THE ACCURATE MEASUREMENT OF TIME. 661 



devices which have been found to do this with anything like accuracy 

 the pendulum and the balance-wheel. Postponing, for the present, the 

 discussion of the balance-wheel, we have before us the pendulum. 



That the pendulum might be used for the measurement of time is 

 a discovery which dates back, at least, to the time of Galileo. A great 

 many, even among well-educated people, suppose that that philosopher 

 discovered that a pendulum of a given length will always oscillate in 

 the same time. This he could not have discovered, for it is not true. 

 The correct statement of the law he discovered is, that a pendulum 

 will always oscillate in the same time through equal arcs, but not 

 through any arc. It has been found that if the curve in which a 

 pendulum swings were a portion of a cycloid instead of a circle, and 

 the pendulum were simple, that is, consisting of a bob or ball sus- 

 pended by means of a thread imagined to have no weight, its oscilla- 

 tions would be in equal times through any arc. To accomplish this, 

 clocks were at one time made with pendulums which were suspended 

 between cycloidal cheeks, and were thus conformed, in swinging, to the 

 cycloidal curve. But this was soon abandoned, as it was found that 

 it was impossible to construct such cheeks without variation or im- 

 perfection sufficient to make a greater error in the pendulum than it 

 would have if allowed to swing on a circular arc. For a short dis- 

 tance the cycloidal curve corresponds quite closely to the circle. 

 Therefore, by adjusting the pendulum to swing but a short distance, it 

 was found to be possible to secure substantial uniformity. This is the 

 plan now universally adopted. 



If the arc of vibration is increased, the clock will lose time. 

 Experience with the common house-clock would seem to contradict 

 this, for every one has noted that when the clock is first wound it will 

 gain time, and then that it will lose as it runs down, and, seemingly, 

 this is due to the difference in the swing of the pendulum. The ex- 

 planation, however, is to be found in the fact that in case the pen- 

 dulum swings farther it is shortened by the curving of the spring 

 by which it is suspended, and also by an effect which the longer 

 swing is found to have upon the escapement, quickening its time. 

 Any ordinary house-clock would keep far better time if the weight of 

 its bob were considerably increased, as this would do much toward 

 equalizing its swing. Ordinarily the weight of the bob is about three 

 ounces ; if the clock is properly put in beat it will carry a bob weigh- 

 ing as many pounds, and all spring-clocks would be greatly improved 

 as time-keepers by such a change. If this is true, some one may ask, 

 Why is it not made by the manufacturers ? The answer is, that any 

 firm who should put such clocks on the market, superior time-keepers 

 as they would really be, would soon find themselves getting the reputa- 

 tion of making a clock that would not run, and all because the public 

 generally would not have the skill or the patience to adjust the beat 

 properly. Let a servant, for example, take such a clock to her room 



