THE DENSITY OF LIQUIDS 431 



otherwise slight errors may occur owing to a change in the 

 weight or in the volume of the bottle, due either to a reduction 

 in the size of the stopper by the repeated grinding, or to 

 a slow change in volume similar to that observed in ther- 

 mometers, or to a slight change in the temperature of the 

 thermostat. Silica pyknometers are sometimes used, and 

 they possess some advantage in certain kinds of work. A 

 specific gravity bottle with a double-bored stopper which 

 allows ready adjustment of level of liquid and minimum ex- 

 posure to air, is sometimes used (Tate's modification). This 

 form of apparatus has advantages when hygroscopic and 

 volatile liquids are used. 



It maybe of interest to refer to the method of determining 

 the density of a liquid at its boiling point. The pyknometer 

 used for such determinations consists of a thin glass bulb pre- 

 ferably of Jena glass, capacity 2-3 ccs. The bent tube attached 

 is of capillary bore. The tube, after cleaning and weighing, 

 is filled as follows : the bulb is first heated and then cooled, 

 and placed in a vessel containing some of the heated liquid 

 of which the specific gravity is required : after a short time 

 the bulb is partly lifted out of the liquid, the open end of the 

 capillary tube being still under the liquid ; liquid is thus 

 forced into the bulb as the latter cools. The liquid in the bath 

 is now heated to a higher temperature and the operation of 

 heating and cooling the bulb repeated. This operation is 

 carried out several times with the bath at a higher tempera- 

 ture each time. The bulb is finally obtained full except for a 

 minute air-bubble. Sometimes filling is done by placing the 

 apparatus in a small vessel which can be partially exhausted, 

 and by dipping the open end of the capillary under the liquid 

 contained in the vessel. On allowing air to enter the latter, 

 some liquid will have entered the pyknometer. This opera- 

 tion is repeated several times until the bulb is practically full. 

 The pyknometer is suspended by a platinum wire in a 

 wide-mouthed boiling flask so that it is just above the surface 

 of the liquid (this liquid being the same as that in the pykno- 

 meter). The flask is next heated until the liquid it contains 

 is raised to its boiling point. With the increase in tempera- 

 ture, the liquid in the pyknometer expands, and all the small 

 air-bubbles are expelled ; the boiling is stopped when the 

 pyknometer and its contents have reached a constant tern- 



