SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.THE important work of the two years (1931-32) under review hasagain for the most part been physical in nature; it has been aneventful time for nuclear physics. A new, possibly ultimate,particle, the neutron, has been discovered, and it has already provedto be a useful weapon in investigations on nuclear structure; ithas been eagerly seized on by theorists interested in the structureof the atom. It has been found that protons when generated withsufficiently high velocities may bring about nuclear disintegrationby bombardment similar to that effected by the a-particle. Theneutron may also act as a projectile in a similar way. Many newisotopes of non-radioactive elements have been found, and theirmasses determined with a greater exactness than heretofore.Inthis work the magnetic-spectrograph and the ordinary spectroscopeare assisting the mass-spectrograph more and more as instrumentsof investigation ; the most remarkable isotope found during theperiod, that of hydrogen with a mass of 2, falls to the credit ofband-spectrum analysis. This combination of methods appearsto be so promising as probably to render superfluous in the nearfuture the older chemical methods of determining atomic weights.There has been a considerable advance in knowledge of the originof the y-ray, especially in its relation to the rare, high-velocity,a-particle ; also of the properties of the penetrating radiation,although its exact nature and origin still elude investigators. Onthe chemical side, the simplicity that ascribed all a-particle radio-activity to atoms of high atomic number has been surprisinglydisturbed in two ways.Element 87, apparently without detectableradioactivity, has been detected both by X-ray analysis and bythe magnetic-spectrograph, while samarium of atomic number 62has been found to be radioactive, expelling cc-particles.Radioactivity of Samarium.G. von Hevesy and M. Pahl1 have made preliminary observationson the radioactivity of the rare-earth element samarium (at. wt.150.4). This is of the a-particle type ; consequently, samariumis the first element outside the range of heavy elements thallium-uranium to show this type of activity. A layer of samarium oxidehas an activity of about one-third of that of a thick layer of potass-ium chloride of equal surface.The radiation is reduced to half-Nature, 1932, 130, 846300 SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.value by aluminium of thickness 1.3 v. Preparations from differentsources showed the same specific activity. Chemical purificationindicated that the activity is not due to known radioactive elements ;it is suggested, however, that it may be due t o the very rare element61, to which i t is chemically very similar.Discovery of Eleinerht 87.Despite the extreme unlikelihood 2 of the existcnce of clcments85 and 87 in nature, on account of the known instability of theneighbouring elements 88, 86, and 84, two pieces of evidence havebeen put forward to support the existence of the missing memberof the alkali metals.worked up 10 kg.of samarskite rich in uranium, containing rubidium and czsium,so as to concentrate the last. The X-ray lines M c c ~ , L a l , L M ~ , Lp,, andL-q, calculated from Moseley’s diagram for element 87, were foundin the concentrate 011 excitation in a Siegbahn apparatus of highdispersion ; the Lp, line, for example, calculated to be 0.8524, wasfound as 0.853 A. The element is regarded as non-radioactive orvery feebly radioactive ; radioactivity, however, has yet to bestudied in detail. L. L. Barnes and R. C. Gibson4 have foundindependent evidence of the new element by examining with aDempster magnetic-spectrograph alkali sulphates known fromX-ray examination to contain traces of element 87.A mass of220 f 1 was the only one which could not be assigned to knownelements. Now it is known on general grounds that the isotopesof element 87 would be 221 and 223 or 221 and 219, an element ofodd atomic number having a maximum of two isotopes each ofodd atomic mass. The mass found is therefore confirmatory ofthe existence of the new element, which can thus be provisionallyassigned masses of 219 and 221 or, if simple, a mass of 219 or of221. F. Allison, (Miss) E. R. Bishop, A. L. Sommer, and J. H.Christensen 5 have described experiments with a magneto-opticalmethod, depending on the time-lag differences of the Faradayeffect behind the magnetic field, on solutions of minerals. Insolutions of pollucite and lepidolite, minima were observed at-tributed to element 87.Six minima are considered to indicatethe probable existence of element 85 also. They have sufficientconfidence in their observations to give names to these two elements,5. Papish and E. WninerAnn. Reports, 1928, 25, 317.J . Amer. Chem. SOC., 1931, 53, 3818; A., 1931, 1348.J . Amer. Chem. SOC., 1932,54, 613 ; A., 317 ; F. Allison and E. J. Murphy,Physical Rev., 1930, [ii], 36, 1097. See ibid., 35,285 ; A., 1931, 1391 ; J . Amer.Chem. SOC., 1932, 54, 405, 616; A . , 353, 355.* Physical Bev., 1932, [ii], 40, 318RUSSELL . 301but this is not yet shared by other workers . Preparations foundt o contain comparatively large amounts of element 87 by themagneto-optical methods were found not to contain i t by X-rayexamination.3 Further.when the method has been applied in aregion where the results are well attested. as for example to thecomplexity of copper or tantalum. 