LETTERS TO THE EDITOR
 
ionization chamber is now being used in order to obtain a more accurate estimate. It seems to be
established, however, that the order of magnitude is one neutron per fission.
     Anderson, Fermi and Hahstein have independently, and by a different method, carried out
experiments on the neutron emission connected with the fission of uranium. Our observations are
consistent with their results, and we wish to thank them for communicating their results to us
before publication.
     While from our observations we can only say that the time delay involved in this
"instantaneous" neutron emission appears to be less than one second, we should expect, for
theoretical reasons, this emission to take place within less than 10-14 second.
          We have also looked for a delayed emission of fast neutrons by performing the following
experiment. The uranium oxide was irradiated for some length of time in the arrangement shown
in Fig. 1. Then the radium was quickly removed from the beryllium block and the cathoderay
oscillograph screen was watched for a period of 15 seconds for an indication of a delayed
emission of fast neutrons. After the radium is removed there is no gammaray background to set a
lower limit for the observable helium recoil energy; the only slight background remaining is due to
electrical fluctuations of the amplifier. In 50 experiments, corresponding to a total observation
time of more than 12 minutes, we observed only two pulses which may or may not have been due
to a delayed emission of fast neutrons. This is to be compared with the emission of 45 fast
neutrons per minute, the number observed while the radium is inside the beryllium block. We
conclude that, if slow neutrons falling on uranium cause a delayed emission of neutrons which are
sufficiently fast for us to observe, their number must be very much smaller than the number of
neutrons which we have observed in the instantaneous emission.
          We are indebted to Dr. S. Seely for his assistance in carrying out some of these
experiments. We wish to thank the Department of Physics of Columbia University for the
hospitality and the facilities extended to us, and also wish to thank the Association for Scientific
Collaboration for enabling us to use one gram of radium in these experiments. 
                                   LEO SZILARD 
                                   WALTER H. ZINN .