ON THE NATURE OF LIGHT AND COLOURS. 467 



colour in the usual manner; thus when the surface is so situated, that the 

 image of the luminous point would be seen in it by regular reflection, the 

 difference will vanish, and the light will remain perfectly white, but in other 

 cases various colours will appear, according to the degree of obliquity. 

 These colours may easily be seen, in an irregular form, by looking at any 

 metal, coarsely polished, in the sunshine; but they become more distinct 

 and conspicuous, when a number of fine lines of equal strength are drawn 

 parallel to each other, so as to conspire in their effects. 



It sometimes happens that an object, of which a shadow is formed in abeam 

 of light, admitted through a small aperture, is not terminated by parallel sides; 

 thus the two portions of light, which are diffracted from two sides of an 

 object, at right angles with each other, frequently form a short series of curv- 

 ed fringes within the shadow, situated on each side of the diagonal, which 

 were first observed by Grimaldi, and which are completely explicable from 

 the general principle, of the interference of the two portions encroaching 

 perpendicularly on the shadow. (Plate XXX. Fig. 445.) 



But the most obvious of all the appearances of this kind is that of the 

 fringes, which are usually seen beyond the termination of any shadow, formed 

 in a beam of light, admitted through a small aperture: in white light three 

 of these fringes are usually visible, and sometimes four; but in light of one 

 colour only, their number is greater ; and they are always much narrower as 

 they are remoter from the shadow. Their origin is easily deduced from the 

 interference of the direct light with a portion of light reflected from the 

 margin of the object which produces them, the obliquity of its incidence 

 causing a reflection so copious as to exhibit a visible effect, however nar- 

 row that margin may be; the fringes are, however, rendered more obvious 

 as the quantity of this reflected light is greater. Upon this theory it follows 

 that the distance of the first dark fringe from the shadow should be half as 

 great as that of the fourth, the difference of the lengths of the different paths 

 of the light being as the squares of those distances; and the experiment 

 precisely confirms this calculation, with the same slight correction only as 

 is required in all other cases; the distances of the first fringes being always a 

 little increased. It may also be observed, that the extent of the shadow itself 

 is always augmented, and nearly in an equal degre^ with that of the fringes : the 