5 5 '9 l9 the results are a t variancewith those obtained by the mass.spectrograph . It is extremelynnlikely also that an element of odd atomic number. like number55. has more than two isotopes .Isotopes and JInss.&ertra .During the period knowledge of the isotopic composition oftwenty-one elements has been extended.' The results are sum-marisecl in Tables I and I1 . In thc latter. inem atomic weightsTABLE I .MinimumAtomic number of Masses (nearest integer) of isotopesElement . number . isotopes . in order of abundance .Hydrogen ......... 1 2 1. 2Beryllium ......... 4 2 9.8Neon ............... 10 4 20.22.21. 23Scandium ......... 21 1 45Rubidium ......... 37 2 85. 87Strontium ......... 38 3 88. 86. 87Niobium ............ 41 1 93Caesium ............ 55 133Barium ............ 56 4 138.137.136. 135Tantalum ......... 7 3 1 181Rhenium ......... 75 2 187. 185Osmium ............ 76 6 192.190.189.188.186. 187Mercury ............ 80 0 202.200.199.201.198.204.196.Thallium ......... 81 2 205. 203Lead ............... 82 8 208.206.207.204.209.210.203.205.Uranium ......... 92 1 238TABLE 11 .Ruthenium ...... 44 (y) 102.101.100.99. (98). 96197. 203Calculated InternationaAtomic Packing atomic weight atomicElement . number . fraction . (0 = 16) . weight . 7aLithium ............... 3 - 6.928 & 0.008 6.94Boron ..................5 - 10.794 & 0.001 10.82Scandium ............ 21 - 7 44.96 A 0.05 45.10Zinc .................. 30 - 9.9 65.38 -J: 0.02 65.38Niobium ............... 41 ca . - 8 92.90 & 0.05 93.3Ruthenium ......... 44 ca . - 6 101.1 101.7Tin ..................... 50 - 7.3 118.72 f 0.03 118.70Caesium ............... 55 - 5 f 2.0 132.92 0.02 132.81Tantalum ............ 73 ca . - 4 180.89 i 0.07 181.4Osmium ............... 76 - 1 4 2.0 190.31 4 0.06 190.8Thallium ............ 8 1 1.8 5 2 204.41 & 0.03 204.39Rhenium ............ 75 - 186.22 0.07 186.316 (Miss) E . R . Bishop. PhysicaE Reu., 1932.40. 16; A., 554 .J . Amer . Chem . SOC., 1931. 53. 1627 . 7u J., 1933. 11 5 302 SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.of twelve elements, calculated from knowledge of the number ofisotopes, their relative abundance, and their packing fractions,are compared with the International values (0 = 16).The new isotope of hydrogen was found by H.M. Urey, F. G .Brickwedde, and G. M. Murphy.8 If isotopes H2 and H3 exist,it is expected thermodynamically that they should be concentratedwhen hydrogen is evaporated near the triple point. In specimensso concentrated, faint Balmer-spectrum lines were found a t thecalculated positions for H2 as broad doublets after an exposure400 times the normal one. The abund-ance of H2 to H1 in ordinary hydrogen was estimated as 1 : 4000.W. Bleakney found this ratio in ordinary electrolytic hydrogenas 1 : 30,000 & 20%. H. Kallmann and W.Lasarev lo found theabundance of ions of mass 3 to those of mass 2 a t low pressuresby a spectrographic method as 1 : 4000. Working with enrichedhydrogen, K. T. Bainbridgell determined the mass of the newisotope as 2.01351 f 0.00018 (0l6 = 16) by comparing the positionsof HiH2+ and He+ on a microphotometer record; he deduced thebinding energy of H2 according as it is imagined as built up of twoprotons and one electron or one proton and one neutron. N. S.Grace l 2 deduced theoretically the mass as 2.0113 f 0.0012. E.W. Washburn and H. C. Urey13 found that H2 was easily concen-trated by the fractional electrolysis of water; the residual waterof cells which have operated for a few years contained a markedincrease in abundance of H2 relative to H1.In the infra-redabsorption spectrum of hydrogen chloride, band lines correspondingwith H2CP5 and HVP7 have been found by J. D. Hardy, E. F.Barker, and D. M. Dennison 14; the abundances of H2C1 in ordinaryand in enriched hydrogen chloride were found to be 1 : 35,000 and1 : 3500 respectively and the mass of H2 was deduced as 2-01367 &-0.0001 in satisfactory agreement with other provisional values.The ease with which H2 can be concentrated in ordinary hydrogenmarks it off sharply from all other isotopic mixtures; the result isto be expected in view of the relatively enormous difference betweenthe two isotopic masses. That H2 could exist at all would havebeen regarded as most extraordinary had not the neutron l 5 almostsimultaneously been found.Although H3, He3, and He5 have been sought and not found:Physical Rev., 1932, [ii], 40, 1 ; A., 554.@ Ibid., 41, 32; A., 894.lo Naturwiss., 1932, 20, 206, 472; A , , 442, 790.11 Physical Rev., 1932, [ii], 42, 1; A., 1185.l2 J. Amer. Chem. Xoc., 1932, 54, 2562; A., 790.l3 Proc. Nut. Acad. Sci., 1932, 18, 496; A., 894.lP Physical Rev., 1932, [ii], 42, 279.l5 J. Chadwick, Nature, 1932, 129, 312; A., 443.No trace of H3 was foundRUSSELL. 