[ i74 ]

cult to render them perfectly void of colour : the man-ganese communicates at first a purplish hue, which oncontinuing the fire disappears, and at the fame time sup-presses or discharges any other tinge that the glass may beimpregnated with : the addition of a little nitre revivesthe purplish colour of the manganese, and Grummet is ofopinion that the colour with which glass becomes tinged,by the admixture of preparations of gold, is no other thanthat of the manganese extricated by the nitrous salt whichthe gold has retained in its precipitation. He affirms thatthe same purplish red colour will be obtained on meltingVenice glass with an eighth part of nitre, without anygold ; that in a hundred repetitions of this experiment, itscarcely fails once ; and that neither nitre nor the gold-precipitate were found to give any thing of the admiredcolour to those kinds of glass which have no manganese intheir composition.

The colours which manganese imparts to glass, it be-longs not to this place to examine : but that precipitatesof gold will communicate, in certain circumstances, apurplish red colour, I have several times experienced ;having myself tinged of this colour fritts composed ofcalcined flint, nitre and borax, without the addition ofany manganese or of glasses containing it. Though gold,dissolved in common aqua regia, exhibits its own yellowcolour; yet, when the menstruum is separated by fire toa certain point, or when the gold is precipitated by tin,or when it is precipitated by alkaline salts and afterwardsmoderately heated, or when the gold is barely divided bymechanical means into subtile powder, and exposed forsome time, in mixture with earthy bodies, to a flightheat, it assumes, in different circumstances, a violet colour,a purple, or a red verging to purple ; in a strong fire, thesecolours vanish, and the gold melts into a mass of its ori-

i ginal