146 



REPOET — 1900. 



It would be very easy to amplify this list ; for example, various other 

 arsenides and antimonides have had formulse assigned to them, and some 

 alloys of aluminium and tin with the rarer metals appear to have 

 been isolated as crystals. Further research will no doubt enormously 

 expand it, though it may also cause the rejection of a few that have been 

 included. 



But as the list now stands it offers matter for the consideration of the 

 student of valency. One sees that the compounds of the metalloids with 

 the metals present formulse that we should expect from the known 

 valencies of the elements, but such bodies as NaHg.j, SnCuj, AlAg.j are 

 more remarkable. The first of these is evidently a well-marked type, 

 which already occurs several times. 



If I rightly understand Professor Kurnakov, he thinks that Mendeleef's 

 law of the total valency of an element for oxygen and hydrogen being 

 8 will find application in the formulre of alloys, the hydrogen being 

 replaced by other metals. In this case, the alkali metals which are 

 monovalent to oxygen should be polyvalent in alloys ; his curves certainly 

 support this view. The freezing-point curves show that the most marked 

 summits — that is, the most stable compounds — occur when a strongly 

 positive metal, such as sodium or aluminium, is alloyed with a metal, such 

 as antimony, lead, or gold, which is far removed from it in the electro- 

 chemical series. 



Tlie Molecular Weights of Metals. 



With the exception of the limited number of vapour -density determi- 

 nations which show that mercury, cadmium, and zinc, and perhaps also 

 sodium and potassium, have monatomic molecules when gaseous, the only 

 evidence as to the molecular weights of metals lies in experiments based 

 on Raoult's methods. 



