NUCLEAR FISSION 285 



(for the present) as those which come out within a few thousandths of 

 a second of the moment of fission. The ratio, according to the Co- 

 lumbia school, is about one to sixty. Delayed electrons and delayed 

 photons have also been observed. Most observations on secondary 

 neutrons are made while the target is being bombarded, and therefore 

 relate to a mixture of the instantaneous with a small proportion of 

 the delayed. 



In energy the secondary neutrons differ greatly from the primary, 

 a remarkable contrast ! This is shown by several neat and pretty ex- 

 periments, in which the secondaries manifest themselves by acting on 

 detectors which cannot perceive the primaries at all. Thus if the 

 primaries are thermal neutrons, an expansion-chamber or an ionization- 

 chamber full of gas can be set among them without showing any sign 

 of them,^^ since they cannot strike hard blows against the molecules 

 therein. Let, however, a piece of uranium be set nearby, and the 

 chamber will show dense trains of ions, produced by nuclei struck very 

 hard and driven out of the molecules by neutrons which are fast. In 

 this manner Halban and Joliot and Kowarski in Paris detected second- 

 aries running in energy up to 11 Mev and beyond, while Zinn and 

 Szilard of Columbia mapped the energy -spectrum up to 3.5 Mev. But 

 also there are detectors able to discriminate between fast and faster 

 neutrons: e.g. phosphorus, which becomes radioactive when bombarded 

 by neutrons if, but only if, these have energy greater than 2 Mev. 

 Dode and others in Paris prepared a source producing neutrons of 

 energy one Mev; placed it next to a uranium target ; surrounded source 

 and target with a tank of liquid carbon disulphide, in which phos- 

 phorus was dissolved ; and the liquid grew radioactive. 



But how many neutrons are released per fission ? This is a question 

 of singular and perhaps of devastating importance, as will presently 

 appear. 



The obvious way to answer it seems to be that elected by Zinn and 

 Szilard, who measured the number of fissions and also estimated, from 

 the number of recoiling nuclei observed in their expansion-chamber, 

 that of the secondary neutrons. Most of the trials have been made 

 by a different method, in which all of the secondary neutrons are 

 reduced to thermal energies before they are detected. (Incidentally 

 there is the advantage, that if neutrons are released with low initial 

 energies they will be counted by this way but not by the other.) 



In this more customary method, the neutron-source and target are 

 close together in the midst of a great tank of water, as large as can 



" Except that if nitrogen is contained in the chamber, the thermal neutrons will 

 react with the nitrogen nuclei so as to release protons (Zinn and Szilard). 



