ANNUAL REPORTSON THEPROGRESS OF CHEMISTRY.RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.DURING the year the number of types and examples of nuclear trans-formations known has been greatly increased by the use of artificialsources of particles and quanta more energetic than those fromnatural radioactive sources. The intensity, also, of the artificialsources is many times greater than that from the biggest naturalsources. a-Particles from the cyclotron have facilitated the ob-servation of a-induced radioactivity in the heavier elements. Withartificial sources of very energetic neutrons and y-rays it has beenfound that the expulsion of a neutron by y-ray or neutron impactis a very general possibility among the elements. The nucleartransmutations known up to the middle of 1937 are discussed veryfully in a review by H.A. Bethe and M. S. Livingston.1The theoretical treatment of the nucleus regards it as made upof neutrons and protons, and the application of quantum theoryto nuclear problems,. with various approximations, has gone onsteadily.2Work on the penetrating radiation has been greatly helped bythe development of the quantum theory of fast electrons. It seemsthat the less penetrating part of the rays consists of electrons whichfit in with the theory, and this can therefore explain most of thecosmic ray phenomena in the atmosphere. The nature of the morepenetrating rays is not yet clear, and a great deal of experimentalattention is now focused on this part of the radiation.ISOTOPIC CONSTITUTION OF THE ELEMENTS.P.W. Aston has made a careful set of comparisons to establishthe isotopic mass of 12C, and finds 12.00355 & 0.00015 for this im-portant mass. He has also made new and accurate mass determin-Rev. Mod. Physics, 1937, 9, 246.H. A. Bethe, ibid., 1936, 8, 82; 1937, 9, 69.Nature, 1937, 139, 6228 RADIOACTIVITY AND SUB- ATOMIC PHENOMENA.ations for 78* 82* 849 86Kr and 129* 132Xe.4 J. Mattauch has madeaccurate determinations of the masses of 86* 87Sr and shown byanalysis of a specimen of strontium obtained from a mica rich inrubidium that 87Sr is produced by the p-ray transformation of thatelement. A. J. Dempster has confirmed Aston's suggestionthat neodymium has isotopes a t 148, 150 and established theexistence of 180W.7 A. 0.Nier has designed a new mass spectro-meter for the purpose of making accurate determinations of relativeabundances of isotopes. The elements Hg, Xe, Kr, Be, I, As, andCs were investigated. No new isotopes were found, but theabundance ratios for 196. 198, 199. 200, 201, 202, 204Hg, 124, 126, 128, 129, 130,131. 132, 134, 136Xe, 78, 80, 82. 83, 84, 86Kr were determined. 8Be was notfound with a detection limit of 1 in 100,000, and caesium, arsenic,and iodine were shown to be simple elements to about this limit.The abundance ratios were also determined for 184, 186s 187, 188. lgo-lg20s ; the 187 isotope was definitely found-it constitutes an iso-baric pair of adjacent elements with ls7Re.The methods available for the separation of isotopes have beendiscussed; l o the diffusion method of separation has been appliedto argon l1 and a chemical method to lithium.12NUCLEAR MOMENTS AND SPINS.The magnetic moment of the proton has been measured by I.Estermann, 0.C. Simpson, and 0. Stern,13 using the deflection in aninhomogeneous field of molecular beams of H, and HD. The valueobtained is 2.46 nuclear magnetons & 3%. The nuclear spin of6Li has been shown to be l.14 An atomic-beam method has beenused by H. C. Torrey l5 to obtain the sign of the magnetic momentof 3sK, which is found to be positive; Le., it has the spin direction.This .ciontradicts the result obtained from hyperfine structure ofspectral lines.16 The atomic-beam method gives the positivesign for 7Li, B5Rb, g7Rb, and 133Cs.17Nature, 1937, 140, 149.Physical Rev., 1937, [ii], 51, 289.Naturwiss., 1937, 25, 170, 180.irbid., 52, 1074.Idem, ibid., p.885. * Ibid., p. 933.lo G. Champetier, Bull. SOC. chim., 1936, [v], 3, 1701 ; Nature, 1937,139, 38.l1 H. Kopfermann and H. Kriiger, 2. PhysiE, 1937, 108, 389; H. Barwickl2 G. N. Lewis and R. T. Mulacdonald, J. Amer. Chem. Soc., 1936, 58, 2519.lS Physical Rev., 1937, [ii], 52, 535.l4 J. H. Manley and S. Millman, ibid., 51, 19.l5 Ibid., p. 501.l6 D. A. Jackson and H. Kuhn, Nature, 1936, 137, 107; R. A. Fisher,l7 S. Millman and J. R. Zacharias, ibid., p. 1049.and W. Schiitze, ibid., p. 395.Physical Rev., 1937, [ii], 51, 887BRADDICK. 9The hyperfine structure of barium lines leads to a value $ forthe spin of the odd isotopes.l*THE GENERAL THEORY OF THE NUCLEUS.The nucleus consisting of protons and neutrons may be treatedtheoretically 19 by an approximate wave-mechanical method anal-ogous to that introduced by Hartree for the extranuclear electrons.The method consists essentially in treating each particle as anindividual in a common field, and may be expected to be a reason-able approximation for the lightest nuclei.The model leads tothe existence of neutron and proton shells and to the appearanceof periodic nuclear properties. In particular, the binding energiesof nuclei show periodicities of the type predicted. Although thewave-functions used do not correspond to pre-formed a-particlesin the nuclei, the energies exhibit a four-shell structure.20 Using theresults of Feenberg and Wigner for the wave-functions of the par-ticles, M.E. Rose and H. A. Bethe 21 have calculated the nuclearspins and magnetic moments of the lightest nuclei. The valueof the spin for 'Li, which can be compared with experiment, is givencorrectly by the theory.This kind of approximation is useless when applied to nucleardynamics, particularly in the case of the heavier nuclei.22 On accountof the short range and high intensity of the forces between nuclearparticles, a particle which strikes the nucleus dissipates its energyamong the nuclear particles. Each particle of the " compoundnucleus " will have some energy, but none will have sufficient energyto escape from the rest, so that the compound nucleus will remainin an excited state until the energy is " by accident " concentratedon one particle and allows it to escape.There are a number ofenergy levels for the compound nuclei ; in light elements the spacingis of the order of hundreds of kv., but in heavy elements it is of theorder of a few volts. These energy levels are responsible for thephenomenon of resonance, e.g., in the capture of protons (see p. 13)and of slow neutrons (see p. 16), for the probability of formationof a compound nucleus is specially high when the energy of thesystem which includes the bombarding particle is nearly equal tothe energy of one of the levels of the compound nucleus.18 A. N. Benson and R. A. Sawyer, Physical Rev., 1937, [ii], 52,1Q H. A. Bethe and R.F. Bacher, Rev. Mod. Physics, 1936, 8, 82.20 E. Feenberg and E. Wigner, Physical Rev., 1937, [ii], 51, 95.21 Ibid., p. 205.22 N. Bohr, Nature, 1936, 137, 344; H. A. Bethe, Rev. Mod. Physics,1127.1937, 9, 6910 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.NUCLEAR DISINTEGRATION BY a-PARTICLES.