200 Mr. H. H. Poole on the Temperature Variation 

 Temperature 159°'5 C. 



X. 



k. 



C. 



X. 



*. 



C. 



0-64 



7-95 



0-027 



1-62 



8-50 



0-413 



0-85 



8-22 



0-047 



1-64 



8-54 



0-436 



0-98 



8-24 



0-072 



1-75 



8-62 



0-553 



1-13 



8-37 



0-115 



1-83 



8-60 



0-688 



116 



8-43 



0-121 



1-93 



8-58 



0-853 



1-30 



8-47 



0-181 



1-95 



8-62 



0-853 



1-37 



8-49 



0-218 



1-97 



8-68 



0-953 



1-39 



8-45 



0-234 



2-00 



8-58 



0-997 



1-42 



8-50 



0-247 



201 



8-62 



1-002 



1-55 



8-56 



0-346 



203 



8-62 



1012 



1-58 



8'25 



374 



2-07 



8-56 



1-105 



1-58 



8-57 



0-407 









The mean values of k at different temperatures are shown 

 below. Here t is the temperature C. recorded by the thermo- 

 meter, t' is that calculated from the conduction current by 

 the formula given later, 8 the difference between these two 

 values, and k the dielectric constant. 



t. 



t'. 



5. 



Je. 



13-o 



— 



— 



7-97 



16 



— 



— 



8-43 



50-5 



52 



-1-5 



8-22 



81 



79 



4-2 



8-16 



84 



84-5 



-0-5 



8-41 



111-5 



111-5 







8-50 



143 



142-5 



+0-5 



8-58 



159-5 



159-5 







8-48 



166 



166 







8-39 



186 



184 



+2 



8-52 



227-5 



227-5 







8-71 



229-5 



232 



-2-5 



9-06 



It will be seen from the iirst of the above tables of figures 

 that at 159°*5 C. k rose slightly with X, tending to an 

 approximately constant limit for large values of X. Similar 

 results were obtained at all temperatures except the two 

 highest, where k, which was approximately constant for 

 small values of X, tended to drop for larger values. The 

 second table shows that the average value of k appeared to 

 rise with the temperature. 



It seems probable that the apparent variations of k may be 

 entirely ascribed to the brush-discharge. We may expect 

 the effect of this to be two-fold, (a) an increase of the 

 effective area of the central disk, and (b) an encroachment 

 on this area by the brush-discharge of the guard-ring, 

 causing a loss of charge and consequent reduction of k when 

 the pressure is sufficiently high. When X is small the effect 



