ON THE CHEMICAL COMPOUNDS CONTAINED IN ALLOYS. 117 



Professor Ramsay and M. Tammann in 1890 showed thab small 

 quantities of various metals dissolved in mercury gave, for the most part, 

 depressions of the vapour-pressure and of the freezing-point, which indi- 

 cated that the dissolved molecule contained one atom of the added metal. 

 At the same time, Mr. Heycock and I found that this was in general true 

 when metals were dissolved in tin. At later dates we extended the 

 generalisation to solutions of metals in the solvents bismuth, cadmium, 

 lead, and zinc, and tables summarising our results are reproduced in the 

 present report. If we could be certain that the dissolved metal did not 

 form a chemical compound with the solvent, these results would afford 

 very strong grounds for holding that the molecules of the dissolved metals 

 were in most cases monatomic. But we know now that chemical com- 

 bination is not uncommon, and it is evident that in dilute solution the 

 dissolved metal a will tend to form compounds of the type A b„„ where b 

 is the solvent. Hence the problem of the chemical compounds formed by 

 metals with the solvent metal must be solved before we can safely 

 dogmatise concerning the molecular weight of the metals when in solution. 

 To take a special case, one atom of copper dissolved in tin produces the 

 molecular depression of the freezing-point, but from Mr. Stead's work we 

 have good reason to attribute this to the presence of a molecule CuSn. 



On the other hand, the abnormal depressions obtained by us in certain 

 cases point to the probability of the compounds Bi„As„ Bi Cu^ Cd Ho-, 

 Cd„Zn,, Cd„Pd„ Cd„K„ Cd„Au„ Cd„As,„ Pb„(Cd, Hg, Bi)^, PblSn;! 

 Pb„Na^, most of which have not at present been studied. It is obvious 

 that n may be zero in any of these. 



The question of the depression of the freezing-point in dilute solutions, 

 is, however, complicated by the probable appreciable solubility of the 

 dissolved metal in the solid crystals of solvent, and by all the thermal 

 difficulties that ISTernst and Abegg have discussed. 



The fascinating question as to the condition of association or dis- 

 sociation of the molecules of the compounds when melted or in solution 

 also comes in when we attempt to interpret our tables, or, indeed, when 

 we examine any freezing-point curve. But it is possible to study inter- 

 metallic compounds without touching this question, and in the present 

 report I have thought it best to do so. 



The vast subject of ternary and more complex mixtures has also been 

 avoided as too complex for the present purpose, although Behrens, 

 Stead, and especially Charpy, have made most interesting studies of such 

 mixtures. 



I have to thank Mr. Heycock for continued assistance in drawing up 

 this report. I have also to thank Mr. Stead and Professor H. Le Chatelier 

 for valuable information and valuable references. 



Depression of the Freezing-point of the Metals Tin, Zinc, Bismuth 

 Cadmium, and Lead, caused by the /Solution of Small Quantities of 

 other Metals. 



The theoretical molecular depressions are calculated froai the latent 

 heat of fusion by means of the formula 



id = 0-02^' 

 A. 



where 6 is the freezing-point of the pure metal, cO the depression,^and 

 X the latent heat of fusion. ^ 2 



