RADIOACTIVITY.The Atomic Weight of Lead.THE theoretical prediction that the atomic weight of lead fromradioactive minerals should differ from 207.1, the internationalfigure, that from uranium minerals tending towards the number206, and that from thorium minerals towards the number 208,lhas now been examined experimentally by several observers. Theresults obtained, although in all cases as yet of a preliminarycharacter only, show clearly that the atomic weight of lead fromradioactive minerals differs from the international value and varieswith the character of the mineral from which the lead is derived.I n the first results to be published, the mineral examined wasCeylon thorite,2 which is uniquely suitable for the investigation ofthe question, as regards thorium, since the ratio of uranium t othorium is exceptionally low, and the amount of lead is so smallthat it may well be all of radioactive origin. The thorium contentis about 55 per cent., the uranium content between 1 and 2 percent., and the lead about 0.4 per cent.If the law of the conservation of mass held during radioactivechanges, and the rigid correctness of this assumption is now opent o doubt (p.271), if all the lead in the mineral were of radio-active origin, and if the lead derived both from the thorium andthe uranium were entirely stable and accumulated linearly withthe lapse of time, it is to be expected that the atomic weight ofthe lead from Ceylon thorite should be rather more than one unithigher than the international figure.The rate of change ofuranium is probably between 3 and 2.5 times that of thorium, sothat, on the above assumption, the lead should be derived at leastten parts from thorium to each part from uranium. The calcu-lated atomic weight of the former is 208.4, and of the latter 206.0,Ann. Report, 1913, 269.F. Soddy and H. Hyman, T., 1914, 105, 1402. This paper was read on May7th: 1914, and an abstract appeared in the Morning Post 011 the following day, andin the Proceedings on May 18th, 1914.esRADIOACTIVITY. 261since they are derived from thorium, 232.4, and radium, 226.0,respectively, by the loss of six and five atoms of helium. Thecalculated atomic weight of the thorite lead should therefore beabout 208.2.I n the experimental work, a kilogram of the mineral was workedup, and only 1.2 grams of purified lead chloride were finally avail-able.It was purified chiefly by precipitation as sulphate and assulphide and by crystallisation as iodide, and its atomic weightwas estimated purely relatively against that of ordinary lead(purified in an identical series of operations), volumetrically by titra-tion with the same silver nitrate solution. The lead chloride wasfinally weigheG in a platinum boat, after fusion in hydrogenchloride and cooling in nitrogen, as recommended by Baxter andWilson. The mean of two determinations showed a difference of1 part in 225 in the volume of silver solution required for equalweights of the two lead chlorides, which is certainly many timesgreater than the possible error of the experiment unless unknownsources of error existed.The atomic weight' of the thorite lead,calculated from that of ordinary lead as 207.1, was 208.4.Photographs of the spectra of the two specimens of lead, takenby the FQry spectrograph, showed complete identity, both in thewavelengths and the relative intensity of the lead lines, with thesingle exception of one line, 4760.1, which was much stronger inthe photographs of ordinary lead than in those of the thoribe lead.As the whole series of experiments are now being repeated de ~ O P Owith the lead derived from 30 kilograms of hand-sorted thorite, dis-cussion of these results may be deferred.I n the next results published,3 the atomic weight of lead fromfive radioactive minerals, uraninite (N.Carolina), pitchblende(Joachimsthal), carnotite (Colorado), thorianite (Ceylon), andpitchblende (Cornwall), and from two commercial products, notradioactive, was determined by thO accurate methods developedin the Harvard laboratories, and used by Baxtter and Wilson intheir previous determination of the atomic weight of lead. Thelead from N. Carolina, uraninite, was a small sample of 3.8 gramsof chloride, separated from 110 grains of the purest selectedmineral by Boltwood and Gleditach ; that from Joachimsthal pitch-blende and Colorado carnotite had been separated by Fajans; tllatT. W. Richards and Max E. Lembert, J. Amw. Chem. Xuc., 1914, 36, 1329 ;Cunzpt. rend., 1914, 159, 248 ; A., ii, 653. These resnlts were riot piiblished by theauthors riiitil after the researches about to be considered, but a preliminnry announce-ment of them was iriade by K.Fajaiis under the title " Naclltrag zu dem Aufsatz' Die Radioelerrierite und das periodische system,' " n'nizcl.u,i~~ensrl,clf(ell, May 29th1914 ; also later iu Sitxungsber. der Heidelbergen h a d . d e ~ Wzss., 1914, Abt. A,,Abh. 11268 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.from Ceylon thorianite was part of 1100 grams of nitrate separatedby Boltwood from 25 kilograms of mineral; and that from Cornishpitchblende was from residues supplied by Sir William Ramsay.The methods of purification used were chiefly the recrystallisationof the nitrate and the chloride, and the precipitation of the dilut,esolution of the latter by hydrogen chloride.The final valuesobtained are shown below:Lead from N. Carolina uraninite ............................... 206.40,, ), Colorado carnotite ................................... 206'59,, ,, Ceylon thorianitc .................................... 206.82 ,, ,) Coroisli pitchblende ................................ 206.86Coiiimon lead .................................................... 207'15), ,) Joachimsthal pitchblende ......................... 206.57I n the nextt serjes,4 the lead from three uranium minerals coil-taining negligible quantities of thorium, pitchblende, carnotite,and yttro-tantalite, from one thorium mineral containing very littleuranium, monazite, and from ordinary galena, was examined.Itwas separated as metal from the crude carbonates by fusion withcyanide, converted into the nitrate and precipitated as sulphate,converted into carbonate and again precipitated as sulphate,separated as oxide by electrolysis, reduced to metal by cyanide andconverted into nitrate, and then into basic nitrate by heat. Thebasic nitrate, dissolved in water, was filtered from the insolublefraction, converted into nitrate, crystallised, and finally obtainedas metal. The metho'd of estimation of the atomic weight was t oweigh the metallic lead in a quartz vessel, to dissolve i t in con-centrated nitric acid, and t o dry to constant weight at 145O t o150°, weighing it again as lead nitrate. This is one of the methodsemployed by Stas, and for relative estimations has advantages overthe more accurate silver method in requiring less manipulation ofthe substance.The lead was purified in each case by a series ofoperations until it showed the same atomic weight after a freshpurification. The values found are shown below :Lead from carnotite ........................................ 206'36 ,, ,, yttro-tantalite ..................................... 206 -54 ,, , , pitchblende ..................................... 206'64,) monazite ............................................. 207'08,, , , galena ............................................... 207.01Unfortunately, the sources of the radioactive minerals are notgiven.Lastly,5 the atomic weight of the lead from Joachimsthal pitch-blende was accurately determined by Baxter's method.It waspurified chiefly by crystallisation of the nitrate from the hot solu-0. Honigschmid and Mlle. S. Horovitz, ibid., 1796 ; A . , ii, 653.4 Maurice Curie, Cornpt. rend., 1914, 158, 1676 ; A . , ii, 563RADIOACTIVITY. 