the planes of principal cleavage; until at last the whole crystal is converted into an aggregate of small crystals of hornblende. From the general appearance of the mineral I take it to be actinolite, which as a non-aluminous variety of hornblende would most readily be formed from ordinary diallage. In many cases it is almost fibrous in structure; then it is paler in colour and feebly dichroic. The specimens which I have examined have not shown me any olivine; yet, as we shall presently see, this mineral abounds in the gabbro further north. This absence seems strange: I have, however, some reason to think that it, too, in this case, has been replaced by actinolite[1].
The "granite vein" headland is a prominence on a rather larger one; beyond this is a little chine and another small headland. Over this space serpentine predominates; but fragments of hornblende schist are included, and intrusive dykes and veins of gabbro are common. Every step shows something new and interesting. At one place the gabbro on the left, and the serpentine on the right, make an almost vertical junction in the cliff. The former includes a long strip of hornblende schist in an upright position; the latter assumes near the junction a rather fissile character. Often it would be hard to say whether the serpentine or the gabbro were the intruder; but here and there may be found conclusive evidence that the latter is the newer. The gabbro is coarse, the diallage crystals being some- times very large, one composite specimen being about 6″ × 2″ × 2″. The foliated structure mentioned above is often seen; and I observed that, as a rule, it was best developed where the gabbro intersected the hornblende schist, especially where it had passed between two masses along the plane of bedding. As the whole mass cooled, these already solid schists would doubtless produce a definite pressure on the crystallizing rock between them at right angles to their bounding surfaces, and so determine its structure. I have seen the mica crystals in a granite vein which had broken through angular fragments of a schist lying at right angles to normals from their surfaces, and have often observed that on the outside of a granite vein the mica plates tend to lie parallel to the surface.
The next headland exhibits both serpentine and gabbro intrusive in schist, with a large felspar vein. But it is needless to carry these details further; so I will select one more section for description, the last which can be reached from the shore. Here a lofty cliff of serpentine is shattered by veins of gabbro, one of which, about ten
- ↑ I have good reason to believe that this replacement of augite or diallage by a form of hornblende has taken place in several of the Welsh "green- stones." It is not precisely a paramorphic process like the formation of uralite, nor a pseudomorphic, because the form of the original crystal is often lost, but a replacement of a mineral by another, which, if not really a dimorphous form of the first, is very closely allied to it. I conceive the change to have been mainly brought about in the "wet way." I have seen the same change in the gabbros from Mont Colon (mentioned above), and the Matterhorn, also in the hypersthenite of Penig (Saxony). MM. Poussin and Renard have observed it in the Ardennes rocks (Roches Plutoniennes de la Belgique, &c. P . 69).
