104 Ohfftvaikns on CJoured Shadow!. 



yellow thsn the cnnHlc-liglit. In the courfe of thefe experiments I obferved that different 

 Oiades of yellow given to the tlay-liglit produced very difFerent and often quite unexpeifled 

 effefls: thus one ftieet of the yellow glafs interpofed before the beam of day-light changed 

 the yellow fliadow to a lively violet colour, and the blue fliadow to a light green ; two flieets 

 of the fame glafs nearly deftroyed the colours of both the ftiadows ; and three flieets 

 changed the fliadow which was originally yellow to blue, and that which was blue to a 

 purplilh yellow colour. 



When the beam of day-light was made to pafs through a flicet of blue glafs, the colours 

 of the fliadows, the yellow as well as the blue, were improved and rendered in the higheft 

 degree clear and brilliant -, but when the blue glafs was placed before the candle, the colours 

 of ihe Ihadows were very much impaired. 



In order to fee what would be the confequence of rendering the candle-light of a ftiU 

 deeper yellow, 1 interpofed before it a flieet of yellow or rather orange-coloured glafs ; 

 ■when a very unexpefted and moil beautiful appearance took place. The colour of the 

 yellow Ihadow was changed to orange, the blue Ihadow remained unchanged, and the 

 whole furface of the paper appeared to be tinged of a mod beautiful violet colour, approach- 

 ing to a light crimfon or pink ; almoft exadlly the fame hue as I have often obferved the 

 diftant fnowy mountains and valleys of the Alps to take about fun-fet. Is it not more than 

 probable that this hue is in both cafes produced by nearly the fiime combinations of coloured 

 light ? In the one cafe it is the white fnow illuminated at the fame time by the pureft light 

 of the heavens, and by the deep yellow rays from the weft ; and in the other it is the white 

 paper illuminated by broad day-liglit, and by the rays from a burning candle, rendered ftill 

 more yellow by being tranfmittcd through the yellow glafs. The beautiful violet colour 

 which fprcads itfelf over the furface of the paper will appear to the greateft advantage if 

 the pane of orange-coloured glafs be held in fuch a manner before the candle, that only a 

 part of the paper, half of it, for inftancc, be affecled by it, the other half remaining white. 



To make thefe experiments with more convenience, the paper, wliich may be about 

 eight or ten inches fquare, fliould be palled or glued down upon a flat piece of board, fur- 

 niflied with a ball and focket upon the hinder fide of it, and mounted upon a (land, and 

 the cylinder fhould be faftened to a fmall arm of wood or of metal, projedling forward 

 from the bottom of the board for that purpofe. A fmall ftand, capable of being made 

 higher or lower as the occafion requires, (hould liLcwife be provided for fupporting the 

 caidlc; and if the board with the paper fallened upon it be furrounded with a broad black 

 frarre, the experiments will be fo much the more (Iriking and beautiful. For dill greater 

 convenience, 1 have added two other (lands for holding the coloured glafs through which 

 the light is occafionaily made to pafs in its way to the white furface upon which the fhadows 

 are projeiTled. It will be hardly ntceflary to add, that, in order to the experiments appearing 

 to the greateft advantage, all light which is not abfoiutely nccefTary to the experiment mull 

 be carefully fhut out. 



Having fitted up a little apparatus, according to the above direftions, merely for the pur- 

 pole of profecutjng thefe enquiries rcfpedting the coloured fhadows, I proceeded to make a great 

 v.iriety of experiments, fome with pointed views, and others quite at random, and merely 

 in hopes of making fome accidental difcovery that might lead to a knowledge of the caufes. 

 of appeaiajiccs which ftill feemed to me to be enveloped iu much obfcurity and uncertainty. 



Having 



