CONTEMPORARY ADVANCES IN PHYSICS 



97 



The data which I have thus far cited pertain to gap-widths con- 

 siderably smaller than the radii of curvature of the electrodes: fairly 

 close approximations to the extreme case of infinite parallel planes. 

 Experience with steady voltages suggests that what really counts is 

 probably not the absolute value of gap-width, but its ratio to the radii 

 of curvature (or to the smaller of the two, if the electrodes are not 

 alike). The foregoing data then show that as this ratio increases, 

 there is a diminution of breakdown-voltage at the higher frequencies, 



o^o 



14 16 18 20 22 

 d IN MILLIMETERS 



24 26 28 30 32 34 36 



Fig. 14 — Curves of sparking-potential vs. gap-width, in air, between spherical 

 electrodes of the indicated radii, at frequencies of the order lO'^ (except the top- 

 most). (Kampschulte, Arch.f. Elektrolech.) 



setting in earlier the larger the ratio is. Continuing in this line of 

 thought, we infer that as we approach the opposite extreme case of 

 sharply-curved or pointed electrodes at distances many times as 

 great as their radii of curvature, the diminution w ill begin at very low 

 frequencies and will be considerable. 



This occurs, and is illustrated by Figs. 14 and 15 (from Kampschulte) 

 the former of which shows the breakdown-potentials between spheres 

 of the indicated radii, over the range of distances indicated along the 



