Neutrons*

J. Chadwick, Cambridge

Introduction. § 1. Bothe and Becker showed that certain light elements, under bombardment by the α-particles of polonium, emit radiation apparently having the character of γ-rays. The element beryllium gives an especially noticeable effect of this kind, and subsequent observations by Bothe, Irène Curie-Joliot and Webster*** showed that the radiation excited in beryllium possesses a penetrating power considerably greater than that of any γ-radiation of radioactive elements known up to the present. In Webster’s experiments the intensity of the radiation was measured with a Geiger–Müller counter and a high-pressure ionization chamber. Webster found that the radiation from beryllium, under the conditions in which his experiments were carried out, has an absorption coefficient in lead of about \(0.22\ \mathrm{cm}^{-1}\). Introducing the necessary corrections for these experimental conditions and using the results of Gray and Tarrant to estimate the relative shares of scattering, photoelectric absorption, and nuclear absorption in the absorption of this penetrating radiation, Webster concluded that the radiation from beryllium must have a quantum with an energy of approximately \(7 \cdot 10^{6}\) electron-volts. In the same way he found that the radiation of boron bombarded by the α-particles of polonium consists in part of radiation considerably more penetrating than the radiation of beryllium, and estimated the energy

* The Existence of a Neutron,—by J. Chadwick. Proc. Roy. Soc., A. 136, 692 (June 1932).
* Bothe und Becker, Z. Physik 66, 289, 1930.
* J. Curie, C. R. 193, 1312, 1931.
* Webster, Proc. Roy. Soc., A. 136**, 428, 1932.

of the quantum of this component of the radiation is approximately \(10\cdot 10^6\) volt-electrons. These conclusions are in good agreement with the idea that the radiation arises by the capture of an \(\alpha\)-particle by the nucleus of beryllium or boron and the emission of the excess energy in the form of a quantum of radiation.

However, the radiation of the light elements exhibits certain peculiar features, and, at my request, several Wilson-chamber photographs were made as this radiation passed through a Wilson chamber. No unexpected phenomena were observed in this, although, as we shall see below, similar experiments have now revealed some remarkable facts. The failure of these early experiments was due in part to the weakness of the polonium source, and in part to the arrangement of the experiment, which, as has now become clear, was not altogether suitable.

Quite recently I. Curie-Joliot and F. Joliot * made a striking observation, namely that these radiations of beryllium and boron prove capable of ejecting, with considerable velocity, protons from substances containing hydrogen. In their experiments the beryllium radiation passed through a thin window into an ionization chamber containing air at atmospheric pressure. If, in front of the window, there was placed a plate of hard paraffin or of some other substance containing hydrogen, the ionization in the chamber increased—in some cases it doubled. It was clear that this effect was due to the ejection of protons, and further experiments by the same investigators showed that the protons have a range in air of up to \(26\ \mathrm{cm}\), which corresponds to a velocity of about \(3\cdot 10^9\ \mathrm{cm/sec}\). They supposed that the energy is transferred from the beryllium radiation to the proton by a process analogous to the Compton effect in the case of electrons, and estimated the magnitude of the quantum of the beryllium radiation at approximately \(50\cdot 10^6\) volt-electrons. **

* J. Curie and F. Joliot, C. R. 194, 273, 1932.

** Many of the arguments in the subsequent discussion apply equally to both radiations, and by the term “beryllium radiation” the radiation of boron is often also meant.

There are, however, two major difficulties in such an explanation. First, it is now firmly established that the frequency* of scattering of high-energy quanta by electrons is given with considerable accuracy by the Klein–Nishina formula, and that this formula should also be applicable to the scattering of quanta by protons. Meanwhile, the observed frequency of proton scattering is several thousand times greater than that predicted by this formula. Secondly, it is difficult to understand how a quantum of \(50\cdot 10^6\) electron-volts can be obtained as a result of the interaction of a beryllium nucleus and \(\alpha\)-particles with kinetic energy \(5\cdot 10^6\) electron-volts. The process capable of providing the greatest amount of energy for radiation consists in the capture of an \(\alpha\)-particle by a beryllium nucleus, \(\mathrm{Be}^9\), and its inclusion in this nucleus with the formation of a \(\mathrm{C}^{13}\) nucleus. The mass defect of \(\mathrm{C}^{13}\) is known both from measurements of the artificial disintegration of boron \(\mathrm{B}^{10}\) and from the study of the band spectrum of carbon; it is approximately \(10\cdot 10^6\) electron-volts. The mass defect of \(\mathrm{Be}^9\) is unknown, but the assumption that it is equal to zero gives the maximum value for the possible change of energy in the reaction \(\mathrm{Be}^9+\alpha \to \mathrm{C}^{13}+\) quantum. From this assumption it follows that the energy of the quantum emitted in this reaction cannot exceed \(14\cdot 10^6\) electron-volts. Of course, one may object that this argument concerning the mass defect is based on the hypothesis that nuclei are built, as far as possible, from \(\alpha\)-particles, i.e., that the \(\mathrm{Be}^9\) nucleus consists of two \(\alpha\)-particles \(+\) 1 proton \(+\) 1 electron, and the \(\mathrm{C}^{13}\) nucleus—of 3 \(\alpha\)-particles \(+\) 1 proton \(+\) 1 electron. Since light nuclei are in question, this assumption is confirmed by facts from experiments on the artificial disintegration of elements, but it is not a general proof.

