Had the objects of nature been rendered visible only by white light, and exercised upon it the same action in refracting and reflecting it to the human eye, all the combinations in the material world, and all the various forms of life, would have displayed no other tint than that which they exhibit in a pencil sketch, a China-ink drawing, or a photographic picture. The magnificent foliage of the vegetable world might have filled the eye with its picturesque and lovely forms, and given protection to its fruit and its flowers, but we should not have rejoiced in the verdure of its youth, nor mourned over the yellow of its age. The sober mantle of twilight would have replaced the golden vesture of the rising and the setting sun. The <176> stars would have twinkled colourless in a grey sky, and the rainbow would have dwindled into a narrow arch of dusky light. The diamond, the ruby, and the sapphire might have displayed to science the nice geometry of their forms, and yielded to the arts their adamantine virtues; but they would have ceased to sparkle in the chaplet of beauty, or adorn the diadem of princes. The human face divine might have expressed all the qualities of the mind, and beamed with all the affections of the heart; but the purple light of love would not have risen on the cheek of beauty, nor the hectic flush have heralded its decay. Life would have breathed and perished in its pale marble; and nature would have sprung and decayed in its russet brown. The material world, however, has been otherwise framed, and those exquisite models of organic and inorganic life, into which the great sculptor has chiselled the furniture of his terrestrial temple, have been enhanced by that ethereal beauty which the play of light and colour can alone impart.

Many attempts were made previous to the time of Newton, to explain the colours of natural bodies; but they all necessarily failed, while philosophers were ignorant of the true nature of colours themselves. In his earliest communications to the Royal Society, Newton had clearly indicated his views respecting the colours of natural bodies; and after showing " that they appear of divers colours, according as they are disposed to reflect most copiously the rays endued by these colours," he proceeds, in the last part of his " Discourse," read to the Royal Society on the 10th February, 1675-6, to consider " the constitution of bodies on which their colours depend." This curious subject continued to occupy the attention of Newton, and he enters upon it more fully in two different parts <177> of his Optics, where " by the discovered properties of light he explains the permanent colours of natural bodies,"[1] and points out the " analogy between such colours, and the colours of thin transparent plates."[2]

After shewing that all bodies, whatever were their colours, exhibited these colours best in white light, or in light which contained their peculiar colour, he proves by experiment, that when coloured bodies are illuminated with homogeneous red light, they appear red, with homogeneous blue light, blue, and so on, " their colours being most brisk and vivid under the influence of their own daylight colours." The leaf of a plant, for example, appeared green in the white light of day, because it had the property of reflecting green light in greater abundance than any other. When the leaf was placed in homogeneous red light, it no longer appeared green, because there were no green rays in the red; but it reflected red light in a small degree, because there were some red rays in the compound green, which it had the property of reflecting. If the leaf had originally reflected a pure homogeneous green, unmixed with red, and reflected no white light (as all leaves do) from its outer surface, it would have appeared quite black, in pure homogeneous red light, as this light does not contain a single ray which the leaf is capable of reflecting. Hence it follows that the colours of natural bodies are owing to the property which they possess of stopping or absorbing certain rays of white light, while they reflect or transmit to the eye the rest of the rays, of which white light is composed. The green leaf, for example, stops or absorbs the red, blue, and violet rays of the white light which falls upon it, and reflects and transmits only those which compose its peculiar green. <178> To this extent the views of Newton are demonstrable, and have been universally adopted; but when he attempts to determine the manner in which the colour of any body is insulated from the other colours which fall upon it, and in which these other colours are stopped or lost; or, in other words, the physical constitution of natural bodies by which these processes are effected, he enters the region of hypothesis and fails in bringing conviction to the mind. His theory, however, is grand and imposing, but standing as it does, and as we shall presently shew, on a perishable basis, it must soon be swept away in the progress of optical discovery.

The following are the principles on which this theory is founded.

