GENETIC CONNEXIONS BETWEEN THE CHEMICAL 

 ELEMENTS. 



By J. MoiR, M.A., D.Sc, F.C.S. 



At intervals all through the nineteenth century theoretical 

 chemists were engaged in trying to trace out those obscure but 

 evident family relationships which exist among the elements, 

 the most comprehensive of these schemes being the well-known 

 Periodic Lav/. Thousands of pages have been written on relation- 

 ships obtained by assuming the atomic weights to be whole members. 

 It is evident, however, nowadays that only a proportion of the 

 whole number of elements have integral atomic weights, and 

 even here, as the author has shown elsewhere,* it becomes necessar 

 to assume a unit for the system different from the cnstoma » 

 ones (H = I, or O = i6), in order that the elements in question 

 shall exhibit an exact integral relationship. 



There is undoubtedly a ' regularity in irregularity ' of a most 

 curious and fascinating character amongst the experimental 

 atomic weights, but it is only recently that determinations have 

 become accurate enough to give any hope of tracing genetic con- 

 nexions from the figures ; since, apart from the iigures, we have 

 nothing to guide us towards the constitution of the atoms except 

 such analogies as we can draw from molecular constitutions {e.g. 

 the configuration of the rubidium atom must closely approximate 

 to that of the ammonium complex). L'ntil a few years ago, how- 

 ever, we had no clue regarding the nature of the sub-elementary 

 bricks out of which atomic edifices are made ; but this clue has 

 been given us by Ramsay's discovery of the spontaneous formation 

 of helium from several of the higher radio-active elements : and 

 indeed there appears to be little doubt that radium and polonium 

 are directly connected, by means of helium., with lead and bismuth. 



The first step in the author's new work on this subject is that 

 already mentioned, the discovery of a basis of calculation on 

 which many of the atomic weights become whole numbers and 

 therefore, very probably, of the same parentage. A brief notice 

 of this will suifice here, for the benefit of those who cannot obtain 

 access to the original papers. The author assumed the existence 

 of an intermediate link between the electron and the atom of 

 hydrogen, in the shape of a sub-atom (apparently 15 electrons) 

 of atomic weight o'oogo, and secondh/ that this entity is the actual 

 physical cause of valency. Each univalent ' element ' contains 

 one, each bi-valent element two, and so on. of this sub-atom, 

 for which the author proposes the symbol ^/. The results are 

 (i) that the atomic weights become whole numbers plus multiples 

 of 1^1 corresponding to the valency: — thus H = i-|-y(/, He=4,. 

 Li = 7-f/;(, C = i2-f-4/i and so on, when the basis of calculation 

 is shifted from = 16 to o = i6-f 2;u = i6"0i8, or, what comes to 



*Journ. Cliem. Sec, Lond., Dec, 1909, and Proc Roy. Soc, S.A., 1909. 