As a result of this I undertook further experiments in order to investigate the properties of the radiation from beryllium. These experiments—

* Here by the word “frequency” is meant the number of elementary acts of scattering per unit time. Translator.

showed that beryllium radiation ejects particles not only from hydrogen, but from all the light elements investigated. The experimental results proved very difficult to explain from the point of view of the hypothesis of the quantum nature of beryllium radiation, but these results followed as direct consequences if one assumes that beryllium radiation consists of particles with a mass approximately equal to the mass of the proton and without an effective charge, i.e.—of neutrons. A brief summary of some of these observations was published in Nature*. The present paper contains a detailed description of the experiments which indicate the existence of the neutron and from which certain properties of these particles may be derived. In the following paper Dr. Feather** sets forth some of his observations, using a Wilson chamber, concerning collisions between beryllium rays and nitrogen nuclei; and, finally, following it is published a paper by Dee*** devoted to experiments on the collision of these rays with electrons.

§ 2. Observation of recoil atoms. The properties of beryllium radiation were at first investigated with the aid of a counter constructed for work on artificial disintegration by $\alpha$-particles and described in detail earlier****. Briefly, it consists of a small ionization chamber connected to a cathode amplifier. The sudden appearance of ions in the chamber when an ionizing particle enters it is detected by means of an oscillograph connected to the amplifier. The deflections of the oscillograph were recorded photographically on a strip of photographic paper.

The source of polonium was prepared from a solution of radium $(D + E + F)$ by deposition on a silver disk. The disk had a diameter of 1 cm and was placed near the disk

* J. Chadwick, Nature, 129, 312, 1932.
** N. Feather, Proc. Roy. Soc. 136, 709, 1932.
*** P. J. Dee, Proc. Roy. Soc. 136, 727, 1932.
**** J. Chadwick, Constable and Pollard, Proc. Roy. Soc. 130, 463, 1931.

of pure beryllium 2 cm in diameter, and both were enclosed in a small cylinder that could be evacuated (Fig. 1). The first of the ionization chambers used had an opening of 13 mm, was covered with aluminum foil with an absorption equivalent to 4.5 cm of air, and had a depth of 15 mm. This chamber had a very small natural effect, giving on average only about 7 spontaneous deflections per hour.

When the cylinder with the source was placed in front of the ionization chamber, the number of deflections immediately increased. For

Fig. 1.

a distance of 3 cm between the beryllium and the counter, the number of deflections was about 4 per minute. Since the number of deflections remained almost the same when thick layers of metal were placed between the source and the counter—up to 2 cm of lead—it was clear that these deflections were caused by radiation issuing from the beryllium. Later we shall show that these deflections were caused by nitrogen atoms set in motion in collisions with the beryllium radiation.

If, in the path of the radiation, directly in front of the counter opening, a plate of solid paraffin about 2 mm thick is placed, then the number of deflections recorded by the oscillograph increases noticeably. This increase is due to particles ejected from the paraffin and entering the counter. By placing absorbing screens of aluminum between the paraffin plate and the counter, it was possible to obtain the absorption curve shown in Fig. 2

(curve $A$). From this curve it is evident that the particles have a maximum range only slightly exceeding $40\ \mathrm{cm}$ of air, if it is assumed that aluminum foil of $1.64\ \mathrm{mg/cm^2}$ is equivalent to $1\ \mathrm{cm}$ of air. If one compares the magnitudes of the deflections (which are proportional to the number of ions produced in the chamber) caused by these particles with the deflections under the influence of protons of approximately the same range, it becomes obvious that these particles are protons. On this basis, from the range–velocity curve for protons, we conclude that the maximum velocity imparted to the protons by beryllium radiation is approximately $3.3 \cdot 10^9\ \mathrm{cm/sec}$, which corresponds to an energy of approximately $5.7 \cdot 10^6$ electron-volts.

Fig. 2.