(a; p) and (a; n) Transformations in the Lighter Elements.-Evidence was found for a small yield in the reactionfH + :He +- iH + in + ;He with thorium-C' a-parti~les.~~ Anattempt to detect protons from 6* iLi + :He --+ 's1;Be + ,H usingradium-C' a-particles gave a null result, though the reverse reactionis well known in the case of 6Li.24 The protons from the reactionlOB(a ; p)i3C were examined,25 since the most energetic group previ-ously known (3.3 M.E.V.) did not carry off the energy to be expectedfrom the nuclear masses.A weak group a t 4.7 M.E.V. was found,in agreement with the masses, but it seems that the production ofan excited final nucleus and a 3.3 M.E.V. proton is far more probable.The probability of the reactions 27Az(ol ; n)30P and 26Mg(a ; p)28A1has been studied as a function of the a-particle energy.26 By com-parison with the excitation function of the reaction 27Al( a ; p)30Si,27Waring and Chang conclude that the resonance observed is corre-lated with easy escape of the neutron or proton rather than witheasy penetration of the a-particle barrier. This is in harmony withBohr's view of the nuclear reaction proceeding by the instantaneousformation and disintegration of an intermediate nucleus.28Experiments have been made on the (a ; p ) transmutation of neon,calcium, and argon;29 the first two gave protons with thorium-C'a-particles, but only one group was detected.It appears that 20Nedoes not fit into the series scheme of 24Mg, 28Si, 32S, which givegroups corresponding to excited states of the product nuclei Meringby roughly 1 M.E.V. Pollard and Brasefield give a table in whichthe energies for (E; p ) reactions are used to interpolate betweennuclear masses obtained by the mass spectrograph and in this waymost of the nuclear masses between %e and 40A are determined.E. Pollard, H. L. Schultz, and G . Brubaker 30 found that chlorine andargon gave a considerable yield of neutrons under a-particle bom-bardment. The suggested reactions are 37Cl(a; p)@K andMA(.; p)43Ca.On the assumption that penetration by thea-particles takes place over the top of the potential barrier, thenuclear radii were calculated as 6.1 x cm. for chlorine, and7.3 x 10-13 cm. for argon.23 H. L. Schultz, Physical Rev., 1937, [ii], 51, 1023.24 W. G. Shepherd, R. 0. Haxby, and E. L. Hill, ibid., 52, 674.25 G. Brubaker and E. Pollard, ibid., 51, 1013.2% J. R. S. Warhg and W. Y . Chang, Proc. Roy. SOC., 1936, A , 157, 652;W. Y. Chang and A. Szalay, ibid., 1937, A, 159, 72.27 J. Chadwick and J. R. Constable, ibid., 1932, A, 135, 48; W. R. Kume,Physical Rev., 1937, [ii], 52, 266.2s E. Pollard and C. J. Brasefield, Physical Rev., 1937, [ii], 51, 8.3O Ibid., p. 140.28 See this vol., p.9BRADDICK. 11ArtiJicial Radioactivity produced by a-Particles in Elements up toslBr,-The cyclotron may be used to accelerate He++ ions and togive a strong beam of energetic or-particles; 11 M.E.V. a-particleshave been obt'ained from the Berkeley cyclotron, and this may becompared with the natural a-particles of thorium-C' which havenearly 9 M.E.V. energy. A number of elements undergo transmut-ation in such a beam, and radioactive elements are sometimesformed by (a ; p ) and sometimes by ( a ; n) reactions. The case ofvanadium is interesting and not fully explained.30 It is possiblethat 64Mn exists in two isomeric forms. The formation of 78Brwas specially checked by A. H. Snel13' in his investigation of thebromine isotopes.TABLE I.Radioactive elements produced by a-particle bombardment (adaptedfrom Ridenour and Henderson 31).Type of activityReaction.and half-life. Other known reactions.l0B(a; rt)13N + 10.3 m. radioactive a's14N(a; n)17F + 1-07 m. radioactive a's 160(d; n)l60(p; y )Z6Mg(a; p)28A1 - 2.56 m. radioactive a's 27Al(d; ~)~7Al(n; y)27Al(a; n)3OP + 2-55 m. radioactive a's 32S(d; ~ ) ~ l P ( y ; n )31P(a; n)34Cl + 32 m. radioactive a's B3S(d; n)W l ( a ; n)38K 4- 7-65 m.41K(a; n)%c + 52 h.32, 34 4Wa(d ; n)40Ca(a ; p)43Sc + 4.0 h.33. 34 43Ca(d; n)4%47Ti(a; p)49. 60V ? 35 m.; 3.7 h.35 Ti(d; n)S0Cr(a; n)63Fe ? +Wr(a; p)66Mn - 160 m.63Cu(a; n)esGa + 9.4 h.3'3WU(U; h)G8Ga + 68 m.36 69Ga(n; Zn), ( y ; n)75Aa(a; n)78Br + 6.3 m.37 'OBr(.n; Zn), ( y ; n)V(a; ?)64Mn ? 1.2 m.; 67 m.8-9 m.66Mn(n; y )Br(a; n)Rb + 1.5m.Br(a; n)Rb + 9.8m.NUCLEAR DISINTEGRATION BY PROTONS.The Production of y-Rays by bombarding the Lighter Elements.-R.G . Herb, D. W. Kerst, and J. L. McKibben38 have examineda number of elements for y-ray emission on proton bombardment.y-Rays were found from Li, Be, B, F, Na, Al, but no certain emissionwas found from C, 0, Si, K, Ca, Ni, Cu, Zn, Mo, Pt, Pb. The emissionfrom lithium is known to be due to 7Li; 39 it consists of a strong line31 L. N. Ridenour and W. J. Henderson, PhysicaE Rev., 1937, [ii], 52, 889.32 D. G. Hurst and H. Walke, ibid., 51, 1033.33 H. Walke, ibid., p. 439.35 Idem, ibid., p. 777.37 Ibid., p. 1007.30 L. H. Rumbaugh and L. R. Hafstad, ibid., 1936, 50, 681.34 Idem, ibid., 52, 400.36 W.B. Mann, ibid., p. 405.38 Ibid., 51, 69112 RADIOACTIVITY AND SUB -ATOMIC PHENOMENA.a t 17 M.E.V.40 and one or more weaker lines around 14 M.E.V.E. R. Gaerttner and H. R. Crane 41 find indications of lower energycomponents, The first reaction is presumably the formation of a8Be nucleus, The y-rays may come from transitions within thisnucleus, or from the excited levels in the a-particles produced bythe disintegration of SBe. The former process seems the morelikely, and it fits in with processes suggested for the y-rays fromberyllium and boron bombarded by protons. The excited 8Behas to have a lifetime for a-particle disintegration long comparedwith its radiation life, and according to G .Breit42 the proton iscaptured to form an excited sBe which is supposed to be odd toexclude disintegration into two g-particles. The excitation functionof this disintegration shows a very sharp resonance maximum at440 kv., a broad maximum at about 1000 kv., and a smooth rise from1200 kv. upwards.The bombardment of beryllium gives y-rays up to about 6 M.E.V.,and the excitation shows a broad resonance a t about 990 kv. Thenuclear reaction is probably :Be@; -)':B, but the y-rays of lowerenergy might be emitted from