The main purpose of this paper is to present a theoretical scheme which des
cribes the solvation and transfer free energies of small molecules and rela
te them to solvent contributions in the biomolecular processes. Several pro
posals, based originally on Flory-Huggins theory, have been made recently t
hat there is a non-negligible solute's volume-proportional term in solvatio
n free energy and the term should be subtracted to obtain solute/solvent co
ntact free energy for biochemical applications. These proposals have result
ed in the revision of the magnitude of the hydrophobic effect in biomolecul
es. The validity has been controversial, since the existence, physical orig
in, and magnitude of the volume-proportional term have been model dependent
. In this paper, we cleared up this problem by using an accurate fused-hard
sphere model and a perturbation scheme in which the compensation between t
he repulsive and attractive interactions has been clarified. The solvation
free energy is shown to be dependent on the solute's surface area and curva
ture: the volume-proportional term is shown to be negligibly small. This di
sproves the basic assumption of the previous theories whose purpose is to "
correct'' the magnitude of the solvation free energy by subtracting volume-
proportional terms. The relationship of our theory to previous theories is
also discussed. (C) 1999 American Institute of Physics. [S0021-9606(99)5010
6-2].