502 The Lunar Atmosphere (October, 
and at present there is no evidence that the moon’s volcanic 
energies are not tenfold more active than on the earth, and 
the importance of the determination of the probability of a 
lunar atmosphere is evident. 
There is another question which is completely controlled 
by that of a lunar atmosphere. Most selenographers, and 
even the most cautious, Beer and Madler, have recognised 
on the moon indications of the existence of some form of 
vegetable life, but, again, vegetation and no atmosphere is 
inconceivable ; waiving the question of moisture as carrying 
us into too large a field, and granting the existence of an 
atmosphere, and most interesting problemsare opened. For 
the researches of M. Paul Bért have shown that vegetable 
life is possible under atmospheric pressures only one-twentieth 
of the normal terrestrial one, and not so much more than is 
possible on the moon ; considering, therefore, the wonderful 
adaptability to circumstances of vegetable life, who can say 
it is impossible at a pressure less than this? The moon’s 
temperature need be no bar, for, although a fatal objection 
near the equator, it is no longer so nearer the poles. 
Similarly with the many more recondite problems that 
the study of the moon’s surface has given rise to—the streaks 
and rays of the full moon, the variability of the markings, 
floor, and craterlets of Plato, the white cloud-like coverings 
of Linne and y Posidonius, the difference in the polarisation 
of the light reflected from the plains and mountains observed 
by Secchi, in their photographic power found by De la Rue, 
in their colour by Birt, in their visibility by Schmidt—all 
involve in their consideration the question as to the existence 
of a lunar atmosphere. It must be at once admitted that 
the existence of a lunar atmosphere facilitating the probable 
explanation of these phenomena proves nothing; for the 
circumstance, that granted a basis and all can be explained, 
is no demonstration of the truth of that basis. But neither 
must it be assumed, as is so often done, that the possibility of 
accounting for phenomena on other grounds shows the 
proposed explanation to be probably wrong; the nature of 
the two must always be considered. For example, it is 
possible to explain the photometric results of Zollner as to 
the brightness of the moon by assuming either that the lunar 
surface is covered by a multitude of very minute steep 
conical hills, each of a definite slope, and arranged in a 
complex manner round two common foci, or else that it is 
covered with extremely small regular furrows of uniform 
angle, running from pole to pole at proper distances; and 
these hypotheses explain most satisfactorily all Zollner’s 
