. [ 86 1

be excellent for rstaking specula, being less subject totarnish in the air than the compositions of which copperis the basis.

It is supposed by many, that gold, melted with othermetals, is always diffused equally through their whole vo-lume, insomuch, that the quantity of gold, obtainablefrom any part of the mixt, shall bear exactly the fame pro-portion to that part, as the whole of the gold does to thewhole mass. There appears, however, in many cafes, asensible inequality in the distribution. M. Hellot, in hisFrench translation of Schlutter’s German work on thesmelting and assaying of ores and metals, gives an accountof an experiment which clearly shewed this inequality :a quantity of silver, amounting to upwards of twentypounds, containing about a sifty-sixth part of gold, wasmelted in a crucible, and poured into cold water, in orderto its being reduced into grains : by dipping at differenttimes an iron ladle into the water, under the stream ofmetal, he received a part of the first running, a part ofthe middle, and a part of the last : the three parcels,assayed separately, were all sound to differ in their contentof gold.

There is a curious experiment of Mr. Homberg’s, re-lated in the French memoirs for the year 1713, which,though I have not yet tried it, I shall venture to insert onaccount of its singularity. Equal parts of gold and silver,melted together and reduced into fine grains, were putinto a crucible, with a mixture of about equal parts ofdecrepitated sea salt and rough nitre under them : thecrucible being kept in a small sire, in a wind-furnace, forabout a quarter of an hour, and then suffered to cool andbroken, the gold was found in one lump at the bottom,and the silver above it in two pieces, with some grains,

wrapped