428 THE YARD, PENDULUM, AND METRE. 



diai?ieter might be more advantageously substituted : but 

 that we have good reason to beUeve the equator to be 

 not strictly circular, but in some degree elliptic, the pro- 

 portion of its greatest and least diameters not being yet 

 precisely known, though very much nearer to equality 

 than that of the equatorial and polar diameters. This 

 hov/ever would not prevent its 7nea?i equatorial diameter 

 from being assumed in preference to its circumference, 

 were not the polar axis, for very obvious reasons, prefer- 

 able to both. Of the latter, and indeed of all three 

 (thanks to the elaborate geodesicai surveys which have 

 been made within the century last elapsed), we possess 

 a knowledge so precise as to render them perfectly avail- 

 able for our purpose. 



(9.) Of lengths which exist not marked by the dimen- 

 sions of any m,aterial object, but which are defined by 

 the nature of things and by physical relations, and which 

 are susceptible of exact determination and of being 

 marked off on a scale, and so of becoming materialized 

 for practical reference ; there have been proposed only 

 three which can be considered theoretically, and of these 

 only one practically available. These are, ist, the vel- 

 ocity of light or the space travelled over by light in some 

 definite time (say the ten-mdllionth part of a second, 

 which would give a modulus of about 100 feet); 2dly, 

 the length of an undulation of a ray of light of somie 

 definite refrangibility a length so minute as to require 

 multiplication a million-fold to give a modular unit ; and 

 3dly, the length of a pendulum vibrating seconds under 

 certain definite and normal circumstances or rather 



