VOL LXVIII.] PHILOSOPHICAL TRANSACTIONS. 421 



fore concludes this paper with a reflection or two on the premises above delivered. 

 Sir Isaac Newton thought it probable, that the mean density of the earth might 

 be 5 or 6 times as great as the density of water ; and we have now found, by 

 experiment, that it is very little less than what he had thought it to be : so much 

 justness was even in the surmises of this wonderful man 1 Since then the mean 

 density of the whole earth is about double that of the general matter near the 

 surface, and within our reach, it follows, that there must be somewhere within 

 the earth, towards the more central parts, great quantities of metals, or such 

 like dense matter, to counterbalance the lighter materials, and produce such a 

 considerable mean density. If we suppose, for instance, the density of metal 

 to be 10, which is about a mean among the various kinds of it, the density of 

 water being 1, it would require l6 parts out of 27, or a little more than one- 

 half of the matter in the whole earth, to be metal of this density, in order to 

 compose a mass of such mean density as we have found the earth to possess by 

 the experiment : or -[I3-, or between 4 and ^ of the whole magnitude will be 

 metal ; and consequently if, or nearly -| of the diameter of the earth, is the 

 central or metalline part. 



Knowing then the mean density of the earth Water 1 



in comparison with water, and the densities of The sun 1^ 



all the planets relatively to the earth, we can Mercury 101. 



Venus . 0' 



now assign the proportions of the densities of The earth 5^ 



them all as compared to water, after the manner Mars 3|. 



of a common table of specific gravities. And The moon — 3^ 



the numbers expressing their relative densities, Jupiter ix 



in respect of water, will be as annexed, sup- *^ 



posing the densities of the planets, as compared to each other, to be as stated in 

 Mr. De Lalande's astronomy. 



Thus then we have brought to a conclusion the computation of this important 

 experiment, and, it is hoped, with no inconsiderable degree of accuracy. But 

 it is the first experiment of the kind which has been so minutely and circumstan- 

 tially treated ; and first attempts are seldom so perfect and just as succeeding en- 

 deavours afterwards render them. And besides, a frequent repetition of the 

 same experiment, and a coincidence of results, afford that firm dependance on 

 the conclusions and satisfaction to the mind, which can scarcely ever be had 

 from a single trial, however carefully it may be executed. For those reasons it 

 is to be wished, that the world may not rest satisfied barely with what has been 

 done in this instance, but that they will repeat the experiment in other situations, 

 and in other countries, with all the care and precision that it may be possible to 

 give to it, till a uniformity of conclusions shall be found, sufficient to establish 

 the point in question beyond any reasonable possibility of doubt. What has 



