The two great doctrines of the different refrangibility of the rays of light, and of the composition of white light, by mixing all the rays of the spectrum, having been established by Newton on an impregnable basis, we come now to describe some of the other results which he obtained regarding the prismatic spectrum and its colours, to point out the errors which he committed, to shew the influence which they had on the progress of optics, and to give an account of the remarkable discoveries which have been made in this branch of science during the last and the present century.

There are few facts in the history of optics more sin <110> gular than that Newton should have believed that all bodies when shaped into prisms produced prismatic spectra of equal length, or separated, or dispersed the red and violet rays to equal distances, when the mean refraction, or the refraction of the middle ray of the spectrum was the same. This opinion, which he deduced from no direct experiments, and into which no theoretical views could have led him, seems to have been impressed on his mind with all the force of an axiom. In one of his experiments he had occasion to counteract the refraction of a prism of glass by a prism of water; and had he completed the experiment, and studied the result of it when the mean refraction of the two prisms was the same, he could not have failed to observe that the prism of water did not correct the colour of the prism of glass, and would have thus been led to one of the most important truths in optics, — that different bodies have different dispersive powers, or produce prismatic spectra of different lengths, when their mean refraction is the same. It is curious to observe, as happened in this experiment, what trifling circumstances often arrest the philosopher when on the very verge of a discovery. Newton had mixed with the water which he used in his prism a little sugar of lead, in order to increase the refractive power of the water; but the sugar of lead having a higher dispersive power than water, made the dispersive power of the water prism equal to that of the prism of glass; so that if Newton had completed the experiment, the use of the sugar of lead would have prevented him from making an important discovery, which was almost in his possession. Had he, on the contrary, increased the angle of his water prism till it produced the same deviation of the mean ray of the <111> spectrum, he would have found that the one prism did not correct the colour of the other, and that the glass had a greater dispersive power than the water, and gave a longer spectrum.

Nor is it less extraordinary that the same discovery escaped from his grasp during his controversy with Lucas. When the Dutch philosopher and his numerous friends who saw his experiments, pronounced his spectrum to be only 312 times its breadth, Newton found it to be at least five times its breadth: it is strange neither party ever thought that this might arise from using different kinds of glass, and never made the least inquiry regarding the material of which their prism was made. It is highly probable that Lucas's prism had a very low dispersive power, which would account for the great difference between his spectra and those of Newton, but whether this was the case or not, Newton, under the blind conviction that all spectra must, cæteris paribus, be of equal length, pronounced "the improvement of telescopes by refractions to be desperate,"[1] and thus checked for a long time the progress of this branch of science.

About two years after the death of Sir Isaac, an individual unknown to fame, broke the spell in which the subject of the spectrum had so long been bound. In the <112> year 1729, Mr. Chester More Hall, of More Hall in Essex, while studying the mechanism of the human eye, was led to suppose that telescopes might be improved by forming their object-glass with two lenses of different refractive powers. He published no memoir on the subject, and has not even left behind him any record of the steps by which he arrived at such a conclusion. It is probable that he may have adopted David Gregory's idea of combining lenses of different density, and as crown and flint glass differed most in this respect, that in combining them he discovered the great difference in their dispersive powers, and was thus led to the invention of the achromatic telescope. Mr. Hall employed working opticians to grind his lenses, and furnished them with the proper radii of their surfaces for correcting the colour arising from the difference of refrangibility in the rays, and the aberration occasioned by the spherical figure of the lenses. Mr. Bass, a well-known working optician, was one of his assistants, and it was probably through him that the knowledge of Mr. Hall's invention has been preserved. About the year 1733 he had completed several achromatic object-glasses, which bore an aperture of more than 212 inches, though their focal length did not exceed twenty inches. One of these telescopes, which in 1798 was in the possession of the Rev. Mr. Smith of Charlotte Street, Rathbone Place, was examined by several gentlemen of scientific eminence, and found to be a genuine achromatic telescope.

