592 SCIENCE PROGRESS 



The diagram shown is that laid in 1902 1 before the Royal Society 

 of London. The appended classified scheme of the elements is 

 but slightly different from that brought before the Royal Society 

 in 1902 and will be understood without further explanation. 

 The Argonides are omitted, as they cannot be classified. Anti- 

 mony and Bismuth are placed according to their atomic weights 

 but might easily be brought into the nitrogen-arsenic column; 

 if this change be made, five of the columns are blank and two 

 have elements only at their head. Cobalt should be put above 

 nickel. 



The recognition of the possibility of arranging elements 

 under families in the order of atomic weight has made it 

 possible to foresee necessary corrections in atomic weights 

 in cases in which the element, when properly arranged in 

 its family, appears to be out of place in weight. Several 

 such forecasts have been justified — there is, however, one 

 obstinate exception, Tellurium. This element should come in 

 advance of iodine, as it belongs to the oxygen-sulphur family, 

 which is in advance of the fluorine-chlorine family. The 

 accepted atomic weight of tellurium, however, is about a unit 

 greater than that of iodine — so that if arranged in the order 

 of its weight it falls after iodine and is out of place. Time 

 after time the atomic weights of tellurium and iodine have been 

 redetermined but without disturbing the relationship. 



It appears to me that a possible explanation of the difficulty 

 that tellurium introduces may be found in the assumption that 

 the genetic connexion between the elements is not necessarily 

 always through one and the same direct line — the conventional 

 assumption. Characteristic of the method of classification I 

 have advocated is the existence, not only in the case of the rare 

 earth elements but in others also, of pockets of related or 

 homologous elements— of sub-families. If such exist, there is 

 no reason why in the cases in which only single elements are 

 not known there should not also be several. It is conceivable, 

 from this point of view, that tellurium belongs to a sub-family 

 of which it is not the head and that iodine is to be regarded as 

 descended in direct succession from the missing head of the 

 tellurium sub-family, not from tellurium itself. 



But the argonides have introduced new difficulties. As 

 they have no chemical properties, in classifying them we have 



1 Proceedings, vol. lxx. p. 86. 



