Curiosities of Light and Sight Part 2

Even as things are, the standard of neutrality is not quite a hard and fast one. We have a tendency to regard any artificial light which we may happen to be using, as more free from colour than it would turn out to be if compared directly with sunlight. If in the middle of the day we go suddenly into a gas-lit room, we cannot fail to observe how intensely yellow the illumination at first appears; in a few minutes, however, the colour loses its obtrusiveness and we cease to take much notice of it.

The effect may be partly a physiological one, depending upon unequal fatigue of the various perceptive nerves of the retina; but I believe that it is to a large extent due to mental judgment. The standard of whiteness, or colour-zero, can apparently be changed within certain limits in a very short time, and, as we shall see later, this is only one of many instances in which our organs of vision seem to be incapable of recognising a constant standard of reference.

And now let us consider how it comes about that each elementary portion of the retina--at least in its central region--has the power of distinguishing so many hundreds of different hues. It is incredible that every little area of microscopic dimensions should be furnished with such a mult.i.tude of independent organs as would be necessary if each of the many colours met with in nature required a separate organ for its perception; and it is not necessary to suppose anything of the kind.

Experiment shows that all the various hues of the spectrum, as well as all (including white) that can be formed from their mixture, may be derived from no more than three distinct colours. There are, in fact, an indefinite number of triads of colours which, in suitable combinations, are capable of producing the sensation of every tone, tint, and shade of colour which the eye of man has ever beheld.

Old-fashioned books, such as an early edition of Ganot's "Physics," tell us that the three "primary" colours are red, yellow, and blue, and that all others are produced by mixtures of these. This was the basis of Sir David Brewster's theory, which attained a very wide popularity, and even at the present time is held as an article of faith among the great majority of intelligent persons who have not paid any special attention to science. But it is not true. A fatal objection to it is the well-ascertained fact that no combination of red, yellow, and blue, or of any two of them, such as blue and yellow, for example, will produce green.

Yet every painter knows that if he mixes blue and yellow pigments together he gets green. That is one of the first things that a child learns when he is allowed to play with a box of water-colours, and no doubt Brewster was misled by the fact.

The truth is, that the colours of all, or almost all, known blue and yellow pigments happen to be composite. An ordinary blue paint reflects not only blue light, but a large quant.i.ty of green as well; while an ordinary yellow paint reflects a large quant.i.ty of green light in addition to yellow. When such paints are mixed together, the blue and yellow hues neutralise one another, and only the green, which is common to both, remains.

The spectrum apparatus will make this clearer. Hold a piece of bright blue gla.s.s before the slit; the light pa.s.sing through the gla.s.s will be a.n.a.lysed by the prism, and you will see that it really contains almost as much green as blue. If a yellow gla.s.s is subst.i.tuted, not only will yellow light be transmitted, but, as before, a considerable quant.i.ty of green. If now both gla.s.ses be placed together before the slit, what will happen? The yellow gla.s.s will stop the blue light transmitted by the blue gla.s.s, the blue gla.s.s will stop the yellow light transmitted by the yellow gla.s.s, and only the green light which both gla.s.ses have the power of transmitting will pa.s.s through unimpeded, forming a band of pure green colour upon the screen.

The combination of simple blue and yellow lights of suitable relative luminosities results in the formation of white or neutral light. If the blue is a little in excess, the combined light will be of a bluish tint; if the yellow is in excess, the combination will have a yellowish tint. It will never contain any trace of green. The combination of simple spectral blue and yellow is easily effected by the colour-patch apparatus, and the result will be found to bear out what has been said.

Since, then, no mixture of red, yellow, and blue, or of any two of them, will produce green, we cannot regard these colours as being, in Brewster's sense of the term, primary ones.

But it is quite possible to find a group of three different hues--and indeed many such groups--which when made to act upon the eye simultaneously and in the right proportions can give rise to the sensation of any colour whatever. Now this experimental fact is obviously suggestive of a possible converse, namely, that almost every colour sensation may in reality be a compound one, the resultant of not more than three simple sensations. a.s.suming this to be so, it is evident that if each elementary area of the retina were provided with only three suitable colour organs, nothing more would be requisite for the perception of an indefinite number of distinct colours.