Many years after the death of Mr. More Hall, Mr. John Dollond and others had turned their attention to the improvement of telescopes. Euler, believing the eye to be achromatic, had endeavoured, but in vain, to discover a combination of media, by which the object-glasses of <113> telescopes could give colourless images. Klingenstierna had endeavoured to shew that refraction without colour might be produced according to the laws of refraction laid down by Newton himself; but none of these philosophers made a single step towards the great discovery which was made by Mr. Dollond, when the previous labours of Hall were unpublished. In 1758, he communicated to the Royal Society an account of his experiments on the different refrangibility of light. In this valuable paper, he proved that glass had a greater dispersive power than water, and attempted to make achromatic object-glasses by inclosing water between two lenses of glass. In this attempt he found the spherical aberration difficult to correct, and he was therefore led to try crown and flint glass, which he found to have such different dispersive powers, that he was at once able to make achromatic object-glasses. In order to secure his right to this invention, Dollond took out a patent; but in consequence of its having been discovered that the same invention had been made before, some of the London opticians tried the question at law, and produced in court the telescope of Mr. Hall. It was in vain to deny the prior claims of Mr. Hall; but as it was certain that Dollond was unacquainted with his labours, and as no achromatic telescope had ever been exposed to sale, Lord Mansfield justly decided the case in favour of Dollond. [2]

It is not easy to explain to the general reader the principle of the Achromatic Telescope; but we think it may be apprehended from an inspection of the annexed diagram. In crown glass the index of refraction is 1.526 for red rays, and 1.547 for violet rays. If LL then be aconvex lens of crown glass, it will refract the violet rays <114> more than the red, the former in the direction L R, and the latter in the direction L V, so that R will be the focus of red, and V that of the violet rays. If we now place behind it a concave lens C C of the same kind of glass and the same curvature, it will by its opposite and equal refractions unite again the rays L R, L V, in the direction L l , so as to form a white ray; but in this case the compound lens acts like a piece of plane glass, or rather like Fig. 9. a watch glass which has no focus. But if we make the concave lens C C of flint glass with less curvature than L L, then since it has a greater refractive and dispersive power than the lens L L of crown glass, it will, notwithstanding its inferior curvature, unite the rays L R, L V, and leave such a balance of refraction in favour of the lens L L, that the rays will be united, and a colourless image formed, at O, so that the double object-glass L L C C Will be an achromatic one.

If the prismatic spectrum formed by crown and flint glass had been exactly the same, that is, if the coloured spaces in each were of the same length, telescopes constructed upon the preceding principle would have been perfect, in so far as colour is concerned; but this is not the case, and consequently in the very best achromatic <115> telescopes, there is left what has been called a secondary spectrum, consisting of green and purple colours, which appear on the border of the images of all luminous objects.

This secondary or residual spectrum, arising from what has been called the irrationality of the coloured spaces in the two equal spectra of crown and flint glass, may be corrected by an ingenious contrivance discovered by Dr. Blair. He found that muriatic acid produced a prismatic spectrum, in which the coloured spaces were nearly the same as in crown glass, and that he could increase its low refractive and dispersive power, by mixing it with metallic solutions, so as to fit it for being used like flint glass for correcting the colour of the crown glass without balancing its refraction. This increase in its refractive and dispersive powers, did not alter the proportion of the coloured spaces in its spectrum, so that it was capable of giving a perfectly colourless image, when placed as a concave lens between two convex ones of crown glass. The metallic solution used by Dr. Blair was muriate of antimony, and in the lens which he constructed, the rays of different colours were bent from their rectilineal course with the same equality and regularity as in reflexion. To this telescope he gave the name of Aplanatic. According to the testimony of Professor Robison, those which he examined surpassed greatly the best ordinary achromatic telescopes; but they have been found difficult to construct, and in so far as we know, there is not in existence a single aplanatic telescope.

