April 7, 1892] 



NATURE 



539 



ment of molecular weakness in the bar, the weight has 

 power to bend it, and the pointer falls. By such experi- 

 ments the exact temperature at which the metal becomes 

 weak, in different varieties of steel, can readily be deter- 

 mined. 



(2) Evidence will now be given in support of the second 

 case it was proposed to treat, and it will be shown that at 

 high temperatures the atoms of metals may truly com- 

 bine with each other ; in f.ict, taking gold as a basis for 

 the experiments, compounds may be formed which would, 

 had they been known centuries ago, have strangely 

 affected the history of science. When the alchemists 

 subjected the metals to high temperatures, their efforts 

 were mainly directed to the discovery of some substance 

 that would either change base metals to the colour of 

 gold, or would give them the brilliancy of silver. The 

 mediaeval chemists believed that there were two distinct 

 substances that would effect this, "one for the white "and 

 another "for the red." Many of their writings might be 

 quoted in support of this view, but a reference to Geber, 

 who wrote in the eighth century, will be sufficient. He 

 pointed out that the transmuting agent "has a tincture of 



and stream of opal," reminding us of the crimson and 

 purple of the poppy, the scarlet and orange of fire and 

 the dawn. No wonder he chides us with turning the 

 lamp of Athena into the safety-lamp of the mmer, and 

 with getting our purple from coal instead of, as of old, 

 from the murex of the sea ; " and thus grotesquely," he 

 says, " we have had forced on us the doubt that held the 

 old world between blackness and fire, and have com- 

 pleted the shadow and the fear of it by giving to a 

 degraded form of modern purple a name from battle — 

 * Magenta.'" 



You will remember that Faraday showed that gold, when 

 finely divided, is brilliantly coloured scarlet and purple. 

 Here is a solution of chloride of gold. Add a little dis- 

 solved phosphorus, and the gold is precipated in an ex- 

 tremely fine state of division, which tinges the solution 

 crimson, but if you try to remove this suspended gold 

 you will only gain a brownish mud. However, I will give 

 you the secret by which anyone who possesses a blow-pipe, 

 ahead of gold, and a fragment of one of the most widely- 

 diffused metals, aluminium, may stain gold purple through 

 and through. But if youadd aluminium to molten gold, you 



itself so clear and splendid, white or red. clean and in- 

 combustible, stable and fixed, that fire cannot prevail 

 against it ; . . . and a property of the medicine is to 

 give a splendid colour, white or intensely citrine," to 

 metals to which it is added. 



That was the effect expected from the transmuting 

 agent, but do not think that the attempt to produce gold 

 arose entirely from the love of gain. The colour of gold 

 and purple impressed men strangely, and the search for 

 the transmuting agent was most eagerly pursued in times 

 when people lived for art, in a dteam of colour. The 

 effort to find the secret of the tint of gold is due to the 

 same impulse which made the French in the thirteenth 

 century manifest a keen " sensitiveness to luminous 

 splendour and intensity of hue," so that, as Sir Frederic 

 Leighton tells us, " a stained glass window, by Cousin, 

 was limpid with hues of amethyst, sapphire, and topaz, 

 and fair as a May morning." The chemists were able 

 to stain glass ruby and purple with gold : why should 

 they not impart the same glories to metals ? I could 

 not hope to interest you in what follows, did I not call 

 artists to my aid ; and many will remember the glowing 

 words Mr. Ruskin uses/ calling purple a "liquid prism 



I "The Queen of the Air," ed. 1887, p. 129 ; Times, December 11, 1891. 



obtain many things, as this coloured diagram and series 

 of specimens show, [This diagram cannot be reproduced 

 without colour.] 



The series of specimens showed that as the proportion of 



aluminium is increased, the golden colour of the precious 



metal is lessened, and when an alloy is formed with about 



ten percent.ofaluminium,the fractured surfaceof the mass 



is brilliantly white: from this point forwards, as aluminium 



is added, the tint deepens, until flecks of pink appear, and 



when seventy-eight parts of gold are added to twenty-two 



of aluminium a splendid purple is obtained, in which 



intense ruby-coloured opaque crystals may readily be 



recognized. Then, as the quantity of aluminium is still 



further increased, the alloys lose their colour, and pass to 



the dull grey hue of aluminium itself. Perhaps the most 



I remarkable point about the purple alloy is its melting- 



I point, which I have shown to be many degrees higher 



j than that of gold itself.^ See diagram, Fig. 7, in 



I which curves of several constants of these alloys are 



j given. This fact affords strong evidence that the 



alloy AuAI.^ is a true compound, having analogies to 



j the sulphides, for in every other series of alloys the 



I melting-points of all the members of the series are 



' ' Proc. Roy Soc.. voL 1., 1891, p. 367. 



NO. I 171, VOL. 45] 



