1890.] 



Lord Rayleiylis Colour Box. 



147 



of the eye have been made by Glan, and it seems to me of importance 

 to repeat and extend his observations. One point, especially, is 

 worth clearing up : in how far are the complementary colours the same 

 for different eyes ? As far as I can judge, according to the view 

 just explained, they should be the same.* 



I have paid some attention to the possibility of a change in the 

 reading of the same observer at different times ; but it is very 

 difficult to obtain decisive evidence in this respect. It is a curious 

 fact, however, that the difference in the reading of the same observer 

 at different times will differ more from each other than one would be 

 led to expect from the consistency of his differing readings taken the 

 same day. As this happens, however, chiefly with observers whose 

 accuracy I have reason to doubt on other grounds, little value can be 

 attached to such differences. It is possible that a careless observer 

 remembers from observation to observation on the same day what he 

 has called a match, though it may be a trifle too green or too red, and 

 in this way the readings may gain an appearance of too great consis- 

 tency. I take, for instance, number 42, whose readings are charac- 

 teristic in this respect. 



Number of Mean 



Date. observations. reading. 



Dec. 4, 1889 5 



Feb. 13, 1890 5 -11 



April 23, 1890 8 -6 



May 6, 1890 8 -10 



Average difference 



between each 



reading and 



mean reading. 



2-0 

 1-9 

 2-6 

 3-1 



On December 4, amongst five readings, he never read lower than 

 4 ; while on February 13, he never read higher that 4. If I had 

 reason to believe number 42 a careful observer, I should take this as 

 a proof of a change in his eyesight ; I am afraid, however, no certain 

 conclusions can be drav/n from his observations. But there are some 

 other cases of marked differences. 



I have already stated that I have constantly compared my sight 

 with that of Mr. Hadley without being able to trace any decided 

 difference. 



The following are a few examples of readings taken at different 

 times with observers in whose judgment I can place reliance : 



* [This is not quite correct, partly owing to the indefinite nature of what we 

 call white. What I meant to say is this : If six pure colours, p, q, r, s, t, u, nre 

 related to each other by the equations ap + bq = cr + ds = et + fu, then a second 

 observer whose eyes only differed by a different absorption in their media should 

 be able to match the six colours so as to obtain the equations a'p + b'q = c'r + d' 

 = e't + fu. If the resulting colour for the normal eye is white p and q, &c., are 

 complementary colours, but the resulting colour for the second observer would not 

 necessarily appear white to him. May 23.] 



