SOME ASPECTS OF THE ATOMIC THEORY 581 



Along these lines we arrive at the view that the periodic 

 law represents a periodic variation of the properties of elements, 

 not with the atomic mass as first concluded, but with the atomic 

 number or nett nuclear charge of the atom. At once there 

 arises the question, what is the relation between atomic weight 

 and atomic charge ? The fact that the vast majority of the 

 atomic weights are approximately whole numbers, in terms 

 of oxygen as 16, suggests that unit change of atomic charge 

 is accompanied by a change in mass, approximately some 

 multiple of the hydrogen atom. In a-ray changes at the 

 extreme end of the periodic table, the change of charge is 

 two and the change of mass four, and this holds fairly well 

 in the early, but not in the subsequent, part of the periodic 

 table where the atomic weight is greater than twice the atomic 

 number. Sir J. J. Thomson advances grounds for believing 

 that as far as the first 18 elements, there are in reality two 

 series, one with even and one with odd atomic weights, the 

 members of each series increasing with the common difference 

 four. There are, however, two exceptions — beryllium, which 

 on this scheme should have an atomic weight of 8 instead of 

 9, and nitrogen, which should be 15 instead of 14. Hydrogen, 

 moreover, does not seem to fit in at all. 



With regard to the exceptions from the rule of the atomic 

 weights being approximate integers, the suggestion is made 

 that the excess or deficiency of mass corresponds with the 

 absorption of liberation of energy in the formation of the element, 

 equivalent to that possessed by the mass gained or lost, moving 

 with the velocity of light. Thus per 35-5 grams of chlorine, 

 the half-gram in excess or deficit of the nearest whole numbers 

 would correspond with 2-25 x io 20 ergs, the amount of energy 

 required to keep the Mauretania going at full speed for a week. 

 The enormous amounts of energy liberated from heavy atoms 

 in radioactive changes certainly suggest such a possibility. It 

 is interesting also to recall that the very recent results of 

 Honigschmid on the atomic weights of uranium and radium, 

 238*15 and 225-95 respectively, make that of the former about 

 o-2 unit in excess of that of the latter plus that of the three 

 helium atoms expelled in the transformation. But it has still 

 to be proved how far the unsuspected existence of isotopic 

 elements may not be accountable for the departure of atomic 

 weights from approximate whole numbers in certain cases. 



