C l 79 J

in fusion, by the continuance os the sire, by the fiuid mat-ter being kept unmoved or stirred with an iron rod, bythe crucible being covered or open, or other like circum-stances, or whether the admixture of a little manganese,though gold will certainly give a ruby colour without it,does not contribute to secure the success, I have not yetdiscovered. The proportion of the gold precipitate tothe vitreous matter is perhaps of principal importance.Solution of gold, as we have seen already, produces noredness with tin unless diluted with a very large quantityof water, in which circumstance the whole mixture ac-quires that beautiful colour which we here want to trans-fer from the watery fluid into fluid glass : it should seemtherefore that the quantity of gold precipitate, for com-municating the admired colour to a certain volume ofglass, ought to be the fame with that, which communicateda like colour to an equal volume of water in the preci-pitation : a quantity extremely minute, and much lessthan that employed in any of my experiments.

I have lately been favoured with some pieces of glass,in greatest part colourless, with one or two large redspots, several small streaks of violet, and some of a lightbrownish yellow. The person from whom I receivedthem informs me that he had “ found that in a heatnot very strong, under a muffle, the glass becomes ofan opake brown, and, if then polished, appears va-riegated like a fine pebble.” I exposed a colourlesspiece to the flame of a lamp, impelled by a blow pipe,and on working it about, sometimes in the smoke andsometimes in the flame, found it change to a true rubyred, perfectly transparent, and free from veins of anyother colour. Another piece, kept for two hours undera close muffle, in such a heat as made it just soft enoughto bend and receive an impression, became on the surface

A a 2 green,