766 DETERMINATION OF THE BOILING POINT. 



neck may be used, such as that represented in fig. 382. 

 The flask is about half-filled with water, and the 

 thermometer is suspended in the flask, either by being 

 tied to a thin bar of wood which is placed across the 

 mouth of the flask, or it may be clamped in the retort- 

 stand, if the stem is sufliciently long. The bulb of the 

 thermometer should in any case not dip into the water it- 

 self, for the temperature 

 of boiling water is not 

 strictly constant, but 

 varies slightly under 

 different circumstances ; 

 whereas the tempera- 

 ture of the steam pro- 

 duced by boiling water 

 is invariable, if the 

 pressure of the atmo- 

 sphere is constant. Care 

 must, however, be 

 taken that the evolu- 

 tion of steam is pretty 

 brisk, and that the gene- 

 rated stream is not 

 farther heated by con- 



FIG. 382 (an. proj. ; real size). 



tact with the hot sides 



of the glass vessel. This may be prevented by support- 

 ing the flask upon a thin plate of metal with a circular 

 aperture ; the heated air is by this means forced to flow 

 off in all directions, and is not allowed to ascend close 

 to the sides of the vessel. To prevent the metal itself 

 from becoming too hot it is covered almost to the edge 

 with a layer of sand, 1 or 2 cm high, all round the flask. 



