C84 Memorandum on the Adjustment of Dynamometers. 
proportional to the distance traversed, therefore the revolutions, 
of the integrating wheel must be proportional to the distance 
multiplied bj the strain ; in other words, the space passed over 
and the strain are multiplied together, and appear as figures 
on the integrating counter, which represent the work done in 
foot-pounds or any other units. 
To arrive at the numerical value of the particular measure 
required, it is necessary to multiply the numbers appearing on 
the counter by a co-efficient, which will be different for different 
springs. To obtain these co-efficients, the instrument — being in 
perfect order, newly cleaned and oiled, every part working freely 
— is placed in its working position, the springs being unloaded, 
and the integrating wheel set exactly over the centre of the disc, 
so that turning the latter will produce no effect on the former. 
And here comes the first difficulty, namely, to determine when 
the springs are completely passive ; for, when they are nearly 
free from strain, the smallest pressure will move them a little 
either Avay, Avhere they will remain, unable to overcome the 
friction of the machine. An experienced operator with a small 
machine can find the zero-point readily by easing the springs 
backwards and forwards, and bringing them to rest by degrees ; 
but in heavy machines the best plan is, by means of a pinch-bar, 
to press the springs over, first one way and then the other, 
marking the points to which they spontaneously return, and 
■then adjusting the integrating wheel to the exact centre of the 
disc, when the springs are set half-way between the two points 
so found. 
The integrating wheel being adjusted, and the counter set to 
zero, or its reading noted, a known weight is put upon the 
springs, either by means of a bell-crank lever, or by hanging 
the instrument up in a crane ; in which latter case the weight 
of the draw-bar, and any other part that tends to affect the 
springs, including, generally, part of their own weight, must be 
ascertained, and considered added to the load. The measuring 
wheel is then turned round till it represents any c\istance, say 
300 feet, passed over, after which the indications of the counter 
are noted. This should be repeated two or three times, the 
springs being relieved of their load each time. The load should 
then be increased, and, after another run of 300 feet, the counter 
should again be noted. A third or fourth run, or more, with 
increased loads should then be made, the greatest loads being 
kept well within tlie limits of elasticity of the springs. 
The results obtained should be treated in the following 
manner: — The several loads should be set off to scale on the 
base-line of a diagram, like Fig. 1, as eg, ok, oZ, and ordi-i 
nates raised on which should be set off tlie corresponding mean' 
