The Discharge of Electricity through Gases. 555 



recombinations and vice versa. Such a tube would show phenomena 

 very similar to stratifications. This is only a suggestion to show that 

 the theory may ultimately be found sufficient to cope with this diffi- 

 culty. At present it seems to me to be an open question whether the 

 stratifications are ever seen in perfectly pure gases. 



The Bark Space. 



No satisfactory explanation has yet been given of the division of 

 the appearance round the kathode into three parts : the first luminous 

 layer, the dark space, and the glow. The division between the dark 

 space and the glow is often very sharp, and it is necessary to discuss 

 how the rapid change in luminosity can be accounted for. It has 

 been suggested that the extent of the dark space represents the mean 

 free path of the molecules. If particles are projected from the 

 kathode at low pressures, comparatively few will impinge in its 

 immediate neighbourhood, but with increasing distance the number 

 of impacts will increase. It has been pointed out by others that the 

 extent of the dark space is really considerably greater than the mean 

 free path of the molecule, calculated according to the ordinary way. 

 My measurements make it nearly twenty times as great. This, 

 however, is not in itself a fatal objection, for, as we have seen, the 

 mean free path of an ion may be different from that of a molecule 

 moving among others. I cannot, however, reconcile the sharpness of 

 the inner boundary of the glow with the explanation given. If the 

 luminosity only depended on the number of impacts, we should 

 expect the parts adjacent to the electrode to be dark and gradually to 

 increase in intensity outwards. The positive ions approaching the 

 kathode would still further reduce the difference in luminosity. I 

 have endeavoured to ascertain the experimental conditions which 

 determine the shape of the boundary of the dark space. The first 

 supposition tested was, whether the boundary was always an equi- 

 potential surface. If so, the velocity of the particle projected from 

 the negative electrode would be the same all over the boundary, and 

 we might imagine that the luminous appearance of the glow depends 

 on some minimum kinetic energy which the impinging particles must 

 possess. The darkness is, however, not limited by an equipotential 

 surface. 



A large cylindrical vessel contained two negative electrodes parallel 

 to the axis of the vessel. The anode was formed by a cylindrical 

 wire netting surrounding the kathodes. Under these circumstances 

 the dark space and glow present some peculiarities, which I shall 

 describe on another occasion. The shadow phenomena described by 

 Goldstein are beautifully seen, and can be photographed, and as the 

 sides of the glass vessel do not interfere (as it now appears they did 



vol. xlvii. 2 s 



