The measurement of the tokamak plasma ion temperature can be obtained by collective Thomson scattering. This method requires a far infrared laser to achieve useful spatial resolution for typical tokamak plasma parameters. The very low Thomson scattering cross section (10-28 m2) implies the use of a pulsed far infrared laser with 1 MW power and 1 µs pulse duration. With the best technology of far infrared heterodyne receivers available today, with a noise equivalent power (NEP) below 10-18 W/Hz, it should be possible to measure the ion temperature with 20 % precision. This thesis describes the development of a far infrared laser and a detection system to prove the feasibility of the ion temperature measurement by collective Thomson scattering on the TCA tokamak. This tokamak is situated at the Centre de Recherches en Physique des Plasmas (CRPP) of the Ecole Polytechnique Fédérale in Lausanne. The aim was to observe a scattered signal from the plasma and to obtain the required practical knowledge for the design of a large Thomson scattering system to measure the ion temperature in TCA. For the experiment, an optically pumped D2O laser is used to produce pulses of 1 µs duration and 150 kW of power, at the far infrared wavelength of 385 µm. The scattered signal is observed with a heterodyne Schottky diode detection system using a quasi continuous CD3Cl far infrared laser as local oscillator. The scattered spectrum is resolved through a multichannel S band microwave filter bank, after which the signals are integrated and digitized. Although the laser power is not sufficient to permit an instantaneous ion temperature measurement, a scattered signal correlated with the presence of the plasma was observed when averaging over ten tokamak discharges. It has been found that perturbations from microturbulence in the plasma are not significant within the useful bandwidth of the scattered spectrum. Hence an increase in the available far infrared laser power and a better sensitivity of the detection system should be sufficient to allow for a single discharge ion temperature measurement. The experiment also showed the necessity to reduce the NEP of the detection system, which currently presents an excessively high mixer conversion loss. This is being improved for the future installation of a multi-megawatt far infrared laser. With this new laser it should be possible to obtain the first direct measurement of the tokamak ion temperature by collective Thomson scattering on a tokamak.