NUCLEAR FISSION 273 



of which the charge added to 56e makes 92e, and of which the mass- 

 number added to 139 makes 239 — that is to say, the nucleus aeKr^"". 



This is an example of the type of process which has been named by 

 borrowing the word "fission" from, biology. The biologists seem not 

 to have found a specific verb to correspond (I am told that they use 

 "divide") and the physicists have had no better inspiration. The 

 dictionaries, however, authorize the use of "fissure" as a verb both 

 transitive and intransitive, and I will henceforth so use it.* 



Now a difificulty looms up, or rather what seems to be a difficulty but 

 is really a great advantage, for the grandeur of the idea depends on it. 

 Mass-numbers are only approximations to true masses, and the true 

 mass of the nucleus U^*^ is greater than the sum of the masses of Ba^*^ 

 and Kr^"". There is a superfluity of mass, and by classical ideas this 

 superfluity might have to vanish, which would indeed be a stumbling- 

 block. However, that stumbling block does not exist, because of 

 something I have now to introduce. // is the rest-mass, in the sense of 

 relativity, of U-^^ which exceeds the sum of the rest-masses of Ba^*^ and 

 j^j.100 Now U^^^ before the explosion is practically at rest, but we are 

 not obliged to make the same assumption about the fragments, and in 

 fact we can assume that the fragments fly apart at just such speeds that 

 their relativistic increase of mass with speed brings up the sum of their 

 masses to exactly the right value. If so, their kinetic energies must be 

 50 to 100 Mev apiece. These on the nuclear scale are immense amounts 

 of kinetic energy, and particles possessing it must be easy to isolate 

 and easy to detect. This is why the idea is a grand, one. 



As it might occur to some reader to go to the tables of constants 

 and look up the mass-values of U^*^ and Ba^^^ and Kr^"", I must say at 

 once that he will not find them. Generally speaking, the mass-spectro- 

 graph cannot be used on radioactive and unstable atoms because one 

 cannot get enough of them together for the experiment (exception 

 being made for very long-lived ones like U^*^ and Th^*^) ji^\\ those 

 three belong in that category, and therefore we have to estimate their 

 masses by extrapolation from those of stable isotopes. The extrapo- 

 lations for Ba'^^^ and U^*^ are so small that the uncertainty is trivial, 

 but Kr^"" is no less than fourteen units heavier than the heaviest stable 

 isotope of krypton, and this is serious. However, one is not so much 

 concerned about conceivable defects in grand ideas when the ideas 

 have already done their work by leading with success to grand experi- 

 ments. I lay emphasis again, for a reason later to appear, on the extent 

 to which Kr^"'' is out of line with the stable krypton isotopes; and now 

 we pass to the experiments. 



^ I am indebted to Dr. Elizabeth Patterson of Bryn Mawr College for this solution. 



