646 Royal Society. 



character and appearance of the dissolving views effected by the magic 

 lanthom. The frame of the window or aperture, and the cross- 

 bars, were always pictured in colours different from those of the 

 panes, besides a fine marginal line of another colour dividing the 

 glass and the frames. These consequential colours, constitute, as is 

 well known, a remarkable feature in the phaenomena. They have 

 generally a certain complementary relation, or tendency to such, to 

 the colours of the primary picture. Thus in the clear green or blue 

 spectrum of a window, derived from strong illumination, the re- 

 mainder of the field of the eye will generally, in the first instance, 

 be covered with a ground of glowing crimson, with cross-bars simi- 

 lar, and purple edgings ; and when the picture changes to crimson 

 or red, the antagonistic tint will also change, perhaps to purple, or 

 orange or brown. The original spectra were found to fade away at 

 intervals, often of tolerable equality, such as of eight or nine seconds, 

 disappearing perhaps for two or three seconds, and then reappearing 

 under, generally, some change of shade or tint, through an extent of 

 very numerous repetitions. The changes of colour from the bright 

 or emerald green, as very frequently traced, went rapidly through 

 yellow-green, yellow, orange, red, scarlet, crimson and brown, or olive. 

 And this series, it is observable, is particularly accordant, in respect 

 to the principal or fundamental colours, with that of the prismatic 

 spectrum from green to yellow, orange and red. These visual photo- 

 graphs, besides having the sharpest definition, and often the most 

 brilliant illuminated colours, were found to possess, under strong 

 intensities of impression, a remarkable degree of permanency — ex- 

 tending sometimes to endurance for an hour or longer after the act 

 of gazing. 



Investigations on the relation of the photochromatic developments 

 to the time of gazing, gave results in many respects corresponding 

 with those derived from differences in the degree of external light. 

 Thus the higher colours of the spectral series elicited by strong light, 

 could, within certain limits, be also developed by more continuous 

 gazing with inferior light : so that the pink- coloured spectrum de- 

 rived from ten to twenty seconds' gazing in low degrees of light, 

 could be elicited by a single glance under bright sunshine. The 

 results, therefore, were clearly in relation to the intensity of the im- 

 pression ; and, taken in the form of a general proposition, we shall 

 not be far wrong, perhaps, in considering the intensity of impression 

 as the product of the time of gazing into the relative quantity of 

 light admitted by the aperture. 



The relation of the colours primarily elicited to the intensity of 

 the impression, yielded (comparatively and roughly taken) the fol- 

 lowing series, — crimson-pink, purple-pink, purple, blue, green, the 

 latter being the produce of the highest intensity tried. 



As in the foregoing researches, the relative degrees of light were 

 but broadly assumed, whilst the comparative experiments comprised 

 a variety of differences affecting the photochromatic results, 

 another series of experiments on the simple effects of degrees of 

 light was instituted, in which all these other differences were elimi- 



