RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.OWING to war conditions it has become increasingly difficult to get a clearpicture of the progress of science. Many periodicals have become inac-cessible and an ever-growing proportion of the research work is being carriedout under the seal of military secrecy. To the Reporter, this seems asuitable time for a general survey of our present knowledge of nuclearprocesses and of the concepts used in explaining and predicting them.This survey is preceded by a historical introduction which briefly recapitulatesthe steps in which this knowledge was acquired.During the last decade our knowledge of the atomic nuclei has enormouslyincreased, with regard both to the number of observed facts and to thedegree of understanding of their significance.(1932) of theneutron, a particle with practically the same mass as the proton but withoutclect'ric charge.The discovery of such a particle did away with the necessityto assume the existence of electrons inside the nucleus (an assumptionwhich caused great theoretical difficulties) in order to explain the fact thatthe atomic weight A is in general greater than the atomic number 2. Instead,it is now generally believed that the atomic nucleus is composed of 2 protonsand A - 2 neutrons, and this assumption, together with plausible assump-tions about the forces acting between these elementary particles, accountsfor most of the known properties of nuclei, such as their size, spin, statisticsand binding energy.Neutrons were first observed in the disintegration of beryllium bycc-particles, and this process is still one of the most convenient ways of pro-ducing them, although much larger intensities can be obtained by meansof a cyclotron.I n passing through matter, neutrons are not influencedby the shell electrons, and they are not repelled by the electric charge ofthe nuclei. Each neutron is therefore bound to hit a nucleus eventually,even though it may have to pass through several centimetres of solid matterbefore this happens. The effectivity of neutrons in causing certain nuclearreactions increases considerably if their energy is reduced, and the studyof " slow neutrons )' (obtained by passing neutrons through water or paraffinwax) has revealed some exceedingly interesting phenomena,2 the inter-pretation of which led to Bohr's theory of the heavy atomic nuclei (1936).The next important step, only a few months later, was the successfulattempt by J.D. Cockcroft and E. T. S. Walton3 (1932) to disintegratenuclei by means of ions which had been accelerated through large electricfields. This " artificial " disintegration is not essentially different fromThe first fundamental step was J. Chadwick's discoveryNatwe, 1932, 129, 312; PTOC. Roy. SOC., 1932, A , 136, 692.See, e.g., E. Amaldi and E. Fermi, PhyekaE Rev., 1936, 50, 899.Proc. Roy. SOC., 1932, A , 137, 229288 RADIOACTIVITY BND SUB-ATOMIC PHENOMENA.the disintegration by natural a-particles, discovered by Rutherford in 1919.It gives, however, a much larger number of disintegrations per unit time andthereby makes their study much less laborious. It also permits the use c;fprojectiles other than a-particles (helium nuclei), thus greatly extendingthis field of research.Thus the discovery of heavy hydrogen (deuterium)was quickly followed by the use of its ion (the deuteron) as a projectileparticularly effective in nuclear disintegration. E. 0. Lawrence’s ingeniousinvention of a device in which ions are given very high energies by beingaccelerated many times in a moderate electric field has made its name“ cyclotron ” a household word in iiuclear laboratories. Lawrence’s latestinstallation produces deuterons of 16 MeV (millions electron volts) andhelium nuclei (“ artificial a-particles ”) of 32 MeV.Particles of suchenergies can overcome the electrical repulsion of even the heaviest nuclei.For the disintegration of light elements, the straightforward accelerationof ions by high electric potentials (up to about 3 MV) has still someadvantages, such as simpler operation and more uniform particle energy.The next great step was the discovery of artificial (or induced) radio-activity.4 Again, it must be said that there is no fundamental differencebetween artificial and natural radioactivity. The great number, however,of radioactive isotopes which can be produced by disintegration of stableelements makes it possible to study many aspects of radioactivity much morecompletely than was possible before. Neutrons, for the reasons given above,have been found particularly effective in producing radioactive substances.Furthermore, new modes of radioactive disintegration were discovered,such as the emission of positrons (positive electrons), the capture ofK-electrons and the transition between isomeric nuclear states.The discovery of nuclear fission is perhaps not of such fundamentalimportance, but it showed, for the first