REPORT OF THE COMMITTEE ON COLOURVISION.

The Committee on Colour-Vision appointed by the Council of the Royal Society on March 20, 1890, and consisting of the following members:-The Lord Rayleigh, Sec. R.S., Chairman; The Lord Kelvin, Pres. R.S.; Mr. R. Brudenell Carter; Prof. A. H. Church, F.R.S.; Mr. J. Evans, Treas. R.S.; Dr. R. Farquharson, M.P.; Prof. M. Foster, Sec. R.S.; Mr. F. Galton, F.R.S.; Dr. W. Pole, F.R.S.; Sir G. G. Stokes, Bart, M.P., F.R.S.; and Captain W. de W. Abney, C.B., F.R.S., Secretary, now submit their Report, with Minutes of the Evidence taken.

The Committee have held 30 meetings, and have examined more than 500 individuals as to their colour-vision. They have tried various methods and apparatus, including Holmgren's wooltest with Dr. Jeaffreson's and Dr. Thomson's modifications, Lord Rayleigh's colour-mixing apparatus and that of Captain Abney, Dr. Karl Grossmann's system, the lantern devised by Mr. F. Galton, and Mr. Lovibond's tintometer. They have taken the evidence of Captain Steele, of the Board of Trade; Mr. Rosser, a private instructor in navigation; Messrs. J. J. Hanbury, A. S. H. Wadden, and Bambridge, connected with the colourtesting departments of certain railways; Captain Macnab, of the Liverpool Board of Trade; Captain Angove, of the Peninsular and Oriental Steamship Company; and the following surgeons and experts in colour-vision testing:-Mr. Priestley Smith, Mr. T. H. Bickerton, Mr. E. Nettleship, Staff-Surgeon T. J. Preston, Dr. G. Lindsay Johnson, and Dr. Edridge Green. The Committee are under great obligations to Captain Abney, not only for having officiated as Secretary, but also for his very considerable labour in the determination of colour-constants, the registration of colours, and the examination, by spectral methods, of particular cases of defective colour-vision.

After weighing the evidence which they have obtained, the Committee have unanimously agreed upon the following recommendations:

1. That the Board of Trade, or some other central authority, should schedule certain employments in the mercantile marine and on railways, the filling of which by persons whose vision is defective either for colour or form, or who are ignorant of the names of colours, would involve danger to life and property.

2. That the proper testing, both for colour and form, of al candidates for such employments should be compulsory.

3. That the testing should be entrusted to examiners certificated by the central authority.

4. That the test for colour-vision should be that of Holmgren, the sets of wools being approved by the central authority before use, especially as to the correctness of the three test colours, and also of the confusion colours. If the test be satisfactorily passed, it should be followed by the candidate being required to name without hesitation the colours which are employed as signals or lights, and also white light.

5. That the tests for form should be those of Snellen, and that they should be carried out as laid down in Appendix VI. It would probably, in most cases, suffice if half normal vision in each eye were required.

6. That a candidate rejected for any of the specified employments should have a right of appeal to an expert approved by the central authority, whose decision should be final. 7. That a candidate who is rejected for naming colours wrongly, but who has been proved to possess normal colour-vision, should be allowed to be re-examined after a proper interval of time.

8. That a certificate of the candidate's colour-vision and formvision according to the appointed tests, and his capacity for naming the signal colours, should be given by the examiner; and that a schedule of persons examined, showing the results, together with the nature of the employments for which examinations were held, should be sent annually to the central authority.

9. That every third year, or oftener, persons filling the scheduled employments should be examined for form

vision.

10. That the tests in use, and the mode of conducting examinations at the different testing stations, should be inspected periodically by a scientific expert, appointed for that purpose by the central authority.

11. That the colours used for lights on board ship, and for lamp signals on railways, should, so far as possible, be uniform, and that glasses of the same colour as the green and red sealed pattern glasses of the Royal Navy, should be generally adopted.

12. That in case of judicial inquiries as to collisions or accidents, witnesses giving evidence as to the nature or position of coloured signals or lights should be themselves tested for colour- and form-vision.

April 28, 1892.

(Signed)

RAYLEIGH,

Chairman.

The reasons on which the Committee have based these recommendations are set forth in the following pages.