excited product nuclei in :Be@ ; a)gLior :Be@; d)8,Be. Boron gave a y-radiation increasing smoothlywith bombarding voltage, but W. Gentner 43 finds a resonance at180 kv. and a further increase a t 360 kv. This may be comparedwith the excitation functions of the reactions 44(1)(2)l;B + iH --+ 3iHeIiB + iH ---+ !Be + :HeThe excitation of the reaction (2), as measured by the emission ofhomogeneous group of a-particles, also shows a resonance at 180 kv.It is therefore likely that the y-rays come from excited levelsof 8Be.The y-radiation from boron was formerly 45 attributed tothe excited levels of I2C formed by proton capture. It does not seemthat the process is excluded, particularly at high bombardingenergies.Fluorine gave a very large y-ray yield, and the excitation function40 L. A. Delsasso, W. A. Fowler, and C. C. Lauritsen, Physical Rev., 1937,4 1 Ibid., p. 49.42 See L. R. Hafstad, N. P. Heydenburg, and M. A. Tuve, ibid., 1936, 50,43 Naturwiss., 1937, 25, 12; W.Bothe and W. Gentner, 2. Physik, 1937,44 J. H. Williams, W. H. Wells, J. T. Tate, and E. L. Hill, Physical Rev.,46 H. R. Crane, L. A. Delsasso, W, A. Fowler, and C. C. Lauritsen, ibid.,[ii], 51, 391; 52, 582.504.104, 685.1937, [ii], 51, 434.1935, [ii], 48, 102BRADDICK. 13shows prominent resonances at 328,892,942,1400 kv., with a broadresonance at 600-700 kv. and indications of resonance at 1700 kv.46The y-radiation is largely monochromatic at about 6 M.E.V.E. R. Gaerttner and H. R. Crane 47 have obtained positron and elec-tron pairs produced by internal conversion of the y-rays in thetarget, and found a second y-ray line at 4 M.E.V. The nuclei W e ,formed by proton capture, and l60, formed by ( p ; a) change,have both been suggested as providing excited levels for the tran-sition. Sodium gave y-rays with a resonance at about 1200 kv.,but the energy of the y-rays has not been measured.The reactionmay be radiative capture, ;;Na(p ; - )2,iMg ; or ;:Na(p ; a)"Ne, with20Ne in an excited state. Aluminium gave a prominent resonanceat 1370 kv. and some weaker resonances.The theoretical significance of the resonances has been discussedby F. Kalckar, J. R. Oppenheimer, and R. Serber.4* In the re-actions considered the primary effect of the hydrogen particle issupposed to be the formation of a compound nucleus.7Li -/- 'H 4 8BeA + Z4He . . . . . . . . (1419F + lH-+ 20Ne,-+ l60 +4He . . . . . ' ( 2 4'Li + IH--+ 8BeB-> 8Be + y+ Z4He . . . . ( l b )19F + 1H 4 20Nes -+ l60, + 4He + l 6 0 + 4He + y (271)l9F + lH -+ 20Ne, -> 20Ne, + y --+ l60 + 4He + y (2c)IIB + lH~----+ 12C, -+ %eA + 4He I_, 34He .. . ( 3 4llB + lH -+ 12CB -> SBe + 4He (3b)llB + lH --+ 12CB --+ I2C + y ( 3 4The sharpness of the resonance in (lb), (2b or c), (3b or c) indicatesa relatively long lifetime for the excited compound nuclei denotedby subscript B. A satisfactory reason for this has been mentionedabove in the case of 8Be. An argument based on the idea that thetotal spin and orbital angular momentum of the neutrons andprotons can only change slowly with time explains the long life ofthe other nuclei against a-emission. A detailed discussion indicatesthat the y-rays obtained by bombarding fluorine come from l60(Zc, above).Proton-induced Radioactivity in Heavier Nuclei.-The elementsSi, Cu, Cr, Mn, Co, Ni, Zn, As, Se, Mo, Cd, In, Sn, and Sb showradioactivity after bombardment with 3.6 M.E.V.protons.49Data of the activities have not yet been published, but in the case46 L. A. Delsasso, W. A. Fowler, and C. C. Lauritsen, Phyeical Rev., 1937,[iiJ, 51, 527.4 7 Ibid., 52, 582.49 S. W. Barnes, L. A. I)u Bridge, E. 0. Wiig, J. H. Buck, and C. V. Strain,4 8 Ibicl., p. 279.ibid., 1937, [ii], 51, 77514 RADIOACTIVITY AND SUB-ATOMIC PHENOMEBA.of manganese it is supposed that a deuteron is emitted and a radio-active isotope of the metal formed.NUCLEBR DISINTEGRATION BY DEUTERONS.The bombardment of deuterium,50* 51 lithium, beryllium, or car-bon 51 with deuterons provides convenient sources of neutrons andhas been studied from this technical point of view (cf.p. 17). Thecase of deuterium has been investigated theoretically ; 52 thetransmutation function and the angular distribution of the pro-ducts can be calculated by application of quantum mechanics tothe nuclear components. The p-activity of SLi, resulting from7Li(&; p ) [an identical substance is obtained from Li(n; isnow found to be accompanied by an a-activity.a The a-particleshave also been detected by W. A. Fowler and C. C. Laurit~en,~5and by L. H. Rumbaugh, R. B. Roberts, and L. R. H a f ~ t a d . ~ ~Nuclear mass data had shown that there was an excess of energyof about 3 M.E.V. over that observed in the particles from the tworeactions7Li(d ; p)sLi, sLi+ sBe + pand no y-rays from excited sBe could be detected. It now seemsprobable that the 8Li always disintegrates into two a-particles and a@-particle, No a-particles could be detected from the bombard-ment of with deuterons or of 7Li with protons, though thesemight be expected if the a-particles were due to excited *Be.Aselection rule must be invoked to forbid the transition of 8Li tounexcited sBe. The a-particles form a continuous distribution,but the energy balance apparently requires the assumption of aneutrino as in ordinary p-decay (see p. 23).The transmutation functions for the reactions W ( d ; m)13N,14N(d ; n)15O, 160(d ; n)17F have been studied.57 They show thatthe potential barriers for deuterons are at 2.8, 3.2, 3.1 M.E.V.incarbon, nitrogen, and oxygen, respectively. For energies abovethese, the transmutation curves are complicated, probably owingto side reactions. The reactions 30Si(d; ~ ) ~ l S i , 3lP(d; p)32P havebeen found to occur.58 The products are previously known @-emitters of life 170 mins., 14.5 days. The very long-lived 22Na50 W. H. Zinn and S. Seely, Physical Rev., 1937, [ii], 52,919; R. B. Roberts,ibid., 51, 810; H. Kellmann and E. Kuhn, Naturwiss., 1937,25,231; R. Dopel,Ann. Physik, 1937, [v], 28, 87.6 1 E. Amaldi, L. R. Hafstad, andM. A. Tuve, PhysicaZReu., 1937, [ii], 51,896.62 L. I. Schiff, ibid., p. 783; M. H. Johnson, aid., p. 779.58 K. S. Knol and J. Veldkamp, Physica, 1937, 4, 166,64 W. B. Lewis, W. E. Burcham, and W.Y. Chang, Nature, 1937, 139, 24.66 PhysicaE Rev., 1937, CiiJ, 51, 1103.67 H. W. Newson, ibid., p. 620.66 Ibid., p. 1106.Idem, iM., p. 624BRADDICK. 