The Achromatic Telescope, on the contrary, even With the imperfection of its secondary spectrum, has undergone great improvements, and promises to rival Reflectors in excellence and power. By the labours of Guinand, Fraun <116> hofer, and M. Bontemps, discs of flint glass of 12, 15, 24, and even 29 inches in diameter, have been made, and we hope soon to see the largest of them converted into a magnificent telescope. The disc of 24 inches has been converted into a telescope by the Rev. Mr. Craig of Leamington.[3]

But while Newton overlooked the remarkable property of the prismatic spectrum, on which the improvement of Refracting Telescopes depends, he committed other considerable mistakes in his examination of the spectrum. It does not seem to have occurred to him that the Solar Spectrum was not the spectrum from which the properties of the sun's rays ought to be deduced, and that the relations of the coloured spaces must depend on the angular magnitude of the luminous body, or of the aperture from which the spectrum is obtained. Misled by an apparent analogy between the length of the coloured spaces and the divisions of a musical chord,[4] which he ascertained "by an assistant whose eyes were more critical than his own," he adopted that division as representing the proportion of the coloured spaces in every dispersed beam of light. Had he studied the prismatic spectrum in Mercury and Jupiter by the same instruments, he would have obtained quite different results. In Mercury, where the sun's apparent magnitude is very large, he would have seen a spectrum without any green, and having red, orange, and yellow at one end, white in the middle, and blue and violet at the other end. In Jupiter, on the contrary, he would have obtained a spectrum in which the coloured spaces were much more condensed, and the pure colours more separated. The Solar spectrum <117> described by Newton, has an intermediate character between these two extremes, and had he examined it under the same circumstances in winter and in summer, he would have found the analysis of the beams more perfect in summer, on account of the sun's diameter being less. We are entitled, therefore, to assert, that neither the number nor the extent, nor the limits of the coloured spaces, as given by Newton, are those which belong to the true prismatic spectrum.

Had Newton received upon his prism a beam of light transmitted through a very narrow aperture, he would have anticipated Wollaston and Fraunhofer in their fine discovery of the lines in the prismatic spectrum. In 1802 Dr. Wollaston, by transmitting the light of the sky through an aperture the twentieth of an inch wide, discovered six fixed dark lines in the spectrum, one in the red, one in the orange, one in the blue, and one in the violet spaces. Without knowing of Wollaston's observations, the late celebrated M. Fraunhofer of Munich, discovered in sun light, nearly 600 lines, the largest of which subtended an angle of from 5″ to 10″. We have found this angle to increase enormously by atmospherical absorption, as the sun passes from the meridian to the horizon, and in a long series of observations we have observed upwards of two thousand lines in the prismatic spectrum formed from the sun's rays.

From his analysis of the Solar spectrum, by examining with the prism its separate colours, Newton concluded, that to the same degree of refrangibility ever belonged the same colour, and to the same colour ever belonged the same refrangibility, and hence he inferred that red, orange, yellow, green, blue, indigo, and violet, were primary and simple colours. This proposition is true in so far as the analysis <118> of the spectrum by the prism is concerned; but we have found another species of analysis, by which the colours of the spectrum may be decomposed. Though we cannot separate the green rays in the spectrum into yellow and blue by the refraction of prisms, yet if we possessed any solid or fluid which had a specific attraction for blue rays, that is, which absorbed them during the passage of the green light through the medium, and allowed the yellow rays to pass, we should then analyze the green into its component elements as effectually as if we separated them by the prism. We have in this way subjected the colours in the spectrum to the analysis of a great variety of solid and fluid bodies of different colours, and we have found that in every part of the spectrum, the colours are more or less changed or decomposed by absorption.

