The classification of the rocks which best harmonizes the field-evidence and the detailed study of their mineralogical composition, is one that arranges these volcanic protrusions into two series. In the one, the orthoclase is sanidine, and the rocks range from the most vitreous pitchstone through perlitic and spherulitic varieties to rhyolite ("quartz-trachyte"). In the other series, which embraces by far the largest proportion of the whole, the orthoclase is always turbid, and in this respect as well as in many others the rocks remind us rather of ancient eruptive masses than of those which have appeared in Tertiary time. They range from flinty felsitic varieties, which are obviously devitrified glasses, through different textures of quartz-porphyry into granophyre, and finally into granite. As I have been unable to recognize any essential difference of structure and composition between these acid Tertiary rocks and those of far earlier geological time, I give them the names which no petrographer would hesitate to apply to them if they were of Palæozoic age. It has long appeared to me that these rocks furnish conclusive evidence of the misleading artificiality of any petrographical nomenclature in which relative antiquity is made an essential element of discrimination.
Granite.—That true granites form part of the Tertiary volcanic series of the British Isles has now been completely established. They occur as bosses and sills which have been intruded into the gabbros and all older rocks. They are thus proved not only to belong to the Tertiary period, but to one of the latest phases of its volcanic history. But besides these granites, the relative age of which can be definitely fixed, there occur others which, standing alone and at some distance from the basaltic plateaux, can only be inferentially classed in the Tertiary series. To this group belong the granite masses of the Isle of Arran and the Mourne Mountains in north-eastern Ireland.
Taking first the unquestionably Tertiary granites which occur as bosses and intrusive sheets, we have to note that the more coarsely crystalline granophyres are hardly to be distinguished externally from granite. As the dark ferro-magnesian constituent of these rocks was generally believed to be hornblende, they were called by the older petrographers "syenite"; that is, granite with hornblende instead of mica. The peculiar micropegmatitic groundmass, which constitutes the distinguishing feature of the granophyres, may occasionally be observed so reduced in amount as only to appear here and there between the other minerals, which are grouped in a granitic structure. From this condition, one step further carries us into a true granite, from which all trace of the granophyric character has disappeared. Such gradations may be traced even within short distances in the same boss of rock. Thus, in the hornblende-biotite-granite boss of Beinn-an-Dubhaich, Skye, a thoroughly granitic arrangement of the component minerals is observable in the centre, while a specimen taken from near the edge on the shore of Camas Malag shows the development of a granophyric groundmass. But, though the large bosses are usually somewhat coarsely crystalline in the centre, and tend to assume finer felsitic textures around their borders, as was observed long ago by Oeynhausen and Von Dechen,[365] the granitic structure is sometimes exhibited even at the very edge, and not only so, but in the dykes that protrude from the bosses into the surrounding rocks. Thus the Beinn-an-Dubhaich mass, at its margin in Camas Malag, sends a vein into the surrounding limestone, but though more close-grained than the main body of the rock, this vein is neither felsitic nor granophyric, but truly granitic in structure.
[365] Karsten's Archiv, i. p. 89.
So far as I have observed, the true granites contain a brown mica and also a little hornblende, both visible to the naked eye, but generally somewhat decomposed. These rocks are thus hornblende-biotite-granites (amphibole-granitites of Rosenbusch). They may be defined as medium-grained aggregates of quartz, orthoclase (also plagioclase), biotite and hornblende, with sometimes magnetite, apatite, epidote and zircon. Dr. Hatch found that in some instances (Beinn-an-Dubhaich) the quartz contains minute inclusions (glass?), bearing immovable bubbles with strongly-marked contours; while in others (Beinn-na-Chro, Skye) this mineral is full of liquid inclusions with bubbles, sometimes vibratile, sometimes fixed. He remarked that the quartz and felspar have consolidated almost simultaneously, but that in some instances (Marsco, Glen Sligachan) there are isolated roughly idiomorphic crystals, of a white, less turbid orthoclase, which belong to a slightly earlier consolidation than that of the more kaolinized felspar of the rest of the rock.
The granite of the island of Arran, in the Birth of Clyde, which is here included in the Tertiary volcanic series, has long been recognized as consisting of two distinct portions, an eastern or coarse-grained, and a western or fine-grained variety. The latter sends veins into the former. These granites contain orthoclase, plagioclase, quartz and dark mica, the quartz being often idiomorphic with respect to the felspar, and a tendency towards a micropegmatitic structure being sometimes observable. A distinguishing characteristic of the Arran granite is the cavernous or drusy structure which it presents, the cavities being often lined with well-crystallized orthoclase and smoky quartz.[366] The granite of the Mourne Mountains in Ireland closely resembles that of Arran. Its druses, with their beautifully terminated minerals, have long been well known.
[366] See Mr. Teall's British Petrography, p. 328.
Microgranite.—This term is applied to certain intrusive masses, which megascopically may be classed with the quartz-porphyries and felsites, but which microscopically are found to possess a holocrystalline granitic groundmass of quartz and orthoclase, through which are scattered porphyritic crystals of the same two minerals, sometimes also with plagioclase, augite, magnetite or apatite. Rocks of this type do not appear to be abundant. They occur as dykes and bosses, but occasionally also as sheets. I have collected them from Skye, Rum and Ardnamurchan.
Granophyre.—Under this name may be grouped the large majority of the acid rocks which play an important part in the geology of the West of Scotland. They are typically developed in the islands of Mull and Skye. Generally pale grey or buff in colour, they range in texture from the true granites, into which, as above stated, they graduate, to exceedingly close-grained varieties like the felsites of Palæozoic formations. In the great majority of them the micrographic intergrowth of quartz and felspar, known as micropegmatite, is their conspicuous structure, and even constitutes most of their substance. They may thus be classed generally as granophyres, in the sense in which this term is employed by Rosenbusch, but without his limitation of it to pre-Tertiary rocks.
The specific gravity of these rocks has been determined from a series of specimens by Mr. A. Harker to range from about 2·3 among the felsites to 2·7 among the granites. No chemical analyses of these rocks have yet been made, but they have been subjected to microscopical examination, and their general structure and composition are now known.