the case of the constellations. To one who has
not noticed this before, it is really quite a revelation to compare the
appearance of one of the large constellations (Orion, for instance) when
high up in the sky and when low down. The widening apart of the various
stars composing the group, when in the latter position, is very
noticeable indeed.

[Illustration: FIG. 17.--Illustrating the author's explanation of the
apparent enlargement of celestial objects.]

Further, if a person were to stand in the centre of a large dome, he
would be exactly situated as if he were beneath the vaulted heaven, and
one would consequently expect him to suffer the same illusion as to the
shape of the dome. Objects fixed upon its background would therefore
appear to him under the same conditions as objects in the sky, and the
illusions as to their apparent enlargement should hold good here also.

Some years ago a Belgian astronomer, M. Stroobant, in an investigation
of the matter at issue, chanced to make a series of experiments under
the very conditions just detailed. To various portions of the inner
surface of a large dome he attached pairs of electric lights; and on
placing himself at the centre of the building, he noticed that, in every
case, those pairs which were high up appeared closer together than those
which were low down! He does not, however, seem to have sought for the
cause in the vaulted expanse. On the contrary, he attributed the effect
to something connected with our upright stature, to some physiological
reason which regularly makes us estimate objects as larger when in front
than when overhead.

In connection with this matter, it may be noted that it always appears
extremely difficult to estimate with the eye the exact height above the
horizon at which any object (say a star) happens to be. Even skilled
observers find themselves in error in attempting to do so. This seems to
bear out the writer's contention that the form under which the celestial
vault really appears to us is a peculiar one, and tends to give rise to
false judgments.

Before leaving this question, it should also be mentioned that nothing
perhaps is more deceptive than the size which objects in the sky appear
to present. The full moon looks so like a huge plate, that it astonishes
one to find that a threepenny bit held at arm's length will a long way
more than cover its disc.

[Illustration: PLATE VIII. THE MOON

From a photograph taken at the Paris