66 Proceedings of the Royal Irish Academy. 



YIII. — Eepokts PBoai the Obseevatoet of TKEfiiT College, DtrsLi^r. 

 By Egbert S. Ball, LL. D., F. E. S., Andrews Professor of 

 Astronomy in the University of Dublin, and Eoyal Astronomer of 

 Ireland. 



ISTo. 1. — 0:^^ the Method of Eegulating a Clock r^*TE^^DED to show 



COKEECT jMEA2f TlilE. 

 [Read, FeLiiiary 26, 1877.] 



Ik- tte following note I describe the method which I have adopted for 

 the regulation of the new mean time clock erected by Messrs. Booth, 

 of Dublin, at the Observatory of Dunsink. This clock controls an 

 electric current, which goes from the Observatory to Dublin, for the 

 purpose of regulating the clock in the Port and Docks Office of Dublin 

 to correct Dublin time. 



In the regulation of the clock the object to be attained is to make 

 the clock show correct time directly. In other words, we try to have 

 the error of the clock always small. The method I have employed is 

 the well-known one of applying small correcting weights to the pen- 

 dulum at the centre of its length. It is possible, however, that the 

 following description of the way to make this correction in a system- 

 atic and orderly manner may be of use : — 



It was found, by experiment, that a weight of 635 milligrammes, 

 placed upon the " shelf," which is fixed on the pendulum rod at the 

 centre of its length, increased the rate of the clock one second per 

 diem. 



I shall suppose that the error of the mean time clock is to be 

 determined every day at mean noon by comparison with the astro- 

 nomical clock. 



Suppose that yesterday at noon the error was E', and that to-day 

 at noon the error is -E, the rate of the clock is therefore JE - E', and 

 consequently the error at noon to-morrow would be 2 ^ - J^'. If, 

 therefore, we wish to have the clock right at noon to-morrow, a 

 number of weights equivalent to 1 E - E' must be added to or taken 

 from the shelf at noon to-day. The principle of the correction is that, 

 by comparison of the errors of noon to-day and noon yesterday, wc 

 endeavour to make the clock right at noon to-morrow. 



In the practical application of this method we may assume that 

 we never have to deal with a clock more than two or three seconds 

 wi'ong, and under these circumstances the correction may be made 

 with the greatest facility. 



We shall assume that half a second is the smallest portion of time 

 of which it is necessary to take cognizance. The following Table will 

 then tell us at once how many weights should be placed on or taken 

 off the shelf:— 



