or Dew-point Instrument, 85 



power of metal than that of glass will lead to the more speedy 

 diffusion over the whole bottle of the cold impression. Mr. John 

 Adie, on comparing the results by DanielFs instrument with 

 those obtained by Dalton's method of transference of cooled 

 water from one vessel to another^ which by most inquirers is 

 admitted to be the most accurate of any, although far from being 

 the least troublesome in point of execution, found the error occa- 

 sionally to amount to 6^° above the result given by Dalton's 

 method, and on an average of twenty-eight observations to 

 reach 2°-9. With a view to this question, I made a set of 

 comparative experiments with DanielFs hygrometer, and that 

 described in this paper, referring to Dalton's mode, as a trust- 

 worthy standard of comparison. The results will be found in a 

 table inserted in my paper in the Transactions of the Royal 

 Society of Edinburgh, and I will here merely mention that my 

 conclusion from the whole was, that the tendency of DanielPs 

 instrument was decidedly in excess, although not to so great aii 

 extent as found by Mr. Adie ; whilst that of the instrument de- 

 scribed in this paper was in the opposite direction ; but that with 

 neither did the error, on an average of sixteen or eighteen obser- 

 vations, reach 1° F. Having found, however, that twelve out of 

 eighteen observations with DanielPs gave a result in excess, the 

 extreme being 2^°, I cannot admit the assumption of DanielPs 

 instrument as a rigidly accurate standard of comparison, and 

 would wish Dalton^s method to be taken as the criterion ; with 

 reference to which I found the deficiency of the new instrument 

 on an average of sixteen observations between 32° and 45°, to 

 be only 0*95° F., which is evidently quite consistent with the 

 employment of the instrument as a dew-point hygrometer. 



The mode of observation which I adopt is that usually followed 

 in regard to dew-point instruments, viz. to observe as accurately 

 as possible the first decided appearance of moisture on the sur- 

 face of the brass ball, and note the temperature indicated by the 

 immersed thermometer at the same instant ; then to stop the 

 process of exhaustion immediately, and again mark the tempera- 

 ture at the moment of disappearance of the moisture ; and lastly, 

 to take a mean of the two observations as the dew-point. 



In constructing the instrument, it is essential that minute 

 attention be paid to the various particulars which have been 

 mentioned regarding the size of the different apertures and pass- 

 ages, the air-tightness of the several screws and of the thermo- 

 meter stopper, the proper condition of the valves, the proper 

 position of the bulb of the thermometer, &c. And it should 

 be remembered that the apparatus ought not only to be well 

 constructed at first, but should be preserved in good condition 

 as regards the several points just mentioned. The oil of the 



