Read the marginal text on the other side.

In proportion as the opening is smaller than the shaded body, so
much less will the images transmitted through this opening intersect
each other. The sides of images which pass through openings into a
dark room intersect at a point which is nearer to the opening in
proportion as the opening is narrower. To prove this let _a b_ be an
object in light and shade which sends not its shadow but the image
of its darkened form through the opening _d e_ which is as wide as
this shaded body; and its sides _a b_, being straight lines (as has
been proved) must intersect between the shaded object and the
opening; but nearer to the opening in proportion as it is smaller
than the object in shade. As is shown, on your right hand and your
left hand, in the two diagrams _a_ _b_ _c_ _n_ _m_ _o_ where, the
right opening _d_ _e_, being equal in width to the shaded object _a_
_b_, the intersection of the sides of the said shaded object occurs
half way between the opening and the shaded object at the point _c_.
But this cannot happen in the left hand figure, the opening _o_
being much smaller than the shaded object _n_ _m_.

It is impossible that the images of objects should be seen between
the objects and the openings through which the images of these
bodies are admitted; and this is plain, because where the atmosphere
is illuminated these images are not formed visibly.

When the images are made double by mutually crossing each other they
are invariably doubly as dark in tone. To prove this let _d_ _e_ _h_
be such a doubling which although it is only seen within the space
between the bodies in _b_ and _i_ this will not hinder its being