1. Bodies that have the highest refractive powers, reflect the greatest quantity of light from their surfaces, and at the confines of equally refracting media there is no reflexion.

2. The least parts of almost all natural bodies are in some measure transparent.

3. Between the parts of opaque and coloured bodies, are many spaces or pores, either empty or filled with media of other densities.

4. The parts of bodies and their interstices or pores must not be less than of some definite bigness, to render them coloured.

5. The transparent parts of bodies, according to their several sizes, reflect rays of one colour, and transmit those of another on the same grounds that thin plates do reflect or transmit these rays.

6. The parts of bodies on which their colours depend, are denser than the medium which pervades their interstices.


7. The bigness of the component parts of natural bodies may be conjectured by their colours.

In illustration of the fifth, or leading proposition of the theory, Newton remarks, " that if a thinned or plated body, which being of an even thickness, appears all over of one uniform colour, should be slit into threads or broken into fragments of the same thickness with the plate, he sees no reason why every thread or fragment should not keep its colour, and by consequence why a heap of those threads or fragments should not constitute a mass or powder of the same colour which the plate exhibited before it was broken. And the parts of all natural bodies being like so many fragments of a plate, must on the same grounds exhibit the same colours."

In order to prove this, Newton proceeds to describe various kinds of colours, to which he considers the theory specially applicable ; but before we follow him in this investigation, we must endeavour to classify all the varieties of colours which are exhibited in the natural world.

Colours may be arranged into seven classes, in each of which the colour has a different origin.

1. Transparent coloured fluids, such as the juices obtained from the coloured parts of plants, and coloured solutions, whether natural or artificial. Transparent coloured solids, such as coloured minerals, glasses, powders, and vegetable tissues.

2. Oxidated films on metals — colours of precious and hydrophanous opal — of Labrador felspar — of the feathers of birds — of the wings, &c., of insects — of the scales of fishes — of the tapetum of animals — of the internal films of mother-of-pearl and various shells — and of decomposed glass.


3. Superficial colours of mother-of-pearl, striated and grooved surfaces, which can be communicated by pressure to other surfaces.

4. Opalescences, or colours dispersed from the particles of different solid and fluid and gaseous bodies, some of which are coloured, and others colourless. These colours appear in ice, in water, in the atmosphere — in fluorspar and several glasses — in solutions of sulphate of quinine, &c., and in the juices of plants and several oils.

5. At the surfaces of media of different dispersive powers, and in which the index of refraction is the same in each medium for certain rays, but different for all the rest.

6. The colours produced by heat, and during combustion.

7. The colours of metals.

The colours referred to in the first of these classes are represented by red and yellow wines — by the coloured fluids shewn in the windows of the apothecary — by the green leaves of plants — by the ruby, the cairngorm, the topaz, and the sapphire — and by the powders of cinnabar, red lead, ultramarine, sulphur, &c.

In all these bodies Newton supposes that the colour peculiar to each, namely, that which passes through its substance, is the tint reflected from the minute particles of which it is composed, the opposite or complementary tint, which is transmitted by the particles being lost within the body by a multitude of internal reflexions. In the ruby, for example, the particles are supposed to have such a size as to appear red by reflexion, and the green light which would be seen by transmission through a single particle, is supposed to be lost by repeated reflexions within the body composed of such particles. If we now examine the ruby, or any other coloured solid or fluid, we <181> shall find that neither red nor green light is reflected from any of its external surfaces, or any of its internal parts. The red light which characterizes the body is seen only by transmission through its substance. If we now analyze the red light by the prism, we shall find that it has not the composition of any of the red rings in Newton's scale of colours.

