Of course, there was the difficulty that neutrons couldn’t be slowed down once they were formed, and as formed they generally had too much energy (according to the new way of looking at things). At least they couldn’t be slowed down by electromagnetic methods—but there were other ways. A neutron didn’t always enter a nucleus that it encountered. Sometimes, if it struck the nucleus a hard, glancing blow, it bounced off. If the nucleus struck by the neutron is many times as massive as the neutron, the neutron bounced off with all its speed practically intact. On the other hand, if the neutron hits a nucleus not very much more massive than itself, the nucleus rebounds and absorbs some of the energy, so that the neutron bounces away with less energy than it had. If the neutron rebounds from a number of comparatively light nuclei, it eventually loses virtually all its energy and finally moves about quite slowly, possessing no more energy than the atoms that surround it. (You can encounter this situation in ordinary life in the case of billiard balls. A billiard ball, colliding with a cannon ball, will just bounce, moving just as rapidly afterward as before, though in a different direction. If a billiard ball strikes another billiard ball, it will set the target ball moving and bounce off itself with less speed.) The energy of the molecules in the atmosphere depends on temperature. Neutrons that match that energy and have the ordinary quantity to be expected at room temperature are called “thermal” (from a Greek word meaning “heat”) neutrons. The comparatively light nuclei against which the neutrons bounce and slow down are “moderators” because they moderate the neutron’s energy. Fermi and his co-workers were the first to moderate neutrons, produce thermal neutrons, and use them, in 1935, to bombard nuclei. He quickly noted how large nuclear cross sections became when thermal neutrons were the bombarding particles. It might seem that hope could now rise in connection with the practical use of energy derived from nuclear reactions. Neutrons could bring about nuclear reactions, even when they themselves possessed very little energy, so output might conceivably be more than input for each neutron that struck. Furthermore because of the large cross sections involved, thermal neutrons missed far less frequently than high-energy charged particles did. But there was a catch. Before neutrons could be used, however low-energy and however sure to hit, they had to be produced; and in order to produce neutrons they had to be knocked out of nuclei by bombardment with high-energy protons or some other such method. The energy formed by the neutrons was at first never more than the tiniest fraction of the energies that went into forming the neutrons in the first place.