MESSRS. W. R. BOUSFIELD AND W. ERIC BOUSFIELD 



heat But within the range of C. to 80 C. an equation of the third degree 

 represents the experimental values of J within the limits of accuracy of 1 part m 

 4000 which is quite within the limits of experimental accuracy. The 

 expression accords with the experimental values within these limits, the differences 

 between observed and calculated values being shown in the table :- 



J/ = 4-2085-0-0015110+0-000026110 2 -0-00000012250 3 . 



TABLE X. 



e. 



13 



27 

 40 

 55 

 73 

 80 



We give the values to four places of decimals, but no value is to be attached to 

 the fourth place except so far as it may serve to determine more accurately the 

 third place. 



To obtain the values of J we have 



which gives 



J, = 4-2085-0'0030220 + 0-000078330 2 -0-000000490 3 . 



From this expression we obtain the following values for J : 



TABLE XL Values of J. 



6. J. 0. J. a J. 



4-2085 30 41751 60 4'2033 



5 4-1953 35 4'1777 65 4'2084 



10 4-1856 40 4-1816 70 4'2127 



15 4-1791 45 4-1865 75 4'2157 



20 41755 50 41920 80 4'2172 



25 41743 55 41977 



The temperature curve for the value of J which results from these figures (together 

 with the curves of other observers, to which reference is made in the next section) is 

 shown in figs. 10 and 11. Our relative value at 80 C. compared with that of 70 C. 

 has some uncertainty, but we think that some drop in the curve as this temperature 

 is approached is a genuine phenomenon, though probably not quite so marked a drop 



