10 MR. 0. W. RICHARDSON ON THE IONISATIOX 



producing other ions by collisions is readily seen when the above curves are compared 

 with similar ones given by H. A. WILSON* for the negative leak from hot plat ii nun 

 in air at different pressures. WILSON used wires of the same thickness and ;m 

 apparatus of about the same dimensions as the author, so that the two sets of 

 observations are quite comparable. It will be noticed that with the negative leak a 

 given P.D. changes the current in a given ratio at a much higher pressure than with 

 the positive leak. In addition, the pressures for the maximum current with a given 

 voltage are much lower for the positive than the negative leak. For instance, when 

 V = 340, the maximum current for the negative leak is somewhere between 12'1 

 and 0'81 millim., whereas for the positive leak it is at a pressure somewhere near 

 0'2 millim. Thus for corresponding effects the value of X/p, where X is the electric 

 intensity and p the pressure, has to be much greater when the positive ions are the 

 active agents than when the negative ions are. The actual difference between the 

 effects of the two kinds of ions is diminished owing to the fact that the negative 

 ions which the original positive ions produce by collisions also act as ionising 

 agents. 



Curves resembling some of the above have been obtained previously by 

 McCLELLANDt, who also explained the increase in the current produced by the 

 electromotive force as being due to the positive ions producing others by collisions. 



The author has calculated the number of positive ions which reach a surrounding 

 cylinder when a given number start from a wire of given radius at its centre and 

 both the positive ions and the negative ions subsequently produced give rise to 

 others by collision, in the hope that it might be applied to the further elucidation of 

 the above experimental results. The expression obtained, however, is so complex 

 that its manipulation, so as to fit the observations, did not seem likely to greatly 

 advance our knowledge of the subject. There does not, however, so far as one can 

 see, seem to be anything in these results contrary to the view that the positive ions 

 from hot metals have about the same power as the positive ions produced by Rontgen 

 rays in air of producing other ions by collisions, and that the other ions so produced 

 consist, in any given case, of an ordinary positive ion together with a corpuscle. 



4. Hysteretic Relations between Current and Electromotive Force. 



The numbers given in Table 4, 3, exemplify the curious fact that in some cases the 

 current increased with diminishing potential difference. The case cited was not 

 found to be an isolated instance, as the following numbers, obtained at a temperature 

 of 826 C. and a pressure of 0'4 millim., testify : 



' Phil. Trans.,' A, vol. 202, p. 253 (1903). 

 t 'Camb. Phil. Proc.,' vol. 11, p. 300 (1902). 



