TIME-RECKONING. 109 



singular consequence, the number of hours were made constant 

 between sunrise and sunset, and instead of being equal in length, the 

 hour varied with the length of daylight. Whatever the moments of 

 sunrise and sunset, the interval of light was divided into liJ parts. 

 If the sun rose at 4 a.m. and set at 8 p.m., according to our notation, 

 each hour would be equal in length to 80 of our minutes. Old habits 

 are so strong that this constantly varying system was adhered to long 

 after mechanical time-keepers were introduced, and attempts were 

 made to regulate clocks to tell the unequal hours. Like the Romans, 

 the Greeks divided the intervals of light between sunrise and sunset, 

 whatever its length, into 12 equal parts, subject to change from day 

 to day. The custom of making the hours variable is still followed 

 by some eastern nations. 



The system of dividing the day by the rising and setting of the 

 sun makes the hours indefinite periods, as they continuously change 

 with the seasons. Except at the equinoxes, the hours of the night 

 and day can never be of equal length. Near the equator the varia- 

 tions are least ; they increase with every degree of latitude until the 

 arctic and antarctic circles are reached, within which a maximum is 

 attained. Even in the latitude of Rome, the length of the hours of 

 daylight and darkness under this system have an extreme difference 

 of 75 minutes. In Spitzbergen the sun sets about the beginning of 

 November, and remains below the horizon for more than three 

 months. It does not set for an equal pei-iod after the middle of May. 



Sun dials had two great defects, they were vmserviceable at night 

 and during cloudy weather. The clepsydi-a or water clock was 

 accordingly introduced at Rome about 158 B.C., by Scipio ISTasica 

 (Jorculum. It measured time by allowing water to escape throuo-h 

 an orifice in a vessel, as sand flows through a modern sand glass. 

 Subsequently some sort of toothed-wheel work was applied to the 

 clepsydra by Ctesibius (A. D. 120). Diurnal and nocturnal time 

 was measured in this or some other rude manner for many centuries. 

 Besides sun dials, gnomons and clepsydrae, all of which appear to 

 have been known to the Egyptians, Indians, Chaldeans, Babylonians 

 and Persians long before their Introduction at Rome, mention is 

 made of a contrivance by which a mechanical figure dropped a stone 

 into a brazen basin eveiy hour, producing a loud sound which for a 

 great distance announced the divisions of time. King Alfred em- 

 ployed as a time-keeper six wax candles, each 12 inches long. Three 



