438 



for another, from which we deduce 



(f'-/')30 



(t-t'Kp-fO-C*-*')^-/') 

 " The subjoined tables contain the experiments ; and ap- 

 plying to them the method of calculation just explained, we 

 obtain eleven values of m. 



Series 1. 



Table 1. 



No. 



t 



t' 



t~t' 



P 



1 



49-6 



44-7 



4-9 



296 



2 



88-5 



62 



26-5 



296 



3 



80-5 



59 



21-5 



29-6 



Table 2. 



No. 



1 



2 



t 



if 



t-t' 



P 



47-2 

 76 



42-5 

 57-5 



47 

 18-5 



3002 

 30-02 



Table 3. 



Table 4. 



No. 



t 



t' \t-f 



P 



1 

 2 



51-3 



82 



45-5 1 5-8 

 59 23 



30-7 

 30-7 



) 1 &2 . m= -011 

 VI &3 



0121 



No. 



* 



*' 



*-*' 



P 



■ 



1 



48-3 



43 



5-3 



29-76 



2 



96 



64 



32 



29-76 





3 



91 



62-5 



28-5 



29-76 





4 



75 



56 



18 



29-76 



- 



fi oc o 'Uizi 



)2&3 -0110 



1 & 2 . m = -0144 



1 &2. ?«= -0117 

 1 & 3 -0122 



1 &4 -0132 



2 & 3 -0035 



2 &4 -0103 



3&4 -0109 



1 & 2 . m— -0106 



Mean value of m — -01 151 



"2. In the general formula given above, if f" =z 0, 



f 30 



= : v X j. . My observations of t and t' in dry 



t - t p — /' f 