269tion by the addition of nitric acid, followed by solution of thechloride i n saturated hydrogen chloride solution and precipitationwith water. The value 206.736 was obtained as the mean of six de-terminations varying over the extremes of 0.03 unit, and of threedeterminations varying over 0.018 unit. The possibility is men-tioned that, by selecting the pitchblende from isolated pieces ofblende, a lead might be obtained of still lower atomic weight.Bearing in mind that two out of the four researches detailedhave been carried out by chemists experienced in atomic-weightdeterminations, and that much of khe mineral examined was nodoubt of very mixed composition, so that not all the lead presentcan be reasonably assumed to have been of radioactive origin, i tis clear that the theoretical predictions have received remarkableconfirmation from this first preliminary experimental examination.Further results with carefully selecte’d minerals must be awaited.That an investigator as experienced in atomic-weight work as Pro-fessor T.w. Richards should regard his results as definitely estab-lishing a variation in the chemical equivalent of lead from differentsources, whereas earlier experimental investigations a t Harvardon this very question, in the case of the elements copper, calcium,sodium, and iron, all gave completely negative results, is perhapsthe chief result gained.The lowest value recorded, 206.40, is for carefully selecteduraninite, and the highest, 208-4, is for thorite containing a nearlynegligible proportion of uranium.The two values, 206.57 and206.74, obtained by equally skilled workers for Joachimsthal pibch-blende, clearly indicate variations in the character of the materialworked on, and the same explanation may possibly cover the varia-tions found for the other thorium-free uranium minerals, but thevalues for the thorium minerals, Ceylon thorianite, 206*S2, andmonazite, 207.08, are lower than would be anticipated if the end-product of thorium is stable and has the calculated atomic weight208.4, aitliough i t must be remembered that, owing t o its greaterrate of change, the uranium would be some three times as efficienta lead producer as the thorium.I n neither case are the essentialdata given as to the ratio of uranium t o thorium in the minerals.For an average thorianite, this ratio would be, perhaps, 1 to 4 or5, and for an average monazite, perhaps 1 to 8 or 10, so that ineach case a value well above 207 might be anticipated.The Stability of Lead from Thorium.Independently of direct experimental data, the instability of theend-product of thorium, the isotope of lead with a calculated atomicweight 208.4, has been presumed. Boltwood, and later Holmes270 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.have concluded from t’he lead content of uranium and thoriumminerals that lead could not be the end-product of thorium, butin the original conclusion old analyses of the minerals, made beforethe real significance of the lead content was appreciated, seem tohave been too Euch trusted.Thus i t may be mentioned that forCeylon thorite itself, actually containing 0.3 to 0.4 per cent. oflead, which corresponds with more than 100 million years’ produc-tion, a blank appears under the column “PbO” in the originalcareful analysis made ten years ago. The thorium minerals arecertainly less rich in lead than the uranium minerals, but then therate of change of uranium is three times greater than that ofthorium.From an old analysis of a Norwegian thorite, returned as con-taining 46 per cent.of thorium, 0.4 per cent. of uranium and lessthan 0.1 per cent. of lead, it has been calculated6 that the upperlimit of the half-period of thorium lead cannot exceed 2 x 107 years,whilst from another analysis, in which 0.02 per cent. of lead wasstated to be present, a figure ten times smaller is indicated. Thesecalculations, however, assume the equilibrium between the thoriumand lead has been reached, and, in the absence of information as t othe age and character of the mineral, may possess little reality,either on the experimental or theoretical side. The same may besaid of the application of the supposed connexion between thestability of isotopes and their atomic weights,7 which atl best is butpartly true, and can only indicate the result t’hat “thorium” leadis likely to be le’ss stable than “uranium” lead if i t is assumed,not only that “thorium ” lead does in fact disintegrate, which isthe question being discussed, but also that it disintegrates, giving8- rather than a-rays.A recent examination of the lead-thorium and lead-uraniumratio of a series of minerals from the Langesund-fjord district,south of Christiania, Norway, all of Devonian age and “almostcertainly (‘ Middle Devonian,” 8 for which, from the lead-uraniumratio, the mean value of the age of the formation had been previ-ously deduced as 370 million years? gives some indirect informa-tion on the question whether the end-product of thorium is lead.The ratio of thorium to uranium varied in the different mineralsover a range of 180.The lead-uranium ratio was, on the whoIe,remarkably constant, but there was no constancy whatever in thelead-thorium ratios. The ratio calculated on the assumption thatlead was the product both of thorium and uranium was less constantK. Fajans, Sitzungsbev. Eeidelberger Akacl. Wi~s., 1914, Abt. A., Abh. 11.7 Ann. Beport, 1913, 269.8 A. Holmes and R. W. Lawson, Phil. Mag, 1914, [vi], 28, 823 ; A., 1915, ii, 5.Ann. Report, 1911, 295RADIOACTIVITY. 2'7 1on the whole than the lead-uranium ratio. Two thorites, however,were notable exceptions, the lead-uranium ratio being between twoand three times as high as the mean for the others. It may bementioned that f o r the majority of the minerals examined theactual lead content was below 0.01 per cent., and the calculationsnaturally depended entirely, in the case of these minerals, on thepossibility of determining such minute quantities with accuracy.Other series of minerals of less definitel geological age are cited insupport of the vietw that the thorium end-product is not' lead.It is difficult a t the present stage t o evaluate the precise bearingof this evidence.Admittedly it shows-the point, indeed, was notin doubt-that some of the lead is derived from the uranium, andtherefore no constancy is to be anticipated in the lead-thoriumratio; but the further point tlhat none is derived from the thoriuminvolves, perhaps, greater trust in the method and in the analysesthan is justifiable.The relatively slow rate of change of thorium,the dubiety regarding the age and unaltered character of themineral, the lack of information regarding the initial lead contentin the mineral, and the exceptiolns referred to, must all be takeninto account, and it may well be doubted whether what is reallyrather a fine point can be settled by such means. The conclusionsappear to be biassed also by the doubtful theoretical deductionthat, since the uranium isotope of lead is stable, the uranium andthorium isotopes cannot both be stable, and therefore the thoriumisotope is unstable. Against such reasoning, that because oneisotope is stable another of different atomic weight cannot be stable,the fact of the difference between the atomic weights of commonlead and of lead of radioactive origin may now be cited.Thequestion is still sub judice until further atomic-weight estimationsare available.The Comervation of Mass in Radioactive Chmge.The recent determinations of Honigschmid on the atomic weightof radium, for which he obtained the value 226.0, lead to acalculated value for the atomic weight of uranium of practically238.0, if mass is conserved in radioactive changes, instead of 238.5,the international number. A revision of the atomic weight ofuranium has now been carried outlo by the method of Richardsand Merigold, except that. quartz vessels were employed insteadof vessels of porcelain and glass, which are slowly att'acked both bythe bromine and the uranous bromide used.The first series ofexperiments with uranous bromide, distilled, melted, and solidifiedin bromine vapour, in which the ratios UBr, : Ag and UBr, : AgBr0. Honigschmid, Compt. rend., 1914, 158, 2004 ; A., ii, 654272 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.were determined, gave as the mean of eight experiments the value238.08. In a second series, the uranous bromide was distilled,melted, and solidified in a current of nitrogen. Fourteen experi-ments gave the mean value 238.175. The first series, where,perhaps, the bromine is in excess, gives a minimum value, and thesecond series the most probable value for the atomic weight.Although considerably nearer the whole number than the inter-national value, it departs from i t sufficiently far to leave thequestion open whether the mass of the atom of radium and thethree a-particles expelled is essctly equal to the mass of theuranium atom from which they are derived.I n view of the fact that so many of the atomic weights approxi-mate to whole numbers in terms of that of oxygen as 16, whilstsome depart notably from whole numbers, this question is of primeimportance in problems concerning the origin and genesis of theelements.It has been suggested11 that in the disintegration or coalescenceof atoms a change of mass may occur proportionate to the changeof energy, a loss of mass corresponding with a liberation of energyand a gain of mass with an absorption of energy, in each caseequal to the energy which the mass lost or gained would possess,if moving with the velocity of light.The Mass and Velocities of the a-Particle.Radium Constants.I n this connexion may be discussed a new determination of themass and velocities of a-particles,l2 which was designed to test asthoroughly as possible whether the mass of the helium atom travel-ling a t high speed, which constitutes the a-particle, was identicalwith that found for the atomic weight of helium, the older deter-minations having indicated a mass some 4 per cent. less. Fromordinary electsochemical data, that is to say, the value of thefaraday, the ratio of the charge to the mass of the hydrogen ion is9670. With 3.998 taken as the mean of recent atomic-weightdeterminations f o r helium, the value for the ‘‘ bivalent ” a-particleshould be 4826 instead of 5070, the experimental value.By theexercise of great care and the experience accumulated in a longacquaintance with the problem, an accuracy of 1 part in 400 wassecured finally in the measurements of this ratio from the magni-tude of the electromagnetic and electrostatic deviations of thea-particles from radium emanation, radium-A, and radium4 in avacuum. Values lying between 4813 and 4826 were obtained, theSir J. J. Thornson, “The Atomic Theory,” Ronianes Lccture, 1914, p. 16.12 Sir E. Rutherfordand H. Robinson, Phil. Mag., 1914, [vi], 28, 552 ; A . , ii, 789RADIOACTIVITY. 