594 Mr. A. M. Tyndall on the 



Discussion of Kesults. 



It would seem at first sight that the drop in field at short 

 distances when P is negative may be explained, as when 

 P is positive, by the theory that at these distances P enters 

 the unclustered region of the discharging N. That is to say, 

 z Q and y would receive a common explanation which would 

 be that they are indications of a region near a discharging 

 point in which some of the ions are unclustered. As a 

 detector of this region the field method would be more 

 sensitive than the wind pressure method because the presence 

 of a very few unclustered ions has a much greater effect on 

 the value of the ionizing field than on the value of the wind 

 pressure. 



There are, however, great difficulties in the way of the 

 acceptance of this theory, and of these the following are the 

 chief : — 



It is not easy to explain on the above view why z is 

 constant for different points. Thus one may imagine that an 

 ion tends to cluster when its velocity has decreased to some 

 critical value ; if the velocity of an ion close to the point is 

 proportional to the field, this implies that growth occurs 

 when an ion enters a certain critical field. Now assuming 

 that the inverse square law of field holds near a point, the 

 values of the fields at say 4 millimetres from the two points 

 A and B may be calculated. Their ratio is about 65 to 1. 

 It is of course probable that the inverse square law does not 

 hold as far from these points as this, but the difference 

 between the two fields is too great to be accounted for in 

 this way. 



Moreover, the following deductions on the growing ion 

 view do not fit the experimental facts. 



When no external ions are supplied to P, glow discharge 

 starts when the positive ions in its neighbourhood ionize. 

 These having just been produced by negative ions are new 

 and consequently unclustered. On this view then, when 

 discharge has once started, the field at such a point is that 

 in which unclustered positive ions ionize. Now when P is 

 negative and external ions are supplied from N at very short 

 distances, the positive ions which start the discharge are 

 again unclustered. The fields in the two cases should there- 

 fore be the same. It is found experimentally, however, that 

 the field at a point opposite a plate is at least six times 

 greater than that at the same point supplied with external 

 ions at small values of y. 



Again, assume that of these two fields the one at a point 