303there is some evidence for Bes. W. W. Watson and A. E. Parker l6found weak satellites in the band spectrum of beryllium hydridewhich they ascribe to Be8; the relative intensities of the hydridesof Be8 and Be9 were estimated as 1 : 2000. A fourth isotope ofneon, Ne23, is claimed by G. Hertz.17 He succeeded in raisingthe ratio of Ne20 to Ne22, normally 10 : 1, to 100 : 1 and in loweringit to 10 : 8, by a diffision process.In a mixture of the latter ratioNe23 in addition to Ne21 was indicated by mass-spectrographicbut not by optical methods. This result, if corroborated, would bea remarkable one. Isobares of odd atomic weight are rare andthe mass 23, common to neon and sodium, would have the addeddistinction of being the lightest isobare known. The remainderof the results of Table I are due to F. W. Aston, obtained oftenby ingenious and unexpected means in the face of great experimentaldifficulties. (Thus, rhenium heptoxide failed to give mass lineseither as vapour or as solid. When gold chloride was excited inthe tube containing rhenium oxide on the walls, rhenium lines wereobtained in great intensity in absence of gold lines.Again, theintensity of oxygen lines was greatly enhanced by exciting themin a mixture containing helium.) Scandium,l8 niobiurn,lg czsium,20tantalum,lg and uranium21 have been found to be simple. Thecontroversy22 about the true atomic weight of czsium seems nowto be decided against the chemical methods. K. T. Bainbridge,23using Dempster’s method of analysis, has shown that the abundanceof a second isotope must be less than 0.3% of that of Csl@; F. W.Aston,20 by producing anode rays of caesium and gas rays of xenonin the same tube, compared their masses to 1 part in lo7 parts andproved conclusively that the packing fraction of caesium has anormal value; his result is given in Table 11.A similar disputeabout tellurium is not yet settled. F. W. Aston’s 24 value, 128.04,may be too high. 0. Honigschmid’s 25 new determination byanalysis of TeBre, 127687 -+ 0.019, is in agreement with the Inter-national value. There is a possibility that minor and light isotopesof tellurium 26 may bring down the higher value. The results onl6 Physical Rev., 1931, [ii], 37, 167; A., 1931, 403.l7 Naturwiss., 1932, 20, 493; A., 790.l8 Proc. Roy. Soc., 1932, [A], 134, 571; A., 210.l* Nature, 1932, 130, 130; A., 895.*O Ibid., 1931, 127, 813; A., 1931, 783.21 Ibid., 128, 725; A., 1931, 1349.22 Ann. Reports, 1928, 25, 305.23 Physical Rev., 1930, [ii], 36, 1668; A., 1932, 6.24 Ann. Reports, 1926, 23, 280; see A., 1925, ii, 618.25 Natumiss., 1932, 20, 659; A., 980.26 K.T. Bainbridge, Physical Rev., 1932, [ii], 39, 1021304 SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.scandium, niobium, and tantalum suggest (Table 11) that thechemica.1 atomic weights are in each case a little too high. Thesimplicity of uranium is provisional ; no second isotope of abundancegreater than 2% of U238 could be detected. Rubidium hasbeen found to have its expected composition; a third isotope hasbeen found for strontium l8 and a fourth for barium.18 Ruthen-ium 27 and osmium 27 were each found to have six isotopes; theabundance of Rug8 is not yet decided. Rhenium 28 and thallium 18, 21both have their heavier isotope in greater abundance, thus differingfrom all other complex elements of odd atomic number greater than7.Two minor isotopes have been found for mercury29 and fourfor lead.3o J3glg6 is certain, Hg203 probable ; the former's abundanceis approximately O.Ol%, the latter's 0.006%. Of the four newlead isotopes, Pb203 and Pb205 are not certain; Pb207 and Pb2l0are related in abundance to Pb2OS as 19 : 1 : 250 according to P.W. Ast~n,~O or as 8 : 1 : 200 according to K. M ~ r a k a w a . ~ ~ Pb204has been found also by H. Schiiler and E. G. Jones 32 in the hyperfinespectrum of ordinary lead. If Pb203 and Zrg6 be confirmed, themass 203 will share with mass 96 the rare property of belongingto three elements : Hg203, TlZo3, PbZo3 and Zr", Mog6, Rug6, re-spec t ively .In Table I1 all the results except that for boron have been obtainedby F.W. Aston. There has been difference of opinion on t'heabundance ratio Li7 : Li6. From band-spectrum work W. R. vanWijk and A. J. van Kceveringe33 find the value 7.2. On themass-spectrograph F. W. Aston 34 found 10.2 & 0.5, which is close to14.9, obtained by M, Aforand35 with a heated anode, and to 10.5,obtained by H. Schiiler 36 from hyperfine structure of the spectrumof Li+. It is likely that the variation in intensities of spectral linesdue to Li6 and Li7 respectively, with the conditions under whichthey are excited, as found by G. Nakamura and T. S~hidei,~' ispartly the cause of the above differences. This does not occurwith positive rays.38 The atomic weight of lithium calculated2 7 Nature, 1931, 127, 233; A., 1931, 280.29 Ibid., 1933, 130, 847.31 Sci.Papers Inst. Phys. Chem. Res. Tokyo, 1932, 18, 245; A., 892.32 Nature, 1932, 129, 833; A., 670.33 Proc. Roy. SOC., 1931, [ A ] , 132, 98; A., 1931, 992; Naturwiss., 1931, 17,34 Nature, 1931, 128, 149; A., 1931, 994.35 Thesis, Paris, 1927; Compt. rend., 1926, 182, 460; A., 1926, 331.36 NaturwiPs., 1931, 19, 772; A., 1931, 207.37 Japanese J . Physics, 1931, 7, 33; A., 667; Nature, 1931, 128, 759; A , ,28 Ibid., p. 591; A . , 1931, 666.Ibid., 129, 649 ; A., 554.894; A., 1931, 1348.1931, 1348.K. T. Batinbridge, J . Franklin Inst., 1931, 212, 317; A., 1931, 1207RUSSELL. 