2 1 6 Love and Smeal : 



No« t t' Dew-point R Wind 



Applying the formula, 



X =/- AB(/- /'), 



the vapour-pressures being taken from Broch's tables, the 

 values of A are calculated. The mean is 0.0007228. with a 

 probable error of 0.0000175, or, approximately, 

 A = (72 ± 2) X 10-'> 



The individual values of A vary rather widely, but ail have 

 been included, since the comparison of the wet-bulbs is not 

 affected by other errors, such as a lack of precision in the 

 indication of the llegnault hygrometer. Some of the largest 

 discrepancies have been traced with great probability to an 

 error of reading amounting to half a degree. The tendency 

 to such a mistake was always present, owing to the method of 

 marking of the thermometers, and its occurrence was actually 

 detected in later observations, when it was watched for. It 

 was not thought advisable, however, to make anj- tentative cor- 

 rections of this nature. The outstanding errors being few, 

 and both positive and Uv^gative, will not affect the mean value 

 by much. It is somewhat high, compared with other observers' 

 results, but this may be partly due to the fact that the values 

 of / — /' are all small, none being above 5"^, although the 

 dry temperatures range from 7.35° to 19.35°. As is 

 seen in the next table, the values of the humidity are corre- 

 spondingly fairly high throughout, being never less than 50 

 per cent. 



With this mean value of A the corresponding values of x 

 were calculated, and are included in Table III. under .r. The 

 humidities calculated from .v and .v are also tabulated as r and 

 r, with tlie difference, or errors, A.v =: x — .v and /sr =^ r ~ r. 