15(positron emitter, 3.0 years) was produced by bombarding mag-nesium with 5-2 M.E.V. deuterons.59 The reaction seems to be!tMg(d ; a)gNa.The reactionsinvolved are probably uCa(d; p)"Ca, g29 93Ca(d; n)41* 439 44Sc,Wa(d ; CC)~~K.The products were separated chemically, and compared withthose of a- and neutron bombardment. The half-lives and types ofemission ascribed to the various isotopes are 45Ca (2-4 h., - ve),4% (53 m., + ve), 43Sc (4.0 h., + ve), (52 h., +ve). No evid-ence for 41Ca from 40Ca ( d ; p ) was obtained, though this reactionwould be expected. This substance may have a very long or a veryshort life.The bombardment of scandium 61 gives 46Sc by a( d ; p ) process. Titanium gives 51Ti witha half-life of 2.8 mins. and a number of isotopes of vanadiumformed by ( d ; n) reactions. These have been compared with otherproducts and assigned as follows : 4*V, 16 days; 49V, 33 mins.;V, 3.7 hrs. Strontium gives active isotopes of strontium andyttrium. There is some evidence that S9Sr has isomeric forms withhalf-lives of 3 hrs. and 55 days, both emitting electrom. It isalso possible that one of the processes observed with strontium isthe capture of a deuteron to formThe bombardment of indium gives periods of 13 secs., 54 mins.,and a few hours ; that of cadmium gives periods of 4.3 hrs., 58 hrs.,and a successive transformation of the latter gives a radioactiveiaotope of indium.The relations of these elements are indicatedin Table II.62At. no. 110 111 112 113 114 115 116 117The activity produced in calcium is complex.60The half-life is 85 days.TABLE 11.'&d 15% 15% 22% 15% 24% 4.3h.J. 16% 58h.$* Percentages are those of stable isotopes, and periods those of radioactiveisotopes ; changes denoted by 4 are #bray changes, and by .f positron changes.Addition of a neutron corresponds to a step 3, and of a proton to a step >.on bombarding palladium with J. D. Kraus and J. M.deuterons, obtained the following radioactive periods :Pd isotopes, 17 mins. (-), 13 hrs. (-)Ag isotopes, 26 mins. (+), 180 hrs. (-)6D L. J. Laslett, PhysicaZ Reu., 1937, [ii], 52, 529.6o H.J. Walke, ibid., 51, 439.62 L. J. Lawson and J. M. Cork, ibid., p. 531 ; J. M. Cork and R. L. Thorn-ton, ibid., 51, 608.e3 Ibid., 52, 763.61 Idem, ibid., 52, 669, 77716 RADIOACTIVITY AND SUB -ATOMIC PHENOMENA.In order to identify these activities they were compared withthose produced by irradiating silver and palladium with fast andwith slow neutrons.64 The results are indicated in Table 111.TABLE 111.At.no. ... 105 106 107 108 109 110 111asPd ...... 23% 27% 2704 13h.J 13.5% 17m.Ja7Ag ...... 180h. ~ 52% 2.3m.$48% 22s.$ 180h.4It may be noted that there are two different long periods in silver,one being produced in a chain reaction. The isotope 106Ag has bothp- and positron-emitting isomers. The reactions in deuteronbombardment include both ( d ; p ) and ( d ; n) types.NUCLEAR DISINTEGRATION BY NEUTRONS.E.Amaldi 65 has given a collective account of the production ofartificial radioactivity by neutron bombardment , which includesa table of the activities obtained up to the middle of 1937. Fewnew cases of the addition of slow neutrons have been found. Thenoble metals 66 have been investigated. Gold gives only the well-known 2.7-day activity; iridium gives, in addition to the 19-hr.and 2-month periods, a l.5-min. period, and platinum gives periodsof 31 mins., 18 hrs., and 3.3 days. Results were also obtained withfast neutrons (see p. 19). It appears that it is necessary to postu-late isomeric nuclei to explain the results.When uranium is bombarded with neutrons, a complicated setof reactions appear^.^' The primary products disintegrate, givingwhole families of radioactive elements, which lie beyond uraniumin atomic number.These have been further investigated by bom-barding uranium with neutrons for different times, so as to bringinto prominence the series beginning with elements of longer orshorter radioactive life. Some of the radio-elements were separatedchemically.26 m.4The families are now found to beB B 9,Eka-Ir %Eka-Pt Eka-Au( 1 )B B B (2) U(n; -)g,U 40s( Eka-Re Eka-0s Eka-IrB (3) U(n; -)92U Eka-Re( '1)64 Cf. p. 19.66 E. McMillan, M. Kamen, and S . Ruben, PhysicaE Rev., 1937, 52, 375.67 L. Meitner, 0. Hahn, and F. Strassmann, 2. Physilc, 1937, 106, 268;6 5 Physikal. Z., 1937, 38, 692.see Ber., 1937, 70, 1374 for chemical aspectsBRADDICK.17The existence of a (n; a) primary reaction 6* has now been dis-proved, though there is evidence of a weak a-emission of long lifesomewhere in the series. The f3-rays of the trans-uranic elementshave been examined in the Wilson chamber, and a y-radiation hasalso been detected.69 The reactions (1) and (2) occur with fast andwith slow neutrons, and the conditions for transmutation, togetherwith the yield obtained, suggest that the process involved is theaddition of a neutron to the abundant isotope ";U. The reaction(3) goes only with slow neutrons and is a very marked case of reson-ance. In this case the reaction can only be a simple addition of aneutron, and the cross-section for capture of slow neutrons shows thatthe abundant isotope is again involved.The phenomena thereforerequire three isomeric ';gU nuclei, and it is not clear how these areto be explained on von Weizsacker's theory of metastable states.70There is some evidence that if the theory applies to these elements,the nuclei of scheme (1) are in metastable excited states.F. A. Heyn 7 1 observed a neutron-induced disintegration of a newtype on bombarding certain elements with neutrons of high energy,obtained by bombarding lithium, beryllium, and hydrogen withdeuterons. The maximum energies of the neutrons me 12, 4.5,and 2.6 M.E.V. respectively. New radio-elements were producedfrom copper and zinc and shown to be isotopic with the originalelements. The character of the copper activity makes it probablethat the reaction isi.e., the removal of a neutron by neutron impact.In the case ofzinc the period the activity is 60 mins., like that produced by slowneutrons. The processes are supposed to beandtXZn(n ; 2n);:Zn (fast neutrons)ttZn(n ; -)$!Zn (slow neutrons)Subsequent work72~73 has shown that this type of activation isvery common, and a number of cases are given in Table IV. Thisincludes also (n; a) and ( n ; p ) reactions now observed with quiteheavy elements on account of the high energy of the lithium neutrons.In the case of scandium,74 the two periods of 4 hrs. and 2 days areidentified with 43Sc and 44Sc, in agreement with the results of68 L. Meitner and 0. Hahn, Naturwiss., 1936, 24, 158; Ann.Reports, 1936,83, 26.6s L. Meitner, Ann. Physik, 1937, [v], 29, 246.70 This vol., p. 21.72 F. A. Heyn, Nature, 1937, 138, 842.73 M. L. Pool, J. M. Cork, and R. L. Thornton, Physical Rev., 1937, [ii],74 Idem, ibid., p. 41.71 Physica, 1937, 4, 160.52, 239.76 Ibid., 51, 43918 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.Element.CN0F 3SiPSc1KC&scTiVCrMn3% coNicuZnGaGeAsSeBrR bSrYTABLE IV.