TIME AND SPACE 281 



wooden frame in which was fastened a perpendicular cylinder closed at 

 the bottom and open at the top. In it was placed a piston and rod, and 

 on the rod a number of cogs. These cogs were geared into the cogs of 

 a pulley. At the end of the axle on which the pulley was fastened was 

 a hand behind which was attached the dial-plate to a wooden frame. 

 On the dial, each numeral from I. to XII. was marked twice, and the 

 hand moved round the whole face once in twenty-four hours. The con- 

 trivance was set in motion by starting a flow of water from a tank into 

 the space between the bottom of the cylinder and the piston. As the 

 piston-rod rose it turned the pulley and the shaft, and of course with it 

 the hand at the end. By regulating the pressure of the water in the 

 tank, the hand could be made to move faster or slower when it was 

 desired to lengthen or shorten the hours to conform to the relative pro- 

 portion of daylight and darkness in the twenty-four. Water-clocks 

 were formerly much in vogue in the east and were sometimes very 

 artistically constructed. Haroun al Easchid presented one to Charle- 

 magne that was provided with a striking mechanism and adorned with 

 movable figures such as are now quite common. The ancient Greek 

 designations for the time of both the day and of the night were very 

 vague : " the full market," " candle-lighting," " the first sleep," and 

 so on. Herodotus says the troops that were dispatched by Xerxes to 

 get in the rear of Leonidas left the camp " about the time of the light- 

 ing of the candles." It would have been more rational to say " about 

 dark," but he evidently used the common phraseology. Cock-crowing 

 was accepted as an indication of time. A well-known example is given 

 in the story of Christ's trial. It is still much relied on by the peasants 

 in some parts of Europe. In the nature of the case the Greek designa- 

 tions did not indicate the same actual time at all seasons of the year, 

 as candle-lighting would be much earlier in the winter than in the sum- 

 mer. Soldiers divided the night into five watches, the length of which 

 also varied with the seasons. It is not probable that they were accu- 

 rately measured. This division of time is doubtless the oldest; it is 

 several times referred to in the old testament. Sun-dials were a good 

 deal used by the ancients. The Greeks seem to have received them from 

 the Babylonians. Only the astronomers regarded the hours as of equal 

 length. So far as can be known they depended upon water-clocks. 

 But they were of much simpler construction than the one described 

 above, usually consisting merely of two vessels each of which had a 

 small orifice in or near the bottom. One of these vessels was placed 

 above the other and the water which had been poured into it allowed to 

 trickle slowly into the one underneath. When the lower vessel was full 

 the orifice in the upper was closed, that in the lower opened and placed 

 uppermost, when the same process would be repeated. The speakers in 

 the assembly were timed by these clepsydrse, as they were called; they 



