Well, when stars are occulted by the moon their rays, on grazing the edge of her disc, do not show the least deviation nor offer the slightest indication of refraction.

It follows, therefore, that the moon can have no atmosphere." Every one looked at the Frenchman, for, this once admitted, the consequences were rigorous.
"In fact," answered Michel Ardan, "that is your best if not only argument, and a _savant_, perhaps, would be embarrassed to answer it.

I can only tell you that this argument has no absolute value because it supposes the angular diameter of the moon to be perfectly determined, which it is not.

But let us waive that, and tell me, my dear sir, if you admit the existence of volcanoes on the surface of the moon." "Extinct volcanoes, yes; volcanoes in eruption, no." "For the sake of argument let us suppose that these volcanoes have been in eruption for a certain period." "That is certain, but as they can themselves furnish the oxygen necessary for combustion the fact of their eruption does not in the least prove the presence of a lunar atmosphere." "We will pass on, then," answered Michel Ardan, "and leave this series of argument and arrive at direct observation.

But I warn you that I am going to quote names." "Very well." "In 1715 the astronomers Louville and Halley, observing the eclipse of the 3rd of May, remarked certain fulminations of a remarkable nature.
These jets of light, rapid and frequent, were attributed by them to storms in the atmosphere of the moon." "In 1715," replied the unknown, "the astronomers Louville and Halley took for lunar phenomena phenomena purely terrestrial, such as meteoric or other bodies which are generated in our own atmosphere.