This is but one proof against the existence of a lunar atmosphere, and, it may be argued, not a very conclusive one; because there may still be an atmosphere, though it be not sufficiently aqueous to condense into clouds and not sufficiently dense to obscure the lunar details. The probable existence of an atmosphere of such a character used to be inferred from a phenomenon seen during total eclipses of the sun. On these occasions the black body of the moon is invariably surrounded by a luminous halo, or glory, to which the name “corona” has been applied; and, further, besides this corona, apparently floating in it and sometimes seemingly attached to the black edge of the moon, are seen masses of cloud-like matter of a bright red colour, which, from the form in which they were first seen and from their flame-like tinge, have become universally known as the “red-flames.” It used to be said that this corona could only be the consequence of a lunar atmosphere lit up as it were by the sun’s rays shining through it, after the manner of a sunbeam lighting up the atmosphere of a dusty chamber; and the red flames were held by those who first observed them to be clouds of denser matter floating in the said atmosphere, and refracting the red rays of solar light as our own clouds are seen to do at sunrise and sunset. But the evidence obtained, both by simple telescopic observation and by the spectroscope, from recent extensively observed eclipses of the sun has set this question quite at rest; for it has been settled finally and indisputably that both the above appearances pertain to the sun, and have nothing whatever to do with the moon. [Illustration: Fig. 11.]