516 PROCEEDINGS OF THE AMERICAN ACADEMY. 



story of failing color vision they did not tell the whole of it, since there 

 are many cases of defective central vision due to tobacco or alcoholism 

 which do not show with the Holmgren skeins by reason of these being 

 sufficiently large considerably to overlap the affected area. F'urther, 

 the psychological effect of testing a man's fitness for employment by 

 setting him at sorting colored wools is bad, for he may not unnaturally 

 think that colored lights would be quite a different matter. Color 

 discs had of course been used in testing but Dr. Williams studied the 

 subject deeply and in 1891 introduced into the C. B. & Q. system a 

 complete scheme for testing the vision of railway men; a set of care- 

 fully prepared cards on Snellen's system with special precautions 

 against the possibilities of memorizing, the Holmgren worsteds with 

 the addition of a scheme of numbering to facilitate the complete record 

 of each case, colored signal flags, and finally a lantern showing colored 

 lights, which he afterwards greatly improved. 



This was at first used only in dubious cases, but more and more 

 reliance became gradually to be placed on it until in its later form it 

 was accepted as the standard test on many important railway systems. 

 The improvement in Dr. Williams' lantern over the colored glass test 

 introduced by his old instructor Professor Donders, and improved by 

 others, lay in the completeness of its correspondence to the working 

 condition of observing railway signals. It includes use of colored 

 spots of varying size and intensity, of those involving contrast of 

 normal and signal lights, the introduction of confusion colors and such 

 as might come from the modifying influence of atmosphere, and the 

 presentation of combinations so diverse as to give absolutely no 

 chance of passing the test by memorizing, or by the judgments of 

 intensities on which the color blind often consciously or unconsciously 

 depend. It has been found in practice that this lantern can detect 

 many cases that escape examination by the worsteds, and therefore 

 much more effectively eliminate a class of employees which is doubly 

 dangerous from its unconsciousness of its own visual failings. 



Dr. Williams also did much to improve the general testing of vision 

 by the design of accurate and uniform test cards and the study of 

 their proper illumination for testing purposes. He also introduced 

 a very ingenious modification of the Snellen types for railway work 

 consisting of characters simulating the various semaphore positions 

 which the railwav man has to note. 



