400 
PROFESSOR 0. REYXOLDS ARD MR. W. H. MOORBY 
per trial, by conduction along the shaft, per unit difference of temperature between 
the stuffing-box and low'er brass, 
Then C is given by the equation 
G75844869 27114.3956 
867995 + 75-6 C ~ 348866 - 22-5 C ~ 
where the numerators represent the sums of the differences of work in the sets 
enumerated above, while the first terms of tlie denominators represent the sums of 
the differences of heat in the same sets, to which the terminal corrections have been 
added. The second term in each denominator represents the correction to be applied 
to the differences of beat for unbalanced conduction along the shaft. 
On solving the equation we get 
C = 12, very nearly. 
This agrees very closely with the value C = 13-61, which may be calculated from 
the dimensions of the conducting shaft, viz., 4 inches diameter and 2f inches long, 
and Forbes’ value of the conduction coefficient for iron, viz. : 
(0-1429 in C.G.S. unit). 
Since nothing was known as to the internal thermal condition of the shaft, the 
figure 12 has been used throughout as a sufficiently close approximation to the 
constant required. 
The corrections to the heat for conduction along the shaft in each trial were then 
obtained by multiplying the fall of temperature between the brake and bearing by 12. 
The sign of the correction varies, of course, wuth the sign of the temperature 
gradient along the shaft. 
Determination of the Loss of Heat hy Radiation. 
44. Under this heading are included all losses of heat not already dealt wdth 
under the headings “ terminal corrections,” “ loss by conduction,” and 'Goss by leakao-e 
of water.” ° 
Radiation in the Unjacheted Trials—Series I. 
45. Determination No. II., consisting of a combination of trials 3 and 4, is omitted, 
for the reasons given. A constant R, representing the loss of heat b}" radiation per 
trial per unit difference of temperature between the brake and surrounding air is 
required. 
In Tables B and C, the corrections to the heat are given for terminal errors and 
conduction along the shaft, the calculation of which has been explained. 