time, a possible way of utilisingnuclear energies on an engineering scale.A detailed discussion of thefission phenomena and of the possibility of nuclear chain reactions wasgiven in this Report for 1939. Further progress in this field is bound tobe slow, and little further evidence has been published since.Radioactivity.This term may be taken to indicate the spontaneous transformationof one nuclear species into another.The transformation, being a uni-molecular reaction, follows the well-known equation N = Noe-Al. Instead ofthe decay constant h, the half-value period (half-life) T = 0.69311 isgenerally used to indicate the rate of decay.The term “ spontaneous ” involves a certain ambiguity, since, strictlyspeaking, the decay of an artificially radioactive substance is not spontaneousbut it consequence of a preceding nuclear disintegration. In practice, thosetransformations which follow the impact of a projectile within a very shorttime are reckoned as part of the disintegration process while those with LL4 IF’. Joliot and IrAne Curie, Ndure, 1934, 183, 201FRISCH . 289macroscopic half-life are regarded as spontaneous (radioactive) transform-ations of the nucleus formed in the disintegration process.a-Radioactivity.-The emission of helium nuclei (a-rays, a-particles) isvery common among the natural radioactive elements, but only onea-active isotope has been produced artificially (by bombarding bismuth withfast He ions from a cyclotron) (see the Report for 1940).The theory ofa-activity 4a was developed in 1929 and has not changed since. It is simplyan application of quantum mechanics to the problem of the motion of ahelium nucleus under the forces which the residual nucleus exerts upon it.In leaving the nucleus the a-particle, after overcoming the nuclear attrac-tion, is subjected to repulsion in the electrostatic (Coulomb) field of thenucleus.The joint effect of these two forces is that the particle mustcross a “ potential barrier,” and the quantum-mechanical treatmentshowed that a particle bass a finife probability of doing so even if its energyis too small for it to surmount the barrier. The probability for this “ tunneleffect ” increases rapidly with increasing particle energy, in quantitativeagreement with the experimental facts which are summarised in the Geiger-Nuttall relation log A = A + B log E, which connects the decay constantA of the substance with the energy E of the cc-particles emitted.a-Radioactivity is observed among the heaviest elements only, samariumbeing the lightest element that shows it. This can be explained from thetrend of the mass defect curve, according to which a-activity becomesenergetically possible only above Z = about 50 and the energy of &in-tegration (and therefore the rate of decay) gets sufEciently high only a tconsiderably higher 2.p-Radioactivity.-The theory of this phenomenon is far less well estab-lished than that of cc-decay.The kinetic energy of the p-particles emittedby any particular substance shows a continuous distribution of valuesbetween zero and an upper limit which is equal to the energy available.This means that always some energy “ disappears,” and to account for thisthe assumption has been made that in addition to the @-particle, a hypo-thetical Encharged particle-called neutrino-is emitted and escapesunnoticed, thanks to its lack of electric or other interaction with matter.This assumption forms the basis of E.Fermi’s theory of @-decay5 and ofseveral variants of this theory, such as the one due to E. J. Konopinskiand G. U. Uhlenbeck.6 For a discussion of these theories and of theirconnection with the recently discovered meson (a particle, occurring in thecosmic radiation, with a mass intermediate between those of electron andproton) we must refer to the Report for 1039.Direct attempts to find effects of the neutrinos from strong radioactivesources have so far been unsu~cessful.~ This does not argue against their*a E. U. Condon and R. W. Gurney, Physical Rev., 1929, 33, 127; G. Garnow,6 2. Physik, 1934, 88, 161.Z . Physik, 1929, 53, 610.6 Physical Rev., 1935, 48, 7.M. E. Nahmias, PTGC.C a d . Phil. Xoc., 1936, 31, 99; H. R. Crane, Physical Rev.,1939, 55, 501.REP.-VOL. XXXVIII. 