The experiment Descartes 
described depends on 
understanding how  the lens 
of an eye—–or any convex 
lens—–creates an image.  The 
lens, when it is shaped for 
optimal focus, bends, or
refracts, all the rays of light 
emanating from the same source
to focus them at a single point. 
From any point on an object,
light is focused by the lens on a 
single point on the retina.  In 
this way, rays emanating from 
the outline of an object are 
focused on the retina so as to
form a corresponding outline 
although, as we shall see, that 
outline will be different from 
the object's outline —– distorted, 
one might say —– and will be 
different from time to time as 
well.

    Because we are concerned 
with perceptual constancy in 
this chapter, not the sharpness 
of an image (which affects 
acuity), we can disregard 
variations in the shape of the 
lens and the action of the 
cornea, the eye's transparent 
outer covering which 
cooperates with the lens to 
focus light, and assume that 
only one ray of light emanates 
from any point of an object. For 
our purposes, then, the image 
on the retina is equivalent to 
that formed on a screen in a 
camera obscura. A camera 
obscura consists of a pinhole 
opening, instead of the larger, 
variable aperture of the eye or
an ordinary camera, and a 
screen for direct viewing, 
rather than the retina or the 
chemically treated, light-
sensitive surface of film. 
Because the pinhole is small, 
for all intents and purposes we 
can say that only one ray of 
light passes through it from any 
point in a scene. Thus, the 
problem of focusing does not 
arise. The shape of the screen 
of the camera obscura is flat, 
not spherical as it is in the eye. 
However, this difference does 
not affect the formation of an 
image or the problem of 
constancy under discussion.