Royal Society. . 647 



nated. In this series the quantity of light was varied by partial or 

 sectional screens of glass, or other transparent or semitransparent 

 substances. The results were particularly satisfactory, — different 

 tints or shades of colour being obtained by the same view and in the 

 same spectrum of a window- aperture, when different thicknesses of 

 window glass were placed in the several sections (six in number) into 

 which the aperture was divided. 



A beautiful example of the chromatic effects of partial and varied 

 screening of light on the optical spectrum elicited, was incidentally 

 obtained by viewing an aperture in the clouds, when the sky was 

 otherwise densely covered. After gazing for a few seconds on the 

 middle of this aperture, the spectrum, as viewed with gently closed 

 eyes, exhibited a singular variety of the richest tints according to the 

 differences in the light screened off by the edges of the cloud and 

 by certain little patches within the aperture. The spectrum resem- 

 bled the variegation and richness of colouring as elicited in certain 

 transparent or semitransparent substances when examined by polar- 

 ized light. 



The experiments on binocular and multiple spectra, as described 

 in Part I. of the author's paper, being repeated under degrees of 

 light adequate for yielding colour, gave pictures, in many cases, of 

 much interest and beauty. The multiple spectra, however, which 

 proved the most strikingly beautiful, were derived from the sun, 

 which was viewed indirectly, and on occasions, near setting, in winter, 

 when the intensity of its light was duly subdued by passing through 

 a dense condition of atmosphere. Under such circumstances, images, 

 sometimes in 100 to 150 repetitions, were impressed on the retina 

 by rapid glances at the sky immediately around the sun. These were 

 taken by quick movements of the head, winking intermediately, at 

 the rate of 60 to 1 20 impressions in the minute ; and the result, 

 when viewed with closed eyes, presented a splendid spectacle like a 

 cluster of coloured stars*, or rather of round planetary discs, brilliant 

 in green, yellow, orange, red, crimson and purple ! 



Besides the experiments thus far described, in which the spectral 

 images were viewed, for the most part, with gently closed eyes kept 

 steadily in the direction in which the objects were gazed on, — the 

 differences, which were often very remarkable, produced by alter- 

 ations in the quantity of light admitted into the chamber of the eye 

 whilst the image was viewed, were also investigated. Sometimes 

 the smallest change in the light thus transmitted was found to alter 

 greatly the character of the spectrum. In certain cases, the com- 

 pressing of the eyelids, or the mere passing of the hands betwixt the 

 eyes and the light, would serve to change a negative picture into a 

 positive, or the colours, as viewed in the usual way, into their com- 

 plementary tints. 



The paper concludes with a considerable series of deductions, ap- 

 plications and general results. — 1. As to the elucidation yielded by 

 these ocular spectra, of the theory of vision. — 2. Of the principles of 

 binocular and simple vision. — 3. Of the action of the retina for the 

 obliterating of impressed images, and the recovery of a normal con- 



