A prism A B C, (fig. 14, Plate IT.) placed in an horizontal position, would project the ray into an oblong form, as has been seen ; apply another horizontal prism, A D B, similar to the former, to receive the refracted light emerging from the first, and having its refracting angle turned the contrary way from that of the former. The light, after passing through both prisms, will assume a circular form, as if it had not been at all refracted.
If the light emerging from the first prism be received by a second, whose axis is perpendicular to that of the for mer, it will be refracted by this trans verse prism into a position inclined to the former, the red extremity being least, and the violet most removed from its for mer position ; but it will not be at all al tered in breadth.
Close to the prism A (fig. 15), place a perforated board, a b, and let the refract ed light (having passed through the small hole) be received on a second board, c parallel to the first, and perforated in like manner ; behind that hole in the second board place a prism, with its refracting angle downward, turn the first prism slowly about its axis, and the light will move up and down the second board ; let the colours be transmitted successively, and mark the places of the different co loured rays on the wall after their refrac tion by the second prism, the red will ap pear lowest, the violet highest, the rest in the intermediate places in order. Here then the light being very much simpli fied, and the incidences of all the rays on the second prism exactly the same ; the red was least refracted, the violet most, &c.
The permanency of these original co lours appears from hence, that they suf fer no manner of change by any number of refractions, as is evident from the last mentioned experiment ; nor yet by re flection ; for if any coloured body be placed in simplified homogeneous light, it will always appear of the same colour of the light in which it is placed, whether that differ from the colour of the body or not ; e. g. if ultra marine and vermilion be placed in a red light, both will appear red ; in a green light, green ; in a blue light, blue, &c. It is, however, to be al lowed, that a body appears brighter when in a light of its own colour than in another; and from this we see that the colours of natural bodies arise from an aptitude in them to reflect some rays more copious ly and strongly than others ; but lest this phenomenon should produce a doubt of the constancy of the primary colours, it is proper to assign the reason of it, which is this ; that when placed in its own co loured light, the body reflects the rays of the predominant colour more strongly than any of those intermixed with it ; therefore the proportion of the rays of the predominant colour to those of the others, in the reflected light, will be greater than in the incident light ; but when the body is placed in a light of a different colour from its own, for a similar reason the contrary effect will follow, i. e.
the proportion of the predominant colour to the others will be less in the reflected than in the incident light, and therefore as its splendour would be greater in the former case, and would be less in the lat ter than if all the rays were equally re flected, the splendour of the predominant colour will be much greater in the former case than in the latter.
As a solar ray was separated into seve ral others of different colours, so, on the contrary, from these homogeneous rays a ray of heterogeneous light may be com pounded, perfectly corresponding both in appearance and properties with the solar rays.
The coloured rays (fig 16) diverging from the prism are received by a double convex lens, at the distance of twice its focal length from the hole ; at the same distance 'behind the lens, where they are collected by its refraction, they are re ceived on a second prism, whose refract ing angle is equal to that of the former ; the divergence of the, homogeneous rays that would otherwise ensue, is counter acted by the second prism, and they are made to proceed parallel to each other from the place of their intersection, and therefbre are all compounded and mixed together in the emergent ray A B, which is exactly of the same appearance with the solar rays, and by experiments made on it similar to those usually made in solar light, is found to possess the same properties.