manner fomewhat fimilar to what is practifed with rivers. This he was afterwards told had been done with fuccefs during the great eruption of Etna in 1669: that the lava was directing its courfe towards the walls of Catania, and advancing very flowly, when they prepared a channel for it round the walls of the town, and turned it into the fea. A fucceffion of men, covered with theep-fkins wetted, were employed to cut through the tough flanks of lava, till they made a paffage for that in the centre, which was in perfect fufion, to dif gorge itself into the channel prepared for it. But this, it is evident, can only take place in fmall ftreams of this burning matter; with that above mentioned it would have been impoffible. It hath been also obferved of the lavas of Etna, that they do not conftantly fall down to the lowest places, but will sometimes afcend in fuch a manner as to make the valleys rife into hills. On this Sir William Hamilton has the following note: "Having heard the fame remark with regard to the lavas of Vefuvius, I determined, during an eruption of that volcano, to watch the progrefs of a current of lava, and I was foon enabled to comprehend this feeming phenomenon, though it is, I fear, very difficult to explain. Certain it is, that the lavas, while in their most fluid ftate, follow always the laws of other fluids; but when at a great distance from their fource, and confequently encumbered with fcoria and cinders, the air likewife having rendered their outward coat tough, they will fometimes (as I have feen) be forced up a fmall afcent, the fresh matter pushing forward that which went before it, and the exterior parts of the lava acting always as conductors (or pipes, if I may be allow ed the expreffion) for the interior parts that have retained their fluidity from not being exposed to the air."

(4.) LAVAS, OBSERVATIONS ON, BY SIR WILLIAM HAMILTON. From 1967 to 1779, this gentleman made many curious obfervations on the lavas of Vesuvius. He found that they conftantly formed channels in the mountain as regular as if they had been made by art; and that, whilft in a ftate of perfect fufion, they continued their courfe in those channels, which were fometimes full to the brim, and at others more or lefs fo according to the quantity of matter thrown out. Thefe channels, after fmall eruptions, were generally from 2 to 5 or 6 feet wide, and 7 or 8 in depth. They were often hid from the fight by a quantity of scoriæ that had formed a cruft over them, and the lava having been conveyed in a covered way for fome yards, came out again fresh into an open channel. Our author informs us, that he had walked in fome of these fubterraneous galleries, which were exceedingly curious, the fides, top, and bottom, being exceedingly smooth and even others were incrusted with what he calls very extraordinary fcoriæ, beautifully ramified white falts, in the form of dropping ftalactites, &c. On view ing a ftream of lava while in its fluid ftate in May 1779, be perceived the operation of it in the channels above defcribed in great perfection. After quitting them, it spread itfelf in the valley, and ran gently like a river that had been frozen, and had maffes of ice floating upon it. The wind hap. VOL. XII. PART II.

pening then to fhift, our traveller was fo incommoded by the smoke, that the guide proposed to crofs it, which was inftantly put in execution without any other inconvenience than the violent heat with which the legs and feet were affected. The cruft was so tough, that their feet made no impreffion upon it, and the motion fo flow that they were in no danger of falling. This circumstance, according to Sir William, points out a method of escape, fhould any perfon happen to be inclosed betwixt two lavas, but ought never to be tried except in cafes of real neceffity; and indeed, if the current of melted matter was very broad, must undoubtedly be attended with extreme danger, both from the heat of the upper cruft and the chance of its breaking and falling down with the paffenger into the burning liquid below. That which Sir William Hamilton croffed was about 50 or 60 feet broad. Having paffed this burning ftream, our travellers walked up along the fide of it to its very fource. Here they faw it boiling and bubbling violently up out of the ground, with a hiffing and crackling noife like that which attends the playing off an artificial fire-work. An hillock of about 15 feet high was formed by the continual splashing up and cooling of the vitrified matter. Under this was an arched hollow, red-hot within, like an heated oven; the lava which ran from it being received into a regular channel raised upon a fort of wall of fcoriæ and cinders, almost perpendicularly, of about the height of 8 or 10 feet, and much refembling an ancient aqueduct. On quitting this fountain of lava, they went quite up to the crater, where, as ufual, they found a little mountain throwing up ftones and red-hot scoria with loud explofions; but the smoke and fmell of fulphur was fo intolerable, that they were obliged to quit the place with precipitation. By the great eruption in August 1779, the curious channels above mentioned were entirely defiroyed, the cone of the mountain was covered with a firatum of lava full of deep cracks, whence continually iffued a fulphureous smoke that tinged the fcoriæ and cinders with a deep yellow, or fometimes white tint. The lava of this eruption appeared to be more perfectly vitrified than that of any former one he had obferved. The pores of the fresh lava were generally full of a perfect vitrification, and the fcoriæ themselves, viewed through a magnifying glass, appeared like a confufed heap of filaments of a foul vitrification. When a piece of the folid lava had been cracked in its fall, without feparating entirely, fibres of perfect glafs were always obferved reaching from fide to fide, within the cracks. The natural fpun glass which fell in fome places along with the ashes of this eruption, and which has likewife been observed in other places, he is of opinion, must have proceeded from an operation of the kind just mentioned; the lava cracking and feparating in the air at the time of its emiffion from the crater, and by that means spinning out the pure vitrified matter from its pores or cells; the wind at the fame time carrying off the filaments of glafs as faft as they were produced. Our author obferved a kind of pumice tone fticking to fome very large fragments of the new lava. On close inspection, however, he found PPPP

