280 Professor Powell on the Nature of the 



cumulation has extended itself very widely. Let it be granted 

 that there are certain classes of facts to which these results have 

 not hitherto been rendered applicable : — it yet remains to be 

 seen, whether, by the introduction of some further condition into 

 the primary supposition as to the sort of reciprocating motion, 

 they may not be made so. There are other facts which are at 

 variance with the results of the first principles, in the form in 

 which they are commonly assumed. The only question is, May 

 not those principles be somehow differently assumed so as to in- 

 clude the cases referred to? " Cases may arise" (as is observed 

 in the paper before quoted), " which require a supplementary 

 theory ; and our only care with optical theories at present must 

 be, that our present assumptions may admit of such a supple- 

 ment at some future time." The investigation of the theory is 

 comparatively recent ; but even now, those first principles have 

 been reduced into such a form as to render them applicable to 

 an immense range of results, for which nothing approaching to 

 an explanation has been offered on any other hypothesis. And 

 every new research evinces its powers more completely. The 

 accumulative argument is, therefore, rapidly advancing in 

 strength. 



We assume only what is a necessary consequence from facts 

 and observations ; but this is quite distinct from any proof of 

 the existence or physical properties of an etherial fluid. When, 

 therefore, we speak of the theory of undulations, or of an ether, 

 let us divest ourselves of all ideas derived from physical hypo- 

 theses, and fix our attention solely on those mathematical mo- 

 tions (so to speak) which are necessary to he supposed in accord- 

 ance with the possibility of two rays destroying each other'' s ef- 

 fect ; and which can alone be some of the various kinds of reci- 

 procating motion, or oscillation, of which mathematical analysis 

 enables us to trace the laws. It has indeed been said, and the 

 authority of great names has been referred to in support of the 

 idea, that it is possible to reconcile the fact of the interferences 

 with the theory of emission. In what particular way this may 

 be done, has not been made to appear ; but one thing is cer- 

 tain, that, in order to accomplish it, the emission theory must 

 be so framed and modelled, as to include the production of some 

 of those reciprocating motions before i-eferrcd to, a problem 



