4fi8 LECTURE XXXIX. 



reason of this circumstance appears to be the gradual loss of light at the 

 edges of every separate beam, which is so strongly analogous to the phenomena 

 visible in waves of water. The same cause may also perhaps have some effect 

 in producing the general modification or correction of the place of the first 

 fringes, although it appears to be scarcely sufficient for explaining the whole 

 of it. (Plate XXX. Fig. 446.) 



A still more common and convenient method, of exhibiting the effects of 

 the mutual interference of light, is afforded us by the colours of the thin 

 plates of transparent substances. The lights are here derived from the suc- 

 cessive partial reflections produced by the upper and under surface of the plate, 

 or when the plate is viewed by transmitted light, from the direct beam 

 which is simply refracted, and that portion of it which is twice reflected 

 within the plate. The appearance in the latter case is much less striking 

 than in the former, because the light thus affected is only a small portion of 

 the whole beam, with which it is mixed; while in the former the two reflected 

 portions are nearly of equal intensity, and may be separated from all other 

 light tending to overpower them. In both cases, when the plate is gradually 

 reduced in thickness to an extremely thin edge, the order of colours may 

 be precisely the same as in the stripes and coronae already described; their 

 distance only varying when the surfaces of the plate, instead of being plane, 

 are concave, as it frequently happens in such experiments. The scale of an 

 oxid, which is often formed by the effect of heat on the surface of a metal, 

 in particular of iron, affords us an example of such a series formed in reflect- 

 ed light; this scale is at first inconceivably thin, and destroys none of the 

 light reflected, it soon, however, begins to be of a dull yellow, Avhich 

 changes to red, and then to crimson and blue, after which the effect is de- 

 stroyed by the opacity which the oxid acquires. Usually, however, the 

 series of colours produced in reflected light follows an order somewhat dif- 

 ferent: the scale of oxid is denser than the air, and the iron below than the 

 oxid; but where the mediums above and below the plate are either both 

 rarer or both denser than itself, the different natures of the reflections at its 

 different surfaces appear to produce a modification in the state of the undu- 

 lations, and the infinitely thin edge of the plate becomes black instead of 

 white, one of the portions of light at once destroying the other, instead of 

 cooperating with it. Thus when a film of soapy water is stretched over a 



