OPTICAL DECEPTION. 39 
fore him; and that without remaining confcious fora mo- 
ment that he has attended to this circumftance: But the 
effe& remains, and will influence his judgment. If on 
looking at a brick hearth he perceives that thofe lines which 
divide the bricks have a dark fhade on that fide oppofite 
to the light, and a bright ftreak on the contrary fide next 
to the light, he muft at the fame time perceive that they 
have the property which he has conftantly obferved in 
ridges, not in furrows. And fince the appearance ot the 
hearth will be fuch, through the glaifes, in confequence of 
their inverting the fituation of its feveral parts, with refpe& 
tothelight, the obferver will inftantly pronounce the chinks 
between the bricks, and every little cavity in them, to be 
fo many perfect elevations above the common {urface, nor 
can any effort of the mind corre& the imagination or alter 
the appearance. 
_ Though I was well fatisfied of the truth of this expla- 
nation, I refolved neverthelefs to bring it to the teft of ex- 
periment, which I did in the following manner. 
In order to give my experiment fair play, I fhut all the 
windows of my chamber excepting one dire@ly oppofite 
tothe chimney. I then took the tube, with two convex 
glaffes, and looking through it at the hearth, all the bricks 
appeared depreffled and the clefts between them elevated, 
as ufual. I then placed a looking-glafs againft the chim- 
ney back, fo that it reflected the light from the window 
upon the hearth, and fet up a {mall board before the hearth 
to intercept the direct light of the window from it. Then 
looking at the hearth through the glaffes, I was much 
pleafed to find it appear in its natural ftate, with the bricks 
elevated. Ithen fat down ona chair at the edge of the 
hearth, and looking through the tube which I held to my 
eye with one hand, whilft with the other I moved the 
board fo as to make it fometimes intercept the dire&t light 
of the window, and at other times the refleCted light of 
the looking-glafs, I conftantly found that when the hearth 
was: 
