28 NEW YORK STATE MUSEUM 



magmatic components as are required to alter marble to garnet 

 gneiss, notably silica and alumina, and a lesser amount of iron. 

 While this is admitted to be in general a theoretic possibility, it is 

 not considered to be the most plausible hypothesis in the present 

 instance, as will appear later. 



Moreover, there is no transition from the garnet gneiss to the 

 granite on the one hand, nor to the limestone on the other. On the 

 side facing the intrusive, the garnet rock is often cut parallel to the 

 banding by sill-like dikes of granite, which may become so nu- 

 merous as to produce a coarse-textured injection gneiss. The ad- 

 mixture is never of the intimate character so common when amphi- 

 bolite is the invaded rock (see page 68), and long narrow sheet- 

 like masses can be seen to have been ripped off the main belt of gar- 

 net gneiss, which has in the main behaved like a country rock already 

 in place, and in substantially its present character when the granite 

 was intruded. Rarely, as at the extreme eastern point of the boss, 

 the granite has partially absorbed a xenolith, resulting in the pro- 

 duction of a hybrid garnetiferous granite (see plate 4, lower figure). 

 On the side facing the limestone the garnet gneiss shows even less of 

 a transition, being succeeded immediately by the normal crystalline 

 limestone; the falling off in the proportion of garnets or other sili- 

 cate minerals of possible contact origin is abrupt, not gradual. 



As regards its mineralogical composition the peripheral gneiss, as 

 typically exposed in a ledge about one-fourth of a mile northwest 

 of the extreme southern point in the granite area, is almost identical 

 with the contorted stratum at Little River. Both are light to 

 medium gray rocks, the former being perhaps a little lighter 

 as a rule on account of the intimate injection with minute pegmatite 

 stringers. Quartz, orthoclase, biotite and garnet are the essential 

 constituents of each, accompanied by accessory pyrite and mag- 

 netite, and scapolite flakes are developed at the expense of feldspar 

 and garnet. Whatever variation there is, is in the proportion ot 

 these constituents; the peripheral gneiss has less scapoHte and is 

 relatively more quartzose than the other, as would be expected from 

 its clbser association with, and injection by, the acid material of the 

 granite intrusion. 



Further it is not noticeably, or at least not essentially, different 

 from the tapering mass of garnet gneiss bordering the northwest 

 side of the reentrant. In this case. the garnet rock is interbanded 

 with normal crystalline limestone. If the peripheral gneiss is a 

 contact rock, so is this. Nevertheless in the latter case it is difficult 

 to admit a degree of selective silicification sufficient to produce two 