After this, the effect obtained when beryllium radiation passes through other substances was investigated. For this purpose an ionization chamber was used with an opening covered by a gold leaf, with an absorption equivalent to $0.5\ \mathrm{mm}$ of air. The elements under investigation were fixed on a clean brass plate and placed very close to the counter opening. In this way lithium, beryllium, boron, carbon, and nitrogen (in the form of paraffin) were tested. In all cases the number of deflections of the counter increased when one or another element was bombarded by beryllium radiation. The ranges of the particles emitted from these elements were very short, on the order of several millimeters of air. The deflections produced by them were of varying magnitude, but many of them were large even in comparison with the deflections produced by slow protons. From this it follows that the particles had a large

ionizing ability and, probably, were in each case recoil atoms of the corresponding elements. The gases were studied by filling the ionization chamber with the gas under investigation by passing this gas through the chamber for several minutes. In this way hydrogen, helium, nitrogen, oxygen, and argon were investigated. Again, in each case deflections were observed, which were attributed to the occurrence of recoil atoms of the corresponding gases. For a given position of the beryllium source relative to the counter, the number of recoil atoms was roughly the same for each gas. We shall return to this peculiarity later. Thus it appears that beryllium radiation can impart energy to the atoms of the substance through which it passes, and that the chances of energy transfer do not change appreciably in going from one element to another.

It was shown that protons are ejected from paraffin with energies up to a maximum of \(5.7\cdot 10^6\) electron volts. If this ejection is attributed to Compton recoil from a quantum of radiation, then the energy of such a quantum must be about \(55\cdot 10^6\) electron volts, for the maximum energy that can be imparted to a mass \(m\) by a quantum \(h\nu\) is equal to:

\[ \frac{2}{2+mc^2/h\nu}\cdot h\nu. \]

The energies of the recoil atoms produced by the same process in other elements can also be readily calculated. For example, the recoil atoms of nitrogen should have energies of at most 450,000 electron volts. Taking the energy required to produce 1 pair of ions in air to be 35 electron volts, we conclude that the recoil atoms of nitrogen should produce no more than 13,000 ion pairs. However, many of the deflections observed in nitrogen correspond to a much larger number of ions; some of the recoil atoms produce from 30,000 to 40,000 ion pairs. In the case of other elements, a similar discrepancy was noted between the observed energies and ranges and the values calculated on the basis of the assumption that the atoms are set in mot—

...by recoil in collision with a quantum of \(55 \cdot 10^6\) volt-electrons. The energies of the recoil atoms were estimated from the number of ions produced in the counter, this number being characterized by the magnitude of the oscilloscope deflections. However, sufficiently good measurements of the ranges can be made either by changing the distance between the element and the counter or by placing thin gold screens between the element and the counter.

The recoil atoms of nitrogen were also investigated in collaboration with Dr. Feather by means of a Wilson chamber. The cylinder with the source was placed directly above a Shimizu-type chamber, so that a considerable fraction of the beryllium radiation passed through the chamber. Over the course of several hours a large number of tracks of recoil atoms were observed. Their ranges, estimated by eye, were about 4 or 6 mm, or, after applying a correction for expansion, approximately 3 mm in standard air. These visual estimates were confirmed by a preliminary series of experiments by Dr. Feather with a large automatic chamber, with which photographs of recoil tracks in nitrogen were obtained. The ranges of nitrogen recoil atoms for different velocities had earlier been measured by Blackett and Lees. Using their results, we found that the nitrogen recoil atoms produced by beryllium radiation can have velocities of at least \(4 \cdot 10^8\) cm/sec, which corresponds to an energy of approximately \(1.2 \cdot 10^6\) volt-electrons. In order for a nitrogen nucleus to acquire such energy in a collision with a radiation quantum, it is necessary that the energy of this quantum be about \(90 \cdot 10^6\) volt-electrons, if the laws of conservation of energy and momentum are obeyed in the collision. But, on the other hand, it had already been shown earlier that a quantum of \(55 \cdot 10^6\) volt-electrons is sufficient to explain the collisions in hydrogen. Generally speaking, the experimental results show that, if the production of recoil atoms is to be explained by collision with a quantum, we must assume an ever greater and greater energy for the quantum as the mass of the nuclei set in motion increases.

§ 3. The neutron hypothesis. It is obvious that we must either regard the laws of conservation of energy and momentum as inapplicable to these collisions, or else make another hypothesis about the nature of the radiation. If we suppose that the radiation is not a quantum radiation, but consists of particles with a mass very close to the mass of the proton, then all the difficulties connected with the collisions, both as regards frequency and as regards the transfer of momentum to different masses, disappear. In order to explain the great penetrating power of the radiation, we must further assume that the particles have no effective charge. We may suppose that they consist of a close combination of a proton and an electron, i.e. of “neutrons,” whose existence was discussed by Rutherford* in his Bakerian lecture of 1920.

