On the Experimental Relations [1857. 



the colour of beaten or pressed metal. It is not, however, 

 pure gold, but an association of it and oxide of iron. Hydro- 

 chloric or other acids remove the iron and reduce the gold to 

 a dark, dense, insoluble set of particles, in very small quantity 

 apparently, yet containing all that was present in the bulky 

 green deposit. If the solution of gold be made slightly acid 

 beforehand, then the change and precipitation is to appear- 

 ance much less ; the reflexion by the particles is feeble, but of 

 a pale brown colour : the general transmitted light is ame- 

 thystine ; in the dark tube the tint is blue ; the particles are 

 much condensed and settle quickly, but occasionally leave a good 

 ruby film on the side of the glass, which has all the characters 

 of the ruby films and particles before described. The loose 

 gold particles quickly adhere together. Hence it appears that 

 the green precipitate often obtained by protosulphate of iron 

 is not pure gold in a divided state ; and that when care is 

 taken to produce such pure divided gold, it presents the ap- 

 pearances of divided gold obtained by other means, the gold 

 being competent to produce the ruby, amethystine, and blue 

 colours by transmission. Usually the gold rapidly contracts 

 and becomes almost insensible, and yet the test of protochlo- 

 ride of tin will show that all has been separated from solution ; 

 it then forms a striking contrast to the depth of colour presented 

 by the same solution of gold precipitated by phosphorus, and 

 most impressively directs attention to the molecular condition 

 of the metal in the latter state. 



A very small quantity of protochloride of tin, added to a 

 dilute solution of gold, gave, first the ruby fluid, showing dif- 

 fused particles by the cone of rays ; this gradually became 

 purple, and if the gold were in sufficient quantity, a precipitate 

 soon began to fall, being the purple of Cassius. If the chlo- 

 ride of tin were in larger quantity, a more bulky precipitate 

 fell and more quickly. Acid very much reduced this in quan- 

 tity, dissolving out oxide of tin, and leaving little else than 

 finely-divided gold, which, when diffused and examined in the 

 dark tube, transmitted a blue colour. I believe the purple of 

 Cassius to be essentially finely-divided gold, associated with 

 more or less of oxide of tin. 



Tartaric acid being added to a weak solution of gold gra- 

 dually reduced it. The amethystine tint produced by diffused 



