently the granular stage represents an arrested development. 

 If the mass were kept long enough with the requisite environment the 

 separated granules of andalusite would draw together, " elbowing 

 aside," as it were, most of the quartz and mica (though sometimes 

 these cannot be expelled, but must be incorporated), so as to form 

 well-developed crystals. 



I may remark here — what, no doubt, has often been observed 

 before — that the difference in the mode of occurrence of the varieties 



