Astronomical and Nautical Collections. 129 



sistent, in all their extent, with the simplest principles of 

 the theory of undulations. Without doubt observation has 

 assisted also in the discovery of this relation ; but observa- 

 tion alone could never have developed them ; while in this 

 case, as well as in many others, the theory of undulations 

 might easily have gone before the experiments, aud have 

 predicted beforehand the precise results in their minutest 

 details. 



The example, which has already been mentioned, suffi- 

 ciently proves that the choice of a theory is not a matter 

 of indifference in the investigation of physical phenomena. 

 Its utility is not confined to the purpose of facilitating the 

 study of facts, by uniting them into groups more or less 

 numerous, according to the relation which they bear to 

 each other. Another no less important end of a good theory 

 must be to contribute to the essential advancement of 

 science, by the discovery of facts, and of relations between 

 classes of phenomena the most distinct, and in appearance 

 the most independent of each other. Now it is obvious that 

 if we set out with an imaginary hypothesis respecting the 

 cause of light, we shall not reach this end so readily, as if we 

 possessed the true secret of nature. A theory depending upon 

 a true fundamental hypothesis, however ill it may be suited 

 for the application of mathematical analysis to the mecha- 

 nical operations which it involves, will still lead us to inti- 

 mate relations between distant facts, which would for ever 

 have remained unknown upon any other suppositions. 

 Hence, to say nothing of our natural curiosity to know the 

 truth in all cases, we see how important it is to the progress 

 of optics, and of the kindred science of chemistry, to know 

 whether luminous molecules are actually projected from the 

 sources of light and enter our eyes, or if light is propagated 

 by the undulations of a continuous medium, to the particles 

 of which, the luminous bodies communicate their vibratory 

 motions. And we have no reason to believe that the de- 

 cision of this question is impossible, because it has long re- 

 mained unaccomplished: we may even venture to assert 

 that, as far as probability goes, .it is already determined ; 

 and that, after comparing attentively the application of the 



JAN.— MARCH, 1827. K 



