Finally in December 1938 the secret broke. Hahn and Strassmann in Germany made a chemical analysis of a uranium target that had been exposed to neutrons. They were far more thorough than previous investigators had been, and they found barium, charge 56, which had not been present in the target material before the experiment. The only possible explanation was the fission process. Within a few weeks the violent kicks caused by the fission products in counters were found, and in the following days this experiment was repeated around the world. There was no doubt that neutrons could induce fission in uranium nuclei. A few more weeks, and it was ascertained that the fission process released neutrons which might lead to more fissions. The chain reaction, however, was still far from a reality. Niels Bohr and John Wheeler proved that a neutron could not cause fission in U²³⁸ unless its energy were greater than about one million electron-volts. When the neutrons are first made in the fission process, many of them do have energies greater than one million electron-volts. But before they can cause a fission, they usually make a few nonfission collisions with uranium nuclei, giving part of their energy to the nuclei and escaping with the remainder. The nuclei are then left with too little energy to undergo fission and the neutrons with too little energy to cause fissions in their next encounters. Thus too few neutrons reproduce themselves and no chain is possible. Bohr and Wheeler suggested, however, that the rare isotope of uranium, U²³⁵, can undergo fission when any neutron, even a slow neutron, hits it. Thus a chain reaction is possible in U²³⁵. This was confirmed experimentally shortly afterwards by John Dunning and Alfred Drier and their co-workers at Columbia University. Why the isotopes 235 and 238 behave so differently, is not difficult to understand. The 235 is more explosive and more prone to undergo fission than 238 because it is smaller and therefore its protons repel each other more strongly. More important still, when a neutron is captured by 235, it acquires a greater kinetic energy by virtue of the short-range nuclear attraction than a neutron acquires when it is captured by 238. This happens for the simple reason that nuclei tend to be more stable when they have an even number of neutrons (or protons) than when they have an odd number. U²³⁵, having an odd number of neutrons, is more eager to receive an additional neutron than 238, which already has an even number of neutrons. Consequently, the capture of a slow neutron by 235 almost always eventuates in the fission process; while in 238, the excess energy, introduced by the neutron, is merely ejected from the nucleus in the form of a gamma ray, and U²³⁸ becomes U²³⁹.