The simplest way of observing these changes is to receive the spectrum in the eye by looking through the prism at a narrow line of light from the sky. If we now interpose between the eye and the prism a plate of purplish blue glass, about the twentieth of an inch thick, we shall see the prismatic spectrum with its bright colours completely metamorphosed. The red part of the spectrum is divided into two red spaces, separated by a dark interval. Next to the inner red space comes a space of bright yellow, separated from the red by a visible interval. After the yellow comes the green, with an obscure space between them, then follow the blue and the violet, the last of which has suffered little or no diminution. Now, in this experiment, the blue glass has absorbed the red rays which, when mixed with the yellow, on one side constituted orange, and the blue rays which, when mixed with the yellow on the other side, constituted green, so that the, insulation of the yellow rays thus effected, and the disap <119> pearance of the orange and of the greater part of the green light, places it beyond a doubt that the orange and green colours in this spectrum are component colours, the former consisting of red and yellow, and the latter of yellow and blue rays of the very same refrangibility. If we compare the two red spaces seen through the blue glass, with the red spaces seen without the blue glass, it will appear that the red has experienced such an alteration in its tint by the action of the blue glass, as would be effected by the absorption of a small portion of yellow light; and hence we conclude that the red of this spectrum contains a slight tinge of yellow, and that the yellow space extends over more than one half of the spectrum, including the red, orange, yellow, green, and blue spaces.

By varying the absorptive media, I have found that red light exists in the yellow space, and we have ocular evidence, that in the violet space red light is combined with the blue rays. From these and other facts, which it would be out of place here to enumerate, I have been led to the conclusion that the prismatic spectrum consists of three different spectra, viz., red., yellow, and blue, all having the same length, all superposed, and each having its maximum intensity at the point where it predominates in the combined spectrum. Hence it follows: —

1. That red, yellow, and blue rays of the same refrangibility exist at every point of the spectrum of intensities, represented by the ordinates of the curve of intensity in each separate spectrum.

2. That the colour of the spectrum at any one point will be that of the predominant ray modified by the smaller quantities of the other two rays; and,

3. That if we could absorb the two predominant at any one point of the spectrum, in such quantities as <120> when mixed with the remaining or unabsorbed ray, would make white light, we should be able to insulate white light indecomposable by the prism.

This view of the structure of the spectrum will be understood from the annexed diagrams, where Figs. 10, 11, and 12, represent the three separate spectra, which are shewn in their combined state in Fig. 13. In all these figures, the point M is the red or least refrangible extremity of the spectrum, and N the violet or most refrangible extremity. The maximum intensity of each spectrum is opposite R, Y, and B, the intensity diminishing to nothing at the extremities M and N. When these three spectra are superposed, they will exhibit the colours shewn in Fig. 13, in which we have inserted the three curves which represent the intensities in each spectrum.

In order to explain how the seven colours, observed by Newton, are produced by the three primitive colours, we shall take the case of the orange, as shewn in Fig. 13, where the three ordinates ax, bx, cx, will indicate the relative intensities of the three colours, combined at any point x of the spectrum. Thus let

The ordinate for red light be ax = 30
"yellow bx = 16
"blue cx = 2
Then ax + bx + cx = 48 rays.

Hence the point x will be illuminated with forty-eight rays, namely, thirty of red, sixteen of yellow, and two of blue light. Now, as there must be certain quantities of red and yellow light, which, when combined with two blue rays, will form white, let us suppose that white light, whose intensity is ten, will be formed by three red, five <121> Fig. 10. Fig. 11. Fig. 12. Fig. 13. <122> yellow, and two blue rays, then it follows that the point x will be illuminated with

Red rays,30 - 3 or 27 rays
Yellow rays,16 - 5 or 11 "
White light + 3 red + 5 yellow + 2 blue,or 10 "
Orange = red + yellow + white,= 48 rays.

That is, the point x will have the colour of orange rendered brighter by a mixture of white light. The blue rays consequently which exist at x will not communicate any blue tinge to the prevailing orange.

