of Simultaneous Contrast- Colour. 



161 



green, though decided, is not strongly marked ; but if viewed 

 through the reflecting-tnbe the ring on the cobalt-blue appears 

 of a rather intense orange, and the ring on the emerald-green 

 appears of a rose-colour. When so viewed the difference in 

 these hues is strongly marked. 



Pieces of cardboard, painted with the following colours, 

 were arranged in pairs. Orange -yellow on the less refran- 

 gible side of D, and orange-yellow on the more refrangible side 

 of D. Greenish-yellow and yellowish-green, both made with 

 gamboge and Prussian blue. Purple of violet end of spec- 

 trum with added red, made with Hoffman's violet R.R.R., 

 and violet between Gf and H, Young and Mu'ller's funda- 

 mental colour. Place a pair of the colours, as given above, 

 side by side and on each put a grey ring, with a reflecting- 

 tube over each ring. When viewed by the unassisted eye 

 the differences in hue are slight and difficult to define ; but 

 when viewed in rapid succession through the adjoining tubes 

 the differences in hue are decided and may be readily matched 

 in colour with disks on the rotator. 



The Duration of the Time of 

 Vision necessary to perceive Con- 

 trast- Colour.— & square screen was 

 made, as shown in fig. 10, with a 

 square of cardboard in its centre, 

 8 centim. on the side, surrounded 

 by a square of translucent white 

 paper, 35 centim. on the side. 

 This screen was placed between 

 the petroleum-flame and the 

 window so that its sides were 

 equally illuminated. On the side 

 facing the window the central 

 square appeared cyan-blue. One 

 half of this square was covered 

 by a strip of paper so tinted with 

 orange-yellow that it appeared 

 grey when in juxtaposition with 

 the other blue half of the square. 



On a rotator were placed two 

 superposed black disks of the 

 same diameter (25 centim.). Near 

 the periphery of each disk was 

 cut out an annular slot, 4 centim. 

 long and 6 millim. wide ? as shown 

 in fig. 11. By turning one of these disks on the other the 

 opening of this slot could be varied from 4 centim. to nothing. 



Fi<J.10< 



Fig.ll, 



