The gases on which the trials were made, and the results of which I am now to give, were those of Edinburgh and Leith, and also that of a town in the north of Scotland, to which I went solely for the purpose of experimenting on this subject, as I knew that a good opportunity there presented itself, owing to the great distance to which the gas is sent. The illuminating power was ascertained chiefly by the chlorine test, because I found it difficult, nay in some cases impossible, to carry an experimental metre with me, and hence one great advantage of this test.

Trials on the Leith Gas.-9th Dec. 1842. The gas was made from Arniston coal, which was found to yield from 510 to 539 cubic feet per cwt., that is, taking the average, 464 from the 100 lb. In my trials with my own apparatus the quantity amounted to 475 feet. By different trials with the chlorine test, the condensation at the works, on an average, amounted to 13 per cent. On the evening of the same day, the gas from the same company was tried at Niddry Street, Edinburgh, a distance of upwards of two miles from the manufactory, and the condensation there was, on an average, 12. In this instance, the flow of gas in the pipes, to supply the immediate neighbourhood, was not great, and I was, therefore, anxious to have other trials made. I accordingly selected George Street, because I knew that, from the works at Leith to this situation, the pipes are laid in a direct line; and, as the flow of gas was considerable, I was convinced that that tried at the works, and in the other place on the same day, would be the same.-16th Dec. At the works the condensation by chlorine was, on an average, 13.7. The gas was from Arniston coal, which had been in use for two days; on the evening of the same day the condensation at George Street amounted to 14.

Trials in the North of Scotland.-26th Dec. 1842. The gas was made from a mixture of two parts of Lesmahago, one of Monkland, and one of Torryburn coal. The average of all the trials at the manufactory, indicated by the chlorine, was 14.75. By the Aldcock burner, the length of flame, by one inch of pressure, was 4.1 inches. The specific gravity was 590. On the same day, the gas was tried at the distance of nearly

six miles. By the chlorine the indication was 13.25. The specific gravity was 653, and the flame by the Aldcock burner, at one inch pressure, was four inches. In this trial the difference is very trifling; and as the gas on which I operated at the works was that made from Saturday evening to Monday morning, on which day the experiments were conducted, there is every reason for believing, that that tried at the works and at the distant point was the same.

Trials on the Edinburgh Gas.-15th Nov. 1842. This gas, tried at my laboratory in Surgeon Square, a distance of nearly a mile from the manufactory, indicated, by the chlorine test, 13.7. On the 9th Dec., it was 13; on the 19th, it was 13.66, giving an average of 13.45; on the 20th Dec., the gas was tried at the works, and on an average was found to be 13.16; on the evening of the same day, the gas at George Square, a distance of upwards of a mile, was 14; on Friday, the 23d Dec., at the works, it was 14; at George Square, on the evening, it was 14. The average of all the trials at the works was 13.5; at a distance, it was 14. I have again, more lately, tried this gas at a greater distance. On the 20th Dec. 1843, the gas at the works indicated 13; at the distance of upwards of three miles on the Glasgow road, the indication was 12.5.

In the above experiments, the illuminating power of the gas is, in some cases, the same at the works and at a distance; in three, it is rather lower, and in one it is rather higher, at the distant point. Taking the average of all the trials, the results are

The difference, as shewn by the above table, is so very trifling, as to be altogether unworthy of notice; consequently proving that the distance to which the gas is conveyed, after

it is properly purified, makes little or no difference in the illuminating power. Of course, in making this remark, it must be understood that it does not at all bear on the loss that may be sustained, by the diminution in the quantity of gas, by leakage in the pipes, and otherwise; a point which it is difficult to decide, where a company disposes of its gas, partly by meter, and partly by contract according to the time, but which does not affect the consumer, who has to pay for the gas that passes through his meter, or by the time his burners are in use.

On the Production of Flames in Volcanoes, and the consequences that may be drawn therefrom. By M. LEOPOLD PILLA.

THE question, whether volcanic phenomena are accompanied with flames, is, in my opinion, of so much importance in the science of the earth, that the attention of natural philosophers cannot be too strongly drawn to it; doubts are still left in the mind respecting it, which ought to be removed. The greater number of men of science who have devoted themselves to the study of volcanoes, deny that there is any manifestation of this phenomenon in volcanic eruptions; and they in general think, that what the vulgar, and even many writers, have called flames, is nothing else than the reflection of the light produced by the burning substances on the walls of the craters, and on the column of smoke which issues from them. I may be permitted to quote, in reference to this, the following passages from most respectable authors.