In submitting to the scientific world this new analysis of light, by absorption, we were fully aware of the difficulties which we had to encounter, and we anticipated the opposition which would be made to it. "Even in physical science," we said,[5] "it is an arduous task to unsettle long-established and deeply-rooted opinions; and the task becomes Herculean when these opinions are intrenched in national feeling, and associated with immortal names. There are cases, indeed, where the simple exhibition of new truths is sufficient to dispel errors the most deeply cherished, and the most venerable from their antiquity; but it is otherwise with doctrines which depend on a chain of reasoning where every step in the inductive process is not rigorously demonstrative; and of this we require no other proof than is to be found in the history of Newton's optical discoveries, and particularly in the opposition they experienced from such distinguished men as Dr. Hooke and Mr. Huygens."

The preceding analysis of the spectrum embraces three propositions, which, to a certain extent, are independent of each other.

1. That the colours of the coloured spaces may be <123> changed by absorbing media, acting by reflexions and transmissions

2. That in pure spectra, white light can be insulated.

3. That the Newtonian spectrum of seven colours consists of three primary spectra, red, yellow, and blue superposed, having their intensity of illumination and maximum at different points, and shading to nothing at their extremities.

The first of these propositions may be true, even though we could not insulate white light at any point of the spectrum; and both the first and second may be true, without our being able to demonstrate that the three spectra have the same length, and diminish in intensity from their maxima to their extremities.

The general proposition that the colours of the spectrum are changed by absorption, has been questioned by classes of critics, — by Mr. Airy,[6] M. Melloni,[7] and Mr. Draper,[8] who have never repeated our experiments, but made some very imperfect ones of their own; — by Dr. Whewell,[9] and the Abbé Moigno,[10] who have made no experiments at all; — and by M. Helmholtz[11] in Prussia, and M. Bernard[12] in France. We have replied to the three of these writers, and shall now make a few observations on the results obtained by MM. Helmholtz and Bernard.

M. Helmholtz has candidly stated, in contradiction of Airy, that "the changes of colour" which we have <124> described, as produced by absorption, "are for the most part sufficiently striking to be observed without difficulty ;" and he adds, that "a careful repetition of at least the most important of my experiments, carried out in exact accordance with my method, and with every precaution hitherto deemed necessary, has indeed taught me that the facts which he affirms to have observed, are described with perfect accuracy."

The change of colour, thus admitted as a physical fact, M. Helmholtz ascribes to two causes: —

1. To the possible admixture of rays scattered from the prism, and the other transparent bodies used in the experiment; and

2. To the mixture of complementary colours produced by the action of the other colours of the spectrum on the retina.

The first of these, as M. Helmholtz almost admits, is wholly influential, and the second, if it does disturb the colorific impressions on retinæ tender and sensitive, had no such effect on ours.

If the subjective perception of colour, when we view the spectrum, or make experiments on which more than one colour reaches the eye, is capable of masking the colours under examination, then all that has been written on colours, thus seen, must be erroneous, and all the gay tints of art or of nature are but false hues under the metamorphosis of a subjective perception. We must not now pronounce a rose to be red, and its leaves green, till we have stared at them through a chink, or torn them from their foot-stalk ! The phenomena of absorption which we have described we have seen, just as Newton saw his seven colours in the spectrum, and Hooke his composite tints in the soap-bubble; and now that our eyes have <125> nearly finished their work, we are not disposed to mistrust, without reason, such good and faithful servants.[13] The observations of M. Bernard, who has repeated only a few of our experiments, differ very little from those of M. Helmholtz. He maintains that the conversion of the blue space into violet arises from the light being diminished. If the colours of the spectrum thus change, as he maintains, by their becoming fainter, we would desire to ask at what degree of illumination are we to see the spectrum in its true colours? Colour cannot depend upon refrangibility, if the blue space is converted into violet either by diminution of light or absorption; and therefore the doctrine of M. Bernard is as fatal to Newton's as to ours. If M. Bernard's experiment be correct, it only proves that the blue rays, when enfeebled, lose their power over the retina sooner than the red.