When these neutrons pass through matter, they sometimes undergo collisions with atomic nuclei and thus give rise to the observed recoil atoms. Since the mass of the neutron is equal to the mass of the proton, the recoil atoms obtained when neutrons pass through substances containing hydrogen must have velocities the greatest of which must coincide with the maximum velocity of the neutrons. Experiments show that the maximum velocity of protons ejected from paraffin is about \(3.3 \cdot 10^9\) cm/sec. This is the maximum velocity of the neutrons emitted by beryllium when bombarded with \(\alpha\)-rays from polonium. From this we can in turn calculate the maximum energy that can be transferred by colliding neutrons to other atoms, and we find that the results are in complete agreement with the energies observed experimentally. For example, a nitrogen atom in a central collision with a neutron of mass 1 and velocity \(3.3 \cdot 10^9\) cm/sec acquires a velocity of \(4.4 \cdot 10^8\) cm/sec, which corresponds to an energy of \(1.4 \cdot 10^6\) electron-volts, and to a range of 3.3 mm in air—

* E. Rutherford, Proc. Roy. Soc., A. 97, 374, 1920 (Russian translation—“Uspekhi fizich. nauk” 2, 192, 1921; also included in the collection “The Structure of the Atom and the Artificial Disintegration of Elements”).

to the spirit and, in number of ions, about 40,000 pairs. Similarly, an argon atom must acquire an energy of \(0.54\cdot 10^{6}\) electron-volts and produce about 15,000 pairs of ions. All these figures are in good agreement with experiment.*

Combining the results of observations in hydrogen and in nitrogen, it can be shown that the mass of the neutron is roughly equal to the mass of the proton. Feather** reports experiments in which about 100 tracks of recoil nitrogen atoms were photographed in a Wilson chamber. Measurement of the track lengths shows that the maximum range of the recoil atoms is \(3.5\) mm in air at \(15^\circ\mathrm{C}\) and 760 mm pressure, which corresponds to a velocity of \(4.7\cdot 10^{8}\) cm/sec, according to the data of Blackett and Lees. If we denote by \(M, V\), respectively, the mass and velocity of the neutron, then in this case the maximum velocity imparted to the hydrogen atom will be:

\[ v_{\mathrm H}=\frac{2M}{M+1}V, \]

and the maximum velocity imparted to nitrogen atoms will be

\[ v_{\mathrm N}=\frac{2M}{M+14}V, \]

whence

\[ \frac{M+14}{M+1}=\frac{v_{\mathrm H}}{v_{\mathrm N}}=\frac{3.3\cdot 10^{9}}{4.7\cdot 10^{8}} \]

and

\[ M=1.15. \]

The total error in estimating the velocity of the nitrogen atom can easily reach 10%, and therefore it is quite permissible to conclude that the mass of the neutron coincides, to a close approximation, with the mass of the proton.

* It should be noted that a small number of recoil nitrogen atoms produces about 50 or 60,000 pairs of ions. These cases probably correspond to cases of nuclear disintegration found by Feather and described in his work.

** N. Feather, Proc. Roy. Soc., A. 136, 709, 1932.

We shall now turn to a consideration of the formation of neutrons from beryllium when it is bombarded with α-particles. We may suppose that the α-particle is captured by the nucleus of Be\(^9\), with the formation of a C\(^ {12}\) nucleus and with the ejection of a neutron. This process is analogous to the well-known artificial disintegration, but in it not a proton, but a neutron, is emitted. The energy relations taking place in this process cannot be established exactly, since the masses of Be\(^9\) and of the neutron are not exactly known. It is easy to show, however, that such a process satisfies the experimental facts. We have:

\[ \mathrm{Be}^9 + \mathrm{He}^4 + \text{kinetic energy of the } \alpha\text{-particles} = \mathrm{C}^{12} + n^1 + \text{kinetic energy of } \mathrm{C}^{12} + \text{kinetic energy of } n^1 . \]

If we assume that the beryllium nucleus consists of two α-particles and a neutron, then its mass cannot be greater than the sum of the masses of these particles, for the binding energy is produced at the expense of the mass defect. Then the energy equation will be:

\[ (8.00212 + n^1) + 4.00106 + \text{kinetic energy of the } \alpha\text{-particles} > 12.003 + n^1 + \text{kinetic energy of } \mathrm{C}^{12} + \text{kinetic energy of } n^1 \]

or:

\[ \text{kinetic energy of } n^1 < \text{kinetic energy of } \alpha\text{-particles} + 0.003 - \text{kinetic energy of } \mathrm{C}^{12}. \]

Since the kinetic energy of a polonium α-particle is equal to \(5.25 \cdot 10^6\) electron-volts, the energy of emission of the neutron cannot be greater than approximately \(8 \cdot 10^6\) electron-volts; therefore the velocity of the neutron must be less than \(3.9 \cdot 10^9\) cm/sec. We have seen that in fact the maximum velocity of the neutron is approximately \(3.3 \cdot 10^9\) cm/sec, so that the proposed process of disintegration agrees with the observations.

