190 CONTINUITY. 



follows, indeed the probability is rather the other way, that the 

 proportions of the substances in the interior of the earth are 

 the same as on the exterior. It might be worth the labour to 

 ascertain the mean specific gravity of all the known minerals 

 on the earth's surface, averaging them on the ratios in which, 

 as far as our knowledge goes, they quantitatively exist, and 

 assuming them to exist without the oxygen, chlorine, &c., 

 with which they are, with some rare exceptions, invariably 

 combined on the surface of the earth : great assistance to the 

 knowledge of the probable constitution of the earth might be 

 derived from such an investigation. 



While chemistry, analytic and synthetic, thus aids us in 

 ascertaining the relationship of our planet to meteorites, its 

 relation in composition to other planets, to the sun, and to 

 more distant suns and systems, is aided by another science, 

 viz. optics. 



That light passing from one transparent medium to another 

 should carry with it evidence of the source from which it 

 emanates, would, until lately, have seemed an extravagant 

 supposition ; but probably (could we read it) everything con- 

 tains in itself a large portion of its own history. 



I need not detail to you the discoveries of KirchhofT, Bun- 

 sen, Miller, Huggins, and others ; they have been dilated on 

 by my predecessor. Assuming that spectrum analysis is a 

 reliable indication of the presence of given substances by the 

 position of transverse bright lines exhibited when they are 

 burnt, and of transverse dark lines when light is transmitted 

 through their vapours, though Pliicker has shown that with 

 some substances these lines vary with temperature, the point 

 of importance in the view I am presenting to you is, that 

 while what may be called comparatively neighbouring cosmi- 

 cal bodies exhibit lines identical with many of those shown by 

 the components of this planet, as we proceed to the more dis- 

 tant appearances of the nebulae we get but one or two of such 

 lines, and we get one or two new lines not yet identified with 

 any known to be produced by substances on this globe. 



Within the last year Mr. Huggins has added to his former 

 researches observations on the spectrum of a comet (comet I 

 of 1866), the nucleus of which shows but one bright line, while 



