MUSEUM OF COMPARATIVE ZOOLOGY. 283 



in some of the basic gabbros. No inference, it seems to me, can be 

 drawn regarding the heat to which a liquid has been exposed from 

 either the pyrognostic or thcrmo-optical characters of the minerals that 

 crystallize out on its cooling, or from the order of this crystallization. 

 Thermo-optic and pyrognostic observations will prove to be of value, 

 doubtless, when applied to those minerals that belong to the first class 

 in volcanic rocks. 



As far as my own observation has gone, and as far also as can be told 

 from the observation of others, no sulphide, no carbonate, nor any 

 hydrous mineral whatsoever, is the direct product of crystallization of 

 the fluid magma, but all these are the products of alteration and infil- 

 tration after the consolidation of the lava, as ai'e also part of the feldspar 

 and quartz, as well as all of the epidote, etc. 



Cases of envelopment are of constant occurrence among the minerals 

 of the second class, but they bear no resemblance to those cases of 

 alteration that Dr. Hunt hjis endeavored to place under the head of 

 envelopment, and which are so abundantly seen in the altered rocks. 



The alterations that take place ai'e not usually attributable to ex- 

 traneous material brought in by infiltrating waters, but are rather 

 molecular changes brought about by the percolating waters within the 

 rock mass. Except, therefore, on the weathered portions or in some 

 rocks exposed to abnormal conditions, while the mineral constituents 

 change through chemical rearrangement, the ultimate chemical consti- 

 tution of the rock, as a whole, remains nearly constant. 



The microscopic and chemical characters of volcanic rocks ai-e opposed 

 to the idea that they came from sedimentary ones, as the inclusions of 

 the first division o? the first class are nearly constant for each species, — 

 a thing which could not happen if they were derived from the fusion of 

 sediments, since a heat which has not obliterated hornblende crystals 

 was certainly not great enough to destroy all the materials of which our 

 sedimentary rocks are composed. While the chemical composition of 

 these rocks indicates a somewhat uniform origin, that of the sedimen- 

 tary ones shows generally a mixed one, as we know is the case with the 

 majority of them. 



The materials of the first division of the first class, too, in each species 

 are only of such rock species as itself or of those that we know have 

 preceded them in order of time ; as, for instance, in rhyolite we find 

 fragments of rhyolite, trachyte, andesite, and mclaphyr or diabase, but, 

 so far as I have seen, never an unaltered basalt. Rhyolite, being the 

 most refractory of the volcanic rocks, is usually the most brecciated ; and 



