PHYSICAL CONSTANTS OF SATUKATED SOLUTIONS. 



191 



m.m 



120 



no 



100 

 90 

 80 



f- 



heating being continued until the bulbs had attained a constant weight. The air 

 current was obtained by means of a Fleuss pump. The same filling of a pyknometer 

 gave, therefore, both the density and the solubility. As a check on the latter, the 

 contents of a pyknometer were occasionally analysed. 



Pyknometers. 



At first Sprengel pyknometers of various shapes and sizes were tried, but were 

 found .to be unsatisfactory. This was because it was almost impossible to prevent the 

 solution from crystallizing in the capillary 

 during the time the level of the liquid was 

 being adjusted to the mark. 



The following was the form finally 

 adopted and found quite satisfactory. A 

 pear-shaped bulb, of about 5 cub. centims. 

 capacity, terminating above in a stem 

 composed of a graduated capillary 120 

 millims. long, and below in a finer capillary, 

 bent as in fig. 2, was used for salts of 

 medium solubility. For somewhat insoluble 

 salts a similar pyknometer, but of about 

 11 cub. centims. capacity, was found to be 

 more suitable, while for very soluble salts, 

 such as sodium sulphate, which have great 

 differences in solubility at different tem- 

 peratures, it. was necessary to have similar 

 pyknometers made of thicker glass, so that 

 when the crystals formed and practically 

 filled the whole of the bulb, the latter 

 would withstand the pressure. It was 

 also found necessary to make the capillaries 

 of a larger internal diameter, so as to be 

 able to fill quickly. And the shape of the lower capillary (see fig. 3) was altered and 

 its end fitted with a glass cap to prevent the solution from " creeping" out when 011 

 the balance. The stem was also fitted with a cap to prevent evaporation. 



Determination of the Capacities of the Pyknometers. 



Before determining the capacities, the pyknometers were heated rapidly and 

 repeatedly to 200 C., being allowed to cool to the temperature of the room between 

 each heating ; by this means it was hoped that the gradual shrinking in volume 

 would be accelerated. The graduated capillaries were then calibrated by the usual 



.'.0 



10 



Fig. 2. 



Fie. 3. 