290 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.existence, since their interaction with matter, as predicted by Fermi’stheory, is many million times below the range of the most sensitive methodsnow available. Positive results have been reported * of an attempt todetect the recoil which a p-active nucleus suffers on account of the emissionof the neutrino. These experiments are, however, exceedingly difficult andtheir interpretation is perhaps not absolutely convincing.It has been suggested9 that neutrinos may be responsible for the so-called new stars (novze). The authors show that, according to Fermi’stheory, production of neutrinos on a vast scale must set in at a certain veryhigh temperature, which may be reached at the centre of some stars.Theneutrinos escape, taking large amounts of energy with them, and this coolingeffect at the centre of the star causes, in the authors’ view, a collapse of thestar and an enormous flare-up in its outer layers.The half-life of p-active substances increases with decreasing transform-ation energy (upper limit of the pray spectrum), though less steeply than inthe case of a-decay. This dependence (Sargent’s rule) lo follows also fromFermi’s theory. There are, however, many substances whose half-life ismuch longer than Sargent’s rule would indicate ; in these cases it is believedthat the spins of the radioactive nucleus and the daughter nucleus differby one or more units, causing the transformation to be “ forbidden,’’ inanalogy to the forbidden transitions between atomic energy levels.The shape of the @-ray spectrum has been studied for a great numberof @-emitters.The interpretation of the results is complicated by the factthat most of the spectra are complex, Le., that the nucleus in question hastwo or more alternative ways of decay, which leave the resulting nucleusin different states of excitation. The excitation energy is nearly alwayswithin a very short time emitted as y-radiation. The so-called coincidencemethod has often been successfully employed in disentangling the sequenceof events. It is, for instance, possible to study the spectrum of thosep-particles only which coincide with (i.e,, are followed within a very shorttime by) a y-quantum, detected by a separate counter. By skilfullycombining such experiments, it has been possible in some caws to analysea complex p-ray spectrum into its simple components.The results, however,do not yet permit a decision between the various p-theories.The latter arealso called positrons, and by contrast the word electron is often used todenote negative electrons only, though sometimes the word negatron isemployed for this.If the term @-decay is to include any process by which a nucleus is trans-formed into an isobaric nucleus (one with the same mass number but differentcharge), then the capture of a K-electron l1 into the nucleus must be mentionedhere. From the point of view of Dirac’s hole theory this process is notThe @-particles may be negative or positive electrons.8 H.R. Crane and J. Halpern, Physical Rev., 1938, 53, 798; 1939, 5e, 732.9 G. Gamow and M. Schoenberg, {bid., 1941, 59, 539.10 €3. W. Sargent, PTOC. Roy. SOC., 1933, A, 139, 659.11 L. W. Alvarez, Physical Rev., 1938, 54, 406FRTSCH . 291fundamentally different from the eniission of a positron. The latter pro-cess can be regarded as the capture of one of the (infinitely many) electronsin states of negative energy; the hole left through its removal is the positronobserved, according to Dirac. The K-capture is a rather curious kind of" radioactivity " : no radiation (save the elusive neutrino) is emitted by thenucleus. The shell electrons, however, have to rearrange themselves afterthe loss of one of their order and X-rays are therefore emitted which betraywhat has happened.The existence of only a limited number of isotopes for each element isintimately bound up with the question of (3-stability .Regarding nucleiwith odd mass number, it has been long known that each mass numberoccurs only once among the stable elements, with a few exceptions. Thisis because, of all the possible isobars of a given odd mass number, only oneis stable against B-decay. I n those few cases where two adjacent isobarsare found in Nature (such as le70s and ls7Re) one of the two is probablyalways unstable, although the decay may be too slow to be detected.I n the case of nuclei with even mass number, pairs and even triplets ofisobars frequently occur, but they occupy only even atomic numbers.Forinstance, the mass number 124 occurs among the isotopes of tin, telluriumand xenon (2 = 50, 52, and 54), but the intervening isobars 124Sb and lz4Iare not found in Nature and have indeed been proved to be (3-active.This marked difference between odd and even mass numbers can beexplained by the assumption (which is supported by other facts) that boththe protons and the neutrons tend to associate themselves in pairs in thenucleus. For any odd mass number, this tendency can never be completelysatisfied, since either a proton or a neutron is always left over. For evenmass number, however, there is a difference between odd and even atomicnumbers. The latter are strongly bound and stable because all the protonsand neutrons are associated in pairs, while in the former one proton and oneneutron is left, making the structure less stable.Isomeric Transitions.These are transitions from an excited state of exceptionally long half-life to the ground state of the same nucleus.The reason for these longhalf-lives is a difference of several units between the spins of the two states(which causes the transition to be highly " forbidden '7 and the