G70 Prof. 0. W. Richardson on the 



are divalent. The results of the measurements of ejm will 

 now be described briefly for each of the salts examined. Q'he 

 method was the same as that described in previous papers. 



Beryllium Nitrate. — The salt used was supplied by 

 Kahlbaum, Berlin. For the freshly heated salt the follow- 

 ing values wore obtained : — e/m = 3G0, and ?n/H = 27. m/H 

 is the mass of the ions which carries the same charge as a 

 hydrogen atom in electrolysis. Additional measurements 

 were made after 16, 23, and 39 hours' heating respectively. 

 They all gave identical values, viz., ejm = 325, and m/H = 29*5. 

 The ionization became too small to measure after this con- 

 tinued heating, and in raising the temperature the strip of 

 platinum melted. 



The values of ??i/H for the ions Be + * and Be ++ would be 

 9*1 and 4*55 respectively. There is thus no evidence of the 

 emission of these ions from beryllium oxide (into which 

 beryllium nitrate is converted when heated). This was 

 rather to be expected, since in the case of the salts of the 

 alkali metals efficiency of thermionic emission diminishes 

 rapidly with diminishing atomic weight, and the salts of the 

 beryllium group are far less efficient than those of alkali 

 metals of neighbouring atomic weight. 



It is probable that the ionization observed with this sub- 

 stance is due to contamination with alkaline impurities (salts 

 of sodium and potassium). It is to be remembered that the 

 thermionic test for these substances is extraordinarily delicate, 

 so that even what are ordinarily known as chemically pure 

 substances may be expected to be quite impure when tested 

 in this way. Quantities of the salts of the alkali metals 

 which would be incapable of detection by the most delicate 

 balance are capable of giving rise to a large thermionic 

 current for a very long time. Mr. C. J. Davisson, working 

 in this laboratory, has recently examined a specimen of 

 beryllium sulphate which was supplied to him by Mr. Parsons, 

 who is one of the leading authorities on the compounds of 

 this element. According to Mr. Parsons the specimen in 

 question is in all probability the purest specimen of a beryllium 

 salt which has ever been prepared. Although Mr. Davisson 

 examined the thermionic emission of this substance most 

 carefully, all the ions which it gave off had a value of ejm 

 exceedingly close to that given by salts of potassium, with 

 which it must have been contaminated. 



More convincing reasons will be given later for believing 



* Throughout this paper a positive ion with a single electronic charge 

 will be denoted by M+, and one with a double charge by M++, where 

 M is the chemical symbol for the atom of the element. 