In the case of the ruby, which we have purposely selected, we are able to apply another test, and one which Newton himself authorizes, when he remarks that changes of colour may be produced by the swelling or shrinking of the tinging corpuscles. In subjecting the balas ruby to a high degree of heat, which must have had the effect of swelling the tinging corpuscles, I found that it became green, which, as the cooling advanced, gradually faded into brown, the ruby resuming its original brilliant red when it had returned to its former temperature. Berzelius observed an analogous fact in the spinelle of Ceylon and Aker, which became brown by heat, then black, and opaque as the heat increased. Upon returning to its former temperature, it passed through a fine chrome green before it recovered its red colour. Hence it is obvious that these colours and changes of transparency, which have no relation to those of thin plates, could not have arisen from the gradual swelling and subsequent shrinking of tinging corpuscles.

A still more striking proof of the want of analogy between the colours of natural bodies and those of thin plates, may be obtained from the prismatic analysis of certain colours in which Newton himself believed that analogy to exist. A green of the third order of colours is, as he observes, " constituted principally of original green, but not without a mixture of some blue and yellow," and <182> contains not a single ray of orange, red, indigo, or violet. He considers the green of all vegetables, to be a green of the third order, not only because this green is the purest and most intense in colour, but because when vegetables wither, some of them turn to a greenish yellow , and others to a more perfect yellow or orange, or perhaps to red, passing first through all these intermediate colours. " Now," he adds, " the green is without doubt one of the same orders with those colours into which it changeth, because the changes are gradual, and those colours, though usually not very full, yet are often too full and lively to be of the fourth order." These changes from green to red, he considers as " effected by the exhaling of the moisture which may leave the tinging corpuscles more dense, and something augmented by the accretion of the oily and earthy part of that moisture."

In order to put these opinions to the test of direct experiment, we examined the brilliant green juice extracted by alcohol from the leaves of twenty different plants, and also the same juice when taken from the leaves in then yellow, orange, and red state, and found that their composition had not the least resemblance to that of the colours of any order whatever, and least of all to those of the greens of the third order. The spectrum obtained from a sunbeam passing through these juices is one of singular beauty, divided by dark spaces into several coloured bands of unequal breadths, and possessing all the colours which ought not to exist in the green of the third order. When the green fluid thus analyzed has stood for three or four days it loses its bright green colour, and becomes of an olive green, which grows more and more of a brownish yellow, till it becomes almost colourless, a <183> series of changes which have no relation whatever to the effects that might be expected to arise from an increase or decrease in the density or size of the tinging corpuscles.

In some plants the green leaf decays in a different manner from that described by Newton. In place of becoming yellow, the green leaves of the privet become of a deep black violet, when they wither; a colour which has no resemblance whatever to any of the colours of thin plates. The fluid obtained from these violet leaves was of a deep red colour, — much deeper than that of the darkest port wine. It divided the red space of the spectrum into two red bands, absorbed the violet and blue spaces generally, and obliterated the middle of the green space. Its action was so different from that of the green juice, that the two tints had no resemblance to those of adjacent colours of the same order[3]

The pale blue of the sky is regarded by Newton as a blue of the first order, produced by the minute particles of " vapours which have not arrived at that grossness which is requisite to reflect other colours;" and while he considers the whiteness of froth, paper, linen, &c., as that which arises " from a mixture of the colours of several orders," that is, from the action of particles of a much greater size than those of vapours which produce the blue of the first order. Now, it is obvious that froth, when seen under a clear blue sky, must have the colour of the sky itself, as it is nothing more than an accumulation of images of the sky reflected from the innumerable aqueous vesicles which compose it. The colour of froth, wherever it is placed, must be the average tint of all the <184> differently coloured rays which fall upon it and are reflected to the eye.

The colours referred to in the second class are undoubtedly analogous to those of thin plates. Newton has himself mentioned the colours of the feathers of some birds as those of thin plates, and the fine colours of the diamond and other beetles obviously have the same origin. The splendid colours of the tapetum, or membrane behind the retina of animals, afford an interesting example of this class of colours. Even when the membrane has been taken out, it exhibits the most beautiful colours by reflexion, but it becomes absolutely black when dry. The colours, however, may be revived by moisture, and, after remaining in the dry state for upwards of twenty years, we have succeeded in restoring the colours by steeping the membrane in warm water. The black passes into a bright blue, the blue into green, and the green into greenish yellow.

