490 Mr. A. B. Meservey on the Potentials required 



does not follow any one curve of figs. 5 to 7. But it is of 

 the same general form as all of them except for the drop at 

 the two smallest diameters, and undoubtedly the change in 

 distribution of force in the field would have the same kind 

 o£ effect on the minimum sparking potential as on the 

 others. Judging from the relative positions of the curves 

 in figs. 5 to 7, the decrease in the critical pressure with the 

 diameter should mnke the positive curve in fig. 9 less steep 

 than the other positive curves, as it is, but this does not 

 seem to account for the drop which occurs at the smallest 

 diameters. But these two diameters are the only ones for 

 which the critical pressure falls below the pressure at which 

 the radiation effect seems to become large, that is, below the 

 point at which the curves cross in fig. 3. The presence of 

 the radiation would explain the lowering of the minimum 

 sparking potential at these diameters for positive inner 

 electrode, and therefore the sudden change in the curve in 

 fig. 9. If, then, we consider the continuous change in the 

 relative positions of the positive and negative curves in 

 figs. 5 to 7 as due to continuous changes in a and /5, we can 

 explain all the curves obtained from these experiments by 

 the collision theory and the presence of radiation at 

 pressures in the vicinity of the critical pressure. 



The results of these experiments may be summarized as 

 follows : — 



1. For every value of the diameter of the inner cylinder, 

 two different sparking potential curves are obtained, cor- 

 responding to the two directions of the field. For smaller 

 diameters, the two curves cross near the critical pressure, 

 and that for a negative inner electrode is higher on the side 

 of the lower pressures, but they tend to coincide as the field 

 approaches that between parallel plates. 



2. The minimum sparking potential depends on the 

 diameter of the inner electrode, and is always lower when 

 that electrode is negative. 



3. The minimum sparking potential is always higher than 

 that for parallel plates when the inner electrode is positive, 

 while the lower values when the inner electrode is negative 

 are below that for parallel plates. The lowest values 

 obtained were 311 volts for air and 240 volts for hydrogen. 



4. Curves drawn to show the relation between diameter 

 and sparking potential at constant pressure exhibit a rise as 

 compared with curves drawn for parallel plates if the inner 

 electrode is positive. The negative curve is lower, and at the 

 higher pressures diverges sharply from the positive for the 

 smaller diameters, but the divergence lessens with the pressure. 



