698 COLOR-VISION AND COLOR-SLINDNESS. 



Such candidates were either improperly rejected on tlie first occasion, 

 or improperly accepted on the second. On English railways there has 

 been no uniformity in the methods of testing ; except (in so far as I am 

 acquainted with them) thatthey have been almost uniformly misleading, 

 calculated to give rise to the imputation of color-blindness where it did 

 not exist, and to leave it undiscovered where it did. In these circum- 

 stances it is not surprising that great discontent should have arisen 

 among railway men in relation to the subject; and this discontent has 

 led, indirectly, to the appointment of a committee by the Eoyal Society, 

 with the sanction of the board of trade, for the purpose of investigating 

 the whole question as completely as may be possible. 



It is perhaps worth while, before proceeding to describe the manner in 

 which the color sense of large bodies of men should be tested for indus- 

 trial purposes, to say something as to the amount of danger which 

 color-blindness produces. A locomotive, as we all know, is under the 

 charge of two men, the driver and the fireman. In a staff of 1,000 of 

 each, allotted to 1,000 locomotives, we should expect, in the absence of 

 any efficient method of examination, to find 40 color-blind drivers and 

 40 color-blind firemen. The chances would be ] in 25 that either the 

 driver or the fireman on any particular engine would be color blind ; 

 they would be 1 in 625 that both would be color-blind. These figures 

 appear to show a greater risk of accident than we find realized in actual 

 working, and it is manifest that there are compensations to be taken 

 into account. In the first place, the term " color-blind " is itself in some 

 degree misleading ; for it must be remembered that the signals to which 

 the color-blind person is said to be "blind" are not invisible to him. 

 To the red-blind, the rod light is a less luminous green ; to the green- 

 blind, the green light is a less luminous red. The danger arises because 

 the apparent differences are not sufficiently characteristic to lead to cer- 

 tain and prompt identification in all states of illumination and of atmos- 

 phere. It must be admitted therefore that a color blind driver may be 

 at work for a long time without mistakes; and it is probable, knowing, 

 as he must, that the differences between different signal lights appear 

 to him to be only trivial, that he will exercise extreme caution. Then 

 it must be remembered that lights never appear to an engine driver in 

 unexpected places. Before being intrusted with a train he is taken ov^er 

 the line, and is shown the precise i>osition of every light. If a light 

 did not appear where it was due, he would naturally ask his fireman to 

 aid in the lookout. It must be also remembered' that to over-run a 

 danger signal does not of necessity imply a collision. A driver may 

 over-run the signal, and after doing so may see a train or other obstruc- 

 tion on the line, and may stop in time to avoid an accident. In such a 

 case he would probably be reported and fined for over-running the sig- 

 nal; and if the same thing occurred again, he would be dismissed for 

 his assumed carelessness, probably with no suspicion of his defect. 

 Color-blind firemen are unquestionably thus driven out of the service 



