Memorandum on the Adjustment of Dynamometers. 
687 
little, and hence the correction must be added : and here again 
the deficiency will be augmented by the distance passed through, 
and not by the total strain put upon the springs. It follows 
that the co-efficient of error, b, must be multiplied by the 
distance passed over to get its total amount, and the general 
equation for the work done will be 
Work done — C ±h x d) = k x. d 
where C — co-e£Bcient of work 
& = „ error 
n = indications of integrating counter 
d = distance run 
h = weight or strain on springs. 
The mean value of b per unit of distance is ascertained from 
the diagram thus : — Let x be the correction o s for a distance 
of y feet, then 
h X y = X .•. h = - ; 
y 
that is to say, the co-efficient b per unit of distance is found by 
dividing the mean value, .r, of the ordinate o s, by the distance 
run in making the trials. Therefore 
k X d = C ± ^^^^ = work done in foot-pounds, 
and from that 
^ _ k X d 
X X d 
n± 
y 
If an instrument is in good order this equation will give 
almost identical values of C for all the trials. 
If there is no error, 
, ^ k X d 
X = 0, and u — , 
equation first given. 
I In using spring-dynamometers, whether rotatory or traction, 
is necessary to test the co-efficients frequently, because the 
prings are liable to take a set. The joints get worn, and the 
|tegrating wheel wears smaller. The disc should also be kept 
the integrating wheel when not actually used to register, 
cause, as the wheel plays backwards and forwards across its 
ce, as long as the instrument is being drawn along, it is apt to 
jrear into flats, and so get out of shape. 
In Fig. 3, a case, which actually occurred in practice with the 
lough-dynamometer, is worked out. The following were the 
lean results of the trials : — 
Load on Springs. J'ean Reading of Index. Distance. 
705 lbs 159-7 .... 192 yards 
593 „ .. .. 129-3 .. .. 192 „ 
360 „ .. .. 68-5 .. .. 192 „ 
