... Fux

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... stream

It is implicitly assumed here that the azimuthal velocity of the Sun is close to the circular velocity of the axisymmetric part of the Galactic potential. This is probably correct within 5-10 kms^-1, as will be argued in Sect. 11.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... potential

As pointed out in D2000, the Figs. 2 and 4 in Fux (2000) wrongly display the OLR as a contour of constant H. This mistake is rectified here throughout the paper.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...1968)

In a realistic disc galaxy, the chaotic orbits not confined within corotation, and in particular the hot chaotic orbits, are not really bound but their escaping timescale is much larger than the Liapunov divergence timescale (see also Sect. 7). This is especially true in the case of our infinite mass logarithmic potential.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...$R/R_{\hbox{\tiny OLR}}\approx 1.1$

More precisely at $R_{\hbox{\tiny OLR}}/R=0.9$.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...u

Doubling the average integration time can reinforce the bimodality, as shown in Fig. 20a.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... resonance

This could be a consequence of the fact that the linear radial oscillation frequency of the quasi-periodic orbits around the least eccentric stable x₁(2) orbits near the OLR radius, i.e. the non-axisymmetric analogue of the epicycle frequency for these periodic orbits, is close to the radial frequency $w_{\rm R}$ of the x₁(2) orbits themselves. A similar argument may also hold for the slow phase mixing noticed in Fig. 11 within the resonant part of the quasi-x₁(1) region.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... contour

In particular, at $R_{\hbox{\tiny OLR}}\approx 1.1$ and $\varphi =25^{\circ }$, where the low angular momentum mode has a chaotic origin, the v squashing of the bimodality reported by D2000 when decreasing his rotation curve parameter $\beta$ is perhaps not the consequence of a local change of the circular velocity slope, but of an implied lower relative value of $\Phi_{\rm eff}(L_{1/2})$.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.