Emanations of Phosphorus through Air and other Media. 395 



The first of these series shows a decrease of conduction 

 from '39 to *33 within two hours, such as would naturally be 

 associated with the waning ionizing power of phosphorus. 

 The succeeding series show similar fluctuations, but the con- 

 duction of the condenser is retained intact after 2 or even 

 3 hours of exposure. It is improbable, therefore, that the 

 number of ions ((/ = ttU<?(A/C<2?)log e) is supplied at a diminish- 

 ing rate in the lapse of time by the phosphorus ; it is much 

 more probable that fluctuations of temperature and like inci- 

 dental causes account for the difference. In the first series 

 the decrease of c' is most marked, being "0007 per minute, 

 while the value "0009 occurs in instances below. Thus in 

 the 15 minutes or less (usually less than 5 minutes) which 

 outlasts the time of a single series of observations, appreciable 

 diminution of the ionizing potency of the source cannot be 

 inferred. 



5. The effect to be ascribed to temperature is much more 

 serious. I made the following observations with the same 

 condenser, endeavouring to obtain the temperature discrepancy 

 by varying the temperature of the room. The observations 

 for c' were completed in the usual way, all being satisfactory. 

 Unfortunately the temperature of the air between the plates 

 is not identical with the superficial temperature of the phos- 

 phorus itself, and it is upon the latter that the ionizing effect 

 depends. Hence the data, some of which are constructed in 

 the curve (fig. 4), often show curious lag phenomena. The 

 observations are difficult, as drafts of air if but slightly too 

 warm are liable to kindle the phosphorus. Different parts of 

 the grid often respond unequally. In a cold room (about 9°) . 

 phosphorus is nearly inert. It then rises rapidly to the large 

 ionizing value between 20° and 30°, but experiments like the 

 present merely emphasize the need of a room of constant 

 temperature. The relation itself is elusive and yet to be 

 found. 



6. After these preliminaries the leakage of the condenser, 

 in the course of time, when the air-space is ionized by 

 phosphorus, is next to be considered. The method was 

 essentially like the preceding, except that longer intervals of 

 time were given to each set of observations. Table II. shows 

 the leakage through the condenser O, P (fig. 2) when the 

 plates are 4 centims. apart, a relatively large distance. The 

 exponential character of the law is clear, but the leakage 

 is throughout faster than can be reproduced by a simple 

 geometric progression. This might plausibly be ascribed to 

 the waning activity of the phosphorus itself, for c' is necessarily 

 a quantity decreasing in the lapse of time with n the number 