*Activity and assignment. Other ramtiom.20 m.10.5 m.2.1 m.108 m.15 h.10 m.15 h.6 m.11 m.3 m.2.5 h.26 m.14 d.33 m.14 d.7.5 m.4 m.1.8 h.4.5 m.4 h.2 d.1.7 h.28 h.4 m.1.8 d.4 m.1.7 1.1.3.6 m.2.5 h.2.5 h.2.5 h.2 h.6 d.10 m.12.5 d.6 m.40 m.12.5 h.20 m.55 m.1.7 h.22 h.1-3 h.20 h.1.1 d.13 d.1 h.7 m.18 m.4 h.11 m.22 h.18 m.3 h.11 m.1.2 h.6.5 h.2.4 d.+I l'C$ 9 lSN +, l 5 0 +, 18F -, 24Na-, 12Mg-, 24Na +, 27Si +, +, -, Wi +, 31s -, 32P +, a4ci -, 3ZP +, SBK+, 39Ca +, 43scf, 44Sc9 sc3 sc-, 52v-, 52v-Y v-, s6Mn-, 5 w - n -, 56Mn-9 -, "A---, 62Cuj-, 64Cu 6 -, 66Cu +, Znf, 64Cu - , 70Ga +, 88Ga$ 9 -, 72Ga-, 72Ge-, 76A~Y se78Br80Br80Br- Y-- 9 -TI Sr-9-9 -, 90YP(y; n)3 m.82Cu(y; n)10.5 mea2Zn(y; n)38 m.82Ga(y; n)20 m.82Ga(y; n)60 m.82Ga(y; n)23 h.82Br(y; m ) 5 m.82Br(y; %)16 m.82Br(y; n)4.5 h.82* From Pool, Cork, and Thornton, ref.73Element.ZrNbMoRuRhPdAgCdInSnSbTeIB&Lac ePrNdGdDYTuReIrPtAuH gT1PbThUBRADDICK.TABLE IV,-Continued.Activity and msignment.Other reactions.10 m.5 h.44 h.7-3 m.3.8 d.17 m.5 d.24 m.3.6 h.4 m.1-1 h.18 m.12.5 h.25.5 m.13 d.33 m.3 h.53 h.1.1 m.54 m.4 h.2 mo.47 m.15.4 m.2.3 d.1.1 h.30 d.26 m.2.5 m.85 m.2.2 h.40 m.3 m.20 h.2.2 h.19 h.19 h.2.5 h.9.1 h.18 h.15 h.1.8 h.3 d.17 m.2.5 d.45 m.5 m.50 m.5 m.1.5 h.5 m.1.4 h.26 m.4 h. -,13 h. - , Eka-0sMo(y; n)17 m.82A g ( y ; n)24 m.82In(y; n ) l . l m.82Sb(y; n)15 m.8819who produced the same isotopes by bombarding calcium withdeuterons and potassium with or-particles.The case of bromine is specially interesting, since three radioactiv20 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.periods are produced whereas bromine has only two stable isotopes.Since it is most unlikely that the y-ray can remove two neutronsfrom the nucleus, this effect is believed to be a case of nuclearisomerism, and SOB, to exist in two forms with different radioactivepr0perties.~6 The nuclear cross-section for this process was estim-ated for the case of copper as 5 x and in the case of theother elements the probability of disintegration was of the sameorder of magnitude.Many elements, however, show an unobservable small effect,so it is probable that a resonance process occurs.The effect is afew hundred times greater than predicted by H.A. Bethe and G .Placzek 77 on their assumption of nuclear resonance levels spacedso closely that selective resonance does not occur.The production of 43Sc by neutron bombardment involves theejection of three neutrons (including the bombarding particle).This is not inconsistent with the Bohr view of an excited intermediatenucleus. It is possible that the reactions 19F(n ; 3n)17F(17FY 1.2 mins.)and 63Cu(n; 3n)61Cu(61Cu, 3.5 hrs.) occur in this way, but theidentification of these activities is not certain.THE NUCLEAR PHOTO-EFFECT.The extraction of neutrons from a nucleus by y-rays ( y ; ntransition or nuclear photo-effect) 78 has hitherto been known onlyfor 2H(y; n)lH and sBe(y; n)8Be. F. A. Paneth and E. Gliickaufhave detected helium by microchemical methods and decided that4He is the main product of the beryllium reaction.79 The nuclearcross-section for the effect in beryllium has been obtained for twovalues of the 7-frequency, and there is satisfactory agreement withcalculation based on the method of Bethe and Peier1s.mThe nuclear photo-effect depends simply on the energy of the y-raybeing sufficient to extract a particle from the nucleus. The reverseprocess (capture of a neutron with emission of a y-ray) *l is known, andthe energies have been estimated.A great extension of the nuclearphoto-effect has been made 82 by the use of the very energeticy-rays (about 17 M.E.V.) obtained by bombarding lithium with' 6 Cf. this vol., p. 21 and refs. 37, 82.7 7 Physical Rev., 1937, [GI, 51, 450.78 Ann.Reports, 1935, 32, 28; 1936, 33, 26.7Q Nature, 1937, 139, 712.L. I. Rusinov and A. N. Sagaidak, Physikal. 2. Sovietunwn, 1936, 10,203; V. I. Mamaschlisow, ibid., p. 214; H. A. Bethe and R. Peierls, Proc.Roy. Xoc., 1935, A , 148, 146.81 R. Fleischmann, Naturwiss., 1936, 24, 77; 2. Physik, 1936, 103, 113;S. Kikuchi, K. Husini, and H. Aoki, Nature, 1936, 137, 992.82 W. Bothe and W. Gentner, 2. Physik, 1937, 106, 236; preliminaryannouncements in Naturwiss., 1937, 25, 90, 126, 191, 284BRADDICK. 21protons.83 A large number of elements were examined, and thepositive results obtained are indicated in Table V. The productsof the photodisintegration were radioactive elements which wereidentified chemically as isotopes of the original elements.In somecases they could be shown by their radioactjive properties to beidentical with the products of previously known nuclear reactions.There is an obvious and close connection between the photoelectricextraction of a neutron from the nucleus and the removal of a neutronby impact in the ( n ; 2n) reactions described on p. 20. There isin most cases a satisfactory identity between the properties of theradioelements obtained in the two ways.TABLE IT.P, Cu, Zn, Ga, Br, Mo, Ag, In, Sb as indicated in Table IV. In addition :Br(y; n)36 h. Ag(y; n)22 s. Te(y; n)60 m.NUCLEAR ISOMERISM.Recent studies of nuclear transmutation in indium,62 rhodium,bromine, uranium,67 silver,63 and iridium and platinum 66 have ledto the conclusion that there exist nuclei identical in charge andmass but differing in radioactive properties.I n the naturalradioactive series, UX, and UZ, have long been believed to be anisomeric pair with 2 = 91 and A = 234.84 C. F. von Weizsacker 85has suggested that the difference between these isomers is due todifferent states of excitation. The y-ray transition between anexcited state of a nucleus may be " forbidden " by selection rulesif it involves a spin change, and the probability of transition isreduced for greater spin changes. It appears that for reasonablevalues of the spin change the probability may be so low that themetastable excited nucleus has a sufficiently long life to accountfor nuclear isomerism.Furthermore, the paray transition is governed by selection rules,so the product nucleus from a P-disintegration of a meta-stablenucleus may itself be in a metastable excited state.This accountsfor the production of parallel families of isomers as in the case ofuranium.I n the cases of 54Mn and Ag 63 it is believed that one nuclearisomer emits positrons, and the other electrons.It has been shown 86 that the production of y-rays by fast neutrons83 L. A. Delsasso, W. A. Fowler, and C. C. Lauritsen, Physical Rev., 1937,[ii], 51, 391, and references therein; this vol., p. 11.84 0. Hahn, 8. physikal. Chem., 1922, 103, 461.85 Naturwiss., 1936, 24, 