1880.] Lowering of the Freezing-Point of Water by Pressure. 535 



metallic manometer which had been chequed on several occasions, 

 records of the first 25 atmospheres' pressure are not reliable. 



Increase. 



Lon on scale 



The 



Galvanometer zero +1. 



25 atmospheres = — 26 posit 



50 „ 



— 



— 43 



75 „ 



— 



— 62 



100 „ 



— 



— 82 



125 





—100 



150 





-120 



175 





-142 



200 





-163 



225 





-185 



250 





-203 



275 





-223 



300 





-243 



17 

 19 



20 

 18 

 20 

 22 

 21 

 22 

 18 

 20 

 20 



This represents a mean deflection of 19*7 on the scale, being 0*18 

 of a degree C. for 25 atmospheres, or a lowering of the freezing-point 

 by 2°1 C. for the total range of pressure. The mean reduction for 

 one atmosphere of pressure is o- 0072. 



These experiments were afterwards repeated, with a stronger pump, 

 to 700 atmospheres. Exactly the same mean deflection was observed, 

 and the galvanometer moved with the same regularity. 



Series II, 



The junction in the flange was placed in a piece of quill tubing, 

 open at top and bottom, in order to insulate it from the sides of the 

 iron bottle, and surrounded with water at 0° C. instead of ice, the 

 iron bottle being afterwards packed, as before, in ice. On applying 

 pressure a very slight decrease of temperature occurred, the whole 

 deflection for 200 atmospheres being only about 4 divisions, or -^w °^ a 

 degree C. This agrees with Joule's* experiments on the compression 

 of water about the same temperature. 



Series III. 



A mixture of finely powdered ice and water was placed round the 

 junction inside the bottle, which was, as usual, packed afterwards in 

 ice. It will be seen, from the following set of readings, that the 

 deflection was about the same as when the junction was frozen into a 

 block of ice, the only difference being that, whereas the galvanometer 

 in the first series assumed its new position immediately after the pressure 

 was applied, the instrument now required from two to three minutes 

 before coming finally to rest, the action being far more sluggish. 



* " On the Thermal Effects of Compressing Fluids." " Prc-c. Key. Soc.," vol. ix, 

 p. 496. 