In placing the internal colours of mother-of-pearl under this class, we must carefully distinguish them from the external colours communicable to wax. By reducing the mother-of-pearl to exceedingly thin plates, we are able to exhibit the action of the colorific films which they enclose, and which, like those of thin plates, give one colour by reflexion, and its complementary colour by transmission.[4]

The splendid colours exhibited by decomposed glass, both in the light which it reflects and transmits, belong also to colours of the second class; and though they are clearly those of thin plates, yet they exhibit peculiarities when produced by a great number of films, which place <185> them in a certain interesting relation with the colours of the first class.[5]

The colours of Labrador felspar, and of precious and hydrophanous opal, which we have shewn to be produced by thin plates and minute pores and tubes, belong also to the second class of colours.[6]

The superficial colours which we have placed in the third class, have obviously no relation whatever to the colours of thin plates. They are spectra produced by interference, and, had he been acquainted with them, they would have been regarded by Newton himself as inexplicable by his theory.[7]

The very remarkable colours produced by internal dispersion, and which have recently excited so much interest from the discoveries of Professor Stokes of Cambridge, form a fourth class, which has not been identified with those of thin plates. The light thus dispersed must be reflected, in cases of ordinary opalescence, from the faces of minute pores in solids, or from particles of different densities disseminated through solids, or suspended in fluids. The beautiful colours exhibited by fluor spar, by solutions of the sulphate of quinine, and various other solids and fluids, are emanations of a phosphoric nature, generated by certain rays in the solar spectrum, and have therefore no analogy with the colours of thin plates. These emanations have all colours, — red in the alcoholic juices of leaves, violet, blue, pink, and whitish in fluor spar, sky-blue in sulphate of quinine, bright green in alco <186> holic solutions of the colchicum autumnale, and in various glasses and oils, and violet in an alcoholic solution of guiacum.[8]

In the fifth class, we have placed a new species of colours which we discovered many years ago, and which we believe have never been studied, or even alluded to by any other person. In the year 1814, when investigating the law of polarisation for light reflected at the separating surface of different media, we had occasion to inclose oil of cassia between two flint-glass prisms, and were surprised to observe that the colour of the reflected light was blue. The cause of this we had some difficulty in discovering. The refractive power of oil of cassia exceeds greatly that of flint-glass for the mean rays of the spectrum, while the action of the two bodies on the less refrangible rays is nearly the same. Hence the red rays must be in a great measure transmitted, while there will be reflected a small portion of the orange, a greater portion of the yellow, and a much greater proportion of the blue and violet, so that the colour of the pencil, formed by reflexion, must necessarily be blue, mixed with some of the less refrangible rays.

By employing different kinds of glass, and different oils, we obtained various analogous results, in which rays of different colours were extinguished from the reflected pencil according to the part of the spectrum where an equilibrium had been established between the refractive powers of the media in contact. When the refractive indices were equal in the blue rays, the colour of the reflected pencil was yellow. As the indices of refraction <187> are the same for all obliquities of incidence,[9] the tint of the reflected pencil, though it must vary in intensity, can never vary in colour; and as that colour is abstracted from the white incident light, its complementary tint must appear, however faintly, in the transmitted pencil. Hence it follows as a general result, that as all reflecting surfaces are the separating surfaces of two media, the pencils which they reflect and transmit must necessarily have a different tint from the incident pencil, excepting in the extreme case, and one not known to exist, where the two bodies in contact have the same refractive powers, or the same differences of refractive power for every ray of the spectrum.

