188

LECTURE XVII.

ON TIMEKEEPERS.

.L HE measurement of time by clocks and watches is a very important and in-teresting department of practical mechanics. The subject is intimately con-nected with the consideration of astronomical instruments, but it is not essen-tially dependent on astronomical principles.

Time is measured by motion; but in order that motion may be a truemeasure of time, it must be equable. Now a motion perfectly free and un-disturbed, and consequently uniform, is rendered unattainable by the resist-ances inseparable from the actual constitution of material substances. l tbecomes therefore necessary to inquire for some mode of approximating to sucha motion. Astronomical determinations of time, which are the most accurate)can only be made under particular circumstances, and even then they a s 'sist us but little in dividing time into small portions.

The first timekeepers somewhat resembled the hour glasses which are stilloccasionally employed; they measured the escape of a certain quantity, not ofsand, but of water, through a small aperture. In these clepsydrae, it ap'pears from Vitruvius 's account that wheehvork was employed, and the houiwas shown on a graduated scale; the graduations were also probably so ad-justed as to correct the error arising from the inequality of the velocity oC 'casioned by the variation of the height of the water in the reservoir. 'I h |Sinconvenience was however sometimes wholly avoided, by means ot a constant steam, which kept the vessel full, or still more elegantly, by the siph° iaof Ilero, which was a bent tube supported by a float, so that its lower orificeat which the water was discharged, was always at a certain distance below tb e

® • d a

surface. Dr. Hooke proposed to keep the reservoir full, by means ot2