594 SCIENCE PROGRESS 



nothing to go upon but their atomic weights — and these we do 

 not know. It has been proposed to group them together 

 vertically in advance of the lithium-sodium family. But argon, 

 assuming its atomic weight to be 40, is out of place before 

 potassium ; moreover, 40 is the atomic weight of calcium 

 and the existence of two isometric elements having such totally 

 different properties is without precedent : nickel and cobalt, the 

 only parallel case, have very similar properties. 



" Que faire ? " My own view is that there is no evidence 

 whatever to justify the conclusion that the argonides are true 

 single-atom molecules and inert as elements but much to 

 support the assumption I advocated in 1895 that the molecules 

 of argon (and therefore those of its companions) with which 

 we are acquainted are inert because they consist of atoms 

 which have intense affinity for each other — of atoms so firmly 

 locked in each other's embrace and so occupied as to be desti- 

 tute of external attractive power or residual affinity. If such 

 be the case, we cannot place them in the table at present, the 

 table being one in which the arrangement is in accordance with 

 atomic weights and with chemical properties — both of these are 

 unknown in the case of the argonides. 



I may here point out that in discussing the properties 

 characteristic of elements we must be careful — as in ordinary 

 life — not to be misled by appearances. The conception formed 

 is obviously an abstraction based on the consideration of the 

 behaviour of the element in its various compounds — for, in few 

 if any cases are we in possession of the elementary material 

 proper in its simplest or atomic form. 



The influence of molecular composition on properties is 

 altogether astounding. As we know it in the atmosphere, 

 nitrogen — which is present in the form of molecules each 

 composed of two atoms — is all but inert, yet we have every 

 reason to believe that, in the atomic state, it is one of the most 

 active of elements. Again, to take the case of carbon in the 

 diamond — which apart from its combustibility is remarkable 

 for its chemical inertness — we are dealing with the least 

 coloured substance that is known to us, whilst carbon in the 

 form of charcoal is the most highly coloured of elementary 

 materials, indeed of known materials ; the diamond is intensely 

 hard and has no metallic properties whatever but graphitic 

 carbon is commonly termed black lead on account of its resem- 



