Before treating of the subject before me, I must describe its history. At a soirée of the Royal Society on April 25, 1877, I showed a number of specimens illustrating a new class of optical properties, applicable to the identification of minerals. These greatly interested the late Sir G. G. Stokes, Bart., and led to a considerable amount of correspondence between us ; and he communicated a paper to the Royal Society on the mathematical part of the subject, and I sent a short one on my apparatus and observations. This was admitted to be very imperfect, since the crystals examined, though showing the general facts perfectly well, had not been cut so as to be suitable for such quantitative measurements as could be compared with Stokes's theoretical determinations.

Among the zeolite specimens from the neighbourhood of Bombay in the Oxford University Collection are several of the well-known type, showing pale green crystals of apophyllite associated with white stilbite and white silky needles labelled as ‘poonahlite’, on a matrix of purplish-brown amygdaloid, with a layer of a bluish-green, earthy mineral lining the amygdules and underlying the zeolites.

The name poonahlite was originally given by H. J. Brooke, in 1831, to a mineral occurring on specimens which he had received from Henry Heuland. He states that the poonahlite formed slender crystals accompanying apophyllite from Poonah, and was at first taken for mesotype or needlestone, but that it had a prism-angle of 92° 20′.

Previous experiments made by myself in the Oxford Mineralogical Laboratory on the crystallization of potash-alum had shown that a drop of supersaturated solution Of potash-alum, when left by itself, crystallized almost invariably in birefringent sphere-crystals, which represent a more soluble and less hydrated variety of that substance. It was further suggested (1) that the spontaneous crystallization of that particular substance always takes place as spherulites ; and (2) that this spherulitie growth is indicative of a sudden change in the solution, which is passing from the metastable state for ordinary alum (in which ordinary octahedra, when introduced, grow by themselves) to the labile state (in which the alum crystallizes as a rectangular network of needles).

Mr. T. V. Barker has lately shown that the parallel growth of isostructural substances upon one another is conditioned by similarity of molecular volume. This very important conclusion invites further investigation from this new point of view, and Professor H. A. Miers suggested to me, while working in his laboratory last summer, that I should make experiments on the parallel growth of sodium nitrate on the pseudo-rhombohedral mineral barytocalcite.

The first experiments, which were made on crystals of barytocalcite in the Oxford collection, were unsuccessful, the crystals employed being small and somewhat impure. Later, I received, from the Museum of Practical Geology in London, some very good and perfectly pure crystals of barytocalcite from Blagilt, near Alston, in Cumberland.

The continuation of the mapping of the aureole of metamorphism round the Bodmin Moor granite has led to the discovery of fresh occurrences of idocrase, developed in the altered, impure limestones. The failure to find this mineral hitherto is now clearly due to the former absence of artificial openings in suitable localities, for the mineral has so far not been found in recognizable form in natural exposures, owing to the rapidity with which these calcareous rocks decompose.

Altered calcareous rocks containing idocrase in considerable abundance have recently been laid open in a small quarry on the west side of the stream at Tvenuth Mill, which is about 400 yards south of the granite margin in the well-known Caradon area.

Until quite recently, in spite of its characteristic features, andalusite had not been met with in the younger detrital rocks of Britain ; a few years ago, however, I remarked on its relative abundance in the older Pliocene sands of St. Erth and St. Agnes in the west of Cornwall. Its presence in these deposits, in which it is associated with pink garnet and cyanite, is easily explained by the close proximity of the areas of deposition to the metamorphic aureoles of the west of England granitemasses, in which andalusite of identical character is exceedingly common. It also occurs in fair quantity in the Pliocene sands of Lenham in Kent accompanied by cyanite, and together with cyanite and garnet in the Red and Norwich Crags and the Chillesford Beds in the east of England.

Vernadsky has pointed out that in certain cases a twin-crystal appears to be a more stable form than the simple polyhedron, as proved by Scacchi's observation that twins sometimes exceed in magnitude and regularity simple crystals growing in the same solution. He suggests that an explanation can be found in the fact that the surfaceenergy may be tess for a twin than for a simple crystal of the same volume. The suggestion is a valuable one, but Vernadsky applies no quantitative test to his theory.

Although more than fifty years have passed since Guiscardi published the first description of guarinite, the true nature of this mineral has never yet been ascertained. As regards the crystalline form, Guiscardi's first idea was that the crystals of guarinite are orthorhombic (he had himself reckoned the ratio a : b = 0.9896 : 1 ; see literature, nos. 3 and 5); but, having observed crystals with a tetragonal habit, he finally decided to describe the mineral as tetragonal-hemihedral. V. v. Lang by his optical investigations appeared to place the orthorhombic symmetry of guarinite beyond doubt, and Des Cloizeaux later published a note on a combination observed by him.

When a mineral fragment is brought under the influence of an electrostatic charge, its behaviour depends chiefly upon the conductivity of the mineral and that of the surface on which it is lying. This fact can be effectively demonstrated by placing fragments of amber (sp. gr. about 1.1) and pyrite (sp. gr. about 5) on a metallic surface, such as that of a copper plate, and noting their behaviour when a rod of vulcanite, which has been electrified by rubbing with cat's skin, is brought near them. Under these conditions a small fragment of dry amber, weighing not more than one milligram, is not attracted to the rubbed rod, although the rod be brought almost in contact with it; whilst a large piece of pyrite weighing as much as 150 milligrams jumps visibly from the plate with the rod held at some distance from it.