246 Mr. A. T. Waterman on the Positive 



correct value, 39. Here the mean value is distinctly 

 above this value. Furthermore, in no other case of heating 

 a salt on a Pt strip have such high jumps in the values of 

 m/H been experienced. Thus it may be possible that an 

 emission of a heavy ion such as Mo + occurred, which most 

 of the time was small compared to the K-emission, but 

 which showed itself in spots where perhaps a fresh piece of 

 Mo$ 2 had come into contact with the hot strip. No satis- 

 faction could be obtained with the slit method, principally 

 because the emission decayed so fast. Another feature of 

 this experiment is the high temperature reached before 

 emission was detected. This maybe due simply to the small 

 amount of the material heated as compared with the experi- 

 ments with molybdenite strips. 



In interpreting these results one notices the more or less 

 constant presence of the familiar K + -emission, which for 

 certainly a large part of the time renders indeterminate any 

 other emission that may occur. 



The next thing to strike the attention is the occasional 

 high value of m/H, pointing to the emission of some heavy 

 ion, the character of which is most probably a singly-charged 

 atom of molybdenum. The average of these high values 

 for m/H agrees with this conclusion, as it lies in the neigh- 

 bourhood of 96. It does not seem possible that an emission 

 of Mo ++ ions was obtained, although if such an emission 

 occurred its presence would be next to impossible to 

 determine. 



This characteristic emission usually attains a measurable 

 value at the start of the break, about 310° C, and may be 

 detected anywhere beyond this point, depending upon the 

 intensity of the accompanying impurity emission. It does 

 not appear to be connected solely with one state or the 

 other, or with the period of transition. More than this 

 cannot be stated with any conviction. 



9. Summary. 



It appears from the foregoing experiments that no 

 characteristic positive emissions are given off by the following 

 salts:— AgCl, AgT, PbC) 2 , PbBr 2 , PtCl 2 , CuCl 2 . The same 

 statement is probably true of A1F 3 . In the case of Cu 2 Cl 2 

 there are indications of Cu ++ . Thus the investigation has 

 not shed much light on the question of the connexion between 

 valency and the charges on the ions mentioned at the 

 beginning of the paper. In nearly every case positive ions 