A further confirmation of the neutron hypothesis was obtained by investigating the radiation emitted by beryllium in the direction opposite to the direction of the bombarding α-particles. The source, Fig. 1, was turned so that the paraffin plate placed in front of the counter was subjected to the action of the “backward” radiation of beryllium-

tion. The maximum range of the protons ejected from paraffin is determined, as before, by counting the number of observed protons after passage through aluminum screens of various thicknesses placed between the paraffin and the counter. The curve obtained is shown in Fig. 2 (curve B). The maximum range of the protons was found to be about 22 cm of air, which corresponds to a velocity of approximately \(2.74 \cdot 10^9\) cm/sec. Since the polonium source was situated at a distance of only about 2 mm from the beryllium, this velocity must be compared not with the velocity of neutrons emitted at an angle of \(180^\circ\), but at an angle somewhat greater than \(90^\circ\) to the direction of the incident \(\alpha\)-particles. A simple calculation shows that the velocity of neutrons emitted at an angle of \(90^\circ\), when an \(\alpha\)-particle with its full range is captured by a beryllium nucleus, should be \(2.77 \cdot 10^9\) cm/sec, if it is assumed that the velocity of neutrons emitted at an angle of \(0^\circ\) in the same process is \(3.3 \cdot 10^9\) cm/sec. The velocity found in the experiment described above should be less than this, since the angle of emission is slightly greater than \(90^\circ\). The agreement with the calculations is satisfactory to the extent that could be expected for these measurements.

§ 4. The nature of the neutron. We have shown that the origin of the radiation of beryllium bombarded by \(\alpha\)-particles, and the properties of this radiation insofar as its interaction with atomic nuclei is concerned, receive a simple explanation if it is assumed that this radiation consists of particles with a mass approximately equal to the mass of the proton and having no charge. The simplest hypothesis that can be made concerning the nature of such particles is that these particles are a close combination of a proton and an electron, so that the resultant charge is zero and the mass is slightly less than the mass of the hydrogen atom. This hypothesis is supported by the study of the data that can be obtained concerning the mass of the neutron.

As we have seen, a rough estimate of the mass of the neutron is obtained from measurements of its collisions with hydrogen and nitrogen atoms, but these measurements are not, for our present purpose, satisfactory.

can be made with sufficient accuracy. We must turn to a consideration of the energy relations in the process in which the neutron is liberated from the atomic nucleus; if the masses of the atomic nuclei taking part in this process are known exactly, then an exact estimate of the mass of the neutron can also be made. The mass of the beryllium nucleus, however, has not been measured accurately and, as shown in § 3, in this case only general conclusions can be drawn. Fortunately, there remains at our disposal the case of boron. In § 1 we pointed out that boron, when bombarded by α-particles of polonium, also emits rays which eject protons from substances containing hydrogen. Further investigation showed that this radiation is in all respects similar to the radiation of beryllium, and therefore it should be assumed that the radiation of boron also consists of neutrons. It is probable that the neutrons are emitted by the isotope of boron \(B^{11}\), since we know that the isotope \(B^{10}\) disintegrates with the emission of a proton.* The process of disintegration in such a case will be:

\[ B^{11} + He \to N^{14} + n^{1}. \]

The masses of \(B^{11}\) and \(N^{14}\) are known from Aston’s measurements, and the remaining data needed to determine the mass of the neutron can be obtained from experiment.

In the source (Fig. 1) the beryllium was replaced by powdered boron deposited on a graphite plate. The range of the protons ejected by the radiation of boron was measured in exactly the same way as in the case of the radiation of beryllium. The observed effect was considerably smaller than in the case of beryllium, and it was difficult to measure the range of the protons accurately. The maximum range was approximately 16 cm of air, which corresponds to a velocity of \(2.5 \cdot 10^{9}\) cm/sec. This, therefore, is the maximum velocity of the neutrons liberated from boron by the α-particles of polonium with a velocity of \(1.59 \cdot 10^{9}\) cm/sec. Assuming that momentum is conserved in the collisions, one can calculate the velocity of motion of the \(N^{14}\) nuclei, and in this case we shall know the kineti-

* Chadwick, Constable and Pollard, l. c.

the energies of all the particles taking part in the disintegration process. The energy equation for this process will be:

\[ \text{Mass of } B^{11} + \text{mass of } He^4 + \text{kinetic energy of } He^4 = \text{mass of } N^{14} + \text{mass of } n^1 + \text{kinetic energy of } N^{14} + \text{kinetic energy of } n^1 . \]

The masses entering into this equation are as follows: \(B^{11}=11.00825 \pm 0.0016\); \(He^4=4.00106 \pm 0.0006\); \(N^{14}=14.0042 \pm 0.0028\). The kinetic energies in units of mass are the following: \(\alpha\)-particle—0.00565; neutron—0.0035; nitrogen nucleus—0.00061. We thus find that the mass of the neutron is equal to 1.0067. The errors in the mass measurements indicated above are given by Aston. They are the maximum errors that can occur in his measurements; the probable error may be taken to be approximately one quarter of these errors* . Taking into account the errors in the mass measurements, we may conclude that the mass of the neutron cannot be less than 1.003 and that it probably lies between 1.005 and 1.008.

