Atoms, Ions and Molecules. 411 



IV. Classification, continued from Part Z* 



Any attempted classification of the elements is necessarily 

 subjected to a severe test through later discoveries. If the 

 classification is in complete accordance with nature, it must 

 offer suitable places for such new elements as may present 

 themselves later. If it does not do this, it must be more or 

 less artificial. 



Mendeleef's remarkable predictions and the fact that the ele- 

 ments next discovered found places marked out for them in 

 advance, caused his periodic law to be received with enthusiasm, 

 It so happened that these new elements missed the weak por- 

 tions of the system. This is now much changed. Within the 

 last year or two, several new elements have been discovered 

 for which the periodic law affords absolutely no place. 



This appears to be admitted by Mendeleef. In an articlef 

 he contends that argon must be polymerized nitrogen and after 

 examining the possible forms that nitrogen might take, he 

 decides in favor of N s . 



But Berthelot succeeded in combining argon with carbon 

 disulphide or with its elements, and from this compound he has 

 been able to regenerate argon with all its characteristic proper- 

 ties. This Berthelot justly calls a capital fact. It is scarcely 

 necessary to say that it is quite fatal to the theory of polyme- 

 rization just mentioned. 



The impossibility of finding a place for argon in the periodic 

 system will become evident from the following considerations. 

 If its atomic weight should prove to be 19*9 its place will 

 necessarily be between fluorine and sodium. Introduced into 

 the periodic system, it would take its place at the head of the 

 iron group. It is to be remarked that in the periodic system 

 the iron group is a somewhat singular one. It does not 

 include any of the metals which we commonly associate with 

 iron, cobalt and nickel, chromium and manganese. Instead of 

 these, iron is made to form a group with ruthenium and 

 osmium, so that if argon should enter in the periodic system 

 with the atomic weight of 19*9 it would form the head of a 

 group consisting of argon, iron, ruthenium and osmium. Or 

 of one consisting of A, Co, Rh, and Ir. Or of another, 

 equally anomalous, A, Ni, Pd, and Pt. 



But argon may prove to have for its atomic weight the num- 

 ber 39*8. For such an element there is absolutely no place in 



*In the first part of this paper (this Journal, May, 1895) in the table at page 

 362, by a vexatious mistake of the printer, the groups have been dislocated so that 

 "the series is no longer in numerical order, thus wholly destroying the sense of the 

 stable. On page 361 the order is correct and the table has been correctly printed 

 in the German translation which appeared in the Zeitschrift fur Anorg. Chemie. 



f Reprinted in the Chemical News for July 12th, 1895. 



