sometimes forming fringes almost like tufts of grass; they make their way sometimes up cracks in the felspar, and appeal even to form as endomorphs. The difference in dichroism between these and the yet unchanged diallage is well exhibited on rotating the polarizer without the analyzer. I find no olivine or grains of serpentine; but a speck here and there resembles the latter; the outline, however, of some of the bunches of actinolite resembles rather that of olivine than diallage. As there certainly has been olivine in the rock, this confirms my suspicion that this mineral sometimes, instead of forming serpentine, contributes to the actinolite. When the latter mineral is formed, it is remarkable how greatly the cohesion of the constituents of the rock seem to have been disturbed; for it not only invades the edges of the felspar in needle-like fringes, but also occurs in blades and tufts anywhere in the heart of that mineral. In fact, where this process has been carried far, it is only now and then that the outline of the original crystals can be traced.
Lastly, as to the trap-dykes, I regret to say that I cannot give precise information with regard to all of these, as I have found it difficult in one or two cases to identify those which I had observed in different visits. The notes made in 1873 and 1875 were taken with a rising tide; and so, as other matters were of still greater interest, I rather hurried over these dykes. In 1876 the tide was very favourable, and I observed more dykes than I had noticed previously; but still, as there was so much to do that I did not collect a specimen from every one, I have a little difficulty in identifying those collected on former occasions with the dykes noted on this. However, the following results are fairly correct. There are two varieties of rock in the dykes:—one, coarse enough to be obviously a crystalline rock, much like an anamesite; the other very compact, of a dark purplish colour, looking almost like a dull serpentine. Thin veins of serpentine often show in the latter; and films of the same not seldom coat joint-faces. The dyke seen in the cliff may be taken as a type of the former. On microscopic examination it proves to be a felspar basalt,—the plagioclase being, as a rule, clear, though with polarized light dull in colour, the augite little changed, and dull green spots replacing probably the olivine. There are also grains of magnetite, with brown stains from one of the other ferric oxides; and cracks are overspread by an infiltrated serpentinous mineral.
The second specimen (collected in 1873 from a dyke "cutting gabbro and serpentine") is quite different. It has an isotropic matrix, which remains black, with crossed prisms, as the stage is rotated, full of rather acicular crystals of actinolite, very pale clear green with ordinary light, slightly dichroic with the polarizer, and often showing brilliant colours with the analyzer. Here and there are patches of the brown hornblende granulated with magnetite (?), like that described from the shore north of Kynance (p. 889). Some of these occur in such a way as to produce a strong impression that they too have been partly replaced by actinolite. One or two portions of unchanged plagioclase crystals remain in the matrix; and
