Rate of Air Movement in Transpiration Experiments . 219 
that the ordinary fan motor is not sufficiently constant in speed but tends 
to slow down slightly. The most satisfactory motor was found to be one 
with a governing mechanism attached made by R. W. Paul of New South- 
gate. 1 With this arrangement the slow air-currents, as low as 5 metres 
per minute, which have been mostly used, are easily obtainable. 
The speed of the air-current is estimated by means of an anemometer 
except in the case of very low speeds, when the rate of movement of smoke 
through the flue is timed by a stop-watch. 
The regularity of the air movement in the apparatus was tested by 
atmometer readings under constant conditions of temperature and humidity. 
An experiment carried out in a dark room will serve to show the variation 
which occurred. The dark room is provided with the usual double doors, 
which are found useful when it is necessary to keep the humidity or tem- 
perature of the room above or below that of the laboratory out of which it 
opens. The table on p. 220 gives an example of the variations in evaporation 
rate in the dark room as determined by half-hourly weighings of a porous-cup 
atmometer exposed in the air-flue to a wind velocity of 7 metres per 
minute. Periodical readings of temperature and humidity were not made, 
1 When the motor reaches a certain speed, centrifugal action causes a weight attached to the 
shaft to move a lever which breaks the circuit through the coils. The lever works against a spring 
which, when the speed of the motor decreases, pulls the lever back to complete the circuit again. 
Thus with a continuous make-and-break mechanism, a regular speed is maintained. This type of 
motor was designed to drive the clock of a recording drum. 
