BETWEEN THE FREEZING AND BOILING-POINTS. 
257 
Sec. 9. —Relation of the Present Measurements to the Work of other Observers. 
It will hardly be necessary for me to enter into a lengthy discussion of the work of 
other observers, more especially as it has been already carefully done in the original 
memoirs of Rowland,* * * § Griffiths,! and Schuster and Gannon. J Since the publi¬ 
cation of these papers, however, a very elaborate and exhaustive series of experiments 
has been made by Reynolds and Moorby§ to determine, by a direct mechanical 
method, using a Reynolds brake and a steam-engine, the energy required to raise 
water from a temperature slightly above. freezing to the boiling-point. The value of 
the mean mechanical equivalent which they obtained is entitled to a great deal of 
weight, from the minute accuracy of their measurements and the careful discussion of 
possible sources of error. 
It is fortunately possible, by means of the present series of experiments, on account 
of their great range, to connect the experiments of Reynolds and Moorby with the 
experiments of Rowland, also by the direct mechanical method, which extends 
between 6° and 36° C. The absolute value of the mean mechanical equivalent 
obtained by Reynolds and Moorby is 4 T8320 joules, which is obviously less than 
the same mean value obtained in the present experiments (i.e., 4*18876 joules) by as 
much as 0T32 per cent. 
This discrepancy in the two results may be caused by an error in the present 
measurements at the extremities of the range, due to the neglecting of some 
correction factor which would cause the variation curve to increase less rapidly than 
it does; blit it is far more probably due to an error in the value of one of the 
constants for the determination of the electrical or heat energy. Of this latter 
possibility the value of the Clark cell is still in doubt, although the value of the ohm 
is fairly well fixed in absolute measure, as defined in the ‘ British Association Report’ 
of 1892. All of the thermal measurements are expressed in our two results to the 
same scale, so that the error resolves itself into an error in the E.M.F. of the Clark 
cell, which, as it enters into the equation for the determination of the electrical 
energy to the second power, has twice the effect. This has been already pointed out 
under the Section devoted to the Clark cell, where it was shown that if all the error 
between the value of the mean mechanical equivalent obtained by the direct 
mechanical method and the value obtained by the electrical method (assuming the 
Clark cell equal to 1*43420 volt and the international ohm equal to 1*01358 B.A. 
units) could be attributed to the Clark cell, the value 1*43420 would have to be 
* ‘ Proc. Amer. Acad.,’ vol. 15, p. 75 (1879). 
f ‘ Phil. Trans.,’ A, vol. 184, p. 361 (1893). 
+ ‘Phil. Trans.,’ A, vol. 186, p. 415 (1895). 
§ ‘Phil. Trans.,’ A, vol. 190, p. 300 (1898). 
2 L 
VOL. CXCIX.—A. 