Such a hypothesis was first proposed by Thomas Young at the beginning of the present century; but it came before its time and met with no attention until fifty years later, when it was unearthed by the distinguished physicist and physiologist, Helmholtz, who accorded to it his powerful support and modified it in one or two important details.

[Ill.u.s.tration: _Fig. 6.--Helmholtz's Curves of Colour Perception._]

According to the Young-Helmholtz theory, as it is now called, there are three different kinds of nerve-fibres distributed over the retina. The first, when separately stimulated, produce the sensation of red, the second that of green, and the third that of violet. Light having the same wave-length as the extreme red rays of the spectrum stimulates the red nerve-fibres only; that having the same wave-length as the extreme violet rays stimulates the violet nerve-fibres only. Light of all intermediate wave-lengths, corresponding to the orange, yellow, green, and blue of the spectrum, stimulates all three sets of nerve-fibres at once, but in different degrees. The proportionate stimulation of the red, green, and violet nerves throughout the spectrum is indicated in Fig. 6, which is derived from the rough sketch first given by Helmholtz. The yellow rays of the spectrum, it will be seen, excite the red and green nerves strongly, and the violet feebly; green light excites the green nerves strongly, and the red and violet moderately; while blue light excites the green and violet nerves strongly, and the red feebly.

[Ill.u.s.tration: _Fig. 7.--Konig's Curves._]

Fig. 7 shows another set of curves given more recently by Dr. Konig as the result of many thousands of experiments made, not only upon persons whose vision was normal, but also upon some who were colour-blind. Konig found that the equations he obtained were best satisfied by a.s.suming as the normal fundamental sensations a purplish red (not to be found in the spectrum), a green like that of wave-length 5050, and a blue like that of wave-length 4700 approximately, the two latter, however, being purer or more saturated than any actual spectrum colour. But Konig's curves are not consistent with every cla.s.s of vision which he examined, and the question as to what are the true fundamental colour-sensations, if such really exist at all, cannot yet be regarded as finally settled.[6]

The Young-Helmholtz theory of colour-vision, whether or not it is destined in the future to be superseded by some other, has at all events proved an invaluable guide in experimental work, and there are very few colour phenomena of which it is not competent to offer a satisfactory explanation. It has at present only one serious rival--the theory of Hering, which, although it seems to be curiously attractive to many physiologists, can hardly be said to present less serious difficulties than that which it seeks to displace. Neither of these competing theories has yet had its fundamental a.s.sumptions confirmed by any direct evidence, and the advantage must rest with the one which best accords with the facts of colour vision. In my judgment the older of the two is to be greatly preferred as a useful working hypothesis.

Certain curiosities of vision with which I propose to deal in a future chapter depend upon the properties of what are known as complementary colours. Two colours are said to be complementary to each other when their combination in proper proportions results in the formation of white.

[Ill.u.s.tration: _Fig. 8.--Stencil Card for Complementary Colours._]

If we produce a compound hue by mixing together the colours of any portion of the spectrum, and a second compound hue by mixing the remainder of the spectrum, it must be evident that these two hues are necessarily complementary, for when they are united they contain together all the elements of the entire spectrum, and therefore appear as white. This may be ill.u.s.trated with the aid of the colour-patch apparatus. Place at H (Fig. 3) a cardboard stencil of the form shown in Fig. 8, and focus upon it a little spectrum, the princ.i.p.al hues of which are indicated by the letters R O Y G B V (red, orange, yellow, green, blue, violet). The two oblong apertures in the card should be of exactly the same height, and the card so placed that one aperture may admit rays extending from the red end of the spectrum to about the middle of the green, while the other admits rays from the remainder of the spectrum. If now the lower aperture be covered, only the red, orange, yellow, and part of the green rays will pa.s.s through the stencil, and these being combined by the lens K (Fig. 3) will form upon the screen a bright patch, the colour of which will be yellow. If the upper aperture be covered, and the rest of the green, together with the blue and violet rays, allowed to pa.s.s through the other, the colour of the patch will become blue; and if both apertures be uncovered at the same time, rays from the whole length of the spectrum will pa.s.s through the stencil, and the patch will, of course, turn white.

