314 
PROFESSOR 0. RErXOLDS ARD MR. W. H. MOORBT 
The Error of Balance of the Brahe. 
Although, when the shaft is running, the brake levers are perfectly free between 
the stops, yieldiug to the slightest force even when carrying a load of 400 pounds in 
addition to the weight of the brake-case of over 300 pounds, yet, when the shaft is 
standing, it requires a moment of some 40 ft.-lbs to move the lever in either direc¬ 
tion, so that the balance can only be obtained as the difference of these moments, 
and this can only be obtained to about 1 foot pound. But, it is to be noticed that 
as long as the distribution of weights are unaltered and the lever is in the same 
position, any error of balance, whatever might be its cause, wmuld be the same for 
all trials, no matter what might be the difference in the suspended load : so that, in 
taking the difference of the trials, the error would be eliminated, and, to insure this, 
the automatic adjustment was so arranged that, by a screw adjustment, the lever 
could be raised or lowered without affecting the automatic adjustment of the valves 
(see fig. 4, p. ;]08). Also an index was arranged adjacent to the end of the lever to 
which it might be always adjusted (showm in Plates 4 and 5). 
The Error of Balance resulting from Eriction of the Automatic Gear, 
Ibis had been a matter of serious consideration in designing the brakes, for, 
although it was obviously possible to so balance the parts of such gear that there 
should be no pressure against (he fixed support arising from the weight of this gear, 
it was not obvious that the friction of these valves and their gear wmuld not allow of 
a steady resistance to motion being maintained—would not allow the brake to lean 
against the fixed support within the limits of friction. However, after careful 
consideration of various contrivances, I came to the conclusion that, if the gearing 
between the support and the valve were inelastic, the joints being an easy fit, the 
tremor of the shaft and the brake, when running, might be depended upon to release 
any frictional resistance in this gear ; so that, after any change, the gear would 
rapidly return to equilibrium. This proved to be the case, even to an unexpected 
extent, as wms shown by the freedom of all the pins. 
It was subsequently found by experiment that, even when the valves wmre so tight 
that it required a moment of 30 ft.-lbs. on the brake to move the automatic gear 
alone, with the shaft standing, in either direction, when the shaft was running auy 
tendency to lean upon the support in either direction was the result of imperfect 
balance in the gear; and that, by adjusting this balance to an extent which would 
not cause a moment on the brake of 0 01 ft.-lb., the tendency of the brake to lean 
either in one direction or the other might be reversed—showing that, with a load of 
600 ft.-lbs., the relative limits of error are < ± O'OOOOIG, and in the difference of the 
trials wmuld be zero. 
