Striated Electrical Discharge, 527 



electric force must consist of a great number of ions which 

 are not at rest, but are moving equally in all directions. 

 This accumulation of ions will be more accurately described 

 as a "nucleus" than as a u slab." There is no longer a 

 sudden discontinuity of densities in the middle of such a 

 structure ; this discontinuity must be replaced by a rapid but 

 continuous change. The number of ions of either kind in a 

 nucleus is being continually increased by arrivals from the 

 adjacent nuclei, and also by the dissociation of some of the 

 molecules which pass into it in the course of their motion ; 

 the number is at the same time decreased by the ions which 

 are carried away from the nucleus, and also by the recom- 

 bination of ions inside the nucleus. When the steady state 

 is reached the losses and gains arising from these four causes 

 exactly balance. 



It will be noticed that the view which we have arrived 

 at, and which regards the discharge as a series of smaller 

 discharges, is almost identical with that put forward by 

 Spottiswoode and Moulton, and others *. 



Comparison of Theory with Experiment. 



§17. The theory which has been developed attempts to 

 predict the arrangement of light and darkness in a discharge- 

 tube, so that a comparison of these predictions with the arrange- 

 ment actually observed will, to a certain extent, supply a 

 test of the truth of the theory. A second and more searching 

 test is afforded by a comparison of a theoretical graph with 

 graphs plotting out the result of actual measurement. A 

 series of measurements of the intensity of electric force at 

 different points of a discharge-tube has been made by 

 Graham f , and there is a second and more recent series by 

 Mr. H. A. Wilson J. Both experimenters have given their 

 results in the forms of graphs for X, so that comparison will 

 be facilitated by imagining the theoretical graph for y changed 

 into a graph for X. 



We are only concerned with the main features of the 

 theoretical graph for y, and it is obvious that these will be 

 reproduced in the graph for X. In particular, the maxima 

 and minima of the two graphs will correspond, although the 

 same is not true of points of inflexion. Unlike the graph for 

 y, the graph for X will always meet the axis of x at right 



* Spottiswoode and Moulton, Phil. Trans. 1880, p. 201. See also J. J 

 Thomson, Recent Researches, § 217 ; E. Goldstein, Phil. Mag. x. pp. 178- 

 190 ; A. Schuster, Phil. Mag. xlvii. p. 554. 



f W. P. Graham, Wied. Ann. lxiv. p. 49. 



% H. A. Wilson, Phil. Mag. xlix. p. 505. 



