480 BELL SYSTEM TECHNICAL JOURNAL 



element have the same weight, which (multiplied b>- the proper uni- 

 versal factor) is the combining weight of the element; whereas now 

 it is known that some of the elements ha\-e two or se\'eral different 

 kinds of atoms apiece, with different weights, of which the observed 

 cf)mbining weight of the element is merely an average. The com- 

 l)ining weight of an element as observed in ordinary chemical experi- 

 ments has no general right to the title of atomic weight; only in special 

 instances may the two be identified. The elements of which the 

 combining weights are integers — meaning, integer multiples of I'n of 

 the combining weight of oxygen — consist of atoms of a single kind, 

 the weight of which is truly and accurately given by the combining 

 weight of the substance. The others, or those of them which ha\e been 

 analyzed, are mixtures of atoms of different kinds, the weight of no 

 one of which isgi\en by the combining weight of the element. \\'here\er 

 it is actually the mass of an atom which is measured by the chemical 

 method, the rule is verified; where the rule is apparently infringed, 

 the quantity measured is merely a misleading average, and not the 

 mass of an atom at all. When, therefore, the rule is restated to apply 

 only to those combining weights which are truly atomic weights, 

 the conspicuous exceptions no longer militate against it, and the 

 supposition that all atoms ma\' I)e built of hydrogen atoms is strongly 

 reinforced. 



When J. J. Thomson developed the technique of his "positive-ray 

 analysis" by which he measured the masses of fast-flying charged 

 atoms and molecules, he was unknowingly preparing the way for 

 ascertaining how many different kinds of atoms belong to a single 

 element. In these classical experiments the ionized particles were 

 those existing in a rarefied gas traversed by an electrical discharge, 

 and drawn to the cathode by the strong field maintaining the dis- 

 charge; through a narrow perforation in the cathode, a thin pencil 

 of the ions passed into a chamber where it was subject to crossed 

 electric and magnetic fields. These fields resolved it into a number 

 of separated and separately-directed pencils, each containing ex- 

 clusively atoms (or molecules) of a single uniform mass, which could 

 be deduced from the location of the trace made by the pencil upon a 

 photogra[)hic plate.'" The method was designed by Thomson as a 

 sensitive, indeed a supersensitive, method of chemical analysis, by 



"The actuality is somewhat more comple.x, as a distinct pencil is obtained for 

 each value of the charge-mass ratio E m, and it is this ratio which is deducihle from 

 the location of the pencil. However E is either the electron-charge r or a small 

 integer multiple of it ((xcasionally, but rarely, as great as 8f), and the iiuilliplicity 

 of pencils corresponding to different values of E and a single value of m seems to 

 be an actual advantage to the experienced interpreter of such data. 



