<„.; 



THE 



AMERICAN JOURNAL OF SCIENCE 



[THIRD SERIES.] 



Art. XII. — On the Use of a Free Pendulum as a Time 

 Standard ;* by T. C. Mendenhall. 



The use of the new half second pendulums in the gravita- 

 tion work of the U. S. Coast and Geodetic Survey has sug- 

 gested the possibility of employing a pendulum with the 

 improved methods for ascertaining its period in terms of a 

 clock or chronometer second, as a standard of time which in 

 constancy and ease of application might go beyond anything 

 now readily attainable. 



The natural though not necessarily invariable unit of time 

 is the sidereal day and while it is sufficiently constant to satisfy 

 all requirements it is inconveniently long for nearly all opera- 

 tions, other than astronomical, in which great precision in 

 time measurement is required. Its subdivision is, therefore, 

 rendered necessary and this is accompanied by rather more 

 uncertainty than usually belongs to the subdivision of a stan- 

 dard. Even the daily rate of our best clocks and chronom- 

 eters is by no means constant and it is safe to say that in most 

 cases little if anything is known regarding their hourly varia- 

 tion from the mean for the day. That such variations, due to 

 fluctuations in temperature, pressure, and other less known 

 causes, exist is well known to all. In the physical, physiolog- 

 ical or other laboratory, in which it is desired to determine 

 intervals of time with a high degree of accuracy it is common 

 to depend upon the daily rate of chronometer or clock, which 

 rate is itself determined by means of clock signals from an 



*Read at the meeting of the National Academy in New York, November 11th, 

 1891. 



Am Jour. Sci.— Third Series, Vol. XLIII, No. 254.— February, 1892. 

 6 



