SCIENTIFIC ABSTRACT: We propose to use ISOCAM and ISOPHOT to study phenomena at the interface between the High Energy Astrophysics and the Physics of Cold Interstellar Gas. If an X-ray binary is inside or near a molecular cloud most of the energy from the relativistic particles streaming out from the compact source must be dissipated by heating of the interstellar material and thermal emission from dust is expected. The Galactic Center region is appropriate to study these phenomena since in there massive stellar remnants and molecular gas are more abundant than anywhere else in the Galaxy. In that region of the sky we wish to study the emission from dust heated by the compact source of annihilation of positrons 1E1740.7-2942 ("Great Annihilator") and by the large scale electric current known under the name of "Snake". Another sources of study for the autumn launch are the radio "ears" in W50 which are brightened by particles streaming out from the classic microquasar SS433, and the runaway binary Circinus X-1 which is moving through the interstellar medium at a speed of several hundred km/s. We point out that our proposed observations of SS433/W50 are of different nature to those proposed by Salama with PHT who aims to determine the spectral shape of the central source. In the event of a spring launch we will observe the black hole candidates NPersei92, NMon75, NMuscae, Cyg X-1, and the peculiar IR bright binary Cyg X-3. OBSERVATION SUMMARY: In order to obtain some spectral information all sources will be observed using CAM01 with the filters LW2 (5-8.5 microns) and LW3 (12-18 microns) in the microscanning mode and PHOT C100 at 90 microns. For all sources we will use a p.f.o.v. of 6", for maximum sensitivity.Typically, for CAM, we aim at detecting fluxes in the range of 0.002 to 0.003 mJy/arcsec2 with a S/N ratio of order 3. For the Great Annihilator, we will cover a field of (2.5'x2.5'),with a raster of (4x4) steps. The same field will be covered with PHOT in one exposure. For the Snake, we will perform a 6x3 position raster , with a displacement of 16 pixels. The area covered for the latter source will be roughly 500x200 arcsec with the highest S/N ratio; the total area covered is of the order of 700x400 arcsec. The PHT concatenated observation will be performed in the raster mode as well, using a linear raster with 4 images displaced by 0.5'. For SS433/W50, we will observe the center of the lobes, over an area of 16'x10', again with steps of 16 pixels. With PHOT, we will survey the same area with a raster of (5x3) positions,with displacements of 1.5'. For Circinus X-1, we will use a raster of (6x6) steps to cover an area in excess of an ISO field of view. This will be followed by a raster with PHOT with 4 positions, with displacements by 1.5' in both directions. For the sources visible in the spring, we plan to use a strategy similar to that used for the Great Annihilator.