Such a value for the mass of the neutron was to be expected if the neutron consists of a proton and an electron, and the result obtained gives good support to this view. Since the sum of the masses of the proton and the electron is 1.0078, the binding energy or mass defect of the neutron is about 1 or 2 million electron-volts. This is a quite possible value. We may suppose that the proton and the electron form a small dipole, or we may draw the more attractive picture in which the proton is embedded in the electron. With such a representation we may expect that the radius of the proton is equal to several units of order \(10^{-13}\) cm.

§ 5. Passage of neutrons through matter. The electric field of a neutron of this type must evidently be extremely small, except in regions situated very close to it, at distances of the order—

* The mass of \(B^{11}\) relative to the mass of \(B^{10}\) was checked by optical methods by Jenkins and Mae Kellaro (Phys. Rev. 39, 546, 1932). Their values agree with Aston’s data to within a unit in the fifth decimal place. This indicates that Aston’s measurements may be regarded with great confidence.

on the order of \(10^{-13}\) cm. In passing through matter, the neutron should undergo no deflection, except in those cases when it comes into close collision with a nucleus. The neutron potential in the field of the nucleus may be roughly represented by Fig. 3. The radius of the collision area for appreciable deflections of the neutron must be smaller than the radius of the nucleus. Further, the neutron must be capable of penetrating easily into the nucleus, and it is possible that the scattering of the neutron is to a considerable extent caused by the internal field of the nucleus, or, in other words, that the scattered neutrons are chiefly those which have penetrated through the potential barrier. According to this view we must expect that a collision of a neutron with a nucleus should occur very rarely and that the scattering should be, roughly speaking, the same in all directions, at least as compared with the Coulomb scattering of a charged particle.

Fig. 3.

These conclusions were confirmed in the following way. A source with a beryllium screen was placed at a distance greater than \(2^{1}/_{2}\) cm from the surface of a closed counter filled with air (Fig. 1). The number of deflections, or the number of nitrogen recoil atoms produced in the chamber, was observed during a known time. The observed number was 190 per hour, if the natural effect is taken into account. Then, between the source and the counter, there was placed a solid piece of lead \(2^{1}/_{2}\) cm thick. The number of deflections fell to 166 per hour. Since the number of recoil atoms must be proportional to the number of neutrons passing through the counter, these observations show that

that 13% of neutrons are absorbed or scattered in passing through \(2^{1/2}\) cm of lead.

Suppose that a neutron which passes at a distance \(p\) from the center of a nucleus is scattered and removed from the beam. Then the fraction removed from the beam in passing through a layer \(t\) of lead will be \(\pi p^2 n t\), where \(n\) is the number of lead atoms per unit volume. Consequently,

\[ \pi p^2 n t = 0.13, \]

whence

\[ p = 7 \cdot 10^{-13}\ \mathrm{cm}. \]

This value for the radius of collision with lead is small, but not implausible. We may compare it with the radii of radioactive nuclei, calculated from the decay constants by Gamow and Houtermans* and amounting, according to their calculations, to \(7 \cdot 10^{-13}\) cm.

Analogous experiments were carried out with the passage of a neutron stream through brass and coal. The values of \(p\), derived from these experiments by the same method, turned out to be respectively \(6 \cdot 10^{-13}\) cm and \(3.5 \cdot 10^{-13}\) cm.

Effective cross sections for some light elements were also compared by another method. For this purpose a second ionization chamber was used, which could be filled with various gases by means of circulation. The position of the source was kept unchanged relative to the counter, and the number of deflections of the oscillograph was observed when the counter was filled successively with hydrogen, nitrogen, oxygen, and argon. Since the number of neutrons passing through the counter was the same in all cases, the number of deflections must be proportional to the effective collision cross section, if one neglects the influence of the substance of the counter and makes a correction for the fact that argon is monatomic. It was found that nitrogen, oxygen, and argon give approximately the same number of deflections; thus the effective cross section of nitrogen and oxygen is, roughly,

* G. Gamow und F. Houtermans, Z. Physik, 52, 453, 1928.

are equal, while the effective cross section of argon is approximately twice as large. With hydrogen the measurements were very difficult, since most of the deflections were very weak owing to the small ionizing power of protons and the low density of the gas. The results show that the effective cross section of hydrogen probably amounts to about two thirds of the cross section of nitrogen or oxygen, but it may also be larger.

As yet we have little information concerning the angular distribution of scattered neutrons. In some experiments, kindly carried out for me by Gray and Dee, scattering by lead was compared in the backward and forward directions by means of ionization in a high-pressure chamber. Gray and Dee found that the amount of scattering is approximately what could be expected on the basis of the measurements just mentioned, and that the intensity per unit solid angle is approximately the same between \(30^\circ\) and \(90^\circ\) in the forward direction and between \(90^\circ\) and \(150^\circ\) in the backward direction. Thus scattering by lead is not appreciably anisotropic.