absence ofany levels in between which might permit a stepwise transition. A fairlyextensive discussion of this phenomenon was given in last year's Report.Nuclear Collisions.After discussing the spontaneous transformations of nuclei, we now turnto those processes which are provoked by the impact of a nuclear particleupon a nucleus.The particular reaction in which the nucleus A is hit bythe particle a, emits the particles by c, d . . . . and is thereby transformedinto the nucleus By is briefly denoted by A(a, b, c, d . . . .)B. If A and a aregiven, B, b, c, d . . . . must fulfil the condition that the sums of their mas292 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.and charge numbers must be equal to the corresponding sums for A and a.But this is not all : the energy liberated in the process must be positive,or if it is negative (if energy is absorbed) it must not be greater than thetotal kinetic energy of A and a, referred to their centre of gravity (if A is,as usually, much heavier than a, this is practically the kinetic energy of A).In most cases it is found that all the possible reactions within the abovelimitations do actually occur, at least with light elements.With heavierelements the emission of charged particles is less probable and nearly allcollisions result in the emission either of one or more neutrons or of y-rays(or of both). In many cases the energy of the bombarding particle has astrong effect on the yield and on the character of the reaction.A complete theory of these phenomena, capable of predicting them indetail, is at present entirely out of the question. Such a theory wouldindeed require the complete mathematical treatment of the motion of a largenumber of particles, all strongly interacting, a task inkitely more difficultthan the corresponding one concerned with the motion of the atomic elec-trons in the Coulomb field of the nucleus, where the mutual interactions ofthe electrons can be regarded as a mere perturbation.Furthermore, theway in which the force between nuclear particles varies with their distanceis still not well known, and doubts have even been raised as to theapplicability of quantum mechanics to problems in which distances as smallas the nuclear radius are involved.It was N. Bohr l2 who showed that a comparatively simple phenomeno-logical theory of nuclear collisions can be developed just because of the stronginteraction of the particles in the nucleus. Because of it, the energy of theimpinging particle becomes rapidly distributed over all the other particlesof the nucleus, and the resulting system, the " compound nucleus," remainstogether until by a chance fluctuation enough energy is concentrated in oneparticle to enable it to break away from the nucleus.If this is the originalparticle or one of its kind, we say it has been scattered; if it emerges withless than its original energy, we speak of inelastic scattering. If it is adifferent particle, we speak of a nuclear disintegration. After the emissionof one particle the nucleus may still retain enough energy to emit a secondone, and even after this a third one. The compound nucleus may also loseenergy by y-radiation or internal conversion (ejection of an inner shellelectron); after this, there may or may not be enough energy left for theemission of a particle, or of further y-radiation.Finally, for the heaviestnuclei, another process, called nuclear fission, is possible ; the nucleusdivides itself into two smaller nuclei of roughly equal size.The important point about it is that these things happen one at a time,and can be regarded separately. The impact of the projectile forms thecompound nucleus, zt system which is characterised by the number of protonsand neutrons in it and by its energy and would have the same propertiesif it had been formed in a different way. The compound nucleus behavesexactly like a radioactive nucleus with several alternative modes of decay,l2 Nature, 1936, 137, 344, 361FRISCW. 293each with its probability per unit time (decay constant). Its half-life is ofthe order of 10-20 t o 10-15 sec., very short on a human scale but long com-pared with the time of about sec.which a particle with several MeVenergy requires to travel its own diameter and which may be regarded as arough “ nuclear time unit.’’ If the compound nucleus emits a particleor a y-quantum, a new system is thereby created, which again has alternativemodes of decay, and so on until a stable nucleus is formed.I n its neglect of details (such as the fate of an individual particle in thecompound nucleus) this theory is essentially thermodynamical, and thethermodynamical analogy can indeed be pushed to a considerable extent.13The impact of the projectile can be compared to the impact of a fast moleculeupon the surface of a very sinall liquid droplet.The formation of thecompound nucleus corresponds to the condensation of the molecule, wherebythe temperature of the droplet is raised, on account both of the heat ofcondensation and of the kinetic energy of the molccule. The droplet canthen lose energy either by the evaporation of one or more molecules or byradiation (analogous to the y-radiation of the nucleus). In order to picturefission as well one would have to endow the droplet with an electrical chargesufficient to lower its effective surface tension almost to zero.I n pursuing this analogy, one must remember that the number of particlesin an atomic nucleus is quite small, vastly smaller than the number of mole-cules in any ordinary thermodynaniical system.Furthermore, the “ nucleartemperatures,” although of the order of 1010 degrees for average excitation,are very low in the sense that only a few of the many degrees of freedomof the nucleus are excited. For these reasons, a nucleus cannot take anyarbitrary energy value but has discrete energy levels, like an atom.There is, however, an important difference between a nucleus and anatom. In an atom the levels come closer and closer together with increasingenergy until we reach the ionisation limit, where the spectrum becomescontinuous : the atom now accepts any amount of energy and immediatelysplits up into an ion and a free electron which carries away the excess energy.A nucleus, however, does not immediately emit a particle even if there isenough energy for this, but has to wait until enough energy happens tobecome concentrated upon one particle. With nuclei the transition t o thecontinuous spectrum is, therefore, gradual : * as the energy is increased, theescape of a particle becomes easier, the energy states get broader, and atsufficiently high excitation they merge into a practically continuous spectrum.Our experimental knowledge of nuclear energy levels is still very in-complete, but as far as it goes it is in good accord with Bohr’s nuclear theory.A good deal of evidence has been accumulated about the lowest levels ofthe lighter elements, by the accurate study of the energy balance in dis-integrations.For instance, bombardment of fluorine with a-particles of5 MeV energy produces several groups of protons, with energies of 5.2,4.0, 2-2 and 1.3 MeV.If we assume that the emission of a proton with 5.2Is See, e.g., R. Peierls, “ Reports of Progress in Physics,” 1941.* In molecules, a similar gradual transition is known ae prediesocbtion294 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.MeV leaves the resulting nucleus in the ground state (the resulting reactionenergy Q = + 1.4 MeV tallies with that calculated from the packingfractions of the nuclei involved), then the emission of the other protonsmust leave the nucleus with an excitation energy of 1 4 , 3.4 or 4.5 MeV.(In calculating the figures, the recoil energy of the nucleus has been allowedfor.) Of course there may be other energy levels in between which are notproduced by this particular disintegration, but this appears unlikely for variousreasons.I n some cases where the same nucleus can be obtained from twodifferent disintegration processes [e.g., l*B (a, H) 13C and 12C (D, H) 13C]the same energy levels have been found to be excited.As long as only the natural a-particles were available these investigationswere restricted to the lightest elements (roughly up to calcium), but the useof artificially accelerated ions should permit their extension to higher atomicnumbers. The difficulty arises, however, that with increasing atomicnumber the emission of neutrons rather than charged particles (which arehampered by the ( ( Gamow barrier ”) becomes prevalent, and energyineasurements on neutrons are laborious and inaccurate.For high atomic numbers some information comes from the y-rays ofthe natural radioactive elements.Their energies have been accuratelymeasured (largely by studying the fast electrons produced by their internalconversion) and level schemes have been deduced. They show that thelowest levels of the heavy elements lie, on the whole, considerably closertogether than those of the light elements. This agrees well with the liquid-drop model if one assumes that the lowest excitations correspond to deform-ation oscillations of the nucleus as a whole ; a large drop has slower oscillations,with correspondingly lower quantum energies. Some investigations on they-rays of artificial radioactive elements support this general trend but showgreat irregularity in the locations of the lowest levels, indicating that theanalogy with a droplet must not be taken too literally.Performed and interpreted in a different way, disintegration experimentscan also give information on much higher energy states, this time not of theresulting but of the compound nucleus.For instance, bombardment of19F with protons results, as the &st step, in the formation of a 20Ne nucleuswith an excitatioii energy equal to the sum of the kinetic energy of the protonand its binding energy of 12-9 MeV. Only