204 THE EAEL OF BERKELEY ON SOME 



The Sulphates. 



The alkali sulphates came from Messrs. MERCK, and were sold as their purest ; the 

 thallium salt came from Messrs. KAHLBAUM. Neither the sodium nor the potassium 

 salts required recrystallising, nor did their solutions require filtering ; analyses of 

 their sulphuric acid contents gave for the sodium salt 67 '37 per cent, (calculated 

 67'57 per cent.) and for the potassium 55'20 per cent, (calculated 55'12 per cent.). 

 Not having purchased enough of the rubidium salt, the balance was made good by 

 treating pure rubidium carbonate (also purchased from Messrs. MERCK) with pure 

 sulphuric acid in just sufficient quantity to neutralise the solution, and then crystal- 

 lising out. The two quantities of salt were then added together and recrystallised, 

 and the crystals examined spectroscopically, in the manner before stated, for potassium 

 and caesium, but with no definite indication of either. An analysis of the sulphuric 

 acid content gave 36'05 per cent, (calculated 35 - 99 per cent.). 



The caesium sulphate was recrystallised three times, and the spectroscopic examina- 

 tion gave no definite indication of either potassium or rubidium. An analysis of the 

 sulphate content gave 26 '62 per cent, (calculated 26'55 per cent.). The thallium 

 sulphate was recrystallised three times and found to be free from lead. An analysis 

 of the thallium content gave 80'96 per cent, (calculated 80'95 per cent.). 



The Nitrates. 



All the salts were Messrs. MERCK'S purest, except the thallium salt, which came 

 from KAHLBAUM. The alkali nitrates were all recrystallised two or three times, and 

 were examined spectroscopically and found to be free from impurities. The thallium 

 nitrate, however, was found to contain some lead ; it was freed from this by repeated 

 recrystallisation. An analysis of the thallium content gave 76'89 per cent, (calculated 

 76-69 per cent,). 



On account of the difficulty of obtaining accurate analyses of the alkali nitrates 

 they were not analysed, but after the first recrystallisation a series of densities and 

 a corresponding series of solubilities at different temperatures were obtained, and 

 these series were compared with similar series obtained from the solution of the 

 crystals of the next recrystallisation. The two differed by no more than the experi- 

 mental errors. 



During the evaporation to dryness in the Jena glass bulbs for the purpose of 

 determining the solubilities, it was found that a trace of nitrate almost invariably 

 came over with the condensed water, and those observations in which more than a 

 trace came over were rejected. It was also noticed that, except in the case of 

 caesium nitrate, the dried salt remaining in the bulbs contained a trace of nitrites. 

 The quantities in both cases were so small that it was not considered necessary to 

 apply any corrections to the resulting solubility. 



