930 Dr. R. D. Kleeman on Disintegration of an Ion 



initial number of free ions were available for the production 

 of further ions by collision. The fields corresponding to this 

 part of the curve therefore assist in the disintegration of the 

 clusters. The effect cannot be due to the positive ions pro- 

 ducing further ions by collision. For it is obvious from the 

 curves A and B that over the region of fields o d the number 

 of ions for which a positive ion is responsible or engenders is 

 extremely small in comparison with the number for which a 

 negative ion is responsible. It may of course be argued 

 that the positive ions drawn through the gauze consist of 

 elementary ions and clusters, of which the former only would 

 be immediately available for engendering the production of 

 further ions by collision, and that the bend of the curve B is 

 therefore further from the origin than would be the case if 

 all the positive ions drawn through the gauze were in an 

 elementary state. The evidence that the effect under dis- 

 cussion is due to the breaking up of negative clusters is the 

 fact that it becomes more evident when the ionization by 

 collision of positive ions decreases in comparison with that by 

 negative ions, which, we have seen, is brought about by 

 decreasing the distance between gauze and plate. Further 

 direct evidence that this is the correct explanation will be 

 discussed later. 



If we calculate n the number of free ions drawn through 

 the gauze that would account for the total leak corresponding 

 to the field 1120 volts per cm., we obtain 



?i =3003/<?' 5xl3 ' 3 = 3'9. 



Now the positive leak obtained for the field 880 volts per cm. 

 is equal to 4, and is approximately proportional to the total 

 number of negative ions drawn through the gauze. Since 

 these numbers are nearly equal to one another, it follows that 



when the quantity — has reached the value 1120 practically 



all the clusters get broken up on their passage from gauze to 

 plate. 



The velocity of a cluster corresponding to the field 1120 

 volts per cm. calculated on the supposition that it is the same 

 as that of an ion as measured in practice and proportional to 



the electric field, is equal to 760 x '85 x 1120 = 7-24 X 10 5 — * 



It is however greater, since we have seen that the velocity 

 of a cluster increases at a greater rate with increase of 

 electric field than proportionally to it. It is thus much 

 greater than the average velocity of translation of a molecule 