273mean being 4820, which agrees with the calculated value within thelimits of experimental error.The new value f o r the initial velocity of the a-particle ofradium-C, namely, 1.922 x loy cm.per second, is some 7 per cent.smaller than the previous one. Before, the experimental value forthe development of heat from radium agreed excellently with thatcalculated from thO thermal equivalent of the kinetic energy of thea-particles expelled. On the new data, however, the observed heateffect is some 7 per cent. greater than that calculated from the massand velocities of the expelled a-particles, including, of course, thekinetic energy of the recoil atoms. It is by no means necessarythat the total heat energy of radioactive change should be the sameas the kinetic energy of the products. It may be either greateror less, according as the change itself is exothermic or endo-thermic.I n other words, the kinetic energy of the a-particles andof the recoil atoms will together equal the heat evolution only ifthey are entirely responsible for the thermal effect. Rutherford 15suggests that in an a-ray change, wherein the magnitude of thenuclear charge of the atom decreases, there is a decrease in theenergy of the electronic system around the nucleus, which presum-ably appears as heat, and calculates that some 10 per cent. of thethermal evolution is due t o this cause, only 90 per cent. beingdue t o the mechanical energy of the expelled a-particle. I n a &raychange, on the other hand, wherein the nuclear charge increases, acorresponding increase of the internal energy of the electronicsystem of the atom occurs. This is supported by the fact thatthe heat evolution of radium-B and -C together, in which j3- aswell as a-rays result, is less than is to be expected from the evolu-tion of the emanation and of radium-A, in the change of whicha-rays only are expelled.The view, on close examination, seems topresent some difficulties of a fundamental chasacter, not so muchduring an a-disintegration where the residual atom may evolveother energy which afterwards appears as heat, as during a &raydisintegration, where a transformation of heat of the surroundingsinto internal atomic energy seems to be required.I n the same paper Rutherford revises all the radium constantsgiven in his " Radioactive Substances and their Radiations " interms of the International Radium Standard.The half-period ofradium is 1690 years, or average-life period 2440 years, calculatedby taking the iiumber of a-particles expelled per gram of radium (byitself) per second as 3.57 x 1O1O and 4.61 x 10-lo E.S.U. as the valueof the atomic charge. The experimental values for the half-period,2000 years (Bolt4wood), 1800 years (Keetman), and 1730 yeamly Sir E. Rutherford, Phil. Mag., 1914, [vi], 28, 320 ; A . , ii, 788.REP.-VOI,. XI. 274 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.(Stefan Meyer) are considerably higher than the calculated value,but the latter depends upon the agreement of so many independentdata that it is unlikely to be seriously in error. The Boltwood-Rutherfoird uranium ratio of radium t o uranium in minerals onthe international st#andard is 3-23 x 10-7, but a more accuratedetermination of this important constant is much t o be desired.a-Particles and ‘( R-Particles.”Rutherford’s nuclear theory of atomic structure leads to someinteresting consequences when collision occurs between the a-particleand an atom of similar or smaller mass, and the effect on thecollision on the atom struck, as well as on the velocity of thea-particle, is important.14 I n the case of helium, where the massesof both atoms are alike, the result follows that no encounter candeflect t’he a-particle from its course through an angle greater than aright angle, whereas in the case of hydrogen the angle of maximumdeflexion is 14O29‘.In the latter case, an undeflect’ed a-particlemay aignify one that has not been in collision at all, or one thathas experienced a perfectly full collision and followed on in itsinitial path behind the more rapidly recoiling hydrogen atom.Oncertain assumptions, i t was estimated that the maximum velocitywhich a recoiling hydrogen atom could thus acquire was 1.6 timesthat of the a-particle striking it, and its “range” in hydrogenshould be 117 cm., or possibly more.In an experimental search for such recoiling hydrogen atoms, or“ H-particles,” 16 in which a movable source of a-particles was placedin a wide tube filled with hydrogen at variable pressure, scintilla-tions, few in number and less intense than those. produced bya-particles, were observed on a zinc sulphide screen, placed a t theend of the tube, far beyond the extreme range of the a-particlesthemselves in hydrogen.I n air, no such scintillations wereobserved. In hydrogen, a suitable aluminium screen, placeddirectly in front of the source of a-particles, suppressed thesescintillations, but when placed near the screen the scintillationswere again seen, showing that they have their origin in thehydrogen, by impact of the a-particles. So far as could be seen,these “H-particles” obey a similar law of absorption in variousmetal foils to that obeyed by the a-particles, but it is to be ex-pected that they will prove much more penetrating, on account oftheir greater velocity and smaller mass.A special search for radiant particles, differing eit-her in mass orl4 C.G. Darwin, Phil. Mag., 1914, [vi], 27, 499 ; A . , ii, 324; Sir E. Rutherford,l5 E. Marsden, ibzd., 824; A., ii, 407.ibid., 488 ; A., ii, 323RADIOACTIVITY. 275charge from the a-particle, in the disintegration of radium emana-tion gave completely negative results.16y-Rays.Foremost in interest in connexion with the physics of radia-tions has been the determination of the wave-lengths in the spectraof the y-rays by reflection from crystal surfaces, the first step ofwhich was announced last year.17 First, however, may be nieiitionedthe completion of the researches on the analysis of y-rays by mealisof their absorption-coefficients, by the examination of the uraniumy-rays.18 These rays come from uranium-X, and -X2, but theirorigin, as between these successive pro’ducts, is undecided, althoughby analogy it may be surmised that only the most penetrating typecomes from uranium-X,.Three types were disiinguished. For thefirst, which comprises 40 per cent. of the total, the value of p / din aluminium is 8.9, for the second 0.26, and for the third 0.052.The first is probably a characteristic A--radiation of the L-series.The value of the atomic weight of uranium-S, 234, and the valueof p / d , 8.9, compared with these values for mesothorium-11, 228and 9.5 respectively, favour the view that radioactinium, withp / d intermediate a t 9.2, has an intermediate atomic weight, 230,and therefore that actinium also has the atomic weight 230, whichis a point of considerable importance.19 However, this would makethe atomic weights of actinium-B and radium-B identical, whereasthe values for p / d of their characteristic X-radiations are, re-spectively, 11.4 and 14.7.Hence the evidence froin this source isconflicting.The examination of the y-rays froin radium-B and -C by reflec-tion froin crystal surfaces,20 in addition to its intrinsic interest, hasalso been the means of putting to an unexpected experimental testthe prediction from the theory of isotopes, that the spectra ofisotopic elements would prove identical.21 The main lines in thespectrum of the soft y-rays of radium-B are reflected froin rock-salt at angles almost exactly loo and 1 2 O , and the wave-lengthscalculated from them, in Angstrom units (10-8 cm.), are 0.982and 1-278 respectively.By extrapolation from Moselcy’s results(p. 277) for the wave-length of the L-series characlerisLic X-rayof gold t o that of lead, i t was calculated that the characteristicl6 Sir E. Butlierforcl atrd H. Bobinson, Phal. Mug., 1914, [vI], 28, 552 ; A., ii, 789.l7 Ann. Report, 1913, 283.H. Richardson, PhzZ. Mag., 1914, [vi], 27, 252 ; L4., ii, 160.Compare Ann. Keport, 1913, 