282 Mr. C. D. Child on Ionization in the 



ions, and that the positive and negative ions are continually 

 recombining, but these statements are so universally accepted 

 that it is hardly necessary to consider them as assumptions. 



To show that some curve similar to that in the figure must 

 hold good, let us first consider the condition when the 

 density is large. In this case the chief obstacle to the 

 passage of the current is the gas itself. The molectules of 

 the gas stop the movement of the ions, so that their average 

 velocity is small. The electrons quickly combine with 

 molecules, since there are many molecules present ; and it 

 is difficult to produce new ions, since the mean free path of 

 the electrons is so small that the electric force must be very 

 large in order to oive the electrons sufficient velocity to 

 produce new ions by their impact. Clearly the electric 

 force must increase as the density increases. 



On the other hand, when the density of the gas is very 

 small, the gas does not obstruct the movement of the ions to 

 any appreciable extent, but there is still recombination 

 between the negative and positive ions, and there is now 

 great difficulty in producing new ions, due to the fact that 

 there are few molecules to be hit by the electrons. When 

 the number of recombinations becomes comparable with the 

 number of collisions between molecules and electrons, a large 

 proportion of such collisions must result in making new ions, 

 and to cause this to occur the electric force must be large. 

 ]f, for example, the number of combinations were equal to 

 the total number of collisions between electrons and mole- 

 cules, it would be necessaiy for each collision to result in 

 ionization. But some of the paths of the elections will be 

 very short, and to meet the needs of such electrons the 

 electric force must be very large, as shown at the left of the 

 curve. In the limiting case it would be necessary for the 

 electric force to be infinite. 



To arrive at some idea of how the electric force varies 

 between these two cases, we must consider the requirements 

 necessary to maintain a current. The essential requirement 

 is that the electric force be sufficiently iarge, so that new 

 ions are produced by the impact of the electrons on the 

 molecules as fast as they recombine. An electric force must 

 be present also in order to move the ions after being formed, 

 but this requirement appears to be met whenever the first 

 one is met. If it were not, we should find that any increase 

 in the current would require a corresponding increase in the 

 electric force, and the force would be directly proportional 

 to the current, as it is in metallic conduction. In reality the 

 electric force decreases slightly whenever the current becomes 



