Since I wrote the above, I have frequently converted a small portion of water in an ounce phial into a weak _Aqua fortis_, by repeated mixtures of common and nitrous air; throwing in alternately the one or the other, according to the circumstances; that is, as long as there was a superabundance of nitrous air, suffering the common air to enter and condense it; and, when that was effected, forcing in more nitrous air from the bladder, to the common air which now predominated in the phial--and so alternately. I have wanted leisure, and conveniences, to carry on this process to its _maximum_, or to execute it in a different and better manner; but from what I have done, I think we may conclude that nitrous air consists principally of the nitrous acid, phlogisticated, or otherwise so modified, by a previous commenstruation with metals, inflammable spirits, &c. as to be reduced into a durably elastic vapour: and that, in order to deprive it of its elasticity, and restore it to its former state, an addition of common air is requisite, and, as I suspect, of water likewise, or some other fluid: as in the course of my few trials, I have not yet been able to condense it in a perfectly dry bottle. NUMBER VI. _A Letter from_ Dr. FRANKLIN. Craven Street, April 10, 1774. Dear Sir, In compliance with your request, I have endeavoured to recollect the circumstances of the American experiments I formerly mentioned to you, of raising a flame on the surface of some waters there. When I passed through New Jersey in 1764, I heard it several times mentioned, that by applying a lighted candle near the surface of some of their rivers, a sudden flame Would catch and spread on the water, continuing to burn for near half a minute. But the accounts I received were so imperfect that I could form no guess at the cause of such an effect, and rather doubted the truth of it. I had no opportunity of seeing the experiment; but calling to see a friend who happened to be just returned home from making it himself, I learned from him the manner of it; which was to choose a shallow place, where the bottom could be reached by a walking-stick, and was muddy; the mud was first to be stirred with the stick, and when a number of small bubbles began to arise from it, the candle was applied. The flame was so sudden and so strong, that it catched his ruffle and spoiled it, as I saw. New-Jersey having many pine-trees in different parts of it, I then imagined that something like a volatile oil of turpentine might be mixed with the waters from a pine-swamp, but this supposition did not quite satisfy me. I mentioned the fact to some philosophical friends on my return to England, but it was not much attended to. I suppose I was thought a little too credulous.