In 1790 Delambre, using all these observations, prepared tables for
computing its position. These worked well enough for a time, but at
last the differences between the calculated and observed longitudes of
the planet became serious. In 1821 Bouvard undertook a revision of the
tables, but found it impossible to reconcile all the observations of
130 years (the period of revolution of Uranus is eighty-four years).
So he deliberately rejected the old ones, expressing the opinion that
the discrepancies might depend upon "some foreign and unperceived
cause which may have been acting upon the planet." In a few years the
errors even of these tables became intolerable. In 1835 the error of
longitude was 30"; in 1838, 50"; in 1841, 70"; and, by comparing the
errors derived from observations made before and after opposition, a
serious error of the distance (radius vector) became apparent.

In 1843 John Couch Adams came out Senior Wrangler at Cambridge, and
was free to undertake the research which as an undergraduate he had
set himself--to see whether the disturbances of Uranus could be
explained by assuming a certain orbit, and position in that orbit, of
a hypothetical planet even more distant than Uranus. Such an
explanation had been suggested, but until 1843 no one had the boldness
to attack the problem. Bessel had intended to try, but a fatal
illness overtook him.

Adams first recalculated all known causes of disturbance, using the
latest determinations of the planetary masses. Still the errors were
nearly as great as ever. He could now, however, use these errors as
being actually due to the perturbations produced by the unknown
planet.

In 1844, assuming a circular orbit, and a mean distance agreeing with