In the past ten years many long carbon chains have been detected in the ISM and the CSM. Cyanopolyynes as long as HC7N, HC9N and HC11N are abundant molecules in sources like TMC1 and IRC+10216 and long carbon chain radicals have been detected in the same sources through their pure rotational spectrum at radio wavelengths. The relatively slow decrease in the abundance of the higher order chains in TMC1 and IRC+10216 suggests that they are produced very efficiently. However, while the abundance of the long and polar carbon chains have been derived in the ISM and the CSM, many light molecules necessary to the formation and synthesis of such chains have not yet been detected, mainly due to their lack of permanent dipole moment and hence of pure rotational transitions. C3 and C4H2 are important molecular species for the synthesis and the growth of long carbon chains and probably they also play an important role in the formation of the first cyclic carbon molecules. Consequenly, a determination of their abundance in the interstellar and circumstellar medium is of a great importance for our understanding of the interstellar and circumstellar chemistry. Both molecules have not permanent dipole moment and hence they dont have a pure rotational spectrum. They can be only detected in the infrared through the emission or absorption of their infrared active bands. We propose to observe the v2 bending mode of C3 (at 63 cm-1), the v8 and v9 bending modes of C4H2, and the v11 and v13 bending modes of C6H2 towards several galactic molecular clouds and a few prototypical circumstellar envelopes. The select clouds cover a large range of physical conditions allowing to detect C3 in emission in the shocked regions of recent low-mass star formation or in absorption/emission towards strong continuum sources at 150 um.