Manchester Memoirs, Vol. xliii. ( 1 899), No. 4. 9 



approximate atomic weight 82, which would follow 

 bromine, and precede rubidium, in the Periodic Table. 

 Any considerable admixture of such a gas would neces- 

 sarily raise the apparent atomic weight of argon ; and it 

 might really have an atomic weight less than that of 

 potassium, instead of greater. The anticipation that 

 there is a still heavier gas mixed with atmospheric argon 

 has turned out to be correct ; but it is present in such 

 infinitesimal amount, that its removal does not materially 

 influence the density of argon. 



Another suggestion, also made at the time of publica- 

 tion of the Argon paper, was that possibly it might 

 contain some diatomic molecules, which would therefore 

 raise the apparent density. This supposition has also 

 turned out to be without support ; for there is no change 

 in the relative density of argon with either rise or fall of 

 temperature. It has been heated to 280° without undue 

 expansion ; and it has also been cooled to the boiling 

 point of liquid oxygen,— 182°, without undue contraction. 

 Had it contained any diatomic molecules, it is almost 

 certain that their number would have been reduced by a 

 rise of temperature, and increased by a fall of temperature 

 such as those to which it has been subjected. The absence 

 of any sign of such increase or reduction makes the 

 existence of diatomic molecules exceedingly improbable. 

 We must therefore be content to admit that its atomic 

 weight is so far anomalous that it is higher, rather than 

 lower, than that of the element which follows it in the 

 Periodic Table. 



We may here inquire whether the atomic weight of 

 argon is the only one which is, so to speak, displaced. If 

 it is not, it need occasion no anxiety to find it high. 

 Now, there are two elements, cobalt, on the one hand, and 

 tellurium, on the other, the atomic weights of which are 



