_Mrs. B._ This can be done by letting the coloured rays, which have been separated by a prism, fall upon a lens, which will converge them to a focus; and if, when thus reunited, we find that they appear white as they did before refraction, I hope you will be convinced that the white rays, are a compound of the several coloured rays. The prism P, you see, (fig. 5.) separates a ray of white light, into seven coloured rays, and the lens L L brings them to a focus at F, where they again appear white. _Caroline._ You succeed to perfection: this is indeed a most interesting and conclusive experiment. _Emily._ Yet, Mrs. B., I cannot help thinking, that there may, perhaps, be but three distinct colours in the spectrum, red, yellow, and blue; and that the four others may consist of two of these colours blended together; for, in painting, we find, that by mixing red and yellow, we produce orange; with different proportions of red and blue, we make violet or any shade of purple; and yellow, and blue, form green. Now, it is very natural to suppose, that the refraction of a prism, may not be so perfect as to separate the coloured rays of light completely, and that those which are contiguous, in order of refrangibility, may encroach on each other, and by mixing, produce the intermediate colours, orange, green, violet, and indigo. _Mrs. B._ Your observation is, I believe, neither quite wrong, nor quite right. Dr. Wollaston, who has performed many experiments on the refraction of light, in a more accurate manner than had been previously done, by receiving a very narrow line of light on a prism, found that it formed a spectrum, consisting of rays of four colours only; but they were not exactly those you have named as primitive colours, for they consisted of red, green, blue, and violet. A very narrow line of yellow was visible, at the limit of the red and green, which Dr. Wollaston attributed to the overlapping of the edges of the red and green light. _Caroline._ But red and green mixed together, do not produce yellow? _Mrs. B._ Not in painting; but it may be so in the primitive rays of the spectrum. Dr. Wollaston observed, that, by increasing the breadth of the aperture, by which the line of light was admitted, the space occupied by each coloured ray in the spectrum, was augmented, in proportion as each portion encroached on the neighbouring colour, and mixed with it; so that the intervention of orange and yellow, between the red and green, is owing, he supposes, to the mixture of these two colours; and the blue is blended on the one side with the green, and on the other with the violet, forming the spectrum, as it was originally observed by Sir Isaac Newton, and which I have just shown you.