SCIENTIFIC ABSTRACT This proposal is to investigate an all sky sample of dense molecular clouds and very cold IRAS sources for evidence of protostellar objects. The prime aim is to use of the high sensitivity of ISO to search for pure inflow sources with no attendant outflows, and hence investigate the very first stages of stellar evolution. We would be able to determine the low end of the luminosity function of embedded IR sources in these clouds, and answer whether or not the initial mass function is truncated at around 0.7Ms as suggested by some in the literature. OBSERVATION SUMMARY For this part of the proposal (part 1) it is intended to observe dense globules or molecular cores for evidence of protostellar condensations. In this case 'detections' at two wavelengths will be sufficient by way of survey. Positive detections must be followed up at a later date. The AOT PHT32 will be used at 100um with C100, and 200um with C200. A strip scan will be made in each case with 4 pointings. For the C100 strip scan 26s integration per raster pointing (13 chopper steps, 2s per chop step) are used, and for the C200 strip scan 14sec integration per raster pointing (7 chopper steps, 2s per chop step). The total spacecraft times quoted in the target lists have been derived using manuals provided.