"An unanswerable proof of the insufficiency of this hypothesis (the disengagement of hydrogen gas in the eruptions of Stromboli) is the following. When the bubbles of the boiling lava burst by the escape of the enclosed gas, who does not see that if this gas consisted of hydrogen, it ought, at that moment, to become inflamed on the surface of the lava? Now, it is very certain that in no eruption do we ever see the slightest flame on the lava.”*

"The different metallic combustibles and metalloids may decompose water, in proportion to the degree of affinity they possess with the oxygen of the latter, and give rise to the series of acids and oxides which appear in volcanoes. We ought, however, to observe, that the hydrogen, on leaving its state of combination, never reaches the apertures which vomit fire, and which are in communication with the atmo

*Spallanzani, Viaggi alle Due-Sicilie, t. iii. cap. 21.

spheric air, because we have never seen flames, either on the crater while in action, nor on the surface of the running lavas."*

"One of the consequences of Davy's hypothesis, and perhaps the most important, would be the disengagement by the craters of volcanoes of an enormous quantity of hydrogen, either free, or combined with other principles, if it really be the water which, by its oxygen, induces volcanic fires. It does not appear, however, that the disengagement of hydrogen is very frequent in volcanoes. Although, during my abode at Naples, in 1805, with my friends Messrs Alexander de Humboldt and Leopold de Buch, I had an opportunity of witnessing at Vesuvius frequent explosions, which threw out broken lava to a height of more than 200 yards, I never perceived any inflammation of hydrogen."+

"The brilliant light reflected by the clouds of aqueous vapour and ashes suspended over the cone, produces this appearance, which is so often described under the erroneous denomination of flames, in the accounts of volcanic erruptions by inexperienced observers, who have no acquaintance with science."‡

Sir Henry de la Beche, when describing the phenomena of an eruption of Vesuvius, expresses himself thus: "The solid substances thrown up by the volcano, appeared like a numerous discharge of red balls, while the light of the burning mass in the interior of the crater, reflected sometimes in a very vivid manner by the column of vapours above, produced, to the view of an observer, placed at a certain distance, the appearance of flames, which there are strong reasons for believing to be illusory. It is at least very certain, that almost all the cases of this nature which have been cited, have no other cause than a reflection of light, which varies in intensity with the activity of the volcano."§

"The vapours illuminated by the incandescent substances which fill the craters, or cover their walls, have often been taken for flames. But this illusion has been combated by a great number of observers, who have affirmed that true flames never issue from the crater of a volcano.”||

I myself, also, drawn away by so many authorities, when I commenced to observe the phenomena of Vesuvius, said, "We ought to take care in such cases not to mistake the luminous radiation produced by the stones and incandescent scoriæ for flames, an error into which many people have fallen ;"¶ and, in fact, when I wrote this sentence, I had never observed flames from Vesuvius.

*Covelli, Storia dei fenomeni del Vesuvio, § xc.

↑ Gay-Lussac, Reflexions sur les Volcans (Ann. de Chim. et de Phys. t. xxii.) Poulet Scrope, Considerations on Volcanoes, cap. 2, § 1.

§ Manuel de Geologie (art. Volcans en activité).

|| Brongniart, Des Volcans et des terrains volcaniques (Art. du Dict. d'Hist. Nat.).

I forbear to quote in this place passages from more ancient authors. It is true that many of them, in describing volcanic phenomena, sometimes mention flames; but it is evident that they paid no particular attention to this phenomenon, and did not distinguish it from the luminous reflection produced by the burning substances.

We perceive, therefore, that geologists, up to the present time, have been of opinion that volcanic eruptions have never been accompanied with flames. This opinion, however, is altogether erroneous. At least, I think that I am entitled to make this affirmation positively in reference to Vesuvius.

Let us commence by stating the facts which may support this general proposition. Of all the phenomena which I have had occasion to observe in regard to Vesuvius for twelve years, I consider those which I am about to notice, and an acquaintance with which I owe to a fortunate accident, as the most important.

On the night of the 2d June 1833, I was within the crater of Vesuvius, in order to observe the phenomena of an eruption, which was approaching its close. In the centre of the crater there was one of those cones of scoria which are formed and disappear with such marvellous rapidity; it was the largest cone I had ever observed, so much so, that it might have been called the little Monte Nuovo. There was a large funnelshaped aperture on its summit, through which the explosions took place. At the moment of which I speak, these had become less frequent, and succeeded each other at intervals of from three to four minutes. This circumstance made me desirous to mount upon the cone, in order to observe the great phenomenon of explosions, which I never before had it in my power to do near at hand, and immediately over the opening. I had frequently, indeed, observed eruptions from the summit of la Punta del Palo; but the distance from the proper opening, or what may be called the air-hole of the volcano, the walls of scoria with which it is usually surrounded during eruptions, the smoke, the ejection of stones, and other circumstances, had always prevented me seeing distinctly what was going on in the volcanic orifice. At the moment of explosion I ascended to the edge of the cone, along with my courageous guide, who shared in my curiosity to observe the appearances. The interior of the opening was almost entirely free from smoke; a small quantity only issued from different points in the walls. This fortunate circumstance enabled me to see very distinctly all the parts of the crater, and everything that was going on there. The bottom of the funnel was open; it lay immediately under my eyes, at a depth of about eighty metres; its circumference was nearly twenty yards; the whole of its burning interior was visible. The view of the phenomena which accompanied the explosions was inexpressibly magnificent. They consisted of the following:

A loud subterranean noise, and a violent shock, announced the explosion; immediately after, and almost at the same time, the mouth opened VOL. XXXVI. NO. LXXII.-APRIL 1844.