In a recent series of papers, Schwinger discussed a process that he ca
lled the dynamical Casimir effect. The key essence of this effect is t
he change in zero-point energy associated with any change in a dielect
ric medium. (In particular, if the change in the dielectric medium is
taken to be the growth or collapse of a bubble, this effect may have r
elevance to sonoluminescence,) The kernel of Schwinger's result is tha
t the change in Casimir energy is proportional to the change in the vo
lume of the dielectric, plus finite-volume corrections. Other papers h
ave called into question this result, claiming that the volume term sh
ould actually be discarded, and that the dominant term remaining is pr
oportional to the surface area of the dielectric. In this paper, which
is an expansion of an earlier Letter on the same topic, we present a
careful and critical review of the relevant analyses. We find that the
Casimir energy, defined as the change in zero-point energy due to a c
hange in the medium, has at leading order a bulk volume dependence. Th
is is in full agreement with Schwinger's result, once the comet physic
al question is asked. We have nothing new to say about sonoluminescenc
e itself.