PRINCIPLES OF THE MECHANICAL THEORY OF HEAT. 275 



Fig. 18. 



point niiisf, through increased pressure, he heightened. The correct- 

 Dt'ss of this consequence lias been cx])eninentally proved by Bniisen, 

 (Pog-o-. Aiiiial., Lxxxr, 18'jO,) and by Hopkins, (Dinj^ler's I'olyt. 

 join-nal, cxxxiv.) Biinsen conducted the experiment in the I'ol- 

 loTVuig- manner : A very thick-walled glass tube about a loot long, 

 and having" a bore of tlie size of a straw, was drawn out at one end 

 into a fine capillary tube from 15 to 20 inches in length, which was 

 made of as accurate calibre as possible. The lower end of this 

 glass tube was also drawn out to a somewhat wider tube, one and 

 a half inch long and curved, as is shown in I'ig, 18. 



By atmospheric pressure the whole apparatus was now filled with 

 quicksilver, and the longer capillary tube was then soldered at a. 

 After cooling, a small quantity of quicksilver was driven out at h 

 l)y gentle heat, and in place of it, while the cooling was renewed, 

 a small quantity of the substance to be tested was imbibed in a state 

 of fusion. The apparatus was now soldered also at 6, the longer 

 capillary tube was then opened at a, and the whole apparatus heated 

 one or two degrees above the melting point of the substance con- 

 tained therein, through which i)rocess a part of the quicksilver is 

 expelled from the open point at a. Finally, after renewed cooling, 

 the range of the quicksilver in the capillary tube at c, together with ||j \\^ 

 the range of the thermometer and barometer, was noted; the point \|L^Z 

 at a was then once more soldered, and thus a column of air of ascer- 

 tained length was included. 



Two such instruments of precisely similar form and contents, one of which was 

 soldered at a, Avhile the other remained open, were now, together with a sensi- 

 tive thermometer, fastened on a small board in such manner that the two little 

 tubes lilled with the substance to be tested might stand close to the bulb of the 

 thermometer. If tTiis ai)}iaratus be immersed in water whose tempeiature is a 

 few degrees above the melting point of the substance, to such a depth that only 

 the tube b shall be submerged, it will be seen that, by gradual cooling of the 

 water, molecular rigidity will ensue simultaneously in both tubes. But were the 

 api)aratus sunk deeper in the warm water, there would follow, through the ex])an- 

 sion of the quicksilver in the closed instrument, a pressure which can bo readily 

 measured by the compression of the air in the cajjillary tube cb, and which may 

 1)0 augmented or diminished at will by depressing the instrument in the warm 

 fluid or partially withdrawing it therefrom. 'J'ho pressure in the open instrii- 

 meiit, on the other hand, remains unchanged during the whole experiment. The 

 difference of temperatm'e at which the Substance grows rigid in the closed instru- 

 ment sooner than in the open one gives the elevation of the melting point for 

 the observed pressure. 



An experiment made with spermaceti gave the following result : 



Prt'ssiire. Point of rig:i(Jity. 



1 atmosphere 47.7° C. 



9G atmospheres AO.T 



156 atmospheres 50.9^ 



The same experiuient repeated with paralline gave : 

 Pressure. Point of ripjidity. 



1 atmosphere 4()..'i° C 



85 atmospheres 48.9° 



100 atmospheres • 49.9° 