305from F. W. Aston’s results is given in the table. The correspondingvalue for boron is calculated from A.Elliott’s 39 abundance ratioBll : BlO, 3-63 & 0.02, for Chilean boron.Despite S. Meyer’s40 advocacy of the value 0l6 = 16.0000 asthe best standard for chemical atomic weights, the InternationalUnion for Chemistry,41 guided largely by F. W. haswisely decided not to depart from the conventional and practicalstandard, 0 = 16, at present in existence, for ,it is very probablethat the abundances of 0l6, OI7, and 0 l 8 in nature is so invariablethat the mean atomic mass of the oxygen atom is as precise aconstant as that of 016. For purposes of atomic and nuclearstructure, radioactivity, mass-spectrography, etc., where a precisionof 1 in 105 is desirable and is expected to be attained, the neutralatom 0l6 = 16.0000 has been chosen from its competitors asstandard by the International Radium-Standards Committee.43The ratio of a mass on this physical standard 0l6 = 16 to that onthe other-the Naud6 correction-has been hitherto taken as1.000125.47 New determinations of this constant have been madeby R.Mecke and W. H. J. Childs 45 and by F. W. A ~ t o n . ~ ~ Theformer find the relative abundances 0l6 : 01’ : 01* as (630 &20) : 0-2 : 1, the latter as 536 : 0-25 : 1. These values raise S. M.Naud6’s correction to approximately 1.0002, an alteration whichcan, of course, have a trifling effect only on the values of atomicweights calculated from mass-spectrograph data, such as are givenin Table 11. The relative abundance W5 : N14, given earlier as1 : 700, has been determined as 1 : 346 by G.M. Murphy and H. C.Urey.48The Neutron.The existence of a neutron, possibly an ultimate particle, ofmass approximately 1 and charge zero, was mooted by J. Chadwick l5as the simplest interpretation of a series of observations initiatedby W. Bothe and H. B e ~ k e r , ~ ~ which were continued and followedup by (Mme.) I. Curie 50 and F. J ~ l i o t , ~ ~ by H. C. Webster,52 and39 Nature, 1930,126,845; A., 1931,15; Z.Physik,1931,67,75; A . , 1931,279.40 Physikal. Z . , 1932, 33, 301; A., 442.41 Ber., 1932, 65, [ A ] , 33; A., 554.43 Phil. Mag., 1931, 12, 609; A., 1931, 1108.44 Ann. Reports, 1930, 27, 310.45 2. Physik, 1931, 68, 362; A., 1931, 543.413 Nature, 1932, 130, 21; A . , 894.4 8 Physical Rev., 1932, [ii], 41, 141 ; A., 980.q9 2.Physik, 1930, 66, 289; A., 1931, 142.50 Compt. rend., 1931, 193, 1412; A., 210.51 Ibid., p. 1415; A., 210; ibid., 1932, 194, 273; A., 210.52 Proc. Roy. Soc., 1932, [ A ] , 136, 428; A., 671.42 Nature, 1931, 128, 731.4 7 Ann. Reports, 1930, 27, 306306 SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.finally by himself. Beryllium under bombardment by a-particles didnot emit protons as did boron or nitrogen, but gave out a weakradiation more penetrating than any y-radiation known. Thiswhen examined by ionisation methods caused material containingcombined hydrogen to emit swift protons. The explanationsuggested for this striking occurrence was that the protons hadgained their energy by a radiation recoil in a process similar tothe Compton effect with electrons; the quantum energy of theradiation was accordingly deduced. J.Chadwick 1 5 9 s3 found thatswift recoil atoms were liberated when the radiation traversed notonly hydrogen-containing material but also helium, lithium, carbon ,air, and argon. His results showed that, if energy and momentumwere conserved in these encounters, the quantum hypothesis ofthe radiation emitted would not hold ; the existence of the neutronwas the simplest explanation of the facts. A neutron in motionwould be expected to produce little if any ionisation in passingthrough matter and to indicate its presence by the recoil of anatomic nucleus with which it collided. Such recoil nuclei wouldbe expected to be easily detected in an ionisation chamber or ina Wilson expansion chamber.The velocity of recoil of a givenatom would be expected to fall off when the radiation was passedthrough increasing thicknesses of an absorbing material such aslead. (This would not be expected if the radiation were a y-radiation.) These expectations have been verified by J. Chadwick.The velocity of the neutron when it is liberated is estimated asone-tenth of that of light. Its mass is found as probably between1-005 and 1.008, suggesting that the neutron may be a small dipolemade up of proton and electron or even a proton embedded in anelectron. Neutrons are found to be emitted by boron as well asfrom beryllium. The processes imagined are Be9 + He4 -+C12 + nf, Bl1 + He4 --+ N14 + nl, d denoting the neutronand the other symbols the nuclei of the elements named.(It isfrom the second expression that J. Chaddck, taking cognizanceof energies and masses, deduced the neutron’s mass.) It is probablethat other processes occur simultaneously, e.g., Be9 + He4 4C13 + y-radiation and B1* + He4 + N1* + y-radiation. N.Feather made an important advance in the work on neutronsby showing that they also could effect artificial disintegration.He obtained disintegration tracks in an expansion chamber re-sulting from collisions of neutrons with nitrogen 54 and with oxygen s5which could be interpreted as n1 + N14 -+ He4 + Bll, the reverseof J. Chadwick’s result, and as nl + 0l6+ C13 + He4. TherePTOC. Roy. XOC., 1932 [ A ] , 138, 692; A., 790.84 Ibid., p. 709; A., 790.6 5 Nature, 1932, 130, 237; A., 081RUSSELL. 