Positive Ionization from Hot Salts. 



(387 



from this substance. According to Mr. Davisson's experi- 

 ments the value of ra/H for the carriers from this material 

 was 22, indicating that they were atoms of sodium. The 

 current-E.M.F. curves were similar to those already obtained, 

 except that the straight part of the curve had a rather steeper 

 slope and there was more variation from one pressure to 

 another than the average. This might be explained if the 

 central electrode were nearer to the platinum tube than 

 usual. 



Two current-pressure curves with -f40 and -f 160 volts 

 were taken at temperatures a little under 800° C. Thev 

 were practically identical. The results of the one with 4- 160 

 volts are exhibited in fig. 3. Curve 1 gives the currents 

 with increasing, and curve 2 with diminishing, pressures. 

 There was a gradual fail of temperature during this experi- 

 ment from 789° C. to 774° C. Curve 3 is taken from Garrett's 



Fi<?. 3. 



!00 



SO 



§60 

 ^40 



20 







4 -""--- - u 













't ^^^ 



^^•^^ 



^^o 





1 \\ 













1 1 ^0 





r^!5*%>"3> 





AlPO* 





2. ' 





J!^^--^^© 1 





Of/*. 





10 15 



S%>£SSUR£ (AfMS.) 



eo 



25 



paper and represents the emission as measured by the strip 

 method at 1005° C. If the data were reduced to the same 

 conditions I think my currents would be larger than those 

 obtained by Garrett at all pressures. Curve 4 will be 

 considered later. 



The chief difference between curve 3 and the mean of 

 curves 1 and 2 is in the behaviour at low pressures. The 

 tube method shows no indication of the large increase to a 

 maximum given by the strip method under these conditions. 

 On the contrary there is a uniform diminution of the emission 



2Z2 



