On a Problem in relation to the Figure of the Earth. 261 



tempted to make a comparison between it and the likewise ex- 

 ceedingly subtile but still luminous matter which composes the 

 nebulae and cometary bodies. This analogy becomes still more 

 striking when we examine closely the appearance presented in a 

 tube containing rarefied hydrogen traversed by the electric dis- 

 charge, by the kind of mist which shows itself when a small 

 quantity of gas is caused to enter the tube, as I described in § 2, 

 and which likewise appears in the dark space when a certain 

 degree of rarefaction has been passed. The gaseous matter is 

 there even more highly rarefied than it is at the other parts of the 

 mass, thus making its resemblance to the luminous matter form- 

 ing the comets and nebulae still more decided. We may add that 

 recent researches on the part of various astronomers have shown 

 that the rays yielded by the prismatic analysis of the light of 

 these celestial bodies are exactly similar to those given by the 

 electric discharge when transmitted through rarefied nitrogen, 

 and especially through hydrogen. Do these gases, then, which 

 intervene in most of the phenomena of terrestrial physics, play 

 an equally or even more important part in the phenomena of cos- 

 mical physics ? There is nothing improbable in this conjecture, 

 especially since the analysis of aerolites has shown that planetary 

 space does not contain any element not found also upon our 

 globe. 



XXXIV. What changes can be made in the arrangement of the 

 materials of the mass of a Body, its external form remaining the 

 same, without affecting its attraction on an external particle ? 

 By Archdeacon Pratt, F.R.S. 



To the Editors of the Philosophical Magazine and Journal. 



On the Indus, 

 Gentlemen, January 30, 1867. 



IT has been observed to me that, in a paper I sent to you last 

 year on the above subject (Phil. Mag. S. 4. vol. xxxii. p. 

 132), I have materially limited the generality of the investigation, 

 by assuming that p the density can be expanded in a convergent 

 series of integral powers of r; also that I seem not to have con- 

 sidered that M , M„ M 2 , . . . may be functions of i. (These 

 letters will be understood by a reference to my former paper. 

 Or they will be understood by what follows.) ,1 now give the 

 demonstration again, modified so as to meet these objections. 



If we subtract the body when under its original arrangement 

 from the body under its new arrangement, particle from particle, 

 we obtain an Imaginary Body, the total mass of which is zero 

 (some parts being of positive and some of negative density), and, 



