EXPERIMENTS WITH IONIZED AIR. 



65 



question an additional and specially severe test as follows, showing that only an 

 insignificant part of the nnclei 2>resent take pai't in the convection of electricity 

 radially across the condenser. 



The question is as yet an open one, whether, after all, the condensation may 

 not be promoted by a distinct set of nuclei from those which conv^ey the electric 

 current. The two, however, must iu the same series be produced propoi'tionally to 

 each other, the nuclei being later stages of growth or clusters of ions, or produced 

 iu other ways one from another, for instance. The factor may vary in different series. 



The observations were made by looking down the color tube, figure 1, through 

 the window «, after adjustment had been made at F' iov a given color and then 



Fig. 3. 



"0 I 2 12301 Z0l2i 



Fig. 3. — Simultaneous Values of the Electric Conduction, and Volumes per Minute of the 

 NEARLY Saturated Emanation Traversing the Condenser. 



alternately charging and discharging the condense)-. For potential differences of 

 60, 150, and 300 volts and for all available colors, not a trace of color fluctuation in 

 the steam tube was to be observed, due to charge or discharge. A slight turn at 

 i^ immediately changes the tint. 



The result admits of quantitative expression. From the chart it appears that 

 color differences of .06 lit/min (blue = 1) can certainly be detected. This 

 amounts in the blue region to 6^ of the total number of nuclei entering. The 

 [)oteutial gi-adient was 300/.141 = 2100 volts/ceutim. Hence even in this strong 

 field the total number of nuclei is still overwhelmingly large as compared with the 

 number conveying current, unless different nuclei discharge the divers functions. 



5. Remarhs on the olservafiotis. — On the other hand, however, the cui-i'ent 

 varies with the number of nuclei i)resent, no matter how small the relative quantity 

 actually promoting electric conduction. This is well shown in the chart, figure 3 ; 

 also in figure 4, which both graphically reproduce table 1. The radial electrical 

 currents or their equivalents, x, are given as ordinates, the corresi)onding volumes 

 per minute ( V) of the air passing longitudinally through the condenser, as abscissas. 

 The lelation of saturated volumes per minute longitudinally, and coulombs per 

 second radially is thus apparent for the 2d and -Ith series. 



