(696) 
Sun-beams upon the Wall , erpecially when the Paper-circle 
was indifferently well illuminated. This Image indeed appear- 
ed contrary to the former as to the fcituation of Colours, that 
is, the Scarlet appearing above, the Violet below , though bur 
faint. But this I wasnot furprized at, having obferv'd upon 
diffefling the eye,that obje6s are painted on the Retina after a 
contrary pofture to what they appear to Sight. Having thus 
rendred the Coloured image much more tiadabie than for- 
merly it was, I conceived good hopes of fome further dif- 
covery in the point mentioned. 
In purfuance then of my former fufpicion, having fixed my 
V Tab ir ^^'^"^ ^ fteady pofture , I caufed the paper C to be 
¥\g.lu6\ applied clofe up to the Paper-circle 4^^: whereupon 
the former Violet^, and Scarlet colour of C vaniflied 
into whitenefs. Next, 1 removed the mentioned Circle from 
the Shuts,and placfed it in the open window, fupported only by 
the edge d : whereupon, to my aftoniflimentjall the former Co- 
lours exchanged poftures in the Retina^ the Scarlet now ap- 
pearing below, the Violet above 5 the intermediate Colours 
fcarce difcernible. And here, on the by, 'tis very remarkable, 
that, during this Obfervation, I clearly perceived both Blew* 
and Scarlet-light to be tranfparent, I being able to difcern 
feveral objefis through both , namely Steeples oppofic to my 
window. Whence it follows , that thefe Colours do in great 
part arife from the neighbouring light. Laftly, 1 placed the 
Paper-circle anew, foas the one half h was faftened to the 
Shuts, the other femicircle a being expofed to the open Air. 
Whereupon the femicirele a became bordered with Violec 
"above, Scarlet below ; but the other femicirele b quite coti- 
irarv. Hence I make the following Inferences. 
Fir/?5Thatnot only the Light reflefled from the Pa per- circle, 
but alfo from the ambient Air , hath great influence upon the 
Coloured imag€,efpecia!ly as to the Violet and Scarlet colours. 
Whence perchance it will not hcreafier feem ftrange, that the 
colowx ^ A Sf elt rum oxi the Wall is lb long, but only that the 
breadth is not greater* Secondly^ Were there a more luminous 
body behind the Sun, we fliould in all likelyhood have the co- 
lours of the Spe51rHm \x\ a contrary fcituation to what they 
appear in at prefent : Whence {thirdly^ i t feems to follow/ bat 
the 
