unlike a decomposed serpentine. Careful examination, however, shows that we have here, highly metamorphosed and entangled in the serpentine, another mass of sedimentary rock, which has once consisted of lenticular bands of a more sandy character, in a mud whose mineral composition somewhat resembled that of hornblende.
The first stage has been the conversion of the former into a kind of granulite, the latter, probably, into a hornblende schist. Torn off and squeezed by the igneous mass (now serpentine), the harder bands have been crumpled up, and in some cases forced into the softer, which at last, by slow action of water, have been converted into a rotten chloritoid and rather serpentinous schist. The serpentine around is also rotten near the junction. In some places the two rocks are so altered by addition and subtraction of mineral constituents that it is almost impossible to fix their precise boundary; still I am convinced that the above explanation is the correct one.
I have examined a slice from the most granitoid part of this rock; and, though highly altered, it quite confirms my view. It consists of quartz, felspar, orthoclase, and some plagioclase, with a little of some variety of magnesia mica. The felspar is full of microlithic alteration products. In the quartz are a good many minute cavities and shapeless microliths—also some microliths of larger size, which may be apatite. Here and there a piece of the felspar (it is not very characteristic) is full of minute branching empty cavities or microliths (I rather think the latter), which would certainly be quoted as canal systems by the opponents of Eozoon. I have seen something similar in a granitoid rock from Holsteinborg (Greenland), but at present can do no more than record the occurrence, hoping to return to the subject on a future occasion. It is the nearest approach to an organic structure that I have ever seen[1].
From Kennack Cove we proceeded along the cliffs to the headland of Karak Clews, about a mile distant. So far as I saw, serpentine continues all the way; and just before reaching the point called Carn Sparnack a small quarry affords some very pretty varieties (no. 11)—one a rich red mottled with dark olive green, the other claret-colour with similar markings, both containing in the green part small crystals of greenish bronzite with a submetallic lustre. Karak Clews is a bold headland formed by the extremity of a great elongated mass or broad dyke of coarse gabbro, which, according to the map, is about two miles long and a furlong broad. The headland terminates in a narrow ridge leading down to a precipitous mass of rock running some little way out to sea; the general direction is nearly N. and S., the mass further inland for most of its course having a N.W. strike.
There are very few places more instructive than this to the student of igneous rocks. Broadly speaking, the ridge consists of gabbro, in which so many pieces of serpentine are entangled that it would
- ↑ Perhaps the structure erroneously described as a Laurentian organism, 'Nature,' xiv. 8. 68. may have been this.
