443 



We are tlius compelled in strictness to omit that result alto- 

 gether in deducing the final mean. Combining, tlierefore, the 

 results of the first and third series, according to the method 

 already laid down, we have 



m = .01140; 



a result which is identical with that of the third series, that 

 of the first, (on account of its large probable error) not 

 affecting the fifth place of decimals.* 



But the second series of experiments, though it cannot 

 properly be combined with the others in deducing the mean, 

 may yet serve another purpose. It may be made, in fact, 

 a test of the accuracy of the different tables of the elastic 

 force of vapour, within the range of temperature belonging 

 to the experiments. With this view, the values of the co- 

 efficient, W2, have been calculated by Dr. Apjohn from his 

 second series of experiments, by means of three separate 

 tables of the elastic force of vapour. The first of these 

 tables is that which has been employed above, as well as in 

 his papers on the Wet Bulb Hygrometer in the Transactions 

 of the Academy, and is that calculated by Dr. Anderson 

 from the experiments of Dalton and Ure. The second 

 table is that deduced by Mr. Kamtz, from his own experi- 

 ments ; and the third is that given in the Report of the 

 Committee of Physics and Meteorology of the Royal Society, 

 and calculated by Mr. Lubbock from a formula of his own. 

 The results are given in the annexed table. 



* The errors of observation in the first series, which are so considerable in 

 comparison with those of the other two, are manifestly owing to the mode of ob- 

 serving/". For it is obvious that the rapidly varying temperature of the thermo- 

 meter in the condensation hygrometer cannot be noted at a precise instant, with 

 the same certainty as that of a thermometer which has arrived at a stationary 

 temperature. 



2v 