307are with nitrogen other possibilities than the ejection of an a-particle : certainly the liberation of a proton, possibly the liber-ation of H2. (Mme.) I. Curie and F. Joliot 56 find that neutronsmay be emitted by lithium, that those emitted by beryllium formtwo groups, and that photons may be emitted simultaneously withneutrons; their results confirm the neutron hypothesis from adifferent angle. Further confirmation comes from the work of3’. Rasetti 57 and of J. L. Destouches.so The excitation of neutronsby radon and their transmission through matter has been studiedhy M. de Broglie and L. Leprin~e-Ringuet,~8 and their penetratingpower by J. Thibaud and F. D. La The existence ofneutrons and of H2 has encouraged theorists to attempt to accom-modate them in the nuclei of light atoms.H. C. Urey’s scheme 61preceded his experimental work.8 Provisional schemes, independentbut in some respects similar, have also been put forward by H. L.Johnston,62 F. Perrin,63 J. H. BartlettYa E. G. Jones,65 W. D.Harkins,66 and others. In one of F. Perrin’s schemes, the “ demi-helion ” is envisaged. This is a particle of mass 2 and charge 1,the union of proton and neutron, and known in the free state asthe heavier isotope of hydrogen. The oldest scheme is due toW. D. Harkins.’Artiscia1 Disintegration by Swift Protons.J. D. Cockcroft and E. T. S. Walton 67 developed the techniqueof producing and using steady high potentials up to 600,000 volts.When lithium oxide was bombarded with a stream of protons, a-particles in pairs were found to be produced.The effect becameappreciable when the protons had been accelerated beyond 120,000volts ; at 250,000voltsadisintegrationparticlewas got for about everylOQ protons striking the lithium. The process imagined is Li7 + H1+ He4 + He4, the symbols representing the nuclei of the elementss6 Nature, 1932, 130, 57; A., 895.6 7 Naturwiss., 1932, 20, 252; A., 556.68 Compt. rend., 1932, 194, 1616; A . , 672; Nature, 1932, 130, 315; A.,b9 Cornpt. rend., ,1932, 194, 1647; A., 672.6o Ibicl., p. 1909; A., 672.131 J . Amer. Chem. SOC., 1931, 53, 2872; A., 1931, 1108; Nature, 1932,130, 403; A., 1074.62 J . Amer. Chem. SOC., 1931, 53, 2866; A., 1931, 1108.us C m p t .rend., 1932, 194, 1343, 2211 ; A., 556, 790.64 Nature, 1932, 130, 166; A., 894.66 Ibid., p. 580; A., 1187.6% J . Amer. Chem. SOC., 1932, 54, 1254; A., 556; Nature, 1933, 131, 23.6 7 Nature, 1932, 129, 649; A., 556; PTOC. Roy. SOC., 1932, [A], 137, 2291073.A., 893; Nature, 1933, 131, 23308 SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.named. A similar disintegration was found to occur less readilywith boron and fluorine, appreciably less with uranium, aluminium,and carbon, and to averyslight extent with a few other elements. Theprocesses here provisionally imagined are Bll + H1 -+ He4 +Be8 or 3He4, F19 + H1 --+ He4 + 0lG, A127 + H1 ---+ He4 +MgZ4, etc., the nature and mass of the residual atom being such thatatomic mass and atomic number are conserved in the process.Itmay be significant that the elements which suffer the emission of theor-particle most easily have masses of form 4a.+3, a being an integer,i.e., have nuclei presumably made up of 3 protons and 2 electrons inaddition to more stable units. It is reasonable to suppose that thecapture of a proton in such nuclei might result in the formation andexpulsion of an a-particle from it. To sum up : Originally, artificialdisintegration of light atoms consisted only in the production ofprotons by swift a-particles. Now a-particles have been shown toproduce neutrons. Each of these processes can also occur in reverse ;protons can produce a-particles, neutrons can produce cc-particles.Neutrons can, in addition, produce protons, but the inverse processhas not yet been demonstrated.Not less in importance is thenature of the resultant residual atom. When the a-particle is theprojectile, the resultant atom is of higher atomic number than thatbombarded; when neutrons or protons bombard, i t is of loweratomic number.Radioactive Constants, Fundamental Constants, and OtherData.A complete survey of radioactive and other atomic constants hasbeen made by the International Radium-Standards Committee.43They give as masses H = 1.0078, proton = 1.0072, He = 4.00216,a-particle = 4.00106, and the electron = 0.000548 (0l6 = 16.0000) ;(on this scale, unity weighs 1.649 x g.) ; Avogadro's constantis given as 6.0644 x 1023 or 6.0265 x 1023 according as c :4.770 x 10-lo or 4-50 x 10-lo, the former being preferred.Thenumber of a-particles expelled per second by 1 g. of radium (freefrom products) is taken as 3.70 x lolo, and the ratio of radium touranium in minerals as 3.4 x In a few cases alternativevalues are given for the half-periods of radioactive products; e.g.,5-0 and 4.9 days for radium-E and 24.5 and 23.8 days for uranium-X,.3.0 x lo5 Years is recommended for uranium-11, the direct measure-ment agreeing with the value deduced from Geiger and Nuttall'srelation. C. H. Collie 68 has, however, since shown, as the result ofthree concordant experiments, that the half-period of this productmust be a t least a million years. He separated electrolytically the68 Proc. Roy. SOC., 1931, [ A ] , 131, 541 ; A., 1931, 891RUSSELL.309uranium-11 arising from the decay of a known quantity of uranium-X and counted the a-particles from the source electrically. 0.Gratias and C. H. Collie 69 found the half-period of uranium- Y to be24.0 & 0-58 hours, the accepted value, 24.6, being ascribed to lackof saturation in the electroscopes used in decay measurements.New determinations of the half-period of uranium-X, and of theactinium-radium branching ratio from uranium by E. Walling 70confirm the International values, the higher alternative in the firstcasc. A. F. Kovarik and N. I. Adams 71 confirm the Internationalvalues for the radium-uranium ratio in minerals and the actinium-radium ratio ; they obtain directly the half-period of uranium-Ias 4.52 x lo9 years, 3% higher than the International value.P.S0ddy,~2 from the growth of radium in uranium purified 25 yearsago, finds the half-period of ionium as 7-41 x lo4 years, 12% lessthan the International value.Great progress has been made in the analysis of groups of a-particles both by the electrical counting methods employed a tCambridge 73 and by the magnetic deviation method used in Paris.73(Lord) Rutherford, F. A. B. Ward, and W. B. Lewis 7* have foundthat the long-range a-particles from radium-C’ may be analysed intonine homogeneous groups of ranges 7-12 cm. The most abundantof these groups has 16.7 particles per million of the ordinary groupof range 6.96 cm., the remainder have abundances varying from 0.2to 1.27 per million.