between looking through the small end of a cone-shaped roll of paper and
looking through the large end; in the former case the eye sees at once
all that is to be seen through the roll (supposed fixed in position), in
the latter the eye may be moved about so as to command the same range of
view, but _at any instant_ sees over a much smaller range.

To return to the arrangement actually employed, which is illustrated by
the common opera-glass. We see that the full illuminating power of the
telescope is not brought into play. But this is not the only objection
to the Galilean Telescope. It is obvious that if the part C D of the
object-glass were covered, the point P would not be visible, whereas, in
the astronomical arrangement no other effect is produced on the
visibility of an object, by covering part of the object-glass, than a
small loss of illumination. In other words, the dimensions of the field
of view of a Galilean Telescope depend on the size of the object-glass,
whereas in the astronomical Telescope the field of view is independent
of the size of the object-glass. The difference may be readily tested.
If we direct an opera-glass upon any object, we shall find that any
covering placed over a part of the object-glass _becomes visible_ when
we look through the instrument, interfering therefore _pro tanto_ with
the range of view. A covering similarly placed on any part of the
object-glass of an astronomical telescope does not become visible when
we look through the instrument. The distinction has a very important
bearing on the theory of telescopic vision.

In considering the application of the telescope to practical
observation, the circumstance that in the Galilean Telescope no real
image is formed, is yet more important. A real image admits of
measurement, linear or angular, while to a _virtual_ image (such an
image, for instance, as is formed by a common looking-glass) no such