The Newtonian doctrine, "that the degree of refrangibility proper to any particular sort of rays is not mutable by refraction, nor reflection, nor by any other cause,"[14] has been recently questioned by Professor Stokes, one of the distinguished successors of Newton in the Lucasian chair. Mr. Stokes[15] found that the chemical rays in the violet space, between the lines G and H of the spectrum, produce, in a solution of sulphate of quinine, light of a sky-blue colour, which he assumes to have the refrangi <126> bility of that portion of the spectrum. By refracting this light through a prism, he converts the sky-blue rays into a spectrum of all colours, and all refrangibilities. Hence he concludes, that the sky-blue light having the fixed refrangibility due to its locality between G and H, is changed by refraction into all the other colours, with their respective refrangibilities. If this conclusion be admitted, our doctrine of the severance of colour and refrangibility is placed beyond a doubt. We have in the first experiment sky-blue light with the refrangibility of violet light between G and H; and, in the second experiment, we have the same blue light changed by refraction into all the colours of the spectrum.

We cannot, however, avail ourselves of this last fact, for, after a careful consideration of Mr. Stokes's important results,[16] we cannot but regard the sky-blue light as a phosphorescence, produced in the quinine solution by the chemical rays, which, like all other phosphorescences, is decomposable by the prism.

[1] Optics, Prop. vii., Book ii., p. 91. In his reply to Hooke, who justly "reprehended him for laying aside the thoughts of improving optics by refractions," he seems to modify his opinion by saying that he tried what might be done "by two or more glasses or crystals, with water or some other fluid between them." "But what the results by theory or by trials have been, he might possibly find a more proper occasion to declare." This was written in 1672, and we can therefore say with certainty that he failed in this attempt, as it was in 1684 that he pronounced the case to be desperate. It is a curious circumstance that David Gregory, in his Lectures delivered in Edinburgh in 1684, suggests that, in imitation of the human eye, the {object-glasses} of telescopes might be composed of media of different density. In Brown's translation of Gregory, the sense of the passage is not brought out. See Gregory's Catoptries, Prop. xxiv., Schol., pp. 110, 111.

[2] See Tilloch's Philosophical Magazine, Nov. 1798, vol. ii. p. 177{,}

[3] See my Treatise on Optics, new edit., p. 506.

[4] Optics, Part, ii., Prop. iii., p. 110.

[5] Edinburgh Transactions, 1831, xii. p. 124.

[6] Phil. Mag. vol. xxx. p. 73.

[7] Bibl. Univers. Août 1847.

[8] Silliman'sJournal, vol. iv. p. 388. 1847.

[9] Hist. of Inductive Sciences, vol. ii. p. 361; and Edinburgh Review, vol. lxvi. p. 136, and vol. lxxiv. p. 288.

[10] Répetoire d'Optique, tom. ii. p. 459.

[11] Poggendorff's Annalen. 1852, No. 8.

[12] Ann. de Chim. et de Phys. tom xxxv. p. 385 &c.

[13] The changes of colour in the spectrum at different seasons of the year, and the different hours of the day, and when formed from different portions of the illuminated sky, as well as from the direct light of the sun, are very remarkable. We have mentioned one or two of them in the Edinburgh Review, vol. lxxiv. p. 284, Jan. 1842. One of these observations is as follows: — "October 23, 1832. 11th, The yellow comes distinctly up to F, and a little beyond it; i.e., the blue has been all absorbed in the green space of Fraunhofer's spectrum from E to F." In another observation on the 5th February 1833, the green space was wholly yellow.

[14] Letter to Oldenburg, Feb. 6, 1672, in Phil. Trans. No. 80, p. 3081, § 3.

[15] Phil. Trans. 1852.

[16] See my Treatise on Optics, new edition, pp. 182, 183.

© 2018 The Newton Project

Professor Rob Iliffe
Director, AHRC Newton Papers Project

Scott Mandelbrote,
Fellow & Perne librarian, Peterhouse, Cambridge

Faculty of History, George Street, Oxford, OX1 2RL - newtonproject@history.ox.ac.uk

Privacy Statement

  • University of Oxford
  • Arts and Humanities Research Council
  • JISC