396 On the Experimental Relations [1857. 



globular portions. A comparatively low heat, however, and one 

 unable to cause separation of the particles, is known to alter the 

 molecular condition of gold, and the gold-beater finds important 

 advantage in the annealing effect of a temperature that does not 

 hurt the skins or leaves between which he beats the metal. 



It might be supposed that the annealed metal, in contracting 

 from the constrained and attenuated state produced by beating, 

 drew up, leaving spaces through which white light could pass, 

 and becoming itself almost insensible through the smallness of 

 its quantity ; and if gold-leaf unattached to glass be heated 

 carefully with oil in a tube, it does shrink up considerably even 

 before it loses its green colour, which finally happens. But if 

 the gold-leaf laid upon glass plates by water only be carefully 

 dried, then introduced into a bath of oil and raised to a tempe- 

 rature as high as the oil can bear for five or six hours, and then 

 suffered to cool, the plates, when taken out and washed, first 

 in camphine, and then in alcohol, present specimens of gold 

 which has lost its green colour, transmits far more light than 

 before and reflects less, whilst yet the film remains in form and 

 other conditions apparently quite unchanged. Being now exa- 

 mined in the microscope, it presents exactly the forms and 

 appearance of the original leaf, except in colour ; the same 

 irregularities appear, the same continuity, and if the destruction 

 of the green colour has not been complete, it will be seen that 

 it is the thicker folds and parts of the mottled mass that retains 

 the original state longest. 



This change does not depend upon the substance in contact 

 with which the gold is heated*. If the leaf be laid upon mica, 

 rock-crystal, silver or platinum, the same result occurs ; the 

 surrounding medium also may change, and be air, oil or carbonic 

 acid, without causing alteration. Nor has the gold disappeared ; 

 a piece of leaf, altered in one part and not in another, was 

 divided into four equal parts, and the gold on each converted 

 by chlorine gas into crystallized chloride of gold ; the same 

 amount was found in each division. 



When the gold-leaf is laid by water on plates of rock-crystal, 



* The disappearance of gold-leaf as metal, when mingled with lime, alumina 

 and other bodies, and then heated, has been already observed ; and referred 

 to oxidation (J. A. Buchner). See Gmelin's * Chemistry/ vi. p. 206, " Purple 

 oxide of gold." 



