xiv.] UNITS AND STANDARDS OF MEASUREMENT. 315 



or to abandon its supposed relation to the earth's dimen- 

 sions. The French Government and the International 

 Metrical Commission have for obvious reasons decided in 

 favour of the latter course, and have thus reverted to the 

 first method of defining the metre by a given bar. As 

 from time to time the ratio between this assumed standard 

 metre and the quadrant of the earth becomes more accu- 

 rately known, we have better means of restoring that metre 

 by reference to the globe if required. But until lost, des- 

 troyed, or for some clear reason discredited, the bar metre 

 and not the globe is the standard. Thomson and Tait re- 

 mark that any of the more accurate measurements of the 

 English trigonometrical survey might in like manner be 

 employed to restore our standard yard, in terms of which 

 the results are recorded. 



The Pendulum Standard. 



The third method of defining a standard length, by 

 reference to the seconds pendulum, was first proposed by 

 Huyghens, and was at one time adopted by the English 

 Government. From the principle of the pendulum (p. 302) 

 it clearly appears that if the time of oscillation and the 

 force actuating the pendulum be the same, the leugth of 

 the pendulum must be the same. We do not get rid of 

 theoretical difficulties, for we must assume the attraction 

 of gravity at some point of the earth's surface, say 

 London, to be unchanged from time to time, and the 

 sidereal day to be invariable, neither assumption being 

 absolutely correct so far as w6 can judge. The pendulum, 

 in short, is only an indirect means of making one physical 

 quantity of space depend upon two other physical quan- 

 tities of time and force. 



The practical difficulties are, however, of a far more 

 serious character than the theoretical ones. The length 

 of a pendulum is not the ordinary length of the instru- 

 ment, which might be greatly varied without affecting the 

 duration of a vibration, but the distance from the centre of 

 suspension to the centre of oscillation. There are no 

 direct means of determining this latter centre, which 

 depends upon the average momentum of all the particles 



