The purpose of this proposal is to probe into the origin of a number of
protostellar candidates isolated from molecular clouds. These sources were
selected from the IRAS point source catalog as candidates for protostars, and
their anti-correlation with molecular materials was found through a comparison
of the IRAS data with our 13CO(J=1-0) map obtained for a search for
molecular clouds in the Cygnus region (80<l<104, -7.5<b<10.5 deg). At least in
this region, the isolated sources are, to our surprise, majority of protostellar
candidates, occupying ~70% of all the IRAS point sources selected as candidates
for protostars. This may indicate that very compact clouds which might
escape detection in our previous 13CO survey are playing a leading role of
star formation in our Galaxy. On the other hand, these objects might be other
astronomical objects such as galaxies or planetary nebulae, although most of
their optical counterparts are absent on the Palomar Sky Survey prints.
 Nothing but the IRAS data is available for these sources at the moment, and
all what is known about them is that they are accompanied by cold dust without
apparent connection to molecular clouds. Using sensitive CAM equipped with
ISO, we propose to make deep imaging of the isolated protostellar candidates
around 10 micron in order to investigate the distribution of the dust emission
around them. We also propose to determine their spectra at far infrared
wavelengths of 160 and 200 micron with ISOPHOT in order to estimate the
dust temperature precisely. These data from ISO are indespensible in order to
investigate the properties of dust associated with these objects,which may give
us a hint for their origin. All the observations can be carried out only by ISO,
since we cannot access to those wavelengths from the ground-based
observations.