302 Rev. P. J. Kirkby : Chemical Effects of the 



curves in fig. 2. Thus the curves in fig. 2 correspond to. those 

 in fig. 3, and apply to the same pressures. 



The curves tend to become straight when the distance 

 between the electrodes exceeds a certain value dependent 

 upon the pressure and current. The explanation of this is 

 that the electric force in the positive column (defined as 

 the region terminating at the anode where the electric force is 

 approximately uniform) is independent of the distance between 

 the electrodes. But that distance must be great enough to 

 allow the positive column to be established. The region where 

 ions are generated by collision lies near the cathode ; and 

 though the highest activity of reproduction within that region 

 is limited to a narrow strip close to the cathode, yet an appre- 

 ciable number of ions must be produced outside the domain 

 of highest activity. This will be especially the case at the 

 lowest pressures when the mean free paths of the negative 

 ions are more considerable. In fact the influence of the 

 cathode fall is propagated to a distance of some centimetres 

 from it when the pressure is so low as 2 mms. (See the curves 

 p. 437, J. J. Thomson, ' Conduction of Electricity through 

 Gases ' ; p. 533 in 2nd ed. For a discussion of the distribution 

 of force see c. xvi. ib.) But when the distance D between the 

 electrodes is so large (5 or cms. in the curves just referred 

 to) that the region near the anode is well removed from the 

 region where ions are generated, so that a positive column 

 (as defined above) exists, any further increase in D merely 

 adds to the length of the positive column. 



Let the electric force in the positive column be denoted 

 by Y, then, if the distance between the electrodes is increased 

 by 8 and their potential difference by YS, the pressure being 

 unchanged, the same current will pass as before and nothing 

 will be altered within the discharge except the length of the 

 positive column, which is increased by S : the positive column 

 remains unchanged in character. (See J. S. Townsend's 

 paper, already referred to, Phil. Mag. March 1905.) 



Thus the electric force in the positive column for the 

 pressures of these experiments can be inferred from the slope 

 of the curves of fig. 3, and others plotted similarly, when they 

 became straight. 



The electric force Y in the positive column at various 

 pressures, determined in this way, are given in the following 

 table. It is obvious, however, that the errors of estimation 

 may be considerable, so that the table cannot claim to be 

 more than roughly true except perhaps for the lowest pres- 

 sures, for the curves corresponding to these are accurately 

 determined by the observations. 



