1902.] 



The Classification of the Elements. 



87 



writing an article for the forthcoming Supplement to the ' Encyclo- 

 paedia Britannica.' The article, I may say, was sent to press in May, 

 1 1900, and the first proof before me is dated November 20, 1900. 



The suggestion of an octave of elements having been made in early 

 days, the tendency to adopt an octavo system has been irresistible. 

 It is difficult now to discover any real justification of such a practice. 

 As there is frequently a difference of (about) a single unit between the 

 atomic weights of contiguous elements, and as the "homologous" ele- 

 ments among those of lower atomic weight differ by (about) sixteen 

 units — I assume that the " elementary difference " may be (about) a 

 single unit, and that the first " horizontal period " comes to an end with 

 oxygen: in other words, that there are sixteen "vertical series" of 

 homologous elements. 



In the table on pp. 88 — 89, the elements are entered under whole 

 numbers without regard to their exact atomic weights. The dominant 

 principle on which the arrangement is based is that of maintaining 

 elements which belong to the same family in the appropriate columns. 

 The chemist cannot adopt any other mode of arrangement in view of 

 the uncertainty which attaches to many of the atomic weights. 



I have not hesitated to assume that the molecules of argon and 

 similar elements are polyatomic like those of nitrogen — their nearest 

 ally in outward behaviour — and have regarded them as diatomic, as 

 the molecules of all the ordinarily gaseous elements are of this order of 

 complexity. It has always been my opinion that they are elements of 

 intense activity, the atoms of which so fully satisfy each other that no 

 residual affinity is manifest in the molecules. From this point of view 

 the appearance of argon immediately after fluorine is quite in order. 



Turning to the several periods, the first calls for little remark. 

 Helium is put next to hydrogen — and it may well be that elements 

 somewhat like helium and argon will be discovered which will fall into 

 positions 3, -1, and 18 ; in fact, this region may well turn out to be the 

 nesting-place of such elements. Beryllium is shown separated from 

 lithium and magnesium, but not to an extent which need give rise to 

 objection. It is perhaps not beyond the region of possibility, however, 

 even in the case of an element with so low an atomic weight, that iso- 

 morphous mixtures may have passed as pure substances : this possi- 

 bility has not been sufficiently taken into account in preparing the 

 material for determinations of atomic weight. 



In the second period, although it does not come immediately below 

 nitrogen, phosphorus is very near to that element — and sufficiently so 

 to correspond with their very close relationship. 



In the third period, when scandium is passed, it is impossible to 

 proceed by units : to bring titanium into position it is necessary to 

 step down five units. Proceeding from titanium, however, vanadium 

 and chromium fall into appropriate positions. 



