58 DEPTH OF THE SEA. [SECT. vi. 



mean of the whole appears not to differ much from 3 g ; that 

 given by the lunar theory has the advantage of being inde- 

 pendent of the irregularities of the earth's surface and of 

 local attractions. The regularity with which the observed 

 variation in the length of the pendulum follows the law of 

 the square of the sine of the latitude proves the strata to 

 be elliptical, and symmetrically disposed round the centre of 

 gravity of the earth, which affords a strong presumption in 

 favour of its original fluidity. It is remarkable how little 

 influence the sea has on the variation of the lengths of the 

 arcs of the meridian, or on gravitation ; neither does it much 

 affect the lunar inequalities, from its density being only about 

 a fifth of the mean density of the earth. For, if the earth were 

 to become fluid, after being stripped of the ocean, it would 

 assume the form of an ellipsoid of revolution whose com- 

 pression is 3U |. S , which differs very little from that determined 

 by observation, and proves, not only that the density of the 

 ocean is inconsiderable, but that its mean depth is very small. 

 There may be profound cavities in the bottom of the sea, but 

 its mean depth probably does not much exceed the mean 

 height of the continents and islands above its level. On this 

 account, immense tracts of land may be deserted or over- 

 whelmed by the ocean, as appears really to have been the case, 

 without any great change in the form of the terrestrial sphe- 

 roid. The variation in the length of the pendulum was first 

 remarked by Richter in 1672, while observing transits of the 

 fixed stars across the meridian at Cayenne, about five degrees 

 north of the equator. He found that his clock lost at the 

 rate of 2 m 28 s daily, which induced him to determine the 

 length of a pendulum beating seconds in that latitude ; and, 

 repeating the experiments on his return to Europe, he found 

 the seconds' pendulum at Paris to be more than the twelfth 

 of an inch longer than that at Cayenne. The form and size 

 of the earth being determined, a standard of measure is fur- 

 nished with which the dimensions of the solar system may 

 be compared. 



