Since the minor planet population as a whole shows a mean proper eccen
tricity, e(p), near 0.1, this paper starts by asking why the Hilda gro
up, some sixty objects whose mean motions lie very close to 3/2 times
that of Jupiter, all have e(p)>0.1. The Hildas execute small oscillati
ons, usually called librations, about this 3/2 ratio and these oscilla
tions have a well defined frequency v, which can, for example, be dete
rmined from periodic variations in a minor planet's semimajor axis or
eccentricity. Another basic frequency, omega, is given by the apsidal
motion. We show that orbits for which the two frequencies differ by sm
all integral or half integral multiples [i.e., v=(n/2)omega, n=3,4,5..
..] are clearly chaotic, with Lyapunov times short enough to suggest i
nstability. Such orbits lie within the 3:2 mean motion resonance and a
ll have e(p)<0.1. Although the low e(p) region is streaked with chaoti
c zones, they are generally narrow, hence generating many chaotic orbi
ts, and possibly depopulating the entire e(p)<0.1 region, would seem t
o require an added process that would slightly alter the loci of the r
esonances over time. At higher multiples of the two frequencies, parti
cularly v=6 omega, orbits are remarkably regular, even when the amplit
ude of libration of successive conjunctions of Jupiter and a minor pla
net about the latter's pericenter is even as large as 90 deg. In fact,
numerical results show that all of the real Hildas having large libra
tions lie at or very close to such commensurabilities, while hypotheti
cal bodies that do not are markedly chaotic. (C) 1995 American Astrono
mical Society.