813.86 G. T. Seaborg, G. E. Gibson, and D. C. Grahame, Physical Rev., 1937,[ii], 52, 40822 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.in a number of elements is due to inelastic collisions in which theneutron raises the nucleus to an excited state and goes on withdiminished energy.THE INTERACTION BETWEEN NEUTRON AND PROTON.It is now generally assumed that neutrons and protons are thefundamental constituents of all nuclei and it is important to obtaindata on the forces between them.The stable existence of thedeuteron shows that this force is attractive and the binding energyof the deuteron has been determined as 2.25 M.E.V.87 The inter-action of neutrons and protons may be studied by investigating thescattered particles when neutrons pass through hydrogen gas in acloud chamber.Former experimental data have been rather conflicting, but therecoil proton tracks produced by the homogeneous 2-6 M.E.V.neutrons from the D(d.3He) reaction have now been investigated byseveral groups of workers.88 The distribution of recoil particles seemsto be nearly spherically symmetrical with respect to the centre ofmass of the moving particles.The theory of these collisions hasbeen worked out 89 under a variety of assumptions. It appearsthat the observed isotropic distribution shows that the interactionbetween neutron and proton is very small a t separations greater than3 x 10-13 cm., and is not consistent with '' exchange forces " betweenthe particles. It is not, however, yet possible to obtain any preciseconclusions about the shape of the potential well. In these calcula-tions it is assumed that the range of the interaction does not dependon the spins of the particles.The binding energy of the deuteron isdue to the interaction energy of the particles with parallel spins,while the large scattering of slow neutrons by hydrogen indicates,according to Wigner,go a weaker interaction with antiparallel spins.J. Schwinger and E. Tellerg1 have shown that if the interactiondepends on the spin of the particles, the scattering of very slowneutrons by ortho-hydrogen (molecule with parallel proton spins)should be much greater than for para-hydrogen (antiparallel spins)87 J. Chadwick, N. Feather, and E. Bretscher, Proc. Roy. SOC., 1937, A ,183, 366.P. I. Dee and C. W. Gilbert, ibid., p. 265; P. G. Kruger, W. E. Shoupp,and F. W. Stallmann, Physkd Rev., 1937, [ii], 52, 678; T.W. Bonner, ibid.,p. 685; G. W. Lampson, D. W. Mueller, and H. A. Barton, ibid., 51, 1021.*O H. S. W. Massey and R. A. Buckingham, Proc. Roy. Soc., 1937, A , 163,281; H. A. Bethe and R. F. Bacher, Rev. Mod. Physics, 1936, 8, 82; P. M.Morse, 3. B. Fisk, and L. I. SchifT, Physical Rev., 1937, 51, 706; S. Shareand J. R. Stehn, ibid., 52, 48.O0 See Bethe and Bacher, ref. 89.s1 Physical Rev., 1937, [ii], 52, 286BRADDICK . 23on account of an inelastic scattering process which appears in theformer case. This effect has been found by e~periment.~2 Thescattering of neutrons by deuterons has been calculated, in fairagreement with experiment .93THE MAGNETIC MOMENT OF THE NEUTRON.It has been found possible to produce partly polarised neutronbeams (Le., beams which contain more neutrons of one direction ofspin than the other) by passing a beam of slow neutrons through aplate of magnetised iron.94 This effect is due to the interaction ofthe magnetic moment of the neutron with that of the ferromagneticatoms.0. R. Frisch, H. von Halban, and J. Kochg5 showedthat a neutron beam polarised in this way should be depolarisedby passing through an axial magnetic field on account of theprecession of the magnetic neutrons, and they found that the mag-netic moment of the neutron was probably 2 nuclear magnetons.The sign of the magnetic moment was shown to be negative,96 andprobably -2 nuclear magnetons as expected from the proton anddeuteron moments. An attempt has been made to detect themagnetic scattering of slow neutrons, by comparing the scatteringof a mixture of manganese and sulphur with that of the sulphidecontaining the manganous ion.No positive results have beenobtained .9THE RAY DISINTEGRATION.The present position of the @-disintegration has formed the sub-ject of a Royal Society discu~sion.~~ The main difficulty remainsthat the @-particles form a continuous energy distribution and thatthis fact cannot be directly reconciled with the conservation ofenergy and the identity of all the atoms of the initial and finalsubstances. A suggestion that the nuclear P-rays might differin mass from other electrons has been tested, with negative results,by C. T. Zahn and A. H. S p e e ~ . ~ ~ The Fermi theory and its deriv-92 J.Halpern, I. Estermann, 0. C. Simpson, and 0. Stern, Physical Rev.,1937, [ii], 52, 142; J. R. Dunning, J. H. Manley, H. J. Hoye, and F. G.Brickwedde, ibid., p. 1076.09 L. I. SchB, ibid., p. 149.94 F. Bloch, ibid., 1936, [ii], 50, 259; J. G. Hoffmann, M. S. Livingston,and H. A. Bethe, ibid., 1937, 51, 214; J. R. Dunning, P. N. Powers, andH. G. Beyer, ibid., 51, 51, 371, 1112.95 Nature, 1937, 139, 756, 1021.0 8 P. N. Powers, J. R. Dunning, H. Carroll, and H. Beyer, Physical Rev.,9 7 M. D. Whitaker, ibid., p. 384; 0. Halpern and M. H. Johnson, ibid.,0 8 C. D. Ellis et al., PTOC. Roy. SOC., 1937, A , 161, 447.09 Physical Rev., 1937, [ii], 52, 524.1937, 52, 38.p. 5224 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.atives 1 postulate that the nucleus contains neutrons and protonswhich are to be considered as states of the same particle.Where itis energetically possible, one of these states may switch over to theother with emission of a P-particle (n + p ) or a positive electron( p --+ n). The energy of the disintegration is represented by themost energetic p-ray emitted, and the simultaneous emission of aparticle called the neutrino is required to explain the continuousenergy distribution. The modifications of this theory depend ondifferent assumptions about the law of interaction between nuclearparticles. The theories give results, which can be compared withexperiment, for the form of the p-ray energy distribution curve andfor the relation between the disintegration constant of t,he nucleiand the energy of the P-particles. The conclusion that the upperlimit of the p-spectrum corresponds with the disintegration energyhas been very well confirmed by measurements of the p-spectrumin light radioactive elements.2 In the case of carbon bombardedwith deuterons the reactions12C 4- 2H -+ I3C + lH + Q1 or 13N + n + Q213N -+ 13c + p+ + &3 + vwhere Q1 and QZ are energies carried off respectively by the protonsand neutrons in the two reactions, Q3 is the energy