PETRO(JrtAl'IIY— SANII)INTRA(^I1YTE. 517 



crystals that may be lioniblciule. Quite remarkable is tlie apjx^arance of 

 fragments of rliyolite, similar to those already described [147], [102], [ICG], 

 in which all the constituents are plainly to be seen, small sanidins and bio- 

 tite in a fine-grained <»r()nndmass. The color of these fragments is (juite 

 different from the surrounding rock nuiss. The groundmass of the rock 

 under examination is granular and has an earthy color, with darker spots 

 through it, being also full of small holes and sanidin crystals, so that is no 

 doubt trachytic. It is a rhyolitic breccia, which has inclosed fragments of 

 both the sandstone and the adjoining rliyolite. 



The sanidin-trachyte [171] from Warren Peaks has a grayish color, 

 faintly mottled with white and red spots, there being, however, no feldspar 

 crystals visible to the naked eye. In the section, its composition appears 

 to be sanidin in large, indistinctly defined crystals in a fibrous groundmass, 

 with a great number of small, prismatic hornblende crystals. 



The greater part of the rock is sanidin in a partly dusty and opaque 

 condition and almost in one mass; it is also in the form of some quite clear 

 crystals, confused in outline with the mass around them. These clear 

 crystals are sometimes twinned and have inclusions of hornblende, while 

 the cleavage lines and fissures are prominent and distinct. The opaque, 

 white crystals have a little more distinct outlines and are most easily seen 

 in reflected light. The hornblendes are very thickly scattered through the 

 rock, seemingly more plentiful between the clear sanidins. They have 

 generally a prismatic shape and ragged or forked terminations, with an 

 occasional large cr3'stal of more distinct outline. They are slightly dichro- 

 itic from brownish-green to a fine deep-green, but the cleavage lines are 

 almost obliterated. The rich green color of the small crystals makes them 

 very (;onspicuous objects in the field of the microscope. Between some of 

 the opaque sanidins was noted a colorless, transparent, isotropic mineral 

 which could not be satisfactorily determined. It may be glass or calcite. 

 Very little magnetite was seen, but considerable brown oxide of iron in 

 spots. The groundmass is almost entirely sanidin aggregated to one mass, 

 with fibrous streaks traversing the rock in every direction. These streaks 

 contain most of the hornblende columns; and in parts of the slide may be 

 observed a slight fluid-like structure around the smaller sanidins. This 



