30 MARGARET FLOY WASHBURN : dark green colour began to appear to force it back into red. It was found (1) that the red image was brighter and lasted much longer than usual ; (2) that the red colour would frequently be brought back even after the image had turned completely blue or green ; and (3), by the subject 0., that when the image had become very dark and indistinct, and was on the point of disappearing, it could be made to look reddish. The subjects were told in the B. II. experiments to shorten the red image and to bring on the dark blue negative stage as soon as possible. This result was obtained, and the blue negative image appeared brighter and lasted longer than usual. The subject 0. found that under these conditions the final stage of the image was quite distinctly blue instead of .green. The attempt in the G. II. experiments was to shorten the red stage, and to turn the final blue negative image green, if possible. This latter result was of course most strongly marked in the case of the subject 0., who found the final image distinctly green. The dark blue negative image was, however, noticeably tinged with green for the other two observers. Another result was that the lines on both red and blue negative images appeared bright green in colour. These results were not uniformly distributed throughout the experiments made in each class. Sometimes only one effect, e.g., the brightening of the blue or green image, would be noticed. Sometimes several effects would appear in the same experiment. In from one to two per cent, of the experiments the effort to visualise produced no result at all. The high degree of strain put upon the attention subjected the results to variations arising from causes too slight and obscure to be traced or calculated, and a table showing the relative frequency of the effect described above reveals an irregular distribution from which no general principles can be deduced. The results themselves seem to fall, broadly speaking, into two classes, which, however, are not separated by any sharply defined line. The first class comprises cases where the effect of visual- ising a certain colour was simply to intensify the traces of that colour already present in the retinal field. Central ex- citation of a given colour meant in these cases merely " calling attention " to that colour whenever it was actually present. In other words, it was an instance of what Prof. James calls " ideational preparation " for attention ; the facilitation of attention to a given conscious contents by expectation The results which fall into this class are : (1) the brightening
