332 ON LIGHT. 



validity, and the soundness of the theory it rehes on, by 

 experiment. There can be . no better presumptive 

 evidence of the truth of a physical theory than its 

 enabling us to predict, antecedent to trial, a result in 

 direct contradiction to what mankind in general would 

 consider as the obvious conclusion of common sense 

 founded on all ordinary experience. This is the case in 

 the present instance. Since the total illumination of one 

 point P on the screen is only that due to the undula- 

 tions which remain outstanding after the mutual destruc- 

 tion of by far the greater proportion of those propagated 

 from the zones (a), (c), (e), &c., (the odd zones, reckon- 

 ing (a) as No. 1), by those emanating from tlie even 

 ones (b), (d), (f), &c., it follows that if all the even zones 

 could be entirely suppressed or rendered ineffective, the 

 illumination at p would be prodigiously increased, and 

 that even the obliteration of a few of them would pro- 

 duce a very material augmentation of brightness at that 

 point. In other words, that by stopping out a large pro- 

 portion of the luminous rays passing through a circular 

 aperture from a bright illuminating point, the illumina- 

 tion of the central point of the image of such aperture 

 thrown on a screen at a certain distance behind it, may 

 be made to exceed by many times what it would he were 

 the whole aperture left open. This strangely paradoxical 

 result is stated by M. Billet* to have been experiment- 



* Billet, Traitc. d'optique Physique, 185S, ii. 55, by far the 

 fullest resume of that subject hitherto published ; only too little ex- 

 planatory, and sadly deficient in facility of reference. Jt deserves a 

 good index. 



