IMIOnrCED BY HOT PL. \TINT\1 IN KIKFKKKNT (lASKs. 



comparable with their superficial an-as. 'I'o allow tin- this. "'> rentim. was 

 subtracted in every case from the length of the wire. The area thus reduced will 

 be referred to as the effective area of the surface of the wire. 



A wire 0*1 millim. in diameter, which had been heated in oxygen for a period of 

 about three months (usually for several hours a day), was found to give the following 

 values of the positive and negative saturation currents at the temperatures stated : 



The pressure of the oxygen in the above experiment was 1'47 millims., and the 

 effective superficial area of the wire was 0'223 sq. centim. In this experiment the 

 negative ionisation was measured with 40 volts P.D. This would probably give 

 rise to ionisation by collisions at the pressure during the experiment, so that the 

 values for the negative ionisation in the table are probably somewhat greater than the 

 true saturation values. 



In proceeding from a low to a high temperature it was usually found that the leak 

 at the high temperature was too big at first and subsequently fell to a smaller value. 

 This effect did not, however, occur in a wire which had been heated in oxygen for a 

 very long time for instance, it was not noticed in the above series of olwervations - 

 ;iiul is probably due to the wire not really having reached a state of equilibrium. 



It will be noticed that the value of the positive ionisation increases very rapidly 

 with the temperature of the wire, though not so rapidly as the negative corpuscular 

 radiation. The relationship between the two is brought out more clearly when the 

 results are exhibited graphically as in fig. U. The continuous curves with points thus x 

 represent the positive, and the dotted curves with points thus o the negative, 

 ionisation. The scale of current is different for the different curves. For curves (1) 

 and (4) the unit is 10~ 13 amp&re, for (2) and (5) 1 = 10' 11 ampere, and for (3) and (6) 

 1 = 10~'" ampere. The two leaks become equal at ataut 1240 C., though the 

 positive is fin- Ki^er than the negative at low temperatui. -. 



Although the negative ionisation increases far more rapidly than the positive, both 

 ilfpfiul on the temperature in the same general kind of way ; in fact, the positive 

 ionisation, so far as its temperature relations are concerned, obeys the law originally 



