CONSTITUTION OF THE GLOBE. 25 



bo large a part of the surface scarcely exceeds 1. To 

 compute the mean density of the superficial parts of 

 the solid earth, is evidently impossible ; hut it probably 

 cannot much exceed 21, as already observed; and as 

 the depth of the ocean has been proved by La Place 

 to be necessarily considerable, while we are equally 

 sure that we have deep access to the earth, from the 

 elevation of mountainous strata, we may safely sup- 

 pose a superficial crust, of many miles in depth, in- 

 cluding the water and the land, not exceeding 2* in 

 weight. But the mean density of the earth itself is 5J 

 or more than double that quantity; and hence the 

 interior parts of the globe must possess a greater spe- 

 cific gravity than 5j, to counterbalance this want of 

 weight at the surface. It is also clear, that this spe- 

 cific gravity must be even such as to be equivalent to 

 that of many metals ; and hence it has been conjectured 

 that, instead of being formed of rock, the interior 

 parts of the earth must be, in some degree at least, of 

 a metallic nature- If, as was also shown from mathe- 

 matical considerations, there is a gradual increase of 

 density towards the centre, it becomes still more pro- 

 bable, that the nucleus of the earth is metallic. This 

 supposition has been supposed to receive additional 

 force from the phenomena of magnetism ; the nature 

 and position of the magnetic centres having been ima- 

 gined to imply the existence of a central mass of iron: 

 and it is thought to be confirmed still further by those 

 chemical facts which teach us that the rocks are formed 

 of metallic oxydes. 



But what the density of the centre of the globe 

 ought to be, on this view of its structure, has also 

 been computed on the theory of compressibility. If 

 5,4 be taken as the mean density, instead of 5,5 as 

 given by La Place, and the ell iplici ties be taken re- 



