[ 209 ]

before it was divided for the cupelling, was weighed inwater, and instead of weighing about nineteen times asmuch as a bulk of water equal to it, its proportion tothat liquor was but that of fifteen and about two thirdsto one; so that its specific gravity was less by about threeand a third than it would if it had been pure gold.From this experiment the author concludes, that an ope-ration very near, if not altogether, as strange as thatwhich is called projection, and in the difficultest pointsmuch of the fame nature with it, may be safely ad-mitted : for the experiment plainly shews that gold,though confessedly the most homogeneous and least mu-table of metals, may in a very short time be exceed-ingly changed, both as to malleablenefs, colour, homo-geneity, and what is more, specific gravity ; and all thisby an inconsiderable proportion of injected powder, notamounting, on the modestest estimate, to a nine hundredand fiftieth part of its weight. He adds, that there is astill stranger effect of this admirable powder, which hehas not mentioned because he must not doit.

On this history it may be remarked, i. That little de-pendance can be had on the conjecture of the weight ofthe powder, as it might possibly not be all distinguishedon the paper it was wrapt in, and as different kinds ofbodies have different weights under equal volumes. 2. Ifno mistake was made in weighing the metal after thefusion, the quantity of powder must have been greaterthan was imagined ; for the lump of metal appeared toweigh as much as the gold employed, exclusive of thevitrefied substance which the bottom of the crucible wasoverlaid with, and of the metallic globules that adheredto it. If these globules had been picked out, or separatedby pounding and washing the pieces of the crucible, andweighed with the lump, it is scarcely to be supposed that

E e the