CONTRIBUTIONS TO SCIENCE. 479 



differs from the standard white light only by being too 

 bright or too dull, all the slits X Y Z must be opened or 

 closed in the same proportion. The result of any observa- 

 tion is expressed by an equation in which each particular 

 colour employed is represented by the corresponding division 

 of the scale placed in brackets, while the breadth of the slit 

 is written as the coefficient of this colour. Thus, the 

 equation 



18-5 (24) + 27 (44) + 37 (68) =W 



means that the breadth of the slit X was 18*5, as measured 

 by the wedge, while its centre was at the division (24) of 

 the scale; that the breadths of Y and Z were 27 and 37, 

 and their positions (44) and (68); and that the illumina- 

 tion produced by these slits was exactly equal, in the 

 estimation of the observer, to the constant white W. 



In the first instance, recorded in the paper above referred 

 to, the observations were made by Professor Maxwell himself 

 (J), and by Mrs. Maxwell (K). Professor Maxwell's com- 

 plexion was dark and his hair black ; Mrs. Maxwell was 

 extremely fair. The observations of each exhibited very 

 small errors from the mean, far less than would have been 

 expected, but there was always a difference apparent between 

 the two observers, and this always in the same direction. 

 A difference of this character was found by Maxwell to be 

 general between dark and fair persons, especially when the 

 light fell on the centre of the retina (the fovea centralis 

 or macula lutea). To this point reference will again be 

 made. 



By considering the errors in the several observations 

 between the standard colours, Maxwell showed that greater 

 accuracy is attainable in the case of red light than in that 

 of green or blue, and that variations in colour are more 

 easily detected than variations in brightness. 



One result of the observations is especially worthy of 

 notice. It follows from a comparison of the equations 

 obtained in the several parts of the spectrum that any tint 

 in the spectrum lying between the divisions (24) and (46) 

 of the scale could be exactly imitated by mixing in proper 



