disc, and judged to be permanent,
were made the basis of a chart. They were not indeed always equally well
seen. They disappeared regularly near the limb, and were at times veiled
even when centrally situated. Some of them had been clearly perceived by
De Ball at Bothkamp in 1882.[820]

Mr. Lowell followed Schiaparelli's example by observing Mercury in the
full glare of noon. "The best time to study him," he remarked, "is when
planetary almanacs state 'Mercury invisible.'" A remarkable series of
drawings executed, some at Flagstaff in 1896, the remainder at Mexico in
1897, supplied grounds for the following, among other, conclusions.[821]
Mercury rotates synchronously with its revolution--that is, once in 88
days--on an axis sensibly perpendicular to its orbital plane. No certain
signs of a Mercurian atmosphere are visible. The globe is seamed and
furrowed with long narrow markings, explicable as cracks in cooling. It
is, and always was, a dead world. From micrometrical measures, moreover,
the inferences were drawn that the planet's mass has a probable value
about 1/20 that of the earth, while its mean density falls considerably
short of the terrestrial standard.

The theory of Mercury's movements has always given trouble. In
Lalande's,[822] as in Maestlin's time, the planet seemed to exist for no
other purpose than to throw discredit on astronomers; and even to
Leverrier's powerful analysis it long proved recalcitrant. On the 12th
of September, 1869, however, he was able to announce before the Academy
of Sciences[823] the terms of a compromise between observation and
calculation. They involved the addition of a new member to the solar
system. The hitherto unrecognised presence of a body about the size of
Mercury itself revolving at somewhat less than half its mean distance
from the sun (or, if farther, then of less mass, and _vice versa_),
would, it was pointed out, produce exactly the effect required, of
displacing the perihelion of the former planet 38" a century more than
could otherwise be accounted for. The planes of the two orbits, however,
should not lie far apart, as otherwise a nodal disturbance would arise
not perceived to exist. It was added that a ring of asteroids similarly
placed would answer the purpose equally well, and was more likely to
have escaped notice.

Upon the heels of this forecast followed promptly a seeming
verification. Dr. Lescarbault, a physician residing at Orgeres, whose
slender oppo