(Lord) Rutherford, C. E. Wynn-Williams, andW. B. Lewis 75 find two groups of long-range particles from thorium-C’, of ranges 9.78 and 11.66 cm., in a ratio 1 : 5.6, the long-rangeparticles having an abundance relative to the ordinary group ofrange 8.62 cm., of 1.9 per million. (S. Rosenblum 73 had foundthem even more complex and since then has extended 76 his result.)The earlier work 73 on the complexity of the particles from actinium-C was confirmed ; actinium47 has two groups in relative abundance0.19 and 1, and relative velocity 0.9737 and 1. The values for thehomogeneous group from actinium-C‘ are on this scale 0.0032 and1.062 respectively. The relative velocities have been confirmed by(Mme.) P. Curie and S. R ~ s e n b l u m , ~ ~ who find 0.973 : 1 : 1.062.W.B. Lewis and C. E. Wynn-Williams 78 have analysed the particlesfrom actinon into two groups, the complexity with this product6s Proc. Roy. Soc., 1932, [A], 135, 299; A., 443.T o 2. Physik, 1932, 75, 425, 432; A., 555.71 Physical Rev., 1932, [ii], 40, 718; A., 790.72 Phil. Mag., 1931, 12, 939; A., 5.74 Proc. Roy. SOC., 1931, [ A ] , 131, 684; A., 1931, 890.j5 Ihid., 133, 351 ; A., 1931, 1349.7 6 Compt. Tend., 1931, 193, 848; A., 5. 7 7 Ibid., p. 33; A., 1931, 995. ’* Proc. Roy. SOC., 1932, [A], 136, 349; A., 671.73 Ann. Reports, 1930, 27, 313310 SUB-ATOMIC PHENOMENA AND RADIOSCTIVITY.being similar to that with actinium-C, with regard t o \Jot11 relativenumbers and relative velocities. As (Mme.) I. Curie 79 had foundthat the a-particles from radioactinium consist of two groups inabout equal numbers, it would appear that there is some recurringcharacteristic of the nucleus underlying this phenomenon.(Mme.P. Curie and S. Rosenblum,so however, have found that radio-actinium is more complex than has been supposed. Its particleshave six or seven groups, two of which are very strong, two strong,and the remainder weak. Actinium-X, however, was found tohave two groups of particles.) S. Rosenblum and (Mlle.) C. Cham% ahave found that radiothorium emits two and possibly three groupsof particles, and S. Rasenblum 82 found two groups from radiumitself. During these investigations, actinium-A , thoron, thorium-A ,radon, and radium-A were found t o give homogeneous particles;many determinations of ranges and energies of a-particles fromevery product except uranium, protoactinium, and thorium havebeen recorded there and elsewhere.83 The accepted values of theranges of uranium-I, thorium, and uranium-I1 have been confirmedby F.N. D. Kurie,B4 G. H. Henderson and J. L. Nickers0n,8~ andS. Bateson 86 respectively.Evidence that the emission of y-rays from radium-C’ is intimatelyconnected with the occurrence of its groups of long-range particleswas given by (Lord) Rutherford, F. B. Ward, and W. B. Lewis; 74it was concluded that the y-rays arise from the transition of ana-particle in an excited nucleus between two levels of differentenergies. This question has been discussed in more detail by (Lord)Rutherford and C.D. Ellis,s7 C. D. Ellis,ss and (Lord) Rutherfordand B. V. BowdexS9 For radium-C’ it is supposed that in thepreceding transformation the emission of an a-particle causes aviolent disturbance in the resulting nucleus which causes some ofthe constituent a-particles to be raised to a much higher level thanthe normal. These, being unstable, are believed to fall back aftera very short interval to normal level, emitting their surplus energyas y-radiation of definite frequency. The ideas of wave mechanics,79 Compt. rend., 1931, 192, 1102; A., 1931, 783.]bid., 1932, 194, 1232 ; A., 555.Ibid., p. 1154; A., 555.82 Ibid., 195, 317; A., 895.83 S. Rosenblum and G. Dupouy, Compt. rend., 1032, 194, 1919;G. H. Briggs, J. Xci. Inst., 1932, 9, 5; Nature, 1932, 130, 1000.84 Physical Rev., 1932, [ii], 41, 701 ; A., 1186.E 5 Ibid., 1930, [ij], 36, 1344; A., 1931, 16.s6 Canadian J.Res., 1931, 5, 567; A., 106.8 7 Proc. Roy. SOC., 1931, [ A ] , 132, 667; A., 1208.88 Ibid., 1932, [A], 136, 396; A., 671. 89 Ibid., p. 407; A., 671.671 RUSSELL. 31 1however, suggest that in this short interval there is a small chancethat some of the a-particles in the higher states can escape from thenucleus. On this view the escaping a-particles are the long-rangeparticles observed, and their energies give the values of the energylevel in the nucleus which they occupied before escape. It was, infact, found that the differences of energies between the variousgroups of a-particles were closely connected with the energies of themost prominent y-rays in the spectrum, and, in general, strongevidence was found that y-rays have their origin in the transitionsof one or more a-particles in an excited nucleus.The energies setfree in transitions are given approximately by the expressionE = pEl-qE,, where El is a difference in energy of two states,Ez a smaller difference of interaction, and p and q integers. Fortyy-rays from radium-C’ and a smaller number from radium-B canbe conveniently