the Sun there have often been noticed
along the limb of the Moon fringes of colour, and dark and bright bands.
This might not necessarily be a real appearance for it is conceivable
that such traces of colour might be due to the telescopes employed not
having been truly achromatic, that is, not sufficiently corrected for
colour; but making every allowance for this possible source of mistake
there yet remains proof that the colour which has often been seen has
been real.

As to whether the Moon's limb can be seen during a partial eclipse, or
during the partial phase of what is to be a total eclipse, the evidence
is somewhat conflicting. There is no doubt that when the totality is
close at hand the Moon's limb can be seen projected on the Corona
(presently to be described); but the question is, whether the far-off
limb of the Moon can be detected in the open sky whilst something like
full daylight still prevails on the Earth. Undoubtedly the preponderance
of evidence is against the visibility of the Moon as a whole, under such
circumstances; but there is nevertheless some testimony to the contrary.
A French observer, E. Liais, said that three photographic plates of the
eclipse of 1858 seen in S. America all showed the outer limb of the
Moon with more or less distinctness. This testimony, be it noted, is
photographic and not visual; and on the whole it seems safest to say
that there is very small probability of the Moon as a whole ever being
seen under the circumstances in question.

What has just been said concerns the visibility of the Moon during quite
the early, or on the other hand during quite the late, stages of a total
eclipse. Immediately before or after totality the visibility of the
whole contour of the Moon is a certain fact; and the only point upon
which there is a difference of opinion is as to what are the time-limits
beyond which the Moon must not be expected to be seen. The various
records are exceedingly contradictory: perhaps the utmost that can be
said is that the whole Moon must not be expected to be visible till
about 20 minutes before totality, or for more than 5 minutes after
totality--but it must be admitted that these figures are very uncertain
in regard to any particular eclipse.

It has been sometimes noticed when the crescent of the Sun had become
comparatively small, say that the Sun was about 7/8ths covered, that the
uncovered portion exhibited evident colour which has been variously
descr