269.2o Sir E. Rutherford and E. N. da C. Andrade, Phil. Mccg., 1914, [vi], 27, 854;21 See Awn. Report, 1912, 322, for the oiiginal snggestion.28, 263 ; A., ii, 408, 698.T 276 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.X-ray of lead should be reflected from a rock-salt crystal a t anangle of 12*07O, whereas the strongest line in the radium-B y-rayspectrum is reflected a t 12.05O.Fleck has shown that radium-$)is isotopic with lead, and, in order t o test directly whether thecharacteristic X-ray spectrum of lead would prove identical withthe y-ray spectrum of radium-B, the latter was compared directlywith that of the secondary y-rays generated in a block of lead byimpact of fl-rays. Only faint lines were obtained in the latter case,but two of them gave reflection angles of 10°2’ and 12OO’, in goodagreement with the radium-B lines. On substituting for the blockof lead one of platinum, quite distinct lines resulted. The atomicweight calculated for radium-B is 214, whereas that of lead is 207,and the authors conclude that the hypothesis that isotopic elementsof different atomic weights have identical spectra is verified in anunexpected manner.I n the further examination of the more penetrating y-rays ofradium-B and -C, new methods of greater accuracy were developed,in which the rays were transmitted through a crystal and fellnormally on a photographic plate.Blank absorption lines andenhanced reflected lines on the plate both gave the angles of reflec-tion required with considerable accuracy. The spectrum ofradium-C y-rays appears to consist of lines reflected from rock-saltat angles 441, loo’, loll’, and 1°24‘, with wave-lengths of 0.071,0.099, 0.115, and 0.137 in Angstriim units. For radium-By thespectrum of the penetrating y-rays appears t o consist of lines re-flected a t 1O371, 1°44’, 2O0’, 2 O 2 0 / , 202W7 20401, 3001, 3O18‘, 4001,4 O 2 2 / , with wavelengths 0.159, 0.169, 0.196, 0.229, 0.242, 0.262,0.296, 0.324, 0.393, and 0.428, of which, possibly, the third andsecond from the last are second-order repetitions.The mostiniportant are the lines reflected a t lo, and the close doublet re-flected a t 1O371 and 1°43/. I n a later paper22 this is regarded asprobably having one component due to radium-B and one t oradium-C (see p. 283).The shortest wave-length, 0.071, that of one of the radiuni-Cy-rays, is seven times shorter than the shortest hitherto measured,namely, the 1‘-series X-ray line of silver, and the authors expresssurprise that the architecture of the crystal is fine and definiteenough to resolve it. Although the range of the thermal agitationof the atoms in the crystal might be expected to be of the sameorder of length, yet placing the crystal in liquid air did notimprove its resolving power.It is probable that one of the lines of wave-length 0.159 and0.169 is a characteristic X-ray line of the K-series of lead(radium-B), but the lines of radium-C cannot belong t o the X -22 Sir E.Rutherford, Phil. Mag., 1914, [vi], 28, 305 ; A , , ii, 789RA DIOAC‘l’I VITY. 217series of bismuth, and are probably in a new and hitherto un-observed series of higher frequency, which is named the H-series.The work on y-rays, although as yet not complete, raises theexpectation that the complicated mixtures of y-radiations givenby the radio-elements will all be resolved into characteristic X-raysof one o r other of three series, the L, K , and N series, the wave-lengths and penetrating powers of the three series being simplyrelated t o one another on the one hand, and on the other to theatomic number of the source, the greater the atomic number, orthe nearer the element is to the end of the Periodic Table, theshorter being the wave-length and the higher the pelletratingpower.The difficult question of the hardening of y-rays of radiuni-G,whereby, by passage through a heavy metal, the rays become morepenetrating to a lighter metal , has been recently re-e~amined.~~The fact that the spectrum of the y-rays of radiuni-C comprisesthree lines is interesting in view of the practically exponentialcharacter of the absorption, and i t is suggested that “hardening”may be due to the weeding out of the two less penetrating rays.Of great general interest in connexion with these advances inour knowledge of y-rays is the light thrown on the practical problemof how to generate X-rays as penetrating as the y-rays of radium,and so t o avoid the use of that extremely expensive substance inmedicine.24 Rutherf ord and Andrade deduce from Planck’s rela-tion between energy and frequency (p.281) that with a fall ofpotential oE 180,000 volts, which is sufficient to give a velocity 0.7of that of light to the electron, it should be possible to generateX-rays as penetrating as the y-rays of radium-C. With a Coolidgetube, this should not now be beyond the range of possible experi-mental accomplishment.This opinion is probably the reverse ofwhat a physicist a year ago might have been inclined to give, andis very significant.X-Ray 8pectra of the Eleme,nts.Last year, Moseley determined the wave-lengths of the character-istic X-rays of the ten consecutive elements in the Periodic Tablefrom calcium to zinc, and found that the spectrum of each con-sisted of two rays, the wave-lengt,h of the stronger rays beingsimply connected with a set of consecutive integral numbers from19 to 29, numbers, in fact, which represent the position of theelement in the Periodic Table, excluding hydrogen, calcium being23 S. Oba, PhiE. Mag., 1914, [vi], 27, 601 ; A . , ii, 409.Compare also-F. Dessauer, Yhysikffil. Zeitseh., 1914, 15, 739 ; A ., ii, 6992’78 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.the 20th and zinc the 30th element in order of ascending atomicweight.25The facts that, in radioactive changes, when two positive chargesare expelled from the atom as an a-ray, the element changes itsplace in the Periodic Table by two places in the direction ofdiminishing atomic mass, and, when one negative charge is ex-pelled as a @-ray, the element changes one place in the oppositedirection, and, moreover, that all the elements occupying the sameplace in the Periodic Table are chemically identical, had alreadysuggested that the successive places in the Periodic Table correspondwith unit differences in the net value of the internal charge of theatom. Moseley’s work gives us the means of establishing, for thefirst time, the relative value of this charge for each element, andthe absolute number of places in the Periodic Table, over the rangestudied.Since then, what amounts to a veritable roll-call of the elementshas been made by this method.26 Thirty-nine elements, withatomic weights between those of aluminium and gold, have beenexamined in this way, and in every case the lines of the X-r,tyspectrum have been found to be simply connected with the integerthat represents the place assigned t o it by chemists in the PeriodicTable.That is to say, giving the atomic number 13 t o aluminium,the 13th element in order of atomic weight, the atomic numbersof all the elements up to gold, the heaviest element yet examined,can be found.The general method is t o plot the square root of the frequencyof the ray in question against the atomic number of the element,when a series of straight lines results.For the lighter elementsbetween aluminium and silver, the X-series of characteristic X-rayswas examined, and for the heavy elements between zirconium andgold, the L-series. I n the first case, the wavelengths range from8.4 to 0.56, and in the second case from 6.09 t o 1-29, Angstromunits. The longer waves are so easily absorbed, that the experi-mental difficulties are considerable, the spectrometer, as well asthe X-ray tube, having to be exhausted. I n the K-series twos lines,and in the L-series five lines, exist, of which three, designated, inorder of decreasing wave-length and decreasing intensity, a, P, y,have been plotted.There is always, also, a faint companion, a’,on the longer wave-length side of a, and in the rare-earth elementgroup a rather faint $-line between P and y. I n addition, thereare a number of very faint lines of longer wave-length than u.When plotted as described above, each of these rays gives a separateline of slope slightly different from the others. For the L-series,The atomic number thus found for gold is 79.25 A m . &?port, 1913, 272.26 H. G . J. Moseley, Phil. Mag., 1914, [vi], 27, 703 j A., ii, 326RADIOACTIVITY. 