expressed by such an equation.The connexion between a-particles and y-rays in thorium-C isdifferent. With radium-C’ the most intense a-particle has the lowestenergy and the long-range particles are rare occurrences ; this is notso with thorium-C. G. Gamow has proposed that here the thorium-C nucleus is initially formed with all the a-particles in the groundstate, not, as with radium-C‘, with some of the a-particles in higherlevels of energy, and that disintegration can sometimes occur insuch a way as to leave the product nucleus excited.found that the y-rays were emitted immediately after the dis-integration of thorium-C, in agreement with G.Gamow’s 9o theory.This provides further proof of the connexion of y-rays with exciteda-particle states in the nucleus. After the discovery that actinonemits two distinct groups of a-particles, it was found by (Lord)Rutherford and B. V. Bowden s9 that the transformation actinon +actinium-A was accompanied by weak @-rays and strong y-rays.From the measurement of the penetrating power of the latter, itwas concluded that the energy of the y-rays is of the right order tobe expected from the difference of energies of the a-particle groups ;again, strong confirmatory evidence that y-rays have their originin transitions of a-particles in an excited nucleus.The controversy as t o the most probable values of the fundamentalconstants, e, h, and the reciprocal of the fine-structure constant,2xe2/hc, continues.W. N. Bond 91 has developed a new way ofreducing the experimental data used in connexion with the determin-ations of e and A which is based on the observa-tion that each group ofO0 Nature, 1930, 126, 397; A., 1339.O1 Proc. Physical Soc., 1932, 44, 374; A., 672; Nature, 1931, 127, 557;A., 1931, 667; Phil.Mag., 1930, 10, 994; A., 1931, 143; ibid., 1931, 12, 632;A., 1931, 1207; Physical Rev., 1932, [ i i ] , 41, 368.C. D. Elli312 SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.experiments connects h and e by an equation of the form h = Aen,where n is 1, 4/3, or 5/3 according to the experiments. He givesto e, h, and hc/2xe2 respectively (significant figures only) the values4.799, 6.558, and 137.02. From the same data, however, R. T.Birge 92 gets 4.769, 6.544, and 137.31. F. Kirchner 93 gets 4-798,6.615, and 137.09 or 4.782, 6.577, and 137.25, depending upon whichalternative data, derived from measurements of the slzort-wavelimit of the X-ray spectrum, are taken. K. Shibag4 has alsocritically reviewed the available data. Me argues that the X-rayvalue of e gives consistent values of h by eight methods.His valuesare e = 4.803, h = 6.624, and hc/2xe2 = 137.03. There is still,therefore, at, the present level of accuracy of measurement, a casefor (Sir) A. S. Eddington's theory,95 which requires for the reciprocalof the fine-structure constant the exact value 137. There appearsto be little, however, for his other theory,gG which requires for theratio of the masses of proton and electron the value 1849.6.Chemistry of Protoactiniurn.A. V. Grosses7 has continued his work98 on the chemistry ofprotoactinium, working with 10-50 mg. of pentoxide free fromother metals. It has been generally assumed that the pentoxidesof tantalum and protoactinium are chemically very similar, as are,for example, the corresponding compounds of barium and radium.They are, however, widely different.Protoactinium is definitelybasic, as it should be from its position in the periodic classification,whereas tantalum is feebly acidic. They are similar in that bothoxides dissolve in 40% hydrofluoric acid and are precipitated byammonia from mineral acid solutions. Protoactinium oxide isinsoluble in molten potassium carbonate (in which tantalum oxideis completely soluble), and almost entirely soluble in molten sodiumbisulphate (in which tantalum is nearly insoluble). Protoactiniumis precipitated from acid solutions by excess of phosphoric acid;there is no corresponcliiig precipitation with tantalum. Theseparabion is, in consequence, a simple matter.0. Gratias 9s hasindependently made similar observations with unweighably smallquantities of protoactinium ; he used an amplifier and an ionisation92 Physical Rev., 1932, [ii], 40, 228; A . , 672.93 Ann. Physik, 1932, [v], 13, 59; A., 556.94 Sci. Papers Inst. Phys. Chern. Res. Tokyo, 1932, 19, 97; A . , 1187.9 5 Ann.. Reports, 1930, 27, 323; A., 1929, 231.9 6 Proc. Camb. Phil. SOC., 1931, 27, 15; A., 1931, 279; Ann. Reports, 1930,O 7 J. Anzer. Chem. Soc., 1930, 52, 1742; A . , 1930, 883.9s Ann. Reports, 2928, 25, 313.99 Thesis, Oxford, 1932; 0. Gratias and C. H. Collie, J., 1932, 987 ; d., 443.27, 324RVSSEU. 