512 Prof. J. J. Thomson on the Relation between the 



atoms sometimes carries a negative charge, sometimes a posi- 

 tive one. But though in the gaseous state the atoms do not 

 restrict themselves to charges of one sign, there are many 

 phenomena which prove that even in this state the atoms of 

 the different elements behave differently with respect to 

 positive and negative charges, v. Helmholtz explained this 

 behaviour of the elements by supposing that there is a specific 

 attraction between the atom and its charge ; that the zinc 

 atom, for example, attracts a positive charge more powerfully 

 than it does a negative one, while an atom of chlorine, on the 

 other hand, attracts a negative charge more powerfully than 

 it does a positive one. 



The connexion between ordinary matter and the electrical 

 charges on the atom is evidently a matter of fundamental 

 importance, and one which must be closely related to a good 

 many of the most important chemical as well as electrical 

 phenomena. In fact a complete explanation of this connexion 

 would probably go a long way towards establishing a theory 

 of the constitution of matter as well as of the mechanism of 

 the electric field. It seems therefore to be of interest to look 

 on this question from as many points of view as possible, and 

 to consider the consequences which might be expected to 

 follow from any method of explaining, or rather illustrating, 

 the preference which some elements show for one kind of 

 electricity rather than the other. 



The following method of regarding the electric field seems 

 to indicate that this effect may be illustrated by simple dyna- 

 mical appliances. The charges on the atoms are regarded as 

 the ends of Faraday tubes (see J. J, Thomson, ' Recent 

 Researches in Electricity and Magnetism,' p. 2): each unit 

 of positive charge is the origin, each unit of negative 

 charge the termination of such a tube. The magnitude of the 

 unit charge is here taken equal to the charge carried by a 

 monovalent ion. When these tubes spread into the medium 

 they give rise to Maxwell's Electric Displacement, and the 

 motion of the tubes through the medium produces a magnetic 

 field. 



Now suppose that the medium forming these tubes pos- 

 sesses rotatory motion: we may imagine, for example, that the 

 tubes are bundles of vortex filaments, the axis of rotation 

 being parallel to the axis of the tube. The total amount of 

 vorticity which starts from any solid totally immersed in a 

 liquid is zero ; to satisfy this condition, we may suppose that 

 there is slipping between the walls of the bundle of vortex 

 filaments and the surrounding liquid, or, what amounts to the 

 same thing, that there is on the surface of the bundle a film 