Hitherto we have supposed the irrationality of the coloured spaces to be simple, but it may be compound, and there may be two, three, or more points in the spectrum of two adjacent media where the indices of refraction are the same, or have equal differences, while in other points they are not the same, or have their differences unequal. In these cases the reflected and transmitted tints will be compound; but as such colours have not been observed, it would be out of place to make any farther reference to such a supposition. It may be sufficient to remark, that even if we never discover spectra of such a character, they may exist in the refractions at the separating surfaces of the tinging corpuscles of Newton, and the media which fill their interstices.

In the sixth class of colours may be ranked those produced by heat in metals and other substances, the colours of different bodies in combustion, and those exhibited in <188> the deflagration of metals. There is no reason to believe that any of these colours have the slightest analogy with those of thin plates, and their nature and origin remain to be investigated.

The colours of metals, which form the seventh class in our enumeration, have been referred by Newton to those of thin plates, but without any plausible reason. The polarisation of light by metallic bodies required to be investigated before the problem of their colour could be solved, and we owe its solution to the recent and beautiful researches of M. Jamin. As it would be foreign to the character of the present work to give an account of the process by which M. Jamin obtained his results, we must content ourselves with presenting them in the following Table :-

Copper, Orange, very red, 69° 56′ 0.113
Brass, Yellow, 103 13 0.112
Bell metal, Orange, yellow, 83 10 0.065
Speculum metal, Orange, very red, 67 25 0.027
Zinc, Blue, 180 57 0.021
Silver, Orange, yellow, 89 00 0.013
Steel, White,
Copper, Red, middle, 42° 29′ 0.812
Brass, Orange, very red, 62 50 0.349
Bell metal, Red, 40 40 0.767
Speculum metal, Red, orange, 53 59 0.292
Zinc, Blue, indigo, 267 58 0.188
Silver, Orange, yellow, 84 32 0.124
Steel, White,

The numbers in the column marked D are the distances of the tint of the metal from the red end of a spectrum, whose whole length is 360° ; and those under the letter C are the intensities of the tints, that of the incident white light being 1.000.


After a careful study of these different classes of colours, philosophers will have no hesitation in concluding that Newton's theory of the colours of natural bodies has only a limited application, and that instead of any general theory such as he contemplated, we must look for a separate explanation of the different classes of phenomena. The first class of our enumeration, which comprehends the largest number of coloured bodies, is the one which presents the greatest difficulty to philosophers; and to it the Newtonian hypothesis is certainly inapplicable. Within the solids, fluids, and gases of this class, certain rays of the intromitted pencil are absorbed or lost, while others are transmitted, or, what is the same thing, the coloured body has different degrees of transparency for different rays, being opaque for different portions of the spectrum at different thicknesses; whereas, in colourless bodies, the rays are absorbed in equal proportions, so that the transmitted beam emerges colourless. The colour of a body, therefore, is not produced by particles having the same colour as itself, but it is the colour which arises from the mixture of all the transmitted rays, and these rays proceed from every part of the spectrum, though in different proportions. Hence we must look for the cause of the colour in the constitution of the body itself, that is, in the manner in which its atoms are combined, and not in the size or nature of the atoms themselves.

In support and in illustration of this opinion, we may mention a few remarkable examples, in which the colour is changed by a change in the condition of its particles. The most remarkable of these is nitrous gas. This body is almost transparent in small thicknesses, and at low temperatures. By heating it, its colour becomes in succession straw yellow, orange, red, and even absolutely <190> black. When phosphorus, which in its ordinary state is of a pale yellow colour, is melted and thrown into cold water, it becomes black, and recovers its original colour when again melted. It is therefore obvious, that in both these cases the blackness could not be produced by any diminution in the size of the particles. A similar change of colour is produced by simple mechanical pressure on the crystals of iodide of mercury, which change their colour by simply pricking them with a sharp point.

The various phenomena of colour in crystallized bodies, and the influence of the continued action of light upon coloured substances, indicate the existence of different causes of colour; and the influence of structure, as one of these causes, is finely shewn in the relation of the colours of dichroitic crystals to their axes of double refraction or crystallization.

