243 



From the condition that this sliall for the eartli not exceeds ay 

 2" per century we tind 



R>10«. 



Then V/^"^<C^^^"^ i*^ actually of the second order compared with 

 x=r25-10-8. This limit of R is still considerably lower than the 

 value which was found above from the displacetnent of the spectral 

 lines. For the planetary motion — and generally for all mechanical 

 problems which do not involve very large values of r — we can 

 therefore in both systems A and B neglect the effect of A entirely. 



