78 



image of the yellow paper insulated by the glass will be seen divided 

 into a yellow half, projecting beyond the blue fringe, and a purplish- 

 or bluish-white one within it, hardly to be distinguished from the 

 image of the white paper, of which it seems a continuation, and 

 which through the glass in question appears a pale blue. This same 

 purplish tint was observed to arise also under the following circum- 

 stances : Laying down in a good diffused light a paper of an ex- 

 ceedingly beautiful ultramarine blue, and beside it, and somewhat 

 overlapping it, another coloured with the same yellow chromate, I 

 set upon the line of junction a sheet of glass inclined to the plane of 

 the papers upwards towards the eye, so as to allow the blue to be 

 seen by transmitted light, while the yellow reflected from the glass 

 was at the same time received into the eye. By varying the inclina- 

 tion of the glass, the yellow reflexion could be made more or less 

 vivid, so as either to be nearly imperceptible or quite to kill the 

 blue of the paper. But at no stage of its intensity, gradually in- 

 creased from one to the other extreme, was the slightest tendency to 

 greenness produced. The colour passed from blue to yellow, not 

 through green, but through a pale uncertain purplish tint, not easy 

 to describe, but as remote from green as could be well imagined. 



Of course in all such experiments one eye only must be used. 

 Stereoscopic superposition of colour, which at first sight would ap- 

 pear readily available, does not satisfy the requisite conditions, and 

 yields no definite results. 



The conclusions from these facts may be summed up as follows : 

 1st. That in no case can the sensation of green be produced by the 

 joint action on the eye of two lights, in neither of which, separately, 

 prismatic green exists ; 2ndly. That the joint action of two lights, 

 separately producing the most lively sensations of blue and yellow, 

 does not give rise to that of green, even when one of them contains 

 in its composition the totality of green light in the spectrum ; 

 and, 3rdly. That all our liveliest sensations of yellow are produced by 

 the joint action of rays, of which those separately exciting the ideas 

 of red and green form a large majority ; and that a decided yellow 

 impression is produced by the union of these only. 



From these premises it would seem the easiest possible step to 

 conclude the non-existence of yellow as a primary colour. But this 

 conclusion I am unable to admit in the face of the facts, 1st, that 



