PHYSICAL CONSTANTS OF SATURATED SOLUTIONS. 199 



perforated to allow the steam to escape, is inserted into the top of the inner tube. 

 When the temperature becomes constant, the pyknometer, with the tap closed and 

 with the end of the lower capillary covered by a filter, is forced through the stopper, 

 so that the bulb and capillary are completely immersed ; when the temperature is 

 again constant, the tap is opened and the pyknometer quickly filled by gentle 

 suction, and the tap closed. The filter is then removed from the end of the capillary, 

 the level of the solution in the stem is read, and both thermometer and pyknometer 

 are taken out of the solution by the withdrawal of the rubber stopper ; the pykno- 

 meter is then washed, dried, and weighed in the usual manner. 



It was found advisable to use the pyknometers described on p. 191 and shown in 

 fig. 3, not only on account of the pressure set up when the salt crystallized out, but 

 because they could be more quickly filled, and therefore less condensed steam formed 

 in the stem ; a further reason for using these pyknometers was that the larger bore 

 of the stem and lower capillary enabled them to be emptied with less difficulty. 



Determination of the Solubility. 



When the pyknometer had been weighed, its contents were emptied into a beaker, 

 and the solution washed into a Jena glass bulb (described on p. 190 and shown in 

 fig. 1), and evaporated to clryness as before. 



Great difficulty was experienced in emptying the pyknometers when filled with 

 the solutions of rubidium nitrate, thallium nitrate, or caesium alum, and the only way 

 of doing so was by alternate heating and cooling when completely immersed in nearly 

 boiling water the operation taking in some cases as long as G hours. It is 

 interesting to note that on testing the boiled saturated solutions of the nitrates of 

 sodium and rubidium for nitrites by means of fuclisine, the conversion of a small 

 quantity of the nitrate into the nitrite was distinctly indicated. 



Modification of Apparatus Necessary to Meet the Case of Extremely Soluble Salts. 



In the case of the nitrates of rubidium and thallium, which are extremely soluble at 

 the boiling-point, the apparatus described on the foregoing page was found to be 

 unsuitable, because a constant temperature could not be maintained for a sufficient 

 length of time to allow the pyknometer to be filled. Two things are essential for 

 maintaining the solutions at their boiling-points : that thorough stirring should take 

 place, and that there should be a sufficiency of undissolved salt left in contact with 

 the solution ; with extremely soluble salts the larger quantity of steam necessary for 

 thorough stirring dissolves so much salt that by the time this stirring is attained the 

 solution is nearly clear, and shortly after, all the salt is dissolved and the temperature 

 begins to fall. 