Two types of collisions are of special interest: the collision of a neutron with a proton and the collision of a neutron with an electron. A detailed study of collisions with elementary particles is of special interest, since it should give information about the structure and field of the neutron itself, whereas collisions with other nuclei are of interest chiefly from the point of view of investigating the structure of these nuclei. Some preliminary experiments by Dee, using a high-pressure ionization chamber—experiments whose purpose was to measure the scattering of neutrons by solid paraffin and liquid hydrogen—indicate that collision with the proton is more frequent than collision with other light atoms. This is not in agreement with the experiments described above; however, the results are not entirely definite. These collisions can be investigated more directly with the aid of a Wilson chamber or by the counting method, and I hope in the near future to carry out such an investigation.

The collision of a neutron with an electron was investigated by two

by two methods: with the aid of a Wilson chamber and with a counter. A description of the experiments with the Wilson chamber was given by Dee in a special article*. Dee studied the total ionization produced by a large number of neutrons in passing through the chamber and the short electron tracks, which must be the result of a close collision between a neutron and an electron. His results show that collisions with electrons are extremely rare in comparison with collisions with nitrogen nuclei, and he estimated that a neutron can produce, on the average, not more than one ion pair in passing through 3 m of air.

In the experiments with the counter, a beam of neutrons passed through a piece of brass 2.5 cm thick, and the maximum range of the protons ejected by the passing beam was determined. From the range found in this way, the maximum velocity of the neutrons after passing through the brass was determined, and it could be compared with the maximum velocity in the incident beam. No change in the velocity of the neutron as a result of their passage through the brass could be detected. The accuracy of these experiments was not great, since determination of the end of the proton range was very difficult. The results show that the loss of energy of a neutron in passing through 2.5 cm of brass is no more than approximately \(0.4 \cdot 10^6\) electron-volts. A path of 2.5 cm in brass corresponds, with respect to collisions with electrons, to a path of approximately \(2 \cdot 10^4\) cm in air, so that this result shows that the neutron loses less than 20 volt/cm of path in air to collisions with electrons. This experiment thus gives general confirmation of the experiments with the Wilson chamber, but it is much less precise. We conclude that the transfer of energy from a neutron to electrons is an exceedingly rare event. This is not unexpected. Bohr** has shown, on the basis of the most general considerations, that a collision of a neutron with an electron must be very rare in comparison with nuclear collisions.

* Dee, Proc. Roy. Soc. A, 136, 727, 1932.
** Copenhagen discussion, unpublished.

Massey,* on the basis of plausible assumptions concerning the neutron field, carried out detailed calculations of the energy loss to electrons, and also found that this loss must be small—no more than 1 pair of ions per 1 m of air.

§ 6. General remarks. It is interesting to investigate whether neutrons are emitted by elements other than beryllium and boron when bombarded with $\alpha$-particles. Experiments performed up to now show that there are no cases in which the effect is comparable with these two. Some indications of neutron emission have been obtained for fluorine and magnesium; however, in this case the effect was very small—less than 1% of the effect obtained in beryllium under the same conditions. There is also the possibility that some elements emit neutrons spontaneously; potassium may belong here, since it is known to emit nuclear $\beta$-radiation accompanied by more penetrating radiation. However, evidence for the presence of neutrons has not been found, and it appears certain that this penetrating radiation of potassium is, as was initially assumed, $\gamma$-radiation.

Although definite proof of neutron emission exists only in two cases of nuclear transformations, we may nevertheless suppose that the neutron is an ordinary component of atomic nuclei. We can try to construct nuclei from $\alpha$-particles, neutrons, and protons, and thereby avoid the presence in the nucleus of uncombined electrons. This has certain advantages, for, as is well known, electrons in the nucleus lose some of the properties that they possess outside the nucleus, for example, their spin and magnetic moment. If the $\alpha$-particle, neutron, and proton are indeed the only components of the structure of the nucleus, then we can calculate the mass defect or binding energy of the nucleus as the difference between the mass of the nucleus and the sum of the masses of its basic components. However, in no case can it be regarded as established that the $\alpha$-particle

* Massey, Nature 129, 469, 691 (1932).

and the neutron are the only complex particles entering into the composition of the nucleus, and therefore the mass defects calculated in this way may also not be the true binding energy of nuclei.

Here it may be noted that the examples of the disintegration of nuclei by neutrons considered by Feather* are not all of the same type, and Feather suggested that in some cases a particle of mass 2 and charge 1 may be emitted, i.e. the isotope of hydrogen recently discovered by Urey, Brickwedde, and Murphy.

Up to now it has been assumed that the neutron is a complex particle consisting of a proton and an electron. This is the simplest assumption, and it is supported by the fact that the mass of the neutron is about 1.006, i.e. somewhat less than the sum of the masses of the electron and the proton. Such a neutron should be the first stage in the combination of elementary particles for the formation of nuclei. It is obvious that this neutron can help us to form a clear picture of the emergence of more complex structures, but we shall not pursue the discussion of these questions further, since such speculations, although they are not entirely idle, are at the present moment still not very fruitful. Of course, one may suppose that the neutron is an elementary particle. At present, however, such a conception is hardly expedient, unless one excludes the possibility of explaining the statistics of nuclei similar to N¹⁴.

