Binney and Tremaine, 4.4, 4.7.1

Optional reading:

Derivations of the collisionless Boltzmann equation (called by other names) are given by:

Chandrasekhar, S. 1942, Principles of stellar dynamics (Chicago: University of Chicago Press)
Ogorodnikov, K. F.  1965, Dynamics of stellar systems (Oxford: Pergamon)

You can also look in almost any statistical mechanics book for Liouville's equation or Liouville's theorem,
although this normally is derived for a 6N-dimensional phase space (compare 8.1.2 in Binney and Tremaine)

Richstone, D. O., and Tremaine, S. 1986, AJ 92, 72 - determining mass-to-light ratios by core fitting (King's method)

Chandrasekhar, S. 1939, An introduction to the study of stellar structure (Chicago: University of Chicago Press) - a comprehensive analytic treatment of the isothermal sphere

King, I. 1966, AJ 71, 64 - King models

Gunn, J. E., and Griffin, R. F. 1979, AJ 84, 752 - a classic study of the dynamics of the globular cluster M3 using Michie models

A more detailed discussion of anisotropic models (constant-anisotropy models and Osipkov-Merritt models) is given in Section 4.3.2 of the second
edition of Binney & Tremaine

Tremaine, S., Henon, M., and Lynden-Bell, D. 1986, MNRAS 219, 285 - entropy in stellar systems

The fact that only spherical systems can have distribution functions that depend only on energy follows
from Lichtenstein's theorem (non-rotating barotropic stars must be spherical). See:
    Tassoul, J.-L. 1978, Theory of rotating stars (Princeton: Princeton University Press)
    Lindblom, L. 1977, J. Math. Phys. 18, 2352