1892.] on Metals at High Temperatures. 517 



Sir Lowthiau Bell showed that the relations between carbonic oxide 

 and iron are of singular interest. Ludwig Mond, Quincke and Langer 

 have since found and isolated most interesting compounds of iron and 

 carbonic oxide,* But to return to the solution of metals in metals. 



The method of taking autographic curves of the cooling of masses 

 of metal has already been indicated in Fig. 4,f and they ought to 

 enable much information to be gained as to what is taking place 

 throughout the mass. Such curves should render it possible to 

 ascertain which of the rival theories as to the nature of solution, as 

 applied to salts, is supported by the behaviour of a metal dissolved in 

 a metal. When, for instance, a little aluminium dissolves in gold, is 

 the analogue of a hydride formed, and, if so, is the curve of freezing- 

 points of a series of aluminium-gold alloys a continuous one ? On 

 the other hand, does the theory advocated by Yan't Hoff, Arrhenius, 

 and Ostwald gain support, and do the molecules of the dissolved 

 metals act independently of the solvent — that is, does osmotic pressure 

 come into play? It will be remembered that the law which regu- 

 lates osmotic pressure has exactly the same form as Boyle's law — 

 that is, the pressure is proportional to the density of the gas or of 

 the solution. Is the view of Arrhenius correct — that, if a solution 

 be very dilute, the molecules of the dissolved substance are disso- 

 ciated, act independently of each other, and behave like a perfect gas ? 



It will require years of patient work before these questions can be 

 answered ; but it appears certain, from the admirable experiments of 

 Heycock and Neville, | to which reference has already been made, 

 that, taking metals with low melting-points (such as tin or lead) as 

 solvents, the lowering of the freezing-point of the solvent is really 

 due to the bombardment exerted by the molecules of the dissolved 

 metals. 



I have extended this investigation by employing as a solvent a 

 mass of fluid gold, which has a high melting-point, and is not liable to 

 oxidation, and the results confirm those obtained by Heycock and 

 Neville. 



There is yet one other question : When metals are added in small 

 quantities to a metallic mass, may the solvent remain inert ? Here is 

 a mass of 1000 grammes of lead, and to it 15 grammes of gold, or 1 • 6 

 atoms for every 100 atoms of lead will now be added. It could be 

 shown that the gold is readily dissolved, and remains dissolved, even 

 if the lead be solidified. Now, to the fluid lead sufficient aluminium 

 will be added to form the purple alloy with the dissolved gold ; the 

 mass will be well stirred, but the aluminium will not unite with the 

 lead ; it will nevertheless find out the gold, and, after uniting with it, 

 will carry it to the surface of the bath. Thence it can be removed, 

 and the purple colour of the alloy identified, or the gold it contains 



* Chem. Soc. Journ. vol. lix. 1891, pp. 604, 1090. 



t Fioc. Koy. Soc. vol, xlix, p. 347, lb91. Loc. cit. 



