I 



1835.] Accidental and Complementary Colours. 285 



persistance of the impressions of the retina, on irradiation, 

 juxta-position of colours, &c.''^ 



3. No theory can, I conceive, be accepted which is founded 

 on a few isolated facts, and which neither agrees with varia- 

 tions of those facts, nor embraces other facts, in the same 

 branch of science, requiring classification. It too often 

 happens, and I may, perhaps, share in the blame now 

 bestowed on others, that a theory is framed from a few 

 observations of phenomena which may be consistent with 

 the theory, as far as it goes, but when other, and it may be, 

 less gifted observers, vary previous experiments, or multiply 

 facts which are in opposition to the theory, there is but one 

 resource to the philosophic mind, which is to modify or 

 reject it, and, in the latter case, to seek another ; and 

 although it may be said, as, indeed, I feel in the present 

 case, how much easier it is to point out discrepancies in one 

 theory, than to offer another which shall be free from fault, 

 still, as theory can only be supported by facts, so facts alone 

 can disturb or destroy it ; and if, after careful examination, 

 doubts still exist in my mind as to the truth of M. Plateau's 

 theory, if I have obtained his principal results as stated in 

 the preceding paper, by far different and simpler means, 

 and if these results have not led me altogether to the same 

 conclusions, my object has been, and is, to prove the non- 

 existence of oscillations in viewing these colours ; that posi- 

 tive and negative states do not exist ; and that, consequently, 

 the analogies upon which he relies with so much confidence, 

 are misapplied. My purpose, also is to consider the view 

 taken by P.C. of these colours, and, as he and M. Plateau 

 somewhat coincide, I must, consequently, state objections 

 which apply to both. I shall also state, hereafter, a theory 

 of my own, and shall endeavour to support it by the only 

 legitimate aid that ought, in the first instance, to be brought 

 to bear, i. e., direct experiment ; because, I repudiate all 

 abstract speculations unsupported by facts, conceiving that 

 they alone ought to support theory, and not, as, unfortu- 

 nately for science, it sometimes happens, that a theory 

 should precede observations and facts. 



4. My principal objection to the existing theories of acci- 



* " Je commencerai par exposer mes id^es sur la nature des couleurs acciden- 

 telles, parceque de la depend I'intelligence de ce qui doit suivre," 