It remains to consider the transformations that occur when an α-particle is captured by a beryllium nucleus Be⁹. The data presented in this article indicate that the principal type of transformation consists in the formation of a C¹² nucleus and in the emission of a neutron. The experiments of Curie-Joliot and Joliot, Auger, and Dee*** show quite definitely that there exists a radiation emitted by beryllium which is capable

* Feather, l. c.

* J. Curie-Joliot et F. Joliot, C. R. Acad. Sciences 194*, 708, 876 (1932).

* F. Joliot, C. R. 194**, 877 (1932).

**** P. Dee, l. c.

create fast electrons when passing through matter*. I carried out experiments with a Geiger counter, with the aim of investigating this radiation, and the results show that fast electrons are produced by γ-radiation. There are two different processes that may give rise to such radiation. First of all, we may suppose that the transformation of beryllium Be⁹ into C¹² sometimes occurs with the formation of an excited C¹² nucleus, which passes into the normal state with the emission of γ-rays. This process is analogous to the transformations which are presumed to occur in some cases when a nucleus is disintegrated with the emission of protons, for example B¹⁰, F¹⁹, Al²⁷; most of these transformations occur with the formation of an excited nucleus, and only in approximately one quarter of the cases does the final state of the residual nucleus arise in the first stage. In such a case we should have two groups of neutrons with different energies and γ-radiation with a quantum equal to the difference of the energies of these groups of neutrons. The energy quantum of this radiation should be less than the maximum energy of the emitted neutrons, i.e. \(5.7 \cdot 10^6\) electron-volts. Secondly, we may suppose that sometimes the beryllium nucleus is transformed into a C¹³ nucleus and the entire excess energy is emitted in the form of γ-radiation. In this case the energy quantum of the radiation should be approximately \(10 \cdot 10^6\) electron-volts.

It is interesting to note that Webster observed soft radiation from beryllium bombarded by α-particles from polonium, the quantum of this radiation being approximately \(5 \cdot 10^5\) electron-volts. This radiation can quite plausibly be attributed to the first of the two processes considered, and its intensity has the correct order of magnitude. On the other hand, some of the electrons observed by Curie-Joliot and Joliot have energies of the order of from 2 to \(10 \cdot 10^6\) electron-volts, and Auger notes one example of an electron with an energy of about \(6.5 \cdot 10^6\) electron-volts. These electrons

* Whereas neutrons produce slow electrons; see above. Translator’s note.

can be produced by hard \(\gamma\)-radiation obtained in transformations of the second type.*

It may be noted that electrons with an energy greater than that indicated above apparently do not exist. This is confirmed by an experiment carried out in this laboratory by Dr. Occhialini. Two Geiger counters were placed in a horizontal plane, and the number of coincidences recorded by them was observed by means of the method developed by Rossi. A beryllium source was then placed in the plane of the counters, so that the radiation passed successively through both counters. No increase in the number of coincidences was found. It follows from this that, even if the production of \(\beta\)-rays with an energy sufficient to pass through the walls of both counters—i.e. through a total of \(4\ \mathrm{mm}\) of brass—is possible, such \(\beta\)-particles arise only in insignificant numbers; the energy in question is greater than approximately \(6 \cdot 10^{6}\) volts. This experiment further shows that neutrons very rarely produce coincidences in the counters under ordinary experimental conditions.**

In conclusion, I shall once more briefly formulate the grounds that lead one to suppose that the radiation whose effects have been considered in this article consists of neutral particles, and not of quanta. First, there are no data indicating that, in electron collisions, radiation with the quantum energy required to explain nuclear collisions could arise. Second, the hypothesis of the quantum nature of this radiation can be preserved only at the cost of abandoning the laws of conservation of energy and momentum. On the other hand, the neutron hypothesis gives an immediate and simple explanation of the experimental facts; it is quite consistent, and it sheds new light on the problem of the structure of the nucleus.

* Although the occurrence of fast electrons can readily be explained in this way, one should not lose sight of the possibility that some of these electrons arise as a result of secondary effects of neutrons.

** See also Rasetti, Nature, 20, 252 (1932).

Neutrons

Summary

The properties of the penetrating radiation emitted by beryllium and boron when bombarded with $\alpha$-particles of polonium have been investigated. It has been concluded that this radiation consists not of quanta, as was initially supposed, but of neutrons, i.e., particles with mass 1 and charge 0. Evidence is presented that the mass of the neutron probably lies between 1.005 and 1.008. This indicates that the neutron consists of a close combination of a proton and an electron with a binding energy of approximately from 1 to $2 \cdot 10^6$ electron-volts. On the basis of experiments on the passage of neutrons through matter, the question of the frequency of their collisions with atomic nuclei and electrons is considered.