Reprinted from THE PHYSICAL REVIEW, Vol. 55, No. 8, 799-800, April 15, 1939
Printed in U. S. A.
Instantaneous Emission of Fast Neutrons in the Interaction of Slow Neutrons with Uranium*
Recently it became known that uranium can be split by neutrons into two elements of
about equal atomic weight. In this fission of uranium the two elements produced have a large
neutron excess; moreover they are probably produced in an excited nuclear state. One might
therefore expect that these excited fragments instantaneously emit neutrons and that perhaps the
number emitted is even larger than one per fission.
One might also expect a delayed emission of neutrons- as was first pointed out by Fermi-
if some of the fragments go through one or more beta-transformations before they emit a neutron.
Delayed emission of neutrons caused by the action of both slow and fast neutrons on uranium has
recently been reported by Roberts, Meyer, and Wang, who find a period of about 12 seconds.
In order to see if there is an instantaneous emission of neutrons from the fission of
uranium we have performed the following experiment. We exposed uranium oxide to neutrons
which were slowed down by paraffin wax, using as a source of neutrons a block of beryllium from
which photoneutrons were liberated by the gamma-rays of radium. A helium-filled ionization
chamber connected to a linear amplifier served as a detector for fast neutrons. The ionization
pulses of the chamber were observed visually by means of a cathode-ray oscillograph and were
recorded by the usual counting arrangement.
Figure 1 shows a diagram of the experimental arrangement. The ionization chamber is
covered by a cadmium sheet cap G which prevents the thermal neutrons from penetrating to the
helium ionization chamber. A cadmium sheet shield H, 0.5 mm thick, is used to cover the
cylindrical box E which contains 2300 g of uranium oxide. The uranium oxide is screened from
the thermal neutrons by this shield and can be exposed to them simply by removing the shield.
We observed about 50 pulses per minute from the helium chamber when we exposed the
uranium oxide to the thermal neutrons in the absence of the cadmium shield H, but obtained only
5 pulses per minute when the uranium was screened from the thermal neutrons by the cadmium
shield. The difference of about 45 pulses per minute we have to attribute to fast neutrons emitted
from uranium under the action of thermal neutrons. It is reasonable to assume that this emission
of fast neutrons is connected with the fission of uranium.
Control experiments were carried out in which uranium was replaced by lead. The effect
of the presence and absence of the cadmium shield H and the cadmium cap G was tested.
In order to estimate the number of fast neutrons emitted per fission under the action of
thermal neutrons we used an ionization chamber lined with a thick layer of uranium oxide having
an area of 25 cm. This uranium chamber was put in place of the helium chamber without
otherwise materially changing the experimental arrangement. Under these conditions the uranium
chamber gave about 45 fissions per minute. Assuming the range of the fission fragments to be
about 0.005 g per cm2 in uranium oxide, the observed 45 fissions per minute should occur in a
surface layer, weighing 0.13 g, of the thick uranium oxide lining. Accordingly, about 800,000
fissions per minute should occur in the 2300 g of uranium oxide which was used in our
experiment. By taking into account the solid angle, the size of the helium chamber and the
pressure used, and by assuming that the "fission neutrons" have an average collision cross section
in helium of 3.5 X 10-24 cm2 we find the number of neutrons emitted per fission to be about two.
This number is of course only a rough estimate; the main cause of uncertainty is the
considerable variation of the cross section of helium with the neutron energy in the region around
one million volts. A hydrogen-filled