of the @-trans-formation, and v is the neutrino mass.Mass-spectrograph resultsexclude the possibility of energy loss by y-radiation. Q3 (+ energyequivalent to v) = 1.08 f 0.1 M.E.V., whereas the estimated end-point of the p-spectrum is a t 1-25 & 0.1 M.E.V.Similar resultswere obtained by the reactions I O B ( d ; n or p ) , 160(d; n or p ) , and27Al( cc ; n or p ) . The results show also that the neutrino mass mustbe small and the limits of error place it as less than one-fifth of theelectron mass.The general form of the energy distribution of the p-particlesfrom radium-E3i4 and 32P395 fits over a large part of the rangewith the theory of Konopinski and Uhlenbeck. The high end-point of the spectrum does not agree with the theory, for this pre-dicts an asymptotic tail which is definitely excluded by theexperiments of Lyman and of Paxton. Similarly, work on thep-spectra of 8Li and 12B shows that the end-point obtained by1 E. Fermi, 2. Physik, 1934, 88, 161 ; E. J. Konopinski and G. E. Uhlen-2 J.D. Cockcroft, Royal Society Discussion, Zoc. cit., ref. 98.3 C. M. Lyman, Physical Rev., 1937, [ii], 51, 1.4 L. M. Langer and M. D. Whitaker, ibid., p. 713; J. S. O’Conor, ibid.,beck, Physical Rev., 1935, [ii], 48, 7, 107.52, 303.H. C. Paxton, ibid., 51, 171.6 D. S. Ragley and H. R. Crane, ibid., 52, 604BRADDICK. 25extrapolation according to Konopinski and Uhlenbeck is much toohigh. The lower-energy part of the thorium-C’ p-spectrum ha,sbeen investigated by using a radioactive vapour in an expansionchamber.’ This method avoids complications due to the absorptionand scattering of the low-energy particles in window or support.The energy distribution does not go to the origin, but shows a largenumber of particles of low energy.This is in accord with theKonopinski-Uhlenbeck formula, if the electrostatic effect of thenuclear charge on the escaping P-particle is taken into account.Similar conclusions follow from Paxton’s work on 32P and B. Z.Dzelopov’s on 28A1, 30P, 152E~, and Ra-E.8 H. 0. W. Richardsonfinds that the shape of the spectrum is given over a wide range by acombination of the Fermi and the Konopinski-Uhlenbeck form~lae.~It is not yet certain how far theory agrees with experiment in thecorrelation of p-ray energies and radioactive lives. A. C. G .Mitchell lo makes Sargent logarithmic plots of the disintegrationconstants against the end-points of the p-ray spectra of a numberof radio-elements as determined by Konopinski-Uhlenbeck extra-polation. He finds that each point lies on one of three lines (Sargentcurves).The lines are supposed to correspond with different spinchanges (“ allowed ” and “ forbidden ” transitions), and thereis a partly successful attempt to fit the spins of a number of nucleiinto a simple scheme.ll Richardson 9 shows that the Konopinski-Uhlenbeck expression for the relation between life and energyis not in agreement with experiment for the P-rays from Th-C”,Th-B, and Ra-D, which are all “ permitted ” transitions, and thatthe Fermi formula fits these cases better. On the other hand,the Konopinski-Uhlenbeck expression gives a correct representationof the relative periods of p-decays of high energy, with end-points 2-12 M . E . V . Richardson finds that a formula containingboth Fermi and Konopinski-Uhlenbeck terms gives a goodapproximation throughout the spectrum.THE PASSAGE OF ENERGETIC p- AND RAYS THROUGH MATTER.The important application of the quantum theory of the inter-action of hard p- and y-rays with matter to the cosmic-ray problemis described on p.26. Several investigations have been made onthe absorption of electrons of moderately high energies (1-11H. 0. W. Richardson and A. Leigh-Smith, Proc. Roy. SOC., 1937, A , 162,391.8 Bull. Acad. Sci. U.R.S.S., 1936, 673.9 Royal Society Discussion, loc. cit. ; Nature, 1937, 139, 505.10 Physical Rev., 1937, [ii], 52, 1.11 Cf. calculations of M. Rose and H. A. Bethe, ibid., 51, 20526 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.M.E.V.) in heavy gases l2 or thin laminae l3 in a cloud chamber.In the experiments with gases it was concluded that large, suddenenergy losses, due presumably to production of radiation, occurredmuch more often than was to be expected from the current theory.The energy loss found by Turin and Crane in a light element, carbon,was in good agreement with theory.Radiation is here small.In lead, they found an average energy loss rather greater thantheoretical, but the discrepancy was not much greater than could beaccounted for by the increased path in the lead due to scattering.Laslett and Hurst, on the other hand, found energy losses markedlygreater than theoretical, and a special investigation of large energylosses showed a frequency many times the theoretical. A recentrevision of the theory 14 does not much alter the theoretical result,so that further investigation will be needed.There is some evidencefor pair production by C3-1-ays.l~ In the case of energetic y-rays,pair production becomes a most important interaction with matter.The angular distribution of the pairs has been studied.16 It hasbeen shown 1 7 that the study in a cloud chamber of the pairs pro-duced by y-rays is a convenient and reliable method of investigatingthe spectrum of the latter.THE PENETRATING RADIATION.Considerable advances have been made in the study of the cosmicradiation as a result of the application of quantum-mechanicalcalculations, known to be more or less valid in the medium-energyrange, to electrons of high energy.l**l9 A fast electron will loseenergy in the form of radiation quanta (" Bremsstrahlung "), andthe quanta will in turn produce positron-electron pairs.A cosmic-ray shower will thus arise as a complicated cascade process, and allthe particles and photons will preserve the general direction of theprimary particle. Direct calculations on these assumptions givethe numbers of emergent particles when a single particle falls on12 H. Klarmann and W. Bothe, 2. Physilc, 1936, 101, 489; L. Leprhce-Ringuet, Ann. Physique, 1937, 7, 5 .1s 5. J. Turin and H. R. Crane, PhysiccsE Rev., 1937, [ii], 52, 63, 610; L. J.Laslett and D. G. Hurst, ibid., p. 1035.J. C. Jaeger, Nature, 1937, 140, 108.lS F. C. Champion and A. Barber, ibid., p. 105; M. MonadjBmi, Compt.rend., 1937, 204, 1560.l6 L.Simona and K. Zuber, Proc. Roy. SOC., 1937, A, 159, 383; H. Adam,Naturwiss., 1937, 25, 13.1 7 L. A. Delsasso, W. A. Fowler, and C. C. Lauritsen, Physical Rev., 1937,[ii], 51, 391. Cf. E. R. Gaerttner and H. R. Crane, ibid., 52, 582.la H. J. Bhabha and W. Heitler, PPOC. Roy. SOC., 1937, A, 159, 432; 5. F.Cerlson and J. R. Oppenheimer, PhysicaE Rev., 1937, [ii], 51, 220.W. Heitler, Proc. Roy. SOC., 1937, A , 161, 261BFtADDICK. 