260 
THE TEACHING OF MECHANICS BY EXPERIMENT. 
carried by three chains. It is free to oscillate, and if the balance is 
not perfect it will oscillate when the motor is set to work. You ob¬ 
serve that the discs are revolving at a rapid pace, but there is no 
oscillation. This is because I have previously placed the weights in 
such positions as to get practically a perfect balance. But if I remove 
one weight leaving the others unaffected the balance is disturbed and 
you now see that the frame wobbles violently when the current is 
turned on. 
Another group of mechanical experiments of great educational value 
are experiments relating to the measurement of power. One meas¬ 
ures the power of engines, of electrical motors, and so on, most 
usually by means of an absorption dynamometer, and for small powers 
no form of absorption dynamometer is more convenient than a simple 
rope brake embracing the wheel. In other experiments one occasion¬ 
ally has to measure power on its way to a machine; one may wish for 
instance to measure the power which is required to drive some 
particular piece of machinery such as a lathe. The most convenient 
way of doing that, where practicable, is to drive the machine from 
an electric motor and to mount the motor on a cradle supported on 
knife edges so that the axis of the motor’s pulley is in the line of the 
kuife edge. If the belt coming away from the pulley to drive 
the machine is exerting equal pulls on both sides, then there is no 
tendency on the part of the cradle to be tilted, but as soon as the belt 
is doing any work it must pull unequally on its two sides and con¬ 
sequently the cradle tends to become tilted by the greater tension of 
the driving side of the belt. That tendency to tilt has to be met and 
counterbalanced by hanging weights to the end of an arm which 
comes out at right angles to the axis of suspension. By observing 
how much weight must be placed there to prevent the cradle from 
tilting we make a direct measure of the mechanical moment which is 
being exerted by the belt, and then by counting the number of 
revolutions also we see how much power is being given off. I have 
lately been applying this particular method of measurement to rather 
an interesting case, namely to measure the power absorbed in the 
rolling of a bicycle wheel. In a bicycle wheel there is a definite 
rolling resistance due to the imperfect elasticity of the tyre. This is 
very much less in the pneumatic tyre than it used to be in the cushion 
tyre, but it is greater in some pneumatic tyres than in others. When 
a bicycle wheel loaded with a weight is made to revolve, work is spent 
upon it. A small part of this work is spent in over coming the friction 
of the ball bearings, but much the greater part of it is due to the im¬ 
perfect elasticity of the tyre itself. If you are in the habit of coasting 
you may, perhaps, have noticed that when two or more riders start fair 
from the top of a hill and come down without using the pedals one 
arrives at the bottom before another. That is because one tyre differs 
from another in its rolling friction. I have experimented with differ¬ 
ent tyres and have found that in some cases the ratio is as much as 
2 to 1 ; it depends upon the particular material with which the tyres 
are lined. But this is a digression from the main subject of the 
