Mr John Adie on Dew-point Instrumenis. 63 



struction. The purer it is the sooner wili a dew-point be obtain- 

 ed, and the farther from the truth. The first I used was filled 

 with the common ether of commerce. This gave a dew-point 

 lower than when filled with that of a purer quality. Some o£ 

 the first observations in the table were made with this instru- 

 ment, but in some cases a dew-point could not be obtained ; 

 and in all cases in which it was obtained, a much greater 

 quantity of ether was expended on the cooling bulb. I had, 

 therefore, an instrument constructed with pure ether, which 

 gave a result with great ease, but it always required a great- 

 er quantity of ether to cool it than the other instruments. 

 The error in this instrument may easily be shown by the fol- 

 lowing experiment : — When the ring of dew is formed round 

 the bulb, and its temperature observed, let the enclosed ether 

 be agitated, keeping up at the same time the process of cooling, 

 and the whole bulb will be dewed over, the enclosed thermo- 

 meter being then observed, it will have sunk several degrees, 

 and will be at the correct dew-point. 



The result obtained by the thermometer having its bulb 

 covered excepting a very small spot is most remarkable. The 

 error in its indications seems to arise from two causes : First, a 

 stratum of mercury in contact with the glass is constantly ex- 

 posed to a depression of temperature which must be commu- 

 nicated to the mass. Secondly, glass being a bad conductor of 

 heat, the space where the deposition is observed requires a cer- 

 tain portion of time to acquire the same temperature as the 

 mercury; so that to render such an instrument perfect, it would 

 be required that both mercury and glass should be instanta- 

 neous conductors of heat ; the principle of correct thermometers 

 being, that all their parts shall have the same temperature ; and 

 where any substance is used as a measure of temperature, it 

 must expand in all its parts simultaneously, or an allowance 

 be made for a known quantity having a different temperature. 

 These effects may be illustrated by using more or less ether in 

 the process of cooling ; for the more quickly the temperature is 

 depressed, the farther will the dew-point given be from the truth, 

 in some cases 5 or 6 degrees ; whereas when it is cooled very 

 slowly, a result may be obtained within 2 or 3 degrees of the 

 correct point. The mean of 28 observations given in the follow- 



ii 



