BETWEEN THE FREEZING AND BOILING-POINTS. 
249 
J, 4*18876 joules, is exactly coincident with the values at 15°*7 and at 68° C. In 
selecting a thermal unit to which the values of the specific heat may be referred, 
it seems desirable to adopt one at a temperature which, if at the same time at a 
convenient part of the scale, may also represent the mean value over the whole 
range. Such a convenient point appears to he indicated at either 15° or 16° C. 
T propose, at the present time, to adopt the value at 16° C., and shall in conse¬ 
quence express the specific heat of water in terms of this unit, which is equal to 
4*1883 joules, and which differs from the mean value by only 1 part in 10,000. 
The following table (p. 252) includes the values of the specific heat of water in 
terms of a unit at 16°. No one simple formula can be fitted to the complete curve 
between 0° and 100° with any degree of accuracy, on account of the change which 
occurs at 37°*5, which is the point of minimum specific heat. Two formulae can be 
Fig. 17. 
fitted together, however, over the range with great accuracy. Between 5° and 37°*5 
the expression representing the specific heat in terms of a unit at 16° C. is 
S* = *99733 + *0000035 (37*5 - tf + *00000010 (37*5 - tf. 
The same expression reads above the minimum point, as far as 55°, in this form : 
S* = *99733 + *0000035 (t - 37*5) 2 + *00000010 (t - 37*5) 3 . 
At 55° and upwards the values diverge more and more from this formula, and follow 
VOL. CXCIX.—A. 2 K 
