piionm-:n i:v HOT PT.ATINTM IN DIFFERENT GASES. 



41 



gram equivalent of each kind of ions being respectively <o + = 4 '92x10* and 

 a*- = 8'97xl0 4 calories. The value of o>_ for oxygen was 13'5xl0 4 calories; the 

 lower value obtained in air is possibly due to 40 volts not being sufficient to 

 produce saturation. Experiments were not made to test this at the time, but some 

 results obtained by H. A. WILSON* indicate that this might have been the case. The 

 value of 4 '92 x 10* for <u + is intermediate between the values 3 '04 x 10* and 7 '12 x 10* 

 found for oxygen and nitrogen respectively. 



Referring to the above table, we find for the maximum current in air the values 

 2'2xlO~ M and 5-2xlO~ u ampere at 893 C. and 900 C. respectively. Taking the 

 mean of these, 37 x 10~" ampere, to be the true value at 897 C., and dividing by 0'66 

 to reduce to unit area, we find for the maximum current at 897 C. the value 

 5'5 x 10~" ampere per square centimetre. The corresponding quantity in oxygen 

 deduced from fig. 9 (here we have to multiply by 3'3 on account of the low pressure) 

 is 5'8 x 10~ 10 ampere per square centimetre, or ten times the corresponding quantity in 

 air. Again, referring to fig. 12, we see that the maximum leak in air at 976 C. is 

 4'5 x lO" 10 ampere per square centimetre, a number which agrees satisfactorily with the 

 value 4 < 95xlO~ 10 ampere per square centimetre given by the table on p. 40. The 

 maximum leak in oxygen deduced from fig. 10 is 12'2xlO~ w ampere per square 

 centimetre, nearly three times the value for air. It is interesting to compare with 

 these numbers what the values for air would be if the nitrogen had no effect. Since 

 air contains 21 per cent, of oxygen by volume, on this basis the leak in air at 

 760 millims. would be the same as that in oxygen at 160 millims. Referring to fig. 7, 

 we see that this should be equal to 0775 of the value in oxygen at 760 millims. 

 This would give 44xlO~ 11 and 9'5xlO~ 10 ampere per square centimetre at the two 

 temperatures instead of the values 5'5x 10~" and 4'5x 10~ 10 actually found. 



This result, that the positive ionisation in air was smaller than if the nitrogen had 

 been removed, seemed at first sight very surprising, and further experiments were 

 made to test the question directly. This was done by letting air and oxygen into 

 the apparatus immediately after each other and measuring the leak at atmospheric 

 pressure. The experiments were carried out at 895 C. The values obtained are 

 given in the subjoined table, that in nitrogen being added for comparison. The 

 experiments were carried out in the order of the table, proceeding from above 

 downwards. The unit of current is 7 x 10~ 1S ampere. 



VOL. CCVil. A. 



* ' Phil. Traiia.,' A, voL 202, p. 266, tig. 5. 

 O 



