ROCK SPECIMENS FROM THE ISLAND OF SOCOTRA. 
279 
bubbles), prismatic microliths of very pale green colour, and numerous colourless hair¬ 
like microliths, scattered irregularly throughout the grain. The resemblance of the 
specimens to examples from the Hebridean series of Scotland, the Laurentian series of 
America, and other Archsean rocks is very striking. [4115], one of a large series from 
the right bank of the upper part of the Gollonsir Valley, contains a fair proportion of 
black mica instead of hornblende. Apatite can be recognised, and the small, almost 
colourless, prismatic microliths already mentioned, some of which may be referred to 
this mineral, though I am of opinion that the majority are different. There is no 
microcline, but a good deal of plagioclase, probably both albite and oligoclase. [4124] 
contains hornblende with a black mica and sphene. ( h) granitoid gneisses, consisting 
mainly of quartz and felspar, belonging to the group for which I have proposed the name 
granitoiditeA In these the quartz and felspar correspond with those described above, 
and the principal difference is that the rock is mainly composed of these two minerals, 
with an occasional grain of iron peroxide and a flake or two of iron glance or a ferru¬ 
ginous mica, hornblende or epiclote. ( c ) hornblendic rocks, which we shall notice in 
the next paragraph, and (d) an impure quartzite from near Kami, north-east of the 
Haggier range. This rock consists of quartz, decomposed felspar, epidote, and 
pefbaps a little hornblende, with, in parts, a good deal of magnetite. 
Diorite and other hornblendic rocks. 
I have made this division somewhat vague for two reasons : one that in rocks of 
igneous origin the hornblende is not unfrequently of secondary origin, having replaced 
augite or diallage, so that the rock is more properly a uralitic or hornblendic diabase 
than a true diorite : the other, that the means of making a thorough study of some 
difficult examples of these rocks—corresi:>onding with those already mentioned in the 
gneisses and granites—are only now being obtained by me.t From the hill near the 
opening of Gollonsir Valley is a series of specimens, which, according to Professor 
Balfour, come from what appears to be a dyke in the Archsean series. Some are 
coarser in texture than others. Two of these [4009] and [4005] have been examined 
microscopically, of which the former was in situ, the latter from a loose block. They 
consist of a decomposed felspar, in which, however, the remains of the twinning 
characteristic of plagioclase can occasionally be discerned, of hornblende, black mica, 
and opacite, with some apatite, and a few grains of epidote. The hornblende is green in 
colour and exhibits very characteristic cleavage. The mica is sometimes altered to a 
greenish mineral, and often contains needles and grains of opacite, generally arranged 
parallel with the principal cleavage planes. In [4053] “from the slope of the hill 
* Quart. Jour. Geol. Soc., vol. xxxv., p. 322. 
f It must be remembered, in excuse for this ignorance, that it is of no use to purchase specimens for 
study of these difficult cases, or, as a rule, to trust the statements which one finds in print. The student 
must collect his own specimens, and to do this it is necessary to visit distant localities and expend much 
time and money. Hence difficulties are but slowly removed. 
