November 2, 1917] 



SCIENCE 



421 



count of its influence upon the further development 

 of the atomic theory this hypothesis must be dis- 

 cussed here, although it but seldom happens that 

 an idea from which important theoretical concep- 

 tions sprang, originated in so faulty a manner as 

 it did. 



However, a careful study of the most ac- 

 curately determined of the recent atomic 

 weights reveals some very remarkable re- 

 lationships. If first of all we make the as- 

 sumption, as a subject for argument, that 

 the heavier atoms are built up from hy- 

 drogen atoms, then it is found that the 

 atoms are in nearly all cases lighter than 

 they should be on the basis of such an hy- 

 pothesis. Thus, if the following atoms of 

 low atomic weight are considered, it is 

 found that nearly all of them weigh 0.77 

 per cent, too little. 



Therefore, if these atoms are built from 

 hydrogen atoms, there must be during their 

 formation a loss in weight, and presumably 

 in mass, equal to 0.77 per cent. This will 

 be called the "packing effect." When all 

 of the 26 elements from helium of atomic 

 number 2, to cobalt (No. 27) are consid- 

 ered, it is found that with the exception of 

 the four elements, beryllium, magnesium, 

 silicon, and chlorine, which have atomic 

 weights higher than the corresponding 

 nearest whole numbers, the average pack- 

 ing effect of the elements is again — 0.77 

 per cent. This constancy of the pack- 

 ing effect suggests that the variation 

 is due to some single cause, though the four 

 exceptional cases cited above, show that 

 there is undoubtedly some other compli- 



cating factor. The discovery by Thomson 

 and Aston that the similar exceptional case 

 of neon is due to the admixture of an iso- 

 tope of higher atomic weight suggests that 

 it may not be impossible to find a cause for 

 the exceptional behavior in the four other 

 cases. 



It has formerly seemed difficult to ex- 

 plain why the atomic weights referred to 

 oxygen (16.00) as a basis are so much 

 closer to whole numbers than those referred 

 to hydrogen as 1.00, but the explanation is 

 indeed very simple from the point of view 

 presented here. The closeness of the atomic 

 weights on the oxygen basis to whole num- 

 bers, is indeed extremely remarkable. 

 Thus for the eight elements from helium to 

 sodium the average deviation is only 0.02 

 unit, or less than the average probable 

 error of the atomic weight determinations, 

 and for all of the first 27 elements the aver- 

 age deviation from a whole number is, 

 though more, increased only to 0.09 unit, 

 when the sign of the deviation is consid- 

 ered. If it is not considered the deviation 

 is reduced to 0.01 unit for 21 elements. 

 The probability that such values as these 

 could be obtained by accident is altogether 

 unworthy of consideration. If an oxygen 

 atom is a structure built up of 16 hydrogen 

 atoms, then according to the ordinary 

 theory that mass and weight are strictly 

 additive, the weight of an atom of oxygen 

 should be exactly 16 times the weight of a 

 hydrogen atom. Now, according to the 

 present system of atomic weights the 

 weight of an atom of hydrogen is taken as 

 1.0078, so the oxygen atom should weigh 

 16.125. However, it is found to weigh 16.00. 

 The difference between 16.125 and 16.000 

 is the value of the packing effect, and if 

 this effect were exactly the same for all of 

 the elements except hydrogen, then the 

 choice of a whole number as the atomic 

 weight of any one of them, would, of neces- 

 sity, cause all of the other atomic weights 



