SCIENTIFIC ABSTRACT The main objective of the LWS proposals on Pre-Main Sequence evolution is the understanding of the evolution of matter from the dilute, cold interstellar state of the dark molecular clouds to the compact, hot stellar state of the main-sequence stars. The present proposal deals with the earlier evolutionary phase. The discovery of a real protostar, i.e. an object that derives the majority of its luminosity from the accretion of infalling material, is one of the most important observations for our understanding of the overall stellar evolution. While a number of plausible candidates do exist among the IRAS Class I sources, fully convincing kinematic evidence for infall is still missing. Far-infrared spectroscopy may be one of the means to provide this evidence. An essential condition for the formation of stars in molecular clouds is the presence of dense clumps of matter, as revealed both from molecular lines (e.g. NH_3 and CS) and millimetre continuum emitted by circumstellar dust. These clumps are o ften coincident with Class I sources. We therefore consider dense molecular cores and the IR sources associated with millimetric emission as the prime candidates for studying the earliest evolutionary stages of star formation. Both spectrometers will be used in low resolution mode, in order to detect the various spectral features expected to trace the different physical conditions around the accreting protostellar objects. From the outside (at scales of hundreds of astronomical units) we expect to detect: - molecular emission from various species (such as water and CO), tracing the parental material; - fine structure lines such as the [OI]63.157 micron, tracing the transition region between fully ionised and molecular gas; - the high quantum number hydrogen recombination lines tracing, together with the IR coronal lines, the hotter and denser gas of the regions close to the accretion shock. - the velocity field can be comparing the observed line intensities with the models of the emitting regions. Using the grating spectrometers the properties of the solid state matter will also be studied directly in the observed shape of the "continuum" and, perhaps, also indirectly from the determination of the relative elemental abundances. In particular, we may obtain unexpected results regarding the phases of the dust in extremely dense interstellar environments, where grain evolution can be quite different from that of the diffuse medium. OBSERVATION SUMMARY Full spectral scans using the LWS in grating mode are planned for all our targets. The "on source" integration time for LWS Grating spectra are computed in order to observe with S/N=10 lines of 2 10^-20 W/cm^2. For the brightest sources this limit is reduced in order to observe with S/N=10 lines that are 1/200 of their 60 um continuum. In presence of an "outflow" the "off source" observation is done carrying out a map (in "raster scan") of the flow with a full grating spectrum in each point of the map. The total integration time spent on all the points of the map will be similar to the one spent on the single "on source" point; the reference "off source" point will be therefore built by binning toge ther all the points of the map. A full SWS grating scan is foreseen for all the objects (with the exception of the objects that will be observed in the SWS guaranteed time) with a 25 um IRAS flux sufficient to give S/N=10 on a line that is 1/30 of the continuum in the SWS01 mode. We also scan with SWS01 mode those points in the flow where evidence of near-infrared shock excited emission exists. We have chosen a few template objects representative of the evolutionary stages addressed in this proposal for FP observations of lines expected to be excited in the different physical regimes.