On Rocks and Minerals from the Karakoram Himalayas. 473 



crushing of the coarser. The matrix has a sub-crystalline aspect, 

 and is variable in character. In one part sub-angular grains appear 

 to be, as it were, set in a matrix composed of small scales of mica 

 (mostly white, but some green), and of a chalcedonic material. This 

 condition closely resembles that described in some Huronian con- 

 glomerates, and probably results from the alteration of a felspathic 

 grit. In other parts a large grain of quartz is occasionally seen, 

 commonly fairly well rounded, and the ground-mass consists of a 

 mixture, resembling that already described, of minute mica with 

 felspathic-looking granules, enclosing larger crystals of green and 

 brown mica and crystals of a second mineral. These are fairly deve- 

 loped, rather elongated prisms (varying from about 1 : 4 to 1 : 8) 

 up to 0'04 in. in length, rather full of microlithic enclosures, very 

 pale yellowish-green with transmitted light, with dichroism almost 

 imperceptible, and moderately bright-coloured with crossed nicols. 

 The mineral shows a rather irregular transverse cleavage with occa- 

 sional hints of one parallel to the sides of the prism, and extin- 

 guishes straight or nearly so with the latter. Though it has a general 

 resemblance to epidote, the cleavage parallel with the above-named 

 sides is not so marked as usual. It is, however, more like this 

 mineral than andalusite, with specimens of which we have compared 

 it. The mica, which occurs in flakes, is dichroic, changing generally 

 from a light straw colour to a brownish-green. It has formed after 

 the epidote, and as it frequently borders the fragments of marble it 

 may possibly be a lime-mica. It produces the impression that it is 

 slowly eating up the ground-mass. Of other minerals present in 

 parts of the slice, rutile, sometimes in geniculate twins, is rather 

 abundant, and is included in both the epidote and the mica. From 

 its mode of occurrence a derivative origin seems probable. A few 

 granules of iron oxide, apparently limonite, occur, also a rounded 

 grain of zircon, and one small crystal of brownish tourmaline, 

 secondary in origin. 



(2.) Remarks on Certain Specimens of Interest. 



We proceed next to describe more particularly those rocks already 

 mentioned which are more specially interesting. First of these are 

 some rocks consisting almost wholly of hornblende. 



Hornblendites. One of them from a fallen block on the west side of 

 the Astor Valley above Dashkin is a dark green, rather friable rock, 

 which consists chiefly of hornblende, mostly in porphyritic crystals 

 about f inch long. More than one variety of this mineral appears on 

 microscopic examination : one is blue-green (for rays along the C axis), 

 and a yellower green (for direction at right angles) ; the other, in larger 

 crystals, is strongly dichroic (a straw colour, b dark bronze-green, c 



