Ionization from Certain Hot Salts. 235 



The values, some forty in number, decreased steadily with 

 rise in temperature, and slightly with time at constant tem- 

 perature, from 36 to 27, and then rose with increasing 

 temperature to 34. The values by the slit method did not go 

 lower than 30, but observed the same decrease and later rise. 



(U) As the PtCl 2 used in the above determination was 

 made in a glass vessel, and as glass is known to be quite 

 soluble in aqua regia, a further quantity of PtCl 2 was pre- 

 pared in a platinum crucible, so as to be more completely 

 rid of this impurity. This does not remove it altogether, 

 however, as glass is soluble to a slight degree in the HC1 

 used. About sixty readings were taken : — 



V. 



H. 



M. 



T. 



e/m. 



m/H. 



204 



2460 



B 



780°' 



256-1 



37-6 









to 



to 



to 









945° 



264-3 



36-5 









to 



to 



to 









1130° 



256-5 



37-5 



V 



•>i 



S 



930° 



to 

 1100° 



256-1 



37-6 



The values were remarkably consistent and did not show- 

 any appreciable decrease until after long heating at 930°. 



Conclusion. — No emission was found characteristic of Pt, 

 unless it be taken as significant that PtCl 2 and a Pt strip 

 give practically the same value of m/H. At any rate there 

 appears to be no connexion with the atomic weight of 

 platinum. 



The presence of glass as an impurity would warrant the 

 lowering of m/H due to Na compounds contained. That 

 this is the case becomes more plausible when in (b) precau- 

 tions are taken to keep the glass out as far as possible. The 

 same decrease occurs but to a much less degree. 



The emission seems to be without doubt due to K as an 

 impurity. 



6. Sources of Impurities. 



As has already been touched upon, it is well known that 

 a Pt strip, in fact apparently any strip of metal, no matter 

 how carefully it is cleaned, gives off a decided positive 

 emission when heated. It is quite persistent and does not 

 appear to be characteristic of the metal, but may most 

 probably be due to potassium in some combination as an 

 impurity. 



Now it has been shown * that a Pt strip which had lost 

 * Horton, Roy. Soc. Proc. A. vol lxxxviii. p. 131 (1912). 