27absorbers of different thicknesses. They reveal the main featuresof the Rossi transition curve in which number of showers is plottedagainst thickness of shower-producing material.lQ~~ They arefurther supported by special transition-curve experiments,21 andby the appearance of obvious cascade processes in many cloud-chamber photographs of showers.Similar calculations may bemade to give the main features of the phenomena of the passageof comnic rays through the upper atmosphere.22I. S. Bowen, R. A. Millikan, and H. V. Neher 23 measured cosmic-ray intensities to great heights in the atmosphere a t two differentlatitudes, and the difference between these sets of measurementsis interpreted as the effect of electrons cut off by the earth's magneticfield at the lower latitude and having energies between 6700 and17,000 M.E.V. These electrons behave in fair accord with thequantum theory predictions. The theory apparently gives a satis-factory account of the behaviour of the soft part of the rays (ab-sorbed in 10 cm. of lead) and of the greater part of the showers.The penetrating particles, which form about 80% of the verticalintensity a t sea-level and can in some cases penetrate hundreds ofmetres of water, have no obvious place in the theory, and their natureis one of the outstanding cosmic-ray problems.Heitler l9 examinesas alternatives the possibilities that they are non-electronic particles,and that the theory breaks down for very high electron energies.It appears that a breakdown of the formule for energies below about10,000 M.E.V. would spoil the explanation of the curve in the upperatmosphere. According to P. M. S. Blackett and J. G. Wilson?very few electrons following the theory a t high energies are observedat sea-level. They measured the energies of a large number ofindividual particles before and after passage through a metd plate,and found that the energy loss of particles of energy below about250 M.E.V.was nearly that given by the quantum theory for elec-trons. The particles above this limit were in the main fax morepenetrating, though a few electrons were also found. The energyloss for the penetrating particles appears to pass through a mini-mum and then increase very slowly with increasing energy. Nopenetrating particles were found with energies lower than thecritical value, and it seems that penetrating particles must appearas electrons when they enter the lower-energy range. In anunpublished experiment 25 the behaviour of particles in a gold plate20 W. H. Furry, Physical Rev., 1937, [ii], 52, 569.21 P.Auger, P. Ehrenfest, A. Freon, and (Mlle.) T. Grivet, Compt. rend.,22 See Ann. Reports, 1930, 33, 33.24 proc. Roy. SOC., 1937, A , 160, 304.26 p. M. S. Blackett, ibid., in the press.1937, 204, 1797.Physical Rev., 1937, [ii], 52, 8028 RADIOACTIVITY AND SUB-ATOMIC PHENOMENA.was studied after preliminary passage through a thick lead filter.The particles after passage through the lead still showed divisioninto a penetrating and a non-penetrating group, with a demarcationat 200-250 M.E.V., and this experiment suggests that penetratingparticles may turn into non-penetrating particles. C. D. Andersonand S. Neddermeyer,26 working in the energyrange 100-500 M.E.V.,found penetrating andnon-penetrating particles throughout the range.Particles occurring singly were penetrating, and electronic particlesoccurred in showers or produced showers.Their experimentalresults are in conflict with those of Blackett and Wilson, whofind only electronic particles over the lower part of the energyrange, and this discrepancy may reasonably be explained by thelower accuracy of their energy determinations. Anderson andNeddermeyer suggest that the penetrating particles have an elec-tronic charge combined with a mass between those of the electronand the proton. Unfortunately, particles can only be distinguishedin mass by their specific ionisation when near the end of theirrange, and there is little evidence of this kind for the existence of aheavy electron. E. C. Stevenson and J. C. Street 27 used a cloudchamber arranged specially to catch particles near the end of theirrange, and observed one proton and one particle estimated to have130 times the electronic mass. These authors2* also showed,from consideration of the range of penetrating particles of measuredenergy, that many particles were neither Heitler electrons norprotons .The very penetrating part of the cosmic rays has been studiedunder 240 m. of water,29 and shower intensities proportional to thevertical intensities were obtained down to this depth. Showers andbursts have been detected under the equivalent of 800 m. of water.30Since Heitler electrons of reasonable energy cannot penetratemore than a few metres of water, these showers must be producedby electrons secondary to the hard radiation, or by an entirelydifferent process-perhaps that suggested by W. Hei~enberg.~~Unpublished work by L. J&nossy32 suggests that both these pro-cesses are operative.The existence of heavy particles, protons,33 and neutrons 3426 Physical Rev., 1937, [ii], 51, 885.27 Ibid., 52, 1003.20 A. Ehmert, 2. Physik, 1937, 106, 751.ao Y. Nishina and C. Ishii, Nature, 1936, 138, 721.31 2. Physik, 1936, 101, 533.32 Cf. P. Auger and G. Meyer, Compt. rend., 1937, 204, 572.3s C. D. Anderson and S. Neddermeyer, Physical Rev., 1936, [ii], 49, 415.34 E. Schopper, Naturwiss., 1937, 25, 557; B. Arakatsu, K. Kimura, and28 Ibid., 51, 1005.Y. Uemura, Nature, 1937, 140, 277BRAD DICK. 29has been reported in cosmic-ray experiments, and these are usuallyregarded as secondaries produced, e.g., by the nuclear photo-effect.A suggestion has been made 35 that the magnetic field of the sunacts to prevent cosmic electrons of low energy from reaching theearth. This explains the observation that the effect of latitudeon cosmic-ray intensity ceases above about 50" N., even at a con-siderable altitude.36 It is not yet certain if the sun's field is adequatefor this effect.H. Alfvbn 37 has suggested a mechanism for accelerating electronsto very high energies in a double star. The electric and magneticfields of the star behave like those in a cyclotron, and the electronsare accelerated by multiple stages. The energies calculated onreasonable assumptions about the field are of the right order ofmagnitude to account for cosmic rays.H. J. J. BRADDICK.85 L. Jhossy, 2. Physik, 1937, 104, 430; M. S. Vallarta, Nature, 1937,86 M. Cosyns, ibid., 1936, 137, 616.87 Ibid., 1936, 138, 761; Compt. rend., 1937, 204, 1180.139, 839