279the relation is not accurately linear, a distinct, although slight,curvature of lines being observe<d.For the a-line of the E-series of the light elements, the relationbetween frequency, v, and atomic number, N , is gven byv=A(N-1)2.For the a-line of the L-series of the heavy elements,v = A ' ( N - 7.4)2.A and A' are constants connected with the fundamental Rydbergfrequency, vo, by the relations:A = (1 / 1 2 - 1 / 22)vO = 3/4vo,A 1 = ( 1 / 22 - 1 / 32) vO = 5 / 3 6v0.It will be noted that it is only for the I{-seriea that the squareroots of the frequencies of the a-line are proportional to integers,which integer is one less than the atomic number or place of theelement in the, Periodic Table.For the L-series, the square rootof the frequencies of the a-line is proportlional t o the atomicnumber diminished by 7.4. I n neither case is the absolute atomicnumber found by the generalisation, and, although for the a-line ofthe L-series the roots of the frequencies are proportional to thenumbers of the positions occupied by the elements in the PeriodicTable if hydrogen is omitted, this is probably not significant.Of greater interest to chemists will be the results as regardselements still missing and those misplaced as regards atomic weightin the Periodic Table.Thus, the elements chlorine and potassiumcorrespond with the numbers 17 and 19, the number 18 beingvacantl for argon, not determined; iron, cobalt, and nickel corre-spond with the numbers 26, 27, 28; molybdenum, ruthenium,rhodium, palladium, silver with the numbers 42, 44, 45, 46, 47;tungsten, osmium, iridium, platinum, gold with the numbers 74,76, 77, 78, 79. The missing numbers 43, 75, obviously correspondwith the two vacancies in the Periodic Table below manganesethe real exist'ence of which the writer, a t least, latterly had beeninclined to doubt.I n the region of the rare-earth elements, this roll-call becomesfascinating.I n this most difficult branch of experimental chem-istry, how many elements have not yet been discovered, how manyhave aliases? Moseley finds only one place clearly vacant, namely,that between neodymium and samarium. In the following list,corrected by the author since the publication of his paper, thenumber in the upper column represents t,he atomic number, andthe symbol in the lower column the corresponding element. Anasterisk indicates that the atomic number has been experimentall280 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.determined from the X-ray spectrum.been inferred :I n the other cases i t has* * *$7 E;”s $9 s“0 61 62 63 64 65 & 67 6: 69 70 71La Cc Pr Nd Sa Eu Od Tb Dy Ho Er Tin Yb LnThe element tantalum, at which the normal course of the PeriodicTable is resumed, has the atomic number 73, and the number 72,i t is thought probable, belongs to the keltium of Urbain.Hence,counting lanthanum and cerium, but not keltium, there are fifteenrare-earth elements possible, of which fourteen are known.There may be a re-shding of some of t4e rare-earth elementsamong the available places when their X-ray spectra have beendetermined, but the fact that the precise number can now bedetermined by this method is alone sufficient t o prove its invalu-ableness.Again, compared with the complicated light spectraemitted by the elements, there is a simplicity and definiteness aboutthese X-ray spectra which ensures that in no very long time theywill supplement, if not displace, ordinary spectroscopic methods asa means of identification of elements. The difference between thetwo kinds of spectra, on the one hand, the light spectrum generatedby the complicated electronic swarm which surrounds and hidesfrom investigation the real material atom within, and, on the other,the X-ray spectrum, probably generated by a few only of the inner-most electrons, is analogous t o the difference between a complexcryptogram and plain writing.Perhaps not the least significantdeduction which follows from this roll-call of the elements is that,on the whole, its evidence seems to be against the existence of thecelmtial elements coronium, asterium, nebulium, etc.If hydrogen is the first element, gold is the 79th, and threeonly between these remain undiscovered, one rareearth elementand the two homologues of manganese. As the writer has shown,27the course of the Periodic Table from tantalum t o uranium is pre-cisely analogous to the course from vanadium to molybdenum, onlytwo places being still vacant, namely, those of the heaviest repre-sentatives of the halogen and alkali-metal families. This gives 92for the atomic number of uranium. Hence, from hydrogen touranium there are 92 possible e1ement.s inclusive (not distinguish-ing between isotopic elements) of which five, with atomic numbers43, 61, 75, 85, and 87, remain t o be found, although the last twomay be too unstable to exist.If the periods are made t o commence from the carbon famiIyrather than from that of the inert gases, we have, before carbon,five elements belonging to the latter part of a first short period,then two complete short periods each of eight members, then t8wo27 “ Chemistry of the Radio-Elements,” Part 11RADIOACTIVITY.281long periods each of eighteen elements, bringing the series tolanthanum. From cerium to lutecium there is the rare-earthelement period of fourteen elements. Next is the missing elementwith atomic number 72, which may be keltium, but may be amissing analogue of zirconium, cerium, and thorium.Includingit, there follows another complete normal long period of eighteenelements, and three elements of the next, before the end of theseries is reached.The discussion of the mathematical theory of atomic structurefavoured by these results, which is still in embryo, cannot beincluded here.28Connexion between the p- and y-Rays.The P-rays consist of electrons moving with definite velocity, and,by the application of a magnetic field, the P-rays of many of theradbelements can be resolved into a magnetic spectrum, consistring of a number of groups of rays having the same velocity, whichare deviated to the same extent and produce definite lines, fromthe position of which the velocity may be deduced.29 The y-rays,on the other hand, have been recently proved t o be light waves ofexcessively short wave-length, which can be determined by reflectionfrom crystal surfaces.That a very close connexion exists betweenthe p- and y-rays has for long been known. Making use of Planck’srelation between the frequency and the energy of radiation, Ruther-ford has sought, with considerable success, for a connexion betweenthe energy of P-rays and the frequency of the y-rays accompanyingthem.It has been found30 that when the magnetic spectrum of asource of P-rays is examined, not by the photographic method, butby Geiger’s counting method, by means of the fitful discharge fromx point kept at a high potential when a- or P-rays traverse the gasin the neighbourhood of the point, that the line-spectrum is ofvery small intensity compared with the continuous spectrum.Inother words, in the case of the &rays of radium-B and -C, for ex-ample, most of the rays are expelled with velocities uniformly dis-taibuted over a wide range, whilst the number producing the linescorresponding with groups of rays of the same velocity is relativelysmall. Probably the photographic plate exaggerates the relativeimportance of the line, as compared with the continuous spectrum.2y J. W. Nicholson, Phil. Mag., 1914, [vi], 27, 541 ; 28, 90 ; A., ii, 325, 643 ;W. M. Hicks, ibid., 28, 139 ; A . , ii, 599 ; J. B. Rydberg, ibid., 28, 144; A.,ii, 599.29 Alin. Beport, 1910, 265.J. Chadwick, IZcr.Deut. physikal. Gcs., 1914, 16, 383 ; A . , ii, 408282 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.Another example of the exaggeration introduced by the photo-graphic plate is discussed under the y-rays of radium-B and -C, inthe spectrum of which the lines due to the soft rays of radium-Bgive much the most intense photographic image, although repre-senting only 0.1 per cent. of the total y-ray energy. Certain radio-elements like radium-E and uranium-X, the latter of which hasbeen recently examined,31 give few or no lines, although, in thelatter case, several broad bands may be distinguished in thespectrum.An investigation of the magnetic spectrum of the &rays excitedby the y-rays of radium-B and -0, in various metals, by the photo-graphic method, showed that these rays also have lines in theirspectrum.When the &rays were excited in lead, the velocities ofthe principal secondary &rays were the same as those of theprimary &rays, but, when excited in gold, the velocities of thesecondary rays were found to be some 2 per cent. greater than thecorresponding primary rays. This is of interest, because of theisotopism of radium-B and lead. The primary P-rays given out byradium-B appear to be of identical character with the secondary&radiation excited by y-rays in its isotope, lead.As a consequence of the great advances made in the last twoyears, Rutherford has modified his theory of the connexioii betweenthe P- and y-rays.32 Instead of supposing that the homogeneousgroups of P-rays were due t o the decrease by quanta of the energyof the primary &particle in exciting y-rays, it is now supposedthat they arise from the collision of y-rays with &rays.A&particle is expelled a t a definite speed from the nucleus of the atom,and, in passing through the outer distribution of electrons, stiff erscollision and shares its energy with these electrons, escaping witha velocity which, as a statistical result with a large number ofatoms, is continuously distributed within certain limits. Thusarises the continuous spectrum of 8-rays, typical. of the /3-rays ofradium-E and uranium-X, and important, as recent work shows,in all cases. I n the next place, it is supposed that there are well-defined regions in the electronic distribution capable of being setinto vibration by the passage of the P-particle, the vibrations con-stituting one o r other of the '' characteristic " y-rays of the atoms.Accompanying the emission of these y-rays are one o r more groupsof /3-rays of definite speed, the latter always accompanying theformer.This is in accordance with the evidence that elements likeradium-E and uranium-X, which give continuous P-ray spectra,3 l 0. von Baeyer, 0. Hahn, and L. Meitner, Phgsikal. Zeitsch., 1914, 15, 649 ;32 A m . Report, 1912, 298 ; Phil. Mng., 1914, [vi], 28, 305 ; A., ii, 789.A., ii, 607RADIOACTIVITY. 