338 



alone furnishes a conclusive argument against the theory con- 

 sidered. 



" It manifestly follows, from the value which has been found 

 for £, that if ci were small, the waters would tend to accumu- 

 late about the equatorial regions ; and if, on the contrary, Ci 

 were large, they would tend to accumulate about the polar 

 regions. If, therefore, from any superficial causes, the earth's 

 jfigure became gradually more oblate, the extent of polar dry 

 land would gradually tend to lessen, while that of the equato- 

 rial regions would at the same time tend to increase. The 

 truth of our fundamental proposition cannot, therefore, admit 

 of any further doubt. 



"4. It may be useful to give still greater force to these 

 conclusions by some additional considerations. With the sup- 

 posed original spherical figure of the earth, the circumambient 

 fluid would, as already remarked, assume, by the action of 

 centrifugal force, a spheroidal form. The fluid would thus 

 tend to accumulate towards the equator, and to recede from 

 the poles. Circumpolar continents might thus be formed, with 

 a great equatorial ocean between them. Some of the fore- 

 going. expressions will assist in determining the conditions of 

 the existence of such continents. 



" Let 9i represent the complement of the latitude at the 

 parallel bounding a circumpolar continent on the surface of 

 the primitive sphere, then the area of this continent will be 



I (1 - COS0i), 



the area of the sphere being unity. But at that parallel r = ai, 

 and, therefore, 



making Ci = in (b). Hence 



