Due to the large columns of gas and dust in the core of the Rho Oph
cloud and the high density of sources, ground-based and spaceborne
infrared observations have thus far been insensitive to two distinct
populations of embedded stars.  The first is the low-luminosity
Class I objects (<1 L(sun)) which are in an early, deeply embedded
phase of evolution and display strong infrared excesses.  The second
group is the diskless Class III sources, which are usually strong
x-ray emitters, but have no infrared excesses and can be difficult to
distinguish from background stars.  Yet these diverse populations have
something in common: they both emit weak radio continuum emission.  We
propose to use ISOCAM at 6.75 and 11.5 microns to identify the infrared
counterparts to a sample of weak radio sources in the Rho Oph core.
These data, when combined with existing or planned near-infrared
observations, will allow us to construct the spectral energy distri-
bution for each counterpart and hence determine the amount of circum-
stellar dust present.  This will give us knowledge of the
extents and distributions of the low-luminosity Class I and Class III
populations embedded throughout the depth of the cloud core.  Mid-
infrared wavelengths are necessary since they are much more sensitive
to thermal emission from circumstellar dust than the near-infrared.
The numbers and luminosities of Class I sources with L < 1 L(sun)
are needed to explore the age, mass function, and accretion rate in
Rho Oph relative to other regions of low mass star formation where
this population has already been completely sampled due to lower
source densities and cloud extinction.  There is strong evidence that
the Class III sources in the core are contemporaries of the T Tauri
stars but have shorter timescales for disk dissipation.  Our data
will allow us to observe their remnant disks and lay important
groundwork for future infrared spectroscopic studies that will
explore their masses and ages.