of the Moist-bulb Hygrometer. 289 
Of the nineteen observations of depression in dry air registered in the preceding 
table, eleven are greater, and eight less than the calculated results. ‘The mean of the 
plus errors of the formula is, .28, and of the minus errors, .4 of a degree; so that 
.28—.40= —.12 of a degree is the mean difference deducible from the whole be- 
tween experiment and calculation. A closer approximation between them than this 
could not, I think, be anticipated, even upon the hypothesis of the strict accuracy of 
the formula. I may also observe that if by means of the equation f’ =m d x x which, 
as we have already seen, belongs to perfectly dry air, we deduce from the preceding 
tables 19 values of m, the mean of all will be found almost accurately equal to 
= a result the more entitled to confidence inasmuch as the mean pressure for the 
19 experiments being but very little over 30, and the air being perfectly dry, neither 
of the corrections which I investigated in my former paper require to be applied. 
If from the experiments already detailed I were to draw the conclusion that the 
1 ‘ : : 
equation f= f — a x 5 will afford the dew-point with a degree of accuracy far 
surpassing ordinary hygrometrical observations, I would, probably, have the concur- 
rence of most of my readers. The evidence adduced in support of the formula 
appears, at least to me, ample and satisfactory. For the purpose, however, of dis- 
pelling any doubts of its accuracy which may exist in the minds of others, I under- 
took another series of test experiments, to the description of which I shall now 
proceed. 
The most direct method of testing our formula consists, as has been already ob- 
served, in comparing its results with dew-points experimentally determined. In 
order, however, that this criterion be decisive, it is not only necessary that the de- 
pressions be considerable in amount, but also, as is obvious, that the dew-points be 
accurately known. Now the registers to which I have had access do not perfectly 
satisfy either of these conditions, the depressions being generally small, and the 
observations made with an instrument—Daniell’s hygrometer, the difficulty of 
observing with which is universally admitted. In reflecting on this matter it occurred 
to me that both difficulties might be evaded in the following simple manner. Let 
air, saturated with moisture, and whose temperature is, therefore, necessarily its dew- 
point, be heated, and let the temperature of the heated air be taken, as also that shewn 
by a moist bulb hygrometer, subjected to the action of a current of it. Let, then, 
by the application of the formula, the dew-point, belonging to the two latter observa- 
tions, be calculated, and from a comparison of it with the original temperature of the 
air when saturated with humidity, we will be enabled to pronounce with confidence 
upon the yalue of our method. 
