PROFESSOR 0. REYFTOLTJS AND MR. W. H. AIOOREF 
through the main regulator this work has been converted into heat, and is measured 
as entering heat by the inlet thermometer, the assumption that the water through 
the branches enters at the pressure 2h, and the temperature given by the inlet 
thermometer, involves no other error than that resulting from radiation, which is 
constant for all trials, and is eliminated in the difference. 
The Regulation of the Tempemture of the Bushes. 
27. In the preliminary trials this temperature was only ascertained by touch, and 
regulated so as to be as nearly as possible that of the laboratory, the branch cocks 
being set with a definite opening, and the excess of pressure maintained as nearly as 
possible constant, a plan which was found to give consistent results. But it also 
appeared that in order to maintain the same temperature in the stuffing-box for the 
large and small trials with the same pressure in the main pipe, it was necessary to 
open the branch cock wider in the large trials. This was to be expected from the 
greater vortex pressure in the large trials. And as owing to the greater resistance 
of the cooler in the large trials there was difficulty in maintaining a great excess of 
pressure over the vortex pre.s£ure, it was decided to run both large and small trials 
with the same setting of the cock, and the same head in the cooling pipe, keeping a 
record until some means was obtained of estimating the comparative slopes of tem¬ 
perature in the shaft in the large and small trials. 
Thr Measurement of the Conparative Slopes of Temperature in the Shcft. 
28. the desirability of some more definite knowledge of the slope of temperature 
in the shaft between the brass of the nearest shaft bearing and the stuffing-box was 
strongly felt, but it was not at first apparent how this might be done, the shaft being 
4 inches in diameter and the gap between the end of the stuffing-box and the brass 
of the bearing being only 3 inches. 
However, as it became more evident, Avith the branch cocks set at a constant 
opennug and the same pressures in the supply pipe, that the temperatures in the 
stuffing-box were greater in the large than in the small trials, and that a small 
difference in the adjustment of the branch cock to the stuffing-box affected the 
apparent loss of heat to the extent of some O'l or 0'2 per cent, of the total heat, I 
determined to try and obtain some definite evidence of the relative slopes of tempe¬ 
rature in the two trials, by measuring the relative temperatures of the brass and the 
stuffing-box as far as was practicable. For this purpose, I liad thick brass tubes, 
radiating outwards, sweated on to the end of the stuffing-box to hold thermometers. 
Two such tubes were necessary on account of the screwing-up of the box, which had to 
be done whenever it began to leak ; and although this was not done during a trial, one 
tube would sometimes lace downwards, which was inconvenient. In a similar manner 
