III.] THE CHEMISTRY OF METAMORPHIC ROCKS. 33 



development of crystals of staurolite and of red garnet. A 

 large amount of argillite occurs in this series ; and when al- 

 tered, whether locally by the proximity of intrusive rock, or 

 by normal metaniorphism, exhibits a micaceous mineral, and 

 crystals of andalusite ; so that it becomes known as chiastolite- 

 slate in parts of its distribution. Granitoid gneiss is abundantly 



associated with these crystalline schists The crystalline 



limestones and ophiolites of eastern Massachusetts, which are 

 probably of this series, resemble those of the Laurentian sys- 

 tem; and the coal beds an that region are in some parts 

 changed into graphite. * .... 



Large masses of intrusive granite occur among the crystalline 

 strata of the fourth series, but the so-called granites of the 

 Laurentian appear to be in every case indigenous rocks ; that 

 is to say, strata altered in situ, and still retaining evidences 

 of stratification. The same thing is true with regard to the 

 ophiolites and the anortholites of both series. No evidences 

 of the hypothetical granitic substratum are met with in the 

 Laurentian system, although this is in one district penetrated 

 by great masses of syenite, orthophyre, and dolerite. Granitic 

 veins, with minerals containing the rarer elements, such as 

 boron, fluorine, lithium, zirconium, and glucinum, are met with 

 alike in the oldest and the newest gneiss in North America. 

 These, however, I regard as having been formed, like metal- 

 liferous veins, by aqueous deposition in fissures in the strata. 



The above observations upon the metamorphic strata of a 

 wide region seem to be in conformity with the chemical prin- 

 ciples already laid down in this paper ; which it remains for 

 geologists to apply to the rocks of other regions, and thus 

 determine whether they are susceptible of a general applica- 

 tion. I have found that the blue crystalline labradorite of 

 the Labrador series of Canada is exactly represented by speci- 



[* See in this connection the prefatory note to this essay, and also Essays 

 XIII. and XV. The carboniferous age of the graphite of eastern Massachu- 

 setts has been generally assumed by geologists, thoiigh without any good 

 reason. The crystalline rocks of this region, embracing New Hampshire and 

 eastern Massachusetts, include representatives of the second, third, and 

 fourth, and probably also of the first series.] 



2* 



