to the temperatures indicated by a wet and dry thermo- 

 meter. 



The observations for the April term day, were those 

 to which it was his intention to advert, as they exhibited 

 higher values for t — t', than those made in any other month. 

 Now upon looking through these which amount to 24, the 

 first fact which at once presents itself is, that in every in- 

 stance but two, the observed dew-points are higher than ob- 

 tained by the formula, and in some instances, by as many as 

 nine degrees Fahrenheit. One or other series, therefore, 

 must be erroneous. That the observed dew-points are inac- 

 curate. Dr. Apjohn inferred, on the ground of their being in- 

 consistent with each other ; for he held it as quite certain, whe- 

 ther the hygrometric expression be correct or not, that when 

 in the case of any two distinct observations, t and t' have the 

 same values, that upon both such occasions the air includes 

 the same amount of moisture, or has the same dew-point. 

 Tried by such a criterion, the results obtained with Daniell's 

 instrument are defective, as is well illustrated by the follow- 

 ing extract from the April observations. 



t t' t" oh. f'calc. 



10^70 51 43.5 34.7 > 



1M71.2 51 34 33.3 > 



7 J 63 47 31 31.6 i 



17^62 47 36.6 32.8 5 



2<55 42.5 30.5 28.9 



24^55 42.2 33 28 



9.5 1.4 



6.6 1.2 

 2.5 0.9 



From what has been just said, it is obvious, that observa- 

 tions 10 and 11 should give nearly the same dew-point. This 

 is true of the dew points got by the formula, but not at all 

 of the direct determinations by the hygrometer, as those 



vapour at t', the temperature by the wet thermometer, and t the temperature of 

 the air. 



