The second standard skein is light purple or pink, and its complementary colour is a green in the spectrum A 5100. The colour is diluted with about 40 per cent. of white. The third test-skein has a colour corresponding with a red of the spectrum ( 6330) diluted with 18 per cent. of white.

Should an accident happen at any time to the standard sealed pattern skeins, the exact hues can be reproduced from the spectrum by a reference to these numbers. The Committee cannot conceal from themselves the fact that the wools are apt to deteriorate with use, both by the constant handling and also, to some extent, by light. In the test as carried out by Holmgren there is but little doubt that almost as much information is conveyed to the examiner by the way in which the different skeins are picked up to match the test-skein as by the absolute matching itself, and this procedure involves handling them and also exposure to light. The assortment of wools which is used in practical testing should therefore be renewed from time to time.

test.

In Jeaffreson's form of this test, which is given in Appendix IV, Jeaffreson's page 392, the handling of the colours is avoided, the match being made as there described. The hesitation evinced by the colourblind in matching the test-colouris, in this instrument, also, of great utility to the examiner; moreover, it has been found practically that as many, or even more persons can be examined in a given time by it than by the original plan. The Committee are therefore of opinion that this modification may be admitted if desired by the examiner.

These wool-tests will detect red-, green-, and violet-blindness, and all other forms of congenital defective colour-vision. The matches of colours will indicate to the examiner the character and extent of the defect.

In cases of appeal the examinations should take a wider range. Examination The test with the spectroscope is decisive, and in Appendix V. on appeal. is described a method of applying it which the Committee think may be convenient and satisfactory.

All tests in which the wools are suspended from a bar, even Tests to be though the test-skeins may be of proper colour and tone, should avoided. be avoided, since the order of arrangement might be ascertained by some means or another by those who are tested. It is quite true that the order might be changed; but in an examination of this character, where large numbers may be under trial, any frequent changing of the order would be impracticable, and hence there would be no security that the test was efficient. The same objection applies to all diagrams of colours which the examinees are required to match with standard colours. Coaching here is even more easily carried out than with the suspended wools, since the diagrams are in the market, and the tints cannot be changed in position.

There are some other efficient tests that are less adapted for Other tests. examining large bodies of men than the wool-tests, but which

may be well applied to demonstrate the presence of colour

Colour-ignor

ance.

blindness in individual cases. Those of Dr. Grossmann are a good example of this class of test. An opinion has been expressed, and with some plausibility, that the only fair tests by which to prove that a man's colour-vision renders him unfit to distinguish coloured lights or signals are the coloured lights themselves when seen under the same circumstances as those under which they would have to be observed. It has already been shown that, with practice, it may be possible for a colourblind person to distinguish between colours by their different luminosities and dilution with white, but it has also been pointed out that such recognition would be rendered uncertain by differing states of the atmosphere and by other conditions. If it were possible to eliminate the chances of correct guessing, which would be very large when using such tests, it would be necessary that the examination should be a prolonged one, being repeated many times with differing conditions of weather. If it were not carried to this extent, it might equally well be conducted in a t sting room, where the apparent size of the signals to the eye could be imitated with great exactness. But the uncertainty of this method, even when the variable factor of weather is absent, is exemplified by the results of the examination of railway employés at Swindon, conducted by the Committee. They found, as already stated (see Appendix VII), that several passed the lamp-test who had failed to pass the wool-test, and that some passed one lamp-test, but failed to pass another similar one on the same occasion. Had the examination of these men been to ascertain their fitness for certain employments requiring normal colour-vision, and been conducted by the lamp-test only, some would have been admitted into the service, and have been a source of danger to the public.

The Committee have had to consider whether what has been called colour-ignorance, that is, ignorance as to the names of colours, is as objectionable as colour-blindness for certain employments. The possibility of the existence of real colourignorance, such as would lead to a non-recognition of the true colour of a signal, appeared to them very doubtful until they had taken the evidence of Staff-Surgeon Preston, R.N.; for it was hard to conceive of ignorance which would lead to confusion in naming a red, a green, and a white signal. His evidence, however, was conclusive of its existence at certain recruiting centres, and more especially in a certain class of recruit. It may be mentioned that in the actual testing of large bodies of men by the Committee, in no case was there a trace of colourignorance exhibited by those possessing normal vision, unless in regard to nondescript colours. Red, green, blue, and white were always correctly named, except where the person examined was proved to be colour-deficient.

There is one type of colour-ignorance which of course may often be encountered; a foreigner on board an English-commanded vessel, would be, practically speaking, colour-ignorant if

he were unable to name the colours in English. It is in evidence before us that in navigation it is often requisite that the look-out man should, without a moment's delay, pass to the officer in charge the name of the colour of a light, and that hesitation, whether caused by true colour-ignorance or from want of knowledge of English terms, might involve disaster. This being the case, the Committee are strongly of opinion that for the marine services the examination for colour-vision should exclude not only men who are colour-blind within the limits Ignorance of already indicated, but also those who are colour-ignorant, whether the names of from defective education or from want of knowledge of the signal colours English names. No man should be accepted as a look-out unless should be a bar to employhe were found capable of naming the signal colours correctly and intelligibly, and without hesitation.

The tests which the Committee recommend for the detection Tests for of colour-ignorance are very simple. After the tests for colour- colourblindness have been satisfactorily passed it would suffice to ignorance. ask the examinees to name the reds and greens of the wool-tests, and if any hesitation was evinced to test them with a lanterntest, such as that proposed by Mr. Galton. Men rejected for colour-ignorance of either type should not be considered permanently ineligible, but only until such time as their education in the subject was perfected, for it must be recollected that, unlike colour-blindness, colour-ignorance is curable.

