relating to Hygrometers and Evaporation. 29 



the pressure. This, no doubt, comes much nearer to my results 

 than to those of Mr Anderson, though it is not -very easy to 

 make the comparison, on account of the different temperatures 

 of the moist ball, at the various pressures in the foregoing 

 tables. 



I must not omit to mention, that, in these tables, the tempe- 

 ratures themselves still require a small correction ; because the 

 thermometers were, as is usual, sealed or close at top, and would, 

 therefore, stand a little too low when under the reduced pres- 

 sure. For, on placing them in a dry state, under a receiver, 

 and exhausting to the utmost, both stood 1°.5 Fahr. lower when 

 the former temperature was restored. Hence, as the entire ba- 

 rometric pressure is to the reduction of pressure, so is 1°.5 to the 

 correction sought. Other thermometers put in with them did 

 not all undergo the same change. For this, there are no doubt 

 various reasons. It is easily shewn, that, within a moderate 

 range, the error will, caeteris paribus, be nearly as the change 

 of pressure, multiplied by the diameter of the bulb, divided by 

 the thickness of the glass. But the sinking of the dry thermo- 

 meter a little, in these experiments, was partly the influence of 

 the cold wet ball on the still confined air. 



It has been long known, that thermometers were affected by 

 pressure ; and to avoid this, a very effectual method, when ap- 

 plicable, was adopted by Professor Leslie, who employed ther- 

 mometers open at top, when he had occasion to use them under 

 a variable pressure. Some, however, give themselves no con- 

 cern about the matter. In experiments on the force of steam, 

 the ball of the thermometer is often included in the boiler with 

 the stem projecting outward. The pressure on the ball may 

 then vary from a small fraction of an inch of mercury to many 

 atmospheres ; and, in such cases, the temperatures must be er- 

 roneous enough *. 



Mr Crichton has already pointed out some serious oversights 

 of a different kind in the Memoir of MM. Dulong and Petit on 



• Large flasks and receivers, if thin, must have their capacities somewhat 

 altered, by varying the pressure. This alteration in similar shaped vessels, 

 will be, cceteris paribus, nearly as the fourth power of the diameter divided by 

 the thickness. 



