SCIENTIFIC ABSTRACT The prime scientific goal is to obtain precise, well spectrally sampled absolute photometry in the 2.5 - 200 microns wavelength range of a complete sample of a compact dense condensations of cold dust and gas, selected by sensitive, high angular resolution surveys of cloud cores at mm/sub mm wavelengths. The high precision FIR/MIR spectra, supplemented by MIR/NIR diffraction limited imaging, will differentiate between various evolutionary stages, allowing true isothermal protostars, disk and cocoon stars to be distinguished from one another. The relative detection rate of these phenomena will reflect the true incidence of these objects within each of the cloud cores surveyed in the submm, and will therby allow statements to be made about the duration of the evolutionary stages of star formation. OBSERVATION SUMMARY For investigation of nearby protostar candidates and their parent clouds: P25 rasters of subfields (3x3 up to 7x7 depending on cloud size) at 200, 100, 60 micron with Tint = 5s. Raster step interval = array repeat length (180, 138 arcsec for C200, C100). To be performed concatenated. For each individual protostar candidate within cloud, perform concatenated: P05 1x11,1x10 rasters at 25 micron, using 10 arcsec aperture and minimal raster step of 10 arcsec. The sum of these will provide 5" sampling over a 100" scanlength. To provide information about the background outside the cloud an additional P05 1x7 raster with 50" raster step is proposed. All P05 rasters use Tint=5s. P32 at 200, 160, 120 micron (4x2)raster, O/S 2, Tint = 10s per S/C raster point. P32 at 100, 70, 60 micron (4x2)raster, O/S 2/3, Tint = 10s per S/C raster point. P40 in a 5x1 raster Tint = 32s. (Raster interval chosen to provide 1 point on protostar candidate, 2 points on parent cloud, 2 points on background) For investigation of distant star forming clouds we propose the following: P25 rasters of subfields (3x3) 200, 100, 60 micron with Tint = 5s. Raster step interval = array repeat length (180, 138 arcsec for C200, C100). Centered at the cloud core, perform concatenated: P05 1x5 raster at 25 micron, using Tint=5s and 23 arcsec aperture (Raster interval chosen to provide 1 point on cloud core, 2 points on cloud, 2 points on background) P40 in an identical 5x1 raster Tint = 32s. P32 at 200, 160, 120 micron (4x2)raster, O/S 2, Tint = 10s per S/C raster point. P32 at 100, 60 micron (4x2)raster, O/S 2/3, Tint = 10s per S/C raster point. VISIBILITY CONSIDERATIONS: Although the source list has been taylored to avoid the holes in sky visibility, some sources of exceptional scientific interest have been retained despite positions of marginal or zero visibility for the standard launch in the 5 degree visibility binning scheme. Specifically, SM1 and RHO OPH B2 may be problematic for the spring launch and the Orion protostar candidates for the autumn launch. Whether or not these sources will be visible will depend critically on the time of night of launch. Therefore, we propose that these sources be retained in the source lists for the present.