We study the evolution of long-period comets by numerical integration of th
eir orbits, a more realistic dynamical approach than the Monte Carlo and an
alytic methods previously used to study this problem. We follow the comets
from their origin in the Oort cloud until their final escape or destruction
, in a model solar system consisting of the Sun, the four giant planets and
the Galactic tide, We also examine the effects of nongravitational forces
as well as the gravitational forces from a hypothetical solar companion or
circumsolar disk. We confirm the conclusion of Oort and other investigators
that the observed distribution of long-period comet orbits does not match
the expected steady-state distribution unless there is fading or some simil
ar physical process that depletes the population of older comets. We invest
igate several simple fading laws, Ne can match the observed orbit distribut
ion if the fraction of comets remaining observable after m apparitions is p
roportional to m(-0.6+/-0.1) (close to the fading law originally proposed b
y Whipple 1962); or if approximately 95% of comets live for only a few (sim
ilar to 6) returns and the remainder last indefinitely, Our results also yi
eld statistics such as the expected perihelion distribution, distribution o
f aphelion directions, frequency of encounters with the giant planets and t
he rate of production of Halley-type comets. (C) 1999 Academic Press.