SUBJECTIVE COLOURS AND THE AFTER-IMAGE. 31 of a coloured image by central excitation of the same colour ; (2) the appearance of patches of the visualised colour on images of other colours. The writer found that in her own case these patches of red on the blue image or of blue on the green were not infrequently present when there was no attempt to visualise red or blue ; and interpreting the reports of the other subjects from her own experience, concludes that they were merely not noticed except when the subject had their colour specially in mind. ^ To the second class of results belong those cases (1) where the image of the visualised colour was brought on sooner than usual, the preceding stage being very considerably shortened ; and (2) where it was held longer than usual, the succeeding stage being as it were thrust back when it began to appear. Here we have apparently not merely an in- tensification of sensations already peripherally excited, but either the production by central excitation of a sensation equal in intensity to the peripherally excited image, or the check- ing of a peripheral by a central excitation. Keally, however, this second class of results is merely an exaggerated instance of the same principle manifested in the first class. Here, too, in all probability, a peripheral excitation is strengthened by a central excitation, only the former is so faint that the resulting sensation seems to have been produced wholly by central excitation. There is always a tendency for the colour of the after-image to fluctuate between stages. Some time before the green stage, for instance, actually passes into the red stage, the red process assumes more or less prominence, and after the red stage has appeared there are still traces of the green process surviving. If the subject is visualising red, the nascent red excitation will be strengthened and made predominant ; if green, the fading traces of green will be held and intensified against the in-coming red. In the case of the subject D. the effect of centrally excited colours on the course of the after-image was very striking. It amounted in many experiments to an entire transforma- tion of the colour series. The subject's own report of one or two experiments will give the best idea of the extent of this effect. The ordinary sequence of colours in the " ring- ing-ofif " was for D. perfectly regular ; blue positive, green positive, red negative, dark blue negative. When in the next experiment, after an interval sufficient to remove all traces of the preceding image from the retinal field, the subject was asked to visualise red, the course of the colour changes was described as follows : First a red image with
