416 BELL SYSTEM TECHNICAL JOURNAL 



that the atoms are elastic rigid spheres with a diameter of about 

 10"' cm. Unlilce as an elastic rigid sphere and a cloud of electrons 

 seem, this agreement between so difTerenth' made estimates is proli- 

 ably no mere coincidence. It will be noticed that all of the figures 

 about sizes at which we have arrived in such various \va>'s (diameters 

 for the electron and the nucleus, for the vacant space inside the 

 electron-cloud, for the entire atom) are quite compatible with one 

 another. If the value derived for the diameter of the interior hollow 

 had been ten times the spacing of atoms in a crystal, or a tenth the 

 diameter of a spherule of electricity with the same electromagnetic 

 mass as an electron, we should indeed be in trouble. 



In the second place, these studies of the deflections of alpha-particles 

 yield numerical values for the nuclear charge : (77.4 ± I)e for platinum, 

 (46.3±0.7)e for silver, (2!).3±0.7)«' for copper, 19e for argon, (iJie 

 for "air" (a sort of statistical average of the \alues for oxygen and 

 nitrogen).'" To these must be added the \alue +2e for the nuclear 

 charge of helium; for we have alrcach' seen the evidence that the 

 alpha-particle is what is left of a helium atom when two electrons are 

 renio\X'd, and these last-cited experiments show that it is itself a 

 nucleus, hence a helium nucleus. This nuclear charge must be bal- 

 anced by negati\e charges within the atom; if this balancing negati\e 

 charge is subdivided into electrons, then the numerical factors of e 

 occurring in these numerical values are equal respectively to the 

 number of electrons belonging to each atom. We thus ha\e f,iirl>- 

 accurate estimates of the number of constituent electrons witiiiii 

 each of four or live atoms. 



These estimates agree, within their experimental uncertainties, 

 with the famous and splendid idea of van den liroek and Moseley : 

 that the number of electrons in each atom, and the nuclear charge 

 measured as a multiple of the electron-charge, "is the same as the 

 niMuber of the place occupied by the element in the periodic tai>ie". 

 'I'his idea is also supporteti by rough measurements of aliiha-particle 

 dellections by a few other atoms, and by the extent to which (iitTereni 

 atoms .scatter X-ra\s; but the most important of the adililional 

 evidence will find its appropriate place in the second section of this 

 article. 



These conclusions are almost .ill tiial c.in be (k< I need from the data. 

 The arrangement of electrons within the electron-cloud is almost 



'" KcfcroiK cs (iir ihcsu ilata an- given in llu' fourth article of this scries. The 

 "lal.i ot.tained hy K. .S. Uielcr (l'r(x-. Ko\'. Sck., 10S.\, pp. 4.?4-4.S(), 19241 show 

 ineiilcnlally, if I do not misread his article, that the nuclear charges of Mg and .Al 

 have the desired values 12e and lie, rcsix-ctively, within a few per cent. Kulher- 

 ford's studies of encounters between alpha-jiarticlcs and hydrogen atoms prove 

 a nuclear charge of e for the latter. 