if this happens to fit one of thelevels of 20Ne can the compound nucleus be formed, and one would expectthe reaction to take place only for certain discrete values of the protonenergy.In fact, however, the reaction occurs for all energies over a considerablerange, although the yield shows pronounced maxima and minima.Thisbroadening-to the extent of partial overlapping-of the levels is due tothe instability of the corresponding nuclear states. According to quantummechanics, the half-width r of any energy level is connected with its decayconstant (The wave function of the unstabltlstate has the character of a damped train of waves, and it is well known thatresonance becomes less sharp if the damping is increased.)bv the relation r = hh/xFRISCH. 295As the excitation energy is increased, the escape of particles from thenucleus becomes easier and therefore the levels increase in broadness. A tthe same time the complexity of motion increases, and therefore the averagedistance between levels gets smaller and smaller. Both trends are clearlyshown in the above-mentioned experiments l4 where the intensity of y-radiation obtained from fluorine under proton bombardment was measuredas a function of the proton energy.At low energies (up to about 1 MeV)the graph shows individual peaks of small but measurable width (a fewKeV), and at the highest voltage used (2.2 MV) the levels have nearlymerged into a continuous mass. Incidentally these peaks have been foundvery useful in calibrating the energies of artificially accelerated protons, muchin the way that spectral lines are used to calibrate optical spectrographs.The increase of level density with excitation energy is shown impres-sively by the fact that the average distance between the lowest levels ofneon is a few MeV, whereas at an excitation of about 15 MeV it is only afew ten thousand electron volts, or a hundred times less.The intermediateregion cannot be observed, since with decreasing proton energy the repulsiveelectric field of the nucleus (the " Gamow barrier ") becomes a greater andgreater obstacle.No such repulsion exists in the case of neutrons, and the nuclear reactionsproduced by slow neutrons offer some of the most striking illustrations ofthe characteristics of nuclear levels, or as it is often called, of nuclearresonance.Neutrons are slowed down by passage through light elements, in par-ticular hydrogen (or hydrogen compounds). The term is generally taken toinclude both those neutrons which have lost all their energy and are inthermal equlibrium with the slowing-down medium (called thermal neutronsor C-neutrons) and those of energies up to a few hundred electron volts.It was found by Fermi et al.that some elements show enormous absorptionfor slow neutrons. Among them, the behaviour of lithium and boron(more exactly of 6Li and l o g ) is particularly interesting. The absorptionof the neutron leads in both cases to the emission of a fast a-particle; it iseasy to detect these a-particles by means of an ionisation chamber and aproportional amplifier, and such a chamber, lined with lithium or boronor filled with boron trifluoride, is a very convenient and sensitive detectorof slow neutrons.Furthermore, from our other experience of the width and spacing ofnuclear levels, we can be certain that there is no marked influence of nuclearresonance in elements as light as lithium and boron, if the energy of theneutrons is varied, say, between 0 and 1000 eV.From such a light nucleusit is very easy for the a-particle to escape, and the levels should have a widthof much more than 1000 eV. In such a case quantum mechanics predictthat the absorption should be inversely proportional to the velocity of theneutrons. This absorption law-often briefly called the 1 /v-law-is equiv-alent to the statement that the probability for a boron nucleus to absorbE. G. Bernet, R. G. Herb, and D. B. Parkinson, Phyeicai Rev., 1938, 54, 398296 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.a slow neutron depends only on the density of neutrons in its neighbuurhood ;for if the velocity of a given stream of neutrons is doubled, their density isobviously halved.It is not possible to visualise the l/w-law by thinkingof collisions between small spheres; the de Broglie wave-length of a slowneutron is much larger than the diameter of a nucleus and the process istherefore rather analogous to the absorption of a light quantum by an atom.For the same reason, the fact that the absorption cross-section for slowneutrons is often many hundred or even thousand times larger than the truesize of the nucleus does not indicate any contradiction to the generallyaccepted view that the forces between nuclear particles are practically zeroat distances larger than 10-12 cm.I n most of the heavier elements, however, the capture of a slow neutronis followed by the emission not of an cc-particle but of y-radiation. Thewidth of the