1912-13.] Error caused by “Lag” in a Recording Instrument. 105 
constructed, and, in order that the effects of “ lag ” might be rendered more 
apparent, the moment of inertia of the wheel was made relatively large, 
while the friction at the bearings was kept as small as possible. The 
turbine is shown diagrammatically in fig. 1. 
Each end of the axle is supported on two anti-friction wheels, and, as 
the counting mechanism is very light, the whole runs very easily. A 
special cistern in which the water level could be kept constant was 
arranged in the room above vertically over the turbine, connection between 
the two being made by a straight-down 2-in. iron pipe. The pipe 
uul- 
fructiffu 
Fig. 1. 
terminates in a key tap and tapering nozzle, with aperture J in. in 
diameter. The key of the tap is moved by a long lever, the far end of 
which travels over a finely graduated scale, and a series of experiments 
was made to ascertain the rate of discharge from the nozzle in terms of the 
graduations. The constancy of the running of the turbine under the same 
conditions being of prime importance, experiments were undertaken to test 
this. The tap being opened to a predetermined extent, and the wheel 
having attained its full speed for that rate of flow, the time required for 
500 revolutions was noted on successive days. It soon became apparent 
that changes were liable to take place in the running, and this effect was 
traced to the fact that the friction varied with the amount of oil present 
in the bearings. Accordingly, an oil-tank was built up round each end of 
the axle, so that the bearings of the anti-friction wheels could be kept 
