It is shown that the neutron densities required for a number of basic experiments with free neutrons can, at present, only be achieved with the thermal-neutron gas produced by a fission reactor. Three basic technological problems in the production of a dense and clean neutron gas were investigated experimentally: 1.(a) measurement of the neutron field in the neutron-gas cavity (flux- and current-densities, anisotropy),2.(b) attainment of a low residual gas density in the neutron-gas cavity in the presence of strong radiation fields, and3.(c) shielding of the measuring device from the primary reactor radiation.In our experiments we use a dense neutron gas in ultra-high vacuum to determine the electron-neutrino angular-correlation coefficient a by a precision measurement of the recoil-proton energy spectrum in free-neutron decay (0–751 eV). A neutron density of 1.42 × 107 neutrons/cm3 was achieved in a cavity with residual gas density of <1 × 103 molecules/cm3 giving a flux of 88 protons/s into the energy spectrometer. Aside from the technological problems mentioned, low magnetic and electric fields in the beam tube are also necessary: disturbing fields of less than 1 mV/cm and 1 mOe were achieved. Preliminary results are discussed.Based on the experience obtained from these experimental studies we discuss proposals for n-n scattering and for scattering of charged particles from a neutron-gas target. The feasibility of n-n experiments with existing reactors, as well as the installation of neutron targets in storage rings is discussed. In connection with a 20 A proton beam n-p scattering with luminosity up to 1031/cm2 s can be performed by means of neutron targets produced from a medium-size research reactor.