The yellow and the blue which were compounded from the two portions of the spectrum are, therefore, in accordance with the definition, complementary colours.

In a similar manner by dividing the spectrum into any two portions whatever--as, for example, by the complicated stencil shown in Fig. 9--we can obtain an indefinite number of pairs of complementary colours.

[Ill.u.s.tration: _Fig. 9.--Stencil Card for Complementary Colours._]

But it is by no means indispensable that both or either of a pair of complementary colours should be compound. To prove this, two strips of card with narrow vertical openings A and B are prepared as shown in Fig.

10. The cards are placed one above the other and can be slipped in a horizontal direction, so that the narrow openings can be brought into any desired part of the spectrum which is indicated in outline by the dotted oblong.

[Ill.u.s.tration: _Fig. 10.--Slide for mixing any two Spectral Colours._]

Bring the opening A of the upper card into the yellow of the spectrum and the opening B of the lower card into the blue. The bright patch formed upon the screen will then be illuminated by simple blue and yellow rays; yet it will be white--not green, as it would be if Brewster's theory were correct. If upon the first trial the white should not be absolutely pure, it can easily be made so by partially covering either A or B--the first if the white is yellowish, the second if it is bluish. Simple spectral blue and yellow are therefore no less truly complementary colours than are the compound hues formed when the spectrum is divided into two parts.

It is noticeable, however, that the white light resulting from the combination of blue and yellow, though it cannot be distinguished by the eye from ordinary white light, is yet possessed of very different properties. Most coloured objects when illuminated by it have their hues greatly altered; a piece of ribbon, for example, which in common light is bright red, will appear when held in the blue-yellow light to be of a dark slate colour, almost black.

If the opening A is placed in any part whatever of the spectrum except the green, it will always be possible, by moving B backwards or forwards, to find some other part where the colour is complementary to that at A. To green there is no simple complementary; a purple is required, which is not found in the spectrum, but may be formed by combining small portions of spectral blue and red. For studying mixtures of three simple colours, a third slide may be added to the two shown in Fig. 10.

The following little table gives the princ.i.p.al pairs of complementary colours.

TABLE OF COMPLEMENTARY COLOURS.

Red Greenish-blue Orange Sky-blue Yellow Blue Greenish-yellow Violet Green Purple

CHAPTER III.

SOME OPTICAL DEFECTS OF THE EYE.

More than one reference has been made to the fact that the sense of sight, even in its best normal condition, is characterised by certain defects and anomalies. Some of these arise directly from causes inherent in the design or structure of the eye itself, and may be broadly cla.s.sified as physical; others are of psychological origin, and result from the erroneous interpretations placed by the mind upon the phenomena presented to it through the medium of the optic nerve and the brain.

Among the numerous physical defects of the eye none is more remarkable than the absence of means for properly correcting chromatic aberration.

This defect is remarkable because it appears--at least to those who are without actual experience in the manufacture of eyes--to be one which might very easily have been avoided. So far as a mere theorist can judge, an achromatic arrangement of lenses would have been just as simple and just as cheap (if I may use the term) as the arrangement with which we find ourselves provided. It is true that we manage to go through life very well with our uncorrected lenses, and indeed it is hardly possible by ordinary observation to detect any evidence of the imperfection. Yet its existence in a glaring degree is undoubted, and can be readily demonstrated by a great variety of methods. The conclusion is inevitable that with achromatic eyes our vision would be improved, but whether there may not possibly exist reasons why such an improvement could only be achieved at a disproportionately high cost is a question which cannot at present be answered.

Without going into matters which are dealt with in every elementary text book of optics or general physics, it may be desirable to explain shortly what is meant by the terms chromatic aberration, and achromatism.

[Ill.u.s.tration: _Fig. 11.--Refraction of monochromatic Light by a lens._]

Let L L, Fig. 11, represent in section a circular convex lens, and P a luminous point, which is most conveniently supposed to be situated on the axis of the lens. Imagine P to be surrounded in the first instance by a gla.s.s shade which transmits only monochromatic red light. So much of the light from P as falls upon the lens will be refracted to a point at the conjugate focus F, and after pa.s.sing this point will diverge again; the refracted light rays will, in fact, form a double cone, of which F is the apex. If a white screen be held at F, there will be focussed upon it a small clearly-defined image of the luminous point. If, however, the screen be moved nearer to or further from the lens, it will cut the cone of light, and the image will then no longer appear as a point, but as a circular red disk, which will be larger the greater the distance of the screen from F. Such a disk is known as a "diffusion circle."

