308 : . HISTORY OF ASTRONOMY. 



fraction of a second is to be estimated by the ear. Now 

 this fraction can not be estimated with the accuracy 

 which is demanded in this kind of comparison. It is 

 found that observers generally estimate the fraction of a 

 second too small when using the ear alone, unassisted by 

 the eye. This error is greatest at the middle date be- 

 tween two clock beats, and is found to vary from 0*06 to 

 0-18 of a second with different observers. 



The experience of a few nights with the preceding 

 method of observation, suggested a second method of 

 comparison which relies on the coincidence of a mean 

 solar and sidereal clock or chronometer. A sidereal 

 clock gains upon a solar clock one second in about six 

 minutes ; and if two such clocks are placed side by side 

 they must tick together once in every six minutes. In 

 order to compare two such clocks, we notice their move- 

 ments, and wait until the beats sensibly coincide, when 

 we know that their difference amounts to an entire num- 

 ber of seconds, which is readily discovered. Chronom- 

 eters generally make two beats in a second; so that 

 between a clock which beats seconds of sidereal time, 

 and a chronometer which ticks half seconds of solar 

 time, there must be a coincidence every three minutes. 



The following is the method pursued in the comparisons 

 between Philadelphia and Jersey City. After transmit- 

 ting a few signals by the former method, so as to deter- 

 mine the difference between the local times of the two sta- 

 tions within a small fraction of a second, the party at the 

 first station commences striking on his signal key every 