In the marine service, it appears that on each stage of promotion Re-testing in an officer is tested as to his colour-vision. On some railways also, the marine on promotion, an employé's eyesight is re-tested. It does not and railway appear that such tests are undertaken with the idea that colour- services. blindness of the congenital type may have become more pronounced, or may have induced it by disease, but rather with the view that those who have been previously tested may have been passed improperly. No doubt these re-examinations are a safeguard; but if the tests already passed had been such as to render detection a certainty, there would be no necessity for repetition except for the detection of such colour-blindness as may be due to disease, injury, or over-use of tobacco. Colour-blindness due to these last causes is at first very seldom appreciated by the sufferer, and is usually only discovered upon his consulting a medical man for impaired form-sense. This raises the question as to whether defective colour-sense other than congenital might not, in some cases, be found in those on whom the lives of passengers and others depend.

disease.

Special tests for colour-blindness induced by disease will very Tests for rarely be necessary if, as should always be the case, every exami- colournation for colour-vision is preceded by one for form. These latter blindness induced by tests are so well known, that the Committee do not think it necessary to enumerate them. If a candidate is found to have defective form-vision of a pronounced type he certainly should be ineligible for the positions of responsibility from which colour-blind persons should be excluded, and the test for form-vision would as a rule

Persons to be entrusted

with examination.

Periodic

therefore exclude the colour-blind of this type (see Appendix VI). It should be remarked that it is quite possible that the Holmgren wool-test might be passed satisfactorily by colour-blind people of this type, more particularly when the diseased area is confined to a small central spot in the retina; in fact, this has happened twice in the presence of the Committee.* The Committee would therefore rely rather on the form-test being stringently carried out, than on instituting another colour-test for this particular class of colour-blindness.

The qualification to be required from the examiners has received the careful consideration of the Committee. An examiner both in the railway and in the marine services would be called upon to carry out not only the tests for colour-vision but also those for form, and the Committee are of opinion that he should be required to obtain a certificate of competency from some duly constituted authority. Testing, such as we have recommended, requires careful training, and is not to be learnt except by practice, for it requires not only a registration of absolute mistakes, but also a ready observation of the manner in which the candidate acts whilst under examination. The Committee would not insist upon the examiner being a medical practitioner, but it is probable that a medical training would be of advantage. They are further of opinion that there should be a periodic inspection of the different testing stations by duly qualified ophthalmic surgeons, who should report upon the condition of the testing appliances and upon the mode in which the tests are carried out; and who might be the authorities to whom an appeal from a rejected candidate should be referred.

In no

case should any test be allowed in substitution of those recommended, though supplementary tests might be tried if desired. The passing or rejection of the candidates should always be based on the tests which have been laid down.

As colour-blindness of the congenital character is never examination. acquired, it is unnecessary that any one who has already been examined for colour-vision by efficient tests should be re-examined. But as tobacco-blindness is not uncommon, the form-sense of those men whose failure in vision would be dangerous to the safety of the public should be tested periodically, say, once every three years.

Persons to be examined..

The Committee are not prepared to give a list of those posts from which the colour-blind should be excluded. Pilots, look-out men and officers on board ship; engine-drivers, firemen and

Captain Abney prepared for the Committee pellets of baked clay of about inch diameter, coated with pigments in distemper of the same hues as those of the wools in the Holmgren test. The images of these small pellets fill such a minute area of the retina that those colour-blind persons were unable to pick out from a small trayful of them correct matches to any of the standard test colours, though they were perfectly able to pick out all those coloured with any shade of blue with ease. As stated above, they passed the ordinary wool-test, the colours being readily distinguished outside the diseased central retinal area.

signal-men on railways, evidently require sight unaffected by defects in colour or form, and there may be other positions, both in the marine service and in that of railways, which should also be included. Some central authority should make a schedule of such positions, and should take measures to enforce the exclusion of colour-blind persons from them.

NOTE (a).

The cause of the different sensations which are conveyed to Cases of the brain is a matter which is still in doubt. It is difficult to abnormal conceive that matter which is so comparatively gross as the rods colourand cones which are situated on the retina can be affected by the merely mechanical action of the vibrations of light.

The little we know about the actual nature of sensations leads us rather to believe that the nervous processes which are the foundation of sensations are, like other nervous processes, the outcome of chemical changes in nervous substances. And it has been suggested that vision originates in the chemical changes of a certain substance (or substances) in the retina, that the chemical condition of this substance, which has been called visual substance, is especially affected by the incidence of light, and that the changes so induced determine the beginnings of visual impulses and thus of visual sensations. We know that light can decompose a substance by acting on its molecules, and thus induce a chemical change in it.

In photographic processes, for instance, we know that the molecules of the sensitive substance are split up by white light, and further, that when these comparatively simple substances are exposed to the spectrum, although it is found that a considerable extent of it produces chemical changes, there is one particular part which acts more strongly than the rest of it. The curve of sensitiveness exhibits the same characteristics as those of the colour sensations in the Young-Helmholtz theory. If it be conceded that the retinal substance acted upon by light is a mixture of three analogous compounds, each having a maximum sensitiveness at a different point of the spectrum, we can account for the three fundamental sensation curves shown in the diagram at page 286.

NOTE (b).

Any complete theory of colour-vision must account not only for normal vision and congenital colour-blindness, but also for those cases of defective colour-sense which are due to disease or injury, and which differ so widely in character from each other.

blindness.

It is somewhat difficult to see how the Young-Helmholtz Difficulties of theory accounts for the last species of colour-blindness. Accord- accounting