218 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1925 



fication at the expense of this heterogeneity? We can not believe 

 that a moment. On the other hand we should be certain of the value 

 of the classification in itself. It is not the question of saving it from 

 a disaster by which it is not menaced, but to restate the principle 

 which certainly contains something of truth although not the whole 

 truth. 



We have heard on many occasions of atomic numbers. First 

 through Rydberg but especially through Moseley. The latter gave 

 us a remarkable law which connects the radiation of the X-ray 

 spectra of each element with its atomic number. But this number 

 designates only the place occupied by each element in the classifica- 

 tion. It has in that sense only as much signification as the principle 

 of increasing atomic weights, one already put to rude test by the 

 respective positions of potassium and argon, iodine and tellurium, 

 nickel and cobalt, so that in this its interest might vanish. The 

 atomic numbers are in themselves only a series of whole numbers 

 without any precise signification. After all, the law of Moseley is 

 only an approximation, since it expresses as a linear relation what 

 experiment shows us more and more is curvilinear. 



One could believe in 1914 (and I sincerely so believed after my stay 

 at Oxford where I worked with Moseley) that, in substituting for 

 the atomic weights the atomic numbers, Mendeleef's theory had a 

 real improvement. Indeed the atomic numbers, susceptible of a 

 measure sufficiently precise according to the distribution of the 

 X-ray in the scale of wave lengths, succeed each oth .r in the definite 

 order conforming precisely to the place occupied by the element or 

 family of elements in the periodical classification. Argon is in its 

 proper place preceding potassium; tellurium and cobalt on the one 

 hand and iodine and nickel on the other are in their proper ranks. 

 All difficulty vanishes. The theory comes out triumphant and 

 strengthen d from the test. 



However, the curvature of the lines of Moseley throws a shadow 

 upon the picture. We might think to find in this succession of whole 

 numbers a certain mystical value. The virtues in numbers which the 

 ancients held has been vigorously decried by }>hilosophers, and their 

 ideas are echoed in modern science. But there is a physical sig- 

 nificance to the numbers of Moseley. Sir E. Rutherford takes one 

 from tlio experiments of Chadwick on the diffusion of the a-rays. 

 But the process is not wholly physical except so far as theory goes, 

 but from a more general point of view metaphysical. On the one 

 hand the diffusion of the a-rays through a bod}^ can lead to the 

 atomic numbers following the ideas of Rutherford; on the other 

 hand these numbers represent the number of positive charges on 

 the nucleus of each atom, or further, the number of electrons which 

 revolve about them. The atoms of Rutherford inspire us with 



