428 REPORTS ON TSfi STATE OF SCIENCE. 



tubes were closed by corks fitted with a sbort tube arid a long syphofl 

 tube the end of which was covered with a piece of cambric to serve as 

 a filter. After two hours, samples were ejected by blowing through 

 the short tubes. 



Tilden and Shenstone 163 determined the solubility of salts at high 

 temperatures by enclosing salt and solvent in specially constructed 

 metal tubes, which were heated to the requisite temperature for four 

 to five hours whilst resting in a slightly inclined position. The end 

 containing the solid was then gradually raised, causing the liquid to 

 drain from the solid and collect at the other end of the tube. 



In carrying out work at low temperatures Etard 107 employed methyl 

 chloride for the bath. He always determined the solubility at the 

 highest temperature the thermometer indicated, recognising that any 

 lowering of the temperature would cause supersaturation. 

 1885 Guthrie llk determined the solubility of salts in fused sodium 



nitrate by adding small quantities of the salt, the whole being 

 well stirred until no more dissolved. The mass was then allowed to 

 cool slowly to allow excess of added salt to separate out; then the 

 liquid was poured on to a cold slab and analysed. 



IQQQ At temperatures about 100° Alexejeff 118 sealed crystals in a 



tube with water and heated until almost all the crystals were 

 dissolved. The tube was then cooled until the remaining crystals 

 began to grow in size, the temperature at which this occurred being 

 taken as the saturation temperature. 



For temperatures between 250°-450° filard 143 used a bath of 



a fused mixture of sodium and potassium nitrates. Salt and 



water were sealed up in thick-walled glass tubes (7 mm. by 15 cm.), 



and the temperature was observed at which the last traces of salt 



disappeared. 



i Qon By means °f an electrically arranged gas cut-off Meyer- 



liofjcr 1S7 maintained the temperature of his thermostat to 0°'l ; 

 in this work the thermostat liquid was not stirred, but the solvent and 

 solute were stirred together in a vessel supported in the middle of the 

 bath. The saturated solution was taken up by means of a warmed 

 Landolt pipette, and was analysed by titration. 



Piieclcer and van Dc venter 10 ° used Victor Meyer's method (vide this 

 section, 1875), modified so that the entire operation, including taking the 

 sample of saturated solution, was carried out under the surface of the 

 thermostat liquid. 



The saturated solutions employed by Noyes 164 were obtained by 

 shaking the solvent with excess of solid for several days in a thermostat. 

 Kolilrausch and Rose 20i calculated the solubility of very 

 sparingly soluble salts from measurements of the electrical con- 

 ductivity of the saturated solutions. 



When determining solubility at high temperatures ~Etard 21i 



used a glass tube bent to an angle of 120° and constricted at the 



middle. Salt and water were placed in one arm and the tube sealed. 



After heating at a fixed temperature the saturated solution was allowed 



to run into the other arm of the tube, and when cooled was analysed. 



Goodwin 2l2 calculated the concentration of one ion in saturated 



