Ionization from Certain Hot Salts. 229 



This value for ejm is the mean of thirty readings in even 

 better agreement than in the previous case. 

 e ,j(c) An attempt was made to see whether freshly prepared 

 AgCl would yield the same results. This specimen was 

 prepared from the purest AgNOo and HC1 obtainable, and 

 kept as far as possible from the light. The salt was pre- 

 cipitated in the form of heavy white flakes, and the greatest 

 difficulty was found in keeping the salt on the strip when it 

 was heated. 



A fresh strip of Pt was inserted and measurements were 

 taken on its emission to check the accuracy of the later 

 experiments with AgCl. It was glowed until no appreciable 

 emission was detected below 1250° C. It was mounted by 

 method (b). 



V. H. 



M. 



T. 



e/m. 



m/H. 



203 2460 



B&S 



1000 ° 



to 



1120° 



790° 



to 

 1130° 



247-5 



38-9 dry salt. 



203 2460 



B 



244-6 



39-4 moistened with 



distilled H 2 0. 



These value 



s were 



the mean 



of a 



number of readings as 



consistent as the former. 



(d) Silver Iodide. This salt was furnished by Eimer & 

 Amend with no specifications as to its purity. A Pt strip, 

 mounting (]:>), was used on which PtCl 2 had previously been 

 tested, but which had been glowed for points below 1200° C. 



The melting point of Agl is given as 527° C. The initial 

 emission was looked for with extreme care, and the first 

 signs of it appeared at 520°. It was far too small to 

 measure, and died away very quickly. A definite emission 

 was found at 575°. 



V. 



H. 



M. 



T. 

 650° 



e/m. 



m/H. 



204 



2460 



B&S 



to 

 1195° 



267-3 



36-2 



These values, about twenty in number, of which these are 

 the mean, were also very consistent although slightly lower 

 than the preceding for AgCl. The difference is probably 

 due to the distance d varying. 



Conclusion. — The values here given for ??i/H for AgCl and 

 Agl do not point to any characteristic silver emission which 

 can be detected, the atomic weight of Ag being 107*88. 

 The numbers were quite consistent and averaged a little 

 lower than the K impurity emission from metals, for which 



