Olftrvathm on Cohiired Sl.-aJowf. 1O5 



Having found that the {hadows correfponding to two like wax candles were coloured, 

 the one blue and the other yellow, by interpofing a (heet of yellow glafs before one of 

 them ; I now tried what the effeA would be when blue glafs was made ufe of inftead of 

 yellow, and I found it to be the fame ; the fliadows were ftill coloured, the one blue and 

 the oth-r yellow, with the difference, however, that the colours of the (hadows were re- 

 verftd, that which with the yellow glafs was before yellow being now blue, and that 

 which was blue being yellow. 



I afterwards tried a glafs of a bright amethyft colour, and was furprifed to find that the 

 "flndows ftill continued to be coloured blue and yellow. The yellow, it is true, had a dirty 

 purple call; but the blue, though a little inclining to green, was neverthelefs a clean, 

 bright, decided colour. . . . , 



Havin- no other coloured glafs at hand to pulh thefe particular enquiries farther, I now 

 removed'the candles, and, opening two holes in the upper ^,arts of the Gutters of two 

 neighbouring windows, I let into the room, from above, two beams of light from dif- 

 ferent parts of the heavens ; and placing the inftrument in fuch a manner that two dillina 

 (hadows were projeaed by the cylinder upon the paper, I was entertained by a fucceffioii 

 of very amufing appearances. The (hadows were tinged with an infinite variety of the moft 

 unexpefted, and often moft beautiful colours, which, continually varying, fometimes flowly 

 and fometimes with inconceivable rapidity, abfolutely fafcinated the eyes, and, commanding 

 the moft eager attention, afforded an enjoyment as new as it was bewitching. It was a 

 windy day, with flying clouds -, and it feemed as if every cloud that paffed brought with it 

 another complete fucceffion of varying hues and moft harmonious tints. If any colour 

 could be faid to predominate, it was purple ; but all the varieties of browns, and almoft all 

 the other colours I ever remembered to have feen, appeared in their turns; and there were 

 even colours which feemed to me to be perfedly new. 



Refleaing upon the great variety of colours obferved in thefe laft experiments, many of 

 which did not appear to have the leaft relation to the apparent colours of the light by which 

 they were produced, I began to fufpeft that the colours of the fliadows might in many 

 cafes, notwithftanding their apparent brilliancy, be merely an optical deception, owing to 

 contraft, or to fome effeft of the other neighbouring colours upon the eye. To determine 

 this faa by a direft experiment, I proceeded in the following manner : Having, by m.aking 

 ufe of a flat ruler, inftead of the cylinder, contrived to render the fliadows much broader, I 

 (hut out of the room every ray of day-light, and prepared to make the experiment with 

 two Argand's lamps, well trimmed, and which were both made to burn with the grcateft 

 pofl-.ble brilliancy; and having afl-ured myfelf that the light they emitted was prcc.fely of 

 the fame colour, by the fliadows being perfedly colourlefs which were projeded upon the 

 white paper, 1 direfled a tube about 12 inches long, and near an inch in diameter, hned 

 with black paper, againft the centre of one of the broad fliadows ; and looking through 

 this tube with one eye, while the other was clofed, 1 kept my attention fixed upon the 

 fliadow, while an affiftant repeatedly interpofed a ftiect of yellow glafs before the lamp, 

 whofe light corrcfponded to the fliadow I obferved, and as often removed it. The refult 

 of ,hc experiment was very ftriking, and fully confirmed my fufpicions with refpea to the 

 fallacy of many of the appearances in the foregoing experiments. So far from being 

 able to obfcrve any change in the (liadow upon which my eye was fixed, 1 was not able 

 Vol- I. — June / 797. P 



even 