283give very little y-radiation in comparison with elements likeradium-B and -C, which give well-marked &ray line spectra andpowerful y-rays.I n the y-rays of the last-mentioned class, how-ever, there is probably, in addition to the vibrations of definite fre-quency, a small part with a continuous spectrum.The above point of view requires one very interesting con-sequence. It is necessary to suppose that the direction of expulsionof the primary &ray is definite with regard to the orientation ofthe internal atomic structure, so that whether or no y-rays areexcited depends, not so much upon whether regions capable ofbeing set into vibration exist within the atom, as upon whetherthese regions are traversed by the expelled P-particle. Thus,radium-B and radium-D are isotopic, and it is to be expected thatthey will give identical y-ray spectra when bombarded by P- orcathode-rays (p.276). That their y-ray spectra are entirelydifferent is only intelligible if the P-particle in the two cases is ex-pelled in different directions, traversing different vibratingregions. Using the term X-ray t o indicate radiations generatedby external bombardment of atoms by p- or cathode-rays, in contra-distinction from the term y-rays t o indicate radiations generatedby internal bombardment of the atom by expelled P-rays, i t is to beexpected that, in the former case, all possible types of character-istic radiation will have a chance of being excited, whilst in thelatter case only those arising from the particular regions traversedby the definitely orientated P-ray will be generated. The energyE of a single y-ray is regarded as emitted in quanta in conformitywith what is now supposed t o be true of radiation in general, andto be connected with the frequency v of thO y-ray by the relationE=hv, where h is Planck's constant.The energies of the y-raysin the line spectra of the penetrating y-rays of radium-B and -Care so found. The strongest line of the radium-C y-ray spectrum,that reflected from rock-salt a t lo, has an energy 1-25 x 1013~,where e is the atomic charge. This is very nearly three times theconstant quantity of energy, 0.4284 x lO13e, integral multiples ofwhich represent the energies of many consecutive lines in the j3-rayspectrum of radium-(7.33 Another such energy quantity is0.74 x lO13e, which, in multiples ranging from 2 to 23, representsthe energies of some thirteen lines in the P-ray spectrum ofradium-C.This corresponds with a y-ray which would be reflectedfrom rock-salt a t nearly 1°40/, and is probably one of the doubletbefore referred to (p. 276) reflected a t that angle, both of tliecomponents of which were a t first ascribed to radium-B.I n the same way, the energies of many of the P-rays of radium-B33 Ann. RepoTt, 1913, 280284 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.agree very closely with that calculated from the frequencies of they-rays reflected from rock-salt at 2O20', 2O28/, 3O18', and possibly1°24/. Lastly, the energies of the groups of 6-rays from radium-Bagree well with certain of the y-rays from radium-B, as is a possi-bility in view of the isotopism of these elements.The intense softy-rays of radium-B, reflected a t loo and 1Z0, do not seem t o beconnected with any of the observed &rays. Although giving themost intense lines, their energy is extremely small, being only some2 per cent. of the total y-ray energy of radium-B.34 In the y-raysof radium-B and -C together, i t was estimated that the relativeenergies of the soft rays of radium-R, the hard rays of radium-B,and the rays of radium-C, are in the ratio 1 : 45 : 639 respectively.E1ectrochemica.J Proofs of the Theory of Isotopes.I n three diff erenf ways, the electrochemical identity of isotopicelements has been put in evidence.35 The decomposition potentiala t which radium-E is deposited on the cathode is altered in thesame direction and quantitatively t o the same amount by theaddition of the isotopic element, bismuth, as, according t o Nernst'stheory, i t would be by adding the same number of radium-E ions,and a similar fact has been established with regard to thorium-Band the isotopic element, lead.Thus, in one experiment with asolution of radium-E, calculated to be 10-9 normal, a sudden in-crease in the quantity of radium-E deposited on the cathodeoccurred a t - 0.24 V , measured against the calomel electrode.When the normality was increased t'o 10-4 by thO addition of theisotope, bismuth, the sudden increase occurred a t - 0.14 T7.According to Nernst's theory, the increase of concentration tentimes should change the decomposition potential Om018'V, so thatthe calculated change is O.O9V, and that found, 0.10V. Witha thorium-B solution calculated to be 10-12N, the addition of theisotope, lead, to a normality of 10-3 decreased the potential atwhich peroxide was deposited on the anode by 0*26V, whereas thedecrease calculated from theory is 0.252 V .Secondly, for the deposition of radium2 and of thorium-B a tpotentials below the decomposition voltage, i t was shown that theaddition of isotopic elements prevents this deposition, other ionsbeing without effect. Thus, approximately 4 per cent.of theradium-E present was deposited on an electrode of area 1 sq. cm.in twenty-four hours, a t a potential of -0*17V, but when thesolution was made N/100 with respect to bismuth ions, this deposi-tion was prevented.Making the solution lO-5N with respect toa4 Miss J. Szmidt, Phil. Mug., 1914, [si], 28, 527 ; A., ii, 792.35 G. v. Hevesy and F. Yaneth, Physikul. Zeitsch., 1914, 15, 797R A D I 0 A @TI V ITY . 285lead lowered the percentage of thorium-B, anodically deposited asperoxide, from 5 to 0.5, whilst in 10-3N-solutions no perceptibledeposition occurred. The presence of thallium or of ions otherthan that of lead had no effect.The last proof depended upon the preparation of pure radium-Din appreciable quantity, and its use in a galvanic chain.36 Thispreparation has been previously attempted, but without completesuccess.37 One curie of emanation, from the large quantity ofradium available a t the Vienna Institute, was sealed in a quartzvessel until its change was complete; the radium-n formed wasdissolved in nitric acid and electrolysed.According to the condi-tions, the deposit could be collected as metal on the cathode or asperoxide on the anode. Preliminary experiments have shown thatin the latter case 10-3 mg. of lead produces a clearly visibleand electrochemically active deposit on a fine platinum wire. I nthis way, a perceptible deposit of pure radium-D peroxide wasobtained, so free from ordinary lead that an artificial contamina-tion by mg. of lead could easily be experimentally detected.The following galvanic chain was set up,Pt,[Ra-D]O, J [ Rs-D](NO,),,HNO,,[Ra-D]O, I KNO, I KCI,Hg,CI,,Hg1 V N 10-3177 sat. N N sat.and found t o have the potential -0.884V.By substituting forthe radium-D peroxide elect'rode a similar one prepared withordinary lead, the potential found was -0.888V. I n anotherresearch lead nitrate was added t o the radium-D solution, andthe change of potential measured for concentrations of 10- 5-, low3-,and 10-1-N with respect t o lead, both for the lead peroxide andthe radium-D peroxide electrode. I n each case the same changeof potential occurred, showing that in Nernst's equation i t is thesum of the conceiitrations of the isotopic ions which fixes thepotential of the electrode, and that lead ions and radium-D ionsare electrochemically identical. These results probably constitutethe most severe test t o which the consequences of the theory ofisotopic elements have as yet been put. Another investigationhas shown that the electrolytic potentials of thorium-B andradium-B are the same as those of lead,36 to an accuracy of 2 x 1 0 - 5volt.New Work o n the Disintegration Series.The preliminary announcement is made of the discovery of anew, long-lived member of the uranium disintegration series,39 iso-3(5 See also Ibid., Bcr., 1914, 47, 2784.'' Ann. Report, 1911, 297.'