Experiments with a Slow Neutron Velocity Spectrometer II

The apparatus previously used for the determination of neutron energy by time of flight method has been completely rebuilt with a view to increasing the accuracy of measurement as well as the high energy limit. The repetition frequency is obtained from a series of relaxation oscillators and a 50-kc oscillator and frequencies of 100, 200, 500, 1000, and 2500 c.p.s. are available. Four detector channels have been built so that neutrons in four different time of flight groups can be counted simultaneously. When used with the highest repetition frequency, this decreases the time to obtain data by a factor of 24 and has thus made the present experiments feasible with the small Cornell cyclotron. The study of the transmission of Ag with a 1.35-g/${\mathrm{cm}}^2$ absorber shows a single strong resonance at 5.8 ev. A re-examination of In with considerably higher resolution than was previously used shows a single resonance at about 1.35 ev. The effective mean life of neutrons in several different sources was examined and a thin paraffin source with Cd backing devised for measurements in the thermal region. The absorption of B has been examined and found to be proportional to $\frac{1}{v}$ within the limits of error of the experiment, from 0.028 to 50 ev. An experiment to determine the B cross section of the Cd stopped neutrons gave 540×${10}^{-24\mathrm{}}$ ${\mathrm{cm}}^2$. The B absorption curve shows that the cross section of thermal neutrons (0.025 ev) is 708×${10}^{-24\mathrm{}}$ ${\mathrm{cm}}^2$. It is concluded that the effective energy of the Cd stopped neutrons is not that of $\mathrm{kT}$ at thermal energy, for the geometry used, but is 0.041 ev. This conclusion is confirmed by the measured resonances in Ag and In which are higher than the values obtained by the boron absorption method. Correction of these values, as measured by Horvath and Salant (reference 4) for the effective energy of the Cd stopped neutrons, leads to 1.32 ev for In and 5.2 ev for Ag, in agreement with the present results. It is concluded that resonances measured by the boron absorption method are in error by an amount which depends upon the geometry of the experiment, and are probably too low by a factor of 0.041/0.025 = 1.64.