The subject of colour-sense and its imperfections is one Introductory. which is necessarily of great scientific interest; but it also has a practical importance, as it affects a definite proportion of the men who are engaged in the two great industries of railway traffic and of navigation. Amongst railway men, at least, if not also amongst sailors, a suspicion has been excited that the methods adopted for testing coloursense are not entirely trustworthy, and have had the effect of excluding some individuals from employments, the duties of which they were well qualified to discharge. On this ground alone, if on no other, it has seemed advisable to the Committee that the reasons for their recommendations should be so stated as to be intelligible, as far as possible, to all those who are interested in the matter.

Every colour, and among colours for convenience sake are included black and white, can be defined by three qualities:-1st, its hue-thus we talk of red, green, violet; 2nd, its purity, or the measure of its freedom from admixture with white-which is expressed by such terms as "deep" or "pale;" and 3rd, its brightness or luminosity-thus we say a colour is "bright," .or "dark." Two colours are identical only when they can be defined as possessing the same three colour qualities, or constants as they are called, and if they differ in any one they are no longer the same. When two objects are compared together for colour, the large majority of persons will agree as to their identity or difference. Their verbal descriptions of the difference may vary slightly, but practical tests show that in reality they recognize the same variations, and hence their vision is termed normal vision. There is, however, not an inconsiderable minority, as will presently be shown, whose perception of colour differs very widely from that of the majority, and, for want of a better term, members of this minority are called "colour-blind." By this term it is not intended to convey the idea that there is absolute insensibility of vision, or even of colour-vision, but merely that the ordinary distinction between certain colours is defective The variations in the amount of this deficiency in colour-perception are numerous, and when small, are often exceedingly difficult to classify.

We have to regard these deviations from normal vision more from a practical than from a theoretical standpoint, and in testing for them we have to take the broad view that the colourblindness which has to be detected is that which may be dangerous to the public in the industries already mentioned.

blindness.

There are some few people who fail to distinguish blue Character of from green, and others, equally few, who only see in mono- colourchrome, but the colour-blindness which is most common, and, therefore, most dangerous, is the so-called red-green blindness, in which there is a failure to distinguish between red and

green; that is to say, a red-green blind person will regard a certain hue of green as identical in colour with some hue of red, another of green as identical with white, and some will also fail to see red at all of another particular hue. When it is considered that on our railways white, green, and red lights are used as safety and danger signals at night, and that the same colours are not unfrequently used for a similar purpose by day, it is very obvious that to place persons who are red-green blind in positions where the colours ought to be correctly recognised may be the cause of disasters. The same objection to the employment of persons with defective colour-vision applies also to navigation, for at night the presence of a green or red light on the port or starboard side indicates the course that a vessel is taking, and if either those in charge, or on the look-out, are colour-blind, serious risks of collisions are run.

Description of It is proposed to enter somewhat minutely into the the spectrum. characteristics of red-green blindness, showing how it may be divided into two species. For this purpose it is necessary to appeal to the spectrum. When a thin slice of white light falls on one or more prisms, or on what is known as a diffraction grating, it is decomposed into a parti-coloured band which we call the spectrum, the principal colours, as given by Newton, being red, orange, yellow, green, blue, indigo and violet. If the light be that from the sun innumerable black lines will be seen interrupting this series of colours, some more marked than others. It is found that these lines always occupy the same position as regards the colour in which they are situated, and hence the more pronounced ones will act to the spectrum as milestones do to a road. Different coloured rays have different lengths of undulations in the all-pervading medium which is called ether, and the wave lengths of the coloured rays which, if present, would occupy the place of the principal black lines have, notwithstanding their minuteness, been determined with extreme accuracy, and this enables the position of any particular hue of spectrum colour to be numerically fixed by a reference to the wave lengths of these lines. We have said that the principal spectrum colours are those stated above, but it must be understood that they are only fully recognized by persons possessing normal vision; for the spectrum would be described by a colour-blind person in very different terms. For instance, some red-green blind would say that the red, orange, and yellow were all yellow; red would be described as dark yellow, orange as less dark, and yellow as bright yellow, whilst the green part of the spectrum bordering on the yellow would be described as yellow diluted with white. In the pure green would be pointed out a white or grey band, and the bluegreen would be described as blue diluted with white; whilst the blue would be called light blue, and the violet dark blue (see No. 2, Plate I). Others, again, whilst similarly describing the blue and violet part of the spectrum would substitute green