instance, which consists of four colours painted on a board in definite positions. A normal-sighted person might tell a colour-blind person the names and positions of the four colours.

The following letter, which appeared in the Railway Press, Dec. 13, 1889, is a good example of the unsatisfactory state of affairs with regard to testing for colourblindness by the railway companies.

SIR,

"COLOUR-BLINDNESS.

"Having read in your valuable paper a very interesting account of how the different companies test their servants as to their eyesight, I will now give you my experience before I was made permanent on the London and North Western and Great Western joint companies, Birkenhead. After being in their employ a few days, I was sent with a medical examination form to the company's doctor, to be filled up by him. He examined me; also asked me a few questions as to health and age, which he put on the certificate as I answered them. Then came the test as to my eyesight. He handed me a small basket containing skeins of coloured wool. He took from the basket three small skeins of different colours, and told me I was to match them to the best of my belief with the others that were left in, at the same time telling me there was none there exactly the same colour as what he gave me. After a little patience I matched them, as I thought, as near as could be. After doing so, he filled up my certificate, signed it, also asked me to do the same. But to my surprise, when I took it to my inspector, he told me I was colour-blind, and that I could not be made permanent. I told him that I was sure that my eyesight was good. I again had to visit the

same doctor, and went through the same performance with the skeins of coloured wool as before, and again he put me down as colour-blind. I was sent to him by my inspector for the third and last time, and he still said I was colour-blind. After this I was sent to our superintendent's office at Shrewsbury, and went through a severe test before my superintendent. First, I was placed by the office-window, and two flags, red and green, were held up, quite a quarter of a mile away, which I distinguished quite easily. Then I was tested with papers of different colours, and lastly with a square board with a quantity of small dots on, similar to the one used by the London and North Western Railway Company, and was successful in answering all questions correctly. I was next sent to the company's doctor in Shrewsbury, and was examined by him the same as I had been by the doctor in Birkenhead. But instead of the basket of skeins of coloured wool, he held up three knots of wool, red, white, and green, and asked me the colour, which I named correctly. After all this testing I was made permanent; and I question whether any other railway servant went through so much testing as what I did. You will now see that extreme care should be taken to really find out who is really colour-blind. The old saying applies in this case that doctors differ,' and that it is necessary that more than one person should judge a matter of this kind. "Yours, etc.,

"WILLIAM VAUGHAN, Signalman.

"Birkenhead, December 9, 1889."

Here a man was tested three times by one of the companies' surgeons, and found to be colour-blind each time. Still the superintendent is not satisfied, and himself tests the man, and then sends him to another of the

companies' surgeons. He is again tested, and the decision of the first surgeon reversed. From reading this letter, I am inclined to think that the writer was colour-blind. At any rate, neither the method of testing employed by the superintendent, nor that of the doctor at Shrewsbury, as described in this letter, is in any way an efficient test for colour-blindness. Why did not the companies have the man examined by an expert? This shows very well the way in which colour-blindness is regarded by the public. The three first examinations apparently counted for nothing. Had the superintendent appreciated the dangers of colour-blindness properly, he would have certainly referred to some expert before keeping the man.

CHAPTER XX.

THE TESTS FOR COLOUR-BLINDNESS.

UNFORTUNATELY nearly all the tests for colour-blindness in use at the present time have been based upon some theory of colour-perception. In this chapter I intend to deal with colour-blindness from a purely practical point of All the observations are based on the facts of colour-blindness, apart from any theory, and so will hold good whatever theory be adopted.

view.

Tests for colour-blindness are of two kinds; namely, those which are used for the purpose of ascertaining whether colour-blindness be present or not, and those which are used when the inquiry is made for some practical purpose. In Chapter XI. I have dealt with the first division of tests for colour-blindness, and it now remains for me to deal with the second.

On account of the arrangement of signals by sea and land, it is necessary that persons employed in the marine and railway services should be able to recognize and distinguish between the standard red, green, and white lights under all conditions in which they are likely to be placed.

It is not only necessary to find out whether a person is able to distinguish between the red, green, and white lights, but to ascertain as well that he thoroughly understands what is meant by colour, and the individual

characteristics of red, green, and white, respectively. Too little attention has been paid to this in constructing tests for colour-blindness, and those who have had much practical experience in testing for this defect, are aware of the ignorance which exists among uneducated persons with regard to colours. Many are under the impression that every shade of a colour is a fresh colour, and others have the most novel ideas with respect to colour. It is necessary that a sailor or engine-driver should be able to recognize a red, green, or white light by its character of redness, greenness, or whiteness, respectively; that is to say, that the examinee has definite ideas of colour, and is able to reason with respect to them. All persons who are not able, through physical defect, to have definite ideas of the standard colours, and to be able to distinguish between them, must be excluded from the marine and railway services.

In constructing a test for colour-blindness, we must not forget the element of colour-ignorance, because an enginedriver or sailor has to name a coloured light when he sees it, not to match it. He has to say to himself, "This is a red light, therefore there is danger;" and this is practically the same as if he made the observation out loud. Therefore, from the very commencement we have colournames introduced, and it is impossible to exclude them. As I have said before, making a person name a colour is an advantage, because the colour-name excludes the element of shade. If, as some persons have said, testing by colour-names is useless, then the whole series of colournames is useless. But if I say to a friend, "That tile is red," and he agrees with me, it is evident that one object, the colour of which is by him classed as red, is also classed as red by me. The ordinary colour-names, red, blue, yellow, and green, form excellent bases for classification.