PHYSICAL CONSTANTS OF SATURATED SOLUTIONS. 205 



The Alums. 



The potassium alum purchased as pure from Messrs. MERCK was found to contain 

 a small quantity of both iron and ammonium, and repeated recrystallisation did not 

 purify it. Pure aluminium and pure potassium sulphates were therefore purchased, 

 and the pure alum made from these. An analysis of the sulphuric acid content of 

 this salt, when recrystallised, gave 40 '33 per cent, (calculated -40 '49 per cent.). The 

 remainder of the alums, purchased as pure from Messrs. MKRCK, were recrystallised 

 several times, and the spectroscopic examination showed no impurities. Analyses 

 gave, for the rubidium salt, 36 - G9 per cent. SO (calculation being 36'89 per cent.), 

 and, for the caesium salt, 33'67 per cent. SO (calculation being 33'80 per cent.). 



The thallium alum was analysed by determining the thallous sulphate content, 

 giving 39 - 29 per cent., the calculated value being 39 '43 per cent. It will be noticed 

 that the tables of results give no values for potassium alum above 60 C., for the 

 rubidium alum above 70 C., and for thallium alum above 60 C. this is because it 

 was found that the prolonged heating at (>8 ( <., at 80 C., and at 7f> ( '. respectively 

 decomposed the solutions, and a white insoluble precipitate was formed. The c;t'siuni 

 alum, however, could be heated to the boiling-point without decomposition. 



The solubilities of the potassium and rubidium alums could not be determined 

 to so close a degree of accuracy as that of the other salts, for it was almost 

 impossible to dry the contents of the pyknometers to a constant weight, without 

 decomposing the salt. The method finally adopted was to evaporate to partial 

 dryness in the Jena glass bxilbs at 115 0., then raise the temperature of the oven 

 gradually to 175 C. (dry air passing the while), and, when all perceptible moisture 

 had been driven oft', to heat the bulb gently with a naked flame, care being taken not 

 to heat to a temperature high enough for the glass to give a sodium flame. With 

 caesium alum a constant weight was obtained by keeping the oven at 130 C. to 

 140 C. 



The Observed Densities and Solubilities. 

 In the following tables the numbers in each column are obtained as follows : 



Column I. gives the temperatures to which the observation recorded in the 

 remaining columns refer. These temperatures are corrected for emergent 

 column, and are, as before stated, the mean temperature corresponding to 

 the meaned density and the meaned solubility. 



Column II. gives the density of the saturated solution, obtained as already 

 stated. It summarizes over 600 observations, excluding the 300 men- 

 tioned on p. 195. 



