[ 622 ]

with powdered nitre, tartar, borax, and arsenic, and thecompound melted with a suitable fire. The proportionsof the ingredients are, 2560 parts of sand, 384 of nitre,240 of tartar, 240 of borax, 28 of arsenic, 5 of tin, and 5of gold.

I have not yet had an opportunity of trying this process,but am convinced, that the mixing of the precipitate withthe sand, &c. in a moist state is a very material circum-stance, if not the principal one upon which the success ofthe operation depends. Perhaps the most certain waywould be, not to wait for any precipitation at ail, but tomoisten the powders with the purple liquor, grinding themwell together, in a moderate warmth, till dry, and if ne-cessary, repeating the humectation. Mr. Potts experi-ments with the microcosmic salt, mentioned in page 619,confirm the utility of this method of mixture.

Page 185. Quantity of gold collected in rivers.

In a paper drawn up by Mr.Guettard, from the obser-vations of Mr. Pailhes, and published in the volume ofthe French memoirs for the year 1761, the gold found inrivers is reckoned an object of more importance than it hasbeen usually represented. It is said that the mint of To-I'ouse received commonly every year two hundred marcs,or one hundred pounds weight, of gold collected from theAriege, Garonne, and Salatj and that since the year 1750,twelve pounds have been carried into the bureau of Pa-nders, though this bureau comprehends at most an extentof only two leagues round, and though the whole of thegold is not sent thither, strangers and hawkers buying it upevery day.