OF LAND AND WATER ON THE SURFACE OF THE GLOBE. 299 



When it is recollected that these calculations were made from the 

 works of two distinct geographers, the agreement is possibly as near 

 as could be hoped for; but it is really closer than, at first sight, can 

 appear; for some little variations were accidentally introduced by the 

 first determinations not having been referred to, until the second had 

 been completed. It was then found that Van Dieman's land had been 

 weighed with New Holland in 1836, but that it had been taken with 

 the other Asiatic islands in 1823. The interior of Africa is so im- 

 perfectly known, that in the last instance it was thought best to 

 omit all attention to the inland lakes that were drawn upon this part 

 of the plates, although they were cut out from the other quarters of 

 the globe. This exception was not made in 1823, which will account 

 for the somewhat smaller quantity which was then found for this 

 continent. The difference in North America is to be attributed to 

 the discoveries which have been made of late years, and the larger 

 allowance which was in consequence assigned to the land within the 

 North Polar Circle. The deficiency for South America is not great, 

 but it was sufficient to make it desirable to ascertain, if possible, a 

 cause by which it might be occasioned. The larger dimension of 

 INIr Addison's plates afforded the means of cutting more deeply into 

 the great rivers of this continent : and Mr Carey, having been con- 

 sulted, pointed out another probable source of variation in the outline 

 of the eastern coast, which, in consequence of i-ecent surveys, is not 

 now laid down exactly as it was in 1823. It certainly is remarkable 

 that the deficiency, which in this case amounts to 0.72, is very nearly 

 the same as occurs for the whole southern hemisphere ; and it may also 

 be stated, that the land of the northern hemisphere came out in 1823 as 

 193.19, which is 4.02 less than is now assigned for it, while the quan- 

 tity (as may be seen above) which was then found for North America 

 Avas also less by 4.60. 



As the force of the present argument depends upon the two trials 

 being not only distant in time, but in every other respect independent 

 of each other, it may not be superfluous to mention that the numbers 

 were also deduced from them in two different ways. On Mr Carey's 

 globe the gores are 20" wide, and they extend each from the itquator 

 Vol. VI. Part II. Q « 