* Z. Klemensiewicz, Compt. rend., 1914, 158, 1889 ; A., ii, 606.39 I<. Fajans and Helene Towma, Naturwissenschaften, 1914, 685286 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.topic with bismuth, and giving u-rays. It was thought thatradium-G, the end-product isotopic with lead, might change withthe emission of P-rays into a further product, “radium-H,” whichgave u-rays, producing an isotope of thallium.This would accountfor the presence of bismuth and thallium in pitchblende. Thebismuth from Joachimsthal pitchblende, after purification, showedan a-activity, several times greater than that. of uranium, that couldnot be removed by a chemical purification which would havesufficed t o remove polonium, or by fractional precipitation ofthe oxynitrate. The range of the a-rays was found to be about3 em., which would correspond with a period of between 105 and106 years. From the a-activity of the preparation, a maximumperiod of 108 years was indicated.No actinium wasfound in a concentrated ionium preparation four years old. Anexamination of pitchblende f o r the supposed a-ray-giving parentof actinium, the heaviest member of the tantalum family, gaveno result.Tantalum, precipitated in the solution and purified,proved to be inactive. It is quite possible that “ekatantalum”departs sufficiently in chemical character from tantalum for thismethod to have been effective.40A thorough examination of the uranium-radium ratio in car-notites of American origin led to the interesting result, that whensamples were taken of large lots, more than 1000 kilos., the ratiowas found to be practically identical with that of pitchblende andother uranium minerals.41 With small samples abiiormal ratios,both higher and lower than the equilibrium ratios, were obtainedApparently, local transportation of radium occurs within the ore-bed which causes differences that are completely equalised whenlarge samples are worked upon.It is pointed out that aridityseems to be a necemary condition for the existence of carnotitebeds, and that the rainfall is small. This may cause local trans-positions rather than complete removal of the radium, as un-doubtedly occurs with a ~ t u n i t e . ~ ~A very accurate determination of the periods of actinium andthorium emanations, in which the decay of the preparations to1/2000th of the initial value was followed, gave the values 3.92and 54.53 seconds respectively for the’ half-periods, or 5.656 and78.69 seconds for the periods of average life.43With regard to the actinium active deposit and the question40 C. Gohring, Phy8ikal. Zeitsch., 1911, 15, 642; A ., ii, 608.41 S. C. L i d and C. F. Whittemore, J. Amer. Chem. Soc., 1914, 36, 2066 ;42 Ann. Report, 1909, 260.43 P. R. Perkins, Phil. Mag., 1914, [vi], 27, 720 ; A., ii, 410.The origin of actinium still remains unsolved.A . , ii, 794RADIOACTIVITY. 287whether the series branches a t the C member, analogously to theother two series,44 it has been reaffirmed that the a-particles, whichhave a range of 6.4 cm., and constitute 0.15 per cent. of the whole,are due t o a branch prod~ct.~5 Confirmation of their existencehas been obtained by an ionisation method, and they have beenfound t o decay with the period of the actinium active deposit.A somewhat disconcerting discovery has been made with regardto the volatility of the various members of the active deposit ofthorium46 Some evidence was obtained in 1912,47 by L.Meitner,of a chemical separation of the two products of the active deposit,giving 35 per cent. and 65 per cent. of the a-rays respectively, byimmersing nickel plates in a solution containing stannous chloride,and comparing the a-activity of the deposit with that of the solu-tion evaporated t o dryness. It appeared that the product giving65 per cent. of the a-radiation alone was deposited, the other beingleft in solution. These results were criticised, and supposed t o beaccounted for by the abnormal volatility of the C-member inthe presence of hydrochloric acid.48I n the present experiments, the active deposit was submittedto carefully regulated temperatures in an electric furnace, and theproportion of thorium-C volatilised was determined.From a-raymeasurements, the curve of percentage volatilised, plotted againstthe temperature, indicated that the C-member was volatilised intwo stages. The first stage commences a t 750°, and correspondswith 35 per cent. of the a-activity, whilst the second stage beginsat 900°. From l3-ray measurements, however, volatilisation of theC-member does not appear to commence below 900°. Thisindicates that the C-member, giving first a-rays and then &rays,is distinct from that giving first &rays and then a-rays, instead ofbeing the same homogeneous element disintegrating dually, in pro-portion 35: 65, respectively, as has been previously assumed.Thorium-B, which is volatile a t 500°, is known to be the productof thorium-C via the a, then @change, which is the 35 per cent.branch.Additional evidence that the a-rays come from two dis-tinct products was derived from the fact that, when heated below900°, the @rays from the preparation, after cooling, show aninitial rise, due t o re-accumulation of thorium-D, but above 900°no such rise occurs. Below 900°, thorium-D, but not its parent, iscompletely volatilised, whilst above 900° both parent and product44 Ann. Report, 1913, 2iO.46 E. Marsden and P. B. Perkins, Phi/. Mag., 1914, [vi], 27, 690 ; A., ii, 410 ;46 A. B. Wood, €'roc. Physical SOC., 1914, 26, 248 ; .4., ii, 6(J6.47 Ann. Report, 1912, 314.48 E. Marsden and R. H. Wilson, PhiE. Mag., 1913, [vi], 26, 354.R.W. Varder and E. Marsden, ibzd., 28, 818 ; A., 1915, ii, 4288 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.are completely volatilised. The authors attempt to explaintheir results by supposing that the 35 per cent. of lower rangea-rays, hitherto regarded as resulting in one mode of the disinte-gration of thorium-C, come from a new product, " Ca," of periodof the order of a minute, according to the scheme:a (5'0 cm.) B and y+ ? +D--- c a -? /r5GyA 1 niin. (1) 3.07 min.(volatile (volatileB (soft) a t 750"). a t 500").+ C -+ B --?a (8'6 cm.)60.3 \ 10'6 hrs.(volatile miiintesat 750"). (volatilea t 900"). %\ & ? 4 1 C' -____10-llsec.It is assumed t h a t thorium-C disintegrates dually as before, butgives &rays, presumably, in both modes.Such an explanatioiiappears to run counter to the periodic law generalisation, for?according t o it, thorium-D should be isotopic with lead rather thanwith thallium. The experiments are very suggestive, and furtherwork promised on volatilisation in atmospheres of different gasesand on the behaviour of the other active deposits will be awaitedwith inberest_. I n the meantime, i t may be noted t'hat there is avery direct tlest of the alleged separation, for a change of the35: 65 ratio of the long- to the shor'crange a-particles should occurin the above deposit, heated a t temperatures between 750° and900°. Upon the constancy of this ratio, under all conditions yettried, the older hypothesis of dual disintegration was primarilybased,49 and it has not yet been possible to establish experimentallysuch a variation.Some interesting experiments have been carried out on thevolatility of thorium-D, showing that the substance, after treat-ment with hydrochloric acid, commences to volatilise at 270°, andis completely volatilised below 500°, whereas the untreated sub-stance commences t o volatilise a t 520° and is completely volatiliseda t 700O.Using the volatility as a test of the s t a b in which thethorium-D was present, the conclusions were drawn that i t recoilsin atomic form from the active deposit which has been treatedwith hydrochloric acid. On the other hand, if obtained by heat-ing the active deposit so treated, it volatilises as chloride a t theAnn. Beport, 1912, 312RADIOACTIVITY, 289lower temperature, showing that in this case combination withchlorine must occur after its production.50Appearance of Helium and Neon in Gases Subjected to t h eElectric Discharge.Numerous experiments, on the same lines as those detailed irrlast year’s Report,51 have been carried out on the presence of tracesof helium and neon in gases subjected t o the electric discharge,and the balance of the evidence appears to be against the viewthat these gases are transmutational products, although much, nodoubt, remains to be explained concerning their apparentlycapricious appearances and non-appearances.Sir J. J. Thomson,whose experiments, criticised here last year, have frequently beenquoted in favour of the view that the appearances are significant,himself now states: “I have never, however, been able to