228 MESSRS. C. T. HEYCOCK AND F. H. NEVILLE 



that the very opaque AuAl 2 which crystallised first has settled to the bottom 

 of the crucible, leaving a mother liquid which afterwards crystallised with the 

 usual fine grain of a eutectic. The shape of the AuAL crystals is distinctly visible 

 in the negative, and is very similar to that of crystals of the same body in alloys 

 between G and H. 



The slowly-cooled alloy (figs. 32 and 33), containing 07 atomic per cent, of gold, 

 is near the bottom of the branch JI, the quantity of AuAl 2 present is insufficient to 

 saturate the aluminium, and, therefore, during the process of solidifying, the 

 aluminium has crystallised first on the walls of the crucible, perceptibly con- 

 centrating the opaque AuAl 2 in the central portion of the ingot. 



The Bodies B, D, E, X, and H. 



The body B, pure at 20 atoms of aluminium, to which we attribute the formula 

 Au 4 Al, was obtained in the form of a brittle rod of white with a faint yellow 

 tinge. It has a silky couchoidal fracture. Although containing little more than 

 3 per cent, by weight of the white aluminium, the colour of the gold is gone, 

 in fact there is no free gold in it. This alloy is more easily attacked by etching 

 reagents than pure gold, or than the D and E bodies, hence, after etching, we 

 never see it as a white body but as yellow or brownish purple from a film of finely 

 divided gold. D and E are pure white bodies, both before and after etching with 

 bromine or aqua regia. Fused caustic potash dissolves out the aluminium from 

 them and leaves them with a golden surface. They also break with a conchoidal 

 fracture, and are hard and brittle. 



The X body, which we think may be AuAl, has not been obtained pure in large 

 masses, but a slowly -cooled alloy with 45 atoms of aluminium contains pure white 

 patches of X ; these are soft. The X body when in contact with Au 2 Al is very 

 rapidly attacked and eaten away by bromine or aqua regia, leaving a grey finely 

 lined surface with the more resistant Au 2 Al in relief. 



The purple AuAl 2 is attacked by hydrochloric acid, a reagent which has little or 

 no effect on the other bodies, the colour being destroyed and an electrolytic deposit 

 of bright gold forming on the surrounding X body. It is curious that the com- 

 pounds containing only a little aluminium should be white, while AuAl, with nearly 

 22 per cent, by weight of the white metal should have the splendid purple of finely 

 divided gold. The identity in the melting point of AuAl 2 and of gold also marks out 

 this compound as worthy of further study. 



