the stage the unaltered part of the crystal becomes (with crossed prisms) as nearly as possible black after it has been moved through 2912° from parallelism with a vertical line; but the other part offers no approach to complete darkness, and, though darkest when near the above position, is not all uniformly at its darkest then. This, however, may be only due to imperfect cohesion of the cleavage-planes.
Slide IV. is cut from the green-coloured variety of the older gabbro. The principal differences of this from I. and II. are:—that the conversion of the olivine is about as complete as in III., but is unattended by any red colour; the serpentine is of a pale yellowish- green tint; the dark dust is formed in the cracks of the olivine, as described above, but appears to recombine, or in some way or other is made partially to disappear at a further stage of the process; for some of the most completely changed greenish grains are very fairly clear. The augite or diallage is scarce, and, as a rule, occurs in very small grains. The resemblance of this rock, macroscopic and microscopic, to the troktolite of Volpersdorf is very remarkable, the only difference being that the felspar in the latter shows brighter colours and the serpentinous part is a shade darker in its green. The cracks traversing the felspar in all these rocks are noteworthy, as they appear to radiate from the serpentinized olivine granules and to imply some strains caused by them in the process of alteration. I may also call attention to the fact that the pyroxenic constituent seldom, if ever, shows any signs of being converted into serpentine.
The Newer Gabbro.—This is generally much coarser in texture than the others, and more decomposed. It pierces the adjacent rock with very irregular veins, which sometimes thin away to mere strings not half an inch thick, but are fairly coarse to the last. The plagioclase, often in crystals nearly an inch across, varies from a dull purplish tint to white. The diallage crystals are frequently from 12 to 34 inch across. In places red spots may be observed; these are evidently altered olivine, as in the last rock, but they are often larger, more friable, and irregular in their occurrence. Included fragments of serpentine are also seen. Spots of the green hornblendic mineral are frequent towards the exterior. As might be expected the weathered surface of the coarser varieties is much rougher than in the older gabbro. The latter rock also has a perfectly close junction with the serpentine, as if it were welded to it; while between that and the newer gabbro is more or less of a crevice.
The result of microscopic examination of a slide of one of the dark purple and green specimens is interesting. It contains abundant plagioclase, apparently labradorite, resembling that in the other gabbros, in fair condition but dull in colour, which here and there seems partially replaced by a clear isotropic pseudomorph. The green spots prove to be altered diallage: here and there small portions of the original mineral with characteristic cleavage &c. remain, but the greater part is converted into a mass of pale green actinolite, the crystals being crowded together like interlacing leaves, and
