THE STRUCTURE OF THE NUCLEUS. 171 



With regard to the permanent deflection, tlie (jiK'stion arises whetlier the 

 positive charge of the capsule leaks into the needle, or whetlier positive charge 

 leaks out of the needle in other ways, leaving a permanent negative charge. This 

 is determined by the recoil, showing the residual charge ol; the needle to be opposite 

 in sign to the charge imparted. It follows also that the permanent deflection de- 

 creases as the recoil increases. 



In the lapse of time the temporaiy deflections soon reach a maximum, and 

 therefore a nearly constant value ; the permanent deflection then steadily decreases 

 while the recoil increases. The fluctuations of the former are reproduced in the 

 latter. 



Experiments equally unsatisfactory were made with the idiostatic method. 



16. Highly cliarged needle. — I now come to the definite form of appai'atus 

 already suggested in ^ 10. Here the needle is put at high voltage by momentary 

 contact with the ca[)sule, which is kept permanently charged. The potential to be 

 measured communicates with the quadrants, either pair being put to earth alter- 

 nately. To compensate for any leakage from the needle, the barriers of the capsule 

 are phosphorus grids. 



The experiments showed that the sensitiveness remains constant so long as the 

 phosphorus lasts, seeing that the reagent deliquesces in the atmosphere and soon 

 vanishes effectively from the instrument. The damping, moreover, was liable to 

 be excessive, 3 minutes being often necessary to reach a permanent deflection. 

 This was remedied by cutting two neai'ly semicircular holes in the vane, leaving 

 only the parts near the stem to insure rigidity. The [)eriod was thus cut down 

 to 30 sec. The silk bifllar was stretched between hard rubber stems to increase 

 the insulation, and the biiilar distance was made a little larger. The table on page 

 172 contains an example of the results. 



Thus, with the proper needle, a deflection of 5 cm. either side of zero was 

 attained per volt, when the charging potential was but 10 storage cells. The 

 maximum sensitiveness attained was thus «= 2.5 about. 



17. Summary. — From the form given to the damping capsule, it appears that 

 the induction of the pi'eceding section is in excess of the effect of conduction 

 throuo-h the ionized air to the insulated needle. Initially both produce like 

 effects, Ijut the permanent deflections are different. The effect of ionization is 

 indirectly or virtually an escape of the chaige of the needle, which may go far 

 enough to reverse the sign of the deflection. The lesult is much like polarization, 

 and increases in the lapse of time. The arrangement is useless for measurement. 



If, how^ever, the needle is first charged by contact, and the ionized air is then 

 used as a mere vehicle for supplying the leakage of a highly chai'ged needle, the 

 method becomes practicable for certain purposes. My chief purpose in this ap- 

 pendix was to call attention to the capsule, which, in the manner shown in §§ 5 

 and 6, has frequently been useful. 