get anyevidence, that I regard as a t all conclusive, that the atom of oneelement could by such means be changed into an atom of adifferent kind; in other words, that by such means we could pro-duce a transmutation of the elements.”62Merton 53 used a tap-free apparatus, essentially identical withone used by the writer in similar kind of work many years ag0.5~Hydrogen, admitted by heating, with a flame, a palladium tubesealed to the apparatus, was subjected t o a heavy discharge, andthen removed by heating the palladium tube with a glowing spiralof platinum wire.When the apparatus was clean, no trace ofargon, neon, or helium was obtained.Minute amounts of argonappearing in the earlier experiments were traced to minute leakageof air, due to a small amount of dirt in the barometric seal. Airwas also found to leak through stopcocks which had every appear-ance of being trustworthy. It is considered doubtful by thisauthor whether stopcocks can be trusted in dealing with suchminute quantities of gases.These results, therefore, are t’o be ranged with those of Strutt,obtained last year, against the view that helium and neon areobtained when due precautions are taken to avoid contamination.Merton’s apparatus was subsequently handed over to Collie,55 whoused it, in conjunction with a cold charcoal apparatus, to demon-strate that uranium gives off, by bombardment with cathode rays,Compare Ann.j0 A. B. Wood, Phil. May., 1914, [vi], 28, 808 ; A . , 1915, ii, 5.51 Ann. Report, 1913, 284.52 Sir J. J. Thomson, Romaneq Lecture, 1914, p. 18.B3 T. R. Mertoii, Proc. Roy. Xoc., 1914, [ A ] , 90, 549 ; A,, ii, 726.j4 F. Soddy aud T. D. Mackenzie, ibid., 1908, [ A ] , 80 92.55 J. N. Collie, ibid., 1914, [ A ] , 90, 554; A . , ii, 727.Report, 1912, 318.REP.-VOL. XI. 290 ANNUAL REPORTS ON THE PROGRESS OF CHEMISTRY.considerable amounts of nitrogen, helium, and neon, althoughnone could be obtained from the uranium by heating. Howeverinteresting such observations may be, they seem confusing t o thepoint a t issue. Deliberately t o put a radioactive source of heliumint,o such an apparatus is simply to spoil the apparatus for thequestion it was designed to test.For, in the writer’s experience,once helium has been used in a discharge apparatus, ever after-wards helium, of the order of magnitude under consideration, maybe obtainable from that apparatus.I n a paper 513 entitled (‘ The Production of Neon and Helium bythe Electric Discharge,” fuller details are given of the experi-mental apparatus by which the previously published results, dis-cussed a t length last year, and some new ones, were obtained,together with a r6sum6 of the results and a discussion of thesources of the gases in question. Many of these experiments alsogave negative results, but the account deals mainly with those thatgave positive results. Apparently, latterly, difficulty has been ex-perienced in obtaining such large yields of helium and neon asformerly, and this is ascribed to’ differences in the interruptersemployed with the coils.A wide diversity of apparatus was used.I n some, transference of gas from the discharge apparatus t o thetesting apparatus was avoided. Electrodes of various metals wereemployed, and in many experiments helium and neon were found,in spite of all precautions t o ensure the purity of the gas and toeliminate air-leakage. Thel absence of argon was considered tobe an extremely delicate proof that the neon and helium foundwere not derived from the air.New experiments with a silica-tube mercury arc apparatus, inwhich considerable quantities of helium and neon were found afterrunning the arc in air, pointed t o the atmosphere in this case asthe source of the gases, for, with the apparatus water-jacketed, nohelium or neon was obtained even after protracted running.Hydrogen, when introduced, disappeared rapidly during the dis-charge, leaving no gaseous residue.Arguments are advanced against the helium and neon in theearlier experiments having been derived from leakage of air, bypermeation through the walls of the discharge tubes, or by previousocclusion of the gases in the materials of the discharge tube, andnegative results were obtained when old glasses were melted in avacuum and when aluminium electrodes were dissolved.Theauthors disclaim the view that th& experiments rigidly excludethe possible sources of the gases discussed, but conclude “that the56 J.N. Collie, H. S. Patterson, and I. Massou, €‘roc. Boy. Soc., 1914, [ A ] , 91,30 ; A., ii, 847RADl OACTLVITY. 291trend of the results is towards conclusions which, if they turn outto be true, would be of very obvious importance.” Clearly, how-ever, the onus of proof now rests with them to show that theseinfinitesimal quantities of helium and neon have a real significance.It is noteworthy, also, that in some of these later experiments thequantities of the gases in question observed must have been ex-tremely near the limit of detectability, for in criticism of Strutt’snegative results, the authors state that they have found i t necessarythat the capillary tubes, used to detect the gases spectroscopically,must, to be sufficiently sensitive, be so fine that mercury can onlybe driven out of them by strong heating: yet the spectrcscopic testfor neon and helium is excsssively delicate.A c t i o i ~ of Radium O ) L t h e Diamond: Chemical A c t i o m .I n a record of various experimentsy extending over a long termof years, on the action of radium on various substances, andespecially on the diamond, some observations were made whichindicate considerable differences between the diamond arid othermaterials rendered radioactive by exposure t o the radium emana-tion.57 First may be mentioned the absence of any trace of radio-activity in a yellow, phosphorescent diamond, which during fortyyears had been used in a Crookes’ tube to show the phosphorescenceproduced by cathode rays.No coloration was produced in adiamond exposed for six months to the P- and y-rays from a sealedtube containing 15 milligrams of radium bromide, but, when sub-jected to the a-rays and emanation by being enclosed in a tubecontaining radium bromide for seventy-eight days, i t acquired abluish-green colour and an enduring radioactivity, comprising a-and P- and y-rays. Such acquired radioactivity, which i t wouldbe natural t o ascribe to radium-l), -E, and -F, proved extra-ordinarily enduring after drastic chemical treat.ment. Neither thecolour nor the activity was affected by prolonged treatment witha hot mixture of fuming nitric acid and potassium chlorate,whereas lead glass which had been rendered active by beingburied for a long period in a radium preparation lost nearly allthe acquired activity by treatment with dilute nitric acid.Cutting the diamond into a brilliant completely removed both theactivity and the colour.Another difference between the diamond and a quartz crystal,for example, rendered active by prolonged contact with a radiumpreparation, was shown in the image produced when laid on asensitive plate. The latter showed the ordinary image of geo-5i Sir \V. Crookts, Phil. T?ans., 1914, [ A ] , 214, 433.u 292 ANNUAL 1tEPORTS ON THE PROGRESS OF CHEMISTRY.metric pattern, due to the superficial deposit of active matter andthe equality of the intensity of the rays in all directions, studiedby Rutherford. I n the case of the diamona, however, the photo-graphs suggested a special discharge of energy from the points andcorners of the crystal. It is not explicitly recorded what precau-tions were taken against phosphorescent light emitted by thediamond to exclude it from contributing to the action.With regard to the very interesting peculiarity shown by thediamond in retaining its activity after drastic chemical treatment,short of supposing a real stimulation of the diamond into1 radio-activity, which would be a phenomenon of an entirely new andrevolutionary character, and is not suggested, possibly the surf aceof the diamond possesses for radium-B and its products a specialattraction, analogous to that shown by palladium for poloniurn.58The subject opened out is new and very attractive.During the year, the decomposition of ammonia,5Q the combina-tion of hydrogen and oxygen,Go and the reduction of carbon mon-oxide by hydrogen61 under the influence of radium emanationhave been studied. I n the latter research, a diminution of volumeof nearly 10 per cent. occurred after nineteen days, correspondingwith a production of methane to the extent of 5 per cent., andethane 0.12 per cent. When stopped at an earlier stage, traces offormaldehyde were found, and the conclusion is drawn that thereaction occurs in two stages, formaldehyde being an intermediateproduct.PREDEI~ICK SODDY.58 Ann. Report, 1913, 274.59 E. Wourtzel, Compt. rmd., 1914, 158, 571 ; A . , ii, 238.6o 0. Scheuer, ibid., 159, 423; A . , ii, i 6 2 .Idem., ibid., 158, 1887 ; A . , ii, 649