426 M. Berthelot on a New Thermometer for measuring 



duced to about 20 or 25 centims. of mercury. The air as it 

 leaves the apparatus passes through the mercury. On reesta- 

 blishing the atmospheric pressure the mercury is driven into the 

 capillary tube. 



In order to ascertain if a sufficient vacuum has been made, the 

 air-reservoir B is plunged successively into a bath of melting ice 

 and into water at the ordinary temperature. The diminution in 

 internal pressure is thus determined by the recession of the mer- 

 curial column contained in the capillary tube. The number of 

 millimetres on the scale through which the mercury column 

 passes ought to be nearly equal to that which expresses the tem- 

 perature of the water employed — that is, if the thermometer is 

 intended to indicate temperatures up to 500°. If the exhaustion 

 has not been sufficient, this number will be too large ; it will be 

 therefore necessary to continue the exhaustion. On the other 

 hand, if the proper degree of exhaustion has been passed — for 

 example, if the pressure has been reduced to 10 or 12 centims., 

 the bulb must be completely emptied of the mercury and the 

 operation recommenced. 



The bulb Q ought to be of considerable diameter, in order that 

 the level of the mercury contained within it may only be inap- 

 preciably affected by the variations in the height of the mercurial 

 column in the capillary tube. The employment of a consider- 

 able quantity of mercury has also the advantage of rendering the 

 mercury in the capillary tube more susceptible of motion. By 

 slightly tapping the instrument the meniscus readily assumes its 

 normal form, and the indications are rendered perfectly regular. 



2. Determination of the Fixed Points. 



In order to graduate the instrument^ I have adopted the follow- 

 ing fixed points : — 



The melting-point of ice . . 



The boiling-point of water . . 100 



„ „ mercury . 350 



„ „ sulphur . 440 



These four points ought to be determined upon the same day, 

 and within an interval of time so short that the barometric pres- 

 sure does not undergo any appreciable variation ; by this I un- 

 derstand a variation of a millimetre. The zero-point is deter- 

 mined by plunging the air- reservoir B, together with a few cen- 

 timetres of the capillary tube 1 1, into melting ice, with the usual 

 precautions. The stem //// must be carefully fixed in a ver- 

 tical position. After a few minutes, and when the mercury 

 column remains stationary in the portion of the tube marked 

 1 1 11 in the figure, the wooden scale is slightly tapped below the 



