THE ANTARCTIC REGIONS. 277 



water on our globe with Mr. Gardner's statement that 

 twenty-six out of twenty-seven parts of the land are 

 antipodal to water, to see that this must be so. There 

 are about 51 millions of square miles of land and 

 about 146 millions of square miles of ocean. Now 

 about 49 millions of square miles of land are antipodal 

 to water, accounting therefore for only 49 millions out 

 of the 146 millions of square miles of ocean surface ; 

 the remaining 97 millions of square miles of ocean 

 are, therefore, not antipodal to land, but one half (any 

 we please) antipodal to the other half. In fact, we 

 have this rather singular result, that the ocean surface 

 of the globe can be divided into three nearly equal 

 parts, of which one is antipodal to land while the 

 other two parts are antipodal to each other. This ob- 

 viously does not force upon us the conclusion that an 

 unknown region must be land because a known region 

 opposite to it is oceanic ; and still less can such a con- 

 clusion be insisted upon when the region opposite the 

 unknown one is itself unknown.* 



* Whether the relation above-mentioned respecting land regions is 

 noteworthy may very well be questioned. It will be seen that Capt. 

 Maury regards it as seemingly a physical law * that land should not be 

 antipodal to land.' Now this is by no means satisfactorily indicated. 

 As a question of probability it is not certain that the present relation, 

 by which twenty-six parts out of twenty-seven of the land are antipodal 

 to water, can be regarded as antecedently an unlikely one, when nearly 

 three-fourths of the whole surface are occupied by water, and when, 

 also, the bulk of the land and water regions consist of such great 

 surfaces as those we call continents and oceans. Granted these pre- 

 liminary conditions, it would appear, indeed, that only by a very 

 remarkable and, as it were, artificial arrangement of land and water 

 could any but a small proportion of the land be antipodal to land. The 