level is therefore only of the order of one eV or less, Thisfigure has been derived from plausible assumptions as to the mechanism ofradiation and agrees with the width of those neutron resonances which havebeen studied.Such a study is easiest in those cases where the nucleus formed by thecapture of the neutron is radioactive.Let us, for instance, consider the caseof gold. If gold foils are exposed to a beam of slow neutrons, under boronabsorbers of varying thickness, their activity is found to decrease at firstrapidly and then more slowly with increasing boron thickness.Analysisof the absorption curves shows that there are roughly two groups of neutrons,with absorption coefficients in boron of about 30 and 3 cm.2/g. The firstgroup has been identified with the thermal neutrons, for instance, from thefact that their absorption in boron depends on the temperature of theslowing-down medium. Since the energy of thermal neutrons (at roomtemperature) is about 0.025 eV, the neutrons in the other group must beten times as fast, or their energy 2.5 eV. They are very strongly absorbedin gold and yet the activity they produce is not very strong. We concludethat there are not many of these particular neutrons, and since we can calcu-late the energy distribution of the slow neutrons in any given slowing-downmedium (e,g., water) from a statistical consideration of their collision in themedium, we can estimate that only neutrons within an energy region ofabout 0.1 eV show this selective absorption in gold.Similar experiments have been carried out with a number of otherelements.In all cases the boron absorption curve shows the presence ofthermal neutrons, and nearly always a group of resonance neutrons, of energycharacterktic of each element. I n some cases there are indications of thepresence of more than one group of resonance neutrons. Of course, eachelement must really have a large number of resonance levels, but only thosewith the lowest energy are readily detected by these experiments. Cadmiumis of particular interest, since it has a resonance level a t about 0.1 eV.Itis therefore a strong absorber for thermal neutrons and, at the same time,practically transparent for neutrons of 1 eV or more. Cadmium sheetsare therefore widelynsed either t o cut out thermal neutrons when they arFRISCH. 297not wanted or to study their properties by taking alternate measurementswith and without a screen of cadmium.It may seem that the experimental evidence for these resonance pheno-mena is somewhat indirect and unconvincing. Actually there are a greatmany more experiments in their support, most of them carried out withvery simple equipment but devised and interpreted with great ingenuity,and the totality of their evidence is very convincing indeed. Furthermore,very direct evidence has been obtained recently by C.P. Baker and R. F.Bacher.15 These authors virtually produced slow neutron beams of homo-geneous velocity by using a modulated neutron source, giving short periodicalbursts of neutrons, and by counting only those which arrived with a, givendelay at the counter, which was placed at some distance from the source. Byvarying the time of delay, they were able to plot the absorption of boron,cadmium and indium as a function of the neutron energy, and their resultsagree well with the conclusions from the earlier, indirect evidence.Nuclear Photo-eJffect .This phenomenon is not essentially different from other disintegrations,if we regard the y-quantum as just another kind of nuclear projectile. Itsabsorption by the nucleus forms a “ compound nucleus,” in this case simplythe original nucleus with an excitation energy equal to the energy of they-quantum. Onlyemission of neutrons has so far been observed, and comparatively little isknown about this “ nuclear photo-effect,” on account of the low yieldobtainable, except for deuterium and beryllium, where the threshold isabnormally low (2-2 and 1.6 MeV respectively, instead of 6 to 10 MeV as inmost elements). Deuterium and beryllium, irradiated with y-rays fromradium or thorium-C, are therefore occasionally used as a neutron source.If the energy is sufficient, a particle may be emitted.Very Light Nwlei.For nuclei containing only a few particles the statistical considerationsof Bob’s nuclear theory are no longer applicable. On the other hand, themathematical difficulties in the way of a complete treatment are less formid-able, and the experimental study of the collisions between the simplestnuclei is our main source of information about the forces acting betweennuclear particles. So far, this information agrees fairly well with the pre-dictions of the meson theory of nuclear forces (see this Report for 1939),but this theory moves at the very edge of quantum mechanics, and realprogress will probably depend on some revolutionary change in the funda-ments of quantum theory. 0. R. FRISCH.l 6 Physical Rev., 1941, 59, 332