that

that this fubftance had been forced out of the minute pores of the folid lava itself; and was a col. lection of fine vitreous fibres of flaments confounded together at the time of their being preffed out by the contraction of the large fragments of lava in cooling, and which had been bent down. wards by their own weight. "This curious fubftance (fays he) has the lightness of a pumice, and refembles it in every respect, except that it is of a dark colour." When the pores of this lava were large, and filled with pure vitrified matter, the latter was fometimes found blown into bubbles on the furface; probably by the air which had been forced out at the time the lava contracted it felf in cooling; and from these thin bubbles it appeared, that this kind of volcanic glafs has much the fame transparency with our common giafs bottles, and like them is of a dirty yellow colour; but when large pieces of it were broken off with a hammer, they appeared perfectly black and opaque. In the lava of this eruption it was obferved, that many detached pieces were in the fhape of a barley-corn or plum-ftone, fmall at each end, and thick in the middle. Some of thefe did not weigh above an ounce; but others could not be less than 60 pounds. Our author took them to be drops from the liquid fountain of fire, which might naturally acquire fuch a form in their fall. There were alfo many other curious vitrifications, different from any he had feen before, mixed with this huge fhower of fcorie and mafies of lava.

(5.) LAVAS, REMARKS ON, BY M. HOUEL, AND DR VAN TROIL. In treating of Mount Etna, M. Houel makes mention of a piece of lava which, after having been once ejected by the volcano, was fwallowed up, and thrown out a fecond time. The intense heat to which it was then fubjected, had fuch an effect upon it, that it appeared all full of chinks to a confiderable depth, and which ran at right angles to one another. He had also an opportunity of obferving to great advantage fome of the hollow channels formed by the lavas of Etna, fimilar to thofe described by Sir William Hamilton, but on a much larger fcale. Here the great eruption in winter of 1755 had overturned, in a vertical direction, an huge tube of this kind for the length of half a mile. The tube itself appear ed to be compofed of enormous maffes, fomewhat refembling planks; each two feet thick, and 12 or 15 in breadth, continued in a straight line through the whole of that space. At the fame time, by the action of the lava, alkind of walls had been formed, from 10 to 16 feet in height, and curved at the top. Some of thefe walls appear rolled together like paper; and M. Houel is of opinion, that thefe various appearances on the furface of the lava when cooled, muft have arifen from particles he terogeneous to the real lava, and which detach themselves from it, rifing to the furface under a variety of forms proportioned to the space of time taken up in cooling. Thefe crufts are formed of different kinds of fcoriæ and dirty lava, mixed with fand or afhes. At the fame place are found alfo great numbers of fmall pieces, like thofe of ice heaped upon one another, after having floated for fome time on a river. Beneath thefe the pure lava is met with, and which has evidently been in a ftate of perfect fufion. This is extremely dense; and

66

hills near Vicenza. M. Dolomieu confiders the chemical analyfis of lava as but of little account. When subjected to the force of fire a second time,↑ they are all of them reducible to the fame kind of glafs; from which it has been concluded, that all volcanic products have been formed of the fame kind of materials, and that the fubterraneous fire has always acted on and variously modified the fame kind of ftone. But an analyfis by fire, he juftly observes, is of all others the most fallacious. The fubftances are all fufible, and we have no proper methods of measuring the intensity of our fire; fo that the fame fubftance which to-day may come out of our furnaces untouched, may tomorrow be found completely altered, even though the fire employed should not appear to us to be any more violent than the former, Analyfes by different menftrua have not been more fuccessful. Sir T. Bergman has indeed analyfed fome lavas with acids; and gives, with aftonishing precision, the following refult, viz. that an hundred parts of lava contain 49 of filiceous earth, 35 of argillaceous earth, four of calcareous earth, and 12 of iron. Thefe experiments, however, our author obferves, give us no information with regard to lavas in general. They only fhow the compofition of the particular fpecimens that he tried; and even after the defcriptions that he has given, we are a good deal at a lofs to discover the fpecies of lava which he abjected to analyfis. "It would be as ridiculous (fays M. Dolomieu) to apply this analysis to every volcanic product, as it would be to believe that the component parts of a fiffile rock were the fame with thofe of every rock compofed of laminæ or thin ftrata." For these reafons he is of opinion, that, in order to understand the nature of lavas, we fhould confider not only that of volcanoes themselves, but of the bases on which they reft. Had this been done, we would have found that the volcanic fires generally exift in beds of argillaceous fchiftus and horn-ftone: frequently in a fpecies of porphyry, the gluten of which is intermediate betwixt horn-ftone and petrofilex; containing a large quantity of fchori, feldt fpar, and greenish quartz or chryfolite, in little rounded nodules. Thefe fubftances, he tells us, would have been found in those mountains which are called primitive, and in ftrata buried under beds of calcareous stone; and among other things, would have convinced us that the fluidity of lavas does not make them lose the distinctive characters of their bafes. In the mountains call. ed primitive, thofe rocks which are affigned as the bafes of the more common lavas are found intermixed with micaceous ones, with gneiss, granite, &c. and they generally reft on mafles of granite. Hence lavas muft confift of all these matters, and the fire must act upon them all whenever it meets with them. Our author as conftantly observed, that volcanoes fituated at the greatest distance from the centre of the chain, or group of mountains on which they are established, produce lavas of a more homogeneous compofition, and lefs varied, and which contain moft iron and argillaceous earth. Thofe, on the contrary, placed nearer the centre, are more diverfified in their product, containing fubftances of an infinite variety of different kinds, Pppp2

The