centres of the
three stars are on the same line with the earth in the middle. But they
are merely annular eclipses, during which the earth, thrown like a
screen across the solar disc, allows the greater part to be seen."

"Why is there no total eclipse?" asked Nicholl. "Is it because the cone
of shade thrown by the earth does not extend beyond the moon?"

"Yes, if you do not take into account the refraction produced by the
terrestrial atmosphere, not if you do take that refraction into account.
Thus, let _delta_ be the horizontal parallax and _p_ the apparent
semidiameter--"

"Ouf!" said Michel, "half of _v_ zero square! Do speak the vulgar
tongue, man of algebra!"

"Well, then, in popular language," answered Barbicane, "the mean
distance between the moon and the earth being sixty terrestrial radii,
the length of the cone of shadow, by dint of refraction, is reduced to
less than forty-two radii. It follows, therefore, that during the
eclipses the moon is beyond the cone of pure shade, and the sun sends it
not only rays from its edges, but also rays from its centre."

"Then," said Michel in a grumbling tone, "why is there any eclipse when
there ought to be none?"

"Solely because the solar rays are weakened by the refraction, and the
atmosphere which they traverse extinguishes the greater part of them."

"That reason satisfies me," answered Michel; "besides, we shall see for
ourselves when we get there. Now, Barbicane, do you believe that the
moon is an ancient comet?"

"What an idea!"

"Yes," replied Michel, with amiable conceit, "I have a few ideas of that
kind."

"But that idea does not originate with Michel," answered Nicholl.

"Then I am only a plagiarist."

"Without doubt," answered Nicholl. "According to the testimony of the
ancients, the Arcadians pretended that their ancestors inhabited the
earth before the moon became her satellite. Starting from this fact,
certain _savants_ think the moon was a comet which its orbit one day
brought near enough to the earth to be retained by terrestrial
attraction."

"And what truth is there in that hypothesis?" asked Michel.

"None," answered Barbicane, "and the proof is that the moon has not kept
a trace of the gaseous envelope that always accompanies comets."

"But," said Nicholl, "might not the moon, before becoming the earth's
satellite, have passed near enough to the sun to leave all her gaseous
substances by evaporation?"

"It might, friend Nicholl,