470 LECTUUE XXXIX. 



but difficult to believe with respect to any of their arrangements constituting 

 the diversities of material substances. 



The colours of mixed plates constitute a distinct variety of the colours of 

 thin plates, which has not been commonly observed. They appear when the 

 interstice between two glasses, nearly in contact, is filled with a great num- 

 ber of minute portions of two different substances, as water and air, oil and 

 air, or oil and water: the light, which passes through one of the mediums, 

 moving with a . greater velocity, anticipates the light passing through the 

 other; and their effects on the eye being confounded and combined, their 

 interference produces an appearance of colours nearly similar to those of the 

 colours of simple thin plates, seen by transmission; but at much greater thick- 

 nesses, depending on the difference of the refractive densities of the sub- 

 stances employed. The effect is observed by holding the glasses between the 

 eye and the termination of a bright object, and it is most conspicuous in the 

 portion which is seen on the dark part beyond the object, being produced by the 

 light scattered irregularly from the surfaces of the fluid. Here, however, the 

 effects are inverted, the colours resembling those of the common thin plates, 

 seen by reflection ; and the same considerations on the nature of the reflec- ■ 

 tions are applicable to both cases. (Plate XXX. Fig. 450.) 



The production of the supernumerary rainbows, which are sometimes seen 

 within the primary and without the secondary bow, appears to be intimately 

 connected with that of the colours of thin plates. We have already seen 

 that the light producing the ordinary rainbow is double, its intensity being 

 only greatest at its termination, where the common bow appears, while the 

 whole light is extended much more widely. The two portions concerned in 

 its production must divide this light into fringes; but unless almost all the 

 drops of a shower happen to be of the same magnitude, the effects of these 

 fringes must be confounded and destroyed: in general, howeyer, they must 

 at least cooperate more or less in producing one dark fringe, which must 

 cut off the common rainbow much more abruptly than it would otherwise 

 have been terminated, and consequently assist the distinctness of its colours. 

 The magnitude of the drops of rain, required for producing such of these rain- 

 bows as are usually observed,- is between the 50th and the 100th of an inch: 

 they become gradually narrower as they are more remote from the common 