Suppose now that we subst.i.tute for the red gla.s.s, surrounding the source of light, a purple one capable of transmitting not only red rays but violet as well. The lens will cause both the red and the violet rays which pa.s.s through it to converge; but since the violet rays are more refrangible--more easily refracted or bent aside out of their straight course--than the red, there will now be two double cones, as shown in Fig.

12, where the contours of the red cones are represented by solid lines and those of the violet by dots.

[Ill.u.s.tration: _Fig. 12.--Refraction of dichromatic Light._]

The focus of the red rays will as before be at F, but that of the violet will be nearer to the lens, as at H, and this being so, it is evident that a well defined image of the purple source of light cannot possibly be formed upon a screen placed anywhere behind the lens. Held in the position indicated by the line C C, where it pa.s.ses through the focus of the red rays, the screen cuts one of the cones of violet light, and the image at F will appear to be surrounded by a violet halo. Held at A A, the screen evidently receives an image with a red halo round it. Only at B B, in the plane where the surfaces of the red and violet cones cut one another, will it be possible to obtain an image without a coloured border; but here good definition is unattainable, for neither the red nor the violet rays are in focus, and the luminous point is represented by a purple disk or diffusion circle of sensible diameter.

If rays of every possible refrangibility are allowed to fall upon the lens, as is the case when the source of light is not shielded by any coloured gla.s.s, there will be formed an indefinite number of pairs of cones, the apices of which will lie along the straight line joining H and F. It is clear that all these cones cannot possibly intersect in a single plane, and consequently no position can be found where the edge of the projected image is perfectly free from colour, though at a certain distance from the lens, where the brightest const.i.tuents of the light--namely, the yellow and green--are approximately focussed, the coloured border is least conspicuous, and is of a purple tint, due to the mixture of the red and violet rays.

For these reasons a single gla.s.s lens cannot, except with h.o.m.ogeneous light, be made to give a perfectly distinct image of a luminous point, nor of an illuminated object, the surface of which may be regarded as an a.s.semblage of points. Such a lens, therefore, is never employed when good definition is required. The confusion resulting from the unequal refrangibility of the differently coloured rays is said to be due to the chromatic aberration of the lens.

In connection with this matter, the history of physical optics contains an interesting little episode. It occurred to Sir Isaac Newton that although a single lens could never be free from chromatic aberration, yet it might be possible to arrange a so-called achromatic combination of lenses in such a manner as to overcome the defect and bring all the rays issuing from a point, whatever their refrangibility, to one focus. Experiments which he undertook for the purpose of testing the matter led him to form the conclusion that such a result could never be attained, the amount of colour dispersion in all substances being, as he stated, always exactly proportional to that of refraction. For this reason he confidently announced that it was useless to attempt the construction of a really good refracting telescope, and so great was the authority attaching to his name that for many years all efforts in that direction were abandoned.

Nevertheless from time to time certain philosophers ventured to surmise that Newton might perhaps have been mistaken, and the curious thing is that they all based their scepticism upon what they considered the self-evident fact of the achromatism of the eye. The system of lenses in the eye, they argued, being unquestionably achromatic, why should not an equally effective combination be constructed artificially?

At length, more than eighty years after Newton had made and published his fundamental experiments, it occurred to a working optician, John Dollond, that it might be worth while to repeat them, and upon doing so he at once found that Newton was wrong in his facts, the results as recorded by him being in direct opposition to the truth. With proper respect for the memory of a great man it is usual to speak of Newton's observation as a "hasty" one, but if in these days a junior science student were to be guilty of a similar lapse, his conduct would not impossibly be stigmatised as grossly careless.

Having established Newton's error, Dollond found little difficulty in constructing achromatic lenses of very satisfactory quality; telescopes of his manufacture long enjoyed the highest reputation, and the best optical instruments of the present day are the direct offspring of his invention.