760 Sir J. J. Thomson on tlte 



any doublet are in the same atom. Thus the moments of the 

 molecular doublets will be greater than those of the atomic 

 ones. We should therefore expect molecules in which the 

 atoms are charged to exert much larger forces on other 

 molecules than molecules in which the atoms are neutral. 

 As large forces between molecules correspond to high in- 

 trinsic pressure, surface tension, and dissociating power, we 

 should expect these qualities to be abnormally developed 

 when the atoms are charged. 



Let us consider the evidence bearing on the question of 

 electrical charges on the atom. 



In my book on ' Rays of Positive Electricity ' I have 

 given reasons for thinking that in the molecules of many 

 gases the atoms are not electrified. I will add to these 

 reasons another derived from recent experiments with these 

 rays. On page 94 of ' Positive Rays ' attention is called to 

 the fact that a molecule of hydrogen may, after passing 

 through the cathode, split up into two atoms, the energy of 

 the molecule being equally divided between the atoms, so 

 that each has one half the normal amount of energy. They 

 will, therefore, experience twice the electrostatic deflexion 

 of the normal atom and show themselves as a beading at 

 twice the normal distance from the vertical on the parabola 

 representing the hydrogen atom. I have observed similar 

 effects when the gas in the tube was CO : when a molecule 

 of CO splits up after passing through the cathode, the 

 energy of the carbon atom will be 12/28, that of the oxygen 

 atom 16/28 of the normal energy; hence the electrostatic 

 deflexion of these atoms will for carbon be 28/12 = 2*3, and 

 for oxygen 28/16 = 1*7 times the normal. The beading 

 corresponding to these atoms can be seen in the photographs 

 of the rays through GO, and the deflexions of the beading 

 are as 2*3 to 1*7, Now if the molecule of CO contained a 

 negatively charged oxygen and a positively charged carbon 

 atom, then the oxygen beading would be mainly on the 

 negative parabolas, the carbon beading on the positive one. 

 The photograph shows that the beading occurs on both the 

 positive and negative parabolas of carbon and oxygen, and 

 seems quite as intense on the positive oxygen as on the 

 negative. This result points to the conclusion that in the 

 molecule of CO the atoms are not electrified. 



There are thus compounds in which the tendency of the 

 electronegative atom to acquire a corpuscle is not sufficiently 

 powerful to drag a corpuscle from the atom of the electro- 

 positive element. 



There is, however, evidence from the phenomena of 



