which, on the side facing St. Michael's, was nearly level with the sea. It was filled with water, at that time boiling, and was emptying itself into the sea, by a small stream about six yards over, and by which I should suppose it was continually filled again at high water. This stream, close to the edge of the sea, was so hot, as only to admit the finger to be dipped suddenly in, and taken out again immediately.

It appeared evident, by the formation of this part of the island, that the sea had, during the eruptions, broke into the crater in two places, as the east side of the small stream was bounded by a precipice, a cliff between twenty and thirty feet high forming a peninsula of about the same dimensions in width, and from fifty to sixty feet long, connected with the other part of the island by a narrow ridge of cinders and lava, as an isthmus of from forty to fifty feet in length, from which the crater rose in the form of an amphitheatre.

This cliff, at two or three miles distance from the island, had the appearance of a work of art resembling a small fort or block house. The top of this we were determined, if possible, to attain ; but the difficulty we had to encounter in doing so was considerable; the only way to attempt it was up the side of the isthmus, which was so steep, that the only mode by which we could effect it, was by fixing the end of an oar at the base, with the assistance of which we forced ourselves up in nearly a backward direction.

Having reached the summit of the isthmus, we found another difficulty, for it was impossible to walk upon it, as the descent on the other side was immediate, and as steep as the one we had ascended; but by throwing our legs across it, as would be done on the ridge of a house, and moving ourselves

forward by our hands, we at length reached that part of it where it gradually widened itself and formed the summit of the cliff, which we found to have a perfectly flat surface, of the dimensions before stated.

Judging this to be the most conspicuous situation, we here planted the Union, and left a bottle sealed up containing a small account of the origin of the island, and of our having Landed upon it, and naming it Sabrina Island.

Within the crater I found the complete skeleton of a guardfish, the bones of which being perfectly burnt, fell to pieces upon attempting to take them up; and by the account of the inhabitants on the coast of St. Michael's, great numbers of fish had been destroyed during the early part of the eruption, as large quantities, probably suffocated or poisoned, were occa→ sionally found drifted into the small inlets or bays.

The island, like other volcanic productions, is composed principally of porous substances, and generally burnt to complete cinders, with occasional masses of a stone, which I should suppose to be a mixture of iron and lime-stone; but have sent you specimens to enable you to form a better judgment than you possibly can by any description of mine.

VIII. On the primitive Crystals of Carbonate of Lime, BitterSpar, and Iron-Spar. By William Hyde Wollaston, M. D. Sec. R. S.

Read February 13, 1812.

WHEN I formerly described to the Society a goniometer* upon a new construction for measuring the angles of crystals, I expressed an expectation that we should thereby be enabled to correct former observations made by means of less accurate instruments. I took occasion to mention one instance of inaccurate measurement in the primitive angle of the common carbonate of lime; and I have had the satisfaction to find the necessity of a correction, in that instance, confirmed by Mons. MALUS, and admitted by the Abbé HAüy, in a work † published nearly at the same time.

It is by no means my design to detract in any degree from the merit of that justly celebrated crystallographer, to the surprising accuracy of whose measurements I could, in various instances, bear testimony. I hope, on the contrary, that in bringing forward two more observations similar to the preceding, and intimately connected with it, I shall offer what will not only appear interesting to crystallographers in general, but will be peculiarly gratifying to the Abbé Haüy.

In his Traité de Minéralogie, and again more recently in his Tableau Comparatif, the same primitive form is assigned

Phil. Trans. 1809.

+ Tableau Comparatif des resultats de la Crystallographie et de l'Analyse Chimique.

to three substances very different in their composition, to carbonate of lime, to magnesian carbonate of lime (or bitter-spar) and to carbonate of iron.

It has been objected to Mons. HAüY, that according to his method identity of form should be accompanied by identity of composition, unless the form were one of the common regular solids. For though in that case any geometrician would readily admit it to be very probable, that many different substances might concur in assuming the same form of cube, of octohedron, or of dodecahedron, &c. there does not appear a corresponding probability that any two dissimilar substances would assume the same form of a particular rhomboid of 105° and a few minutes, to which no such geometric regularity or peculiar simplicity can be ascribed.

But though so accurate a correspondence, as has been hitherto supposed to exist in the measures of the three carbonates above-mentioned, might be justly considered as highly improbable, no degree of improbability whatever attaches to the supposition, that their angles approach each other by some difference, so small as hitherto to have escaped detection. And this in fact I find to be the case.

Since the angles observable in fractures of crystalline substances are subject to vary a little at different surfaces, and even in different parts of the same surface (as is evident from the confused image seen by reflection from them), I shall not at present undertake to determine the angles of these bodies to less than five minutes of a degree. This, indeed, is the smallest division of the goniometer that I usually employ, as I purposely decline giving so much time to these inquiries, as would be requisite for attempting to arrive at greater precision,

The most accurate determination of the angle of carbonate of lime is probably that of Mons. MALUS, who measured it by means of a repeating circle, and found it to be 105° 5′. And this, indeed, is the result to which I formerly came by a different method.* If it differ in any respect from this quantity, I am inclined to think that it will more likely be found to be deficient by a few minutes, than to exceed the measure here assigned; and accordingly to differ still more widely from those angles which I am about to mention.

In the magnesian carbonate of lime, or bitter-spar, the primitive form is well known to be a regular rhomboid, as well as that of carbonate of lime, and so nearly resembling it, as to have been hitherto supposed the same. I find, however, a difference of 1° 10' in the measures of these crystals; for that of the magnesian carbonate is full 1064°, as I have observed with uniformity in at least five different specimens of this substance obtained from situations very distant from each other.

The primitive angle of iron-spar is still more remote from that of the carbonate of lime, which it exceeds by nearly two degrees. I have examined various specimens of this substance, some pure white, others brown, some transparent, others opake. That which gives the most distinct image by reflection is of a brownish hue, with the semi-transparency of horn. It was obtained from a tin mine, called Maudlin Mine, near Lostwithiel in Cornwall. By repeated measurement of small fragments of this specimen, the angle appears to be so nearly 107°, that I cannot form any judgment whether in perfect crystals it will prove to be greater or less than that angle.

In this instance the carbonate of iron is nearly pure, and so * Phil. Trans. 1802, p. 385.