130 
PROFESSOR HUGH L. CALLENDAR ON 
which the results mainly depend. The corrections calculated from the actual 
numbers in the tables are, as a rule, slightly larger than those calculated from the 
formula for the heat-loss, but the latter are usually within 1 in 10,000 of those 
observed. 
It may be interesting to remark that formula (16) gives a mean rate of increase of 
the heat-loss which is proportional to the fourth power of the absolute temperature 
over the range 0° to 80°. The fourth power law is well known to represent heat-loss 
by radiation with considerable accuracy over a moderate range at ordinary 
temperatures under conditions similar to those of this experiment. Values of the 
1 eat-loss calculated according to the fourth power law starting from the same value 
at 30° C., are given, for comparison, in the third column, headed E0h They 
represent the experimental numbers rather better than the linear formula (16) at 
temperatures below 30°, but give results which are a little too high at 80° and 90°. 
The corrections in the table have been calculated from the linear formula, but the 
difference would be very slight over the greater part of the range. 
The values of the specific heat given in the column headed “Results, 1900, 
Corrected,” are obtained from those calculated by the formuke given by Barnes, 
‘ Proc. Roy. Soc.,’ 1900, p. 242, by adding the correction in calories given in the 
previous column, and expressing the results in terms of a unit at 20° C., instead of 
the unit at 16° employed by Barnes. Although the values of the corrections are 
worked to the next figure, I have not considered it desirable to give the values of 
the specific heat beyond 1 jiart in 10,000. 
It happens, by a curious coincidence, that the correction for the variation of the 
temperature gradient is very nearly equal and opposite to the correction required to 
reduce the results to the hydrogen scale, if calculated, as previously explained in 
Section 23, from the observations of Joule and Thomson for air at a constant 
pressure of 76 centims. The values given in the last column, reduced to the 
hydrogen scale, are practically identical with those calculated directly from Barnes’ 
formulae without correction, except that they are expressed in terms of a unit at 
20 ° C., and are corrected for an obvious misfit of the formulae at 55° C. (see §46). 
Part V.—Discussion of Results. 
(38.) Meaning of the Term “ Specific Heat.” 
The term “ specific heat ” is here employed as an abbreviation for the phrase 
“ specific capacity for heat,” or “ thermal capacity of unit mass, ’ i.e., the quantity of 
heat per unit mass per degree required to raise the temperature of a substance. 
In a similar manner, “specific electrical resistance” or “ resistivity” of a substance is 
understood to mean the resistance of the material per unit area of section per unit 
length. On this understanding, specific heat may be measured in terms of any con- 
