4G0 Proceedings of Boyal Society of Edinlurgh. [july 16. 
have assumed that a slower speed is unfavourable to red, the edge 
should be violet and not purple. But there was a good deal of red 
in the light and no violet, as was shown by the spectroscope ; and 
the purple might have been got by the superposition of this red light 
upon a violet phase in our sense. Or again, it have might been due 
to the superposition of a contrast red, and a contrast violet upon a 
violet phase. Or lastly, it might have been due to contrast alone. 
If the third supposition be true, all we can say is, that the main 
phenomena (in the middle part) cannot be due to contrast, and that 
some of the appearances at the edges are certainly not explained by it. 
Experiment 8. — Is precisely analogous to Experiment 5, as is also 
Experiment 10. 
It is not difficult to apply to 9 an explanation similar to that 
already given for 3 and 6. 
11, 12, and 13 demand somewhat more notice. In Experiment 
11 we are virtually dealing with a mixture of green, blue, and violet 
light, the red end of the spectrum being almost cut out. The solu- 
tion in the beaker is practically a cylindrical lens, and therefore 
the central part of the image is very much brighter than the edges. 
Doubtless, owing to this, the changes of colour were better seen than 
in almost any of the other experiments. The bluish-violet appear- 
ance in the central part with the slow rotation must be explained as 
before, the violet having the advantage with the long time of 
stimulation (fig. 4). The edges at this stage are green, because 
where the light is less intense the speed of rotation is enough to give 
the second (greenish) phase, into which, with increased speed, the 
central part enters. With further increase of speed the reddish phase 
is not got, evidently, I think, because of the absence of red rays. 
In Experiment 12 there are no well-marked coloured edges, 
because with the feeble light there is not enough difference of 
intensity between the middle and the sides of the image to allow of 
appreciable differences of phase with the same speed. 
Experiment 13.' — This is interesting as an experiment with light 
of the same wave-length. The yellow light will stimulate chiefly 
the red and green fibres. There is therefore no trace of the first or 
violet phase. The other two are, however, to be seen with the 
light near the mirror (stronger stimulation), but not when it is 4 
feet away (weaker stimulation). Even with the strongest light used, 
