REGULATION OF MOTION BY FLY WHEELS. 60 



The case is the same witli a machine moved by 

 water, or by a weight ; for though the strength of 

 these does not exhaust itself like that of an animal, 

 yet the yielding of the parts of the machine renders 

 the impulse much less after it begins to move ; 

 hence its velocity is accelerated for some time, 

 until the impulse becomes so small, that the ma- 

 chine requires an increase of power to keep up the 

 necessary motion. Then the machine slackens its 

 pace, the water meets with more resistance, and of 

 consequence exerts its power more fully, and the 

 machine recovers its velocity. 



But when a fly is added to the other parts, this 

 acts first as a power of resistance, so that the 

 machine does not acquire the velocity it would 

 otherwise do. When it next begins to yield to the 

 pressure of the water, and the impulse of course to 

 slacken, the fly communicates part of its motion to 

 the other parts; so that if the machine be well 

 made, there is very little difference in the velocity 

 perceptible. 



The t^'uth of what is here advanced will easily 

 be seen, from considering the inequality of motion 

 in a clock, when the pendulum is offj and how 

 very regularly it goes when regulated by a pen- 

 dulum, which here acts as a fly. 



Flies are particularly useful in any kind of work 

 which is done by alternate strokes, as the lifting 

 of large pestles, pumping of w r ater, &c. In this 

 case, the weight of the wheel employed is a prin- 

 cipal object ; and the method of calculating this is 

 to compare it with the weight to be raised at each 

 stroke of the machine. Thus, suppose it is re- 

 quired to raise a pestle 30 pounds weight to the 

 height of one foot, 60 times in a minute ; let the 

 diameter of the fly be seven feet, and suppose the 



vol. i. f 



