ON TT-TE MECHANiCAE EQUJVAl.ENT OF HEAT. 
333 
The Conduction hy the Levers. 
This lagging only extended over the body of the brake covering all the brass- 
work, leaving: the levers and balance wcig'hts on the levers bare. 
These levers being: in metallic contact with the brass of the brake assumed at 
these points the temperature of the brake, and would conduct the heat along to 
the balance weights till it was lost by radiation. As the temperatures were 
constant in all the trials this loss of heat would merely form part of the radiation 
and be eliminated as the rest; but, owing to the masses of the balance weights and 
the length of the levers, it must take a long time for the balance weights and the 
further parts of the levers to arrive at a steady temperature, a flict which would 
account for a greater loss of heat in the first trial made in the day. 
In order to obtain assurance that this delay produced no error it was arranged 
that after the completion of the series of trials with the brakes lagged, coi'responding 
to that with the naked brakes, that the balance weights should be removed, and 
only the load at 4 feet from the brake left, and a third series of trials made. 
Starting and, Stopping the Trials. 
34. Having adopted an hour as the length of each trial, and 300 revolutions as the 
normal speed, the engines having been running for an hour previously, while the 
water entering the brake was being adjusted, and afterwards, so as to insure the 
temperature, not only of the brake, but of the surrounding objects, having become 
approximately steady at the time of starting the trial, all that was necessary was 
that the counter should be pushed into gear, and at the same time the water-switch 
pushed over, and the reverse operation at the end of the trial. These operations, 
simple as they were, entailed errors, which arose partly from the impossibility of instan- 
Fig. 10. 
taneous engagement of the counter simultaneously with the switching of the water. 
In order to diminish these as far as possible, the spindle of a counter, on which was 
the worm which drove the worm wheel, was wrapped with a spiral S23rlng of steel 
wire, w'hich gripped the spindle so tight that it would not slip, the end of the wire 
being bent, so as to form a clutch standing off the shaft half-an-inch, the end of the 
wire being pointed, the shaft of the counter projecting a little beyond the wire. Facing 
the end of this shaft, and in line with it, was a socket in the end of the engine shaft, 