The great diversity in the constitution of coloured bodies is peculiarly shewn in the diversity of their action on the different rays of the spectrum; and it is therefore probable that the cause of their difference of colour may be found in the diversity of action exercised upon light by their particles or elementary atoms. In describing the colours of the fifth class, we have already mentioned an experiment with flint glass and oil of cassia, and its indication of a new theory of the colours of natural bodies of the first class. In Fraunhofer's spectrum, the principal black lines which it contains are represented by the letters A, B, C, D, E, F, G, H, I; — AI being nearly the whole of its length. If a, b, c, d, e, f, g, h, i, represent the same lines formed by any fluid or solid different from that which produces the spectrum AI, then though ai be equal to AI, it frequently happens, and we venture to say, always happens, that ab is not equal to AB, nor cf to CF, while <191> ac may be equal to AC, and dh to DH. The equal spectra may coincide in particular points, that is, individual lines in the one, indicating particular colours, may coincide with individual lines marking the same colours in the other spectrum, and yet other lines may not coincide, indicating different colours. When a ray of white light, therefore, is incident on the separating surface of the two media which give these two spectra, a very large portion, or rather the whole of the colours, indicated by the coincident lines, will be transmitted, while a very small portion of the colours indicated by the non-coincident lines will be reflected, the greatest quantity of the colours being reflected where the non-coincidence is greatest, and the greatest quantity being transmitted at the points of coincidence. Where there are many separating surfaces, and many elements in the body, the spectrum obtained by the prismatic analysis of the transmitted light will be cut up by obscure portions exactly as it is found to be in all coloured media.

When the constitution of any coloured body is altered by heat or pressure, the refractive and dispersive power of its elements are changed, and the resulting colour altered, according to the ratio in which the refracting forces are changed in the elementary molecules. Changes of this kind are finely exhibited in the growth of certain coloured crystals. In the tourmaline, for example, we have sometimes a red nucleus which absorbs one of the doubly refracted pencils, namely, the green one, and transmits only the red. When this nucleus was completed, some change had taken place in the circumstances under which the crystallization was proceeding,and the molecules, though still combining as tourmaline, combine in such a in such a manner as to produce no colour — no difference in the tint <192> of the pencils — and no absorption of one of them. At a subsequent stage, the structure which produces the red colour again appears and disappears, forming in succession coloured and colourless laminæ round the original nucleus !

Another example of great interest is afforded by certain specimens fluor spar, in which the colours of the fourth class are produced.[10] The structure which produces a white phosphorescence, is succeeded by one which produces a coloured phosphorescence, and this again by a structure which produces no phosphorescence at all. The changes of structure to which these different effects are owing, arise, in all probability, from a change in the arrangement of the atoms in the molecular groups of which the body is composed.

[1] Optics, Book i., Part ii., Prop. 10.

[2] Optics, Book ii., Part iii.

[3] A full account of these experiments, with coloured drawings of the spectra, will be found in the Edinburgh Transactions, 1833, vol. xii. pp. 538-545.

[4] See Phil. Trans. 1814, p. 397, and 1836, pp. 55, 56.

[5] See Layard's Discoveries in the Ruins of Nineveh, 1853, pp. 674-676; and Phil. Trans. 1837, p. 249.

[6] See Edinburgh Transactions, 1829, vol. xi. p. 322; and Reports of the British Association, 1844, p. 9.

[7] See Phil. Trans. 1814, p. 397; and 1829, p. 301.

[8] See Edinburgh Transactions, 1833, vol. xii. p. 512; and 1846, vol. xvi. p. 111; Reports of the British Association, 1838, pp. 10-12; Phil. Trans. 1845, p. 143; and 1852, p. 463.

[9] In his Memoir on Diffraction, Fresnel has thrown out the idea that, at great incidences, and with very thin laiminæ, the law of refraction may not follow the proportionality of the sines.

[10] See Edinburgh Transactions, vol. xvi. p. 112.

© 2020 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