THE EEUPTIVE ROCKS. 311 



granite nearly 4 feet thick border the larger granite veins, or are inter- 

 calated in the amphibolite, and at last the whole greatly resembles the 

 "syenite" — here a diallage-biotite-gabbro. 



The eastern hornblendic band comes south as a sharp synclinal fold of 

 hornblende-schist, embracing a band of mica-schist, and becomes changed 

 to resemble the tonalite, while the inclosed schist continues far sovith into 

 the tonalite, metamorphosed into a highly crystalline fibrolite-schist. 



On the west side of the river broad bands of liornblende-schist and 

 limestone can be traced to contact with the tonalite, and isolated fragments 

 appear on the latter across its Avhole lengtli. Farther west, beyond the 

 influence of the hornblende-schist, the tonalite changes to biotite-granite, 

 and still farther west to muscovite-granite. Biotite-granite becomes the 

 prevailing rock of the batholites, where they are contained in the Conway 

 garnetiferous schists. 



Two circumstances are very peculiar in the distribution of the rock. 

 The first is the barrenness of the great central masses as compared with the 

 richness in minerals of the smaller bordering dikes ; the other, the degree to 

 which the granite is confined to the mica-schist and avoids the gneiss which 

 bounds it east and west and in all probability underlies it. This association 

 is so marked that when a narrow strip of the Conway mica-schist appears 

 east of the river in Northfield there are associated with it dikes of pegmatite 

 having secondary veins of albitic granite canying cleavelandite, spodumene, 

 columbite, and beryl. 



The western line of Pelliam and its prolongation northward and south- 

 ward through Leverett and Belchertown is the eastern boundary of the 

 disturbed area, and in the gneiss east of it granite dikes are few and unim- 

 portant, rarely, as at the Monson quarry, caiTying garnet and beryl. 



I have given much thought to those theories which would trace the 

 granite down to the subjacent gneiss which, entirely melted, is supposed to 

 have been "extravasated" into the subjacent rocks; but I find no good 

 reason for inferring any intimate relation between the gneisses of the region 

 and the pegmatite. Many chemical and microscopical peculiarities of the 

 gneiss militate against that relation, such as the large content of quartz, 

 calcium, and iron and the small content of potassium, the uniform distribu- 

 tion of biotite and titanite, and the absence of tounnaline and musco%ate. 



Further (exception being made of the small secondary veins with 