313counter to detect the radioactive material’s presence in the chemicaloperations. He showed that the product of decay of uranium-Yemitted a-particles, and had approximately the same half-periodand the same chemical properties as protoactinium.He has thusestablished directly by experiment what hitherto has been merelyassumed on general grounds, namely, that uranium-Y is the directparent of protoactinium.The Penetrating Radiation.The problem of the cosmic or penetrating radiation has definitelyadvanced towards solution during the period under review, althougha t first sight it would appear that opinion about i t could hardly bemore widely divided; its nature has been variously describedas quantum, neutron, electron, and positively charged particle.There has been a gradual change, however, from the older quantumview of R. A. Millikan to the view that the radiation is a very high-energy particle. The former view has been reconsidered by (Sir) J.H.JeansY3 who rejects the view that the radiation can be anythingbut y-radiation on the grounds that a charged particle would bedeflected in the laboratory by a magnetic field, which was apparentlynot the case,4 and could not fall evenly on the earth, as it does,5owing to the influence of the earth’s magnetic field. He has cal-culated the penetrating power of the radiation on the assumptionof R. A. Millikan that part of it is generated by the formation ofnuclei of iron from the necessary protons and electrons, and on hisown assumption, by the annihilation of one proton or four protonsby their respective electrons. In this calculation he has used theformula of 0.Klein and Y. Nishina,6 the scattering electrons beingtaken as all the electrons in the atom and not, as is generally done,the extra-nuclear electrons only. The two hardest constituents ofpenetrating radiation, as found by E. Regener,’ have penetratingpowers very close indeed to the calculated values on the assumptionthat four protons and one proton have been annihilated; thesynthesis of iron gives much too soft a radiation in this calculation.(Sir) J. H. Jeans has pointed out that if the radiation had such anorigin there is no need to assume, as R. A. Millikan has done, thatthe process is still occurring in the depths of space. It may becalculated that the hardest constituent of the radiation is so pene-trating that it woulcl not be reduced to l i e of its initial intensityAnn. Reports, 1930, 27, 322.Nature, 1931, 128, 104; 127, 594; A4., 1931, 666.P. Epstein, €‘roc.h7at. Acad. Xci., 1930, 16, 658.K. Grant, Nature, 1931, 127, 924.Ibid., 1931, 127, 233; A., 1931, 408.Ibid., 1928, 25, 321.ti Ibid., 1928, 122, 395314 SUB-ATOMIC PHENOMENA AND RADIOACTIVITY.until after 5 x 1015 years, a period greater, so far as is known, thanthe age of the universe. The penetrating radiation of to-day maytherefore be the result of the annihilation of matter (possibly nearthe surfaces of astronomical bodies, more probably in unattachedatoms or molecules in free space) a remote period ago. A similarview has been tentatively advanced by E. Regener.*The replacement of the ionisation vessel by the Wilson cloud-chamber and the Geiger-Miiller particle-counter has given resultswhich have modified the above views. The newer workers regardthe radiation as a particle.C. D. Ander~on,~ P. M. S. Blackettand G. Occhialini lo and others l1 have devised apparatus so thatthe penetrating ray itself actuates the Wilson cloud-chamber,and a photograph of what is occurring may be taken a t the timeand not at random. In the device of P. M. S. Blackett and G.Occhialini the cloud-chamber is inserted between two Geiger-Muller counters in line. Passage of the penetrating radiationthrough both counters actuates the cloud-chamber within 0.01 see.,which is sufficient interval to enable the track to be photographed.They found that only about 10% of the tracks were markedly bentin a field of 2000 gauss, so that if the radiation was an electron itsenergy would be 106-107 volts. The remainder, unbent, corre-sponded with electrons of 6 x lo8 volts or protons of 2 x los volts.C.D. Anderson found a much smaller proportion of unbent tracks ;they were quite rare. He observed pairs of tracks frequently, oneof which was always that of an electron. He ascribed these to thedisruption of a single atomic nucleus by the penetrating radiation.Sudden bursts of ionisation, as though from a shower ofionisingparticles from a violently bursting nucleus, have been observed byE. G. Steinke,12 H. Schindler,13 A. H. Compton,14 and others.These appear to be greater than those given by any a-particle andto be more frequent at high altitudes. The radiation also behavespeculiarly when it traverses successively thicknesses of two differentmetals; a peculiar secondary radiation is set up related to theprimary as are 8-rays to or-particles. H. Geiger l5 interprets thispuzzling occurrence by regarding the radiation as protons with veryhigh energy. The older observation that the radiation comes* Nature, 1931, 12'7, 869.lo Nature, 1932, 130, 363.11 L. M. Mott-Smith and G. L. Locher, Physical Rev., 1931, [ii], 38, 1399;1932, 39, 1883; A., 5; T. J. Johnson, W. Fleisher, and J. C. Street, ibid.,1932, 40, 1048.Physical Rev., 1932, [ii], 41, 405.l2 Physikal. Z., 1930, 31, 1019; 2. PIr?y.sik, 1932, 75, 115; A . , 566.l3 Naturwiss., 1932, 20, 491 ; A . , 791.14 Physical Rev., 1932, [ii], 41, 681. l6 Nature, 1931, 127, 785RUSSELL. 315equally from all parts of the sky has been confirmed.16 While,however, V. F. Hess l7 has found that the sun does not contributemore than 0.5% of the total intensity a t 2.5 km. above sea-level,A. H. Compton l8 found that the intensity a t 3-9 km. was 1.5 -40.25% greater between 8 a.m. and 4 p.m. than between the corre-sponding night hours. The same observer,lS in accord with J.Clay 2o but in discord with earlier observation^,^ found that theintensity of the radiation is in general higher the greater the angle ofmagnetic dip.A. S. RUSSELL.l6 E. Regener, Nature, 1932, 130,364; 2. Physik, 1932, 74, 433 ; A., 1072 ;l7 Nature, 1931, 127, 10; A., 1931, 143.ao Proc. K . Akad. Wetensch. Amsterdam, 1930, 7, 711.A. Piccard, Comnpt. rend., 1932, 195, 71.Physical Rev., 1932, [ii], 41, 111. l9 Ibid., p. 681